structural and thermal performance of cold-formed steel...
TRANSCRIPT
School of Urban Development
Queensland University of Technology
Structural and Thermal Performance of Cold-formed Steel
Stud Wall Systems under Fire Conditions
By
Prakash Nagaraj Kolarkar
BE (Civil) (Govt. College of Engineering Pune, India)
ME (Structures) (Govt. College of Engineering Pune, India)
A Thesis Submitted to the School of Urban Development, Queensland University of Technology in Partial Fulfillment of the Requirements for
the Degree of DOCTOR of PHILOSOPHY
September 2010
ABSTRACT
Cold-formed steel stud walls are a major component of Light Steel Framing (LSF)
building systems used in commercial, industrial and residential buildings. In the
conventional LSF stud wall systems, thin steel studs are protected from fire by placing
one or two layers of plasterboard on both sides with or without cavity insulation.
However, there is very limited data about the structural and thermal performance of
stud wall systems while past research showed contradicting results, for example,
about the benefits of cavity insulation. This research was therefore conducted to
improve the knowledge and understanding of the structural and thermal performance
of cold-formed steel stud wall systems (both load bearing and non-load bearing) under
fire conditions and to develop new improved stud wall systems including reliable and
simple methods to predict their fire resistance rating.
Full scale fire tests of cold-formed steel stud wall systems formed the basis of this
research. This research proposed an innovative LSF stud wall system in which a
composite panel made of two plasterboards with insulation between them was used to
improve the fire rating. Hence fire tests included both conventional steel stud walls
with and without the use of cavity insulation and the new composite panel system.
A propane fired gas furnace was specially designed and constructed first. The furnace
was designed to deliver heat in accordance with the standard time temperature curve
as proposed by AS 1530.4 (SA, 2005). A compression loading frame capable of
loading the individual studs of a full scale steel stud wall system was also designed
and built for the load-bearing tests. Fire tests included comprehensive time-
temperature measurements across the thickness and along the length of all the
specimens using K type thermocouples. They also included the measurements of load-
deformation characteristics of stud walls until failure.
The first phase of fire tests included 15 small scale fire tests of gypsum plasterboards,
and composite panels using different types of insulating material of varying thickness
and density. Fire performance of single and multiple layers of gypsum plasterboards
was assessed including the effect of interfaces between adjacent plasterboards on the
thermal performance. Effects of insulations such as glass fibre, rock fibre and
cellulose fibre were also determined while the tests provided important data relating
to the temperature at which the fall off of external plasterboards occurred.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions i
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions ii
In the second phase, nine small scale non-load bearing wall specimens were tested to
investigate the thermal performance of conventional and innovative steel stud wall
systems. Effects of single and multiple layers of plasterboards with and without
vertical joints were investigated. The new composite panels were seen to offer greater
thermal protection to the studs in comparison to the conventional panels.
In the third phase of fire tests, nine full scale load bearing wall specimens were tested
to study the thermal and structural performance of the load bearing wall assemblies. A
full scale test was also conducted at ambient temperature. These tests showed that the
use of cavity insulation led to inferior fire performance of walls, and provided good
explanations and supporting research data to overcome the incorrect industry
assumptions about cavity insulation. They demonstrated that the use of insulation
externally in a composite panel enhanced the thermal and structural performance of
stud walls and increased their fire resistance rating significantly. Hence this research
recommends the use of the new composite panel system for cold-formed LSF walls.
This research also included steady state tensile tests at ambient and elevated
temperatures to address the lack of reliable mechanical properties for high grade cold-
formed steels at elevated temperatures. Suitable predictive equations were developed
for calculating the yield strength and elastic modulus at elevated temperatures.
In summary, this research has developed comprehensive experimental thermal and
structural performance data for both the conventional and the proposed non-load
bearing and load bearing stud wall systems under fire conditions. Idealized hot flange
temperature profiles have been developed for non-insulated, cavity insulated and
externally insulated load bearing wall models along with suitable equations for
predicting their failure times. A graphical method has also been proposed to predict
the failure times (fire rating) of non-load bearing and load bearing walls under
different load ratios. The results from this research are useful to both fire researchers
and engineers working in this field. Most importantly, this research has significantly
improved the knowledge and understanding of cold-formed LSF walls under fire
conditions, and developed an innovative LSF wall system with increased fire rating. It
has clearly demonstrated the detrimental effects of using cavity insulation, and has
paved the way for Australian building industries to develop new wall panels with
increased fire rating for commercial applications worldwide.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions iii
TABLE OF CONTENTS
Abstract i
Table of Contents iii
List of Figures vi
List of Tables xx
Statement of Original Authorship xxii
Acknowledgements xxiii
Chapter 1.0: Introduction 01-13
1.1: Cold-formed Steel Members 01
1.2: Need for Fire Resistant Structures 04
1.3: Fire Resistance of LSF Stud Wall Systems 06
1.4: Problem Definition 08
1.5: Aims of this Research 09
1.6: Research Method 11
1.7: Contents of Thesis 13
Chapter 2.0: Literature Review 14-51
2.1: Experimental Research 14
2.2: Analytical Research 30
2.3: Mechanical and Thermo Physical Properties of Steel Stud Wall Assembly Components at Elevated Temperatures
39
2.4: Literature Review Findings Relevant to this Research 49
Chapter 3.0: Experimental Work to Determine the Mechanical Properties of G500 Cold-Formed Steel at Elevated Temperatures.
52-74
3.1: Introduction 52
3.2: Experimental Investigation 54
3.3: Comparison of Reduction Factors with Results as Obtained by Other Researchers and as Recommended by Steel Design Codes
66
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions iv
3.4: Conclusion 74
Chapter 4.0: Thermal Performance of Gypsum Plasterboards and Composite Panels
75-130
4.1: Introduction 75
4.2: Test Setup and Procedure 76
4.3: Test Specimens 78
4.4: Conclusion 128
Chapter 5.0: Thermal Performance of Non-Load Bearing Wall Systems 131-188
5.1: Introduction 131
5.2: Test Specimens 132
5.3: Construction Details of Test Specimens 134
5.4 Test Set-up and Procedure 142
5.5 Observations, Results and Discussion 144
Chapter 6.0: Structural and Thermal Performance of Load Bearing Wall Systems
189-332
6.1: Introduction 189
6.2: Test Specimens 190
6.3: Construction Details of Test Specimens 195
6.4: Test Set-up and Procedure 205
6.5: Observations and Results 218
6.6: Discussions
332
Chapter 7: Discussions and Recommendations 333-376
7.1: Discussions 333
7.2: Simplified Method for the Determination of Failure Times of Wall Specimens 357
7.3 Essential Points to Consider for Thermal Modeling 371
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions v
7.4: Conclusion: 376
Chapter 8: Summary 377-381
8.1: Main Research Outcomes 379
8.2: Recommendations to the Construction Industry 380
8.3: Future Research 381
References 382-388
List of Figures Page No.
Chapter 1 1-13
Figure 1.1: Commonly Used Cold-formed Steel Structural Shapes 01
Figure 1.2: Strength of Steel at Elevated Temperature Relative to Yield Strength at Ambient Temperature
02
Figure 1.3: Applications of Cold-formed Steel Products 03
Figure 1.4: Transportable Houses 03
Figure 1.5: House Frames 03
Figure 1.6: Steel Stud Wall System 04
Figure 1.7: Wall Panel Showing Steel Channels Sections and Plasterboards 05
Figure 1.8: Fire Ratings of Some Exterior Wall Systems of Boral 07
Figure 1.9: New LSF Stud Wall System using a Composite Panel 10
Chapter 2 14-51
Figure 2.1: Construction of Assemblies 18
Figure 2.2: Typical Steel Frame Fabrication Layout for Wall Specimens 22
Figure 2.3: Location of Temperature Measurements and Simulation Boundaries
22
Figure 2.4: Structural Failure Modes 27
Figure 2.5: Total Horizontal Deflection for Load-bearing Systems 31
Figure 2.6: Thermal Bowing and Secondary Deflection 34
Figure 2.7: Stud End Conditions 34
Figure 2.8: Gypsum Plasterboard 39
Figure 2.9: Thermal Conductivity of Gypsum Plasterboard 43
Figure 2.10: Specific Volumetric Enthalpy of Gypsum Plasterboard 43
Figure 2.11: Specific Heat of Type X Gypsum Board 44
Figure 2.12: Thermal Conductivity of Type X Gypsum Board 45
Figure 2.13: Density Variation of Type X Gypsum Plasterboard on Heating 45
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions vi
Figure 2.14: Mass Loss in Gypsum Plasterboard on Heating 47
Chapter 3 52-74
Figure 3.1: Tensile Test Specimen 54
Figure 3.2: Furnace Details 56
Figure 3.3: Details of Test Rig and its Components 57-58
Figure 3.4 EDCAR (Experimental Data Collection and Recorder) 59
Figure 3.5 Strain Measurement using LSE 60
Figure 3.6: Typical Speckle Output for Strain Measurements 61
Figure 3.7: Comparison of Stress-Strain Curves using Strain Gauges and Laser Speckle Extensometer
62
Figure 3.8: Determination of (a) Yield strength and (b) Elastic modulus. 63
Figure 3.9: Stress-Strain Graphs at Different Temperatures 65
Figure 3.10: Graph Showing Strength Reduction Factors associated with Various Percentages of Yield Strength as Obtained from Tests
66
Figure 3.11: Yield Strength Reduction Factors 67
Figure 3.12: Modulus of Elasticity Reduction Factors 67
Figure 3.13: Variation of Yield Strength Reduction Factors with Temperature
68-69
Figure 3.14: Comparison of 0.2% Strength Reduction Factors with AS 4100 (SA, 1998) Recommendations
70
Figure 3.15: Comparison of Modulus of Elasticity Reduction Factors with AS 4100 (SA, 1998) Recommendations
70
Figure 3.16: Comparison of Yield Strength Reduction Factors with Test Results and Predictive Equation
72
Figure 3.17: Comparison of Elastic Modulus Reduction Factors with Test Results and Predictive Equation
72
Figure 3.18: Comparison between Ranawaka (2009) Equation and Predictive Equation in the determination of Yield Strength Reduction Factors
73
Figure 3.19: Comparison between Ranawaka (2009) Equation and Predictive Equation in the determination of Elastic Modulus Reduction Factors
73
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions vii
Chapter 4 75-130
Figure 4-1: View Showing Adapter Attached to Large Furnace for Carrying Out Fire Testing of Small Scale Specimens
77
Figure 4-2: Adapter Details 77
Figure 4-3: View Showing Plasterboard Specimen Installed for Fire Testing 78
Figure 4-4: Pressure Transducer used for Determining Furnace Chamber Pressure during Testing
78
Figure 4-5: Thermocouples on the Ambient Side of the Specimen 80
Figure 4-6: Instrumentation for Test Specimen 1 81
Figure 4-7: Fire Testing of Test Specimen 1 83
Figure 4-8: Time-Temperature Profile of Test Specimen 1 84
Figure 4-9: Temperature-Depth Profiles of Test Specimen 1 84
Figure 4-10: Instrumentation for Test Specimen 2 85
Figure 4-11: Fire Testing of Test Specimen 2 86
Figure 4-12: Time-Temperature Profile of Test Specimen 2 87
Figure 4-13: Temperature-Depth Profiles of Test Specimen 2 87
Figure 4-14: Instrumentation for Test Specimen 3 88
Figure 4-15: Fire Testing of Test Specimen 3 90
Figure 4-16: Time-Temperature Profile of Test Specimen 3 91
Figure 4-17: Temperature-Depth Profiles of Test Specimen 3 91
Figure 4-18: Instrumentation for Test Specimen 4 92
Figure 4-19: Fire Testing of Test Specimen 4 93
Figure 4-20: Time-Temperature Profile of Test Specimen 4 94
Figure 4-21: Temperature-Depth Profiles of Test Specimen 4 95
Figure 4-22: Instrumentation for Test Specimen 5 95
Figure 4-23: Time-Temperature Profile of Test Specimen 5 96
Figure 4-24: Temperature-Depth Profiles of Test Specimen 5 97
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions viii
Figure 4-25: Construction of Test Specimen 6 99
Figure 4-26: Instrumentation for Test Specimens from 6 to 15 99
Figure 4-27: Construction of Test Specimen 7 100
Figure 4-28: Construction of Test Specimen 8 101
Figure 4-29: Construction of Test Specimen 9 101
Figure 4-30: Fire Testing of Test Specimen 6 103
Figure 4-31: Time-Temperature Profile of Test Specimen 6 103
Figure 4-32: Temperature-Depth Profiles of Test Specimen 6 104
Figure 4-33: Test Specimen 7 Installed in the Furnace for Testing 104
Figure 4-34: Time-Temperature Profile of Test Specimen 7 105
Figure 4-35: Temperature-Depth Profiles of Test Specimen 7 105
Figure 4-36: Fire Testing of Test Specimen 8 106
Figure 4-37: Time-Temperature Profile of Test Specimen 8 106
Figure 4-38: Temperature-Depth Profiles of Test Specimen 8 107
Figure 4-39: Fire Testing of Test Specimen 9 107
Figure 4-40: Time-Temperature Profile of Test Specimen 9 108
Figure 4- 41: Temperature-Depth Profiles of Test Specimen 9 108
Figure 4-42: Construction of Test Specimen 11 110
Figure 4-43: Fire Testing of Test Specimen 11 111
Figure 4-44: Time-Temperature Profile of Test Specimen 10 112
Figure 4-45: Temperature-Depth Profiles of Test Specimen 10 113
Figure 4-46: Time-Temperature Profile of Test Specimen 11 113
Figure 4-47: Temperature-Depth Profiles of Test Specimen 11 114
Figure 4-48: Construction of Test Specimens 12, 13 and 14 116
Figure 4-49: Fire Testing of Test Specimen 11 119-120
Figure 4-50: Time-Temperature Profile of Test Specimen 12 121
Figure 4-51: Temperature-Depth Profiles of Test Specimen 12 121
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions ix
Figure 4-52: Time-Temperature Profile of Test Specimen 13 122
Figure 4-53: Temperature-Depth Profiles of Test Specimen 13 122
Figure 4-54: Time-Temperature Profile of Test Specimen 14 123
Figure 4-55: Temperature-Depth Profiles of Test Specimen 14 123
Figure 4-56: Construction of Test Specimen 15 125
Figure 4-57: Fire Testing of Test Specimen 15 126
Figure 4-58: Time-Temperature Profile of Test Specimen 15 126
Figure 4-59: Temperature-Depth Profiles of Test Specimen 15 127
Figure 4-60: Time-Temperature profiles for interface Ins-Pb2 of Test Specimens 6 to 15
128
Figure 4-61: Average Time-Temperature profile for interface Ins-Pb2 of Test Specimens 6 to 9 compared with Time-Temperature profile of Pb1-Pb2 interface temperature of Test Specimen 4
130
Chapter 5 131-188
Figure 5-1: Typical steel wall frame used to build NLB test wall specimens 134
Figure 5-2: Construction of Test Specimen 1 134
Figure 5-3: Thermocouple Locations for Test Specimen 1 135
Figure 5-4: Test Specimen 2 with a Joint in the Exposed Plasterboard over the Central Stud and Thermocouple Locations
136
Figure 5-5: Thermocouple Locations for Test Specimen 3 136
Figure 5-6: Construction and Placement of Test Specimen 4 in the Furnace 137
Figure 5-7: Thermocouple Locations for Test Specimens 4, 5 and 6 138
Figure 5-8: Construction of Test Specimen 5 138
Figure 5-9: Construction of Test Specimen 6 139
Figure 5-10: Thermocouple Locations for Test Specimens 7, 8 and 9 140
Figure 5-11: Construction of Test Specimen 9 141
Figure 5-12: Test Specimen placed in the specially built adapter in the large furnace
142
Figure 5-13: Test Specimen subjected to fire on one side 143
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions x
Figure 5-14: Test Specimen 1 after the fire test 145
Figure 5-15: Test Specimen 2 after the fire test 145
Figure 5-16: Time-Temperature Profile for Test Specimen 1 (No joints in plasterboard)
146
Figure 5-17: Time-Temperature Profile for Test Specimen 2 (With a joint in the exposed plasterboard over the central stud)
146
Figure 5-18: Time –Temperature Profiles of the Flanges in Stud No.1 of Test Specimens 1 and 2
148
Figure 5-19: Time –Temperature Profiles of the Flanges in Stud No.2 of Test Specimens 1 and 2
148
Figure 5-20: Time –Temperature Profiles of the Flanges in Stud No.3 of Test Specimens 1 and 2
149
Figure 5-21: Average Unexposed Surface Temperature of Test Specimens 1 and 2
150
Figure 5-22: Time-Temperature Profiles of Cavity facing surfaces of Specimens 1 and 2
151
Figure 5-23: Lateral Deflections of the Central Studs in Test Specimens 1 and 2
152
Figure 5-24: Test Specimen 3 after the fire test (no cavity insulation) 154
Figure 5-25: Test Specimen 4 after the fire test (glass fibre cavity insulation) 154
Figure 5-26: Test Specimen 5 after the fire test (rock fibre as cavity insulation)
155
Figure 5-27: Test Specimen 6 after the fire test (cellulose as cavity insulation)
156
Figure 5-28: Time-Temperature Profiles of Plasterboard surfaces in Test Specimen 3 (No Cavity Insulation)
159
Figure 5-29: Time-Temperature Profiles of Plasterboard surfaces in Test Specimen 4 (Cavity Insulation – Glass Fibre)
159
Figure 5-30: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 5 (Cavity Insulation-Rock Fibre)
160
Figure 5-31: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 6 (Cavity Insulation – Cellulose Fibre)
160
Figure 5-32: Time-Temperature Profiles across Studs in Test Specimen 3 (No Cavity Insulation)
164
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xi
Figure 5-33: Time-Temperature Profiles across Studs in Test Specimen 4 (Cavity Insulation – Glass Fibre)
164
Figure 5-34: Time-Temperature Profiles across Studs in Test Specimen 5 (Cavity Insulation-Rock Fibre)
165
Figure 5-35: Time-Temperature Profiles across Studs in Test Specimen 6 (Cavity Insulation – Cellulose Fibre)
165
Figure 5-36: Time-Temperature Profiles across the Cross-section of Test Specimen 3 (No Cavity Insulation)
167
Figure 5-37: Time-Temperature Profiles across the Cross-section of Test Specimen 4 (Cavity Insulation – Glass Fibre)
167
Figure 5-38: Time-Temperature Profiles across the Cross-section of Test Specimen 5 (Cavity Insulation – Rock Fibre)
168
Figure 5-39: Time-Temperature Profiles across the Cross-section of Test Specimen 6 (Cavity Insulation – Cellulose Fibre)
168
Figure 5-40: Deflection-Time Profiles of Test Specimen 3 (No Cavity Insulation)
170
Figure 5-41: Deflection-Time Profiles of Test Specimen 4 (Cavity Insulation – Glass Fibre)
170
Figure 5-42: Deflection-Time Profiles of Test Specimen 5 (Cavity Insulation – Rock Fibre)
171
Figure 5-43: Deflection-Time Profiles of Test Specimen 6 (Cavity Insulation – Cellulose Fibre)
171
Figure 5-44: Test Specimen 7 after the fire test (Glass fibre as external insulation)
174
Figure 5-45: Test Specimen 8 after the fire test (Rock fibre as external insulation)
174
Figure 5-46: Test Specimen 9 after the fire test (Cellulose fibre as external insulation)
175
Figure 5-47: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 7 (External Insulation-Glass Fibre)
176
Figure 5-48: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 8 (External Insulation-Rock Wool)
177
Figure 5-49: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 9 (External Insulation-Cellulose Fibre)
177
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xii
Figure 5-50: Time-Temperature Profiles across Studs in Test Specimen 7 (External Insulation-Glass Fibre)
181
Figure 5-51: Time-Temperature Profiles across Studs in Test Specimen 8 (External Insulation-Rock Fibre)
182
Figure 5-52: Time-Temperature Profiles across Studs in Test Specimen 9 (External Insulation-Cellulose Fibre)
182
Figure 5-53: Time-Temperature Profiles over the Entire Cross-section of Test Specimen 7 (External Insulation-Glass Fibre)
184
Figure 5-54: Time-Temperature Profiles over the Entire Cross-section of Test Specimen 8 (External Insulation-Rock Fibre)
184
Figure 5-55: Time-Temperature Profiles over the Entire Cross-section of Test Specimen 9 (External Insulation-Cellulose Fibre)
185
Figure 5-56: Lateral Deflection -Time Profiles of Test Specimen 7
(External Insulation-Glass Fibre)
186
Figure 5-57: Lateral Deflection -Time Profiles of Test Specimen 8 (External Insulation-Rock Fibre)
186
Figure 5-58: Lateral Deflection -Time Profiles of Test Specimen 9 (External Insulation-Cellulose Fibre)
187
Chapter 6 189-332
Figure 6-1 (a): Basic Local Failure Modes 189
Figure 6-1 (b): Basic Global Failure Modes 190
Figure 6-2: Test Wall Frame 191
Figure 6-3: Stud to Plasterboard Connections 192
Figure 6-4: Protection of Joints 193
Figure 6-5: Stud to Track Connection at the Top 195
Figure 6-6: Construction of Test Specimen 2 196
Figure 6-7: Fixing of Face Plasterboard on the Ambient Side of Test Specimen 3
197
Figure 6-8: Construction of Test Specimen 4 Using Glass Fibre as Cavity Insulation
198
Figure 6-9: Construction of Test Specimen 5 using Rock Fibre as Cavity Insulation
199-200
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xiii
Figure 6-10: Construction of Test Specimen 6 using Cellulose Fibres as Cavity Insulation
201
Figure 6-11: Construction of Test Specimen 7 using Glass Fibres as External Insulation
202
Figure 6-12: Construction of Test Specimen 8 using Rock Fibres as External Insulation
203
Figure 6-13: Construction of Test Specimen 9 using Cellulose Fibres as External Insulation
204
Figures 6-14: Details of Furnace Operation and Components 206-207
Figure 6-15: Loading Frame 209
Figure 6-16: Loading Arrangement 209-210
Figure 6-17: Hydraulic Pump and its Connections 211
Figure 6-18: Test Set-up for Ambient Temperature Test 212
Figure 6-19: LVDTs Used in the Measurement of Axial Shortening and Out-of-plane Deflection of Test Specimen Wall
213
Figure 6-20: Thermocouple Locations for Load Bearing Wall Specimens 215
Figure 6-21: Infrared Gun Used for the Measurement of Ambient Side Temperatures
217
Figure 6-22: Test Specimen Complete with all its Instrumentation Ready for Fire Test
217
Figure 6-23: Failure of Test Specimen 1 218
Figure 6-24: Load Vs Axial Deformation - Profiles of Test Specimen 1 at Ambient Temperature
219
Figure 6-25: Fire Performance Test of Specimen 2 220
Figure 6-26: Detachment and Opening of Plasterboard joints Caused by Shrinkage
221
Figure 6-27: Test Specimen 2 after Removing the Exposed Plasterboard Layer
222
Figure 6-28: Stud Failure Initiated by Plasterboard Fall-off 223
Figure 6-29: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 2
225
Figure 6-30: Time-Temperature Profiles across Studs 1 to 4 of Test Specimen 2 226-227
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xiv
Figure 6-31: Axial Deformation Plots for Studs of Test Specimen 2 228-229
Figure 6-32: Lateral Deflection -Time Profiles of Test Specimen 2 at Mid-Height
229
Figure 6-33: Close up of Test Specimen 3 after Removing the Exposed Plasterboards
234-235
Figure 6-34: Studs of Test Specimen 3 after the Fire Test 235
Figure 6-35: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 3
237-238
Figure 6-36: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 3
239-240
Figure 6-37: Time-Temperature Profiles across Central Studs at Mid-height in Test Specimen 3
241
Figure 6-38: Axial Deformation Plots for Studs of Test Specimen 3 242-243
Figure 6-39: Lateral Deflection-Time Plots of Test Specimen 3 243-244
Figure 6-40: Axial Load -Time Profile of Test Specimen 3 during Fire Test 245
Figure 6-41: Test Specimen 4 after the Fire Test 248-250
Figure 6-42: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 4
252-253
Figure 6-43: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 4
254-255
Figure 6-44: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 4
255
Figure 6-45: Outward Lateral Deflection of Test Specimen 4 at Failure 257
Figure 6-46: Axial Deformation Plots for Studs of Test Specimen 4 258
6-47: Lateral Deflection-Time Plots of Test Specimen 4 259-260
Figure 6-48: Axial Load -Time Profile of Test Specimen 4 during Fire Test 260
Figure 6-49: Test Specimen 5 after the Fire Test 263-264
Figure 6-50: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 5
267-268
Figure 6-51: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 5
269
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xv
Figure 6-52: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 5
270
Figure 6-53: Specimen Behaviour during the Test 272-273
Figure 6-54: Ambient Side of Wall after Fire Test 273-274
Figure 6-55: Axial Deformation Plots for Studs of Test Specimen 5 274
Figure 6-56: Lateral Deflection-Time Plots of Test Specimen 5 275
Figure 6-57: Axial Load -Time Profile of Test Specimen 5 during Fire Test 276
Figure 6-58: Test Specimen 6 after the Fire Test 279-281
Figure 6-59: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 6
283
Figure 6-60: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 6
284-285
Figure 6-61: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 6
286
Figure 6-62: Inward Thermal Bowing of Test Specimen in the Initial Period of the Test
287
Figure 6-63: Axial Deformation Plots for Studs of Test Specimen 6 288
Figure 6-64: Lateral Deflection-Time Plots of Test Specimen 6 289-290
Figure 6-65: Axial Load -Time Profile of Test Specimen 6 during Fire Test 290
Figure 6-66: Test Specimen 7 after the Fire Test 292-294
Figure 6-67: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 7
297
Figure 6-68: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 7
298-299
Figure 6-69: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 7
299
Figure 6-70: Axial Deformation Plots for Studs of Test Specimen 7 300
Figure 6-71: Lateral Deflection-Time Plots of Test Specimen 7 301-302
Figure 6-72: Axial Load -Time Profile of Test Specimen 7 during Fire Test 302
Figure 6-73: Test Specimen 8 after the Fire Test 305-307
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xvi
Figure 6-74: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 8
310
Figure 6-75: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 8
311-312
Figure 6-76: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 8
313
Figure 6-77: View of Loading Arrangement 314-315
Figure 6-78: Axial Deformation Plots for Studs of Test Specimen 3 315-316
Figure 6-79: Lateral Deflection-Time Plots of Test Specimen 8 316-317
Figure 6-80: Axial Load -Time Profile of Test Specimen 8 during Fire Test 318
Figure 6-81: Test Specimen 9 after the Fire Test 320-322
Figure 6-82: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 9
324-325
Figure 6-83: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 3
326-327
Figure 6-84: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 9
327
Figure 6-85: Axial Deformation Plots for Studs of Test Specimen 9 328-329
Figure 6-86: Lateral Deflection-Time Plots of Test Specimen 9 329-330
Figure 6-87: Axial Load -Time Profile of Test Specimen 9 during Fire Test 331
Chapter 7 333-376
Figure 7-1: Time-temperature Profiles for the Central Stud in Non-Load Bearing Wall Test Specimens 4 and 7 (Glass fibre insulation)
334
Figure 7-2: Time-temperature Profiles for the Central Stud in Non-load Bearing Wall Test Specimens 5 and 8 (Rock fibre insulation)
335
Figure 7-3: Time-temperature Profiles for the Central Stud in Non-Load Bearing Wall Test Specimens 6 and 9 (Cellulose fibre insulation)
335
Figure 7-4: Time-temperature Profiles for the Central Stud Hot Flanges in Non-Load Bearing Wall Test Specimens 4 to 9
336
Figure 7-5: Hot Flange Temperatures of the Central Stud in NLB Wall Test Specimens 3 to 9
337
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xvii
Figure 7-6: Temperature Difference across the Central Studs C/S in NLB Wall Test Specimens 3 to 9
338-339
Figure 7-7: Ambient Side Temperature of External Plasterboard 2 in Test Specimens 3 to 9
339-340
Figure 7-8: Temperature on the Ambient Face of the Non-load Bearing Wall Test Specimens 3 to 9
340-341
Figure 7-9: Fall off Times of Plasterboard 2 in Non-Load Bearing Wall Test Specimens 4 to 9
341
Figure 7-10: Ambient Side Time-temperature Profiles of Test Specimens 4 to 9
342
Figure 7-11: Average Time-temperature Profiles for the Central Studs in Load Bearing Wall Test Specimens 5 and 8
343
Figure 7-12: Average Time-temperature Profiles for the Central Studs in Load Bearing Wall Test Specimens 6 and 9
344
Figure 7- 13: Average Hot Flange Temperatures of the Central Studs of Load Bearing Wall Test Specimens 3 to 9
345
Figure 7-14 : Average Temperature Difference across the Central Studs for Load Bearing Wall Test Specimens 3 to 9
346
Figure 7-15: Average Temperature Difference across Central Studs and their Lateral Deformations versus Time for Test Specimens 5 and 8
347
Figure 7-16: Average Temperature Difference across Central Studs and their Lateral Deformations versus Time for Test Specimens 6 and 9
347
Figure 7- 17:Average Lateral Deformations of the Central Studs in Load Bearing Wall Test Specimens
348
Figure 7-18: Average Time-temperature Profiles of Hot Flanges for the Central Studs in Test Specimens 4 to 9
349
Figure 7-19: Average Time-temperature Profiles of Cold Flanges for the Central Studs in Test Specimens 4 to 9
350
Figure 7-20: Temperature Difference across the Central Studs in Cavity Insulated and Externally Insulated Specimens
350
Figure 7-21: Average Time-temperature Profiles of Pb2-Cav Surface in Test Specimens 3 to 9
352
Figure 7- 22: Average Time-Temperature Profiles on the Ambient Side of the Exposed Base Layer Plasterboard 2 in LBW Test Specimens 3 to 9
353
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xviii
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xix
Figure 7- 23: Ambient Side Temperatures of LBW Test Specimens 3 - 9 354
Figure 7-24: Variation of Yield Strength Reduction Factor of 1.15 mm G500 Steel with respect to Temperature
357
Figure 7-25: Idealized Hot Flange Temperatures of Load Bearing Test Specimens 2 to 9
360
Figure 7-26: Development of Hot Flange Failure Times for a Given Load Ratio
360
Figure 7-27: Determination of Hot Flange Failure Times using Load Ratio 362
Figure 7-28: Load Ratio Vs Critical Hot Flange Temperatures at Stud Failure 365
Figure 7-29: Load Ratio Vs Stud Failure Times for Test Specimen 2 366
Figure 7-30: Load Ratio Vs Stud Failure Times for Test Specimen 3 366
Figure 7-31: Load Ratio Vs Stud Failure Times for Test Specimen 4 367
Figure 7-32: Load Ratio Vs Stud Failure Times for Test Specimen 5 367
Figure 7-33: Load Ratio Vs Stud Failure Times for Test Specimen 6 368
Figure 7-34: Load Ratio Vs Stud Failure Times for Test Specimen 8 368
Figure 7-35: Load Ratio Vs Stud Failure Times for Test Specimen 9 369
Figure 7-36: Load Ratio Vs Stud Failure Times for all Test Specimens Using Predictive Equations
369
List of Tables
Chapter 2 14-51
Table 2.1: Fire Resistance of Typical Floors, Walls and Partitions (From SCI, 1993)
16
Table 2.2: Small Scale Assembly Parameters and Fire Test Results (Sultan & Lougheed, 1994)
17
Table 2.3: Full Scale Fire Test Specimens (Used by Gerlich, 1995) 19
Table 2.4: Summary of Fire Resistance Tests on Load-Bearing LSF Walls by Kodur et al. (1999)
25
Table 2.5 Summary of Fire Resistance Tests on Load-Bearing LSF Walls by Alfawakhiri (2001)
26
Table 2.6: Mechanical Properties of Australian Manufactured Plasterboards Goncalves et al., (1996)
42
Chapter 3 52-74
Table 3.1: Mechanical Properties of 1.15 mm G500 CFS at Ambient and Elevated Temperatures.
64
Table 3.2: Reduction Factors for Yield Strength and Modulus of Elasticity of 1.15 mm G500 Steel
65
Chapter 4 75-130
Table 4-1: Details of Plasterboard and Composite Panel Test Specimens 79
Table 4-2: Time–Temperature profile of the ambient side of the insulation (Ins-Pb2 interface) in Test Specimens 6, 7, 8 and 9 using glass fibre as insulation material
109
Table 4-3: Time–Temperature Profile of the Ambient Side of the Insulation (Ins-Pb2 interface) in Test Specimens 10 and 11 using Rock Fibre as Insulation Material
115
Table 4-4: Time–Temperature profile of the ambient side of the insulation (Ins-Pb2 interface) in Test Specimens 12, 13 and 14 using cellulose fibre as insulation material
124
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xx
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xxi
Chapter 5 131-188
Table 5-1: Details of non-load bearing wall specimens 133
Table 5-2: Central Stud Temperatures of Test Specimens 1 and 2 150
Table 5.3: Hot Flange Temperature versus Time for the Central Stud 166
Table 5-4: Failure Times of Wall Components in minutes 172
Table 5-5: Hot Flange Temperature versus Time for the Central Stud 183
Table 5-6: Failure times of Wall Components in Minutes 187
Chapter 6 189-332
Table 6-1: Details of Test Specimen Configuration 194
Chapter 7 333-376
Table 7-1: Failure Times of Test Specimens 355
Table 7-2: Stud Reversal Times for Cavity Insulated and Externally Insulated Specimens along with the Corresponding Temperatures
356
Table 7-3: Comparison of Predicted Hot Flange (HF) Failure Times of Load Bearing Wall Specimens with Actual Local Buckling of HF (minutes) at a Load Ratio of 0.2
361
Table 7-4: Comparison of Predicted Failure Times of Non-load Bearing Wall Specimens with their Actual Failure Times.
363
Table 7-5: Predictive Equations for Obtaining Stud Failure Times for Different Wall Models
370
Table 7-6: Table Comparing the Actual Stud Reversal Times and Wall Failure Times with the Predicted Values
371
Statement of Original Authorship
The work contained in this thesis has not been previously submitted for a degree or
diploma at any other higher education institution. To the best of my knowledge and
belief, the thesis contains no material previously published or written by another person
except where due reference is made.
Prakash Nagaraj Kolarkar
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xxii
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions xxiii
Acknowledgements
The research described in this report was carried out in the School of Urban
Development, Queensland University of Technology, Australia.
I would like to thank my supervisor Prof. Mahen Mahendran of the Queensland
University of Technology for his inspiration, guidance and enthusiasm. Thanks also to
my fellow researchers Banduka Heva and Gunalan for their help during my experimental
work and to the technical staff in the laboratory, Arthur, Brian and Jim Hazelman for
their help in the testing and data acquisition work.
I would also like to thank my parents, especially my mother who has been a constant
source of encouragement, my wife and son who exhibited immense patience and
willingness to work around my schedule, and my sister who was happy to manage all my
commitments back in my home country. Their support has been of immense help in the
completion of this thesis.
Chapter 1: Introduction
1.1: Cold-formed Steel Members
Steel members are widely used in buildings due to their advantages of high strength,
good ductility and fast fabrication and erection. Two types of structural steel are used
in the building industry, i.e. hot-rolled and cold-formed steels. In cold-formed steel
products, the strength comes from the material and how it is shaped. The load bearing
capacity of a thin flat sheet of steel can be greatly increased if it is formed into an
efficient multi-sided cross-section. The strength to weight ratio of cold-formed steel
products is very favourable in comparison to the thicker hot-rolled steel products.
Figure 1.1 shows some of the commonly used cold-formed steel structural shapes.
The depth of the members generally ranges from 50 mm to 300 mm while their
thicknesses are in the range of 0.75 to 3 mm.
Plain C-section Lipped C-section Swage Beam Hat Section
Z – Section Back to Back C-Sections Box C-Section Figure 1.1: Commonly Used Cold-formed Steel Structural Shapes
There are many differences between hot-rolled and cold-formed steel sections.
1) Many complex structural shapes are possible with cold-formed steel. There is a
gain in material strength and hardness due to cold working effects.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 1
2) The high strength to weight ratio of cold-formed steel makes it much easier and
economical to mass produce, transport and install cold-formed products.
3) The thicker hot-rolled sections generally prevent the occurrence of local
buckling before yielding. In contrast, local buckling may become a concern for
cold-formed steel sections as it can occur at stresses well below the yield point.
4) Cold-formed steel loses strength more rapidly than hot-rolled steel when
exposed to increasing temperatures. According to Sidey and Teague (1988) hot-
rolled steel retains its full strength up to 4000C, beyond which the strength
quickly decreases. The loss of strength in the case of cold-formed steel is 10 –
20% more than that of hot-rolled steel as shown in Figure 1.2
Figure 1.2: Strength of Steel at Elevated Temperature Relative to Yield
Strength at Ambient Temperature (NAHB Research Centre, 2002)
5) Cold-formed steel exhibits superior corrosion resistance than hot-rolled steel due
to the improved galvanising and other coating technology. The protective
coating system is not damaged during the cold-forming process (Davies 2000).
6) Cold-formed steel products exhibit more accurate complex shapes of precise
lengths due to the recent progress in rolling and forming technologies.
The main applications of cold-formed steel products have been in elements such as
purlins and sheet rails, cladding and decking, pallet racking and shelving. Their
strength, lightweight, versatility, non-combustibility, ease of prefabrication and
handling has made cold-formed steel members very popular in the building industry.
Trusses, wall Frames, posts and beams made of cold-formed steel as shown in Figure
1.3 are being regularly used. Cold-formed steel is an attractive alternative over other
materials for constructing the entire buildings as shown in Figures 1.4 and 1.5.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 2
a: Trusses b:Wall Frames
c: Posts d: Beams
Fig.1.3: Applications of Cold-formed Steel Products (Steelbuilt Kit Homes, 2005)
Figure 1.4: Transportable Houses (Steelbuilt Kit Homes, 2005)
Figure 1.5: House Frames (Steelbuilt Kit Homes, 2005)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 3
1.2: Need for Fire Resistant Structures
Fire resistant barriers play an important role in maintaining building integrity and
reducing the spread of fire from the room of origin to adjacent compartments. The
traditional method of stud wall construction is with light timber framing and sheet
material linings. However, there has been an increasing demand for prefabricated light
steel frame systems (LSF). Because of its high strength and yet good forming
properties, the material generally used is galvanised mild steel. Steel track and stud
are seen as an environmentally friendly, recyclable alternative to timber stud system.
The replacement of timber with steel becomes more prevalent in regions where timber
resources are scarce and also in commercial or community applications where other
advantages such as speed of assembly and fire retardance are more important.
Figure 1.6: Steel Stud Wall System (Gyprock, 2005)
Partition wall panels composed of a cold-formed steel frame lined with one- or two-
side sheathing (for example plasterboards) have been widely used in building
constructions since 1940s. The panels are typically constructed by first connecting
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 4
studs and tracks with rivets to form the frame, and then connecting sheathing boards
to the frame with self-drilling screws (see Figure 1.6). These panels can be easily
assembled to fabricate load-bearing as well as non-load-bearing partition walls.
Plasterboard
Stud
Plasterboard
Figure 1.7: Wall Panel Showing Steel Channel Sections and Plasterboards
Cold-formed thin-walled (CF-TW) steel channels are the predominant sections used
as load bearing wall studs in light-weight steel construction. Under fire conditions,
because of their thinness, CF-TW steel sections heat up quickly resulting in fast
reduction in their strength and stiffness. However, if gypsum boards are combined
with thin-walled steel channels as shown in Figure 1.7 to form steel stud walls, their
fire resistant performance will improve since the gypsum boards can protect the steel
studs from fire exposure, thus delaying temperature rises in the steel studs.
When the walls are used as part of a fire resistant construction, they should satisfy
three fire resistant requirements, namely stability, insulation and integrity.
a) Load-bearing Capacity (Stability)
For load-bearing elements of structure, the test specimen shall not collapse in such a
way that it no longer performs the load-bearing function for which it was constructed.
The purpose of stability requirements in fire is two-fold. Internally, to maintain the
viability of escape routes for a sufficient period to allow safe evacuation and search
and rescue; and externally, to prevent toppling of the building which would endanger
people in the vicinity. The stability requirements to ensure safe egress are independent
of storey level but the external risk from overbalancing of the building depends both
on the height of the structure and the level on which the fire occurs.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 5
b) Insulation
For elements of structure such as walls and floors which have a function of separating
two parts of a building.
a) The average temperature of the unexposed face of the specimen shall not increase
above the initial temperature by more than 140°C.
b) The maximum temperature at any point of this face shall not exceed the initial
temperature by more than 180°C.
c) Integrity
Initial integrity failure shall be deemed to have occurred when a cotton pad is ignited
or when sustained flaming, having duration of at least 10 s, appears on the unexposed
face of the test specimen. It is required to maintain structural integrity during a fire to
avoid structural collapse and to prevent spread of flame and smoke into adjacent
areas. Ultimate integrity failure shall be deemed to have occurred when collapse of
the specimen occurs or at an earlier time based upon integrity and insulation criteria.
When LSF stud wall panels are used as load-bearing walls, sufficient fire resistance is
essential to
prevent or delay the spread of fire within the building or to another building
prevent sudden collapse of building components for the safety of the people and
the fire fighting personnel and assure integrity over a specific interval of time to
facilitate the safe evacuation of the people and allow the fire fighters to operate
safely is a major issue.
The extent to which these walls can withstand fire conditions without losing on
integrity, insulation and stability is known as fire resistance rating.
1.3: Fire Resistance of LSF Stud wall Systems
Non-load bearing LSF stud wall systems have an established history of use, mainly in
light industrial and commercial partitioning. Advantages over timber framing include:
Light-weight nature and dimensional stability of the frame, Speed and ease of frame
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 6
erection (often friction fit connections of studs to top and bottom tracks), No lining
delays due to high moisture content of framing, Aesthetic quality of finished wall and
Demountability. These advantages have resulted in a ready acceptance of non-load
bearing LSF stud wall systems as ‘infill’ partitioning in buildings, which have a
conventional structural shell, such as reinforced concrete or masonry construction. In
response to a market demand for fire separations in this area of light industrial and
commercial partitioning, lining manufacturers have developed, tested and published a
range of fire resistance ratings. In Australia the details of tested non-load bearing LSF
stud wall systems are published by Boral and Gyprock (see Figure 1.8). They have
prescribed steel stud walls with plasterboard linings achieving fire resistance ratings
ranging from 60 to 120 minutes. These systems are based on full scale fire resistance
tests against the standard IS0 834 fire curve.
Figure 1.8: Fire Ratings of Some Exterior Wall Systems of Boral (Gyprock, 2005)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 7
Load bearing LSF stud wall systems are less likely to be used as ‘infill’ commercial
partitioning, and will more likely form part of a total LSF construction system. With
the developing use of LSF in load bearing applications the demand for fire resistance
ratings has increased.
1.4: Problem Definition
Cold-formed thin-walled (CF-TW) steel channels are the predominant sections used
as load bearing wall studs in light-weight steel construction. Under fire conditions,
CF-TW steel sections (high section factor) heat up quickly resulting in a rapid
reduction in their strength and stiffness. The use of high strength steels is also
becoming popular in load bearing LSF stud wall construction. The structural
behaviour of high strength steel stud walls is yet to be researched. Also, in Australia
there is no data available on the fire ratings of load-bearing steel stud wall systems.
In parallel with the growing interest in LSF stud wall systems, the understanding and
application of specific Fire Engineering Design is used increasingly for the fire safety
design of buildings. To more accurately apply Fire Engineering Design, a better
understanding of the fire performance of components constituting the LSF stud walls
systems is required. With increasing demand for higher fire ratings of LSF stud wall
systems, the current practice is to prescribe more than two layers of gypsum
plasterboard lining on either side of the cold-formed steel frame making the entire
construction process more labour oriented and expensive. Therefore there is an urgent
need to develop innovative LSF stud wall systems made with improved plasterboard
and insulation systems and verify their improved fire performance.
Several researchers have carried out investigations to determine the impact of
different types of cavity insulations on the thermal performance of stud wall systems.
Sultan and Lougheed (1994) noted that rock and cellulose fibres when used as cavity
insulation improved the fire ratings of the wall systems whereas glass fibres hardly
contributed to any improvement in the thermal performance of the stud wall system.
Feng et al. (2003) reported that the thermal performance of non-load bearing wall
specimens improved with the use of cavity insulations. However, Sultan (1995)
observed that glass fibre cavity insulation has no impact on the thermal performance
of stud walls whereas cellulose fibre cavity insulation actually reduces the fire
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 8
resistance. Hence, as very limited data is available about the thermal performance of
non-load bearing and load bearing wall systems and past research has only provided
contradicting results about the benefits of cavity insulation to the fire rating of stud
wall systems, it is necessary to conduct further research by undertaking fire tests on
both non-load bearing and load bearing wall models with and without the use of
cavity insulations to increase the knowledge in this field and provide definitive
methods of improving the fire ratings of the stud wall systems.
The current practice of constructing non-load bearing and load bearing walls is not
very favourable when considering their role as fire resistant barriers. The use of glass
fibres and mineral wool as cavity insulation has only resulted in decreasing the
stability of load-bearing walls due to increased temperature gradients across the wall
and thus promoting larger lateral deflections leading to an early collapse of the wall.
By undertaking a detailed investigation into the structural behaviour of high strength
cold-formed steel studs in load bearing walls under simulated fire conditions and also
studying the fire performance of non-load bearing and load bearing wall panels with
and without insulation using both small scale and large scale fire tests it is proposed
we can fully understand their thermal and structural performances and hence develop
simple design rules that will contribute to the improvement of fire safety design.
1.5: Aims of This Research
Overall aim of this research is to improve the knowledge and understanding of the
structural and thermal performance of both conventional and innovative cold-formed
high strength steel stud wall systems (load bearing and non-load bearing) under
simulated fire conditions and develop reliable and simple methods to predict their fire
resistance rating.
Specific tasks of this research are:
1) To design and build a custom made gas furnace suitable for the standard fire
tests of both small and large scale LSF stud wall systems.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 9
2) To design and build a compression loading frame capable of loading the
individual studs of a large scale LSF stud wall specimen.
3) To conduct both small scale and large scale fire tests to determine the fire
performance and thermal deformations of non-load bearing and load-bearing
stud wall systems using the developed fire test rig.
4) To investigate the structural and thermal performance of currently used cold-
formed high strength steel (LSF) stud wall systems (both load bearing and non-
load bearing) with and without cavity insulations under simulated fire
conditions with temperatures of up to 10000C using the developed fire test rig.
5) To develop new cold-formed steel stud wall systems (both load bearing and
non-load bearing) with improved fire performance based on a new composite
panel in which insulation is located externally between two plasterboards (see
Figure 1.9), and investigate their structural and thermal performance
Composite panel with insulation between two layers of plasterboard
Figure 1.9: New LSF Stud Wall System using a Composite Panel
6) To determine the fire resistance rating (failure times) of load bearing and non-
load bearing cold-formed steel stud wall systems under fire attack from one
side based on full scale fire tests.
7) To determine experimentally the fire performance of Gypsum plasterboards
8) To determine the fire performance of different types of insulations such as
glass fibre, rock fibre and cellulose fibre including the effect of insulation
density and thickness on the fire performance of stud wall systems.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 10
9) To determine experimentally the temperature effects on the mechanical
properties of cold-formed steel and develop empirical equations to predict the
yield strength and modulus of elasticity at elevated temperatures.
10) To identify deficiencies in the conventional stud wall systems and the
improvements provided by the new stud wall systems assembled with the
composite panels shown in Figure 1.9.
11) To develop idealised time-temperature profiles for existing and proposed cold-
formed steel stud wall systems exposed to standard (cellulosic) fire curve.
12) To provide accurate structural and thermal performance data for the numerical
modelling of cold-formed steel stud wall systems under fire conditions.
13) To provide simple methods to determine the fire resistance rating of LSF load
bearing walls under different loading conditions.
1.6: Research Method
The research method essentially consisted of the following steps.
Step 1 - Literature Review: A comprehensive literature review on previous
experimental and analytical works, mechanical and thermo-physical properties of
steel stud wall assembly components at elevated temperatures, cold-formed steel, and
cavity insulations.
Step 2 - Establish a fire research laboratory at QUT: A propane fired gas furnace was
specially designed and constructed. The furnace was designed to deliver heat in
accordance with the standard time-temperature curve as proposed by AS 1530.4 (SA,
2005) or ISO 834-1 (ISO, 1999). A compression loading frame capable of loading the
individual studs of a full scale steel stud wall system was also designed and built for
conducting load bearing tests.
Step 3 - Conduct Fire Tests: To determine the fire resistance rating, standard fire tests
were carried out by exposing one side of the wall specimens to heat in the specially
designed furnace with controlled fuel input to achieve the specified time-temperature
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 11
curve and simulate the specimen’s exposure to heat in a real fire. The test specimens
were representative of the construction elements subject to loadings and end
constraints similar to the conditions of actual components. Such standardised furnace
tests provided good comparative data for systems tested under identical conditions.
Following tests were conducted to improve the knowledge and understanding of the
structural and thermal performance of cold-formed steel stud wall systems.
1) The mechanical properties of 1.15 mm G500 cold-formed steel at elevated
temperatures using steady state tests in an electric furnace. Tensile tests were carried
out in the elevated temperature range of 1000C to 8000C at intervals of 1000C.
2) The thermal properties of Gypsum Plasterboards (FireSTOP, Boral Industries) was
studied in a temperature range from 200C to 11000C using the gas furnace.
3) Fire tests on different types (Based on number of plasterboards, type of insulation,
number of joints) of small scale non-load bearing wall specimens with and without
cavity insulations. An adapter for the large propane fired gas furnace was designed
and constructed for these tests. The adapter requires the use of only a single burner for
the small scale tests as against the six burners for the full scale tests.
4) Fire tests of specially designed composite panels to improve the fire resistance
rating of wall systems.
5) Ultimate load test on large scale steel stud wall specimen at ambient temperature. A
loading frame was specially designed and built to load the individual studs.
6) Fire tests in accordance with AS 1530.4 (SA, 2005) were carried out on large scale
load bearing wall specimens to study their thermal and structural response at elevated
temperatures. A total of 16 transducers (to measure the axial shortening and lateral
displacement of the wall) and 45 to 57 K type thermocouples (to measure the
temperature at various locations across the wall) were used.
Step 4 - Develop new improved wall systems: New cold-formed LSF wall systems
with increased fire resistance rating and lower lateral deformations than the
conventional cavity insulated systems was achieved by the use of external insulation.
The improvement was validated by conducting several fire tests of both small scale
and large scale wall systems.
Step 5: -Develop simple methods to predict fire performance of LSF stud wall
systems: Simple predictive models for the mechanical properties of high strength
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 12
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 13
steels at elevated temperatures, the idealised time-temperature profiles of studs in LSF
walls under fire conditions and their fire resistance rating as a function of varying
arrangements of plasterboards and insulation were developed using simple methods to
predict the failure times of non-load bearing and load bearing wall models.
1.7: Contents of Thesis
Chapter 2: Past research on thermal performance of steel stud wall systems at elevated
temperatures is presented covering a range of research papers, reports and thesis.
Chapter 3: Presents detailed experimental work carried out to determine the
deterioration of mechanical properties of high grade (G500) cold-formed steel with
increasing temperatures. The chapter focuses on high grade steel as it is fast gaining
popularity in the construction industry.
Chapter 4: Deals with a series of small scale experiments performed to determine the
thermal performance of gypsum plasterboard specimens and their composite panels
using different types of insulating material of varying thickness and density.
Chapter 5: This chapter examines and compares the thermal performance of nine
small scale non-load bearing wall specimens built using cold-formed steel frame,
gypsum plasterboards and various types of insulating material.
Chapter 6: This chapter examines and compares the structural and thermal
performance of nine large scale load bearing wall specimens built in a manner similar
to the small scale non-load bearing wall specimens.
Chapter 7: This chapter presents the outcomes of the tests performed on non-load
bearing and load bearing wall specimens and compares the structural and thermal
performance of the conventional wall models with the new models proposed in this
thesis. It then develops idealised time-temperature profiles and simple fire design
methods for various LSF stud wall systems considered in this study
Chapter 8: Presents the main findings and recommendations.
Chapter 2: Literature Review
This chapter presents a detailed literature review that covers a range of research
papers, reports and theses in the field of fire performance of light gauge steel frame
systems.
2.1: Experimental Research
Son and Shoub (1973) carried out fire endurance tests on two load-bearing stud wall
assemblies with glass fibre batt cavity insulation. Each assembly consisted of double
module walls of gypsum board and steel studs. The outer plasterboards were type X
Gypsum boards 15.9 mm thick while the inner ones facing the cavity between the
walls were 12.7 mm in thickness. Studs used were lipped channel sections (76.2 x
44.5 x 12.7 x 1.21mm). The glass fibre insulation used in assembly two was thicker
than the one used in assembly one. A uniformly distributed load of 15 kN/m was
applied to each wall. On exposure to fire from one side, the structural failure of the
fire exposed wall in assembly 1 occurred in 42 minutes as compared to 67 minutes in
assembly two. In both assemblies, the structural failure occurred only after the
collapse of the exposed plasterboard. It was also observed, that as compared with
assembly one the heat penetration in the second assembly was much slower. This was
attributed to the thicker insulation used in assembly two.
The investigators recommended the use of two layers of plasterboard with staggered
board joints to eliminate the direct passage of heat onto the steel studs when the joints
open out in the fire.
Klippstein (1978) carried out tests on ten wall panels exposed to ASTM E119 fire.
The first seven of these tests were sponsored by American Iron and Steel Institute
(AISI) and conducted at Underwriters’ Laboratory (UL). The other tests were
sponsored by U. S. Steel Corporation (USSC). The objective of these tests was to
empirically determine the variation of the stud temperature and the lateral deflection
of the stud during the test up to the failure of the wall, which would serve as inputs in
predicting the structural behaviour of the studs when exposed to ASTM E119 or
similar fires.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 14
All panels consisted of C–shaped steel studs of varying thickness and dimensions,
spaced at 600 mm centres. One to three layers of (12.7 mm or 15.9 mm) gypsum
boards were attached on the fire side. One gypsum board was attached to the cold side
of the panels. Out of the ten wall panels, four wall panels had fibreglass insulation
placed between studs and claddings as cavity insulation. The average load per stud
ranged from 15.12 kN to 44.7 kN.
The steel studs closer to the wall ends were seen to be at lower temperatures than the
central ones, possibly due to the flow of cold air from outside into the furnace
chamber caused by a negative pressure inside the furnace. The central studs being at a
higher temperature than the studs at the wall ends, expanded more and consequently
attracted more load during the initial phase of the fire test. In the later phase of the test
the load was redistributed to the studs farther away from the central ones and the
failure times of the wall panels varied from 37 minutes to 127 minutes, with the
higher failure times generally seen associated with greater number of gypsum boards
on the fire side and lower wall loads.
Kwon et al. (1998) carried out four fire tests on two types of load bearing exterior
walls (Wall-1 and Wall-3) at the fire Insurers Laboratories of Korea (FILK). The wall
specimens were 3 m long and 2.4 m high. They consisted of C-shaped lipped steel
studs 140 X 40 X 1 mm spaced at 450 mm on centres. Rockwool insulation was
placed between the studs and the claddings as cavity insulation. Another 10 mm thick
unspecified insulation material was used as the base lining layer on the exterior side
of the wall specimens. Two layers of 12.5 mm thick type X gypsum plasterboards
made in Korea were attached on both sides of the framing for the Wall-1 specimens.
Out of the two Wall-1 specimens, one specimen was exposed to fire from outside and
the other specimen was exposed to fire from inside. The exhibited structural failure
times were 28 minutes and 40 minutes, respectively.
Wall-3 specimens were lined on either side with one layer of 15.9 mm thick type X
Gypsum boards made in Canada. When tested for fire endurance from outside and
inside, the structural failure times observed were 25 minutes and 30 minutes,
respectively. The researchers observed that the fire resistant properties of the steel
stud walls depended mainly upon the fire resistant properties of the Gypsum boards
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 15
and to have a fire rating of one hour, at least two layers of 12.5 mm thick type X
gypsum boards would be required on either side of the framing.
SCI Publication (1993)
The publication in its section dealing with “Building Design using Cold-Formed Steel
Sections: Fire Protection” presents the fire ratings of cold-formed steel sections using
planar protection with respect to various parameters such as the number of
plasterboards, protection thickness, type of plasterboard and insulation as reproduced
in table 2.1 below.
Table 2-1: Fire Resistance of Typical Floors, Walls and Partitions Comprising Cold-Formed Steel Sections and Planar Board Protection, and Heated from One
Side Only (From SCI, 1993) Fire Resistance (hours) Form of
construction
Number
of layers
of board
Protection
thickness
(mm)
Plasterboard Fire
resistant
board†
Notes
1 12.5 - 0.5 - Floors with
ceiling
protection
2
2
12.5
15
0.5
-
1.0
1.5
+ 60 mm glass
wool mat**
-
1
1
1
12.5
12.5
15
0.5
0.5
0.5
0.5
1.0
1.0
-
+ 25 mm glass
wool mat*
-
Non-load
bearing
walls
(partitions)
(number of
layers per
face)
2
2
2
12.5
12.5
15
1.0
1.0
1.5
1.5
2.0
2.0
-
Boxed section
depth > 60 mm
-
1 12.5 - 0.5 - Load
bearing
walls
2
2
12.5
15
0.5
-
1.0
1.5
-
-
† ‘Fireline’ or ‘Firecheck’ board or similar
* Glass wool mat is required for insulation purposes for more than 30 minutes fire resistance
** For floors, the glass wool mat is only necessary for fire resistant suspended ceilings
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 16
Sultan and Lougheed (1994) performed several small scale fire resistance tests on
gypsum board clad steel stud wall assemblies using glass fibres, rock fibres and
cellulose fibres as cavity insulation. The test specimens were 914 mm in height and
914 mm in width with depth depending upon the number of layers of gypsum board
used. The small scale wall assemblies were constructed using two types of gypsum
boards (regular and Type X). Details of the test specimens are as shown in Table 2.2
and Figure 2.1
Table 2.2: Small Scale Assembly Parameters and Fire Test Results (Sultan & Lougheed, 1994)
Assembly Number
Gypsum Board Layers (Exp/Unexp)
Gypsum Board
Thickness (mm)
Gypsum Board Type
Insulation Type
Insulation Thickness
(mm)
Point Failure (min)
Average Failure (min)
S - 09 1 X 1 12.7 X None - 46 46 S – 22 1 X 1 12.7 X GF 90 46 48 S – 14 1 X 1 12.7 X RF 40 69 72 S – 15 1 X 1 12.7 X CF 90 69 71
S – 10 1 X 2 12.7 X None - 86 86 S – 23 1 X 2 12.7 X GF 90 88 93 S – 26 1 X 2 12.7 X RF 90 114 117 S – 18 1 X 2 12.7 X CF 90 134 135
S – 12 2 X 2 12.7 X None - 129 129 S – 25 2 X 2 12.7 X GF 90 139 139 S – 27 2 X 2 12.7 X RF 90 160 162 S – 21 2 X 2 12.7 X CF 90 157 163
S – 01 2 X 2 12.7 RL None - 82 84 S – 32 2 X 2 12.7 RL GF 90 74 76 S – 33 2 X 2 12.7 RL RF 90 98 101 S – 34 2 X 2 12.7 RL CF 90 102 ***
X – Type X Gypsum Board 12.7 mm thick (7.83 kg/m2)
RL – Regular lightweight gypsum board with glass fibre in gypsum core (7.35 kg/m2)
GF – Glass Fibre Insulation RF – Rock Fibre Insulation
CF – Cellulose Fibre Insulation
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 17
Fire Exposed Side Fire Exposed Side
Unexposed Side Unexposed Side
1 x1 Assembly 1x2 Assembly (a) (b)
Fire Exposed Side
Unexposed Side
(c)
Figure 2.1: Construction of Assemblies
The authors observed that compared to uninsulated wall assemblies, the cavity side of
the exposed gypsum board of insulated wall assemblies heated up more rapidly
reaching temperature levels of 7000C far earlier as compared to the temperature rise of
the exposed gypsum board in an uninsulated wall assembly. Compared to the
uninsulated assemblies, the assemblies with cavity insulation recorded much higher
temperatures on the exposed side of the cavity just after the calcination of the exposed
board.
The authors observed that, in the case of type X gypsum board, the temperature
increase was primarily due to the burning of combustible material used in the
insulation, whereas with regular gypsum boards the temperatures on the exposed side
of the cavity were comparable to the furnace temperatures implying a rapid and
extensive failure of the gypsum board. The advantage gained in the use of cavity
insulations was that, the board on the ambient side remained at a much lower
temperature for a longer time as compared to the board in the uninsulated wall
assembly.
After the failure of the gypsum board on the exposed side, the cavity insulation helped
in providing some initial protection against fire to the gypsum board on the ambient
side. This protection offered was around 5 – 10 minutes with glass fibres, 10 – 15
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 18
minutes with rock fibres and 25 - 30 minutes with cellulose fibre insulation. The
increase in temperature of the unexposed gypsum after the initial protection period
was found to be most rapid in case of assembly with glass fibre insulation in the
cavity. The temperature in the cavity was seen to exceed even those measured in the
uninsulated assembly, thus giving a neutral effect on the fire resistance of assemblies
constructed with type X gypsum board. For regular gypsum board assemblies, this
increased temperature led to an earlier failure of the boards, thus in fact, lowering the
fire resistance rating of the assembly below that of the uninsulated assembly.
The authors remarked that the Rock and Cellulose fibre cavity insulations, gave
approximately a 30 minute improved fire resistance when compared with uninsulated
wall assemblies.
Gerlich (1995) conducted tests on LSF load bearing walls lined with Gypsum plaster
boards at elevated temperatures at the fire testing laboratory of BTL (Building
Technology Limited) pertaining to BRANZ (Building Research Association of New
Zealand). Gerlich carried out three full-scale furnace tests on specimens as detailed in
Table 2.3
Table 2.3: Full Scale Fire Test Specimens (Used by Gerlich, 1995)
Fire Test Number FR2020 FR2028 FR2031 Wall Height (mm) 2850 3600 3600 Steel Grade (MPa) 300 450 450 Framing Type C-section Lipped C-section Lipped C-section Stud size (mm) 76 X 32 X 1.15 102 X 51 X 1.0 102 X 51 X 1.0 Stud spacing (mm) 600 600 600 Nog spacing One row central One row central One row central Frame connections Welded Welded Welded No. of load-bearing studs 4 4 4 Load (kN/stud) 6 16 12 Lining exposed (mm) 16.0 12.5 12.5 Lining unexposed (mm) 16.0 12.5 9.5 Fire curve ISO 834 ISO 834 ‘real’
The three wall frames were actually made of six cold-formed studs welded to the top
and bottom channels. The flanges of the top channel were cut in the end bays to
minimise load transfer to the cooler edge studs, and thus have only central 4 effective
load bearing studs. The steel frames were lined on both the faces by a single layer of
glass-fibre reinforced gypsum plasterboard. The sheets were fixed vertically to all the
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 19
studs with long self-drilling drywall screws spaced at 300 mm c/c. The vertical joints
formed over studs were tape reinforced and plaster stopped. The sheet length covered
the full height so that horizontal joints were not required.
The two wall specimens named FR 2020 and FR 2028 were exposed to the standard
ISO 834 time-temperature curve whereas the third specimen FR2031 was exposed to
a much severe time-temperature curve to simulate a real fire condition. Gerlich
observed that in the actual fire test, the furnace temperatures were considerably
different from those required as he experienced some difficulties in driving the
furnace.
Tests were carried out in accordance with AS 1530: Part 4 (SA, 1990). They were
well instrumented giving detailed information regarding temperatures and deflections.
In all the tests vertical thermal expansion of the steel-framing members was allowed
to take place freely, so that additional axial loads would not build up when the studs
expand. Vertical displacement of the loading platen indicated the thermal expansion
of the studs.
Horizontal displacements in all the tests were observed to be towards the furnace. No
evidence was found of significant double curvature along the length of the studs in
any of the test specimens implying that thermal deformations override any rotational
end restraint, justifying the assumption of hinges at the stud ends in a typical fire test
set up.
The failure of all the tested walls was governed by the structural collapse of load
bearing studs through buckling of the compression flange on the ambient side of the
wall assembly near midspan. The reversal of the vertical displacement of the loading
platen indicated the failure of the test specimens. Lateral buckling about the minor
axis and flexural torsional buckling was prevented by the lateral support provided by
the unexposed linings in tests FR2020 and FR2028. In test FR2031 torsional buckling
was noticed as the thinner unexposed lining at elevated temperature significantly
degraded and could not provide sufficient lateral support and prevent this buckling
mode. The structural failure times were observed at 72, 44 and 32 minutes for the
specimens FR2020, FR2028 and FR2031, respectively. In all the experiments it was
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 20
the structural failure, which led to integrity failure. All the wall specimens satisfied
the insulation criterion right up to structural failure.
It was observed that walls with low levels of axial load may perform better in fire
tests than in actual fire situation because frictional restraints (between specimen and
specimen holder) and redistribution of load can enhance the test results.
In the work done by Gerlich (1995)
1) Information regarding furnace pressures is not documented adequately.
2) Actual furnace temp-time control was not satisfactory.
3) Vertical thermal expansions of the studs were allowed at elevated
temperatures to occur freely, which is unlikely in real life conditions.
4) It would have been better to have horizontal joints instead of vertical joints of
the plasterboard on the exposed face, because when the boards shrink due to
loss of water the joints will open out and expose the steel stud underneath over
the entire length to direct fire.
5) Since too many parameters affecting the performance of the wall have been
altered between the specimens it is difficult to draw any correlation between
the samples tested.
Kodur et al. (1999) conducted three fire resistance tests on load bearing LSF walls at
FRM of IRC/NRC. They studied the behaviour of the wall assemblies (designated as
W1, W2 and W3) exposed to standard ISO 834 fire conditions and having Glass fibre,
Rock fibre and cellulose as cavity insulation material, respectively.
The wall assemblies tested were 3048 mm high, 3658 mm long and 157 mm deep
with each assembly consisting of a single row of galvanized cold-formed steel studs
(minimum yield strength 230 MPa) protected with two layers of fire resistant gypsum
boards on each side. The studs used were lipped C – sections 92.1 mm deep by 41.3
mm wide with 12.7 mm stiffening lips. They were spaced at 406 mm, 610 mm and
406 mm in the three specimens W1, W2 and W3, respectively.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 21
Figure 2.2: Typical Steel Frame Fabrication Layout for Wall Specimens (Kodur et al.’s (1999) Work Reproduced By Alfawakhiri, 2001)
Figure 2.3: Location of Temperature Measurements and Simulation Boundaries (Kodur et al.’s (1999) Work Reproduced By Alfawakhiri, 2001)
The lateral stability of each wall specimen in its plane was provided by top and
bottom channel tracks, diagonal cross bracings, two rows of bridging and solid
blocking in four locations as shown in Figure 2.2. Nine resilient channels spaced at
406 mm c/c were attached perpendicular to the studs on the fire exposed side of all the
frames.
The gypsum boards on the fire exposed side were oriented horizontally and attached
to the resilient channels (giving horizontal joints). On the unexposed side, the boards
were oriented vertically and attached to the steel studs (giving vertical joints). All the
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 22
gypsum board joints were taped and covered with two applications of joint compound.
The testing of the wall specimens was carried out in a propane fired wall furnace of
NRC.
Total vertical load applied to the specimens W1 and W3 was 73.3 kN and to the
specimen W2 was 46.8 kN. This load was achieved by incremental loading and then it
was held constant for 45 minutes before furnace ignition and then throughout the fire
endurance test. The number of load bearing studs was 10, 07 and 10 in the assemblies
W1, W2 and W3, respectively. The test set up was very well instrumented giving
detailed information pertaining to measured temperatures and deflections. A slight
negative furnace pressure around –30 Pa at the bottom probe and neutral at the top
probe was maintained throughout the tests.
In all the tests, it was observed that the structural failure was rather abrupt resulting in
the overall out of plane buckling of the walls in the direction away from the furnace
with compression failure (local buckling) of the hot flange. In all the tests the lateral
deflections were initially positive (towards the furnace) due to thermal bowing effects.
After reaching the critical temperature the lateral deflections for the lower ¼ portion
of the wall showed reversal of displacement (i.e. away from the furnace) due to loss of
stiffness of the hot flange leading to local compression buckling near the first web
perforation in the web of the stud at 0.2 H level. Average hot flange temperatures just
prior to structural failure were about 550oC, 800oC and 650oC for specimens W1, W2
and W3, respectively. Lateral deflections of the end studs though not measured in any
of the tests were observed to be less than the central studs. This was because the end
studs were heated much slower and experienced smaller temperature gradients than
the central studs. Vertical displacements of the loading beam indicated thermal
expansion of the studs. Gradual downward movement was observed just prior to
structural failure.
The structural failure in all the tests occurred before heat penetration failure of the
unexposed side could take place. Gypsum boards on the fire-exposed side of the
specimen were observed to fall off several minutes prior to structural failure. Visual
inspection of cavity insulation material after the tests revealed that glass fibre
insulation had undergone only limited damage in certain areas, rock fibre insulation
was in good condition, but cellulose insulation was totally burnt out. Structural failure
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 23
before integrity failure in all the tests indicated that much higher fire resistance ratings
was possible for similar non load-bearing LSF assemblies.
The time-temperature curves indicated that the two layers of board on the exposed
face provided about 40 minutes of delay in the temperature rise of the hot flange. This
duration was insensitive to type of insulation used in the stud walls. The temperature
rise in steel studs after this initial period was the fastest in wall W1 with glass fibre
insulation and the slowest in wall W3 with cellulose fibre insulation. This Kodur et al.
(1999) observed was probably due to the lower bulk density and associated lower
thermal capacity of the glass fibre insulation and the higher bulk density of the
cellulose fibre insulation.
The structural failure times of wall specimens W1, W2 and W3 were observed to be
55, 73 and 70 minutes, respectively. Kodur et al (1999) also observed that the higher
fire resistance of wall W2 as compared to W3 was probably due to greater stud
spacing and therefore lower total load on specimen W2 as compared to specimen W3.
The structural failure of all specimens clearly highlighted the significant role played
by loading in determining the failure times at elevated temperatures.
Following points could be noted in the work of Kodur et al. (1999)
1) Failure of studs in all the test specimens was due to the local compression
failure of the hot flange as opposed to the compressive failure of the cold
flange near stud mid height with the studs buckling towards the furnace as
observed by Gerlich (1995).
2) The negative pressure within the furnace, sucked in cool outside air
leading to a drop in temperature of the edge studs as compared to the
central ones leading to unequal thermal expansions of the studs and thus
contributed to the building up of internal stresses within the framework.
3) Vertical thermal expansions of the studs were allowed.
4) Magnification of the thermal bowing deflections due to axial compression
was ignored.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 24
Alfawakhiri (2001) conducted three fire resistance tests on load bearing LSF walls
with plasterboard linings to study the effects of cavity insulation, stud spacing and
effect of resilient channels on failure times. His three specimens were based entirely
upon the previous work carried out by Kodur et al (1999) on specimens W1, W2 and
W3. Alfawakhiri named his three specimens as W4, W5 and W6.
He suggested:
1) Test W4 as a duplicate of test W1, but without cavity insulation to
establish effect of cavity insulation on the fire resistance of load bearing
LSF walls.
2) Test W5 as a duplicate of test W2, but with 406 mm c/c stud spacing to
establish the effect of stud spacing on the fire resistance of load bearing
LSF walls.
3) Test W6 as a duplicate of test W4, but without resilient channels to note its
effect on the fire resistance of load bearing LSF walls.
Tables 2.4 and 2.5 help in comparing the test set-ups and results of both Kodur et al.
(1999) and Alfawakhiri (2001).
Table 2.4: Summary of Fire Resistance Tests on Load-Bearing LSF Walls by Kodur et al. (1999)
Fall of Time of Gypsum board
*
(min)
Temperature rise +
(oC)
Sp. No.
Stud Spacing
(mm)
Insulation Type
(Fibre)
Resilient Channels*
Load **
(kN/m)
Face layer
Base layer
Structural
Failure Time
(Min) Max Ave.
W1 406 Glass Yes 21.5 50 In place
55 52 36
W2 610 Rock Yes 14.3 57 67 73 50 42
W3 406 Cellulose Yes 21.5 57 In place
70 42 37
* On exposed side ** Including Self Weight + On unexposed side, under pads at failure time.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 25
Table 2.5 Summary of Fire Resistance Tests on Load-Bearing LSF Walls by Alfawakhiri (2001)
Fall of Time of Gypsum board on
Exposed side (min)
Temperature rise on unexposed
side, under pads at failure time
(oC)
Sp. No.
Stud Spacing
(mm)
Insulation Type
(Fibre)
Resilient Channels
on exposed
side
Load Including
Self-weight
(KN/m)
Face layer
Base layer
Structural Failure Time
(Min)
Maximu Avera
W4 406 - Yes 21.5 58 In place
76 64 60
W5 406 Rock Yes 21.5 53 In place
59 37 26
W6 406 - No 21.5 In plac
In place
83 76 69
Alfawakhiri (2001) used the same method and material as Kodur et al (1999) to
construct his wall specimens. The test set-up and instrumentation of specimens W4,
W5 and W6 was done in a manner similar to specimens of Kodur et al. so that
comparisons could be drawn accurately.
He observed that
1) All the three wall specimens exhibited structural failure before any significant heat
penetration to the unexposed side could occur indicating the possibility of much
higher fire ratings for non-load bearing LSF wall assemblies.
2) The delay of approximately 40 min in the temperature rise of the hot flanges of the
studs in all the tests due to the protection offered by dual layers of fire resistant
gypsum boards indicated that the delay could be regarded as a stable property of the
boards as it remained unaffected by the parameters of the test series.
3) The comparison of the tests suggested that the insulation placed in the wall cavity
reduced the fire resistance of load bearing LSF walls. The insulation restricted the
passage of heat through the cavity causing an accelerated rise in the temperature of
the hot flange and a delayed temperature rise in the cold flange on the ambient side.
Hence in insulated walls there was a high temperature gradient across the steel section
as compared to a very low temperature gradient (almost uniform) across the steel
cross section of uninsulated walls.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 26
4) Structural buckling of the uninsulated wall frames was due to the overall buckling
of the studs towards the furnace, finally leading to local compression failure of the
cold flange near the mid height. In case of insulated walls the lateral deflections were
seen to be initially positive i.e. towards the furnace followed by an early reversal in
the direction of deflection at a height of 0.25H from base leading to structural failure
by local compression failure of the hot flange near that level as shown in Figure 2.4
He concluded that
1) A comparison of W2 and W5 showed that wider stud spacing improved the fire
resistance of load bearing LSF walls in standard tests.
2) A comparison of W4 and W6 showed that use of resilient channels reduced the
fire resistance of load bearing LSF walls because it reduced the ability of fire exposed
gypsum boards to remain in place.
Figure 2.4: Structural Failure Modes: (a) Uninsulated Walls (b) Insulated Walls (Alfawakhiri, 2001)
Following points could be noted in the work of Alfawakhiri (2001)
1) Test W5 was intended to be a duplicate of Test W2 with the objective of
testing the effect of stud spacing on fire resistance of LSF wall assemblies. To
achieve this Alfawakiri changed the stud spacing from 610 mm c/c (in W2) to
406 mm c/c in specimen W5. He also changed the loading from 14.3 kN/m (in
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 27
W2) to 21.5 kN/m (in W5), which actually is detrimental for the fire resistant
property of the wall. Hence the early failure of wall W5 as compared to wall
W2 cannot be entirely attributed to the decreased spacing of the studs.
2) Vertical thermal expansions of the studs were allowed.
3) The end studs remained significantly cooler than the central ones and
exhibited smaller thermal bowing than the studs in the central part developing
undesirable internal stresses within the framework.
Feng et al. (2005) conducted eight tests on loaded full-scale steel stud walls, two at
ambient temperature and six exposed to the standard fire condition on one side. The
tests were carried out in the UMIST fire-testing laboratory. Each wall panel was of
size 2200 X 2000 mm. Studs used were lipped channel sections (either 100 X 54 X 15
X 1.2 mm or 100 X 56 X 15 X 2 mm ) with a minimum yield strength of 350 MPa
spaced at 750 mm centres. Two elongated service holes were provided at the centre of
the webs, one at 300 mm from the top and the other at 300 mm from the bottom.
Unlipped C - sections of size 100 X 56 X 2 mm were used as top and bottom tracks.
Lateral bracing was provided by 4 flat bars fixed horizontally, two on either side of
the framing, one placed at 650 mm from the top and the other at 650 mm from the
bottom. The frame assembly was lined with one layer of 12.5 mm thick Fireline
Gyprock board on either side. Isowool 1000 was used as cavity insulation. The
Gyprock plasterboards were fixed horizontally to the steel studs by screws at 300 mm
centres. Three different load levels, being 0.2, 0.4 and 0.6 times the load carrying
capacity of the same panel tested at ambient temperature, were applied during the six
fire tests on the three panels using each type of lipped channels respectively. The
instrumentation consisted of 6 control thermocouples, placed inside the furnace to
regulate the average furnace temperature in accordance with BS 476 fire time-
temperature relationship, 30 thermocouples placed within the wall specimen to obtain
the time-temperature history of the assembly, 9 displacement transducers to record the
vertical in plane movement of the panel and nine displacement transducers placed at
three locations along each of the three studs to measure the out of plane movements of
the test panel.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 28
Load was applied using three hydraulic jacks, located directly above the three lipped
channels of the test panel. The load was transmitted onto the panel through a moving
steel beam of the reaction frame. The load was kept unchanged during the fire test and
the specimen was considered to have failed when the applied loads could not be
maintained.
The investigators concluded that the failure mode was generally observed to be due to
global buckling of the studs about the major axis, with the flexural buckling of the
steel channels being restrained by the unexposed gypsum board. The investigators
recommended the use of non-combustible cavity insulation as it was observed that in
some of the specimens that the panel failure took place shortly after the burning out of
the insulation material. The investigators also observed that the fall of the gypsum
boards on the fire side was an effect and not a cause of the panel structural failure. It
was observed that under the same load ratio, panels using thinner channels had lower
fire ratings than panels using thicker channels. At higher load ratios (0.4 and 0.7)
panels using the thinner (1.2 mm) channels did not achieve even a 30-minute standard
fire resistance.
Sultan (2010) conducted 41 full-scale wall fire resistance tests at the National
Research Council of Canada, in accordance with ULC-S101 standard fire exposure, to
determine the gypsum board fall temperatures from the wall panels. The tests used
assemblies with wood and steel studs protected with either one or two layers of Type
X gypsum board and with or without insulation in the wall cavity. The temperature
criterion recorded for the fall off of the plasterboards was based on the sudden
temperature rise measured on the back side of the fire exposed gypsum board caused
by its failure. The parameters studied included resilient channels, spaced either 406
mm o.c. or 610 mm o.c. installed between the gypsum board and framing for sound
reduction purposes. The insulations used were glass and rock fibre batts and
cellulosic fibre insulation either spayed wet or dry blown in the wall cavity.
Plasterboards used were of Type X gypsum board 12.7 mm or 15.9 mm thick. The fall
off temperatures for assemblies with a single layer of gypsum board, with and without
insulation in wall cavity, and with different screw spacing was observed to be in the
range of 7550C to 7850C. The fall off temperatures for both single and two layers of
gypsum board was observed to have very little difference.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 29
2.2: Analytical Research
Gerlich (1995) proposed a temperature sensitive structural analysis spreadsheet in
accordance with the AISI design manual (AISI, 1991) with strength and stiffness
expressions suitably modified to take into account the temperature effects. He
assumed the studs were free to rotate and expand at both ends. The thermal
deformations induced due to the temperature gradient across the steel stud (Thermal
bowing) was determined using the following expression (Cooke, 1987)
1 = L2T / 8D (2.1)
Where,
1 = Horizontal deformation at midspan due to thermal bowing
= Expansion coefficient for steel.
L = Length of member
T = Temperature difference across the member
D = Member depth
Mean stud temperature was used to calculate the expansion coefficient of steel. The
expression gave reasonably good predictions upto temperatures less than 400oC. At
higher temperatures of steel, the temperature difference across the steel would reduce
giving lower calculated deflections. Actual deflections would not return to the
calculated values due to plastic deformation of steel. Gerlich (1995) conservatively
assumed the calculated deformations (1) to remain constant when temperature
gradients (T) decreased. It was achieved by not allowing the value of calculated 1 to
be less than the one calculated for the previous step. The total horizontal deflection for
the system was calculated by adding the thermal deflection (1) and the deflection
(2) due to P- effects.
To obtain the deflection due to P- effects, the deflection 1 (Stress free thermal
deformation) was treated as the initial eccentricity. The initial bending moment P1
then gave an additional horizontal deflection 2
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 30
The P- component was predicted analytically by solving the moment equilibrium
equation
ET Ix d2 2/dZ2 = Pa (1 + 2) (2.2)
Figure 2.5: Total Horizontal Deflection for Load-bearing Systems
(Gerlich et al., 1996)
The solution of the above equation yielded deflection 2 at mid-height as
2 = 1 {[1/cos ( L/2)] -1} (Gerlich et al. 1996) (2.3)
Where,
XTa IEP / (mm-1)
With ET = Modulus of elasticity of steel in MPa at average stud temperature.
Ix = Second moment of area of the stud cross-section in mm4
Pa = Applied axial stud load in N
1 = Initial eccentricity at stud ends in mm (Taken equal to thermal bowing
deflection at midspan)
2 = P- deflection in mm
L = member length (wall height) in mm
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 31
The total horizontal deflection () was expressed as = 1 + 2. The studs were then
analysed as compression members subjected to axial load P and bending moment P.
A critical temperature was determined such that the maximum permissible stud load
was equal to the applied axial load. This temperature was then compared with the
compression flange temperature on the ambient side of the wall assembly to find the
failure time.
To model the heat transfer and steel framing temperatures, Gerlich (1995) used a
commercially available heat transfer model TASEF (Sterner and Wickstorm, 1990).
TASEF was found to give good correlations with measured values on the exposed
face of the wall. Discrepancy was noted on the ambient side of the wall, with TASEF
values lower (unconservative) than the actual measured values. This was attributed to
the inability of TASEF to model mass transfer and consider ablation or degradation
(with opening of joints) of the exposed lining allowing sudden rise in measured
temperature on the ambient face.
It was observed from specimen FR2031 that the temperature predictions of TASEF
for fires significantly hotter than ISO 834 were unconservative giving low values. As
TASEF under-predicted the lining and framing temperatures on the ambient side, it
gave a greater temperature difference across the stud cross section leading to a higher
calculated horizontal deflection due to thermal bowing effects (Since 1 is
proportional to T). Gerlich (1995) observed that the use of TASEF gave greater steel
stresses (due to higher predictions of deflection) and thus conservative failure time
predictions within 80 to 90% of test results.
Gerlich (1995) used the TASEF and temperature sensitive structural analysis
spreadsheet model to generate data and correlate the failure times of LSF wall
specimens with the load ratios and thickness of plasterboard linings when exposed to
ISO 834 fire conditions. The load ratio versus failure time curves drawn for various
board thicknesses were found to give 62% to 73% accuracy when compared with
measured values.
Milke (1999) observed that fire resistance of structural assemblies depends on fire
exposure conditions, material properties at elevated temperatures, thermal response of
the structure and structural response of the heated assembly. Heat transmission limits
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 32
were established to prevent the ignition of combustibles in contact with the unexposed
side of the assembly (Schwartz and Lie, 1985).
Alfawakhiri (2001) used a computer program TRACE (Temperature Rise Across
Construction Elements) to model heat transfer through LSF walls exposed to fire. This
program developed by him was based on an explicit finite difference integration
algorithm (Sultan 1991) to solve one dimensional transient heat transfer equations.
The presence of the steel frame was neglected in the heat transfer simulations.
A large number of numerical trials was conducted on uninsulated wall specimens (W4
and W6) and the properties of gypsum board were calculated by matching the
measured and simulated temperature histories at all the simulated boundaries of the
wall specimens. Further numerical trials were conducted on insulated walls (W1, W3
and W5) using the calibrated gypsum board properties so as to obtain the properties of
insulation materials (Glass fibre batts, loose fill cellulose and Rock fibre batts). The
apparent thermal properties so calculated, also accounted to some extent the physical
phenomena other than heat transfer, such as mass transfer, phase change, etc. This
Alfawakhiri (2001) observed, was due to the fact that the temperature rise in LSF
walls exposed to fire was affected by processes not described by heat transfer, such as
migration of moisture vapours within the board, penetration of cool ambient air or hot
furnace gases into the cavity etc.
The other parameter affecting the simulated temperature histories was the fall off of
gypsum boards. The TRACE model used by Alfawakhiri models the spalling of
gypsum boards by removing it from the simulation at a user specified time. It was
observed that the simulated temperature histories were not very sensitive to the choice
of emissivity coefficients and even less sensitive to the choice of convection
coefficients.
In formulating the structural model Alfawakhiri (2001) made the same basic
assumptions as suggested by Klippstein (1978), except for the assumption of eccentric
loading as shown in Figures 2.6 and 2.7, which in part models rotational end
restraints.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 33
According to Alfawakhiri the eccentricity ‘e’, developed due to the following reasons
1) Shift of the centre of steel section towards the cold flange due to the
deterioration of the stiffness of the hot flange with the increase in the
temperature gradient across the stud.
2) Shift of the load towards the hot flange due to rotation of the end studs
associated with thermal bowing.
3) Internal Stresses caused due to non-linear thermal strain gradients.
Figure 2.6: Thermal Bowing Figure 2.7: Stud End Conditions And Secondary Deflection (a) Uniform Heating, (b) Non-uniform Heating (Alfawakhiri, 2001) (Alfawakhiri, 2001)
Alfawakhiri (2001) gave the expression for eccentricity ‘e’
as e = (1-KR) Øβ-2 (2.4)
Where Ø = Thermal bowing curvature = αTδT/D and β2 = P/EI*
KR is a reduction coefficient
αT = Thermal expansion coefficient for steel (Lie 1992)
= (12 + 0.004 TA)10-6 (2.5)
D = Stud section depth
δT = TH - TC = Temperature difference across stud section in 0C
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 34
TA = 0.5 [ TH + TC] = average stud temperature in 0C
P = Vertical load applied at stud ends
E = 203000 MPa = Modulus of Elasticity of steel at room temperature
I* = Elastic modulus- weighted moment of Inertia of the unreduced stud section
about the neutral axis parallel to flanges.
Alfawakhiri (2001) gave the shape of the stress free initial imperfection y1(Z) caused
by thermal bowing as y1(Z) = 0.5 ØZ(H -Z)
The secondary lateral deflection y2(Z) caused by the vertical load ‘P’ with eccentricity
‘e’ was obtained by solving the differential equation
EI* d2y2(Z)/dZ2 = P[y1(Z) + y2(Z) –e] (2.6)
Total lateral deflection ‘y(Z)’ was expressed as
Y(Z) = y1(Z) + y2(Z) (2.7)
= (Øβ-2 - e) [tan(0.5βH)Sin(βZ) + Cos(βZ) – 1]
Substitution of Eq. 2.3 in Eq. 2.4 gave the expression for the lateral deflection ‘Δ’ at
the mid-height of the stud,
Δ = KR Øβ-2 {[1/cos(0.5 βH)] -1} (2.8)
Due to the temperature variation from the hot flange to the cold flange, the modulus
of elasticity varies across the steel stud section. Alfawakhiri used the reduction
coefficient ‘nT’ as expressed by Gerlich (1995) based upon Klippstein’s (1978, 1980)
work.
nT = ET/E
= 1.0 - 3T(10-4) + 3.7T2(10-7) – 6.1T3(10-9) + 5.4T4(10-12) (2.9)
The variation of ‘E’ across the flange was accounted by the ‘modulus-weighted’
moment of inertia ‘I*’
I* was quantified numerically by dividing the stud section into sufficiently large
number ‘q’ of two-dimensional elements, so that
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 35
q
I* = ∑ ni [Ii + Ai (xi – c)2] (2.10)
i = 1
Where
ni = Reduction factor for temperature Ti, calculated using expression for nT
from Eq. 2.9
Ii = Moment of inertia of element ‘i’ about its own neutral axis parallel to flanges,
Ai = Area of element ‘i’.
xi = Distance of element ‘i’ from the extreme fibre of the cold flange.
Ti = Temperature of element ‘i’ calculated from
Ti = Tc + (δT xi / D) (2.11)
And c= ∑ ni Ai xi / ∑ ni Ai
(2.12)
q q
i = 1 i = 1
Alfawakhiri (2001) incorporated the above expressions in the computer program
STUD that had been developed to model the structural behaviour of load bearing LSF
walls in fire. The program assumed the calculated deflections to remain constant
whenever δT decreased in time. In STUD simulations, the values of Δ and Y(Z) at any
time were not allowed to be less than their values in the previous steps to account for
the creep and stress relaxation in steel studs at temperatures higher than 400oC. The
simulated mid-height lateral deflections were found to be in good agreement with the
measured values for all the specimens when KR was taken equal to 0.6
In case of insulated walls it was seen that owing to a large temperature difference
across the studs cross-section, a large eccentricity ‘e’ would develop at the stud ends
causing negative secondary deflections y2(Z).
The numerical simulations suggested that stress free thermal bowing gets restrained
(reduced), due to restraints at end studs and by nonlinearity of thermal strain gradients
across the stud cross sections. The compressive stresses in the hot flanges were seen
to reach critical levels when δT values exceeded 350 oC leading to compressive failure
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 36
of hot flange. STUD checked for this failure mode at the perforated section (Z= 0.2H)
using the expression
ƒH = nH [ (P/A*e ) + P(e-y0.2H)/S*eH ] ≥ FyH (2.13)
where
ƒH = Compressive stress at the extreme fibre of the hot flange
nH = Reduction factor for the temperature TH, calculated using the Eq. 2.9
FyH = Yield strength of steel at temperature TH
A*e = Elastic modulus weighted effective stud section area in compression,
S*eH = Elastic modulus weighted effective stud section modulus in bending that
causes compression of hot flange.
The stud cross-section calculations were done in accordance with S136-94. The
effective cross-sectional dimensions were assumed to be insensitive to temperature
and were based on steel properties at room temperature and compressive stress ƒ = Fy.
The effective cross-sections were then used in the calculation of the temperature
dependent ‘modulus weighted’ properties A*e and S*eH as per Equations given below
A*e = ∑ ni Ai
(2.14)
q
i = 1
S*eH = ∑ ni [Ii + Ai (xi – c)2] / (D-c) (2.15)
q
i = 1
For uninsulated walls (W4 and W6), the section at mid-height was checked by STUD
for the compressive failure of cold flange at perforated section (Z= 0.4H) using the
expression
ƒc = nc [ (P/A*e ) + (P y0.4H)/S*ec ] ≥ Fyc (2.16)
where,
ƒc = Compressive stress at the extreme fibre of the cold flange
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 37
nc = Reduction factor for the temperature Tc, calculated using Eq. 2.9
Fyc = Yield strength of steel at temperature Tc
S*ec = Elastic modulus weighted effective stud section modulus in bending that causes
compression of cold flange calculated from
S*ec = ∑ ni [Ii + Ai (xi – c)2]/ c (2.17)
q
i = 1
The section properties A*e, S*eH and S*ec were based on three different effective cross-
sections, as the configuration of compression elements was different in each case.
Also the value of ‘e’ was taken as zero and was not used in the expression (2.16). This
was because as δT decreases in the final stages of tests (on uninsulated walls) the
heating rate in cold flange becomes higher than in hot flange. This combined with
creep and stress relaxation in steel at temperatures greater than 400oC causes gradual
reduction of eccentricity leading to an assumed value of zero near failure time in an
non-insulated wall specimen.
The STUD program carried out structural failure checks considering both failure
modes at every time step. It was seen that for uninsulated walls (W4 and W6) the
predictions of failure times from STUD showed reasonable agreement with test
structural failure times. For insulated walls (W1, W3 and W5) predicted failure times
agreed well with the initiation of the structural failure in central studs. This was
because of the quasi-elastic approach used in the STUD model formulation and the
effect of load redistribution to colder studs at wall ends in the final phases of the tests.
Predictions based on measured temperatures showed a better agreement with test
results than predictions based on simulated temperatures. The STUD simulations
showed that the major part of deflections observed in the fire resistance tests was due
to thermal bowing and not due to deterioration of steel stiffness.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 38
2.3: Mechanical and Thermo Physical Properties of Steel Stud Wall Assembly
Components at Elevated Temperatures.
The fire resistance behaviour of load bearing, gypsum board protected, stud wall
assemblies can be modelled only with a thorough understanding of the mechanical
and thermo-physical response of individual components of the assembly at elevated
temperatures.
2.3.1 Gypsum Plasterboards
2.3.1.1: Introduction
Gypsum wallboards have been in use since the early 1900’s. They are widely used as
wall or ceiling linings in domestic housing or commercial buildings. The core of these
wallboards or plasterboards is made up of Gypsum i.e. calcium sulphate dihydrate
(CaSO4.2H2O), a naturally occurring non-combustible mineral. The core is
sandwiched between two layers of paper (see Figure 2.8), which are chemically and
mechanically bonded to the core to form flat sheets available in a range of sizes. The
papers provide sufficient tensile strength to the board to assist in handling and
transportation. Gypsum plasterboards have become very popular due to their non-
combustible core and fire resisting properties. Most gypsum boards are made with a
thickness between 10 and 20 mm.
Paper covers Gypsum Core
Figure 2.8: Gypsum Plasterboard
2.3.1.2: Types of Gypsum Plasterboards
The plasterboards are generally available in three main varieties i.e. Regular, Type X
and Special Purpose Boards.
1) Regular Plasterboards generally do not have any fire resistance rating and are made
up of low density Gypsum core without use of reinforcing fibres. They are mostly
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 39
used in constructing non-load-bearing walls. When exposed to fire the regular boards
tend to crack up and fall off soon after the burning up of paper facings, which takes
place at around 300oC.
2) Type X Board is a generic term that describes a Gypsum board with a specially
formulated core to provide a greater fire resistance than a regular board of the same
thickness. All type X boards contain some additives such as Vermiculite and Glass
fibre reinforcing to enhance the fire resisting properties. Vermiculite expands when
exposed to heat and thus partly helps in compensating the shrinkage of the Gypsum
core during calcination (i.e. dehydration). Glass fibres improve the mechanical
properties of the board, reduce shrinkage and ablation and thus enhance the stability
and integrity of the board, when exposed to fire.
3) Special Purpose Boards (some called as Type C Boards) are proprietary products
made by manufacturers to obtain superior fire or structural performance over Regular
or Type X boards. For the Gypsum boards to stay in place, in the stud wall assembly
they should possess sufficient tensile ductility to accommodate the thermal strain
incompatibility with steel studs i.e. as the boards tend to shrink and the studs tend to
elongate with rise in temperature. Special additives are used in proprietary
formulations to reduce shrinkage and enhance strength and ductility characteristics of
the Gypsum core.
2.3.1.3 Chemical Properties of Gypsum Plasterboard Linings
Gypsum contains approximately 21% by weight chemically bound water of
crystallization and about 79% calcium sulphate, which is inert below a temperature of
1200oC (Goncalves et al., 1996). In addition to water of crystallization it is found that
approximately 3% free water is also present inside Gypsum plaster, depending upon
the ambient temperature and relative humidity (Buchanan, 2001).
The fire retarding property of the gypsum board primarily stems from this water
content (Free water and water of crystallization). When the gypsum board is exposed
to fire, the free water and water of crystallization is gradually released and evaporated.
The dehydration i.e. release of water occurs in two phases. In the first phase also
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 40
known as calcination, Gypsum dehydrate loses some amount of water to yield
Gypsum hemihydrate (CaSO4.1/2H2O) commonly known as Plaster of Paris.
CaSO4.2H2O CaSO4.1/2H2O + 3/2H2O (2.18)
The above reaction (calcination) is endothermic in nature occurring between 100oC to
120oC and consumes large amounts of energy in order to evaporate the free water and
release as steam the chemically bound water of crystallization.
This absorption of energy delays the heat transmission through the board and causes a
temperature plateau on the unexposed face of the lining. The length of this plateau is a
function of the lining thickness, density and composition and is commonly referred to
as the ‘Time Delay’ (Gerlich et al., 1996).
Calcination leads to shrinkage and loss of strength of the sheet material. The progress
of calcination through the sheet thickness is retarded by the exterior layer of calcined
Gypsum on the fire exposed side which acts as a protective layer and adheres well
with the inner uncalcined layers.
The second phase of dehydration, i.e. complete dehydration occurs when the Gypsum
hemihydrate is transformed to Gypsum anhydrite.
CaSO4.1/2H2O CaSO4 + ½ H2O (2.19)
This reaction occurs at about 210oC according to Andersson and Jansson (1987) and
at about 600oC according to Sultan (1996) and at about 225oC according to Bakhtiary
et al. (2000). The temperature at which the second phase occurs much depends upon
the rate of heating. (Thomas, 2002).
2.3.1.4 Mechanical Properties
Most gypsum boards have a density between 550 and 850 kg/m3 .
Goncalves et al., (1996) reported the mechanical properties of Gypsum plasterboards
subjected to fire. The results were based on the tests conducted on plasterboards from
three Australian companies; Boral, C.S.R and Pioneer. The tensile strength
characteristics and modulus of elasticity (MOE) of plasterboard at 300oC and 500oC
are summarised in Table 2.6.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 41
Table 2.6: Mechanical Properties of Australian Manufactured Plasterboards (Goncalves et al., 1996)
Board Type Failure Stress at 3000C (MPa)
Failure Stress at 5000C (MPa)
MOE at 3000C (MPa)
MOE at 5000C (MPa)
Boral 0.121 0.098
C.S.R 0.163 0.098
Pioneer 0.107 0.117
8.5 – 18.8 6.0 – 9.9
2.3.1.5 Thermo-Physical Properties of Gypsum
Gerlich et al. (1996) have quoted Thomas et al. (1994) who summarised the data
measured by Mehaffey (1991) for the thermal conductivity and enthalpy of glass fibre
reinforced gypsum plasterboard as a function of temperature. Thomas’s values for
thermal conductivity and enthalpy are shown in Figures 2.9 and 2.10, respectively.
The enthalpy values represent the summation of the product of specific heat and
temperature, expressed per unit volume. The authors have used the enthalpy values in
modelling to avoid numerical instabilities resulting from sharp peaks that may occur
in the specific heat of materials containing water, due to evaporation of moisture.
Sultan (1996) conducted tests at NRCC to obtain the thermo-physical properties of
Type X Gypsum Board. Measurements were carried out at a heating rate of 2oC/min
as it provided the maximum specific heat at approximately 100oC. The author has
given the results in the form of equations for Specific heat, Thermal conductivity and
Density. The equations are represented in the form of graphs in Figures 2.11, 2.12 and
2.13, respectively.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 42
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Temperature [°C]
Th
erm
al c
on
du
ctiv
ity
[W/m
K]]
Figure 2.9: Thermal Conductivity of Gypsum Plasterboard (Franssen, 1999)
0
200
400
600
800
1000
1200
1400
1600
1800
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Temperature [°C]
En
thal
py
[MJ/
K]]
Figure 2.10: Specific Volumetric Enthalpy of Gypsum Plasterboard
(Franssen, 1999)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 43
0
2 000
4 000
6 000
8 000
10 000
12 000
14 000
16 000
18 000
20 000
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Temperature [°C]
Sp
ec
ific
He
at
[J/k
gK
]
Figure 2.11 Specific Heat of Type X Gypsum Board (Franssen, 1999)
The two peaks in the specific heat curve indicate the dehydration of Gypsum, which
appears at temperatures around 100 oC and 650oC. The first peak is the main cause of
delay in the temperature rise of protected steel studs in the stud wall assembly. The
area under the peak gives the energy consumed per kg of Gypsum board to drive out
the water of crystallization and free water from its core.
At this stage (i.e. around 100oC) there is a drop in density and thermal conductivity of
the board. There is a steady rise in thermal conductivity beyond 400oC. Thermal
conductivity also depends upon density variations of Gypsum board (Clancy, 1999).
Thermal conductivity values above 400oC get affected by the presence of shrinkage
cracks in the board which depend upon the type and composition of board and the
nature of fire (Buchanan, 2001).
Conductivity increases on account of radiative heat transfer caused due to the opening
of cracks in the Gypsum boards at high temperatures and also due to ablation, a
process in which thin layers of calcined gypsum due to their cohesionless nature tend
to fall off the board. Cracking is more severe in fire with greater initial temperature
gradient. According to Manzello et al. (2005), cracks will be propagated in the order
of opening at plasterboard joint, cracks at screw points along the stud and transverse
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 44
cracks. Manzello et al.’s (2006) study recommends incorporating plasterboard
contraction and crack formation in thermal models.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Temperature [°C]
Th
erm
al c
on
du
ctiv
ity
[W/m
K]]
Figure 2.12 Thermal Conductivity of Type X Gypsum Board (Franssen, 1999)
0
100
200
300
400
500
600
700
800
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Temperature [°C]
Den
sity
[kg
/m³]
]
Figure 2.13: Density Variation of Type X Gypsum Plasterboard on Heating (Franssen, 1999)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 45
Harmathy (1988) gave a value of 0.88 kJ/kg.k as the specific heat of gypsum board
at ambient temperature with a peak of 7.32 kJ/kg.k at about 1000C giving
approximately 201 kJ/kg of gypsum as the area under the peak.
Mehaffey et al. (1994) gave a base value of 0.95 kJ/kg.k for specific heat. The
authors used 2oC/min and 20oC/min as the two scanning rates in their differential
scanning calorimeter. For 2oC/min they obtained a peak of 29kJ/kg.k at 95oC and for
20oC/min they got a peak of 14kJ/kg.k at 140oC. The area under both the peaks is
about 490 kJ/kg. Mehaffey et al. measured specific heat only upto 200oC and thus did
not record the second peak (Thomas, 2002).
Andersson and Jansson (1987). The authors did not mention a base value of specific
heat but reported that at 1000C, 75% of bound water is evaporated requiring 515 kJ/kg
of gypsum for the process to occur, and the balance 25% of bound water being driven
off at 2100C requiring 185 kJ/kg of gypsum (Thomas, 2002).
Thomas (2002) studied the two consecutive dehydration reactions that gypsum
undergoes when heated and suggested modifications for the thermo-physical
properties, so as to obtain smooth curves for enthalpy and thermal conductivity
suitable for input into a finite element heat transfer model. The values were calibrated
and validated using furnace and fire test data. Experiments were conducted on
plasterboards similar to the North American type C boards. The author assumed the
base value for specific heat of gypsum as 0.95 kJ/kg.k. This value was adopted from
Mehaffey et al. (1994). The energy required to complete the two stages of dehydration
was adopted from Andersson and Jansson (1987) and was taken to be 515 kJ/kg and
185 kJ/kg of gypsum, respectively. The first is assumed to occur between 100oC and
120oC and the second between 200oC and 220oC. Figure 2.13 gives its density
variation at elevated temperatures and Figure 2.14 gives the mass loss in gypsum
plasterboard undergoing heating (Thomas, 2002)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 46
Figure 2.14: Mass Loss in Gypsum Plasterboard Undergoing Heating (Thomas,
2002)
Baux et. Al. (2008) conducted several fire tests on gypsum plasterboards using
varying proportions of silica fume as filler introduced into the plaster. They observed
that in the case of conventional gypsum plasterboards, the dehydration on exposure to
elevated temperatures leads to large thermal shrinkage due to the loss of bound water
molecules. This resulted in rapid development of cracks allowing passage of heat. To
reduce this tendency of crack formation the researchers added silica fumes as filler
material to the hemihydrates before hydration with an amount ranging from 10 to 60
wt%. From the fire tests they observed that the number of cracks on the fire exposed
face of the gypsum plasterboard decreased with the increase in the filler content with
no cracks developing at a 40 wt% filler amount. The reduction in cracking also was
observed to reduce the overall shrinkage of the plasterboard during the fire tests.
A serious drawback, however, was a decrease of the latent heat effect due to the
substitution of the plaster by silica fume. It was observed that as compared to pure
plaster, the higher the filler content, lower was the latent heat, thus reducing the heat
absorbing capacity of the plasterboard. A 30 wt% of silica fume was found to be
acceptable at temperatures below 10000C. Above this temperature reactions between
the binder and filler material led to melting and geometric instability. To overcome
this problem they are currently trialling aluminosilicate filler in place of silica fumes.
However, they have observed that higher density and thermal conductivity could be
the setback for this material.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 47
Ghazi and Hugi (2009) observed that the thermo-physical properties of Gypsum
Plasterboards are affected by the chemical composition of the ingredients (i.e. the
various carbonates in the plasterboard). Depending on different ingredients different
endothermic reactions were observed to occur between room temperature and 9000C
impacting the sensitivity of thermal conductivity, effective heat capacity and density
with respect to temperature.
Frangi et. al. (2010) carried out experimental and numerical analysis on the fire
behaviour of protective cladding made of gypsum plasterboards at ETH Zurich. 17
small-scale fire tests were performed with non-loaded specimens lined with gypsum
plasterboards and subjected to ISO fire exposure. The fire tests were carried out in the
EMPA’s horizontal small furnace with the internal dimensions of 1.0 x 0.8 m.
Gypsum plasterboards of type A and F according to EN 520 and gypsum fibreboards
according to EN 15283-2 were studied. Gypsum plasterboards of type A are similar to
regular type of plasterboards with a porous gypsum core and no reinforcement or filler
material. Type F plasterboards are similar to type X used in North America with
gypsum core reinforced with glass fibres or fillers to improve the core cohesion at
higher temperatures. Gypsum fibreboards have a gypsum core reinforced with paper
fibres and are usually denser than plasterboards of type A and X.
The researchers observed that the overall thermal behaviour of different types of
gypsum board with different density, fibres and fillers was quite similar although a
general improvement was noticed in the mechanical properties (shrinkage, cracking,
and ablation) of the boards after complete dehydration. The tests indicated that the
layer backing the gypsum board may have a strong influence on the thermal behaviour
of the gypsum board. Insulating batts were observed to cause the fire exposed gypsum
boards to heat more rapidly and fail sooner.
2.3.1.6 STRUCTURAL BEHAVIOUR
The plasterboards play an important role in providing lateral stability to the steel
studs. They provide adequate restraint against torsional buckling and flexural buckling
of the stud about the minor axis. When the assembly is exposed to fire, this ability of
the plasterboard to provide lateral restraint reduces, due to the calcination of the
gypsum board on the fire exposed side, whereas the plaster board on the ambient side
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 48
of the assembly continues to provide lateral support as it is less affected by
temperature.
Sultan (1996) reported that fall-off of plasterboard occurs when the unexposed face of
the board reaches about 600°C (Buchanan and Gerlich, 1997). The temperature at
which the gypsum boards lose their restraining capacity depends on the type of board
used. However, according to Ranby (1999) a common temperature of 550°C was
proposed. In the numerical study of Kaitila (2002), the boundary conditions providing
lateral restraints at both flanges were assumed to be valid until 600C. Thermal
properties of gypsum plasterboard are required to determine the extent to which the
plasterboards offer lateral support to the cold-formed steel studs at elevated
temperatures resisting buckling about the minor axis.
2.4 Literature Review Findings Relevant to this Research
Researchers have attempted to improve the fire ratings of wall systems by using
different types of insulations in the wall cavities. Their observations, however, were
found to be contradictory. Sultan and Lougheed (1994) observed that by use of rock
or cellulose fibre as cavity insulation the fire resistance improved by approximately
30 minutes when compared with uninsulated wall assemblies. Sultan (1995) remarked
that only rock fibre when used as cavity insulation in non-load bearing wall
assemblies gave an improvement in the fire performance whereas assemblies using
cellulose fibre actually showed reduced fire resistance. Feng et al. (2003) observed
that the fire performance of non-load bearing wall panels improved with the use of
cavity insulation.
As other researchers have not been able to conclude the effect on the fire ratings of
wall specimens using cavity insulation, it is considered necessary to conduct further
detailed experimental studies to fully understand the benefits or drawbacks of the
traditional method of wall construction using cavity insulation and to recommend new
wall models to improve the fire performance.
In the research conducted by Kodur et al. (1999) and Alfawakhiri (2001), the studs
had perforations in the web and they were found to be failing at these particular
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 49
points. All the frames in their study were cross-braced and it was assumed that the
flexural-torsional buckling and weak axis buckling failure modes were prevented by
this lateral restraint. Gerlich (1995) tested frames with a central row of nogs. Feng et
al. (2005) conducted studies using studs with service holes at the centre of the webs.
The studs were braced laterally by four flat bars fixed horizontally, two on either side
of the framing. No tests were conducted on unperforated studs or frames without
lateral bracing. Therefore further study is necessary to investigate the behaviour of
stud panels without perforations and/or bracing.
Gerlich (1995) study was limited to steel grades of 300 and 450. Kodur et al.’s (1999)
and Alfawakhiri’s (2001) studies involved the use of cold-formed steel studs of yield
strength 230 MPa. Feng et al. (2005) extended their study to 1.2 mm and 2 mm
thickness steels with the grade of S350. Hence it can be concluded that past research
on the behaviour of LSF stud wall panels at elevated temperatures was limited to
lower steel grades. Therefore further research is needed on LSF wall systems made of
higher steel grades.
Uniform temperature values were assumed for flanges and lips of lipped channel studs
on both the hot and cold sides in the studies of Kodur et al. (1999), Alfawakhiri
(2001) and Feng et al. (2005). These observations demonstrate the need to study the
true temperature profiles across and along the studs under fire conditions and to
recommend a simplified temperature profile for use in numerical and theoretical
studies of LSF wall systems.
Most of the numerical models of the past tests were not fully validated due to lack of
experimental results and the complexity of the problem. Also the previously
developed elevated temperature calculation methods are extremely complex. Hence
further study is required to recommend a simplified temperature profile for use in
numerical and theoretical studies and to develop simple models
In the study of Gerlich (1995), too many variables were incorporated in the tests
conducted on only three test specimens. More than one variable between the tests
made the comparison difficult to draw any specific inferences. Also, only one layer of
lining was used in this study. There is very limited data about the effect of multiple
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 50
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 51
plasterboards, joints in plasterboard as well as the influence of density and thickness
of cavity insulation on the fire performance of the wall assemblies. Therefore a series
of fire tests to establish the fire performance characteristics of the plasterboards, non-
load bearing wall assemblies and load bearing wall assemblies using different types of
insulations of varying density and thickness needs to be undertaken.
In most of the previous experiments it has been noted that the actual Standard time-
temperature heating profile in the furnace could not be controlled satisfactorily. Also,
the instrumentation has not been detailed enough to provide a complete understanding
of the temperature gradient across the thickness and along the height of the wall
specimen. Information regarding furnace pressures is not documented adequately. A
negative furnace pressure in most cases has led to a drop in temperature of the edge
studs as compared to the central ones leading to unequal thermal expansions of the
studs. Also, the commonly observed loading of the studs using a steel beam does not
ensure equal load on all the studs. The problem is further compounded by unequal
expansion of the studs leading to inaccurate collapse load predictions.
To overcome all these problems it was considered necessary to build a special furnace
capable of delivering accurately the required heating regime along with a facility to
monitor and control the furnace pressure, choose appropriate instrumentation to
measure the temperature development across the wall specimens, custom build a
loading frame to enable the application of load to individual studs and maintain a
constant load ensuring free thermal expansion during the test procedure. Also, since
most of the research has been carried out outside Australia, representing specific
materials and method of construction used in those countries and as there has been no
research on LSF load bearing stud walls in Australia, it is important that these
investigations are carried out to assess the behaviour of Australian LSF stud wall
systems and provide recommendations to improve their fire ratings.
Chapter 3: Experimental Work to Determine the Mechanical
Properties of G500 Cold-formed Steel at Elevated Temperatures
3.1: Introduction
The deterioration of the mechanical properties of steel at elevated temperatures is the
primary cause of concern in the design of steel structures exposed to fire. The problem
is even more severe when thin sections made of high grade cold-formed steel are
used.
However, the mechanical properties of cold-formed steel at elevated temperatures
have not yet been fully understood and appropriate design values are not available to
the designers. Research carried out in the past has mostly focused on the reduction in
mechanical properties of hot-rolled steels at elevated temperatures. Consequently the
reduction factors for mechanical properties adopted in various steel design standards
are based on the results of hot-rolled steels.
The reduction factors for the mechanical properties of hot-rolled steels are considered
to be different from those of cold-formed steels. Sidey and Teague (1988) state that
the strength reduction factors of cold-formed steels at elevated temperatures may
differ by as much as 20% than those of hot-rolled steels at corresponding
temperatures. Most steel design standards do not provide the reduction factors for
mechanical properties of cold-formed steels at elevated temperatures except for
Eurocode 3: Part 1.2 (ECS, 2001) and BS 5950: Part 8 (BSI, 1990). However,
Eurocode 3: Part 1.2 (ECS, 2001) adopts the same reduction factors for both cold-
formed and hot-rolled steels at elevated temperatures, while BS 5950: Part 8 gives the
strength reduction factors for cold-formed steels at 0.5%, 1.5% and 2.0% strain levels
over a limited temperature range of 2000C to 6000C. The 0.2% proof stress which is
the most commonly used yield strength value in the designs is not included in BS
5950: Part 8 (BSI, 1990).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 52
Gerlich (1995), Makelainen and Miller (1983), Outinen et al. (2000), Lee et al.(2003),
Chen and Young (2004), and Ranawaka and Mahendran (2009) have investigated the
mechanical properties of cold-formed steels at elevated temperatures. However,
limited data is available for the mechanical properties at elevated temperatures of
cold-formed steels of Grade 500 and thickness 1.15 mm, which is fast gaining
popularity in the construction industry.
Use of high grade cold-formed steel is becoming popular in the construction industry
in Australia. G500 is being increasingly used in the construction of steel stud wall
systems. It is necessary to determine accurately the mechanical properties of cold-
formed steel at elevated temperatures to be able to determine the load carrying
capacity of stud wall systems under fire conditions. The mechanical properties of
cold-formed steels such as the yield strength, elastic modulus, ultimate strength and
ultimate strain at different temperatures can be obtained from the corresponding
stress-strain curves of steels at elevated temperatures. Therefore a series of tensile
tests was conducted to determine the mechanical properties of G500 steel at different
temperatures using steady state conditions.
Tensile testing was preferred over compression testing due to its simplicity, as past
research by Ranawaka and Mahendran (2009) has shown that the mechanical
properties obtained from tension and compression tests show minimal differences. In
steady state tests the specimen is heated up to a specified temperature and then the
tensile test is carried out by controlling either the loading rate or the strain rate.
Alternatively, the mechanical properties of steel at elevated temperatures could also
be determined by using the transient state method, wherein the load on the tensile
specimen is kept constant and then the temperature is raised until the specimen fails.
The transient state tests are considered to be more realistic in predicting the behaviour
under fire conditions than the steady state tests, because, in a real fire scenario, the
structural members are exposed to varying temperature under constant loads in which
the creep effect is also included (Outinen and Makelinen, 2002, Chen and Young,
2007, Lee et al., 2003). The creep effect is time dependent and influenced by both the
applied load and the temperature. As the tests are conducted generally within an hour
the effect of creep may be ignored. In transient test methods the temperature-strain
curves need to be converted into stress-strain curves to obtain the mechanical
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 53
properties of steel. In this conversion some approximation in the mechanical
properties of steel cannot be avoided.
In this study the steady state test method under controlled strain rate was preferred
due to the simplicity it offered in developing the stress-strain curves along with
accurate data acquisition.
3.2: Experimental Investigation
3.2.1 Test Specimens
Test specimens were prepared in accordance with the Australian Standard AS 2291
(SA, 1979). They were cut from structural steel sheets in the longitudinal direction
giving test specimens with dimensions as shown in Figure 3.1.
(a) Dimensions
(b) Strain Gauged Test Specimen
Figure 3.1: Tensile Test Specimen
The specimen included two holes at the ends to enable fixing to the loading shafts
located at the top and bottom ends of the furnace.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 54
Specimens were obtained from steel sheets of grade G500 with a thickness of 1.15
mm. The actual cross-sectional dimensions of the coupons were determined using a
micrometer after the removal of the zinc coating from the specimens. The coating was
removed by immersing the specimens in a dilute hydrochloric acid bath. The base
metal thickness (BMT) was then used in the determination of the mechanical
properties. The specimens were tested at ambient and eight elevated temperatures,
giving a total of 9 temperatures. Three specimens were tested per temperature, giving
a total of 27 test results.
3.2.2 Test Rig
An electrical furnace was used for the simulated fire tests to determine the mechanical
properties of G500 steel (see Figure 3.2). Four glow bars positioned at the four
corners of the furnace generated heat. Tensile specimen was located at mid-height of
the furnace equidistant from the four glow bars to ensure uniform heating. A
programmable logic controller was used to regulate the furnace temperature. The
furnace could deliver a maximum temperature of 11000C. Two internal thermocouples
located inside the furnace were used to measure the furnace temperature.
An additional thermocouple kept in contact with the specimen measured the surface
temperature of the specimen. The specimen was mounted between the two loading
shafts made of 253 MA stainless steel for satisfactory operation at elevated
temperatures.
Figure 3.3 shows the details of the test rig and its components. The upper loading
shaft passing through an insulated hole in the roof of the furnace was connected to a
hydraulic actuator of capacity 45 kN. A load cell of capacity one ton was connected as
shown in Figure 3.3 (b) for load determination. The hydraulic actuator was rigidly
connected to a cross-head at the top. The bottom loading shaft passing through an
insulated hole in the base of the furnace was rigidly connected to the floor of the
structural laboratory (see Figure 3.3 (c)). Special care was taken to align the loading
shafts in order to avoid eccentric loading of the test specimen. Figure 3.3 (e) shows
the installation of the test specimen in the furnace.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 55
Tensile load
Hydraulic actuator
Load cell
Top loading shaft Glow bars Thermocouples Specimen Bottom loading shaft
Figure 3.2: Furnace Details
The hydraulic actuator was connected to a Multi-Purpose Test Ware System. The
furnace and specimen temperatures were recorded using an automatic data acquisition
system (EDCAR) at intervals of one minute. Data channels from the load cell, strain
gauges and control device of laser speckle extensometer were connected to EDCAR
to process the data and obtain the stress-strain curves (see Figure 3.4).
The elongation of the specimen was measured in the middle portion of the specimen
using an advanced Laser Speckle Extensometer (LSE) as resistance type strain gauges
are good at only room temperatures and get easily damaged at elevated temperatures.
Also the use of mechanical extensometers to measure the strains at elevated
temperatures was considered unsatisfactory as its accuracy largely depended upon the
precision of the connecting devices. The LSE comprises of a PC based video
processor, laser diodes (class 3A), video cameras, lens and a frame grabber. The
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 56
frame grabber digitalizes the analogue video signal and these digitized images are
displayed on the computer screen. The video processor is capable of continuously
measuring the displacements of two speckle patterns developed at a specified distance
apart, by the laser beams and recorded by the video cameras on the specimen gauge
length.
(a) Furnace (b) Loading shaft at the top connected to the actuator
(c) Loading shaft connected to floor (d) LSE mounted on a frame outside the glass window
Figure 3.3: Details of Test Rig and its Components
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 57
(e) Specimen installed in furnace (f) LSE being used to measure the extension of test specimen
Camera lens
(g) Main Components of LSE
Figure 3.3: Details of Test Rig and its Components
Laser outlets
Pivot mounted cameras
Data transferring
cable
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 58
Figure 3.4 EDCAR (Experimental Data Collection and Recorder)
The displacement of the speckle patterns are converted to a strain signal and sent to an
external control system. The instrument was set up in front of the observation window
built of special heat resistant glass, located on one side of the furnace (see Figure 3.3
(d)). The gauge length of the extensometer is the distance between its two cameras,
which was set at 50 mm. Infrared filters were used in front of the camera lens at
temperatures greater than 5000C for better visibility of the test coupons. Figure 3.3 (f)
shows the instrument in use and Figure 3.3 (g) shows the main components of LSE.
The extensometer was calibrated before testing. The laser beam was located behind
the furnace such that the cameras could be directed onto the specimens gauge length
through the special window made from fire resistant glass. In this method of strain
measurement, two laser beams along with two cameras are oriented targeting the
specimens as shown in Figure 3.5. The upper camera is referred to as the slave and the
lower camera as the master.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 59
Figure 3.5 Strain Measurement using LSE
The measuring principle of the extensometer is based on tracking laser speckle
patterns by the use of two upper and lower cameras named as slave and master,
respectively. The extensometer works on the following principle. When a coherent
laser beam is directed on to an optically rough surface, the light gets diffused in
different directions. If the diffused light rays travel through the original beam, the
light is spatially eliminated, resulting in a granular looking speckle pattern as shown
in Figure 3.6. Each camera records unique speckle patterns relevant to the zones it is
directed. Speckle patterns corresponding to the two zones on the specimen which are
separated from each other by a predetermined distance in the elongation direction are
initially stored as reference speckle patterns. When the tensile load is applied to the
specimen, targeted zones of the two cameras change causing a shift in the speckle
patterns observed by the cameras. The video processor is able to locate the new
position of a stored reference pattern and calculate the distance it has moved between
images. Figure 3.6 (a) shows typical speckle patterns of master and slave cameras
before the test while Figure 3.6 (b) shows typical speckle patterns during the test.
Before the tests, the laser-speckle extensometer was calibrated with a special
calibration method, which enables accurate strain measurements. Since the distance
between initial reference patterns given by the two cameras is adjusted to 50 mm and
is stored in the program before starting the testing, the processor is able to calculate
the strain at any time using Equation 3.1.
50 mm
Slave camera
Laser beams
Specimen
Master camera
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 60
o
masterslave
l
dd ………………………………………………….............….... (3.1)
where, - Sum of displacements on Slave Camera slaved
masterd - Sum of displacements on Master Camera
ol - Distance between initial reference patterns
Selected speckle patterns
Output of slave camera Output of
slave camera
(a) Speckle Output before the Test
Distance travelled from master camera
Distance travelled from slave camera
(b) Speckle Output during the Test
Figure 3.6: Typical Speckle Output for Strain Measurements
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 61
The LSE was validated by using the strain measurements taken at room temperatures
using strain gauges. Figure 3.7 shows a good correlation between the readings
obtained by the LSE and the conventional strain gauges.
Figure 3.7: Comparison of Stress-Strain Curves using Strain Gauges and Laser Speckle Extensometer
3.2.3 Test Procedure
Tensile tests were carried out based on the steady state test method. The specimens
were heated at a rate of 20-250C per minute from ambient temperature up to the pre-
selected temperature and then loaded up to failure while maintaining the same
temperature. A very small tensile load was maintained in the specimen during the
heating phase in order to eliminate the development of compressive forces in the
specimen due to thermal expansion of the specimen and the connecting shafts. After
reaching the specified temperature, a time of approximately 10 minutes was allowed
to elapse for the temperature to stabilize before the application of load. A constant
displacement rate of approximately 0.2 mm/min was adopted for the loading shaft
during the tests. This was equivalent to a strain rate of 0.0033/minute, which was
within the range of 0.001 to 0.005/minute according to the testing standard for
metallic materials, SFS-EN10002-5 (ECS, 1992).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 62
The tests were carried out on the specimens in the elevated temperature range of
1000C to 8000C at intervals of 1000C.
3.2.4 Test Results
The yield strengths at strain levels of 0.2%, 0.5%, 1.5% and 2.0% were obtained for
the purpose of comparison as these strain levels are accepted widely. The 0.2% yield
strength (f0.2) is the intersected value of the stress-strain curve and the proportional
line, offset by 0.2% strain whereas, the yield strengths of f0.5, f1.5 and f2.0 at the strain
levels of 0.5%, 1.5% and 2.0% respectively are obtained from the intersected values
of the stress–strain curve and the non-proportional vertical lines drawn from the
specified strain levels as shown in Figure 3.8(a).
(b) (a)
Figure 3.8: Determination of (a) Yield strength and (b) Elastic modulus. (Lee et al., 2003)
The yield strength values determined based on 0.2% proof stress were used in
deriving the reduction factors for the yield strengths at elevated temperatures. The
modulus of elasticity was determined from the tangent modulus of the initial elastic
linear part of the stress-strain curve, as shown in Figure 3.8(b).
Yield Strength: Table 3.1 presents the yield strengths at various percentages of strain
(f0.2, f0.5, f1.0, f2.0), ultimate strength (fu), percentage strain at failure (εu) and modulus
of elasticity (E) at ambient and elevated temperatures.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 63
From Table 3.1 and Figure 3.9, it is observed that the mechanical properties are not
much affected up to a temperature of 3000C. Beyond 3000C the drop in mechanical
properties is observed to be very rapid with about 90% of the strength lost at around
6000C. Table 3.1 shows that the ductility values increase with increasing temperature.
The ductility is observed to increase very rapidly for temperatures greater than 4000C.
Table 3.1: Mechanical Properties of 1.15 mm G500 CFS at Ambient and Elevated Temperatures.
Temperature f0.2
(MPa) f0.5
(MPa) f1.5
(MPa) f2.0
(MPa) fu
(MPa) εu
(%) E
(GPa)
27 569.00 583.25 583.50 584.5 589.00 10.70 213.52
100 565.00 578.25 577.83 579.81 590.60 7.10 209.92
200 560.00 570.50 580.20 580.72 600.00 6.40 193.55
300 539.00 556.59 582.17 585.03 606.50 14.35 166.61
400 400.00 408.27 455.13 455.91 457.50 13.10 154.15
500 219.00 221.63 274.25 280.56 283.00 10.03 85.25
600 69.00 70.00 70.02 70.14 74.00 30.00 63.16
700 39.50 40.82 40.85 46.76 45.00 30.00 43.18
800 23.00 23.33 29.18 29.225 28.00 47.00 11.88
The Stress-Strain graphs for 0.2% Proof stress obtained at different temperatures are
shown in Figure 3.9.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 64
Figure 3.9: Stress-Strain Graphs at Different Temperatures
The reduction factors (fy,T/fy,20) determined as the ratios of different yield strengths at
elevated temperatures to that at ambient temperature corresponding to the strain levels
of 0.2%, 0.5%, 1.5% and 2.0% are presented in Table 3.2 along with the reduction
factors (ET/E) determined for the modulus of elasticity at different elevated
temperatures to that at ambient temperature.
Table 3.2: Reduction Factors for Yield Strength and Modulus of Elasticity of 1.15 mm G500 Steel
Temperature (0C)
27 100 200 300 400 500 600 700 800
f0.2,T/f0.2 1.00 1.00 0.98 0.95 0.7 0.39 0.12 0.07 0.04
f0.5,T/f0.5 1.00 0.99 0.98 0.95 0.70 0.38 0.12 0.07 0.04
f1.5,T/f1.5 1.00 0.99 0.99 0.99 0.78 0.47 0.12 0.07 0.05
f2.0,T/f2.0 1.00 0.99 0.99 1.00 0.78 0.48 0.12 0.08 0.05
1.15
mm
G500
ET/E 1.00 0.98 0.91 0.78 0.72 0.4 0.3 0.2 0.06
Figure 3.10 shows the Strength Reduction Factors (more appropriately Strength
Retention Factors) corresponding to various percentages of yield strength. The yield
strength values are seen to drop suddenly from 3000C to 6000C. The drop in strength
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 65
values for cold-formed steel is observed to be more rapid than in hot-rolled steel
probably due to the loss in strength initially gained by the cold-forming process.
Figure 3.10: Graph Showing Strength Reduction Factors associated with Various
Percentages of Yield Strength as Obtained from Tests
Note:
fy (0.2%): 0.2% Proof Stress
fy (0.5%): Yield strength corresponding to a total strain of 0.5%
fy (1.5%): Yield strength corresponding to a total strain of 1.5%
fy (2.0%): Yield strength corresponding to a total strain of 2.0%
3.3: Comparison of Reduction Factors with Results as Obtained by Other
Researchers and as Recommended by Steel Design Codes
Figures 3.11 and 3.12 show the comparison of test results, with the reduction factors
as obtained by other investigators for 0.2% proof stress and modulus of elasticity,
respectively.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 66
Figure 3.11: Yield Strength Reduction Factors
Figure 3.12: Modulus of Elasticity Reduction Factors
The yield strength as observed by other researchers is seen to drop progressively from
ambient temperature with increase in temperature. The yield strength as obtained by
most researchers is seen to be unconservative beyond 4500C. The reduction in
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 67
modulus of elasticity of the test specimens and as observed by other researchers is
seen to observe a progressive reduction in its value as the temperature increases.
Figures 3.13 ((a) to (c)) show the reduction in the yield strength of cold-formed steel
test specimens as compared to that given by BS 5950.8 (BSI, 1990). It is observed
that BS 5950.8 (1990) gives values on the conservative side up to temperature of
around 3250C, beyond which the yield strength of cold-formed steel is seen to fall
rapidly making the values given by BS 5950.8 (1990) more and more unconservative.
Similar observations can be made in Figure 3.14 where the yield strength of cold-
formed steel test specimens corresponding to 0.2% proof stress is compared with the
yield strength of hot-rolled steel as given by AS 4100 (SA, 1998).
(a) Comparison of 0.5% Strength Reduction Factors as Suggested by BS 5950-8 (BSI, 1990) with Test Results
Figure 3.13: Variation of Yield Strength Reduction Factors with Temperature
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 68
(b) Comparison of 1.5% Strength Reduction Factors as Suggested by BS 5950-8 (BSI, 1990) with Test Results
(c) Comparison of 2.0% Strength Reduction Factors as Suggested by BS 5950-8 (BSI, 1990) with Test Results
Figure 3.13: Variation of Yield Strength Reduction Factors with Temperature
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 69
Figure 3.14: Comparison of 0.2% Strength Reduction Factors with AS 4100 (SA, 1998) Recommendations
Figure 3.15: Comparison of Modulus of Elasticity Reduction Factors with AS 4100 (SA, 1998) Recommendations
The yield stress ratios determined from the test results compared well with the values
given in AS 4100 (SA, 1998) up to a temperature of 2150C. Beyond this temperature
and up to 3800C the values given by the code were seen to be conservative. However
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 70
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 71
beyond 3800C the code values were seen to be unconservative right up to 8000C with
a notable difference at 6000C, at which point the code considers a yield strength ratio
of 0.44 (i.e. a 56% reduction in yield strength) for design purposes, whereas the test
results at the same temperature give a yield stress ratio of 0.12 (i.e. a 82% reduction in
yield strength). The strength and stiffness values adopted by AS 4100 (SA, 1998) are
seen to be grossly unconservative in the steel temperature range of 5000C – 7500C as
can be seen in the Figures 3.14 and 3.15, respectively. The modulus of elasticity ratios
of AS 4100 (SA, 1998) are seen to be unconservative from 1400C onwards when
compared with the test results. At 6000C the code adopts a reduction factor of 0.5 (i.e.
a 50% reduction in stiffness) whereas the test results give a value of 0.3 (i.e. a 70%
reduction in stiffness)
As the mechanical properties of high grade cold-formed steel are significantly
different from the values proposed by AS 4100 (SA, 1998), following empirical
equations based up on the tensile coupon tests have been proposed to better represent
the behaviour of high grade cold-formed steels at elevated temperatures. The
developed equations relate the reduction factors of strength and stiffness with respect
to temperature.
Although various strain levels (0.2%, 0.5%, 1.5% and 2.0%) were considered in
determining the yield strength, only the reduction factors based on the 0.2% proof
stress method were used in deriving the empirical equations. This is because the yield
strengths based on other strain levels have not been accepted widely.
Equations 1((a) to (c)) give the yield strength reduction factors with respect to
temperature.
(fyT / fy20 ) = -1.891 X 10-4 T + 1.012 27 < T ≤ 300 Eq. 1(a)
(fyT / fy20 ) = -2.8 X 10-3 T + 1.8 300 < T ≤ 600 Eq. 1(b)
(fyT / fy20 ) = -4 X 10-4 T + 0.356 600< T ≤ 800 Eq.1(c)
Equation 2 gives the reduction factors of modulus of elasticity with respect to
temperature.
ET/E = -5.733 X 10-7 T2 – 8.419 X 10-4 T + 1.0637 100< T ≤ 800 Eq.2
Figures 3.16 and 3.17 show a good agreement between the values predicted from
these equations and the test results.
Figure 3.16: Comparison of Yield Strength Reduction Factors from Test Results and Predictive Equation
Figure 3.17: Comparison of Elastic Modulus Reduction Factors from Test Results and Predictive Equation
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 72
Figures 3.18 and 3.19 show the comparison of the proposed equations based on the
experimental work for G500 grade of steel with the generalised equations as given by
Ranawaka and Mahendran (2009).
Figure 3.18: Comparison of Ranawaka and Mahendran’s (2009) Equation and Predictive Equation in the Determination of Yield Strength Reduction Factors
Figure 3.19: Comparison of Ranawaka and Mahendran’s (2009) Equation and Predictive Equation in the Determination of Elastic Modulus Reduction Factors
The yield strength reduction factors as obtained by Ranawaka and Mahendran’s
(2009) equation are slightly unconservative in the temperature range of 3500C to
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 73
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 74
6000C as compared to the proposed predictive equation. Also the strength reduction as
observed in the experimental work is seen to be linear with increase in temperature
instead of being non-linear as given by Ranawaka and Mahendran’s (2009) equation.
The elastic modulus reduction factors given by the proposed predictive equation are
observed to conform closely to Ranawaka and Mahendran’s (2009) equation.
3.5: Conclusion
To address the lack of reliable mechanical property data for cold-formed steel at
elevated temperature, detailed experimental work was carried out as part of this study
in the Structural laboratory of Queensland University of Technology. Tensile coupon
tests of G500 cold formed steel were undertaken at temperatures ranging from 1000C
to 8000C under steady state conditions to obtain the reduction factors (more
appropriately retention factors) for strength and modulus of elasticity.
The current Australian and European Standards do not present accurate reduction
factors for the yield strength and elasticity modulus of cold-formed steels at elevated
temperatures. These factors are observed to be grossly unconservative at temperatures
beyond 3500C. To provide more accurate mechanical properties and facilitate safer
design of cold-formed steel structures at elevated temperatures, predictive equations
have been developed based on the results from this research for calculating yield
strength and elastic modulus. These equations were seen to compare well with the test
results and conservatively predict the yield strength and elastic modulus at elevated
temperatures as compared to the existing standards.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 75
Chapter 4: Thermal Performance of Gypsum Plasterboards and
Composite Panels
4.1: Introduction
Fire resistance of non-load bearing and load bearing wall systems depends to a large
extent on the level of protection offered to the steel frame against fire attack. The
most popular method of providing this protection to the steel frame is by attaching
gypsum plasterboard sheets on either side of the frame. To improve the fire resistance
of the wall, multiple sheets are attached on either side. When the wall is exposed to
fire from one side the plasterboards form the first line of defence by protecting the
steel from the intense heat generated by the fire. A temperature gradient inevitably
develops across the depth of the wall. A substantial drop in temperature occurs across
the thickness of each layer of plasterboard used. With prolonged exposure to fire the
plasterboards calcine and develop cracks allowing the heat to penetrate, eventually
leading to the failure of the wall.
To better understand the thermal performance of the gypsum plasterboards many
experiments were conducted in the Fire Research Laboratory of Queensland
University of Technology. Fire tests were performed on Type X gypsum plasterboards
supplied by the company Boral Plasterboards under the product name FireSTOP.
Thermal performance of single, double and triple layers of plasterboards was studied.
Different types of insulations were also used to help improve the fire performance.
Composite panels were developed with a layer of insulation between two sheets of
plasterboard.
This chapter presents the details of a series of fire tests performed on individual and
multiple layers of plasterboard. It also examines and compares the thermal
performance of composite panels developed from different insulating materials of
varying densities and thicknesses and makes suitable recommendations.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 76
4.2: Test Setup and Procedure
Fire tests were carried out by exposing one face of the specimens to heat in a propane-
fired vertical gas furnace. An adapter was specially designed to fit into the large
furnace so as to isolate a single burner to facilitate the testing of small scale specimens
(see Figures 4-1 and 4-2). The internal dimensions of the adapter were 1290 mm x
1010 mm. The specimen to be fire tested was mounted on a platform extending from
the base of the adapter such that the specimen would enclose the open furnace
chamber. On starting the furnace, the specimen was exposed to heat from one side
only. The furnace temperature was measured using four Type K mineral insulated and
metal sheathed thermocouples, each being placed at the centre of the four quarters
formed by the horizontal and vertical centre lines. Care was taken to ensure that the
distance of the hot junction of all the furnace thermocouples from the fire surface of
the test specimen was about 100 mm. The average temperature rise of these
thermocouples served as the input to the computer controlling the furnace heat
according to the cellulosic fire curve (Standard time-temperature curve) given in AS
1530.4 (SA, 2005), which is similar to ISO 834-1 (1999) and ASTM E119 (1995).
Additionally four more thermocouples were uniformly distributed in the chamber so
as to give a reliable indication of the average temperature of the furnace chamber in
the vicinity of the test specimen. These thermocouples were connected to the data
logger and used for the plotting of the furnace time-temperature graphs.
The adapter was able to utilise the existing exhaust vent built into the left side wall of
the large furnace as shown in Figure 4-1. To establish proper hot air circulation within
the small chamber another exhaust vent was built by drilling a hole into the right side
wall of the adapter. Both the exhausts were fitted with control gates to monitor the
exhaust opening sizes and achieve greater control on the convection currents within
the chamber. This provided a more uniform temperature over the height of the
chamber. The specimens were installed in the furnace as shown in Figure 4-3. A
pressure transducer was also used to measure the pressure inside the furnace chamber
during the fire test (see Figure 4-4). Time-temperature profiles at various locations
across the thickness of the Test Specimens were plotted from the data available to
help assess their fire performance. The tests were stopped once the plasterboard paper
on the ambient side of the specimen started to burn.
Large Furnace
Existing left vent in the large furnace
Right vent built in the adapter
Adapter for fire testing of small scale specimens
Figure 4-1: View Showing Adapter Attached
to Large Furnace for Carrying Out Fire Testingof Small Scale Specimens
Supports for attaching top clamps to hold the specimen in place
Valve to control vent opening
150 mm thick ceramic fibre lining
Slots to attach platform for supporting small scale specimens
Figure 4-2: Adapter Details
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 77
Figure 4-3: View Showing Plasterboard Specimen Installed for Fire Testing
Figure 4-4: Pressure Transducer used for Determining Furnace Chamber Pressure during Testing
4.3: Test Specimens
Fire tests were conducted on gypsum plasterboard specimens each measuring 1350
mm x 1080 mm in dimensions. The specimens were built using either single or
multiple plasterboards. Composite panel specimens using different types of
insulations placed between the plasterboards were also built. Insulation densities were
varied to study their effect on the fire performance. K type wire thermocouples were
inserted within the body of the plasterboard by drilling holes so as to measure the
temperature variation at different depths across the thickness of the plasterboard when
exposed to fire from one side. The holes were drilled normal to the plane of the
plasterboard to the required depth at mid-height of the specimen. The hot junction of
the wire thermocouple was then inserted into the hole, and the hole was then sealed
off using moist powdered gypsum plasterboard. Minimum of two holes were drilled
and thermocouples installed to determine the temperature profile at any particular
depth. Fifteen test specimens were built as shown in Table 4-1. The position of
thermocouples is indicated by the coloured dots. Construction details of each
specimen have been discussed after the table.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 78
Table 4-1: Details of Plasterboard and Composite Panel Test Specimens
No. Configuration Specimen Description
1 Pb = 13 mm
2
Pb = 16 mm
3
Pb1 = 13 mm (Fire Side)
Pb2 = 16 mm (Ambient Side)
4
Pb1 = 16 mm (Fire Side)
Pb2 = 16 mm (Ambient Side)
5 Pb1 = 16 mm (Fire Side)
Pb2 = 16 mm (central)
Pb3 = 16 mm (Ambient Side)
6, 7,
8, &
9
Pb1 = 16 mm (Fire Side) Insulation: Glass Fibre of varying thickness, density and type. Pb2 = 16 mm (Ambient Side)
10 &
11
Pb1 = 16 mm (Fire Side) Insulation: Rock Fibre of varying thickness, density and type. Pb2 = 16 mm (Ambient Side)
12, 13
& 14
Pb1 = 16 mm (Fire Side) Insulation: Cellulose Fibre of varying thickness, density and type. Pb2 = 16 mm (Ambient Side)
15 Pb1 = 16 mm (Fire Side) Insulation: Isowool. Pb2 = 16 mm (Ambient Side)
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 79
To measure the temperature of the ambient surface of the Test Specimens, five
thermocouples were attached to the unexposed surface of the plasterboard, one
thermocouple at the centre of the specimen and one at the centre of each quarter
section of the specimen as shown in Figure 4-5.
Thermocouples
Figure 4-5: Thermocouples on the Ambient Side of the Specimen
Fifteen Test Specimens were built and tested to achieve the following objectives:
1) To study the thermal performance of a single layer of plasterboard including
the temperature distribution across the thickness of the plasterboard.
2) To determine the influence of chemically bound and free water present in the
body of the plasterboard on the ambient side temperature of the plasterboard
specimens.
3) To study the effect of joints between plasterboards on the temperature
distribution across the thickness of the plasterboards.
4) To study the effect of joints between plasterboards on the ambient side
temperature of the plasterboard specimens.
5) To study the effect of multiple boards on the thermal performance of
specimens.
6) To study the effect of various types of insulation with varying density and type
on the thermal performance of the composite boards.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 80
4.3.1: Test Specimen 1
4.3.1 A): Construction Details
This specimen was made of a single layer of 13 mm thick gypsum plasterboard.
Thermocouples were located on the specimen as shown in Figure 4-6. Five additional
thermocouples were attached on the ambient face to measure the temperature of the
unexposed surface of the wall. Thus a total of nine Thermocouples (4+5) were used to
measure the temperature across the specimen.
Fire Side (Exposed Surface)
13 mm Plasterboard
Ambient Side (Unexposed Surface)
Two thermocouples on the fire side at mid height
Two thermocouples at 7 mm from the fire side at mid height
Five thermocouples on the ambient side as shown in Figure 4-5
Figure 4-6: Instrumentation for Test Specimen 1
4.3.1 B): Observations, Results and Discussions
One side of Test Specimen 1 was subjected to the standard time-temperature heating
regime in the furnace (see Figure 4-7 (a)). By the end of 3 minutes smoke was seen to
start coming from the edges of the specimen. This was on account of the burning of
the plasterboard paper on the exposed side. The smoke subsided after the paper got
completely burnt out. By the end of 6 minutes steam was seen to come out from the
specimen and condense on the top front face of the furnace adapter. By the end of 12
to 13 minutes the steam subsided and the specimen soon was seen to burn steadily
without letting out smoke or steam. By the end of 18 minutes the ambient side paper
of the plasterboard started to discolour. The specimen was also seen to bow laterally
outward (see Figure 4-7 (b)). This was probably caused due to the shrinkage of the
surface exposed to furnace following the expulsion of water. By the end of 33 minutes
the outside paper had started to burn and the test was stopped. The specimen was seen
to have lost most of its strength and had developed deep vertical cracks on the
exposed surface, although the cracks were not visible on the ambient surface.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 81
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 82
Figure 4-8 shows the time-temperature profile across the plasterboard thickness for
Test Specimen 1. The time-temperature profiles obtained within the thickness of the
specimen at a distance of 7 mm from the exposed surface (fire side) and on the
ambient surface show the development of the temperature in three phases. The first
phase displayed a steady rise in temperature from the ambient temperature to around
1000C. This was followed by the second phase where the temperature was maintained
close to 1000C thus giving a plateau. In this phase the heat energy was primarily used
up to convert the free and chemically bound water in the plasterboard to steam. The 7
mm depth and 13 mm depth profiles have their second phase extending up to
approximately 6 and 12 minutes respectively. The third phase started when moisture
in the plasterboard was no longer available for conversion into steam. The
temperature in the third phase was seen to increase gradually reaching 4500C at 7 mm
depth and 2750C on the ambient surface towards the end of the test.
The test was stopped on account of the paper on the ambient side burning. The
thermocouple at 7 mm depth from the exposed surface did not record any sudden rise
in temperature in the third phase up to the end of the test. This implies that the layer of
plasterboard up to 7 mm depth, though calcinated, was still intact and prevented any
sudden ingress of heat. A temperature difference of approximately 3500C was
observed from the exposed surface to a depth of 7 mm. With a further drop of
approximately 2000C from 7 mm to 13 mm depth giving a temperature of 2750C on
the ambient surface.
Figure 4-9 shows a graph of temperature versus depth plotted at intervals of ten
minutes. The profile at the end of ten minutes shows the curve linear from the
exposed surface up to the 7 mm depth and then gradually flattening out and merging
with 1000C line at a depth of 8.5 mm. This implies that, at the end of 10 minutes of
fire exposure, the specimen still had moisture from 8.5 mm to 13 mm depth. This can
be verified from Figure 4-8 where it can be noticed that, at the end of 10 minutes, the
7 mm depth profile had entered the third phase but the 13 mm depth profile was still
in the second phase. The 20 minute profile in Figure 4-9 shows the temperature at all
the depths above 1000C implying that moisture had been completely driven out of the
entire thickness of the specimen.
With the expulsion of water from across the thickness of the specimen, the
plasterboards became more and more calcinated with the development of several
shrinkage cracks over the surface and within the body of the plasterboard. At this
stage, the graph of temperature vs depth (see Figure 4-9) approaches linearity. The
two profiles, 20 minutes and 30 minutes, in Figure 4-9 obtained after the expulsion of
water from the plasterboard body are seen to run almost parallel to each other (i.e. the
slopes are almost the same) suggesting that the thermal properties of the calcinated
plasterboard does not undergo much change with respect to temperature as long as the
integrity is maintained.
(a) Test Specimen 1 at the start of the test (b) Thermal bowing visible towards the end of the test
Figure 4-7: Fire Testing of Test Specimen 1
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 83
0
100
200
300
400
500
600
700
800
900
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS 7 mm Amb
Figure 4-8: Time-Temperature Profile of Test Specimen 1
Note:
AS 1530.4: Standard time-temperature curve from AS 1530 Part 4
F.S.: Temperature profile of the exposed surface of the specimen (Fire Side surface)
7 mm: Temperature profile at a depth of 7 mm from the exposed surface
Amb: Temperature profile of the unexposed surface of the specimen
0
100
200
300
400
500
600
700
800
900
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Depth (mm)
Tem
per
atu
re (
oC
)
10 min 20 min 30 min
Figure 4-9: Temperature-Depth Profiles of Test Specimen 1
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 84
4.3.2: Test Specimen 2
4.3.2 A): Construction Details
This specimen was made of a single layer of 16 mm thick gypsum plasterboard.
Thermocouples were located on the specimen as shown in Figure 4-10. A total of 13
thermocouples (8+5) were used to measure the temperature across the specimen.
Fire Side (Exposed Surface)
16 mm thick plasterboard
Ambient Side (Unexposed Surface)
Two thermocouples on the fire side at mid-height
Two thermocouples at 4 mm from the fire side at mid-height
Two thermocouples at 8 mm from the fire side at mid-height
Two thermocouples at 12 mm from the fire side at mid-height
Five thermocouples on the ambient side as shown in Figure 4-5.
Figure 4-10: Instrumentation for Test Specimen 2
4.3.2 B): Observations, Results and Discussions
The specimen was fire tested for about 78 minutes. The observations pertaining to the
evolution of smoke and steam were similar to that of Test Specimen 1. Figure 4-11 (a)
shows Test Specimen 2 mounted in the small adapter for the fire test. By the end of 29
minutes, the paper on the ambient surface started to discolour uniformly. By 40
minutes, the ambient surface had become quite dark. Towards the end of the test, the
paper was partially burnt and the specimen had begun to bow laterally in the outward
direction (see Figures 4-11 (b) and (c)).
The plateaus for the 4, 8, 12 and 16 mm depth profiles in Figure 4-12 were seen to
extend approximately up to about 3, 7, 12 and 18 minutes respectively. That is, on
average, 1 min of fire exposure was required to expel water from 1mm thickness of
the plasterboard. Thus, at the end of 10 minutes, a plasterboard thickness of
approximately 10 mm from the exposed surface would have its water expelled. This
means, the plateau of the unexposed surface, in the case of 13 mm plasterboard would
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 85
extend up to 13 minutes and for 16 mm plasterboard it would extend approximately
up to 16 minutes.
In Figure 4-12 it can be clearly seen that the 12 mm depth profile has entered the third
phase at the end of 15 minutes whereas the 16 mm depth profile was still in the
second phase implying the presence of moisture in the last few mm thickness of the
plasterboard. The 75 minutes profile in Figure 4-13 is seen to approach linearity. It is
seen to be almost parallel to the 30 minute profile signifying very little change in the
thermal properties of the plasterboard over that duration of time.
(a) Test Specimen 2 at the Start of Test (b) Test Specimen 2 at the End of Test
(c) Thermal Bowing Visible at the End of Test Figure 4-11: Fire Testing of Test Specimen 2
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 86
0
100
200
300
400
500
600
700
800
900
1000
1100
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS 4 mm 8 mm 12 mm Amb
Figure 4-12: Time-Temperature Profile of Test Specimen 2
0
100
200
300
400
500
600
700
800
900
1000
1100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17Depth (mm)
Tem
per
atu
re (
oC
)
15 min 30 min 45 min 60 min 75 min
Figure 4-13: Temperature-Depth Profiles of Test Specimen 2
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 87
4.3.3: Test Specimen 3
4.3.3 A): Construction Details
This specimen consisted of two layers of gypsum plasterboards, one of 13 mm
thickness and the other of 16 mm thickness attached to each other by 30 mm screws
spaced at 300 mm centres along the periphery. The 13 mm thick plasterboard was
labelled as Pb1 and formed the exposed side of the specimen, whereas the 16 mm
thick plasterboard was labelled as Pb2 and formed the ambient side of the specimen.
Thermocouples were located on the specimen as shown in Figure 4-14. Thirteen
thermocouples were used to measure the temperature across the specimen.
Fire Side (Exposed Surface)
13 mm thick Plasterboard (Pb1)
16 mm thick Plasterboard (Pb2)
Ambient Side (Unexposed Surface)
Two thermocouples on the fire side (No. 2)
Two thermocouples at 7 mm from the fire side and within Pb1 at mid-height
Two thermocouples at 13 mm from the fire side i.e. in the interface of Pb1-Pb2 at mid-height
Two thermocouples at 21 mm from the fire side and within Pb2 at mid-height
Five thermocouples on the ambient side as shown in Figure 4-5
Figure 4-14: Instrumentation for Test Specimen 3
4.3.3 B): Observations, Results and Discussions
Test Specimen 3 was exposed to fire for a period of 171 minutes. The fire side paper
of the exposed plasterboard caught fire by the end of 3 minutes when the temperature
of the exposed surface was around 4000C. The smoke was soon followed by steam
which continued for 4 to 5 minutes. This was followed by a period of steady burning
during which time there was hardly any emission of smoke or steam. By the end of 20
minutes, smoke reappeared. This was probably due to the plasterboard paper on the
ambient side of Pb1 (Plasterboard 1) burning. By the end of 62 minutes the
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 88
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 89
plasterboard on the unexposed surface of the specimen started to discolour, when its
temperature was about 2000C.
Figure 4-15 (a) shows Test Specimen 3 with all its instrumentation at the start of the
test. Towards the end of the test the paper on the ambient surface of the specimen was
noticed to have blackened uniformly (see Figure 4-15 (b)). Horizontal and vertical
folds in the paper indicated that deep cracks in the plasterboard body had reached the
ambient surface with only the paper holding the pieces together. The test was stopped
at this stage. The specimen had undergone lateral deformation in the outward
direction towards the end of the test. The plasterboard pieces of the test specimen
when closely observed showed that the glass fibres used in the making of Type X
gypsum boards were intact only in the layers close to the unexposed surface (see
Figures 4-15 (c) and (d)) of the specimen.
Plasterboard 1 had undergone severe calcination with a large network of shrinkage
cracks. The glass fibres in this plasterboard had completely melted. In spite of the
severe calcination of Plasterboard 1 (13 mm board) compared to Plasterboard 2 (16
mm board) towards the end of the test, the drop in temperature across the thickness of
the specimen, (slope of the 150 minutes profile in Figure 4-17) was seen to be almost
uniform, again suggesting that the thermal properties of the plasterboard do not
undergo much change even when it is severely calcinated.
In Figure 4-16 the plateaus for the 7, 13, 21 and 29 mm depth profiles were seen to
extend up to about 6, 14, 31 and 54 minutes, respectively. As observed from the
previous specimens, Plasterboard 1 had expelled its moisture across the thickness of
13 mm in approximately 13 – 14 minutes. However, Plasterboard 2 showed extended
periods of plateau. This was probably because of the heat that was allowed to escape
from the interface Pb1-Pb2 thus lowering the severity of the fire on Plasterboard 2.
The 13 mm curve in Figure 4-16 shows the time-temperature profile of the interface
(Pb1-Pb2). A temperature drop of approximately 4000C was observed from the
exposed surface to the ambient side of Plasterboard 1 (i.e. Pb1-Pb2 interface) and a
further drop of approximately 5500C up to the unexposed surface of the specimen
towards the end of the test. Both plasterboards remained intact until the end of the
test. The thermocouples recording the fire side temperature displayed a sudden
increment in temperature at around 32 minutes. This has been assumed to be on
account of the malfunctioning of the wire thermocouples at temperatures approaching
9000C.
(a) Test Specimen 3 at the Start of Test (b) Test Specimen 3 at the End of Test
(c) Glass Fibres Intact Close to the (d) Glass Fibres Seen at the Top
Unexposed Surface of the Specimen Right Corner
Figure 4-15: Fire Testing of Test Specimen 3
The joint in the plasterboard is seen to have an influence on the length of the plateau
(second phase) in the time-temperature profile of the ambient surface of the
plasterboard. The plateau is seen to last up to 44 minutes as against the expected 29
minutes which is the combined thickness of the two plasterboards (see Figure 4-16).
The extra 15 minutes is probably on account of the possible recondensation of steam
in the interface.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 90
0100200300400500600700800900
1000110012001300
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS 7 mm 13 mm 21 mm Amb
Figure 4-16: Time-Temperature Profile of Test Specimen 3
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30Depth (mm)
Tem
per
atu
re (
oC
)
30 min 60 min 90 min 120 min 150 min
Figure 4-17: Temperature-Depth Profiles of Test Specimen 3
The 7 mm and 13 mm depth temperature profiles of Specimen 3 are seen to display
higher temperatures than the equivalent depth temperature profiles of Specimen 1 at
corresponding times. This is due to the influence of Plasterboard 2 in Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 91
which blocks the escape of heat and redirects most of it back onto Plasterboard 1
causing it to heat up faster.
4.3.4: Test Specimen 4
4.3.4 A): Construction Details
This specimen consisted of two 16 mm thick plasterboards attached to each other by
40 mm screws spaced at 300 mm centres along the periphery. Thermocouples were
positioned to measure the temperature profiles of the exposed surface i.e. the fire side
surface (FS), the interface between the two plasterboards (Pb1-Pb2), the unexposed
surface, i.e. the ambient side (amb), and also within the thickness of the plasterboards
as shown in Figure 4-18.
Fire Side (Exposed Surface)
Pb1 (16 mm thick)
Ambient Side (Unexposed Surface) Pb2 (16 mm thick)
Two thermocouples on the fire side at mid-height
Two thermocouples at 8 mm from the fire side and within Pb1 at mid-height
Two thermocouples at 16 mm from the fire side and in the interface of Pb1-Pb2 at mid-height
Two thermocouples at 24 mm from the fire side and within Pb2 at mid-height
Five thermocouples on the ambient side as shown in Figure 4-5
Figure 4-18: Instrumentation for Test Specimen 4
4.3.4 B): Observations, Results and Discussions
Test Specimen 4 was subjected to the fire test for approximately 222 minutes. The
behaviour of the specimen was very much similar to that of Test Specimen 3. After
intermittent evolution of smoke and steam, the ambient side of the specimen started to
discolour at the end of 78 minutes. The test was continued for some time even after
the burning of the ambient side paper. The specimen displayed a small amount of
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 92
lateral deformation in the outward direction. The test was finally stopped when most
of the ambient side paper started to peel and burn.
Figures 4-19 ((a) and (c)) show Test Specimen at the start and towards the end of the
test. The plateaus for 8, 16, 24 and 32 mm depth profiles in Figure 4-20 were seen to
extend up to 6, 23, 40 and 60 minutes, respectively. Plasterboard 2 showed much
extended periods of plateau. The interface is seen to increase the plateau on the
ambient side by approximately 28 minutes more than the expected duration of 32
minutes. Towards the end of the test a temperature difference of approximately 3200C
was noticed across Plasterboard 1 and approximately 6300C across Plasterboard 2.
Both plasterboards were intact up to the end of the test.
Figure 4-20 shows the temperature-time profiles for various depths across the
specimen cross-section while Figure 4-21 shows the temperature-depth profiles at
specific intervals of time.
As expected the 8 mm and 16 mm depth temperature profiles are seen to display
higher temperatures than the equivalent depth profiles of Test Specimen 2 at
corresponding times due to the heat redirected by the ambient side plasterboard.
(a) Test Specimen 4 at the Start of Test (b) Evolution of Smoke and Steam from Test Specimen 4
Figure 4-19: Fire Testing of Test Specimen 4
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 93
Clamps permitting vertical movement of the specimen
Uniform discolouration of paper
K type thermocouple wires
(c) Test Specimen 4 Showing Thermal Bowing at the End of Test
Figure 4-19: Fire Testing of Test Specimen 4
0
100
200300
400
500
600700
800
900
10001100
1200
1300
0 20 40 60 80 100 120 140 160 180 200 220Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS 8 mm 16 mm 24 mm Amb
Figure 4-20: Time-Temperature Profile of Test Specimen 4
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 94
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35
Depth (mm)
Tem
per
atu
re (
oC
)
30 min
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 95
60 min 90 min 120 min 180 min 210 min
Figure 4-21: Temperature-Depth Profiles of Test Specimen 4
4.3.5: Test Specimen 5
4.3.5 A): Construction Details
This specimen consisted of three plasterboards, each of 16 mm thickness, firmly
attached together using 50 mm long screws spaced at 300 mm centres along the edges.
Thermocouples were located on the surface as well as between the plasterboards to
measure the interface temperature. Eleven thermocouples were used to measure the
temperature profiles across the specimen as shown in the Figure 4-22.
Fire Side (Exposed Surface)
Pb2 (16 mm thick)
Pb1 (16 mm thick)
Ambient Side (Unexposed Surface) Pb3 (16 mm thick)
Two thermocouples on the fire side at mid-height
Two thermocouples at 16 mm from the fire side and in the interface of Pb1-Pb2 at mid-height
Two thermocouples at 32 mm from the fire side and in the interface of Pb2-Pb3 at mid-height
Five thermocouples on the ambient side as shown in Figure 4-5
Figure 4-22: Instrumentation for Test Specimen 5
4.3.5 B): Observations, Results and Discussions
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 96
ing regime for a period of slightly
over 3 hours. The test was stopped when the ambient side plasterboard paper started
temperature reaching 9000C
by the end of 155 minutes of test. At about 165 minutes Plasterboard 1 must have
Test Specimen 5 was subjected to the furnace heat
to burn. Figure 4-23 shows the time-temperature profiles across the specimen
thickness at various depths. The 16, 32 and 48 mm depth temperature profiles show
their second phases extending up to 23, 62 and 120 minutes, respectively. The
interfaces are seen to increase the plateau on the ambient side by approximately 64
minutes more than the expected duration of 48 minutes.
Plasterboard 1 was seen to heat up quite rapidly with its
partially or fully collapsed as the curve is seen to rise rapidly and merge with the fire
side (FS) curve. At the end of the test, the temperature across the Pb2-Pb3 interface
had reached 7500C and the unexposed surface had crossed 2000C.
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS 16 mm 32 mm Amb
Figure 4-23: Time-Temperature Profile of Test Specimen 5
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50Depth (mm)
Tem
per
atu
re (
oC
)
30 min 60 min 90 min 120 min 150 min 180 min
Figure 4-24: Temperature-Depth Profiles of Test Specimen 5
Figure 4-24 shows the temperature-depth profiles at different time intervals. With the
passage of time the curves are seen to become more and more linear up to 120
minutes. Beyond 150 minutes Plasterboard 1 starts to deteriorate very rapidly forcing
the portion of graph between 0 to 16 mm (representing Pb1) to become horizontal. At
180 minutes the initial portion of graph from 0 to 16 mm was horizontal, signifying
that Pb1 had collapsed and was no longer effective. The temperature drop from 16
mm to 32 mm and from 32 mm to 48 mm signifies the presence of Pb2 and Pb3 until
the end of the test.
The advantage of three layers of plasterboard over two layers is observed only during
the initial two hours of the test due to the extended plateau of the temperature profile
on the ambient surface of the specimen. After two hours the advantage starts reducing
rapidly and at around three hours both specimens are equivalent and display similar
thermal performance.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 97
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 98
4.3.6: Test Specimens 6, 7, 8 and 9
4.3.6 A): Construction Details
Test Specimen 6 consisted of a composite panel formed by sandwiching a layer of
insulation between the plasterboards. The insulation was laid in the cavity formed on
Pb1 by bordering it with plasterboard strips along the edges. This was achieved by
first fixing 16 mm plasterboard strips of 50 mm width along the periphery of the
board to form the cavity. The desired depth for the cavity was obtained by choosing
the appropriate thickness and number of plasterboard strips to be used along the
border. In the making of this test specimen a cavity of 32 mm was decided as it could
be easily provided using two 16 mm strips mounted on each other. The strips were
then fixed to the exposed plasterboard (Pb1) using 50 mm screws spaced at 300 mm
centres along the centreline of the frame.
Glass fibre insulation in the form of a mat was then cut to appropriate dimensions and
laid inside the cavity (see Figures 4-25 (a) and (b)). The mat used was 50 mm in
thickness and of 13.88 kg/m3 density. The mat was then compressed to a thickness of
32 mm (depth of the cavity) by firmly pressing it down with the help of a second layer
of 16 mm thick plasterboard (Pb2) which formed the ambient side of the test
specimen. This plasterboard was then screwed to the frame by 70 mm long screws
running along the periphery at 300 mm centres and staggered with the screws initially
applied to form the frame. The compressing of the 50 mm thick glass fibre mat to 32
mm thickness increased its density from 13.88 kg/m3 to 21.68 kg/m3 (ρ2 = ρ1 x t1/t2 =
13.88 x 50/32 = 21.68 kg/m3)
Thermocouples were installed during the construction process at various locations
across the thickness of the test specimen as shown in Figure 4-25. The wires were
carefully installed in the interfaces formed between plasterboard and insulation so as
to obtain the temperature profile on either side of the insulation. This would help in
understanding the behaviour of the insulation and its effectiveness at high
temperatures during the fire test. Eleven thermocouples were used to study the fire
performance of the test specimen.
(a) 50 mm Thick Glass Fibre Mats (b) Laying of Single Layer Glass Fibre Mat
Figure 4-25: Construction of Test Specimen 6
Fire Side (Exposed Surface)
Ambient Side (Unexposed Surface)
Insulation
16 mm Plasterboard (Pb2)
16 mm Plasterboard (Pb1)
Two thermocouples on the fire side at mid-height
Two thermocouples at the interface of Pb1-Ins at mid-height
Two thermocouples at the interface of Ins-Pb2 at mid-height
Five thermocouples on the ambient side as shown in Figure 4-5
Figure 4-26: Instrumentation for Test Specimens from 6 to 15
Test Specimen 7 was built in a manner similar to Test Specimen 6. However two
glass fibre mats each of 50 mm in thickness were laid in a cavity of depth 32 mm. The
glass fibre mats were held pressed down to a thickness of 32 mm by the use of
washers held down by screws passing through the base plasterboard (see Figure 4-27).
The compressed glass fibre mat was then further compressed by fixing the second
layer of plasterboard in a manner similar to the previous specimen. The compressing
of the glass fibre mats from a combined thickness of 100 mm to 32 mm increased its
density from 13.88 kg/m3 to 43.4 kg/m3. This test specimen was built to study the
effect of insulation density on the fire performance of the panel. Thermocouples were
installed on the interfaces and on the plasterboard surfaces to measure the temperature
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 99
profiles across the panel during the fire test as in Test Specimen 6. Eleven
thermocouples were used to study the fire performance of the panel.
Figure 4-27: Construction of Test Specimen 7
(a) Laying of Two Layers of 50 mm thick Glass Fibre Mats
(b) Glass Fibre Mats Compressed to 32 mm Thickness
Test Specimen 8 was built in a manner similar to Test Specimens 6 and 7. In this
specimen semi-rigid glass fibre mat of density 37 kg/m3 was used as insulation
material. The mat was 25 mm in thickness and easier to cut and lay. The cavity for
laying this insulation was formed by using two strips of 13 mm thick plasterboard as
border along the periphery giving a total depth of 26 mm. The construction of the
border frame was identical to that in Specimens 6 and 7. The glass fibre mat was then
cut to fit into the cavity as shown in Figure 4-28. This was followed by the fixing of
the ambient side plasterboard (also 16 mm in thickness as the exposed plasterboard).
Thermocouples were fixed in the same manner as Test Specimen 6 to record the
temperature variation across the body of the panel. Eleven thermocouples were used
to study the fire performance of the panel.
Test Specimen 9 was built similar to Test Specimens 6 to 8 with the only variation of
using a different thickness of glass fibre mat. This specimen was built using glass
fibre board of thickness 13 mm and density 168 kg/m3. The cavity was formed using a
single layer of 13 mm thick plasterboard strip to form the frame border with the
construction being similar to the previous specimens (see Figure 4-29). After the mat
was cut and placed inside the cavity, it was covered by the ambient side plasterboard
of 16 mm thickness and fixed to the frame to form the panel. Eleven thermocouples
were used to determine the temperature gradient across the panel. The instrumentation
was identical to that of Test Specimen 6.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 100
(a) Laying of Single Layer of 25 mm (b) Border Frame of two 13 mm Thick Thick Semi-Rigid Glass Fibre Mat Plasterboard Strips to facilitate the
attachment of the Ambient Side Plasterboard
Figure 4-28: Construction of Test Specimen 8
Washers to hold the insulation in place
13 mm thick Glass fibre mat
Plasterboard strip of thickness 13 mm and width 50 mm
Figure 4-29: Construction of Test Specimen 9
4.3.6 B): Observations, Results and Discussions
Test Specimens 6, 7, 8 and 9 were exposed to the standard time-temperature heating
regime for slightly over three hours. The initial behaviour of all the specimens was
similar to the previously tested specimens. All the specimens displayed a small
amount of thermal bowing in the outward direction towards the end of the test. The
ambient surface of these specimens showed uniform discolouration after about 110
minutes of testing (see Figures 4-30 (a), 4-36 (a) and 4-39 (b)). The tests were stopped
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 101
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 102
when the paper on the ambient surface started to burn. When the specimens were
inspected after the test, it was noted that the glass fibre insulation in all the specimens
had been almost completely consumed by heat with only small amounts still visible
along the edges of the specimens.
The time-temperature graphs show that the interface Pb1-Ins in all the specimens
showed a very rapid rise in temperature in the third phase crossing 6000C at about 35
minutes from the start of the test. The temperature profile of the interface (Pb1-Ins)
tended to become horizontal when its temperature approached 7000C. The glass fibre
insulation at this temperature began to disintegrate and lose its insulating properties as
could be seen from the temperature-depth profiles (see Figures 4-32, 4-35, 4-38 and 4-
41). The central portions of the graphs representing the insulation tended to become
horizontal from 90 minutes onwards for Test Specimens 6, 7 and 8 and 109 minutes
for Test Specimen 9 indicating that the insulation was no longer capable of bringing
about a temperature drop across its thickness.
As the heat energy was probably used up for disintegrating the glass fibre insulation,
the temperature on the ambient side of Pb1 (i.e. Pb1-Ins) did not rise. Also less heat
was getting redirected due to the continuous loss of insulation. These factors kept the
temperature on the ambient side of Pb1 steady (under 7000C) almost up to the end of
the test. The temperatures of the two interfaces Pb1-Ins and Ins-Pb2 are seen to merge
together soon after the disintegration of the glass fibre insulation due to direct
transmission of heat by radiation.
Regardless of insulation thickness and density, it was seen that the glass fibre
insulation became ineffective at about 7000C making the composite panels follow
similar temperature-time profiles up to the end of the test. In all these specimens Pb1
and Pb2 were found to remain intact until the end of the test.
(b) Glass Fibre Insulation Consumed by Heat
(a) Uniform Discolouration of Paper on Ambient Side
Figure 4-30: Fire Testing of Test Specimen 6
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-31: Time-Temperature Profile of Test Specimen 6
Note:
Pb1-Ins: Temperature profile of the interface between Pb1 (exposed plasterboard) and the insulation
Ins-Pb2: Temperature profile of the interface between the insulation and Pb2 (unexposed plasterboard)
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 103
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70Depth (mm)
Te
mp
era
ture
(oC
)
30 min 60 min 90 min 120 min 150 min 180 min
Figure 4-32: Temperature-Depth Profiles of Test Specimen 6
Figure 4-33: Test Specimen 7 Installed
in the Furnace for Testing
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 104
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-34: Time-Temperature Profile of Test Specimen 7
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70Depth (mm)
Te
mp
era
ture
(oC
)
30 min 60 min 90 min 120 min 150 min
Figure 4-35: Temperature-Depth Profiles of Test Specimen 7
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 105
(a) Uniform Discolouration of Paper (b) Glass Fibre Insulation is Totally
on Ambient Side Consumed Leaving a Dark Stain Behind
Figure 4-36: Fire Testing of Test Specimen 8
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-37: Time-Temperature Profile of Test Specimen 8
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 106
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55 60Depth (mm)
Te
mp
era
ture
(oC
)
30 min 60 min 90 min 120 min 150 min 180
Figure 4-38: Temperature-Depth Profiles of Test Specimen 8
(a) Test Specimen 9 Installed (b) Uniform Discolouration of Paper in the Furnace on the ambient Side for Testing of the Specimen
Figure 4-39: Fire Testing of Test Specimen 9
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 107
0
100200
300
400
500600
700
800
9001000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-40: Time-Temperature Profile of Test Specimen 9
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50Depth (mm)
Tem
per
atu
re (
oC
)
30 min 60 min 90 min 120 min 150 min 180 min
Figure 4- 41: Temperature-Depth Profiles of Test Specimen 9
The temperature development on the ambient side of the insulation in Test Specimens
6, 7, 8 and 9 has been shown in Table 4-2 over 10 minute intervals. Table 4-2 clearly
shows that the thickness, number of layers or the density of glass fibre insulation does
not significantly affect the temperature development of the Ins-Pb2 interface. In the
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 108
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 109
initial stages the Insulation in Test Specimen 9 was seen to perform slightly better
than the insulations in the other test specimens. However, this advantage was lost by
the time the fire side temperature of the insulation reached 7000C.
Table 4-2: Time–Temperature profile of the ambient side of the insulation (Ins-Pb2 interface) in Test Specimens 6, 7, 8 and 9 using glass fibre as insulation
material
Time
(min)
Test Specimen 6
T = 32 mm
P = 21.7 kg/m3
Test Specimen 7
T = 32 mm
P = 43.4 kg/m3
Test Specimen 8
T = 25 mm
P = 37 kg/m3
Test Specimen 9
T = 13 mm
P = 168 kg/m3
30 220 190 180 120
40 300 250 240 200
50 350 300 300 250
60 390 320 350 300
70 450 380 410 370
80 520 470 500 430
90 580 540 560 500
100 600 600 660 600
110 620 600 660 700
120 650 600 670 700
130 660 640 660 680
140 680 680 670 670
150 690 690 690 670
160 700 700 700 700
170 710 700 705 720
180 700 720 740
The temperatures on the ambient side of the insulation in all the test specimens (6 to
9) were seen to be in close comparison after the exposed surface of the insulation
crossed 7000C. Hence, for all practical purposes the thermal performance of the glass
fibre insulated composite panels can be assumed to remain unchanged regardless of
the thickness or density of the insulations used. However, the use of semi-rigid glass
fibre mats is recommended as the construction of Test Specimens 8 and 9 was much
easier than Test Specimens 6 and 7 due to the ease of handling of the insulation mats.
4.3.7: Test Specimens 10 and 11
4.3.7A): Construction Details
Test Specimen 10 was built using rock wool insulation of density 100 kg/m3 and
thickness 25 mm. The construction and instrumentation of this specimen was identical
to that of Test Specimen 8. Test Specimen 11 was built using 13 mm thick rock wool
insulation of density 114 kg/m3 (Figure 4.42). The construction and instrumentation
of this specimen was similar to Test Specimen 9.
Rockwool strips of width 225 mm and thickness 13 mm
Figure 4-42: Construction of Test Specimen 11
4.3.7 B): Observations, Results and Discussions
Both test specimens (10 and 11) were subjected to the fire test for nearly three hours.
Slight thermal bowing in the outward direction was noted in both specimens towards
the end of the test. Figures 4-44 and 4-46 show the time-temperature profiles across
the specimens at various depths.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 110
In Test Specimen 10 (Figure 4-44), the Pb1-Ins profile was seen to rise rapidly in the
third phase crossing 6000C by the end of 32 minutes, beyond which it flattened out,
with the temperature gradually increasing to 9000C by the end of 147 minutes.
Around this time Plasterboard 1 must have collapsed as the curve jumps rapidly to
merge with the FS curve. The Ins-Pb2 curve is seen to rise gradually up to 6000C by
the end of 147 minutes at which time the temperature is seen to rise sharply on
account of the collapse of Plasterboard 1. The profile of the interface Ins-Pb2
continued to maintain a temperature difference of over 2500C with the FS curve even
beyond 147 minutes implying that the insulation was still intact and functional. The
150 mm depth temperature profile in Figure 4-45 shows a horizontal segment from
exposed surface to 16 mm depth signifying the collapse of Plasterboard 1. Beyond 16
mm and up to 41 mm depth the temperature drop is brought about by the 25 mm layer
of insulation and beyond 41 mm up to 56 mm the drop in temperature is on account of
Plasterboard 2.
(a) Test Specimen 11 at the Start of Test (b) Thermal Bowing at the End of Test
(c) Rock Wool Insulation Intact even after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 111
Figure 4-43: Fire Testing of Test Specimen 11
In the case of Test Specimen 11 (Figure 4-46), the Pb1-Ins graph crossed 6000C by
about 45 minutes and then increased gradually to reach 9000C at about 174 minutes.
The graph was seen to rise sharply at this stage merging with the curve signifying the
collapse of Pb1. The graph of Ins-Pb2 maintained a profile well below that of Pb1-Ins
giving a minimum temperature difference of 2000C across the thickness of the
insulation up to the end of the test signifying the presence of the insulation until the
end of the test. This can be verified from the temperature-depth graph in Figure 4-47
where the 180 minute profile shows a horizontal segment from exposed surface to 16
mm depth signifying the collapse of Plasterboard 1. Beyond 16 mm and up to 29 mm
depth the temperature drop is brought about by the 13 mm layer of insulation and
beyond 29 mm up to 45 mm the drop in temperature is on account of Plasterboard 2.
Contrary to glass fibre insulation, the rock wool insulation showed greater resistance
to disintegration. The physical presence of the insulation was blocking and redirecting
the heat flow back to Plasterboard 1. This resulted in the rising of the temperature of
Pb1-Ins to values beyond 7000C and steadily kept rising up to 9000C when Pb1 started
to breach. Even after getting directly exposed to fire after the collapse of Plasterboard
1, the insulation remained intact and continued to offer protection to Plasterboard 2.
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-44: Time-Temperature Profile of Test Specimen 10
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 112
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55 60Depth (mm)
Te
mp
era
ture
(oC
)
30 min 60 min 90 min 120 min 150 min
Figure 4-45: Temperature-Depth Profiles of Test Specimen 10
0100
200300400
500600700800
90010001100
12001300
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-46: Time-Temperature Profile of Test Specimen 11
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 113
0100
200300
400500600
700800900
10001100
12001300
0 5 10 15 20 25 30 35 40 45 50Depth (mm)
Te
mp
era
ture
(oC
)
30 min 60 min 90 min 120 min 150 min 180 min
Figure 4-47: Temperature-Depth Profiles of Test Specimen 11
Table 4-3 shows the temperature profile of the ambient side of the insulation in Test
Specimens 10 and 11. The thermal performance of both insulations is seen to be
nearly the same in spite of the thicknesses being different. The temperatures were
seen to differ only after 145 minutes from the start of the test after the collapse of the
external plasterboard (Pb1) in Test Specimen 10. Temperatures shown in red indicate
the absence of Plasterboard 1 at that time.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 114
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 115
Table 4-3: Time–Temperature Profile of the Ambient Side of the Insulation (Ins-Pb2 interface) in Test Specimens 10 and 11 using Rock Fibre as Insulation
Material
Time (min)
Test Specimen 10 T = 25 mm
P = 100 kg/m3
Test Specimen 11 T = 13 mm
P = 114 kg/m3
30 160 180
40 240 230
50 290 300
60 310 340
70 330 400
80 390 480
90 450 520
100 500 520
110 520 540
120 540 550
130 550 580
140 580 590
150 700 600
160 850 605
170 880 650
180 860
4.3.8: Test Specimens 12, 13 and 14
4.3.8 A): Construction Details
Test Specimens 12, 13 and 14 were built using cellulose fibre as insulation. The
insulation was wet sprayed onto the plasterboards using a special wall nozzle (see
Figure 4-48). Insulation thickness and density were varied to study their effect on the
fire performance of the composite panels. Test Specimens 12, 13 and 14 were built
with an insulation thickness of 32 mm (density = 102 kg/m3), 25 mm (density = 108
kg/m3) and 20 mm (density = 131 kg/m3), respectively. Eleven thermocouples (6+5)
were used to study the temperature gradient across each composite panel.
Wall Nozzle
Wet cellulose spray
Figure 4-48: Construction of Test Specimens 12, 13 and 14
4.3.8 B): Observations, Results and Discussions
The tests for the cellulose fibre composite panels lasted slightly over two hours when
the tests were stopped following the burning of the paper on the ambient side of the
composite specimens. All the specimens displayed outward thermal bowing at the end
of the test (see Figure 4-49(d)). The paper on the ambient side started to discolour
after about 100 minutes. The discolouration in all three specimens was observed to be
non-uniform (see Figure 4-49 (b)) indicating the burning of cellulose fibre within the
specimen in certain areas creating pockets of high temperature. This allowed the heat
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 116
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 117
to penetrate the insulation compromising the integrity of the composite panel. The test
was stopped soon after the burning of the ambient side paper (see Figure 4-49(c)).
Closer inspection after the fire test revealed that in all the test specimens Plasterboard
1 had collapsed fully or partially and the cellulose fibre had completely burnt out at
the end of the test leaving behind traces of ash sticking to the fire side of Plasterboard
2 (see Figure 4-49(e)).
Figures 4-50 to 4-53 show the temperature profiles across Test Specimens 12 and 13,
respectively. The profiles of both the composite panels were observed to be almost
identical with the plateaus of the Pb1-Ins interface extending up to 18 minutes and
crossing 6000C at about 35 minutes. By the end of 120 minutes the ambient side
temperature of Plasterboard 1 (Pb1-Ins) in both the specimens had reached
approximately 9000C. The third phase of the Ins-Pb2 profile in both test specimens
started at about 36 minutes and by the end of 120 minutes of fire test a temperature
difference of approximately 2000C was recorded by the thermocouples across the
thickness of the insulation (i.e. the difference in the interface temperatures of Pb1-Ins
and Ins-Pb2) indicating the presence of insulation. The sudden increase in temperature
of the Pb1-Ins interface at 125 minutes for Test Specimen 12 and at 119 minutes for
Test Specimen 13 indicate the breaching of the exposed plasterboard. This was soon
followed by the burning of the paper on the ambient side of the composite panel and
the test was terminated.
The temperature profile of the ambient side as seen in both test specimens show the
plateau extending up to 90 minutes. Beyond 90 minutes the temperature (the average
of the five thermocouples mounted on the ambient side on the composite panel)
however was seen to climb quickly crossing 2000C by 120 minutes and 112 minutes
for Test Specimens 12 and 13, respectively. This rapid rise in temperature was
probably on account of the insulation burn out in certain areas creating pockets of
high temperature and allowing the heat to penetrate the composite panel (see Figures
4-50 and 4-52).
Figures 4-51 and 4-53 show the temperature-depth graphs for Test Specimens 12 and
13, respectively. The graphs in both test specimens are almost linear at 90 minutes.
Beyond this time deterioration in the insulation can be noted in Test Specimen 12 as
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 118
the profile tends to become horizontal in the central portion, although a drop of 2000C
across the insulation thickness can still be noted at 120 minutes. However in Test
Specimen 13, beyond 90 minutes, the deterioration of Plasterboard 1 seems to have
started along with the disintegration of the insulation as the initial portion of the
profile from 0 mm to 16 mm (representing Pb1) at 120 minutes has become almost
horizontal suggesting the cracking of the exposed plasterboard, whereas a temperature
drop of about 3000C can still be seen from 16 mm to 41 mm in the profile
(representing the 25 mm thick insulation) signifying the physical presence of the
insulation.
Test Specimen 14 was seen to deteriorate more rapidly when compared to Test
Specimens 12 and 13. Figure 4-54 shows the temperature profile across the composite
panel. The plateau for the Pb1-Ins interface was seen to extend up to 21 minutes
beyond which the temperature was seen to increase rapidly crossing 6000C by 40
minutes. Beyond 6000C the temperature rise was gradual reaching 8000C by about
100 minutes at which time the plasterboard appeared to have partially collapsed as the
temperature of the interface (Pb1-Ins) increased sharply merging with the fire side
curve at 120 minutes from the start.
The 60 minute profile in Figure 4-55 is almost linear beyond which the central portion
of the graph is seen to gradually flatten out indicating the disintegration of the
insulation in the composite panel. At 120 minutes the profile is seen to be horizontal
from 0 mm to 36 mm indicating the collapse of the exposed plasterboard and the
complete burnout of the insulation.
The temperature profile on the ambient side of the composite panel was seen to have
its plateau extending up to 80 minutes beyond which it started to increase quickly
crossing 2000C by about 112 minutes. The test was terminated following the burning
of the ambient side paper.
(a) Test Specimen 11 at the Start of Test (b) Non-Uniform Discolouration of Paper
on the Ambient Side
(c) Burning of Ambient Side Paper
Figure 4-49: Fire Testing of Test Specimen 12
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 119
(d) Thermal Bowing of Specimen (e) View Showing Collapse of Pb1 and
Disintegration of Cellulose Fibre Insulation
(f) Cellulose Fibre Sample Before and After the Fire Test
Figure 4-49: Fire Testing of Test Specimen 12
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 120
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-50: Time-Temperature Profile of Test Specimen 12
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70Depth (mm)
Tem
per
atu
re (
oC
)
30 min 60 min 90 min 120 min
Figure 4-51: Temperature-Depth Profiles of Test Specimen 12
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 121
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-52: Time-Temperature Profile of Test Specimen 13
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55 60
Depth (mm)
Tem
per
atu
re (
oC
)
30 min 60 min 90 min 120 min
Figure 4-53: Temperature-Depth Profiles of Test Specimen 13
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 122
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-54: Time-Temperature Profile of Test Specimen 14
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55
Depth (mm)
Tem
per
atu
re (
oC
)
30 min 60 min 90 min 120 min
Figure 4-55: Temperature-Depth Profiles of Test Specimen 14
In the case of Test Specimens 12 and 13 the 120 minute profile showed a fall in
temperature across the entire thickness of the composite panel signifying that both the
plasterboards and the insulation were still intact, whereas in the case of Test Specimen
14 the 120 minute profile was horizontal from 0 mm to 36 mm indicating the collapse
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 123
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 124
of the exposed plasterboard (Pb1) and the complete disintegration of the cellulose
fibre.
The temperature development on the ambient side of the insulation in Test Specimens
12, 13 and 14 is shown in Table 4-4
Table 4-4: Time–Temperature profile of the ambient side of the insulation (Ins-Pb2 interface) in Test Specimens 12, 13 and 14 using cellulose fibre as insulation material.
Time (min)
Test Specimen 12 T = 32 mm
P = 102 kg/m3
Test Specimen 13 T = 25 mm
P = 108 kg/m3
Test Specimen 14 T = 20 mm
P = 131 kg/m3
30 90 100 90
40 150 120 140
50 230 205 250
60 300 300 340
70 350 340 450
80 400 340 530
90 460 400 600
100 510 450 650
110 600 520 880
120 670 620 1000
The thermal performance of these specimens has been seen to differ with the change
in thickness and density of insulation, unlike the specimens built using glass fibre and
rock fibre. It is assumed that the influence of density is more dominant than the
influence of thickness in the case of cellulose fibres. As there is less control in
maintaining the density of the insulation layer over the entire interface it is likely that
certain areas burn faster than others leading to the formation of hot pockets leading to
an early insulation failure of the specimen.
4.3.9: Test Specimen 15
4.3.9 A): Construction Details
Test Specimen 15 was built using 25 mm thick Isowool insulation. The purpose of
using this high quality insulation, which is used as lining material in the interior of the
furnace, was to only compare the performance of other insulations in relation to
Isowool insulation. The construction and instrumentation methods were similar to
Test Specimen 8.
Figure 4-56: Construction of Test Specimen 15
4.3.9 B): Observations, Results and Discussions
Test Specimen 15 was subjected to the fire test for slightly over three hours. The
specimen displayed slight thermal bowing towards the end of the test. The ambient
side of the composite panel was seen to discolour uniformly from about 140 minutes
(see Figure 4-57).
Figure 4-58 shows the temperature profiles at various depths across the width of the
composite panel. The interface temperature (Pb1-Ins) had its plateau extending up to
21 minutes beyond which it was seen to increase sharply crossing 6000C at 35
minutes and reaching 7000C at 45 minutes. Beyond this point the temperature was
almost constant up to 80 minutes and then increased very gradually reaching 9000C at
160 minutes. The exposed plasterboard (Pb1) must have breached at this time as the
temperature of the interface (Pb1-Ins) increased suddenly merging with the fire side
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 125
curve. The temperature drop across the thickness of the insulation at this time was
about 4000C indicating that the insulation had maintained its integrity until the end of
the test. The ambient side temperature was seen to have its plateau extending up to
120 minutes beyond which it increased gradually crossing 2000C at about 170
minutes. The test was stopped following the burning of the ambient side paper.
(a) Test Specimen 15 at the Start of Test (b) Uniform Discolouration of Paper on the Ambient Side
Figure 4-57: Fire Testing of Test Specimen 15
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 FS Pb1-Ins Ins-Pb2 Amb
Figure 4-58: Time-Temperature Profile of Test Specimen 15
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 126
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55
Depth (mm)
Te
mp
era
ture
(oC
)
30 min 60 min 90 min 120 min 150 min 180 min
Figure 4-59: Temperature-Depth Profiles of Test Specimen 15
Figure 4-59 shows the temperature-depth profile of the specimen at intervals of 30
minutes. The 150 minute profile is seen to be linear whereas the 180 minute profile is
seen to have its initial portion of the curve (from 0 mm to 16 mm) horizontal
indicating the collapse of the exposed plasterboard (Pb1). The temperature of the
interface (Ins-Pb2) and the ambient side temperature of the test specimen are seen to
rise rapidly soon after the collapse of the external plasterboard.
Figure 4-60 shows the interface temperature (Ins-Pb2) of Test Specimen 15 in
comparison with that of other specimens using glass fibre, rock fibre and cellulose
fibre as the insulation material. Isowool insulation being capable of withstanding very
high temperatures is seen to perform better than all other types of insulation in the
initial stages of the test. However, the rising temperature of the interface (Pb1-Ins) on
account of the redirected heat forced the exposed plasterboard to heat up rapidly
leading to its collapse. The advantage gained by the use of superior insulation was lost
soon after the collapse of the external plasterboard (Pb1).
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 127
0
100
200
300
400
500
600
700
800
900
1000
1100
0 20 40 60 80 100 120 140 160 180
Time (min)
Te
mp
era
ture
(oC
)
Sp. 6 Sp. 7 Sp. 8 Sp. 9 Sp. 10
Sp. 11 Sp. 12 Sp. 13 Sp. 14 Sp. 15
Figure 4-60: Time-Temperature profiles for interface Ins-Pb2 of Test Specimens
6 to 15
4.4: Conclusions
This chapter has described the details of 15 fire tests on the thermal performance
of Plasterboards and Insulations. Following is a list of the main findings.
1) The time of exposure to the cellulosic fire curve determines the approximate depth
up to which the free and chemically bound water present in the gypsum plasterboard
gets expelled. On an average, 1 minute of fire exposure is required to expel water
from 1 mm thickness of plasterboard. Hence in the case of 13 mm thick plasterboard
exposed to standard time-temperature curve from one side, the temperature on the
ambient surface would be maintained at about 1000C up to 13 minutes and in the case
of 16 mm plasterboard it would be maintained for up to 16 minutes.
2) In spite of the numerous shrinkage cracks which are seen to develop over the
surface and within the body of the plasterboard due to the expulsion of free and
chemically bound water, the thermal gradient across the thickness of the plasterboard
when exposed to the standard time-temperature heating regime is seen to be
unaffected by the period of fire exposure and rising plasterboard temperature up to
about 900oC beyond which the plasterboard is seen to lose integrity probably because
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 128
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 129
the cracks get interconnected and reach up to the ambient surface allowing the
passage of heat.
3) Use of multiple boards introduces interfaces between adjacent plasterboards. These
interfaces between plasterboards are seen to help in improving the fire performance of
the wall system i.e. two plasterboards would give a better thermal performance than a
single plasterboard of double thickness. The interface leads to an improved thermal
performance by extending the duration of the plateau (second phase) of the
temperature profile on the ambient side of the specimen. The interface, however, does
not influence the linear variation of temperature across the specimen thickness after
the water is completely driven out from the plasterboards as transmission of heat
across the joint is very rapid on account of radiation.
4) Test Specimen 5 built using three layers of plasterboard is seen to perform better
than the Test Specimen 4 built using two layers only up to two hours from the start of
the test. Beyond two hours the external plasterboard of the triple layered specimen fell
off probably because of the bending of the heat softened screws under the dead weight
of the plasterboards making it equivalent to a double layered specimen. Hence it
appears that better fixing methods are needed if multiple layers (more than two) of
plasterboards are to be used.
5) Glass fibre insulation is seen to disintegrate at about 700oC. This resulted in similar
time-temperature profiles for all the composite panels using glass fibre insulation of
varying thickness and density.
6) In the case of Test Specimens 6, 7, 8 and 9 both plasterboards Pb1 and Pb2 were
found to be intact until the end of the test. This was probably because the temperature
of the Pb1-Ins interface did not climb beyond 7500C on account of the disintegration
of the glass fibre insulation from 7000C onwards. In the case of Test Specimens 10 to
15 the external plasterboard Pb1 fell off on account of the continuous build up of the
Pb1-Ins interface temperature reaching 9000C on account of the redirected heat from
the longer lasting insulation. Following the collapse of the external plasterboard Pb1
the temperature on the ambient side of the insulation and subsequently on the ambient
side of the test specimen was observed to build up rapidly.
7) Compared to Test Specimens using double plasterboards, the composite panels
using glass fibre insulation are seen to perform better as can be seen in Figure 4-61.
The average Ins-Pb2 interface temperature is used for the plotting of the temperature
profile for the specimens using glass fibre insulation (Test Specimens 6 to 9)
Figure 4-61: Average Time-Temperature profile for interface Ins-Pb2 of Test
Specimens 6 to 9 compared with Time-Temperature profile of Pb1-Pb2 interface
temperature of Test Specimen 4
8) Composite panels made of rock fibre insulation of varying density and thickness
also did not display any appreciable difference in their thermal performances although
the insulation lasted nearly up to the end of the test. Rockwool insulation showed
much greater resistance to disintegration when compared with glass fibre and
cellulose fibre insulation.
9) Fire resistance of cellulose insulation appears to depend upon its density. The
higher density cellulose fibre in Test Specimen 14 was totally burnt, whereas the
lower density cellulose fibre in Test Specimens 12 and 13 lasted longer time. The
distribution of cellulose fibre within a specimen could also be non-uniform as the
technique of spraying of wet cellulose fibre onto the plasterboard is not a standardized
process and can lead to areas of varying density within a specimen. This was probably
the cause of the burning of cellulose fibre within the individual specimens at different
rates giving rise to pockets of high temperature and thus lowering the integrity.
P.N.Kolarkar: Structural and Thermal Performance of Cold‐formed Steel Stud Wall Systems under Fire Conditions 130
Chapter 5: Thermal Performance of Non-Load Bearing Wall
Systems
5.1 Introduction
Fire safety of light gauge cold-formed steel frame (LSF) stud wall systems is critical
to the building design as their use has become increasingly popular in all areas of
construction throughout Australia. Partition wall panels composed of a cold-formed
steel frame lined with one or two plasterboards as side sheathing are being widely
used as they are very easy to assemble, thus improving the speed of construction. In
Australia, plasterboard lining manufacturers provide fire resistance ratings of non-
load bearing LSF stud wall systems. They have prescribed steel stud walls with single
or multiple plasterboard linings achieving fire resistance ratings, ranging from 60 to
120 minutes. These systems are based on full-scale fire resistance tests using the
standard fire curve recommended by ISO 834 and AS 1530.4 (SA, 2005). With
increasing demand for higher fire ratings of these walls, more than two layers of
plasterboard linings are being prescribed, which not only make the construction
process very laborious but also the resulting walls become very heavy.
Efforts have also been made to improve the fire ratings of the wall systems by using
different types of insulations in the wall cavities, but contradicting results were
obtained. Sultan and Lougheed (1994) performed several small scale fire resistant
tests on gypsum board clad steel wall assemblies (914 mm x 914 mm) using glass
fibres, rock fibres and cellulose fibres as cavity insulation. They noted that the rock
and cellulose fibre cavity insulations improved fire resistance rating by approximately
30 minutes when compared with non-insulated wall assemblies, whereas only a small
benefit was noted in the case of specimens using glass fibres. The cavity side of the
exposed gypsum board of insulated wall assemblies heated up more rapidly reaching
temperature levels of 7000C much earlier when compared to that in non-insulated wall
assemblies. Following the calcination of the exposed plasterboard, the exposed side of
the cavity recorded higher temperatures when compared to that in non-insulated wall
assemblies. Sultan (1995) carried out full scale fire resistance tests on non-load
bearing gypsum board wall assemblies and noted that when rock fibre was used as
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 131
cavity insulation the fire resistance rating increased by 54% over the non-insulated
wall assembly. Use of glass fibre as cavity insulation did not affect the fire
performance while cellulose fibre insulation reduced the fire resistance. Feng et al.
(2003) conducted fire tests on non-load bearing small scale wall systems and reported
that the thermal performance of wall panels improved with the use of cavity
insulation.
In summary, past research has produced contradicting results about the benefits of
cavity insulation to the fire rating of stud wall systems and hence further research is
needed. There is also a need to develop new wall systems with increased fire rating.
This chapter introduces a new wall system that uses a thin insulation layer between
two plasterboards on each side of stud wall frame instead of cavity insulation. It then
presents the details of a series of fire tests of non-load bearing (NLB) walls, examines
and compares their thermal performance, and makes suitable recommendations.
5.2 Test Specimens
Fire tests were conducted on nine small scale wall assemblies each measuring 1280
mm in width and 1015 mm in height. The wall assemblies typically consisted of three
commonly used cold-formed steel lipped channel section studs (90 x 40 x 15 mm)
spaced at 500 mm. The studs were fabricated from galvanized steel sheets (G500)
having a nominal base metal thickness of 1.15 mm and a minimum yield strength of
500 MPa. Test frames were built (see Figure 5-1) by attaching the studs to the top and
bottom tracks made of 1.15 mm G500 steel unlipped channel sections (92 x 50 mm)
using 12 mm long self-drilling wafer head screws. Test specimens were built by lining
the test frames with one or two layers of gypsum plasterboards manufactured by Boral
Plasterboard under the product name of FireSTOP. All the plasterboards used were
1280 mm in width and 1015 mm in height with a thickness of 16 mm and a mass of
13 kg/m2. The nine wall specimens built were divided into four categories as shown in
Table 5-1.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 132
Table 5-1: Details of non-load bearing wall specimens
Category Specimen
No. Configuration Objective
1 To study the effect of single layer of plasterboard (1x1) on both sides of frame on the fire rating of wall specimens
I
2
To study the effect of vertical joint in the exposed plasterboard over the central stud.
II 3
To study the effect of dual layers of plasterboard (2x2) on both sides of frame on the fire rating of wall specimens
4 To study the effect of glass fibre used as cavity insulation in a wall specimen with two layers of plasterboard (2x2).
5
To study the effect of rock fibre used as cavity insulation in a wall specimen with two layers of plasterboard (2x2).
III
6
To study the effect of cellulose fibre used as cavity insulation in a wall specimen with two layers of plasterboard (2x2).
7
To study the effect of glass fibre used as external insulation between the two layers of plasterboard on each side of a wall specimen with two layers of plasterboard (2x2)
8
To study the effect of rock fibre used as external insulation between the two layers of plasterboard on each side of a wall specimen with two layers of plasterboard (2x2)
IV
9
To study the effect of cellulose fibre used as external insulation between the two layers of plasterboard on each side of a wall specimen with two layers of plasterboard (2x2)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 133
1.15 mm G500 steel unlipped channel track: 92 x 50 mm
1.15 mm G500 steel lipped channel stud: 90 x 40 x 15 mm
Figure 5-1: Typical steel wall frame used to build NLB test wall specimens
5.3 Construction Details of Test Specimens
Test Specimen 1
Test steel frame shown in Figure 5-1 was lined on both sides by a single layer of
plasterboard (1x1 assembly) covering the entire frame without any joints (see Figure
5-2).
Thermocouple wires
Figure 5-2: Construction of Test Specimen 1
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 134
The plasterboards were attached to the three studs of the steel frame by 25 mm long
self-drilling bugle head screws at 300 mm centres and to the top and bottom tracks at
250 mm centres along the top and bottom edges of the board.
K type thermocouple wires were located on the steel frame, three on each stud at mid-
height to measure the temperatures of the hot flange (flange attached to the exposed
plasterboard), web, and the cold flange (flange attached to the ambient plasterboard).
These thermocouples allowed the determination of the average stud temperature and
the temperature gradient across the stud at mid-height. Additional thermocouples were
attached at the mid-height of the plasterboard to measure temperatures inside the wall
cavity and on the fire exposed surface (Figure 5-3 shows the locations of 15
thermocouples used across the wall assembly). To measure the temperature of the
ambient surface of the wall assembly, five more thermocouples were attached to the
unexposed surface of the plasterboard, one thermocouple at the centre of the wall and
one at the centre of each quarter section of the assembly giving a total of 20
thermocouples.
Pb1
Pb2
Figure 5-3: Thermocouple Locations for Test Specimen 1
Test Specimen 2
Construction of Test Specimen 2 was identical to that of Test Specimen 1, but with a
vertical joint in the exposed plasterboard located on the hot flange of the central stud.
A screw spacing of 200 mm on centres was adopted along each of the two
plasterboard edges forming the joint. The screw positions along both the edges were
staggered giving a screw spacing of 100 mm along the stud. The joint was taped and
covered by two applications of joint compound. The instrumentation was identical to
Test Specimen 1 (see Figure 5-4).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 135
Vertical Joint
Pb1
Pb2
Figure 5-4: Test Specimen 2 with a Joint in the Exposed Plasterboard over the Central Stud and Thermocouple Locations
Test Specimen 3
Test Specimen 3 was built by lining the steel frame with two layers of plasterboard on
either side (2x2 assembly). The base layer plasterboards were first attached to the
three studs by 25 mm long self-drilling bugle head screws at 300 mm centres. The
face layer plasterboards were then attached by 45 mm long self-drilling bugle head
screws spaced at 300 mm centres and penetrating the studs midway between the base
layer screws. Nineteen thermocouples were used to measure the temperature variation
across the mid-height of the wall assembly as shown in Figure 5-5. Five additional
thermocouples were used as in the previous specimens to measure the temperature of
the ambient surface of the wall assembly.
Thermocouple
Figure 5-5: Thermocouple Locations for Test Specimen 3
Test Specimen 4
This specimen was built similar to Test Specimen 3, but with the cavity filled with
two layers of 50 mm thick glass fibre mats of original density 13.88 kg/m3
compressed to 90 mm thickness (i.e. the depth of the cavity) giving the insulation a
density of 15.42 kg/m3 (ρ2 = ρ1 x t1/t2 = 13.88 x 100/90 = 15.42 kg/m3). Figure 5-6
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 136
shows the laying of the glass fibre mats in the cavity of the wall assembly between the
studs. Care was taken to pack the cavities of the individual studs and tracks with
insulation so as not to leave any air pockets in the wall cavity. Insulation was also
packed into the wall cavity beyond the end studs to establish conditions as close to the
central stud as possible. The instrumentation used was identical to that used for Test
Specimen 3 (see Figure 5-7).
(a) (b)
Non-load bearing wall specimen using glass fibre as cavity Insulation
packed into the wall cavity
beyond the end studs
Thermocouple wires
Steel platform to support the Test Specimen
(c)
Figure 5-6: Construction and Placement of Test Specimen 4 in the Furnace
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 137
Cavity Insulation
Figure 5-7: Thermocouple Locations for Test Specimens 4, 5 and 6
Test Specimen 5
Test Specimen 5 was built similar to Test Specimen 4, but with rock fibre of density
100 kg/m3 used as cavity insulation. Two mats each of 25 mm in thickness were
placed in the cavity of the wall. This left a gap of 40 mm between the insulation and
the cavity facing side of Plasterboard number three (Base layer plasterboard on the
ambient side) (see Figure 5-8). The thermocouple locations and numbers were
identical to that of Specimens 3 and 4 (see Figure 5-7).
Figure 5-8: Construction of Test Specimen 5
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 138
Test Specimen 6
Test Specimen 6 was built similar to Test Specimens 4 and 5, but with cellulose fibre
used as cavity insulation. The insulation was wet sprayed into the cavity until it was
filled completely (see Figure 5-9). The calculated density of cellulose insulation in the
cavity was 125 kg/m3. This was obtained by using the expression ρ = {[weight of Test
Specimen 6 (with insulation) - weight of Test Specimen 3 (without
insulation)]/volume of cavity} Instrumentation was the same as Specimens 3, 4 and 5.
Figure 5-9: Construction of Test Specimen 6
Test Specimen 7
In Test Specimen 7, a layer of 25 mm thick glass fibre insulation of density 37 kg/m3
was sandwiched between the two plasterboards, thus forming composite panels on
either side of the steel frame. The face plasterboard layer was attached through the
insulation layer to the base layer and the frame with 65 mm long drywall screws with
bugle heads, spaced at 300 mm centres along the studs and 250 mm centres along the
top and bottom edges connecting to the tracks. A total of 28 thermocouples including
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 139
five of them on the ambient surface were used to measure the temperature profiles
across the wall assembly (see Figure 5-10).
Figure 5-10: Thermocouple Locations for Test Specimens 7, 8 and 9
Test Specimen 8
Test Specimen 8 was built similar to Specimen 7, but with a 25 mm layer of rock fibre
of 100 kg/m3 used as insulation to form the composite panels. The methods of
construction and instrumentation used were the same as for Test Specimen 7.
Test Specimen 9
Test Specimen 9 was built similar to Specimens 7 and 8, but with a 25 mm layer of
cellulose fibre of density approximately 108 kg/m3 used as insulation to form the
composite panels. The fibre was wet sprayed on to the base plasterboard layer and
then covered by the face layer (see Figure 5-11). Sixty mm wide plasterboards were
used along the edges to enable the inclusion of cellulose insulation within a firmly
connected specimen. The instrumentation used was the same as for Test Specimen 7.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 140
Border frame built using two layers of plasterboard strips each of 60 mm width and thickness 13 mm
Specimen surface ready for wet spraying of cellulose fibre
Thermocouple wires
(a)
Facing Plasterboard on ambient side with thermocouple wires
Wet spray up of cellulose fibre
(b)
Figure 5-11: Construction of Test Specimen 9
The thermocouple wires in all the test specimens were taken directly to the outside of
the wall (ambient side) through small holes drilled in the unexposed plasterboard.
This was done so as to have minimum length of thermocouple wires within the wall
thus minimizing the possible contact of the lead wires forming hot junctions at
locations other than at those where the measurements are desired. The ends of each
wire were colour coded so as to be able differentiate between them once the wall
assembly was completed.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 141
5.4 Test Set-up and Procedure
The custom built adapter described in Chapter 4 was used to test the non-load bearing
specimens. Fire tests were carried out by exposing one face of the specimens to heat
in this propane-fired vertical furnace as shown in Figure 5-12. The specimens were
subjected to a heating profile, which followed the standard-time temperature fire
curve as given in AS 1530 Part 4 (SA, 2005).
Large Furnace
Vents on both sides of furnace
Gate to control exhaust opening
Adapter
Figure 5-12: Test Specimen placed in the specially built adapter in the large furnace
Similar to the procedure followed to test the plasterboard panels (as described in
chapter 4) the furnace temperature was measured using four thermocouples
symmetrically placed about the horizontal and vertical centre lines and the average
temperature of which was used by a software to control the furnace heat according to
the cellulosic fire curve (Standard time-temperature curve) given in AS 1530.4 (SA,
2005). Additional thermocouples were placed within the furnace to measure the
chamber temperature. The average temperature of these thermocouples was used as
furnace temperature for the plotting of graphs. The specimens were installed in the
furnace as shown in Figure 5-13.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 142
Test specimen with vertical edges free to deflect laterally
Gaps sealed by Isowool
Horizontal wooden beam to attach the LVDTs
(a)
Top clamps that allow vertical expansion of specimen
Pressure Transducer
Data Logger
LVDTs
(b) Figure 5-13: Test Specimen subjected to fire on one side
Specially designed clamps positioned at the top to hold the specimen in place allowed
the specimen to expand freely during the test. The vertical edges of the specimen were
kept free to allow lateral deformations. All the gaps and openings around the
specimen were sealed using Isowool. Three Linear Variable Differential Transducers
(LVDTs) mounted on a wooden beam were used to measure the mid-height lateral
deflections of the studs. One LVDT was positioned parallel to the specimen at the
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 143
centre to measure axial deformations. A pressure transducer was used to measure the
chamber pressure variations during the test. The failure of the small scale test
specimens was based on the integrity and insulation criteria in AS 1530.4 (SA, 2005).
The furnace and specimen temperatures were recorded using an automatic data-
acquisition system at intervals of one minute.
Based on AS 1530.4 (SA, 2005), the assembly was deemed to have failed if any one
of the following occurred first:
1. A single point temperature reading on the unexposed surface of the specimen
exceeded the ambient temperature by 180 0C;
2. The average of the five thermocouples on the unexposed surface of the specimen
exceeded the ambient temperature by 140 0C
3. Passage of flame or smoke for a minimum duration of 10 s through the unexposed
surface of the specimen.
5.5 Observations, Results and Discussion
5.5.1 Test Specimens 1 and 2
5.5.1.1 Visual Observations
Both specimens were exposed to the cellulosic fire curve in the furnace for slightly
more than three hours (Specimen 1-200 minutes and Specimen 2-190 minutes). When
the specimens were examined at the end of the test, it was noted that both the exposed
and ambient side plasterboards were severely affected in both specimens, but were
intact i.e. did not fall off during the test. The ambient surface of the unexposed
plasterboard in both specimens showed discolouration of paper and folds indicating
development of cracks on the cavity facing surface of the ambient plasterboard. All
the studs in both the frames were in good condition as seen in Figures 5-14 and 5-15.
The joint in the exposed plasterboard of Test Specimen 2 had opened up about 5 to 10
mm over the height of the stud. A small part of the joint at the bottom of the central
stud is visible in Figure 5-15.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 144
Cavity facing side of ambient
Figure 5-14: Test Specimen 1 after the fire test
Joint
Figure 5-15: Test Specimen 2 after the fire test
(Exposed plasterboard fell-off after the test during handling)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 145
5.5.1.2) Time-Temperature Profiles
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 Furnace FS HF WebCF Pb1-Cav Pb2-Cav Amb
Figure 5-16: Time-Temperature Profile for Test Specimen 1 (No joints in
plasterboard)
0
100200
300400
500600
700
800900
10001100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 Furnace FS HF Web
CF Pb1-Cav Pb2-Cav Amb
Figure 5-17: Time-Temperature Profile for Test Specimen 2 (With a joint in the
exposed plasterboard over the central stud)
Note:
AS 1530.4: Cellulosic fire curve (standard time-temperature curve) given by Australian Standard 1530 Part 4
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 146
Furnace: Average time-temperature curve followed by the furnace
FS: Average time-temperature profile of exposed surface of Plasterboard 1
HF: Average time-temperature profile of the hot flanges
Web: Average time-temperature profile of the webs
CF: Average time-temperature profile of the cold flanges
Pb1-Cav: Average time-temperature profile of the cavity facing surface of Plasterboard 1
Pb2-Cav: Average time-temperature profile of the cavity facing surface of Plasterboard 2
Amb: Average time-temperature profile of the unexposed surface of the wall
Figures 5-16 and 5-17 show the time-temperature profiles across Test Specimen 1 and
Test Specimen 2, respectively, when exposed to fire from one side. The steel stud
temperatures are seen to remain in a very narrow band in Specimen 1 (giving almost a
uniform temperature distribution across the cross-section) whereas a slight dispersion
is seen in Specimen 2. The central studs were critical in both the specimens as they
showed higher temperature than the end studs over the entire test. The temperature
rise of the studs was seen to occur in three phases. The first phase (which consisted of
the initial 7-8 minutes) showed a rapid gain in the stud temperatures up to about
1000C. Around this temperature the second phase started with the heating rate
decreasing and almost becoming zero due to the presence of moisture (free and
chemically bound water) in the gypsum plasterboard. The duration of this phase
depends on the time required to vaporize and drive away the moisture across the
thickness of the shielding plasterboard. The third phase (this was around 20 minutes
for both specimens) started soon after the evaporation of the moisture leading to a
rapid increase in the stud temperatures. Both specimens followed the same pattern
identically without showing any influence of the joint on the heating rates of the studs
as seen in Figures 5-18 to 5-20.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 147
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Te
mp
era
ture
(oC
)
Sp1-S1HF Sp1-S1CF Sp2-S1HF Sp2-S1CF
Figure 5-18: Time –Temperature Profiles of the Flanges in Stud No.1 of Test Specimens 1 and 2
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Tem
per
atu
re (
oC
)
Sp1-S2HF Sp1-S2CF Sp2-S2HF Sp2-S2CF
Figure 5-19: Time –Temperature Profiles of the Flanges in Stud No.2 of Test Specimens 1 and 2
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 148
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Te
mp
era
ture
(oC
)
Sp1-S3HF Sp1-S3CF Sp2-S3HF Sp2-S3CF
Figure 5-20: Time –Temperature Profiles of the Flanges in Stud No.3 of Test Specimens 1 and 2
Sp1-S1/2/3HF: Time-temperature profile followed by the hot flange
of Stud 1/2/3 in Specimen 1
Sp 1-S1/2/3CF: Time-temperature profile followed by the cold flange
of Stud 1/2/3 in Specimen 1
Sp2-S1/2/3HF: Time-temperature profile followed by the hot flange
of Stud 1/2/3 in Specimen 2
Sp 2-S1/2/3CF: Time-temperature profile followed by the cold flange
of Stud 1/2/3 in Specimen 2
The opening of the vertical plasterboard joint, caused by the shrinkage of the gypsum
plasterboard (calcination), appears to affect the central stud after the initial period
(time required for the weakening of the joint) of 70 minutes of fire exposure, as up to
that time the hot flange temperatures of both specimens were identical (see Figure 5-
19). A sharp increase in the temperature of the central stud in Test Specimen 2 is seen
with the deterioration and opening of the joint beyond this initial period (see Figure 5-
19 and Table 5.2). Table 5.2 compares the central stud temperatures of both the
specimens up to the end of the test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 149
Table 5-2: Central Stud Temperatures of Test Specimens 1 and 2
Test Specimen
Time
Temp.
30
(min)
60
(min)
90
(min)
120
(min)
150
(min)
180
(min)
HF 282 524 573 603 634 677
W 231 465 531 576 606 643
1
(Without
Joint)
CF 182 416 495 560 591 628
HF 263 504 683 847 930 957
W 212 439 553 648 720 782
2
(With
Joint)
CF 164 390 517 648 724 787
Note: HF – Hot Flange, W – Web, CF – Cold Flange
In spite of the increase in the central stud temperatures in Test Specimen 2, the
ambient side (plasterboard) temperatures of both specimens (i.e. with and without the
joint) remained almost identical up to 130 minutes from the start of the test (see
Figure 5-21)
0
25
50
75
100
125
150
175
200
225
250
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Tem
per
atu
re (
oC
)
Sp1-Amb Sp2-AmbCritical avr temp. 170 Insulation Failure Time (Sp-1) 89 minInsulation Failure Time (Sp-2) 92 min
Figure 5-21: Average Unexposed Surface Temperature of Test Specimens 1 and 2
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 150
Note:
Sp1/2-Amb: Average time-temperature profile on the ambient surface of specimen 1/2
This implies that although the joint had begun to open after the initial period (70
minutes) exposing the hot flange of the central stud in Test Specimen 2, the
plasterboards at the joint had not been detached from the screws. This allowed the
plasterboards to remain attached to the studs and prevented any sudden ingress of heat
from the furnace chamber into the wall cavity (see Figure 5-22).
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Tem
per
atu
re (
oC
)
Sp1 Pb2-Cav Sp2 Pb2-Cav
Figure 5-22: Time-Temperature Profiles of Cavity facing surfaces of Specimens 1
and 2
Note: Sp1/2 Pb2-Cav: Average time-temperature profile on the cavity facing surface of Plasterboard 2
5.5.1.3) Specimen Behavior
Lateral deflections of both specimens showed the same pattern with the end studs
deflecting less than the central stud. After the initial 20 minutes of protection offered
to the studs by the single layer of plasterboard, the hot flange temperatures of the
studs started rising sharply creating a temperature gradient across the cross-section of
the studs. This caused the studs to elongate more on the hot side than on the cold side
and forced the studs to bow towards the fire. The lateral deflection profiles of the
central stud in both specimens shown in Figure 5-23 are almost identical implying
little or no effect of the joint on the stud deformation. The lateral deflections, although
very small, were seen to be negative (i.e. towards the furnace) for both specimens
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 151
with a maximum of 2.75 mm occurring at approximately 45 minutes from the start of
the test.
-3.00
-2.50
-2.00
-1.50
-1.00
-0.50
0.00
0.50
0 10 20 30 40 50 60 70 80 90 100
Time (min)
De
fle
cti
on
s (
mm
)
Sp1-S2 Sp2-S2
Figure 5-23: Lateral Deflections of the Central Studs in Test Specimens 1 and 2
Note:Sp1/2-S2: Lateral deflection profile of stud 2 (central stud) in specimen 1/2
5.5.1.4) Wall Failure
Insulation failure (i.e. thermal failure) of Test Specimens 1 and 2 occurred at 89 and
92 minutes, respectively. At this time the average temperature of the unexposed
plasterboard surface of test specimens exceeded the ambient temperature of 300C by
140 0C.
5.5.1.5) Significance
A single layer of 16 mm FireSTOP (Boral) plasterboard on the fire side was seen to
offer an initial protection of around 20 minutes to the studs after which the stud
temperatures increased rapidly. Test Specimen 2 also showed consistent results. Test
Specimen 1, although built without joints in the plasterboard, did not show any
improvement in the fire rating when compared with the performance of Test
Specimen 2 built with a vertical central joint in the exposed plasterboard. Test
Specimen 2 had failed by insulation, before the effect of the joint, could be noticed on
its ambient surface. However, the vertical joint is likely to reduce the fire rating of
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 152
load bearing walls as the rapidly rising temperatures in the studs is likely to cause a
premature structural failure of the studs.
5.5.2 Test Specimens 3, 4, 5 and 6
5.5.2.1 Visual Observations
Test Specimen 3 (No cavity insulation), Test Specimen 4 (Glass fibre as cavity
insulation), Test Specimen 5 (Rock fibre as cavity insulation) and Test Specimen 6
(Cellulose fibre as cavity insulation) were subjected to heat in the furnace for slightly
more than 3 hours. Inspection soon after the test showed that the Plasterboards 1 and 2
(Fire side plasterboards) in Specimen 3 were still intact whereas they had partially
fallen off in the case of Specimens 4, 5 and 6. They fully collapsed due to their
extreme brittleness when they were removed from the furnace and placed on the
laboratory floor for inspection. Plasterboard 3 (base plasterboard on the ambient side)
in Specimen 3 was not as severely damaged as it was in the case of Specimens 4, 5
and 6. The cavity insulations of Specimens 4 and 6 were seen to be completely burnt
out with only traces of cellulose fibre ashes left behind in Specimen 6, whereas the
rock fibre insulation in Specimen 5 was still visible although it had completely lost its
integrity. Studs of Specimen 3 were seen to be in good condition whereas the studs in
the cavity insulated specimens were seen to be severely damaged, in particular the
ones in Specimen 6 using cellulose fibre as cavity insulation (see Figures 5-24 to 5-
27). The unexposed surface of all the specimens showed no signs of damage or effect
of temperature until the end of the test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 153
Studs in good condition
Exposed plasterboards were still intact at the end, but fell off during handling
Cavity facing surface of base layer plasterboard on ambient side.
Figure 5-24: Test Specimen 3 after the fire test (no cavity insulation)
Figure 5-25: Test Specimen 4 after the fire test (glass fibre cavity insulation)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 154
(a) Front view of specimen after fire test (b) Side view of specimen after fire test
(c): Central stud showing a missing (d): Damaged central stud flange in the middle portion
Figure 5-26: Test Specimen 5 after the fire test (rock fibre as cavity insulation)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 155
Central stud has totally disappeared in the middle portion.
(a) Side view of test specimen after fire test
(b) Badly deformed end stud showing (c) Central stud totally consumed separation of web from the flanges.
Figure 5-27: Test Specimen 6 after the fire test (cellulose as cavity insulation)
5.5.2.2 Time-Temperature Profiles
a) Plasterboard Surfaces: (Figures 5-28 to 5-30)
i) Average temperature of the interface surface between the exposed
Plasterboards 1 and 2 (Pb1-Pb2)
The temperature on this surface was seen to increase in all the specimens (in three
phases) from approximately four minutes after the test was started. In the initial
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 156
couple of minutes (first phase) all the specimens showed a sharp increase in the
temperature from ambient to around 800C beyond which the temperature did not
increase much (second phase) due to the hydration of the exposed plasterboard with
the temperature gradually increasing to 1200C by the end of 20 minutes. By this time
all the free and chemically bound water was probably driven out with the gypsum
board starting to dry and shrink developing and propagating fine shrinkage cracks.
Beyond 20 minutes (third phase) all the specimens showed a sharp increase in the
temperature of the interface (Pb1-Pb2), which continued until the end of the test,
except in the case of Specimen 3 where the temperature rise became gentler after
crossing 7400C at 90 minutes. By the end of 130 minutes the interface temperature in
Specimen 3 had reached 8000C whereas this temperature was reached 40 minutes
earlier by the cavity insulated specimens, ie. at about 90 minutes (95 minutes for
Specimen 6). This sustained rise in the temperature of the interface in cavity insulated
specimens was considered to be due to the heat being blocked and redirected back to
the exposed plasterboards by the insulation in the cavity.
The exposed plasterboard (Pb1) must have fallen off from Specimens 4, 5 and 6 after
about 132 minutes, 124 minutes and 137 minutes, respectively, as can be seen in the
sudden jump in temperature of the interface. In the case of Specimen 3 the increase in
temperature gradient at about 180 minutes probably indicates the loss of integrity of
Plasterboard 1. The temperature of the interface between Plasterboards 1 and 2 (Pb1-
Pb2) was seen to be between 9000C and 10000C in all the specimens when
Plasterboard 1 started to disintegrate.
ii) Average temperature on the cavity facing surface of the exposed Plasterboard
2 (Pb2-Cav)
An initial increase in temperature (Phase 1) was followed by a plateau (Phase 2)
extending up to 60 minutes in Specimen 3 and 55 minutes in the cavity insulated
specimens. The temperature was under 1200C at the end of the plateau. Pb2-Cav side
showed a much longer plateau than the one displayed by the ambient side of
Plasterboard 1 (Pb1-Pb2). This was because the fire side of Plasterboard 2 was
exposed to a fire curve defined by the interface temperature between Plasterboards 1
and 2, which was much less severe than the standard time-temperature curve
experienced by the fire side of Plasterboard 1.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 157
In Specimen 3 the plateau was followed by a gradual increase in temperature (Phase
3), reaching 4000C by the end of 115 minutes. Specimens 4 and 5 showed a much
more rapid rise in temperature from 55 to 70 minutes with the temperature crossing
4000C beyond which it became gentler (but much steeper when compared with
Specimen 3). In Specimen 4 the temperature reached 7000C in 124 minutes and
crossed 10000C by 170 minutes whereas in Specimen 5, it reached 7000C in about 120
minutes and crossed 10000C by 145 minutes (25 minutes earlier than Specimen 4).
Specimen 6 also showed a rapid increase in temperature from 55 to 75 minutes
reaching 4100C beyond which it continued to rise gradually reaching 7000C by the
end of 140 minutes. This was followed by a very rapid rise in temperature crossing
10000C by 155 minutes.
The temperature in Specimen 3 did not rise as rapidly as in the case of cavity
insulated specimens probably because in Specimen 3 the base layer plasterboard on
the fire side was allowed to lose heat via radiation in the empty cavity. The fast
passage of heat across the cavity and into Plasterboard 3 which served as a heat sink
to the fire side plasterboard checked its temperature escalation. In contrast, in
Specimens 4, 5 and 6, the insulation in the cavity due to its very low conductivity was
blocking the flow of heat and redirecting it back to the cavity facing surface of the
exposed plasterboard thus forcing a sharp and sustained rise in its surface temperature.
Fastest increase in temperature was noted in Specimen 5 with rock fibre insulation in
the cavity whereas the responses of Specimens 4 and 6 were nearly the same.
Specimens 5 and 6 show a rise in temperature gradient at about 124 minutes and 137
minutes, respectively, which coincides with the fall off times of Plasterboard 1 in
these specimens. The fall off times of Plasterboard 2 in the case of Specimens 4, 5 and
6 appears to be around 148 minutes, 145 minutes and 146 minutes as seen in the sharp
rise in the temperature recorded by the thermocouple on the cavity side of
Plasterboard 2 (Pb2-Cav). Plasterboard 2 in Specimen 3 was intact throughout the
test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 158
0100200300400500600700800900
100011001200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 Furnace FS Pb1-Pb2Pb2-Cav Pb3-Cav Pb3-Pb4 Amb
Figure 5-28: Time-Temperature Profiles of Plasterboard surfaces in Test Specimen 3
(No Cavity Insulation)
Figure 5-29: Time-Temperature Profiles of Plasterboard surfaces in Test Specimen 4
(Cavity Insulation – Glass Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 159
0100200300400500600700800900
100011001200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 Furnace FS Pb1-Pb2Pb2-Ins Ins-Pb3 Pb3-Pb4 Amb
Figure 5-30: Time-Temperature Profiles of Plasterboard Surfaces inTest Specimen 5
(Cavity Insulation-Rock Fibre)
0
100
200
300
400
500600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 Furnace FS Pb1-Pb2Pb2-Ins Ins-Pb3 Pb3-Pb4 Amb
Figure 5-31: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 6
(Cavity Insulation – Cellulose Fibre)
Note:
Pb1-Pb2: Average time-temperature profile of the interface between Plasterboards 1 and 2
Pb3-Pb4: Average time-temperature profile of the interface between Plasterboards 3 and 4
Pb2-Ins: Average time-temperature profile of the interface between Plasterboard 2 and insulation
Ins-Pb3: Average time-temperature profile of the interface between insulation and
Plasterboard 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 160
iii) Average temperature on the cavity facing surface of the ambient side
Plasterboard 3 (Pb3-Cav)
In Specimen 3 the initial rise in temperature of this surface almost coincided with the
rise in temperature of Pb2-Cav surface, probably due to the fast transmission of heat
via radiation across the cavity. The plateau extended up to 65 minutes (as compared to
60 minutes on Pb2-Cav surface) reaching a temperature of 1280C. The time-
temperature profile of Pb3-Cav surface followed the corresponding profile of Pb2-
Cav very closely but with a slight lag that never exceeded 700C.
In Specimen 4 the initial temperature rise on Pb3-cav surface started 6 minutes after
the initial rise on Pb2-Cav surface due to the protection offered by the insulation in
the cavity. The plateau was seen to extend up to 66 minutes (11 minutes longer than
that of Pb2-Cav surface). The plateau was followed by a gradual increase in
temperature up to 124 minutes reaching a temperature of 2800C when the temperature
on the Pb2-Cav surface had reached 7000C. The disintegration of the glass fibre
insulation must have started around this temperature (7000C) from across the cavity
reducing its density and thickness as the Pb3-Cav side started recording a very rapid
rise of temperature reaching 7850C by 153 minutes and crossing 10000C by 170
minutes. At this point of time the glass fibre insulation must be considered totally
useless as the time-temperature curve merged with that of the Pb2-Cav surface. Once
the disintegration of the glass fibre started at around 124 minutes, it took 29 minutes
for the insulation to become totally ineffective.
In Specimen 5 the temperature started rising with an 8 minute delay, reaching 760C by
around 75 minutes. From 75 minutes to 130 minutes the temperature rose gradually,
reaching 2260C by 130 minutes at which time the temperature across the insulation on
the Pb2-Cav surface was around 8500C developing a difference of 6240C across the
cavity. This temperature difference across the insulation was maintained fairly
constant up to 145 minutes, beyond which it started to converge gradually with the
decrease in the insulation integrity, crossing 10000C at 170 minutes before merging
with the time-temperature curve of Pb2-Cav. Disintegration of rock fibre insulation
started at about 145 minutes and took approximately 25 minutes to become totally
ineffective.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 161
In Specimen 6, the temperature started rising with a 9 minute delay and reached 980C
by about 94 minutes. This plateau was the longest when compared with 66 minutes of
Specimen 4 and 75 minutes of Specimen 5, indicating the superior initial insulating
properties of the cellulose fibre over the other two types. Beyond 94 minutes the
temperature was seen to rise gradually reaching 2150C at 110 minutes at which instant
the temperature across the cavity on the Pb2-Cav surface was 6300C giving a
temperature of 4150C across the cavity. A very sharp increase in temperature was
noticed beyond 145 minutes reaching 10000C in less than 10 minutes before merging
with the Pb2-Cav surface curve. The cellulose fibre insulation in the cavity of
Specimen 6 was intact for almost 145 minutes beyond which it disintegrated in just 10
minutes as against the 29 minutes taken by glass fibre insulation and 25 minutes taken
by the rock fibre insulation.
iv) Average temperature on the ambient side of unexposed Plasterboard 3
(Pb3-Pb4)
Specimen 3 showed a plateau up to 131 minutes beyond which it rose very gradually
to 2200C by 170 minutes. Specimen 4 showed a plateau up to 144 minutes beyond
which it rose very rapidly crossing 7000C by 187 minutes. Specimens 5 and 6 both
showed a plateau up to 160 minutes beyond which the graphs showed a sharp rise
crossing 7000C by 196 and 187 minutes, respectively. The gradual increase in the
temperature of Pb3-Pb4 surface in Specimen 3 was because the fire side plasterboards
were still intact without loosing their integrity whereas the rapid increase in
temperature in the cavity insulated specimens was probably due to the extreme
damage suffered by the fire side plasterboards due to their accelerated calcination
caused by the redirected heat from the cavity insulation. This led to the early collapse
of the fire side plasterboards and the subsequent quick disintegration of the cavity
insulation, leaving the ambient side plasterboards to face the full impact of fire from
the furnace.
v) Average temperature on the ambient side of unexposed Plasterboard 4
The average temperature on the ambient surface of the cavity insulated specimens was
only marginally lower than that recorded by Specimen 3 with no insulation up to
approximately 130 - 150 minutes beyond which the cavity insulated specimens
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 162
showed a sharper increase in temperature following the disintegration of half the wall
on the fire side and surpassed the ambient side temperature of Specimen 3.
b) Steel Surfaces (Figures 5-32 to 5-35)
The first phase involving the initial rapid temperature gain from ambient to about
1000C in the studs of all the specimens occurred in the first 20 minutes. This was
followed by the second phase wherein the temperature gradient became almost zero
up to about 60 minutes in Specimen 3 and 50 minutes in the case of Specimens 4, 5
and 6. The length of this plateau depended upon the time required to evaporate the
water contained in the two external plasterboards. Once this water had vaporized and
been driven out a rapid increase in the stud temperatures (Phase 3) was noted in all the
specimens until the end of the test.
The steel temperatures in the specimens depended upon the Pb2-Cav surface
temperature history. The studs of Specimen 3 picked up heat from this surface by
conduction (through the physical contact of the studs with the plasterboard surface),
convection (movement of hot air within the wall cavity) and by direct radiation. In
Specimens 4, 5 and 6 whose cavity was filled with insulation, the heat was passed on
to the studs from the plasterboard surface by conduction alone through steel and
insulation material. As the heat transfer via radiation is the fastest, the studs of
Specimen 3 had a more uniform temperature gradient across the studs as compared to
the large temperature variations across the stud cross-sections in the cavity insulated
wall specimens. The large temperature gradients across the cross-sections of studs in
cavity insulated specimens was due to the low conductivity of the insulation in the
cavity, which reduced the heat flow towards the cold flanges of the studs and
accelerated the temperature rise of the hot flanges due to the additional heat redirected
from the surface of insulation. This caused the hot flanges of the studs in cavity
insulated specimens to heat up more rapidly than those of Specimen 3. The
temperatures of the studs remained high over the entire test period leading to their
earlier damage. This is why severe damage and burn-out was observed in the studs as
shown in Figures 5-26 and 5-27.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 163
0
100
200
300
400
500
600
700
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Te
mp
era
ture
(oC
) S1-HF
S2-HF
S3-HF
S1-W
S2-W
S3-W
S2-CF
S3-CF
Figure 5-32: Time-Temperature Profiles across Studs in Test Specimen 3
(No Cavity Insulation)
Note:
S1/2/3-HF: Time-temperature profile followed by the hot flange of Stud 1/2/3
S1/2/3-W: Time-temperature profile followed by the web of Stud 1/2/3
S1/2/3-CF: Time-temperature profile followed by the cold flange of Stud 1/2/3
0
100200
300400
500600
700800
9001000
11001200
1300
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Tem
per
atu
re (
oC
)
S1-HF
S2-HF
S3-HF
S1-W
S2-W
S3-W
S1-CF
S2-CF
S3-CF
Figure 5-33: Time-Temperature Profiles across Studs in Test Specimen 4
(Cavity Insulation – Glass Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 164
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Tem
per
atu
re (
oC
)
S1-HF
S2-HF
S3-HF
S1-W
S2-W
S3-W
S1-CF
S2-CF
S3-CF
Figure 5-34: Time-Temperature Profiles across Studs in Test Specimen 5
(Cavity Insulation-Rock Fibre)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Tem
per
atu
re (
oC
)
S1 HF
S2 HF
S3 HF
S1 W
S2 W
S3 W
S1 CF
S2 CF
S3 CF
Figure 5-35: Time-Temperature Profiles across Studs in Test Specimen 6
(Cavity Insulation – Cellulose Fibre)
The central studs in all the specimens showed higher temperatures at any time than
the end studs. This was probably because the end studs could dissipate heat in to the
atmosphere faster as they were closer to the end of the walls and thus had lower
confinement when compared to the central stud.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 165
Table 5.3 shows the times taken by the hot flanges of the central studs of Specimens 3
to 6 to attain temperatures ranging from 4000C to 7000C.
Table 5.3: Hot Flange Temperature versus Time for the Central Stud
Time in Minutes Hot Flange Temperature
(oC) Specimen 3 Specimen 4
(Cav. Ins. GF) Specimen 5
(Cav. Ins. RF) Specimen 6
(Cav. Ins. CF)
400 100 78 74 91
500 144 91 82 106
600 218 107 97 125
700 119 115 139
The hot flange of Specimen 5 (with rock fibre as cavity insulation) heated up the
fastest whereas Specimen 6 (with cellulose fibre as cavity insulation) was the slowest
to heat up amongst the cavity insulated specimens. Specimen 3 with no insulation in
the cavity showed the best results with the studs remaining at temperatures much
lower than the cavity insulated specimens over the entire test period. The results of
Specimen 3 however could not be compared beyond 180 minutes as the furnace
heating regime in the case of Specimen 3 showed a sudden deviation from the
standard time-temperature curve profile due to some errors in the settings of the auto-
controlled valves regulating the flow of gas in the burner. The problem was fixed and
did not recur when Specimens 4, 5, and 6 were tested (see Figures 5-29 to 5-31).
In the cavity insulated specimens the sudden rise in the temperature of the studs was
observed within few minutes of the partial collapse of Plasterboard 1 and the severe
calcination and cracking of Plasterboard 2. As the exposed plasterboards were more
severely affected in the central portion, the middle stud was the first to show a sudden
increment in temperature. Studs 1 and 3 followed with a time delay of approximately
5 to 10 minutes. As the plasterboards of Specimen 3 were intact throughout the test,
the studs always had a gradual increase in temperature.
5.5.2.3 Entire Wall: Time-temperature graphs of Specimens 3 to 6 shown in Figures
5-36 to 5-39 display the temperature histories across the entire wall thickness with
plasterboard and steel taken together. Steel temperatures used in these figures are the
average temperatures of the three studs.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 166
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
FS Pb1-Pb2 Pb2-Cav H F Web
CF Pb3-Cav Pb3-Pb4 Amb
Figure 5-36: Time-Temperature Profiles across the Cross-section of Test Specimen 3
(No Cavity Insulation)
0
100200
300
400
500600
700
800
9001000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
FS Pb1-Pb2 Pb2-Cav HF WebCF Pb3-Cav Pb3-Pb4 Amb
Figure 5-37: Time-Temperature Profiles across the Cross-section of Test Specimen 4
(Cavity Insulation – Glass Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 167
0
100200
300
400
500600
700
800
9001000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
FS Pb1-Pb2 Pb2-Cav HF W
CF Pb3-Cav Pb3-Pb4 Amb
Figure 5-38: Time-Temperature Profiles across the Cross-section of Test Specimen 5
(Cavity Insulation – Rock Fibre)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Time (min)
Tem
per
atu
re (
oC
)
FS Pb1-Pb2 Pb2-Cav HF WebCF Pb3-Cav Pb3-Pb4 Amb
Figure 5-39: Time-Temperature Profiles across the Cross-section of Test Specimen 6
(Cavity Insulation – Cellulose Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 168
5.5.2.4 Behaviour of Specimens:
Figures 5-40 to 5-43 show the lateral deflection and axial deformation (A.D.) of studs
with time. Specimen 3 when exposed to the heating regime was observed to bow
away from the furnace reaching a maximum lateral deflection of 4 mm (in Stud 1) at
approximately 190 minutes from the start of the test. The axial deformations
(elongations) of studs were observed to start after 65 minutes of fire exposure. This is
consistent with the start of Phase 3, which is the steep increase in the stud
temperatures following the end of plateau in the time-temperature profiles of the
studs. The axial elongation steadily increased to reach a maximum of 4.2 mm by 200
minutes. Unlike Specimen 3 for which the temperature gradient of stud was small, the
cavity insulated specimens were seen to bow towards the furnace with the maximum
lateral deflections of the central studs being 3.7 mm, 5.5 mm and 5.6 mm, respectively
(negative deflection means towards the furnace). The deflections of the central studs
were seen to reverse sharply past these points in the cavity insulated specimens. This
was probably induced by the higher temperatures in the hot flanges leading to the loss
of their strength and stiffness faster than the cold flanges and thus undergoing local
and flexural buckling and/or deformations. For larger non-load-bearing walls this is
expected to occur earlier due to the greater slenderness of the studs and the larger self
weight of the walls. Specimen 3 also moved away from the furnace at the end of
testing for the same reasons given above for other specimens.
The axial elongations were measured only near the central stud for all the specimens.
In Specimens 4 and 5 the axial elongations were observed to start beyond two hours
of fire exposure reaching a maximum of 7.8 mm for Specimen 4 and 6.5 mm for
Specimen 5 in 170 and 167 minutes, respectively, whereas in Specimen 6 the
elongations started from approximately 80 minutes reaching 10.5 mm in 170 minutes.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 169
-1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Def
lect
ion
(m
m)
L.D. Stud 1 L.D. Stud 2 L.D. Stud 3 A.D.
Figure 5-40: Deflection-Time Profiles of Test Specimen 3
(No Cavity Insulation)
Note: L.D. Stud 1/2/3: Lateral-deflection time profile of Stud 1/2/3 A.D.: Axial deformation of studs
-4.00
-3.00
-2.00
-1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Def
lect
ion
(m
m)
L.D. Stud 1 L.D. Stud 2 L.D. Stud 3 A.D.
Figure 5-41: Deflection-Time Profiles of Test Specimen 4
(Cavity Insulation – Glass Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 170
-8.00
-6.00
-4.00
-2.00
0.00
2.00
4.00
6.00
8.00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Def
lect
ion
(m
m)
L.D. Stud 1 L.D. Stud 2 L.D. Stud 3 A.D.
Figure 5-42: Deflection-Time Profiles of Test Specimen 5
(Cavity Insulation – Rock Fibre)
-6.00
-4.00
-2.00
0.00
2.00
4.00
6.00
8.00
10.00
12.00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Def
lect
ion
(m
m)
L.D. Stud 1 L.D. Stud 2 L.D. Stud 3 A.D.
Figure 5-43: Deflection-Time Profiles of Test Specimen 6
(Cavity Insulation – Cellulose Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 171
5.5.2.5 Wall Failure
Table 5-4 shows the times at which the different portions of the wall were severely
affected contributing to the failure.
Table 5-4: Failure Times of Wall Components in minutes
Specimen Pb1:Fall
off time
Pb2:Time at partial/full collapse
Period of insulation failure
Local buckling of Hot Flange of central stud
4 132 148 124 to 150 125
5 124 145 145 to 170 145
6 137 146 145 to 155 145
All the specimens (3 to 6) were very stable with the ambient side temperature well
below the insulation failure temperature of 1650C (Ambient temperature was 250C)
throughout the test i.e. no insulation failure. If the reversal of the lateral deformations
of the studs could be considered as the failure of the steel frames caused due to the
softening and consequent local buckling of the hot flanges, then the failure times of
Specimens 4, 5 and 6 would be 125, 145 and 145 minutes, respectively. Specimen 3
showed no signs of failure until the end of the test.
5.5.2.6 Significance
1) Two layers of 16 mm FireSTOP gypsum plasterboard used in a 2x2 non-load
bearing wall construction were seen to offer around 60 minutes of initial protection to
the steel frames before losing the moisture content across both the plasterboards. This
period is more than twice that of a single board (20 minutes) as the second
plasterboard was subjected to a much less severe fire curve than the outer plasterboard
layer and hence took a longer time to expel the water held inside of it.
2) Heat transfer in the cavity of walls not filled with insulation took place via
conduction, convection and radiation. As a result of the faster transmission of heat
mostly through radiation, the temperatures across the stud cross-sections were
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 172
generally uniform, thereby resulting in minimum lateral deformations (ie. reduced
thermal bowing)
3) Use of cavity insulation was seen to be detrimental to the fire rating of walls. It not
only led to higher temperatures in the steel studs, but also to larger temperature
gradients across their depth (and increased thermal bowing effects).
4) Cavity insulated specimens were seen to bow initially towards the furnace. The
lateral deflections reversed sharply with the hot flanges of the studs undergoing local
buckling leading to the failure of the wall.
5) Among the three types of insulations used, rock fibre developed the maximum
temperature gradient across the studs whereas cellulose fibre developed the minimum.
The hot flange temperatures in the specimen using rock fibre insulation were more
than those of other specimens at any given time.
6) The heat trapped in the cavity by the insulation led to extensive stud damage in the
case of cavity insulated specimens. This was in contrast to the relatively good
condition of the studs in the non-insulated Specimen 3.
5.5.3 Test Specimens 7, 8 and 9
5.5.3.1 Visual Observations: (Figures 5-44 to 5-46)
The exposed Plasterboards 1 and 2 in all the three specimens fell off completely while
the insulation layer between the exposed plasterboards was totally consumed by the
furnace heat. Only small pieces of rock fibre insulation could be seen lying in the
debris whereas the other two insulations had completely vanished. The base layer
plasterboard on the ambient side (Pb3) had also collapsed in the central portion of all
the specimens. The ambient side plasterboard (Pb4), although cracked on the fire side,
was still intact and standing in one piece in all the specimens. The unexposed side of
the wall (ambient side of Plasterboard 4) showed no signs of any damage or
discolouration until the end of the test. Traces of glass fibre and cellulose fibre
insulation could be seen along the periphery between Pb3 and Pb4. In the case of
Specimen 8 the rock fibre insulation was found to be intact on one half of the
specimen. On the other half a portion of it had fallen off from the central area.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 173
The cold-formed steel frames were not twisted or bent. The upper and lower tracks
were in good condition. The central stud in all the three specimens was the most
affected. The central stud in Specimen 9 (using cellulose fibre insulation) showed the
maximum damage. The long screws used to hold the external exposed plasterboards
onto the steel frame were seen to be bent downwards, probably due to the weight of
the external plasterboards acting on the heat softened screws during the test. The
yielding of the screws also may have accelerated the collapse of the external
plasterboards.
Figure 5-44: Test Specimen 7 after the fire test (Glass fibre as external insulation)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 174
Figure 5-45: Test Specimen 8 after the fire test (Rock fibre as external insulation)
Figure 5-46: Test Specimen 9 after the fire test (Cellulose fibre as external insulation)
5.5.3.2 Time-Temperature Profiles
a) Plasterboard Surfaces: (Figures 5-47 to 5-49)
i) Average temperature of the interface surface between the exposed
Plasterboard 1 and Insulation (Pb1-Ins)
Similar to the cavity insulated specimens, the thermocouples on the ambient side of
Plasterboard 1 in all the specimens having external insulation used in the form of
composite panels responded in about three minutes showing a rapid rise in
temperature during the first phase. This phase displayed a sharp increase in
temperature reaching about 800C in less than 2 minutes, after which the temperature
increased gradually to 1200C by the end of 18 minutes. In the case of Specimen 9,
using cellulose fibre as composite insulation, the second phase lasted only 12 minutes
(6 minutes less than that observed in Specimens 7 and 8). The third phase in the case
of these specimens was much different than the cavity insulated specimens. Unlike the
cavity insulated specimens which displayed an almost constant temperature gradient
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 175
up to failure, the temperature gain was very rapid up to 34 minutes in Specimens 7
and 8 crossing 5500C (whereas the cavity insulated specimens recorded around 3000C
by the end of 35 minutes) and 23 minutes in Specimen 9 crossing 4500C, beyond
which the graph became very gentle with the temperature of the interface in
Specimens 7, 8 and 9 crossing 8000C by the end of 145 minutes, 115 minutes and 125
minutes, respectively. The initial steep rise in temperature in the 3rd phase was
considered to be due to the heat being blocked and redirected by the following layer
of insulation.
The fall off times (either partial or complete) of Plasterboard 1 in Specimens 7, 8 and
9 were considered to be 167 minutes, 145 minutes and 125 minutes, respectively, as
the temperature after these times showed a rapid rise (vertical), merging with the fire
side curve.
0
100200
300
400
500600
700
800
9001000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace FS Pb1-Ins Ins-Pb2
Pb2-Cav Pb3-Cav Pb3-Ins Ins-Pb4 Amb
Figure 5-47: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 7
(External Insulation-Glass Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 176
0
100200
300400
500600
700800
9001000
11001200
1300
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace FS Pb1-Ins Ins-Pb2
Pb2-Cav Pb3-Cav Pb3-Ins Ins-Pb4 Amb
Figure 5-48: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 8
(External Insulation-Rock Wool)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace Pb1-Ins Ins-Pb2 Pb2-CavPb3-Cav Pb3-Ins Ins-Pb4 Amb
Figure 5-49: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 9
(External Insulation-Cellulose Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 177
ii) Average temperature of the interface surface between the insulation and
exposed base layer Plasterboard 2 (Ins-Pb2)
This interface (in Specimens 7, 8 and 9) responded to the initial rise in temperature
(Phase 1) in under 4 minutes of exposure to furnace heat, reaching around 800C
rapidly and then remained almost constant (second phase) up to 22 minutes in
Specimens 7 and 8. In Specimen 9 however the second phase lasted only up to 15
minutes. The rate of temperature rise in the third phase was almost uniform in
Specimen 7 up to 85 minutes, beyond which it showed a sudden increase. This was
considered to be due to the rapid disintegration of the glass fibre insulation as the
temperature on the interface of Pb1- Ins had reached 7000C at this stage. From 85 to
95 minutes most of the glass fibre insulation would have burnt out. Beyond 95
minutes the rate of temperature rise became gentler and was constant up to 151
minutes during which time the insulation must have been totally consumed as the
graph of Ins-Pb2 surface merged with the graph of Pb1-Ins interface.
In Specimen 8, the rate of temperature rise was uniform up to 165 minutes beyond
which a sudden rise in temperature was noticed, probably due to the collapse of the
external Plasterboard 1 at about 145 minutes. A temperature difference of
approximately 2000C between the two sides of the insulation (i.e. between Pb1-Ins
and Ins-Pb2) indicates that the rock fibre insulation was still intact and functional
almost up to 180 minutes, beyond which it began to lose its integrity. Specimen 9
showed a change in the rate of temperature rise, becoming gentler after 23 minutes at
around 2300C. It continued with a uniform rate up to 125 minutes beyond which it
rose sharply, suggesting the collapse of both the external Plasterboard 1 and the
insulation disintegration.
iii) Average temperature on the cavity facing surface of the exposed Plasterboard
2 (Pb2-Cav)
The initial increase in temperature was followed by a plateau extending up to 84
minutes, 90 minutes and 70 minutes in Specimens 7, 8 and 9, respectively (compared
with 55 minutes observed in the cavity insulated specimens). This extended period for
the second phase must be due to the additional protection offered by the insulation
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 178
placed between the external plasterboards. Specimen 7 observed a more or less
uniform temperature gradient up to 173 minutes beyond which the gradient increased
as a result of the falling off of the external plasterboard and the disintegration of the
insulation. The sudden jump in the temperature at 198 minutes implies the falling off
of the base layer plasterboard (Pb2) on the fire side.
Specimen 8 also observed an almost uniform temperature gradient up to 200 minutes
beyond which it rose sharply suggesting the collapse of Plasterboard 2. Specimen 9
showed a change in gradient at about 131 minutes becoming steeper on account of the
collapse of the external plasterboard and cellulose fibre insulation at 125 minutes.
Beyond 163 minutes there was a further increase in the temperature gradient implying
a breach in the base layer Plasterboard 2, followed by its collapse.
iv) Average temperature on the cavity facing surface of the ambient Plasterboard
3 (Pb3-Cav)
In the absence of insulation, the transmission of heat across the cavity by radiation
was very quick, forcing the Pb3-Cav surface to heat up almost instantaneously and
trace very closely on the underside of the time-temperature profile of Pb2-Cav surface
with the maximum temperature difference between the two cavity surfaces being
290C, 500C and 580C in Specimens 7, 8 and 9, respectively.
v) Average temperature of the interface surface between base layer Plasterboard
3 and insulation (Pb3-Ins)
The thermocouples positioned in this surface started sensing a rise in temperature
from 12 minutes, 11 minutes and 18 minutes in Specimens 7, 8 and 9, respectively.
The temperatures climbed gradually reaching around 1000C and remained almost
constant up to approximately 130 minutes, 150 minutes and 135 minutes in
Specimens 7, 8 and 9 respectively, beyond which the third phase started with the
temperatures increasing rapidly. In Specimens 7, 8 and 9, a sharp increase in
temperature gradient was noted after about 200 minutes, 204 minutes and 168
minutes, respectively, indicating the possible breaching of Plasterboard 3 in all the
specimens. In Specimen 9, the change in gradient at about 168 minutes is not large
suggesting that cracks developed in Plasterboard 3 at this stage may not have been
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 179
very severe but progressed and widened rapidly leading to a collapse at about 184
minutes.
vi) Average temperature of the interface surface between the insulation and
Plasterboard 4 (Ins-Pb4)
The time-temperature graphs of this interface in the three specimens followed very
closely, but on the underside of the Pb3-Ins graph. The initial rise in temperature was
followed by a plateau which lasted up to 140 minutes, 175 minutes and 145 minutes
in Specimens 7, 8 and 9, respectively.
In Specimen 7, the plateau was followed by a gradual increase in temperature up to
200 minutes giving a temperature difference of 2000C on either face of the glass fibre
insulation. Beyond 200 minutes the insulation was seen to disintegrate very rapidly on
account of the steep increase in the temperature of the ambient side of Plasterboard 3
(fire side of insulation ). This led to a sharp rise in the temperature at about 204
minutes on the Ins-Pb4 interface.
In Specimen 8, the temperature rise following the plateau was very gentle up to 205
minutes, with a temperature difference 4000C being developed across the insulation.
After 210 minutes the rock fibre insulation must have lost its integrity as the
temperature of the Ins-Pb3 interface increased suddenly from 1870C to 7000C and
merged with the temperature profile of the ambient side of Plasterboard 3.
In Specimen 9, the constant temperature plateau lasted up to 145 minutes beyond
which the temperature started rising quickly. The cellulose fibre insulation must have
been intact up to 180 minutes as it maintained an almost stable temperature difference
of around 2000C between the ambient side of Plasterboard 3 and fire side of
Plasterboard 4. After about 185 minutes the temperature on the ambient side of
Plasterboard 3 crossed 7500C, leading to the burning up of the insulation. This led to a
quick rise in the temperature profile of the Ins-Pb4 interface merging it with the Pb3-
Ins profile implying the total disappearance of the insulation. The temperature-time
profile of the Ins-Pb4 interface in Specimen 8 with rock fibre insulation was the
lowest when compared with the interface profiles in Specimens 7 and 9. This clearly
indicates the superior insulating properties of rock fibre insulation over the glass fibre
and cellulose fibre insulations.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 180
vii) Average temperature on the ambient side of unexposed Plasterboard 4
The average temperature on the unexposed wall surface was below 1000C for all the
three specimens until the end of the test. This clearly indicates that the failure of these
wall specimens would be due to the structural failure of the steel frame rather than the
thermal insulation failure caused by the heat penetration through the wall.
b) Steel Surfaces (Figures 5-50 to 5-52)
The time-temperature graphs of the studs in Specimens 7, 8 and 9 showed the profiles
of the hot flanges, webs and the cold flanges to lie in a narrow band unlike the wider
profiles that were seen in the cavity insulated specimens. This was due to the quick
transmission of the heat across the cavity by radiation leading to a more uniform
temperature variation across the individual studs in each of the walls. The initial
temperature rise in all the specimens was over within the first 20 minutes. This was
followed by a period of almost constant temperature (plateau) up to 70 minutes for
Specimens 7 and 9. However, the plateau in the case of Specimen 8 using rock fibre
insulation extended up to 80 minutes.
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 20 40 60 80 100 120 140 160 180 200 220
Time (min)
Tem
per
atu
re (
oC
)
S1-HF
S2-HF
S3-HF
S1-W
S2-W
S3-W
S1-CF
S2-CF
S3-CF
Figure 5-50: Time-Temperature Profiles across Studs in Test Specimen 7
(External Insulation-Glass Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 181
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 20 40 60 80 100 120 140 160 180 200 220 240
Time (min)
Tem
per
atu
re (
oC
)S1-HF
S2-HF
S3-HF
S1-W
S2-W
S3-W
S1-CF
S2-CF
S3-CF
Figure 5-51: Time-Temperature Profiles across Studs in Test Specimen 8
(External Insulation-Rock Fibre)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 20 40 60 80 100 120 140 160 180 200
Time (min)
Te
mp
era
ture
(oC
)
S1-HF
S2-HF
S3-HF
S1-W
S2-W
S3-W
S1-CF
S2-CF
S3-CF
Figure 5-52: Time-Temperature Profiles across Studs in Test Specimen 9
(External Insulation-Cellulose Fibre)
In the third phase of the time-temperature graphs following the plateau, the stud
temperatures increased sharply at about 200 minutes for Specimens 7 and 8 at which
time Plasterboard 2 in these specimens fell off. In Specimen 9 the temperature
gradients became sharp at around 160 minutes which was consistent with the fall off
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 182
of Plasterboard 2 at 163 minutes. Akin to the cavity insulated specimens, the central
studs of the composite panel specimens recorded higher temperatures than the end
studs at all times. Table 5-5 shows the times taken by the hot flanges of the central
studs of Specimens 7, 8 and 9 to attain temperatures ranging from 4000C to 7000C.
The hot flange of Specimen 9 (using cellusic fibre (CF) as external insulation) heated
up the fastest whereas Specimen 8 (using rock fibre (RF) as external insulation) was
the slowest to heat up. Specimen 8 gave the best results as the rock fibre insulation
effectively maintained the steel temperatures lower than other specimens for a longer
period and displayed better insulating properties than cellulose and glass fibre.
Table 5-5: Hot Flange Temperature versus Time for the Central Stud
Time in Minutes Hot Flange Temperature
(0C) Specimen 7
(Insulation:GF) Specimen 8
(Insulation:RF) Specimen 9
(Insulation:CF)
400 117 142 122
500 148 160 132
600 159 178 140
700 175 189 151
5.5.3.3 Entire Wall: (Figures 5-53 to 5-55)
Time-temperature graphs of Specimens 7, 8 and 9 display the temperature histories
across the entire wall thickness with plasterboard and steel taken together. Steel
temperatures used are the average temperatures of the three studs. Time-temperature
graphs of Specimens 7, 8 and 9 display the temperature histories across the entire wall
widths with plasterboard and steel taken together. The hot flange temperatures of the
studs were seen to be higher than the cavity facing surface of Plasterboard 2 (Pb2-
Cav). This was probably caused by the close proximity of the hot end of the
thermocouples measuring the flange temperatures with the screws fixing the external
plasterboards to the studs. Due to the thermal bridging a small region of the flange
around the screw would be at a higher temperature, thus causing the thermocouples in
that region to record slightly higher temperatures than actual. On the other hand the
hot end of the thermocouple used for measuring the cold flange temperatures was
placed between the flange and the base layer plasterboard on the ambient side, thus
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 183
shielding it from direct radiation coming from Pb2-Cav surface. This probably caused
it to measure slightly lower values than the actual taking the profile slightly lower
than the profile of Pb3-Cav surface.
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220
Time (min)
Tem
per
atu
re (
oC
)
FS Pb1-Ins Ins-Pb2 Pb2-Cav HF WCF Pb3-Cav Pb3-Ins Ins-Pb4 Amb
Figure 5-53: Time-Temperature Profiles over the Entire Cross-section of Test Specimen 7
(External Insulation-Glass Fibre)
Figure 5-54: Time-Temperature Profiles over the Entire Cross-section of Test Specimen 8
(External Insulation-Rock Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 184
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
FS Pb1-Ins Ins-Pb2 Pb2-Cav HF WCF Pb3-Cav Pb3-Ins Ins-Pb4 Amb
Figure 5-55: Time-Temperature Profiles over the Entire Cross-section of Test Specimen 9
(External Insulation-Cellulose Fibre)
5.5.3.4 Behaviour of Specimens: (Figures 5-56 to 5-58)
Specimen 7 started exhibiting slight deformations after 2 hours of fire exposure. The
specimen was seen to bow away from the furnace with the lateral deformation
reaching its maximum value of 5.8 mm in 213 minutes. The axial deformation was
not significant up to 140 minutes, beyond which axial shortening was noticed
probably caused by the lateral deformations.
Specimen 8 showed no significant lateral or axial deformations throughout the test.
The maximum axial shortening was noted to be 2 mm at which time the maximum
lateral deformation of 1 mm was reached. The thermocouples were removed from the
wall specimen after 3 hours.
Specimen 9 displayed a delayed lateral bowing towards the furnace with the central
stud reaching a maximum of 6.2 mm at around 188 minutes, at which time the axial
deformation was around 9 mm.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 185
-6.00
-4.00
-2.00
0.00
2.00
4.00
6.00
8.00
10.00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220
Time (min)
Def
lect
ion
(m
m)
L.D. Stud 1 L.D. STUD 2 L.D. Stud 3 A.D.
Figure 5-56: Lateral Deflection -Time Profiles of Test Specimen 7
(External Insulation-Glass Fibre)
-6.00
-4.00
-2.00
0.00
2.00
4.00
6.00
8.00
10.00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Def
lect
ion
(m
m)
L.D. Stud 1 L.D. Stud 2 L.D. Stud 3 A.D.
Figure 5-57: Lateral Deflection -Time Profiles of Test Specimen 8
(External Insulation-Rock Fibre)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 186
-6.00
-4.00
-2.00
0.00
2.00
4.00
6.00
8.00
10.00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Def
lect
ion
(m
m)
L.D. Stud 1 L.D. Stud 2 L.D. Stud 3 A.D.
Figure 5-58: Lateral Deflection -Time Profiles of Test Specimen 9
(External Insulation-Cellulose Fibre)
5.5.3.5 Wall Failure
Table 5-6 shows the times at which the different portions of the wall were severely
affected contributing to the failure.
Table 5-6: Failure times of Wall Components in Minutes
Specimen Pb1: Fall off time
Period of insulation failure between Pb1 and Pb2
Pb2: Fall off time
Pb3: Fall off time
Period of insulation failure between Pb3 and Pb4
7 167 85-95 198 200 204-210
8 145 180-190 200 204 205-210
9 125 125 163 184 185-188
The ambient side of the plasterboard in Specimens 7, 8 and 9 showed no signs of
exceeding the insulation failure temperature during the entire course of the test. The
tests were discontinued after about 3 hours of exposure to the furnace heat. For all
practical purposes the failure of Plasterboard 2 would suggest the commencement of
wall failure as the steel stud temperatures would rise quickly leading to the failure of
the cold-formed steel frame.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 187
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 188
5.5.3.6 Significance:
1) Use of external insulation offered greater thermal protection to the studs resulting
in a near uniform temperature distribution across their cross-sections thereby
producing minimum early lateral deformation (thermal bowing).
2) The difference in temperature of the individual studs in the externally insulated
specimens was not significant as the radiation of heat in an open cavity is very fast
leading to a quick balance of temperatures in the individual studs. This would help in
reducing the building up of internal stresses in the frame caused by the unequal
expansions of the individual studs.
3) The wall can be considered to have failed when the studs reverse in lateral
deformation or when the external plasterboards collapse, whichever occurs first.
CHAPTER 6: STRUCTURAL AND THERMAL PERFORMANCE
OF LOAD BEARING WALL SYSTEMS
6.1: Introduction
A detailed experimental study was conducted in the Fire Research Laboratory of
Queensland University of Technology to evaluate the fire resistance of full scale, load
bearing steel stud wall assemblies. It included nine test specimens. One specimen was
tested at room temperature to determine its ultimate load bearing capacity while the
remaining eight specimens were exposed to the standard fire condition on one side
under a constant load to assess their fire performance.
This chapter presents the details of the experimental study into the thermal and
structural performance of the load bearing wall assemblies lined with single or dual
layers of plasterboard with or without cavity insulation. The insulations used were
glass, rock and cellulose fibres. Three new stud wall systems were built with the
insulation sandwiched between the plasterboards on both sides of the steel wall frame
instead of being placed in the cavity. Details of the results, including the temperature
and deflection profiles, measured during the tests are presented along with the stud
failure modes. Figures 6-1 (a) and (b) show the likely local and global stud failure
modes that could be encountered during the test.
Basic Section Flange Buckling
Web Buckling Distortional Buckling
Figure 6-1 (a): Basic Local Failure Modes
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 189
Flexural Buckling about Major Axis Flexural Buckling about Minor Axis
Torsional Buckling Lateral Torsional Buckling
Lateral Distortional Buckling
Figure 6-1 (b): Basic Global Failure Modes
6.2: Test Specimens
All the steel frames used in the large scale load bearing wall models were built to a
height of 2400 mm and a width of 2400 mm to represent a typical wall in a building.
All the studs and tracks used were fabricated from galvanized steel sheets having a
nominal base metal thickness of 1.15 mm and a minimum specified yield strength of
500 MPa. The frames were made of four vertical studs having 90 x 40 x 15 x 1.15 mm
lipped channel sections as shown in Figure 6-2. The studs were spaced at 600 mm
centres. Test frames were made by attaching the studs to the top and bottom tracks
made of 92 x 50 x 1.15 mm unlipped or plain channel sections using 12 mm long self
drilling wafer head screws.
The steel frames were lined on both sides by single or multiple layers of gypsum
plasterboards manufactured by Boral Plasterboard under the product name of Firestop.
The plasterboards supplied were 1200 mm in width by 2400 mm in length with a
thickness of 16 mm and mass of 13 kg/m2. The sheets were manufactured to the
requirements of Australian Standard AS/NZS 2588 – “Gypsum Plasterboard” (SA,
1998).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 190
(a)
Stud Wall Frame
(b) (c) Stud Section (External Dimensions) Track Section
Figure 6-2: Test Wall Frame
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 191
(a) Isometric View
Butt Joints between Plasterboards
Butt Joints between Plasterboards
(b) Plan View Figure 6-3: Stud to Plasterboard Connections
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 192
The plasterboards were installed vertically on both sides of the steel frame to build the
single layer wall models. The vertical butt joints were located over the centre line of
stud flanges for proper fixing. The boards were attached by 25 mm long drill point
screws which require much less effort than needle point screws thus reducing the
chance of stud distortion. These screws were spaced at 200 mm centres along the
plasterboard edges and 300 mm centres along the intermediate studs in the field of the
plasterboard as shown in Figure 6-3(a).
The butt joints between the plasterboards on one side were offset in relation to the
corresponding joints on the other side, with the offset being equal to a single stud
spacing of 600 mm (see Figure 6-3(b)). Plasterboards ‘c’ and ‘d’ were connected to
studs only along one edge whereas the other plasterboards had multiple stud supports.
A minimum edge distance of 15 mm was maintained for all the screws from the
plasterboard edges. For wall models requiring double layers, the second layer or the
outer layer consisted of plasterboard sheets installed horizontally with the joint at
mid-height of the wall. The outer layer plasterboards were attached by 45 mm long
self-drilling bugle head screws spaced at 300 mm centres in the field of the
plasterboard and penetrating the studs. The exposed screw heads were given two coats
of joint compound. The joints were sealed with 50 mm wide perforated chamfered
edge joint reinforced paper tape and covered with two coats of joint compound as
shown in Figure 6-4.
Figure 6-4: Protection of Joints
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 193
Table 6-1 gives an overview of the nine load bearing test wall specimens used in this
study.
Table 6-1: Details of Test Specimen Configuration
Test
No.
Configuration Test Insulation Objective
1 Ambient None To determine the ultimate load bearing capacity of specimen at ambient temperature
2 Fire None To study the fire performance of a 1x1 LBW
3
Fire None To study the fire performance of a 2x2 LBW
4
Fire Glass Fibre (Cavity
Insulation)
To study the fire performance of a 2x2 LBW with glass fibre as cavity insulation
5
Fire Rock Fibre (Cavity
Insulation)
To study the fire performance of a 2x2 LBW with rock fibre as cavity insulation
6
Fire Cellulose Fibre
(Cavity Insulation)
To study the fire performance of a 2x2 LBW with cellulose fibre as cavity insulation
7
Fire Glass Fibre (External
Insulation)
To study the fire performance of a 2x2 LBW with glass fibre as external insulation
8
Fire Rock Fibre (External
Insulation)
To study the fire performance of a 2x2 LBW with rock fibre as external insulation
9
Fire Cellulose Fibre
(External Insulation)
To study the fire performance of a 2x2 LBW with cellulose fibre as external insulation
Note:
LBW: Load Bearing Wall
1 x 1: Single layer of plasterboard on both sides
2 x 2: Two layers of plasterboard on both sides
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 194
6.3: Construction Details of Test Specimens
6.3.1: Test Specimen 1
The test steel frame shown in Figure 6-2(a) was lined on both sides by a single layer
of plasterboard (1 x 1 assembly) covering the frame as shown in Figure 6-3.
Thermocouple wires were not installed as the specimen was tested for its ultimate
axial compression capacity at ambient temperature.
6.3.2: Test Specimen 2
Construction of Test Specimen 2 was similar to the construction of Test Specimen 1
in all respects, except for the type of connection adopted at the top end of the studs. In
Test Specimen 2, a gap of 15 mm was left between the stud and the upper track as
shown in Figure 6-5.
16 mm Plasterboard
Figure 6-5: Stud to Track Connection at the Top
Screws were not used to connect the top end of the studs to the upper track. Instead
friction fit connections were adopted to allow for the vertical expansion of the studs
when exposed to elevated temperatures. Studs 2 and 4 had vertical plasterboard joints
on the fire side.
K type thermocouple wires were installed to measure the temperature variations
across the wall (over the plasterboard and steel surfaces) and along the stud lengths
(see Figure 6-6). Their locations on the wall are given in Section 6.4.6(B). The
thermocouple wires were attached to the hot flange, web and cold flange of the stud
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 195
by passing the hot junctions of the wire through small holes drilled in the steel
elements. These wire ends (hot junctions) were then pressed flat against the steel to
measure the surface temperature. The wires were then drawn to the ambient side
through tiny holes drilled in the unexposed plasterboard as shown in Figure 6-6(a). A
total of 50 thermocouple wires were installed, out of which 36 were used for
measuring the stud temperatures at three different levels (Figure 6-6b) and 14 were
used for measuring the temperatures on the plasterboard surfaces.
(a) Close up view of TC wires (b) View showing 36 TC wires attached attached to the flanges and web to the studs over 3 levels
(c) Fixing of Ambient Side Plasterboard
Figure 6-6: Construction of Test Specimen 2
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 196
6.3.3: Test Specimen 3
Test Specimen 3 was constructed with two layers of plasterboard on both sides of the
steel frame. The first layer was installed vertically while the second layer was
installed horizontally. Friction fit joints (similar to Test Specimen 2) were adopted at
the top end of the studs. Fifty six thermocouple wires were installed in the wall to
measure the temperature variations across the width of the wall and length of the
studs, when the wall specimen was exposed to furnace heat from one side. The
thermocouple wires were pulled across the width of the wall and onto the ambient
side through small holes drilled in the plasterboards. The holes in the second layer
(face layer) of the plasterboard were made to align exactly with the hole locations in
the first (base) layer so as to avoid any damage to the thermocouple wires (Figure 6-
7). Some of the thermocouple wires were bent with their hot junctions sandwiched
between the plasterboard surfaces to measure the interface temperature between the
plasterboards.
Thermocouple wires passed through aligned holes of the ambient side plasterboards
Face layer applied horizontally on ambient side
Base layer applied vertically on ambient side
Figure 6-7: Fixing of Face Plasterboard on the Ambient Side of Test Specimen 3
6.3.4: Test Specimen 4
The construction of Test Specimen 4 was very similar to that of Test Specimen 3. The
only difference was in the use of cavity insulation. After fixing the two plasterboards
on the fire side along with their associated thermocouples the cavity in the wall
between the studs was filled with two layers of 50 mm thick glass fibre mats (with an
original nominal density of 13.88 kg/m3) compressed to 90 mm thickness (depth of
the cavity) giving a density of 15.42 kg/m3. Figures 6-8(a) and (b) show the
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 197
installation of glass fibre mats. The cavities of the individual studs and tracks were
also packed with the same insulation to avoid the formation of air pockets within the
wall cavity. Figure 6-8(c) shows the temporary positioning of the base layer
plasterboard on the ambient side to facilitate the passing of thermocouple wires
through holes drilled in it at appropriate places. After fixing the base layer
plasterboard, the face layer plasterboard on the ambient side was fixed in a manner
similar to Test Specimen 3. Fifty six thermocouple wires were used to measure the
thermal response of the wall model when subjected to fire from one side.
(a) (b)
Laying of Glass Fibre Mats in the Wall Cavity
(c) Temporary Position of Base Layer Plasterboard on the Ambient Side for Passing Thermocouple Wires through Aligned Holes
Figure 6-8: Construction of Test Specimen 4 Using Glass Fibre as Cavity Insulation
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 198
6.3.5: Test Specimen 5
Construction of Test Specimen 5 was identical to that of Test Specimen 4 in all
respects, except for two layers of rock fibre each of 25 mm in thickness and density
100 kg/m3 being used as cavity insulation. Figure 6-9(a) shows the installation of rock
fibre mats in the wall cavity whereas Figure 6-9(b) shows the passing of thermocouple
wires through small holes drilled into the base layer plasterboard of the ambient side.
Once all the wires were passed through the holes the base layer was lowered (with the
thermocouple wires being gently pulled simultaneously) and fixed onto the steel
frame. The face layer on the ambient side was then positioned at right angles (applied
horizontally) over the base layer with the holes in the face layer aligned perfectly with
the pattern of holes in the base layer. The thermocouple wires were then passed
through the aligned holes of the face layer after which the face layer was lowered
gently and fixed onto the studs through the base layer.
(a) Laying of Rock Fibre Mats in the Wall Cavity
Figure 6-9: Construction of Test Specimen 5 using Rock Fibre as Cavity Insulation
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 199
(b) Passing of Thermocouple Wires through the Base Layer Plasterboard on the Ambient Side
Figure 6-9: Construction of Test Specimen 5 using Rock Fibre as Cavity Insulation
6.3.6: Test Specimen 6
Test Specimen 6 was built similar to Test Specimens 4 and 5, but with cellulose fibre
used as cavity insulation. After fixing the two layers of plasterboard on the fire side
along with their associated thermocouple wires the cavity was fine sprayed with plain
water to just moisten the cavity facing surface of the plasterboard. This was quickly
followed by a wet spray of cellulose fibre. Spraying was stopped after the complete
filling of the wall cavity with an approximate insulation density if 100 – 110 kg/m3.
The ambient side plasterboards were then subsequently laid and fixed to complete the
specimen construction.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 200
(a) Preparing the Cavity Surface (b) Spraying of Wet Cellulose Fibres For the Cellulose Spray into the Wall Cavity
(c) Cavity Filled Completely (d) Construction of Test Specimen 6 With Cellulose Fibres Completed with the Fixing of Ambient Side Plasterboard
Figure 6-10: Construction of Test Specimen 6 using Cellulose Fibres as Cavity Insulation
6.3.7: Test Specimen 7
The construction of Test Specimen 7 required the insulation to be laid not within the
cavity as in the previous specimens but outside the cavity (referred from here on as
external insulation) and between the base and face layer plasterboards on both sides of
the wall. To achieve this, the plasterboard layer on the fire side was attached along
with its associated thermocouples to the steel frame in a manner similar to the
previous specimens. This was followed by fixing the base layer plasterboard of the
ambient side along with its thermocouples to the steel frame thus closing the wall
cavity. Before fixing this plasterboard, all the thermocouples attached to the base
layer plasterboard on the fire side were carefully passed on to the ambient side
through small holes drilled at appropriate locations in the base layer plasterboard on
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 201
the ambient side. After fixing the base layer plasterboard and closing the wall cavity,
13 mm plasterboard strips of 60 mm width were fixed to the base layer plasterboards
on either side along the periphery of the wall to generate a cavity for external
insulation. To increase the depth of the cavity a second layer of plasterboard strips of
the same width and thickness was mounted on the previous strip giving a total cavity
depth of 26 mm. This was followed by the placing of a single layer of 25 mm thick
glass fibre mat of density 13.88 kg/m3 in the cavity formed on either side of the wall
along with additional thermocouples to measure the temperatures on either side of the
insulation during the fire test. Finally the face layer plasterboards were fixed
horizontally on either side of the wall (sandwiching the insulation between the face
and base layer plasterboards) while taking care to pass all the thermocouple wires
onto the ambient side of the wall through holes in the ambient side plasterboard.
(a) Thermocouple Wires being Passed (b) View Showing Plasterboard Strips through the Base Layer Plasterboard Attached along the Border of the Base layer with the Glass Fibre Mat Installed in Position
Figure 6-11: Construction of Test Specimen 7 using Glass Fibres as External Insulation
6.3.8: Test Specimen 8
The construction of Test Specimen 8 was identical to that of Test Specimen 7 in all
respects, except for a single layer of 25 mm thick rock fibre insulation of density 100
kg/m3 used as external insulation.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 202
(a) Single Layer of Rock Fibre Insulation Laid as External Insulation over the Base Layer
(b) Fixing of the Face Layer Plasterboard over the External Insulation
Figure 6-12: Construction of Test Specimen 8 using Rock Fibres as External Insulation
6.3.9: Test Specimen 9
The fixing of the base layer plasterboards to the steel frame and the strips along the
periphery to develop the cavity for external insulation was identical to that used in
Test Specimens 7 and 8. In this specimen 25 mm thick cubical spacers cut from
plasterboard strips were also positioned in the field of the cavity. This was done to
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 203
provide a firm support to the face layer plasterboard which was subsequently attached
after wet spraying the cavity with cellulose fibre insulation of density 100 – 110
kg/m3 (see Figures 13 (a) to (c)).
Spacer blocks
Wet spray of cellulose insulation
Border strips attached to base layer Plasterboard
(a) Spraying of Wet Cellulose as External Insulation on Ambient Side
(b) Spraying of Wet Cellulose (c) Fixing of the Face Layer Plasterboard as External Insulation on the Fire Side On the Fire Side
Figure 6-13: Construction of Test Specimen 9 using Cellulose Fibres as External Insulation
The outer layer plasterboards in Test Specimens 7, 8 and 9 were fixed to the steel
frame by 70 mm long plasterboard screws (with bugle heads) spaced at 300 mm
centres in the field of the plasterboard. As the screws used were not of the self drilling
type, it was necessary to pre-drill the studs before fixing the screws.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 204
6.4: Test Set-up and Procedure
6.4.1: Furnace
A propane fired gas furnace was specifically designed to carry out the tests on the
wall specimens. The furnace has internal dimensions of 2.1 m width, 0.3 m depth and
2.4 m height. The front face of the furnace was left open thus exposing all the burners
(see Figures 6-14 (a) and (b)). The furnace was mounted on a carriage so that it could
roll on wheels (Figure 6-14(c)).
To start the test the carriage was moved forward to make contact with the frame
holding the test wall specimen, thereby completing the combustion chamber. On
starting the furnace the wall was exposed to heat from one side as desired.
The gas burners are nozzle mixing units with a high velocity, spinning, air flow,
creating a negative vortex at the refractory block mouth. When gas enters the vortex it
mixes rapidly producing intense combustion. Also the inverted parabolic shape of the
burner block port works with the vortex and pulls the flame flat on to the furnace wall
(Figure 6-14d). This protects the wall from any localised flame impingement and
ensures a more uniform distribution of the temperature over the wall surface, mostly
by radiation.
Furnace Specifications
1) Structural steel furnace shell lined with ceramic fibre insulating material.
2) Six pyronics model SW 3 infrared flat flame gas burners including ignition
pilot burners.
3) 7.5 KW, 415-volt centrifugal combustion air fan complete with air distribution
manifold.
4) Gas control safety train and six individual control gas trains.
5) Control panel consisting of fan stop, start, programmable temperature control,
oven temperature control and burner controls (see Figure 6-14e).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 205
Gas Burner
Exhaust
Steel Shell
Ceramic Fibre Insulation
(a) Front View of Furnace before Ignition
(b) Front View of Furnace before and after Ignition
Figures 6-14: Details of Furnace Operation and Components
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 206
Specimen in Loading Frame Furnace
Control
Data Logger
(c) Side View Showing the Furnace on Wheels
(d) View of Radial Flame (e) Control Panel (Back View of Furnace)
Furnace
Thermocouple
Burner
Viewing Port 100mm x 200mm
450 mm
200mm
200 mm
(f) Rear wall of Furnace Housing the Thermocouples and Viewing Ports
Figures 6-14: Details of Furnace Operation and Components
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 207
Furnace Temperature:
The furnace was designed to deliver heat in accordance with AS 1530.4 (SA, 2005) as
given by the following equation
Tt – To = 345 log 10 (8t+1)
Where, t = Elapsed time in minutes
Tt = Furnace temperature (0C) at time t
To is the ambient temperature (0C) at the start of the test.
Furnace Pressure
The specimen holder containing the wall assembly was sealed against the furnace in
order to maintain the furnace pressure by at least 2 Pascal greater than the atmospheric
pressure over the top two thirds of the wall specimen. The positive pressure helped in
preventing the drawing of outside cold air into the combustion chamber.
Observation Ports
Four observation ports were provided on the rear side of the furnace as shown in
Figure 6-14(f) for observing the structural response of test walls.
Instrumentation Ports
Eight ports were provided on the rear wall to facilitate the introduction of
thermocouples to record the furnace temperatures during the test. Three ports were
also provided along the side (top, centre and bottom) to record the internal pressure
(see Figure 6-14(f)).
6.4.2: Compression Loading Frame
Loading Arrangement Used in Ambient and Fire Tests:
The loading frame was specially designed to load the individual studs of a wall
specimen in compression directly from the bottom side. It consisted of two columns
firmly bolted to the ground and a universal beam (UB) connecting the two columns to
form an ‘H’ shaped portal frame. A second universal beam was bolted to the floor.
Four jacks each of 45 kN capacity were mounted on this beam at a spacing of 600
mm. The shafts of the jack were co-axially guided through a hollow sleeve running
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 208
across a rectangular hollow section (RHS) attached between the two columns and
fixed parallel to the universal beam as shown in Figure 6-15.
Figure 6-15: Loading Frame
Test Specimen
RHS
Jack
Bottom Universal
Beam
(a) Loading of Each Stud in the Specimen using Jacks
Figure 6-16: Loading Arrangement
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 209
Bottom Track of wall
Gap to allow for stud elongation
Cement Plasterboard
Loading Plate
Sliding Plate
RHS
Jack
(b) Close up View of Loading Assembly
Figure 6-16: Loading Arrangement
The RHS supporting the hollow sleeve ensured a vertical movement of the jack.
Loading plates with collars to house the shafts coming out of the RHS were mounted
on top of each jack. The loading frame was built such that, when the test specimen
was mounted into the frame, the bottom track rested on the four loading plates and the
upper track was pressing against a cement board clamped firmly to the underside of
the top UB (the cement board was used to minimize the heat loss from the upper track
of the test specimen to the top UB). The test specimen was mounted in a manner to
ensure that the centroids of the studs aligned with the centroids of the loading plates.
This was made possible as the spacing of the jacks was identical to the spacing of the
studs. A special arrangement was also made using a sliding plate to facilitate
movement of any individual jack along with its shaft and sleeve to the extent of 20
mm on either side so as to achieve a better accuracy in lining up of the centroids of
the loading plate and the studs.
All the jacks were connected to a single hydraulic pump as the aim was to determine
the failure load of the wall specimen. A load cell was attached to the pump to obtain
directly the load being delivered to the studs by the jacks. The use of a single pump
ensured equal loading on all the studs as the same hydraulic pressure operated all the
jacks (see Figure 6-17).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 210
(a) Hydraulic Pump
connecting the Four Jacks
(b) Excess Load Release Valve
Figure 6-17: Hydraulic Pump and its Connections
To test the wall specimens subjected to fire, it was necessary to prevent the heat loss
from the bottom track into the loading plates. This was achieved by inserting cement
plasterboard pieces between the bottom track and the loading plate. The plasterboard
pieces were cut to the same size as the loading plates.
6.4.3: Ambient Temperature Test Procedure
Figure 6-18 shows the test specimen installed into the loading frame. The studs were
centred over the individual jacks and the wall was checked for its verticality using
spirit levels. After proper positioning of the wall, the top track was fastened to the top
beam using G - clamps on either side. This was done to retain the correct positioning
of the wall during testing. The studs were loaded initially to about 5 to 10 kN and then
unloaded. This was done twice to remove any residual strains and initial slackness
which may be present in the system during the assembly of the wall specimen. All
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 211
four studs were then reloaded simultaneously in increments of 2 kN. Load and
displacement readings were recorded by the Edcar software at the end of each load
increment. The specimen was assumed to have failed when the oil pressure in the
jacks could not be maintained. This was also confirmed via the Edcar load-
displacement graph which showed rapid load reductions (unloading).
Portal Frame
Wooden beam to mount LVDTs
LVDT
Figure 6-18: Test Set-up for Ambient Temperature Test
6.4.4: Instrumentation for Ambient Temperature Tests
To measure the axial shortening of the studs four Linear Variable Displacement
Transducers (LVDT) were used with each LVDT placed under the loading plate and
as close as possible to the stud as shown in Figures 6-19 (a).
Eight LVDTs were used to measure the out-of-plane movements of the wall
specimen. The transducers were attached to the wooden beams in front of the
specimen as shown in Figure 6-19(c) and were placed at 0.25H, 0.50H and 0.75H
along the height (H) of the two central studs and only at mid-height for the outer studs
as shown in Figure 6-18.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 212
Test Specimen
LVDT
JACK
(a) LVDTs Measuring Axial Shortening
(b) LVDTs (c) LVDTs Measuring Measuring Axial Shortening out-of-plane Deflection
Figure 6-19: LVDTs Used in the Measurement of Axial Shortening and Out-of-plane Deflection of Test Specimen Wall
6.4.5: Elevated Temperature Test Procedure
Test specimens were installed in the loading frame in the same manner as for the
ambient temperature test specimen. The furnace was then rolled forward towards the
wall specimen to close the gap between them and thus complete the combustion
chamber of the furnace with the wall forming the fourth side of the chamber facing
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 213
the burners. This arrangement ensured that only one face of the test specimen was
exposed to elevated temperatures at the start of testing. The width of the wall was
designed to be less than the width of the furnace opening by 20 mm such that there
would be a gap of 10 mm on either side of the wall, to ensure free vertical edges. The
gap was then packed with Iso wool, a non-restraining and non-combustible mineral
fibre such that the lateral displacement of the wall was not restricted due to frictional
forces. An axial compression load of 15 kN was applied gradually to each stud at a
constant rate by the hydraulic jacks. This load was based on a load ratio of 0.2, i.e. 0.2
times the ultimate capacity of each stud at ambient temperature obtained from Test 1.
The load was held constant at room temperature for about 45 minutes before the
furnace was started. The load of 15 kN per stud was maintained throughout the fire
endurance test. This allowed free vertical expansion of the wall when exposed to
elevated temperatures. During the fire test, the furnace temperature was regulated
such that the average temperature recorded by the control thermocouples inside the
furnace followed the standard cellulosic temperature-time curve in accordance with
AS 1530.4 (SA, 2005). During the fire test the vertical and lateral displacements of
the wall, the temperature readings from all the thermocouples and the furnace pressure
readings were taken at intervals of 1 minute. The test was stopped immediately
following the failure of the wall. The time to failure was then recorded.
6.4.6: Instrumentation for Elevated Temperature Tests
A) To measure displacements
To measure the axial shortening of the studs the location of the transducers was
identical to that adopted for the ambient temperature test. For the out-of-plane
movements of the wall specimen the transducers were placed at 0.25H, 0.50H and
0.75H along the height (H) of all the four studs.
B) To measure temperatures
K type thermocouples were used to measure the temperature development across the
wall specimens. The stud temperatures were measured at three levels, namely at 0.25
H, 0.50 H and 0.75 H (where ‘H’ is the height of the Test Specimen).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 214
Fire Exposed Side
Unexposed Side
(a) Thermocouple Locations for 1x1 LBW Specimen
Fire Exposed Side
Unexposed Side
(b) Thermocouple Locations for 2x2 LBW Specimens with and without Cavity Insulation
Fire Exposed Side
Unexposed Side
(c) Thermocouple Locations for 2x2 Wall Specimens with External Insulation
Studs
Horizontal joint of face layer plasterboard
Thermocouples
(d) Positions of Thermocouples to Measure the Average Temperature Rise on the Ambient Surface of the Test Specimen
Figure 6-20: Thermocouple Locations for Load Bearing Wall Specimens
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 215
At each level three thermocouples were attached per stud to measure the temperatures
of the hot flange, web and cold flange thus giving a total of nine thermocouples per
stud (over the three levels) and 36 thermocouples per frame (as the frame is made up
of four studs). These thermocouples allowed the determination of average stud
temperature, temperature gradient across the stud cross-section and also along the stud
length.
Additional thermocouples were attached at the mid-height of the assembly between
the studs to measure the temperatures on the plasterboard surfaces facing the cavity
and also on the exposed face of the wall to measure the temperature of the
plasterboard surface subjected to fire thus giving a total of 45 thermocouples (36 + 9)
for a 1x1 LBW Specimen as shown in Figure 6-20 (a).
For 2x2 LBW Specimens with or without cavity insulation, additional thermocouples
were installed between the plasterboard surfaces on either side at mid-height as shown
in Figure 6-20 (b) giving a total of 51 thermocouples (36 + 9 + 6) to measure the
temperature variations across the test specimen. For 2x2 LBW Specimens using
external insulation six more thermocouples were installed to measure the temperature
across the insulation layers at mid-height thus giving a total of 57 thermocouples (36
+ 9 + 6 + 6) as shown in Figure 6-20 (c)
To measure the average temperature on the unexposed face of the wall, five
thermocouples were positioned on the unexposed face, 1 at the centre of the area and
one at the centre of each quarter section as mentioned in AS 1530.4 (SA, 2005). The
temperature measured by these thermocouples indicated the heat penetration across
the specimens (see Figure 6-20 (d)).
To measure the temperature at various other points on the ambient face an infrared
gun was used (Figure 6-21). Figure 6-22 shows the complete instrumentation of a test
specimen before the fire test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 216
Figure 6-21: Infrared Gun Used for the Measurement of Ambient Side Temperatures
.
Isowool Insulation along the periphery to close the gap between the specimen and the furnace
LVDTs
Pressure Transducer
EDCAR Data
Logger Thermocouple wires
Jacks
Figure 6-22: Test Specimen Complete with all its Instrumentation Ready for Fire Test
Eight K – type furnace thermocouples were symmetrically placed inside the furnace
chamber within a vertical plane 100 mm from the exposed surface of the specimen to
record the temperature of the furnace (see Figure 6.14 (f)).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 217
6.5: Observations and Results
6.5.1: Test Specimen 1
6.5.1.1) Visual Observations and Specimen Behaviour
The wall specimen showed no visible signs of deformation up to a load of 52 kN.
Beyond this load the web elements of Studs 1 and 4 which were the only visible parts
of the steel frame showed local buckling waves developing in the web portion. This
local deformation progressed with increasing load. When the load approached 79
kN/stud there was a sharp noise and the specimen failed. The load dropped rapidly as
the oil pressure could not be maintained. Lateral movement of the wall specimen was
not visible at any stage of the test. The failure was due to local buckling of the flange
and web elements at the base as shown in Figure 6-23.
(a) Local Buckling at the base of Stud 1 (b) Local Buckling at the base of Stud 4
Figure 6-23: Failure of Test Specimen 1
The gypsum plasterboards showed no damage and were seen to be successful in
effectively restraining the studs from torsional buckling and flexural buckling about
the minor axis. The screws connecting the plasterboards to the studs were seen to have
pulled through the plasterboard at the base close to the locally buckled stud sections.
Axial deformations in the range of 13 mm to 16 mm were noted in the studs as seen in
Figure 6-24.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 218
0
10
20
30
40
50
60
70
80
90
0 2 4 6 8 10 12 14 16 18
Axial Shortening (mm)
Lo
ad (
kN)
Stud 1 Stud 2 Stud 3 Stud 4
Figure 6-24: Load Vs Axial Deformation - Profiles of Test Specimen 1 at Ambient Temperature
The wall failed at a load of 79 kN/stud (total load of 316 kN on the frame) by the local
crushing of the stud channels at the base near the loading plates.
6.5.2: Test Specimen 2 (1x1 LBW without insulation)
6.5.2.1) Visual Observations and Specimen Behaviour
The specimen was subjected to the standard time-temperature heating regime in the
furnace. During the test the Edcar software crashed for a period of 8 minutes from 9 to
17 minutes from the start of the test resulting in the loss of readings. Also it was
uncertain whether the applied load of 15 kN was maintained after 9 minutes of test.
The temperature time graphs could be plotted accurately. However the portion of the
graphs between 9th minute to 17th minute was unavailable due to the failure of the
software.
The specimen showed no signs of lateral displacement during the initial application of
the compression of 15 kN load. After 3 minutes of starting the furnace smoke was
seen coming out from the top of the wall specimen (Figure 6-25(a)). This was
probably due to the burning of the plasterboard paper on the exposed surface. At the
end of 11 minutes thick smoke and steam were seen to escape from the outer edges
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 219
from the top of the wall. The presence of steam in the mixture of escaping gases was
confirmed by the heavy condensation of steam into water on the bottom flange and
web of the top UB of the loading frame (see Figures 6-25(b) and (c)).
(a) Smoke and Steam Escaping from the Top Side of Test Specimen 2
(b) Condensation of Steam (c) Condensation of Steam on the On the Bottom Side of Top UB Web of the Top UB
Figure 6-25: Fire Performance Test of Specimen 2
By 32 minutes the lateral displacement or bowing of the wall towards the furnace was
prominently noticeable. At the end of 40 minutes soft crackling sounds were heard
from within the wall. These crackling sounds were probably the result of energy being
released by the propagation of shrinkage cracks in the plasterboard exposed to fire.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 220
The wall failed to support the applied load (Structural Failure) at the end of 53
minutes and the test was stopped. On rolling the furnace back to expose the
plasterboard on the fire side, it was noticed that the exposed plasterboard strip over
Stud 4 (Figure 6-28(a)) had fallen off as it was attached only along one edge.
Shrinkage of the exposed plasterboard had caused it to detach from the fasteners,
opening the joints and exposing the studs as shown in Figure 6-26(b).
(a) Detachment of Plasterboards (b) Close up of Open Joint along the joint
Figure 6-26: Detachment and Opening of Plasterboard joints Caused by Shrinkage
The joints opened up from 20 mm at the base to about 35 mm at the top of the stud,
indicating the greater severity of the plasterboard shrinkage at the top, caused
probably due to the higher temperatures in the chamber at the top due to upward
movement of hot air. Due to the opening of joints the studs immediately behind the
joints not only lost their lateral support from the plasterboard on the fire side but were
also severely affected by higher temperature. As seen in Figure 6-27 Studs 2 and 4 of
the specimen were seen to be more affected than Studs 1 and 3 as they had vertical
plasterboard joints on the fire side. Time-temperature graphs of Studs 2 and 4 (see
Figures 6-30 (b) and (d)) clearly show a much higher temperature of the hot flange at
failure when compared to the hot flanges of Studs 1 and 3 (Figures 6-30 (a) and (c)).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 221
(a) Front View of Test Specimen 2 after Removing the Exposed Plasterboards
Stud 4
Stud 3
Stud 2 Stud 1
(b) Side View Showing Detachment of Studs from the Plasterboards by Screw Pull-out
Figure 6-27: Test Specimen 2 after Removing the Exposed Plasterboard Layer
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 222
The collapse of exposed plasterboard over Stud 4, on the fire side towards the end of
the test led to a rapid increase in the temperature of the stud resulting in the plastic
deformation of the top portion of the stud as shown in Figure 6-28 (b).
Stud 4 Stud 4
(a) Detachment and Collapse of (b) Stud Failure Exposed Plasterboard
Figure 6-28: Stud Failure Initiated by Plasterboard Fall-off
Visual inspection revealed that Stud 4 was the first to fail followed by Stud 2. Studs 2
and 4 with vertical joints of plasterboard on the fire side underwent larger thermal
bowing deformations due to higher thermal gradients across the cross-section. This
caused them to separate from the ambient side plasterboard, pulling the screws
inwards as shown in Figure 6-27(b). The reduced lateral support on either side of the
studs coupled with decreasing mechanical properties at elevated temperatures caused
the studs to undergo flexural torsional buckling about the minor axis.
The friction fit connections provided at the top end of each stud, led to a greater
degree of instability in the frame as they allowed the top end to shift in the plane of
the wall during the thermal deformations. Also at the beginning of the test when the
specimen was loaded at room temperature, the gaps provided in the joints (for the
purpose of allowing free thermal expansion of the studs when exposed to fire) closed
up, thus defeating the main purpose. Figure 6-31(a) shows the axial deformation of
the studs when subjected to loading at ambient temperature. The large variations in
the profiles of the individual studs were probably on account of the closure of the
unequal gaps in the joints before the studs underwent actual axial shortening due to
the applied loads.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 223
The plasterboard on the ambient side was seen to be in good condition with the paper
on the cavity facing surface burnt only in few locations, thus maintaining the integrity
of the wall. Heat penetration failure (Insulation failure) was also not detected as the
temperature on the ambient face of the unexposed plasterboard was much lower than
the standard failure criteria (maximum average temperature of 1400C above the
ambient or a maximum temperature of 1800C at any location on the ambient surface)
until the end of the test as recommended by AS 1530.4 (SA, 2005).
The cause of failure of the wall specimen could be attributed to the structural failure
of the frame precipitated by the opening of plasterboard joints and partial collapse of
plasterboard on the fire side.
6.5.2.2) Time-Temperature Profiles
The time-temperature profile of the furnace and the exposed face of the wall (FS)
were seen to follow the standard time-temperature curve defined by AS 1530.4 (SA,
2005)
a) Plasterboard Surfaces
1) Average temperature on the cavity facing surface of the exposed Plasterboard
(Pb1-Cav)
The temperature on this surface developed in three phases. In the first phase the
temperature was quick to rise from about 3 minutes to approximately 1000C by the
end of 5 minutes. In the second phase, which lasted from 5 minutes to 20 minutes, the
temperature on the plasterboard surface was maintained constant at about 1000C due
to the energy consumed in converting the free and chemically bound water present in
the plasterboard into steam. The data logger failed due to some glitch in the software
during the period from 9 to 17 minutes and no readings were taken in this time period.
The beginning and end of the plateau (second phase) can be observed in Figure 6-29.
In the third phase beyond 20 minutes the temperature of the plasterboard surface
reached 4000C by 40 minutes. Towards the end of the test the temperature had
reached 5000C. A sharp rise in temperature beyond 5000C indicates the breaching of
the exposed plasterboard.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 224
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 225
2) Average temperature on the cavity facing surface of the ambient Plasterboard
(Pb2-Cav)
The temperature profile of this surface followed very closely the profile of Pb1-Cav
until the end of second phase. A maximum temperature difference of about 1100C
across the cavity was observed at the end of 30 minutes. Beyond this time the
temperature difference was seen to gradually decrease (due to the degradation of the
exposed plasterboard) with the Pb2-Cav profile almost merging with the Pb1-Cav
profile towards the end of the test.
3) Average temperature on the ambient side of unexposed plasterboard 2
The temperature on this surface was below 800C during the test.
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 6
Time (min)
Tem
per
atu
re (
oC
)
0
AS 1530.4 Furnace FS Pb1-Cav Pb2-Cav Amb
Figure 6-29: Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 2
b) Steel Surfaces
Figures 6-30 (a) to (d) show the time-temperature profiles of Studs 1 to 4,
respectively, during the fire test. Time-temperature profiles of the studs develop in
three stages in phase with the Pb1-Cav temperature, as the temperature of the hot
flanges in the studs primarily depend up on the heat conducted through the Pb1-Cav
surface. The temperature rise of Studs 2 and 4 is observed to be more rapid than that
of Studs 1 and 3 as a result of the plasterboard joints present along their lengths (see
Figure 6-3 (b) for joint locations).
Figure 6-31 (a) shows the axial deformations measured during the loading of the studs
at ambient temperature. The initial readings up to 4 kN load is primarily due to the
closure of the expansion gap provided in the friction fit end connections of the studs
during the construction of the wall. Axial shortening of approximately 1 to 1.5 mm
was observed when the studs were fully loaded. Figure 6-31 (b) shows the axial
deformations measured during the fire test. Axial deformation profiles of Studs 2 and
4 reverse suddenly at about 51 minutes, suggesting the sudden buckling of these studs
at this stage. Figure 6-28 (b) shows the plastic deformation and local buckling in Stud
4 and Figure 6-27 (b) shows the global buckling of Stud 2, with the plasterboard
screws pulled through the ambient side plasterboard. A maximum lateral deformation
of approximately 28 mm was observed at the mid-height of the wall in line with Stud
2 as can be seen from Figure 6-32.
0
100
200
300
400
500
600
700
800
900
1000
0 5 10 15 20 25 30 35 40 45 50 55 60
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace Hot FlangeWeb Cold Flange Failure Time
(a) Time-Temperature Profiles across Stud 1
Figure 6-30: Time-Temperature Profiles across Studs 1 to 4 of Test Specimen 2
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 226
0
100
200
300
400
500
600
700
800
900
1000
0 5 10 15 20 25 30 35 40 45 50 55 60
Time (min)
Tem
per
atu
re (
oC
)
AS 1350.4 Furnace Hot Flange
Web Cold Flange Failure Time
(b) Time-Temperature Profiles across Stud 2 (A Vertical Plasterboard Joint Runs along the Length of This Stud)
0
100
200
300
400
500
600
700
800
900
1000
0 5 10 15 20 25 30 35 40 45 50 55 60
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 Furnace Hot Flange
Web Cold Flange Failure Time
(c) Time-Temperature Profiles across Stud 3
Figure 6-30: Time-Temperature Profiles across Studs 1 to 4 of Test Specimen 2
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 227
0
100
200
300
400
500
600
700
800
900
1000
0 5 10 15 20 25 30 35 40 45 50 55 60
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace Hot Flange
Web Cold Flange Failure Time
(d) Time-Temperature Profiles across Stud 4 (A Vertical Plasterboard Joint Runs along the Length of This Stud)
Figure 6-30: Time-Temperature Profiles across Studs 1 to 4 of Test Specimen 2
-14
-12
-10
-8
-6
-4
-2
0
2
0 2 4 6 8 10 12 14 16Load (kN)
Def
orm
atio
n (
mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Load Profiles of Studs at Ambient Temperature
Figure 6-31: Axial Deformation Plots for Studs of Test Specimen 2
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 228
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 229
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 6
Time (min)
Def
orm
atio
n (
mm
)
0
Stud 1 Stud 2 Stud 3 Stud 4
(b) Axial Deformation -Time Profiles of Studs at Elevated Temperatures
Figure 6-31: Axial Deformation Plots for Studs of Test Specimen 2
-35
-30
-25
-20
-15
-10
-5
0
0 10 20 30 40 50 6Time (min)
De
fle
cti
on
(m
m)
0
Stud 2 Stud 3 Stud 4
Figure 6-32: Lateral Deflection -Time Profiles of Test Specimen 2 at Mid-Height
6.5.2.3) Specimen Report
a) Test Specimen Number: 2
b) Date of Test: 23/06/07
c) Description of Specimen: Large scale load bearing wall comprising lipped steel
channels (90 x 40 x 15 x 1.15 mm) lined on both sides by single layer Gypsum
(FireSTOP) plasterboard 16 mm thick.
d) Overall Thickness of Wall: 122 mm
e) Severity of Test: 100%
The severity of fire exposure in a test is determined by comparison of the area under
the curve of the mean measured furnace temperature with the area under the standard
ISO 834 curve for the same period.
f) Specimen Temperature:
The average temperature of the unexposed surface of the test specimen towards the
end of the test was 700C indicating a rise of 570C above the ambient temperature of
130C. The maximum temperature of the unexposed surface at that time was 850C. The
maximum temperature of the stud hot flange was in the range of 500 to 6000C with a
temperature difference of 100 to 2000C across the stud depth.
g) Specimen Behaviour:
Most of the fire side plasterboard 1 remained attached to the steel frame until the
failure of test specimen. The lateral deflection of the test specimen was towards the
furnace and the maximum deflection at mid-height of the wall just prior to failure was
28 mm. The total thermal expansion of the studs during the fire test was in the range
of 5 to 7.5 mm.
h) Failure Criterion:
The test specimen was deemed to have failed at approximately 53 minutes from the
start of the test when the specimen could no longer sustain the applied load.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 230
The failure of studs occurred due to the opening of plasterboard joints and partial
collapse of fireside plasterboard. Higher stud temperatures (i.e. reduced mechanical
properties) and reduced lateral support led to the studs failing by local and overall
buckling modes. The presence of plasterboard joints along the stud height affected the
behavior and failure of studs.
Stud 4 failed by local buckling at a hot flange temperature of 5500C
The value of the actual load acting on the wall specimen cannot be stated as the load
counter was reset during the experiment when the software failed temporarily.
i) Performance:
Performance observed in respect of the following criteria:
Structural adequacy - Failure at 53 minutes
Integrity - No failure at 53 minutes
Insulation - No failure at 53 minutes
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 231
Test Specimens 3 (without insulation), 4, 5, 6 (cavity insulated) and 7, 8, 9
(Externally insulated)
All the load bearing wall specimens tested for fire resistance showed similar initial
response during the test. They showed hardly any signs of stress when subjected to a
total axial load of 60 kN (15 kN/stud, giving a load ratio of 0.2). None of the
specimens showed any lateral deformation. Upon starting the furnace, smoke and
steam was seen to come out from the periphery of the specimen at the end of three to
four minutes. The smoke would indicate the burning of the paper on the exposed face
of the external plasterboard, and the steam was due to the escape of moisture (both
free and chemically bound) from the plasterboards. The presence of steam could be
easily noted as it would condense on the inner sides of the loading frame producing
streaks of water running down the column faces. The specimens would display
periods of steady burning with little or no smoke or steam. This would happen after
the complete burning of the paper and the complete conversion of water into steam
from the plasterboard. The smoke and steam would reappear with subsequent layers
of plasterboard heating up. There were periods of thick smoke ensuing continuously
from the specimens for almost 30 to 45 minutes. This would probably indicate the
burning of the insulating material used in the walls. Amongst the three insulations
used, cellulose fibre was seen to produce the maximum smoke and rock fibre the
minimum.
Lateral deflections were visible in the cavity insulated specimens after about 70
minutes of fire test. In the case of externally insulated specimens, the lateral
deflections became noticeable only towards the end of the test. In both cases, the
deflection of the wall was initially towards the furnace. Near the end of the test, a
reversal of lateral deflection was observed for all the specimens forcing the wall to
bow in the outward direction. Maintaining the applied load on the wall was difficult at
this stage with the hand pump controlling the jacks being operated more frequently.
The failure was sudden in all the specimens with the load quickly dropping off with
the studs buckling in the outward direction cracking the plasterboards on the ambient
side.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 232
The ambient surface of each wall specimen recorded values well below the insulation
failure temperature throughout the test. The failure of the walls in every test was due
to the structural failure of the studs.
6.5.3: Test Specimen 3
6.5.3.1) Visual Observations
The specimen was exposed to the furnace heat for a period of 112 minutes. The test
was stopped when the specimen could no longer maintain the applied load. Visual
inspection of the fire tested specimen reveled that the plasterboards 1 and 2 (exposed
plasterboards) though severely calcined were still intact offering protection to the
studs. The screws connecting the plasterboards to the studs were seen to have been
pulled in through the plasterboard thickness at the top 1/3rd portion of the wall.
On the removal of exposed plasterboards (Pb1 and Pb2) it was noticed that the studs
had been laterally displaced at the top end. The friction fit joints provided at the top
end of each stud had failed to prevent the slipping of the studs in the lateral direction
at elevated temperatures. This bending of the studs about the minor axis near the top
portion of the wall caused the screws to pull out from the plasterboard body.
The central studs also displayed distortional buckling in the top 1/3rd portion of their
lengths. The ambient side plasterboards (Pb3 and Pb4) were seen to be in a fairly
good condition. The unexposed surface of the specimen showed no visible signs of
wall failure.
6.5.3.2) Time-Temperature Profiles
a) Plasterboard Surfaces
i) Average temperature of the interface surface between the exposed
Plasterboards 1 and 2 (Pb1-Pb2)
The temperature on this surface was seen to follow three phases of development. The
first phase was highlighted by a quick rise in temperature from the ambient
temperature at about 2 minutes to approximately 1000C by the end of 5 minutes. This
was followed by a second phase, a plateau during which with the temperature was
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 233
constant at 1000C until 16 minutes. During the second phase the free and chemically
bound water in the plasterboard was expelled out in the form of steam. Beyond 16
minutes the third phase started with the temperature rising sharply from 1000C to
8000C by the end of 83 minutes. The temperature gradient was seen to be almost
constant until this point. Beyond 83 minutes the temperature gradient decreased
slightly and the curve seemed to run parallel to the fire curve maintaining a difference
of approximately 1500C to 2000C until the end of the test.
Plasterboard 3 intact although slightly damaged at screw locations due to pull out caused by stud distortion in the top portion
Distortional and flexural buckling @ minor axis in the top third portion for studs 1, 2 and 3
Stud 1
Stud 2
Stud 3
Stud 4
(a) Side View of Test Specimen 3 after Removing the Exposed Plasterboards
Figure 6-33: Close up of Test Specimen 3 after Removing the Exposed Plasterboards
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 234
Stud 3 Stud 2
(b)
Figure 6-33: Test Specimen 3 after Removing the Exposed Plasterboards
(a) Stud 1 (b) Stud 2
(c) Stud 3 (d) Stud 4
Figure 6-34: Studs of Test Specimen 3 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 235
Feng et al. (2005) observed in his experiments on load bearing wall panels that about
25% of the chemically bound water in the plasterboard which was not released during
the second phase of the graph was released at about 7000C giving a second plateau in
the time-temperature profile of the plasterboard. However, this was not observed in
the current experiment. The temperature of the interface (Pb1-Pb2) showed no change
in gradient as it crossed the 7000C mark. Also the slight drop in gradient beyond 83
minutes could be the result of a balance being achieved between the inflow and
outflow of heat across the thickness of the plasterboard, thus maintaining a
temperature difference of approximately 2000C across the 16 mm thick FireSTOP
gypsum plasterboard.
The external plasterboard appeared to be intact until the end of the fire test as the
temperature (between the fire curve and the Pb1-Pb2 interface) was maintained until
the failure of the frame.
ii) Average temperature on the cavity facing surface of the exposed plasterboard
2 (Pb2-Cav)
The temperature on this surface was seen to increase from the ambient at about 5
minutes to approximately 800C by the end of 13 minutes representing the end of the
first phase of the curve. The second phase constituting the plateau lasted until 55
minutes with the temperature hovering about 1000C, beyond which the third phase
started with a sharp rise in temperature. The temperature crossed 4000C by about 90
minutes and 5000C by about 110 minutes. During the third phase the temperature
gradient was fairly constant and maintained a temperature difference of approximately
4000C with the Pb1-Pb2 interface temperature graph which served as the fire curve for
the Pb2-Cav surface. This meant a temperature difference of 4000C across the
thickness of plasterboard 2 (i.e. the second layer on the fire side). The maintenance of
this temperature difference until the end of the test indicates the continued integrity of
plasterboard 2 until the failure of the specimen.
iii) Average temperature on the cavity facing surface of the ambient side
plasterboard 3 (Pb3-Cav)
The temperature profile of this surface coincided almost identically with that of the
Pb2-Cav surface for about 60 minutes. This was probably due to the almost
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 236
instantaneous transmission of heat across the cavity by radiation. Beyond 60 minutes
the temperature increased rapidly to 4000C in 102 minutes. The temperature growth
rate was seen to be linear and lagged behind the Pb2-Cav surface profile by
approximately 500C -1000C until the end of the test.
iv) Average temperature on the ambient side of unexposed plasterboard 3
(Pb3-Pb4)
The temperature of this interface remained below 1000C until the end of the test.
v) Average temperature on the ambient side of unexposed plasterboard 4
The temperature on the unexposed face of the wall was below 700C until the end of
the test.
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace FS Pb1-Pb2Pb2-Cav Pb3-Cav Pb3-Pb4 Amb
(a): Average Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 3
Figure 6-35: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 237
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Tem
per
atu
re (
oC
)
FS-L FS-M FS-R Pb1-Pb2-L Pb1-Pb2-M
Pb1-Pb2-R Pb2-Cav-L Pb2-Cav-M Pb2-Cav-R Pb3-Cav-L
Pb3-Cav-M Pb3-Cav-R Pb3-Pb4-L Pb3-Pb4-M Pb3-Pb4-R
(b) Time-Temperature Profiles across the left, middle and right sections of
Plasterboard Surfaces in Test Specimen 3
Figure 6-35: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 3
b) Steel Surfaces
The studs were well protected from the furnace heat by two layers of plasterboard on
the fire side. Figure 6-36(a) shows the temperature profiles of the 6 thermocouples
reading the hot flange temperatures of the central studs. The temperatures of the
middle portion of the studs are seen to be higher than the top and bottom level
temperatures. This could be due to the thermal bowing of the wall panels towards the
furnace bringing their central portions closer to the furnace burners, thus causing the
central portion of the wall to heat up faster than the top and bottom levels. All
thermocouples except the one at the middle of Stud 2 are seen to fall in a very narrow
band signifying an almost constant temperature distribution over the length of the
studs.
The temperature profiles were seen to develop in three phases. The first phase saw the
temperature increase from the ambient at about 5 minutes to approximately 900C by
20 minutes, which was followed by the second phase of almost constant temperature
of 1000C for about 40 minutes. In the third phase the temperatures increased steadily
until the failure of the specimen. The temperature growth in the hot flanges
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 238
corresponded well with that of the Pb2-Cav surface profile with the only exception
being the thermocouple located at the centre of Stud 2, which recorded higher
temperatures. Towards the end of the test at about 112 minutes the lower and middle
portions of Stud 2 recorded a steep increase in temperature, which coincided with the
suspected breach of plasterboard 2 on the fire side at about the same time.
Figures 6-36 (b) and (c) show the temperature profiles of the webs and cold flanges of
the central studs measured at six locations. The plateau in the second phase extended
until 50 minutes for the webs and 60 minutes for the cold flanges, beyond which the
third phase included a steady temperature rise at almost the same rate (only the lower
cold flange thermocouple on Stud 2 deviated from the group recording higher
temperatures throughout).
Figure 6-37 shows the temperature profiles of the six thermocouples located at mid-
height of Studs 2 and 3 measuring the hot flange, web and cold flange temperatures
for each stud thus giving the temperature variation across the depth of the central
studs during the test. The first two phases demonstrate an almost uniform temperature
gradient across the central studs. The third phase shows a temperature difference
ranging from 1000C to 1700C (i.e. hot flange temperature-cold flange temperature)
across Stud 2 and about 800C to 1000C across that of Stud 3 until the end of the test.
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90 100 110 120Time (min)
Tem
per
atu
re (
oC
)
S2-T-HF S3-T-HF S2-M-HF S3-M-HF S2-L-HF S3-L-HF
(a) Time-Temperature Profiles on Hot Flange Surfaces of Central Studs in Test Specimen 3
Figure 6-36: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 239
0
100
200
300
400
500
600
700
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Te
mp
era
ture
(oC
)
S2-T-W S3-T-W S2-M-W S3-M-W S2-L-W S3-L-W
(b) Time-Temperature Profiles on Web Surfaces of Central Studs in Test Specimen 3
0
100
200
300
400
500
600
700
0 10 20 30 40 50 60 70 80 90 100 110 120Time (min)
Tem
per
atu
re (
oC
)
S2-T-CF S3-T-CF S2-M-CF S3-M-CF S2-L-CF S3-L-CF
(d) Time-Temperature Profiles on Cold Flange Surfaces of Central Studs in Test Specimen 3
Figure 6-36: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 240
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90 100 110 120Time (min)
Te
mp
era
ture
(oC
)
S2-M-HF S3-M-HF S2-M-W S3-M-W S2-M-CF S3-M-CF
Figure 6-37: Time-Temperature Profiles across Central Studs at Mid-height in Test Specimen 3
Note:
S2/3-T-HF/W/CF: Time-temperature profile followed by the hot flange/web/cold flange of Stud No.2/3 at the top level
S2/3-M-HF/W/CF: Time-temperature profile followed by the hot flange/web/cold flange of Stud No.2/3 at mid-height
S2/3-L-HF/W/CF: Time-temperature profile followed by the hot flange/web/cold flange of Stud No.2/3 at lower level
6.5.3.3) Behaviour of Specimen
The loading of the specimen at ambient temperature resulted in an initial displacement
of the studs in the axial direction caused by the closure of the 15 mm gap in the
friction fit connection between the studs and the top track. As the studs were not
screw connected with the top track, the gap closed immediately on the initial
application of the load. On further application of load to 15 kN/stud (giving a load
ratio of 0.2) an axial shortening of about 10 mm was observed for each stud. Figure 6-
38 (a) shows the graph of axial deformation versus load, in which the axial
displacement includes the 15 mm rigid body displacement of the studs in the axial
direction with the rest being the axial deformation for each stud at ambient conditions
when subjected to a load of 15 kN/stud. The wall was able to take the load without
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 241
any sign of visible lateral deformations. On starting the furnace, with the applied load
held constant, the studs were seen to elongate due to thermal expansion.
Figure 6-38 (b) shows the axial deformations of the studs under sustained loading
when exposed to heat. The maximum axial expansion of almost 9 mm was noticed in
Stud 2 by 80 minutes.
Figures 6-39 (a), (b) and (c) show the lateral deformations of the wall during the fire
test at the top, middle and bottom level, respectively (middle and quarter points). A
reversal in the direction of lateral deflection of the wall is seen in Figure 6-39 (a) by
67 minutes allowing the wall to straighten out. About this time the average
temperatures across the hot flanges, webs and cold flanges were 2500C, 1750C and
1400C, respectively. The wall continued to maintain the applied load of 60 kN until
111 minutes at which time the wall suddenly failed. The sudden increase in lateral
deflection as seen in Figure 6-39 (a) and the drop in load in Figure 6-40 confirm the
time of failure.
-30
-25
-20
-15
-10
-5
0
0 2 4 6 8 10 12 14 16
Load (kN)
Dis
pla
cem
ent
(mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Load Profiles of Test Specimen 3 at Ambient Temperature
Figure 6-38: Axial Deformation Plots for Studs of Test Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 242
-35
-30
-25
-20
-15
-10
-5
0
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Def
orm
atio
n (
mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Time Profiles of Test Specimen 3 at Elevated Temperatures
Figure 6-38: Axial Deformation Plots for Studs of Test Specimen 3
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(a) Lateral Deflection -Time Profiles of Test Specimen 3 at Upper Level
Figure 6-39: Lateral Deflection-Time Plots of Test Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 243
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
De
fle
cti
on
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Lateral Deflection -Time Profiles of Test Specimen 3 at Middle Level
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(c) Lateral Deflection -Time Profiles of Test Specimen 3 at Lower Level
Figure 6-39: Lateral Deflection-Time Plots of Test Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 244
10
11
12
13
14
15
16
17
18
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Lo
ad (
kN)
Figure 6-40: Axial Load -Time Profile of Test Specimen 3 during Fire Test
6.5.3.4) Specimen Report
a) Date of Test: 27/08/07
b) Severity of Test: 100%
c) Specimen Temperature
The average temperature of the unexposed surface of the test specimen towards the
end of the test was 690C indicating a rise of 490C above the ambient temperature of
200C. The maximum temperature of the unexposed surface at that time was 720C.
d) Specimen Behaviour
The fire side plasterboards 1 and 2 remained attached to the steel frame until the
failure of the test specimen. The lateral deflection of the specimen was initially
towards the furnace and then reversed in direction at the end of 67 minutes (more
pronounced at the top) from the commencement of the test. The maximum deflection
at mid-height of the wall at that time was 16 mm.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 245
e) Failure Criterion
The test specimen was deemed to have failed at approximately 111 minutes from the
start of the test when the specimen could no longer sustain the applied load.
Partial fall off of exposed plasterboards near Studs 2 and --- occurred suddenly that
led to overall buckling failure of those studs and the failure of test specimen.
f) Performance
Performance observed in respect of the following criteria:
Structural adequacy - Failure at 111 minutes.
Integrity - No failure at 111 minutes
Insulation - No failure at 111 minutes
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 246
6.5.4: Test Specimen 4 (LBW-Cavity Insulation-GF)
6.5.4.1) Visual Observations:
Test Specimen 4 was subjected to heat in the furnace for 101 minutes. Visual
inspection soon after the test showed that plasterboards 1 and 2 (Fire side
plasterboards) had fallen off in the lower right hand portion of the wall. In the
remaining area of the wall plasterboard 1 had fallen off in some portions of the wall
whereas plasterboard 2 was still intact though severely damaged. The external
plasterboards fully collapsed under their own self weight when the specimen was
removed from the loading frame for further inspection (see Figure 6-41 (b)).
The cavity insulation was totally burnt out at the lower right hand portion whereas in
the remaining portion of the wall the cavity insulation was still intact though it had
warped and shrunk to some extent exposing certain parts of the cavity facing surface
of plasterboard 3. The glass fibres had melted on the exposed surface of the
insulation. The inner layers of insulation were still in good condition (see Figure 6-41
g). On removing the remaining pieces of exposed plasterboard and the cavity
insulation the paper on the cavity facing surface of plasterboard 3 was found to be
more or less intact although burnt out at the lower right hand portion (Figure 6-41
(c)).
The front view of the studs clearly shows that torsional failure along with flexural
buckling of the steel channels about the minor axis was fully prevented by the lateral
support offered by the plasterboards on both sides. The ambient surface of the wall
was not affected by the heat of the furnace although it had cracked up horizontally at
the centre when the wall failed by bowing in the outward direction. Studs 1, 2 and 3
from the right side of the wall had failed by local buckling (compressive failure) of
the hot flange close to the mid-height of the wall (see Figure 6-41 (d)) resulting in the
reversal of lateral displacement and causing the outward movement of the wall. Stud 4
was seen to be relatively undamaged. Stud 1 although placed symmetrically to Stud 4
showed greater damage, probably because the base layer plasterboard on the fire side
had a vertical joint running over its length, whereas the plasterboard was continuous
over Stud 4. Local buckling of the hot flange of Stud 3 was seen to occur between the
screws connecting the hot flange to the plasterboards on the fire side indicating a good
support offered by the fire side plasterboards to Stud 3 until its failure. Stud 2 (see
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 247
Figure 6-41 (e)) had its hot flange buckle locally at the screw location after the screw
got pulled out from the plasterboard thus doubling the effective length for the flange
buckling. The upper and lower tracks supporting the studs were relatively undamaged
and were seen holding the studs firmly in place (see Figure 6-41 (f)). The
plasterboards on the ambient side were intact giving good lateral support to the studs
throughout the test.
(a) View of Test Specimen 4 after the Fire Test
(b) Front View Showing Cavity Insulation (glass fibre) Burnt Out in the Lower Right Region
Figure 6-41: Test Specimen 4 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 248
Stud 4 Stud 3 Stud 2 Stud 1
(b) Front View After Removing Exposed Plasterboards and Cavity Insulation
Stud 3 Stud 4
Stud 2
Stud 1
(c) Side View Showing the Overall Buckling away from Furnace
Figure 6-41: Test Specimen 4 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 249
(e) Local Buckling of Hot Flange (Stud 2)
(f) Top Track
(g) Glass Fibre Mat Used as Cavity Insulation
Figure 6-41: Test Specimen 4 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 250
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 251
6.5.4.2) Time-Temperature Profiles
a) Plasterboard Surfaces (see Figures 6-42 (a) and (b))
Figures 6-42 (a) and (b) show the time-temperature profiles of the plasterboard
surfaces as observed during the fire test. Figure 6-42 (b) shows the detailed time-
temperature profiles as recorded by all the individual thermocouples installed across
the left, middle and right sections of the wall, the average of which gives the profiles
in Figure 6-42 (a).
i) Average temperature of the interface surface between the exposed
plasterboards 1 and 2 (Pb1-Pb2)
The temperature on this surface was seen to develop in three phases. The first phase
involved a quick rise in temperature from the ambient at about 3 minutes to about
800C by the end of 5 minutes. Beyond this time the second phase started with the
temperature being held constant at about 1000C for about 20 minutes. Beyond 20
minutes the third phase started with the temperature rising quickly from about 1000C
to 8500C after 90 minutes. The temperature gradient was almost constant until this
point (similar to specimen 4, this graph too did not show any change in gradient upon
crossing the 7000C mark). Beyond 90 minutes the curve flattened out and ran with an
almost constant temperature difference of approximately 1500C with the fire curve
until the end of the test (indicating a temperature difference of 1500C across the
thickness of the fire side plasterboard). This constant temperature difference across
the plasterboard thickness also suggests no loss in the integrity of the exposed
plasterboard until the failure of the frame.
ii) Average temperature on the cavity facing surface of the exposed plasterboard
2 (Pb2-Cav)
The thermocouples on this surface responded at about 4 minutes and recorded a
temperature of about 900C by the end of 8 minutes. The second phase (plateau) lasted
until 55 minutes beyond which the third phase started with the temperature rising very
rapidly to 4000C by 65 minutes and 5000C by 70 minutes. The rate of temperature rise
for this surface decreased beyond 70 minutes and seemed to have reached a steady
state of heat flow maintaining a temperature difference of 2000C across the thickness
of the plasterboard until the end of the test. This also indicated that plasterboard 2
maintained its integrity and continued to offer protection to the steel frame until its
failure.
ii) Average temperature on the cavity facing surface of the ambient side
plasterboard 3 (Pb3-Cav)
The cavity insulation kept the temperature of this surface below 900C until 70 minutes
from the start of the test. A linear growth rate in temperature rise was observed
beyond this time. The temperature on this surface was below 2000C at the end of the
test. The temperature across the thickness of glass fibre insulation in the cavity was
about 5000C when the test was terminated. This indicated that the glass fibre
insulation maintained its integrity and was not burnt through until the failure of the
specimen.
iv) Average temperature on the ambient side of unexposed plasterboard 3
(Pb3-Pb4)
The temperature of this interface remained below 1000C until the end of the test.
0
100
200
300
400
500
600
700
800
900
1000
1100
0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace FS Pb1-Pb2
Pb2-Cav Pb3-Cav Pb3-Pb4 Amb
(a) Average Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 4
Figure 6-42: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 4
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 252
0
100
200
300
400
500
600
700
800
900
1000
1100
0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace Pb1-Pb2-L Pb1-Pb2-M Pb1-Pb2-RPb2-Cav-L Pb2-Cav-M Pb2-Cav-R Pb3-Cav-L Pb3-Cav-MPb3-Cav-R Pb3-Pb4-L Pb3-Pb4-M Pb3-Pb4-R
(b) Time-Temperature Profiles across the left, middle and right sections of
Plasterboard Surfaces in Test Specimen 4
Figure 6-42: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 4
b) Steel Surfaces
Figures 6-43 (a), (b) and (c) show the temperature profiles of the hot flanges, webs
and cold flanges, respectively for the central studs. The rate of temperature increase in
the first phase was seen to decrease with the increase in the distance of the
thermocouples from the Pb2-Cav surface, giving the fastest temperature growths in
the hot flanges and the lowest in the cold flanges. The second phase extended until 50
minutes in the case of hot flanges, whereas it was seen to last until 56 minutes and 62
minutes for the webs and cold flanges, respectively. The third phase was marked by a
rapid rise in stud temperatures with the temperature growth rate of the hot flanges
being the maximum with the central hot flange temperature of Stud 2 following the
Pb2-Cav profile closely. A temperature difference was noticed along the length of the
studs. Maximum temperatures were recorded at mid-height and minimum at the top.
A temperature difference ranging from 1300C to 2300C was noticed along the length
of the studs in the hot flanges towards the end of the test. Similar patterns were
observed for web and cold flange, with the temperature differences along the stud
lengths ranging from 700C to 1800C in Figure 6-43 (b) and 1000C to 2600C in Figure
6-43 (c) at the end of the test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 253
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90 100 110Time (min)
Te
mp
era
ture
(oC
)
S2-T-HF S3-T-HF S2-M-HF S3-M-HF S2-L-HF S3-L-HF
(a) Time-Temperature Profiles on Hot Flange Surfaces of Central Studs in Test Specimen 4
0
100
200
300
400
500
600
0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
Tem
per
atu
re (
oC
)
S2-T-W S3-T-W S2-M-W S3-M-W S2-L-W S3-L-W
(b) Time-Temperature Profiles on Web Surfaces of Central Studs in Test Specimen 4
Figure 6-43: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 4
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 254
0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80 90 100 110Time (min)
Tem
per
atu
re (
oC
)
S2-T-CF S3-T-CF S2-M-CF S3-M-CF S2-L-CF S3-L-CF
(c) Time-Temperature Profiles on Cold Flange Surfaces of Central Studs in Test Specimen 4
Figure 6-43: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 4
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90 100 110Time (min)
Te
mp
era
ture
(oC
)
S2-M-HF S3-M-HF S2-M-W S3-M-W S2-M-CF S3-M-CF
Figure 6-44: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 4
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 255
Figure 6-44 shows the temperature profiles across the depth of the central studs
measured at mid-height. In the first two phases the temperature distribution across the
studs was seen to be almost uniform. After the calcination of the exposed
plasterboards, the temperatures started rising rapidly (third phase) in the studs. The
presence of cavity insulation shielded the cold flanges from direct heat introducing a
large temperature variation across the depth. Maximum temperature differences of
3600C and 3300C were observed across Studs 2 and 3, respectively, at the end of the
test.
6.5.4.3) Behaviour of Specimen
Figure 6-46 (a) shows the axial deformations of each stud loaded to 15 kN at ambient
temperature. Rigid body displacements of the studs as witnessed in Test Specimen 3
were not observed as the studs were screw connected to the upper and lower tracks.
The studs were seen to deform from 4mm to 6 mm towards the end of the loading. At
this stage, the wall had no visible lateral deformations.
Figure 6-46 (b) shows the thermal expansions of the individual studs from the time
the furnace was started until the end of the fire test. A total expansion of
approximately 10 mm was observed in the studs at the end of the test.
Figures 6-47 (a), (b) and (c) show the lateral deformations of studs with respect to
time taken at three different levels. Lateral deformations towards the furnace could be
noticed from the start with the central studs deflecting more than the end studs making
them the critical studs. Until 55 minutes the lateral deformations were seen to develop
slowly. Beyond this time the deformations became more rapid, with Stud 2 recording
the maximum deflection of approximately 32 mm at the centre by the end of 85
minutes with its hot flange recording a temperature of 5700C and with a temperature
difference of approximately 3300C across the cross-section. About the same time the
lateral deflection of Stud 3 was approximately 30 mm at the centre, with its hot flange
measuring a temperature of about 4700C and a temperature difference of 2500C across
its depth. Beyond 85 minutes, Stud 2 reversed its direction of lateral deformation and
started to straighten out. Stud 3 maintained its deflection profile from 85 to
approximately 95 minutes before reversing its lateral deformation and begin to
straighten out. By this time its hot flange had reached a temperature of 5700C
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 256
(temperature at which Stud 2 had started to deform in the outward direction) and had a
temperature difference of approximately 3100C across its depth. Meanwhile Stud 2
had deflected by 12 mm in the reverse direction and had a hot flange temperature of
6600C. Beyond 95 minutes the lateral deformations progressed rapidly in both the
studs leading to failure at 101 minutes as seen from Figure 6-48 which gives the exact
time of failure of the wall. The ambient side plasterboards were cracked open by the
outward thrust offered by Stud 2 at the centre (Figure 6-45).
Figure 6-45: Outward Lateral Deflection of Test Specimen 4 at Failure
The sudden deformation of the wall at about 101 minutes must have caused the partial
collapse the exposed plasterboards and led to the burning out of the cavity insulation.
The presence of the external plasterboards and the insulation can be verified from the
time-temperature profiles as seen in Figures 6-42 (a) and (b).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 257
-8
-6
-4
-2
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16
Load (kN)
De
form
ati
on
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Load Profiles of Test Specimen 4 at Ambient Temperature
-8
-6
-4
-2
0
2
4
6
8
10
12
0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
Def
orm
atio
ns
(mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Axial Deformation -Time Profiles of Test Specimen 4 at Elevated Temperatures
Figure 6-46: Axial Deformation Plots for Studs of Test Specimen 4
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 258
-35
-30
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
Def
lect
ion
(m
m)
Stud 2 Stud 3 Stud 4
(a) Lateral Deflection -Time Profiles of Test Specimen 4 at Upper Level
-35
-30
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Lateral Deflection -Time Profiles of Test Specimen 4 at Middle Level
Figure 6-47: Lateral Deflection-Time Plots of Test Specimen 4
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 259
-35
-30
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(c) Lateral Deflection -Time Profiles of Test Specimen 4 at Lower Level
Figure 6-47: Lateral Deflection-Time Plots of Test Specimen 4
10
11
12
13
14
15
16
17
18
0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
Lo
ad (
kN)
Figure 6-48: Axial Load -Time Profile of Test Specimen 4 during Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 260
6.5.4.4) Specimen Report
a) Date of Test: 11/04/08
b) Severity of Test: 100%
c) Specimen Temperature
The average temperature of the unexposed surface of the test specimen towards the
end of the test was 540C indicating a rise of 340C above the ambient temperature of
200C. The maximum temperature of the unexposed surface at that time was 600C.
d) Specimen Behaviour
The fire side plasterboards 1 and 2 remained attached to the steel frame until the
failure of the test specimen. Portions of plasterboard in the lower right hand area of
the wall fell off after the structural failure of the wall.The lateral deflection of the test
specimen was initially towards the furnace and then reversed in direction at the end of
85 minutes from the commencement of the test. The maximum deflection at mid-
height of the wall at that time was 32 mm.
e) Failure Criterion
The test specimen was deemed to have failed at approximately 101 minutes from the
start of the test when the specimen could no longer sustain the applied load.
Fire side plasterboards appeared to have provided sufficient restraint to the studs and
hence the studs did not undergo any buckling failures associated with twisting. The
main failure mode was due to local buckling / crushing of hot flanges and flexural
deformations about the major axis.
f) Performance
Performance observed in respect of the following criteria:
Structural adequacy - Failure at 101 minutes.
Integrity - No failure at 101 minutes
Insulation - No failure at 101 minutes
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 261
6.5.5: Test Specimen 5 (LBW-Cavity Insulation-RF)
6.5.5.1) Visual Observations:
Test Specimen 5 was subjected to heat in the furnace for 107 minutes. Both of the fire
side plasterboards had fallen off at the end of the test as seen in Figure 6-49 (a). The
rock fibre cavity insulation was almost fully intact with only the outer layer of
insulation having lost its integrity at certain locations as seen in Figure 6-49 (b). On
stripping the cavity insulation off from the wall, it was noted that Plasterboard 3 had
remained in good condition until the end of the test. Only the paper on the cavity
facing surface of Plasterboard 3 was burnt in certain locations as seen in Figure 6-49
(c).
The front view (see Figure 6-49 (c)) shows that torsional failure was prevented in
Studs 2, 3 and 4. However, Stud 1 displayed a combination of local compressive
failure and torsional buckling of the hot flange. The torsional buckling of the hot
flange probably occurred as the exposed plasterboard 1 had partially collapsed in that
region and the severely calcined base layer plasterboard (Pb2) was unable to provide
sufficient lateral restraint all by itself. Figure 6-49 (d) shows the local compressive
failure of the hot flanges of Studs 1, 2 and 3. Figures 6-49 (e) to (g) show the close up
views of the individual studs.
The ambient side plasterboards, although in good condition until the end of the test,
had cracked up when the wall failed by bowing in the outward direction. Studs 2 and
3 from the right side of the wall failed by local compressive failure of the hot flange at
the mid-height between the screws, whereas Stud 1 failed by local buckling of hot
flange initiated by screw pull out. Stud 4 was seen to be in good condition. The tracks
were seen to be in good condition and maintained good contact with the studs
throughout the fire test (Figures 6-49 (h) and (i)).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 262
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 263
(a) Front View Showing Cavity Insulation (Rock Fibre) after the Fire Test
(b)Front View Showing Cavity Insulation (Rock Fibre) after the Fire Test
Figure 6-49: Test Specimen 5 after the Fire Test
Stud 1 Stud 2 Stud 3 Stud 4
Studs are numbered 1 to 4 from right to left
Stud 1 displaying local compressive failure and torsional buckling of the hot flange
Burn-out of paper on the cavity facing surface of Pb3
(c) Front View after Removing Exposed Plasterboard Pieces and Cavity Insulation (Rock Fibre) after the Fire Test
Stud 3 Stud 2 Stud 1
(d) Side View Showing Overall Buckling of Wall Away From Furnace
Figure 6-49: Test Specimen 5 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 264
(e) Local Compressive Failure (f) Local Compressive Failure of Hot Flange (Stud 1) of Hot Flange (Stud 2)
(g) Local Compressive Failure (h) View of Top Track of Hot Flange (Stud 3)
(i) View of Bottom Track (j) Rock fibre mats after Fire Test
Figure 6-49: Test Specimen 5 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 265
6.5.5.2) Time-Temperature Profiles:
a) Plasterboard Surfaces (see Figures 6-50 (a) and (b))
i) Average temperature of the interface surface between the exposed
Plasterboards 1 and 2 (Pb1-Pb2)
The first two phases of the time-temperature profile of this interface was almost
identical to that of Specimens 3 and 4. The third phase started from about 19 minutes
with the temperature rising sharply from 1000C to 9000C by 90 minutes while the
temperature gradient was maintained almost constant over the entire period. Beyond
90 minutes the temperature was almost constant with the fire curve giving a
temperature difference of approximately 500-1000C across the plasterboard thickness
until the end of the test. Figure 6-50 (b) shows the Pb1-Pb2-L temperature profile
merging with the fire side curve at about 88 minutes signifying gradual collapse of
parts of exposed plasterboard in the left section of the wall i.e. between studs 1 and 2.
ii) Average temperature of the cavity facing surface of the exposed Plasterboard
2 (Pb2-Cav)
The first phase started at approximately 4 minutes from the ignition of the furnace
taking the temperature from the ambient to about 500C by the end of 6 minutes
beyond which the profile with a very gentle gradient representing the second phase
started. The temperature in the second phase reached 1100C by the end of 55 minutes,
beyond which the third phase started with the temperature rising sharply and crossing
4000C by 70 minutes and 5000C by 78 minutes. The gradient of the curve beyond 65
minutes reduced with the graph stabilizing itself and maintaining a temperature
difference of 2500C – 3000C across the thickness of the second layer of exposed
plasterboard until the end of the test indicating the physical presence and the
protection offered by the plasterboard until the failure of the specimen.
The absence of any sharp temperature rise in the profiles of the Pb2-Cav as observed
in Figure 6-50 (b) suggests that the exposed base layer plasterboard (Pb2) was intact
in all the three wall sections until the end of the test. Pb2-Cav-L showed a small
uptrend as compared to Pb2-Cav-M and Pb2-Cav-R probably due to the partial
collapse of Pb1 in that region at about 88 minutes. The base layer plasterboard in this
region must have got more severely calcinated as compared to other regions as it
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 266
collapsed at about 106 minutes following the structural failure of the wall as can be
seen by the sudden temperature rise of Pb2-Cav-L profile at 106 minutes in Figure 6-
50 (b).
iii) Average temperature on the cavity facing surface of the ambient side
Plasterboard 3 (Pb3-Cav)
Similar to Specimen 4, the temperature on this surface was kept below 900C until 70
minutes by the protection offered by the rock fibre insulation in the cavity. Beyond 70
minutes the temperature growth rate was linear reaching 2000C by the end of the test.
The rock fibre insulation seemed to have maintained its integrity throughout the test
as no sudden increment in temperature was noted in the Pb3-Cav surface temperature.
The temperature across the insulation was approximately 5500C at the end of the test.
The Pb3-Cav-L profile in Figure 6-50 (b) shows the maximum gradient as compared
to Pb3-Cav-M and Pb3-Cav-R. This is in confirmation to the partial collapse of
plasterboard 1 in this region.
0
100
200
300
400
500
600
700
800
900
1000
1100
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 Furnace FS Pb1-Pb2Pb2-Cav Pb3-Cav Pb3-Pb4 Amb
(a) Average Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 5
Figure 6-50: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 5
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 267
0
100
200
300
400
500
600
700
800
900
1000
1100
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Te
mp
era
ture
(oC
)
AS 1530.4 Furnace Pb1-Pb2-L Pb1-Pb2-M Pb1-Pb2-RPb2-Cav-L Pb2-Cav-M Pb2-Cav-R Pb3-Cav-L Pb3-Cav-MPb3-Cav-R Pb3-Pb4-L Pb3-Pb4-M Pb3-Pb4-R
(b) Time-Temperature Profiles across the left, middle and right sections of Plasterboard Surfaces in Test Specimen 5
Figure 6-50: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 5
iv) Average temperature on the ambient side of unexposed Plasterboard 3
(Pb3-Pb4)
The temperature of this interface remained below 750C until the end of the test.
v) Average temperature on the ambient side of unexposed Plasterboard 4
The temperature on the unexposed face of the wall remained below 550C (well below
the insulation failure temperature) during the fire test.
b) Steel Surfaces
The time-temperatures profiles of the hot flanges, webs and cold flanges taken along
the central studs are shown in Figures 6-51 (a), (b) and (c). The temperatures along
the stud lengths were seen to be more uniform than Test Specimen 4, with the
maximum temperature difference along the stud length being less than 1000C.
The sudden rise in temperature of S2-L-HF (see Figure 6-51 (a)) at about 92 minutes
confirms with the partial collapse of plasterboard 1 in that region.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 268
The second phase of the time-temperature profile was seen to extend until 50 minutes
for the hot flanges, 60 minutes for the webs and 65 minutes in the case of cold
flanges.
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Te
mp
era
ture
(oC
)
S2-T-HF S3-T-HF S2-M-HF S3-M-HF S2-L-HF S3-L-HF
(a) Time-Temperature Profiles of Hot Flange Surfaces of Central Studs in Test Specimen 5
0
50
100
150
200
250
300
350
400
450
500
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Tem
per
atu
re (
oC
)
S2-T-W S3-T-W S2-M-W S3-M-W S2-L-W S3-L-W
(b) Time-Temperature Profiles of Web Surfaces of Central Studs in Test Specimen 5
Figure 6-51: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 5
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 269
0
50
100
150
200
250
300
0 10 20 30 40 50 60 70 80 90 100 110 120Time (min)
Te
mp
era
ture
(oC
)
S2-T-CF S3-T-CF S2-M-CF S3-M-CF S2-L-CF S3-L-CF
(c) Time-Temperature Profiles of Cold Flange Surfaces of Central Studs in Test Specimen 5
Figure 6-51: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 5
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Tem
per
atu
re (
oC
)
S2-M-HF S3-M-HF S2-M-W S3-M-W S2-M-CF S3-M-CF
Figure 6-52: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 5
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 270
Note:
S2/3-T/M/L-HF: Time-temperature profile followed by the hot flange at top/middle/lower level of Stud No. 2/3
S2/3-T/M/L-W: Time-temperature profile followed by the web at top/middle/lower level of Stud No. 2/3
S2/3-T/M/L-CF: Time-temperature profile followed by the cold flange at top/middle/lower level of Stud No. 2/3
Figure 6-52 shows the temperature profiles of the thermocouples placed across the
depth of the central studs at mid-height. Similar to Test Specimen 4, a large
temperature variation is noted across the stud cross-sections. Studs 2 and 3 were seen
to have temperature differences of 3800C and 3300C, respectively, over their depths at
the end of the test.
6.5.5.3) Behaviour of Specimen
The specimen was seen to deflect laterally towards the furnace during the early stages
of the test and then outwards during the final stages (see Figure 6-53 (a) and (b))
finally leading to the cracking of the plasterboard on the ambient side of the wall( see
Figure 6-54).
Figure 6-55 (a) shows the axial deformations suffered by the studs during the loading
of each stud to 15 kN at ambient temperatures. The studs were seen to deform on an
average by approximately 6 mm. Figure 6-55 (b) shows the axial expansion of the
studs brought about by increasing the stud temperatures under constant load. A total
expansion of approximately 10 mm was observed in the studs by the end of the test.
Figures 6-56 (a), (b) and (c) give the lateral deflections of the studs taken at the top,
middle and lower levels. The wall started bowing slowly towards the furnace from the
start of the test and was seen to deform rapidly beyond 60 minutes. The deformations
in stud reached a maximum of 36 mm by 92 minutes at which time the hot flange
temperature of the stud had reached 5500C at the centre and the temperature
difference across the stud was 3300C. Beyond 92 minutes Stud 2 underwent a quick
reversal in lateral deformation and started to straighten out. This affected the axial
deformation plot.
A sharp rise in the expansion rate of Stud 2 can be observed from about 94 minutes
which was caused by the release of the excess oil pressure in the jacks to maintain the
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 271
load constant, thus allowing Stud 2 to reverse its lateral deformation and straighten
out without accumulating temperature stresses. The release of the oil pressure did not
affect the load carried by the other studs as only the excess oil pressure developed by
Stud 2 was released. This could be verified by the smooth continuation of the load
versus time graph (Figure 6-57) until failure.
Stud 3 continued to deflect laterally until 103 minutes reaching a maximum of
approximately 34 mm. At this time its hot flange temperature at the centre had
reached 5500C and the temperature across the stud was 3300C (same as that observed
in Stud 2 at the end of 92 minutes) and started to reverse laterally in the outward
direction. A response similar to the one observed for Stud 2 at 94 minutes can be seen
for Stud 3 at 104 minutes in the axial deformation graph (Figure 6-55 (b)). By 102
minutes Stud 2 had moved in the reverse direction by 24 mm and continued to deform
rapidly outwards. This was simultaneously followed by a rapid increase in the lateral
deformation in the outward direction by Stud 3 leading to the failure of the wall at 107
minutes as seen from Figure 6-57.
(a) Inward Bowing of Specimen during Early Stages of Test
Figure 6-53: Specimen Behaviour during the Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 272
(b) Outward Bowing of Specimen during Final Stages of Test Figure 6-53: Specimen Behaviour During the Test
Figure 6-54: Ambient Side of Wall after Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 273
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 274
-7
-6
-5
-4
-3
-2
-1
0
0 2 4 6 8 10 12 14 1
Load (kN)
Def
orm
atio
n (
mm
)
6
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Load Profiles of Test Specimen 5 at Ambient Temperature
-8
-6
-4
-2
0
2
4
6
8
10
12
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (mm)
Def
orm
atio
ns
(mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Axial Deformation -Time Profiles of Test Specimen 5 at Elevated Temperatures
Figure 6-55: Axial Deformation Plots for Studs of Test Specimen 5
-40
-35
-30
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
De
fle
cti
on
(m
m)
Stud 2 Stud 3
(a) Lateral Deflection -Time Profiles of Test Specimen 5 at Upper Level
-40
-35
-30
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Lateral Deflection -Time Profiles of Test Specimen 5 at Middle Level
Figure 6-56: Lateral Deflection-Time Plots of Test Specimen 5
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 275
-40
-35
-30
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
De
fle
cti
on
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(c) Lateral Deflection -Time Profiles of Test Specimen 5 at Lower Level
Figure 6-56: Lateral Deflection-Time Plots of Test Specimen 5
10
11
12
13
14
15
16
17
18
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Lo
ad (
kN)
Figure 6-57: Axial Load -Time Profile of Test Specimen 5 during Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 276
6.5.5.4) Specimen Report
a) Date of Test: 16/04/08
b) Severity of Test: 100%
c) Specimen Temperature
The average temperature of the unexposed surface of the test specimen towards the
end of the test was 520C indicating a rise of 370C above the ambient temperature of
150C. The maximum temperature of the unexposed surface at that time was 530C.
d) Specimen Behaviour
The fire side plasterboards 1 and 2 remained attached to the steel frame until the
failure of the test specimen. The lateral deflection of the test specimen was initially
towards the furnace and reversed in direction at the end of 92 minutes from the
commencement of the test. The maximum deflection at mid-height of the wall at that
time was 35 mm.
e) Failure Criterion
The specimen was seen to fail deflecting laterally in the outward direction (away from
the furnace) due to the local compressive failure of the exposed flanges under a
flexural bending action about the major axis.
The Test Specimen was deemed to have failed at approximately 107 minutes from the
start of the test when the specimen could no longer sustain the applied load.
f) Performance
Performance observed in respect of the following criteria:
Structural adequacy - Failure at 107 minutes.
Integrity - No failure at 107 minutes
Insulation - No failure at 107 minutes
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 277
6.5.6: Test Specimen 6 (LBW-Cavity Insulation-CF)
6.5.6.1) Visual Observations
The fire test was stopped at 112 minutes when the specimen failed to maintain the
applied load and it was observed that the fire side plasterboards had fallen off from
the central and lower portion as seen in Figure 6-58 (a). Cellulose fibre used in the
cavity had completely burnt out leaving behind a black residue. The paper on the
cavity facing surface of Plasterboard 3 had burnt out completely. The surface had also
undergone deep calcination showing multiple shrinkage cracks (Figure 6-58 (b)).
The studs stayed connected to the ambient side plasterboards over the entire height.
The ambient surface of Plasterboard 4 did not show any temperature effects although
it had lost its integrity due to frame failure. The upper track connecting the studs was
heat affected and showed distortional buckling along its length (Figure 6-58 (g)). The
lower track was still in good condition probably because the convection currents carry
the heat in the upward direction keeping the top track at a temperature higher than the
bottom track.
The central studs were seen to be severely damaged. Stud 1 showed a small amount of
local buckling of the hot flange at 370 mm from the bottom (Figure 6-58 (c)). Stud 4
was undamaged. In Studs 2 and 3 the local buckling of the hot flange involved screw
pull out through the plasterboard thus doubling the buckling length of the flange
leading to rapid outward movement of the studs causing the punching of the
plasterboard on the ambient side (Figure 6-58 d). Stud 3 also displayed local crushing
of the cross-section near the top support in the frame (Figure 6-58 (e)).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 278
(a) View of Test Specimen 6 after the Fire Test
(b) Front View of Test Specimen 6
Note: In all the test specimens the studs are numbered 1 to 4 from right to left
Figure 6-58: Test Specimen 6 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 279
(c) Local Buckling of Hot Flange in Stud 1 at 370 mm from base
(d) Local Buckling of Hot Flange in Stud 2 at 350 mm from base
(e) Local Crushing Near the Top Support
Figure 6-58: Test Specimen 6 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 280
(f) Side View of Specimen (g) View of the Top Track
Figure 6-58: Test Specimen 6 after the Fire Test
6.5.6.2) Time-Temperature Profiles a) Plasterboard Surfaces (see Figures 6-59 (a) and (b))
i) Average temperature of the interface surface between the exposed
Plasterboards 1 and 2 (Pb1-Pb2)
The first two phases of this interface was seen to follow the same profile as that of
Specimens 3, 4 and 5. The third phase started from about 20 minutes with the
temperature rising sharply from 1000C to 9000C by about 85 minutes from the start of
the test. The temperature gradient in the third phase was fairly constant until this time.
Beyond 85 minutes the graphs started to flatten out, but continued to converge with
the fire side curve (suggesting the rapid deterioration of the fire side plasterboard) and
indicating a breach or partial collapse of the plasterboard (see Figure 6-59 (a)).
Figure 6-59 (b) gives the temperature profiles of the individual thermocouples placed
in the three regions of the wall (left, middle and right). Temperature profile of Pb1-
Pb2-L meets the fire side curve at about 79 minutes indicating a partial collapse of
Plasterboard 1 in this region (i.e. between Studs 1 and 2).
ii) Average temperature on the cavity facing surface of the exposed Plasterboard
2 (Pb2-Cav)
The first phase started from about 5 minutes taking the temperature up to 900C by 11
minutes from the start of the test. Beyond this time in the second phase the
temperature was maintained at about 1000C until about 50 minutes beyond which the
third phase started with the temperature rising quickly and crossing 4000C by 68
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 281
minutes and 5000C by 72 minutes. Beyond 70 minutes the curvature of the
temperature gradient started to decrease with the graph stabilizing itself and
maintaining a temperature difference of approximately 2500C across the thickness of
the second layer of the exposed plasterboard until the end of the test. Figure 6-59 (b)
shows a sudden rise in the temperature of Pb2-Cav-R profile close to the end of the
test at about 106 minutes signifying the breaching of Plasterboard 2 in this region.
iii) Average temperature on the cavity facing surface of the ambient side
Plasterboard 3 (Pb3-Cav)
Similar to the other cavity insulated specimens, the temperature on this surface was
maintained below 1000C until 70 minutes. The temperature started rising fast beyond
70 minutes crossing 2000C by 88 minutes, 3000C by 97 minutes and 4000C by 110
minutes. The temperature difference across the thickness of the insulation was close to
2000C at the end of the test.
A sharp rise in the temperature of the Pb3-Cav surface after 110 minutes indicates the
quick deterioration of the cellulose fibre in the cavity of the wall. A steep rise of Pb3-
Cav-L and Pb3-Cav-R in Figure 6-59 (b) indicates the burn-out of the cellulose
insulation in these regions at the end of the test.
iv) Average temperature on the ambient side of unexposed Plasterboard 3
(Pb3-Pb4)
The temperature of this interface remained below 850C until end of the test.
v) Average temperature on the ambient side of unexposed Plasterboard 4
The temperature of the unexposed face of the wall remained under 600C (well below
the insulation failure temperature) during the fire test.
Figure 6-59 (b) shows the detailed time-temperature profiles as recorded by all the
individual thermocouples installed across the left, middle and right sections of the
wall.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 282
0100200300400500600700800900
100011001200
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace FS Pb1-Pb2Pb2-Cav Pb3-Cav Pb3-Pb4 Amb
(a) Average Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 6
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace Pb1-Pb2-L Pb1-Pb2-M Pb1-Pb2-RPb2-Cav-L Pb2-Cav-M Pb2-Cav-R Pb3-Cav-L Pb3-Cav-MPb3-Cav-R Pb3-Pb4-L Pb3-Pb4-M Pb3-Pb4-R
(b) Time-Temperature Profiles across the left, middle and right sections of Plasterboard Surfaces in Test Specimen 6
Figure 6-59: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 6
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 283
b) Steel Surfaces
Similar to the other cavity insulated specimens the temperatures along the length of
the central studs in the web and cold flange portions were seen to be almost uniform
(see Figures 6-60 (b) and (c)). Larger temperature variations along the length were
observed in the hot flanges with the maximum temperature occurring at the centre (6-
60 (a)).
Figure 6-61 shows the temperature variations across the depth of the central studs at
mid-span. The temperature difference between the hot and cold flanges can be seen to
increase from about 1000C at 60 minutes to about 3500 to 3700C at the end of the test.
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120Time (min)
Tem
per
atu
re (
oC
)
S2-T-HF S3-T-HF S2-M-HF S3-M-HF S2-L-HF S3-L-HF
(a) Time-Temperature Profiles of Hot Flange Surfaces of Central Studs in Test Specimen 6
Figure 6-60: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 6
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 284
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Tem
per
atu
re (
oC
)
S2-T-W S3-T-W S2-M-W S3-M-W S2-L-W S3-L-W
(b) Time-Temperature Profiles of Web Surfaces of Central Studs in Test Specimen 6
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120Time (min)
Te
mp
era
ture
(oC
)
S2-T-CF S3-T-CF S2-M-CF S3-M-CF S2-L-CF S3-L-CF
(c)Time-Temperature Profiles of Cold Flange Surfaces of Central Studs in Test Specimen 6
Figure 6-60: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 6
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 285
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Tem
per
atu
re (
oC
)
S2-M-HF S3-M-HF S2-M-W S3-M-W S2-M-CF S3-M-CF
Figure 6-61: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 6
6.5.6.3) Behaviour of Specimen
The Test Specimen on loading until 60 kN (i.e. 15 kN/Stud) showed no visible signs
of lateral deformation at ambient temperature. However, the studs showed an average
axial shortening of about 7 mm on reaching the applied load (Figure 6-63 (a)). On
starting the furnace the studs started expanding due to heat and by the end of the test
the average expansion of the studs was seen to be about 8 mm (Figure 6-63 (b)). The
wall was observed to bow towards the furnace from the start. With the passage of time
the lateral deformations continued to increase (see Figure 6-62) as the temperature
difference across the depth of the wall increased. Interface temperature Pb1-Pb2-L
(Figure 6-59 (b)) shows a jump in temperature at about 78 minutes, indicating a
partial breach or collapse of Plasterboard 1 between Studs 1 and 2. By 96 minutes
Stud 2 was seen to have laterally deformed by 35 mm towards the furnace beyond
which its direction was seen to reverse (Figure 6-64 b). The rest of the studs continued
to bow towards the furnace. About 107 minutes, a step in the temperature profile of
interface Pb1-Pb2-R indicated the cracking up of Plasterboard 1 between Studs 3 and
4 (Figure 6-59 (b)). Within a couple of minutes the exposed base layer plasterboard in
the same place cracked up as shown by Pb2-Cav-R profile exposing the studs and
cavity insulation (cellulose fibre) to direct furnace heat in this region. A temperature
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 286
difference of about 3300C across the cavity at this time (Pb2-Cav R – Pb3-Cav R)
indicates the physical presence of cellulose fibre in the cavity. By about 107 minutes,
Stud 3 had deformed laterally towards the furnace by 44 mm (Figure 6-64 (b)). Its
direction was seen to reverse abruptly beyond this time with the deformations
progressing rapidly and finally leading to the failure of the wall at about 110 minutes
as confirmed by Figure 6-65.
Similar to Test Specimen 5, the structural failure in this specimen also occurred
within 15 minutes of the first stud reversal. In this period large lateral deformations
were responsible for the braking of the fragile and calcinated exposed plasterboards
Figure 6-62: Inward Thermal Bowing of Test Specimen in the Initial Period of the Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 287
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 288
-15
-10
-5
0
5
10
15
0 2 4 6 8 10 12 14 1Load (kN)
De
form
ati
on
(m
m)
6
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Load Profiles of Test Specimen 6 at Ambient Temperature
-15
-10
-5
0
5
10
15
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
De
form
ati
on
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Axial Deformation -Time Profiles of Test Specimen 6 at Elevated Temperatures
Figure 6-63: Axial Deformation Plots for Studs of Test Specimen 6
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (mm)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(a) Lateral Deflection -Time Profiles of Test Specimen 6 at Upper Level
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Lateral Deflection -Time Profiles of Test Specimen 6 at Middle Level
Figure 6-64: Lateral Deflection-Time Plots of Test Specimen 6
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 289
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(c) Lateral Deflection -Time Profiles of Test Specimen 6 at Lower Level
Figure 6-64: Lateral Deflection-Time Plots of Test Specimen 6
10
11
12
13
14
15
16
17
18
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (min)
Lo
ad
(k
N)
Figure 6-65: Axial Load -Time Profile of Test Specimen 6 during Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 290
6.5.6.4) Specimen Report
a) Date of Test: 08/04/08
b) Severity of Test: 100%
c) Specimen Temperature
The average temperature of the unexposed surface of the test specimen towards the
end of the test was 580C indicating a rise of 370C above the ambient temperature of
210C. The maximum temperature of the unexposed surface at that time was 620C.
d) Specimen Behaviour
The fire side Plasterboard 1 breached partially at about 78 minutes between Studs 1
and 2, whereas Stud 2 remained attached to the steel frame until the failure of the Test
Specimen. The lateral deflection of the test specimen was initially towards the furnace
and reversed in direction at the end of 96 minutes from the commencement of the test.
The deflection at mid-height of the wall at that time was 35 mm.
e) Failure Criterion
The test specimen was deemed to have failed at approximately 110 minutes from the
start of the test when the specimen could no longer sustain the applied load.
f) Performance
Performance observed in respect of the following criteria:
Structural adequacy - Failure at 110 minutes.
Integrity - No failure at 110 minutes
Insulation - No failure at 110 minutes
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 291
6.5.7: Test Specimen 7 (LBW-External Insulation-GF)
6.5.7.1) Visual Observations
At the end of fire test, the exposed plasterboards were seen to have fallen off near the
centre of the wall exposing the Pb3-Cav surface (Figure 6-66 (a)). The glass fibre
insulation used between Plasterboards 1 and 2 had completely disappeared leaving
only some molten glass traces along the periphery of the specimen. The base layer
plasterboard on the fire side was seen to have shrunk, opening the plasterboard joints
by 10 -15 mm thus exposing the studs to fire (Figure 6-66 (d)).
The exposed plasterboards were stripped off and the debris removed to expose the
frame (Figure 6-66 (b)). A local buckling wave in the web and flanges was visible
over the middle third length of Stud 3 (Figure 6-66 (c)). Stud 2 showed local
compressive failure of the entire cross-section close to the mid-span suggesting the
complete mobilization of the capacity of the cross-section (Figure 6-66 (f)). Figure 6-
66 (e) shows the local buckling wave in the flange and web near mid-height of Stud 2.
Stud 1 was seen to display local buckling of hot flange and web near mid-height. Stud
4 was seen to be undamaged (Figure 6-66 (c)).
(a) Front View Showing the Partial Collapse of Exposed Plasterboards
Figure 6-66: Test Specimen 7 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 292
The front view (Figure 6-66 (b)) shows that torsional buckling and flexural buckling
about minor axis were fully prevented by the lateral support provided by the
plasterboards. From the side view (Figure 6-66 (c)) one can see that even the flexural
buckling about the major axis was almost negligible for all the studs.
Stud 4 Stud 3 Stud 2 Stud 1
(b) Front View after Removing the Exposed Plasterboards and Remains of External Insulation
Stud 4 Stud 3
Stud 2 Stud 1
(c) Side View Showing the Slight Outward Buckling of Wall
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 293
Figure 6-66: Test Specimen 7 after the Fire Test
(d) 10-15 mm Wide Opening (e) Local Buckling Wave in of Plasterboard Joints Flanges and Web at mid-height of Stud 2
(f) Local Buckling of Flanges
and Web in Stud 2
Figure 6-66: Test Specimen 7 after the Fire Test
6.5.7.2) Time-Temperature Profiles
a) Plasterboard Surfaces (see Figure 6-67)
i) Average temperature of the interface surface between the exposed
Plasterboard 1 and Insulation (Pb1-Ins)
As for the cavity insulated specimens, the temperature profile of this surface showed a
rapid rise in temperature within three minutes of starting the test (first phase) reaching
a temperature of about 900C by the end of 5 minutes. Beyond 5 minutes the second
phase started with the temperature increasing gradually to about 1200C by the end of
15 minutes. This was followed by a very rapid rise in temperature (third phase)
crossing 4000C in 25 minutes and 5500C in 30 minutes. In comparison for Specimen 4
using glass fibre as cavity insulation, the ambient surface of Plasterboard 1 recorded
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 294
the temperature of 4000C after 42 minutes and crossed 5500C by 59 minutes from the
start of the test. The sudden increase in the temperature of the interface of Specimen 7
in contrast to that of Specimen 4 in the early stages of Phase 3 was probably caused
by the heat blocked and redirected by the adjoining layer of insulation. Beyond 30
minutes the temperature gradient was reduced with the temperature crossing 6500C by
about 40 minutes. From this point onwards a difference in temperature of
approximately 2500C was maintained across the thickness of Plasterboard 1 until 72
minutes at which time the furnace malfunctioned causing the fire curve to drop as
seen in Figure 6-67. This caused the temperature of the Pb1-Ins interface to fall
simultaneously and follow the fire curve with a small lag in temperature. The fire
curve was stepped up by manually operating the furnace from 150 minutes onwards
which caused the interface temperature to also rise. The temperature of the interface
was about 9500C by the end of the test.
ii) Average temperature of the interface surface between the insulation and
exposed base layer Plasterboard 2 (Ins-Pb2)
This interface responded to the initial rise in temperature (phase 1) in under 4 minutes
from the start of the furnace and reached a temperature of approximately 800C rapidly
by about 6 minutes and then remained constant (second phase) until about 25 minutes.
Beyond this time the third phase started with the temperature rising almost linearly
with respect to time touching 7000C by about 72 minutes. A closer look at the
individual thermocouples revealed that Ins-Pb2-L thermocouple recorded very rapid
temperature gains and intersected the Pb1-Ins-L curve at 55 minutes indicating the
complete burning of the insulation at mid-height between Studs 1 and 2. The
intersection of curves Pb1-Ins-M and Ins-Pb2-M at about 70 minutes indicates the
burning out of the insulation at mid-height of the wall between Studs 2 and 3. This
was followed by the disintegration of the glass fibre insulation at mid-height of wall
between Studs 3 and 4 at about 78 minutes.
iii) Average temperature on the cavity facing surface of the exposed Plasterboard
2 (Pb2-Cav)
The initial rise in temperature of this surface was followed by a plateau extending
until 80 minutes (compared with the 55 minutes observed in Specimen 4, using glass
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 295
fibre as cavity insulation). The third phase started beyond 80 minutes with the
temperature increasing gradually and crossing 5500C towards the end of the test. A
temperature difference of approximately 5000C was present across the thickness of
Plasterboard 2 towards the end of the test indicating the sustained integrity of the
exposed base plasterboard layer until the failure of the specimen.
iv) Average temperature on the cavity facing surface of the ambient Plasterboard
3 (Pb3-Cav)
The transmission of heat across the cavity was almost instantaneous due to radiation
making the temperature profile of Pb3-Cav surface follow very closely but just on the
underside of the Pb2-Cav surface profile.
v) Average temperature of the interface surface between base layer Plasterboard
3 and insulation (Pb3-Ins)
The temperature of this surface increased gradually reaching about 1000C and
remained almost constant until 125 minutes. The third phase started beyond this time
with the temperature increasing linearly with respect to time reaching 4000C by 180
minutes. Beyond 160 minutes it was observed that the temperature difference across
the plasterboard was approximately 750C until the failure of the specimen showing the
integrity of the ambient base layer plasterboard during the test.
vi) Average temperature of the interface surface between the insulation and
Plasterboard 4 (Ins-Pb4)
This surface was well protected from the fire and remained under 1000C until 170
minutes from the start of the test. The temperature was just below 1100C towards the
end of the fire test.
vii) Average temperature on the ambient side of unexposed Plasterboard 4
The temperature on the unexposed face of the wall remained under 500C (well below
the insulation failure temperature) during the fire test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 296
0
100
200300
400
500
600
700
800900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace FS Pb1-Ins Ins-Pb2
Pb2-Cav Pb3-Cav Pb3-Ins Ins-Pb4 Amb
(a) Average Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 7
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace FS-L FS-M FS-RPb1-Ins-L Pb1-Ins-M Pb1-Ins-R Ins-Pb2-L Ins-Pb2-MIns-Pb2-R Pb2-Cav-L Pb2-Cav-M Pb2-Cav-R Pb3-Cav-LPb3-Cav-M Pb3-Cav-R Pb3-Ins-L Pb3-Ins-M Pb3-Ins-R
(b) Time-Temperature Profiles across the Left, Middle and Right Sections of Plasterboard Surfaces in Test Specimen 7
Figure 6-67: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 7
b) Steel Surfaces
Figures 6-68 (a), (b) and (c) show the time-temperature profiles of hot flanges, webs
and cold flanges of the central studs. The temperatures along the length are seen to be
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 297
almost uniform in all three cases. Figure 6-69 gives the temperature variation across
the central stud cross-sections at mid-height. The temperature difference is observed
to be under 800C throughout.
0
100
200
300
400
500
600
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Te
mp
era
ture
(oC
)
S2-T-HF S3-T-HF S3-M-HF S2-L-HF S3-L-HF
(a) Time-Temperature Profiles of Hot Flange Surfaces of Central Studs in Test Specimen 7
0
100
200
300
400
500
600
700
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
S2-T-W S3-T-W S2-M-W S3-M-W S2-L-W S3-L-W
(b) Time-Temperature Profiles of Web Surfaces of Central Studs in Test Specimen 7
Figure 6-68: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 7
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 298
0
100
200
300
400
500
600
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Tem
per
atu
re (
oC
)
S2-T-CF S3-T-CF S2-M-CF S3-M-CF S2-L-CF S3-L-CF
(c) Time-Temperature Profiles of Cold Flange Surfaces of Central Studs in Test Specimen 7
Figure 6-68: Time-Temperature Plots of Flange and Web Surfaces of Central Studs in Test Specimen 7
0
100
200
300
400
500
600
700
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200Time (min)
Te
mp
era
ture
(oC
)
S3-M-HF S2-M-W S3-M-W S2-M-CF S3-M-CF
Figure 6-69: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 7
6.5.7.3) Behaviour of Specimen
The studs when loaded at ambient temperature up to the applied load (15 kN/stud)
showed an average axial shortening of 5 mm (Figure 6-70 (b)). There were no signs of
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 299
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 300
lateral deformations. On starting the furnace the wall was observed to start bowing
towards the furnace with Studs 2 and 3 reaching a maximum central deflection of 16
mm at 172 and 180 minutes, respectively (Figure 6-71 b)). Near the end of the test
this was reversed and both studs deformed very rapidly in the outward direction
leading to the wall specimen failure at 181 minutes (Figure 6-72).
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
1.0
0 2 4 6 8 10 12 14 1
Load (kN)
Def
orm
atio
n (
mm
)
6
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Load Profiles of Test Specimen 7 at Ambient Temperature
-8.0
-6.0
-4.0
-2.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Time (min)
Def
orm
atio
n (
mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Axial Deformation -Time Profiles of Test Specimen 7 at Elevated Temperatures
Figure 6-70: Axial Deformation Plots for Studs of Test Specimen 7
-20
-15
-10
-5
0
5
10
0 20 40 60 80 100 120 140 160 180 200
Time (min)
Def
lect
ion
(m
m)
Stud 2 Stud 3 Stud 4
(a) Lateral Deflection -Time Profiles of Test Specimen 7 at Upper Level
-20
-15
-10
-5
0
5
10
0 20 40 60 80 100 120 140 160 180 200Time (min)
De
fle
cti
on
s (
mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Lateral Deflection -Time Profiles of Test Specimen 7 at Middle Level
Figure 6-71: Lateral Deflection-Time Plots of Test Specimen 7
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 301
-15
-10
-5
0
5
10
0 20 40 60 80 100 120 140 160 180 200
Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(c) Lateral Deflection -Time Profiles of Test Specimen 7 at Lower Level
Figure 6-71: Lateral Deflection-Time Plots of Test Specimen 7
10
11
12
13
14
15
16
17
18
0 20 40 60 80 100 120 140 160 180 200
Time (min)
Lo
ad (
kN)
Figure 6-72: Axial Load -Time Profile of Test Specimen 7 during Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 302
6.5.7.4) Specimen Report
a) Date of Test: 30/04/08
b) Severity of Test: Less than 80 %
c) Specimen Temperature
The average temperature of the unexposed surface of the test specimen towards the
end of the test was 530C indicating a rise of 380C above the ambient temperature of
150C. The maximum temperature of the unexposed surface at that time was 570C
d) Specimen Behaviour
The fire side Plasterboards 1 and 2 remained attached to the steel frame until the
failure of the test specimen. The lateral deflection of the test specimen was initially
towards the furnace and reversed in direction at the end of 174 minutes from the
commencement of the test. The maximum deflection at mid-height of the wall at that
time was 15 mm.
e) Failure Criterion
The test specimen was deemed to have failed at approximately 181 minutes from the
start of the test when the specimen could no longer sustain the applied load.
f) Performance
Fire performance could not be specified as the test specimen was not subjected to the
standard time temperature curve due to the malfunctioning of the furnace during the
test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 303
6.5.8: Test Specimen 8
6.5.8.1) Visual Observations
The specimen was subjected to 112 minutes of furnace exposure after which the test
was stopped following the failure of the specimen. Figure 6-73 (a) shows the
specimen after the fire test. Exposed Plasterboard 1 (Pb1) had completely fallen off.
Rock fibre insulation had disintegrated near the lower right corner of the wall.
Exposed base layer plasterboard (Pb2) had also collapsed in this area exposing the
cavity surface of the ambient side plasterboard (Pb3).
The external insulation had undergone overall shrinking leading to the opening of the
joints and exposing Plasterboard 2 (Figure 6-73 (b)). The base layer plasterboard had
also undergone extensive calcination with the joints opening up to expose the hot
flanges of the studs (Figure 6-73 (c)).
The front view of the specimen (Figure 6-73 (d)) after removing the external
plasterboards and insulation shows the studs without any torsional or flexural
buckling about the minor axis as the studs were laterally supported adequately.
Figures 6-73 (e) and (h) show the local buckling wave in the central part of Stud 1.
The local buckling wave can be clearly seen in the hot flange and web elements. Studs
2 and 3 also exhibited local buckling of flange and web elements (Figure 6-73 (i)).
Severe local buckling was seen near the mid-height, that led to an outward movement
of the studs and thus breaking of the ambient side plasterboards (Figure 6-73 (g) and
(h)). The tracks were seen to maintain good contact and connection with the studs.
(Figure 6-73 (k)).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 304
Rock fibre insulation intact in the top half of the specimen
Collapse of external plasterboards and insulation in the lower right portion of the specimen
Pb3-cav surface
(a) View Showing Partial Collapse of Exposed Plasterboards and External Insulation
(b) Shrinkage Gap (c) Shrinkage Gap in Base in Insulation Joint Exposing Plasterboard Exposing Steel Stud
Base Plasterboard
Figure 6-73: Test Specimen 8 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 305
Stud 4 Stud 3 Stud 2 Stud 1
(d) Front View after Removing Exposed Plasterboards and External Insulation (Rock Fibre)
Stud 4
Stud 3
Stud 2 Stud 1
(e) Side View Showing the Failure Modes of Studs 1, 2 and 3
Figure 6-73: Test Specimen 8 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 306
(f) Local Buckling Wave (g) Local Buckling of in Flanges and Web Complete cross-section in Stud 2
Stud 3 Stud 2
(h) Local Buckling of (i) Side View Showing Local Buckling Complete cross-section in Stud 3 Wave near Failure Points in the
Studs
(j) View of Top Track (k) Joint between Stud and Track
Figure 6-73: Test Specimen 8 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 307
6.5.8.2) Time-Temperature Profiles a) Plasterboard Surfaces (see Figure 6-74)
i) Average temperature of the interface surface between the exposed
Plasterboard 1 and Insulation (Pb1-Ins)
The temperature on this surface developed in three phases as in Specimen 7 with the
first phase involving a quick rise in temperature from the ambient at about 2 minutes
to 900C by the end of 5 minutes (see Figure 6-74 (a)). The second phase involving the
plateau lasted until 18 minutes by which time the temperature had gradually risen to
about 1100C. Beyond 18 minutes the third phase took off with a sharp rise in
temperature crossing 4000C by 24 minutes and 5500C by 30 minutes (almost identical
with Specimen 7). Beyond 30 minutes the curve seemed to flatten out with the
temperature of the interface increasing gradually and reaching 9000C by about 105
minutes. A temperature difference of 100 - 1500C was noticed at about this time
across the thickness of the plasterboard which was maintained until the end of the test.
A temperature difference of 1000C across Plasterboard 1 towards the end of the test
suggested that the integrity of the fire side plasterboard was maintained until the
failure of the specimen. Figure 6-74 (b) also shows gentle gradients for the Pb1-Ins
profiles signifying the presence of Plasterboard 1 until the end of the test.
ii) Average temperature of the interface surface between the insulation and
exposed base layer Plasterboard 2 (Ins-Pb2)
The temperature of this interface gained rapidly (first phase) from the ambient
temperature at 4 minutes and climbed rapidly to about 900C under 5 minutes (see
Figure 6-74 (a)). This was followed by a plateau (second phase) extending until 25
minutes, beyond which the temperature gained almost linearly with respect to time
(third phase) reaching 4000C by 60 minutes and 8000C by 110 minutes. After this time
a constant temperature difference of about 1000C was maintained with the Pb1-Ins
curve, which signifies a temperature difference of 1000C across the thickness of the
insulation until the end of the test. This indicated the intactness of the rock fibre
insulation used as external insulation between Plasterboard 1 and Plasterboard 2 until
the failure of the specimen.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 308
Figure 6-74 (b) shows a temperature difference between the profiles of Pb1-Ins and
that of Ins-Pb2 at any time during the test signifying the presence of insulation which
caused the drop in temperature until the test was stopped.
iii) Average temperature on the cavity facing surface of the exposed Plasterboard
2 (Pb2-Cav)
The temperature profile of this surface was almost identical with that of Specimen 7
until 80 minutes, beyond which the third phase started with an almost linear
temperature growth rate crossing 3000C by 105 minutes (compared to 7200C by the
same time in Specimen 5 using the same insulation in the cavity). The surface
recorded a temperature of 4500C by 135 minutes near the end of the test which also
happened to be the temperature difference across the thickness of the plasterboard
demonstrating its integrity until the failure of the specimen.
iv) Average temperature on the cavity facing surface of the ambient Plasterboard
3 (Pb3-Cav)
In the absence of insulation, the transmission of heat across the cavity by radiation
was very quick, forcing the Pb3-Cav surface to heat up almost instantaneously and
trace very closely on the underside of the time-temperature profile of Pb2-Cav surface
with the maximum temperature difference between the two surfaces not exceeding
500C until the end of test.
v) Average temperature of the interface surface between base layer Plasterboard
3 and insulation (Pb3-Ins)
The temperature of this interface remained at about 1000C until almost 120 minutes,
beyond which it started rising gently. A temperature difference of 2500C was
noticeable across the thickness of the Plasterboard 3 near the end of the test signifying
the continued integrity of the plasterboard until the failure of the specimen.
vi) Average temperature of the interface surface between the insulation and
Plasterboard 4 (Ins-Pb4)
The ambient side base layer plasterboard and the rock fibre insulation continued to
give very good protection to this surface keeping its temperature below 1000C during
the test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 309
0100200300400500600700800900
100011001200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Tem
per
atu
re (
oC
)
AS 1530.4 Furnace FS Pb1-Ins Ins-Pb2
Pb2-Cav Pb3-Cav Pb3-Ins Ins-Pb4 Amb
(a) Average Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 8
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Te
mp
era
ture
(oC
)
AS 1530 Furnace FS-L FS-M FS-RPb1-Ins-L Pb1-Ins-M Pb1-Ins-R Ins-Pb2-L Ins-Pb2-MIns-Pb2-R Pb2-Cav-L Pb2-Cav-M Pb2-Cav-R Pb3-Cav-LPb3-Cav-M Pb3-Cav-R Pb3-Ins-L Pb3-Ins-M Pb3-Ins-R
(b) Time-Temperature Profiles across the Left, Middle and Right Sections of Plasterboard Surfaces in Test Specimen 8
Figure 6-74: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 8
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 310
vii) Average temperature on the ambient side of unexposed Plasterboard 4
The temperature on this surface remained below 700C until the end of the test thus
excluding the possibility of insulation failure at any stage of the fire test.
b) Steel Surfaces
The time-temperature profiles of hot flanges, webs and cold flanges of the central
studs are shown in Figures 6-75 (a) to (c) with the second phases extending until 70,
80 and 85 minutes respectively (compared to 50, 60 and 65 minutes in Test Specimen
5 using rock fibre as cavity insulation). As in the case of other specimens, the
temperature at the mid-height of the wall was found to be the maximum. Beyond 80
minutes and until the end of the test the difference in temperature in the hot flanges
along the stud lengths was found to be about 1000 – 1300C (Figure 6-75 (a))
(compared to the temperature difference of 700 – 1000C along the hot flanges of the
central studs in Test Specimen 4 beyond 70 minutes and until end of the test (Figure
6-43 (a)). The maximum recorded hot flange temperatures at 60 and 90 minutes were
1200C and 2500C, respectively (compared to 2100C and 5400C in the cavity insulated
specimen at the same times).
0
100
200
300
400
500
600
700
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Tem
per
atu
re (
oC
)
S2-T-HF S3-T-HF S2-M-HF S3-M-HF S2-L-HF S3-L-HF
(a) Time-Temperature Profiles of Hot Flange Surfaces of Central Studs in Test Specimen 8
Figure 6-75: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 8
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 311
0
100
200
300
400
500
600
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Tem
per
atu
re (
oC
)
S2-T-W S3-T-W S2-M-W S3-M-W S2-L-W S3-L-W
(b) Time-Temperature Profiles of Web Surfaces of Central Studs in Test Specimen 8
0
100
200
300
400
500
600
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Tem
per
atu
re (
oC
)
S2-T-CF S3-T-CF S2-M-CF S3-M-CF S2-L-CF S3-L-CF
(c) Time-Temperature Profiles of Cold Flange Surfaces of Central Studs in Test Specimen 8
Figure 6-75: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 8
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 312
0
100
200
300
400
500
600
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Tem
per
atu
re (
oC
)
S2-M-HF S3-M-HF S2-M-W S3-M-W S2-M-CF S3-M-CF
Figure 6-76: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 8
Figure 6-76 shows the temperature profiles across the depth of the central studs
measured at mid-height. The maximum temperature differences (HF temperature – CF
temperature) across the central studs at 60 and 90 minutes were observed to be 200C
and 1200C, respectively, compared to 1000C and 3150C in the cavity insulated Test
Specimen 5 at the same times). Towards the end of the test the temperature difference
across the central studs had dropped below 1000C.
6.5.8.3) Behaviour of Specimen
The specimen was loaded to 15 kN at ambient temperature when there were no signs
of lateral deformation. Axial deformations of the studs were seen to be about 3 mm as
shown in Figure 6-78 (a). During the fire test the studs expanded due to heat. The total
average elongation of the studs towards the end of the test was about 15 mm.
Figures 6-79 (a) to (c) give the lateral deflections of the studs at the top, middle and
lower levels. The wall was found to bow towards the furnace within a few minutes of
starting the fire test. The deflections were maximum at mid-span and started
developing faster beyond 90 minutes. Stud 2 underwent the maximum lateral
deformation of 22 mm by 114 minutes (Figure 6-79 (b)), beyond which it stopped
bowing any further until 124 minutes at which time the hot flange temperature of the
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 313
stud had reached 4700C at the centre (Figure 6-75 (a)) and the temperature difference
across the stud was 1200C (Figure 6-76). At this time, Stud 3 had reached its
maximum lateral deflection of 22 mm (Figure 6-79 (b)).
Stud 2 reversed its direction of deformation and started to straighten out from about
124 minutes. This was soon followed by Stud 3 straightening out at 127 minutes at
which time its hot flange temperature was about 4600C at centre (Figure 6-75 (a)) and
the temperature difference across the depth was 1300C (Figure 6-76). Both central
studs continued to deform rapidly in the outward direction. By 132 minutes Stud 2
had exhausted the free expansion scope provided for it in the loading frame as the
loading plate of Stud 2 came into contact with the lower beam rendering the jack non-
functional (see Figure 6-77). This meant that beyond 132 minutes and until the failure
of the specimen, the load on Stud 2 could have gone up well beyond 15 kN due to the
temperature stresses induced in the stud on the prevention of its thermal expansion.
This was soon followed by the other studs also running out of free thermal expansion
scope leading to the collapse of the frame due to increased thermal stresses by about
136 minutes. All the plasterboards were seen to be intact throughout the fire test
(Figure 6-74). As observed for the previous specimens the structural failure occurred
within 15 minutes of the first observed reversal in lateral deformation of the central
studs.
(a)
Figure 6-77: View of Loading Arrangement
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 314
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 315
(b)
Figure 6-77: View of Loading Arrangement
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0 2 4 6 8 10 12 14 1Load (kN)
Def
orm
atio
n (
mm
)
6
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Load Profiles of Test Specimen 8 at Ambient Temperature
Figure 6-78: Axial Deformation Plots for Studs of Test Specimen 3
Allowance for free thermal expansion
Jack
Loading Plate
-5
0
5
10
15
20
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Def
orm
atio
n (
mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Axial Deformation -Time Profiles of Test Specimen 8 at Elevated Temperatures
Figure 6-78: Axial Deformation Plots for Studs of Test Specimen 3
-30
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Def
lect
ion
(m
m)
Stud 2 Stud 3 Stud 4
(a) Lateral Deflection -Time Profiles of Test Specimen 8 at Upper Level
Figure 6-79: Lateral Deflection-Time Plots of Test Specimen 8
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 316
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
De
fle
cti
on
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Lateral Deflection -Time Profiles of Test Specimen 8 at Middle Level
-30
-25
-20
-15
-10
-5
0
5
10
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(c) Lateral Deflection -Time Profiles of Test Specimen 8 at Lower Level
Figure 6-79: Lateral Deflection-Time Plots of Test Specimen 8
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 317
10
11
12
13
14
15
16
17
18
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Lo
ad (
kN)
Figure 6-80: Axial Load -Time Profile of Test Specimen 8 during Fire Test
4) Specimen Report
a) Date of Test: 18/04/08
b) Severity of Test: 100%
c) Specimen Temperature
The average temperature of the unexposed surface of the test specimen at the end of
120 minutes from the commencement of the test was 570C indicating a rise of 410C
above the ambient temperature of 160C. The maximum temperature of the unexposed
surface at that time was 590C.
d) Specimen Behaviour
The fire side Plasterboards 1 and 2 remained attached to the steel frame until the
failure of the test specimen. The lateral deflection of the test specimen was initially
towards the furnace and reversed in direction at the end of 123 minutes from the
commencement of the test. The maximum deflection at mid-height of the wall at that
time was 23 mm.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 318
e) Failure Criterion
The test specimen was deemed to have failed at approximately 136 minutes from the
start of the test when the specimen could no longer sustain the applied load.
f) Performance
Performance observed in respect of the following criteria:
Structural adequacy - Failure at 136 minutes.
Integrity - No failure at 136 minutes
Insulation - No failure at 136 minutes
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 319
6.5.9: Test Specimen 9
6.5.9.1) Visual Observations
The specimen was subjected to fire for 124 minutes until the wall specimen failed.
Figure 6-81 (a) shows the specimen after the fire test. Plasterboards 1 and 2 had fully
collapsed and the cellulose fibre between them was totally burnt out (Figure 6-81 (a)).
Figure 6-81 (b) shows the outward movement of the studs caused by the local
compressive failure of the hot flanges of the studs. Studs 1, 2 and 3 were severely
damaged by the local buckling of the hot flange and web whereas Stud 4 was seen to
be unaffected (see Figure 6-81 (c)). Out of plane deformations of Studs 2 and 3 had
also ruptured the ambient side plasterboards (Figures 6-81 (d) and (e)). The tracks
were in good condition providing good support to the studs (Figure 6-81 (f)).
Remains of the external cellulose fibre insulation
Plasterboard 1
Plasterboard 2
Pb3-Cav surface
(a) Front View Showing the Collapse of External Plasterboard and Burn-out of External Insulation
(Cellulose fibre)
Figure 6-81: Test Specimen 9 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 320
Stud 2 Stud 1 Stud 3 Stud 4
(b) Side View of Wall after Removing the External Plasterboards
Stud 4 Stud 3
Stud 2 Stud 1
(b) Failure Modes of Studs 1, 2 and 3
Figure 6-81: Test Specimen 9 after the Fire Test
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 321
(d) Local Compressive Failure of Flanges and Web in Stud 2
(e) Local Compressive Failure (f) Top Track in Good of Hot Flange and Web in Stud 2 Condition
Figure 6-81: Test Specimen 9 after the Fire Test
6.5.9.2) Time-Temperature Profiles
a) Plasterboard Surfaces (see Figures 6-82 (a) and (b))
i) Average temperature of the interface surface between the exposed
Plasterboards 1 and Insulation (Pb1-Ins)
The initial response of this interface was very much similar to Specimens 7 and 8. The
first phase displayed a quick temperature rise up to 800C by the end of 5 minutes (see
Figure 6-82 (a)). The second phase of near constant temperature lasted until 18
minutes, beyond which the temperature gained sharply in the third phase reaching
4000C by 29 minutes and 5500C by 48 minutes. The rate of temperature growth was
reduced after 30 minutes. The thermocouple positioned at the centre along mid-height
between Studs 1 and 2 in this interface showed a continuous increase in its
temperature finally merging with the fire curve at about 88 minutes indicating a
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 322
breach at that time of the external plasterboard in this area (refer Figure 6-82 (b)). The
other two thermocouples measuring the interface temperature and placed at the mid-
height of the wall between the remaining studs merged with the fire curve at 97
minutes and 125 minutes, respectively.
ii) Average temperature of the interface surface between the insulation and
exposed base layer Plasterboard 2 (Ins-Pb2)
The temperature profile of this interface almost coincided with that of Pb1-Ins curve
until about 20 minutes. The temperature gained beyond 20 minutes, reaching 4000C
by 53 minutes. Figure 6-82 (b) shows the merging of Ins-Pb2-L curve with Pb1-Ins-L
at about 84 minutes indicating the total disintegration of the insulation at mid-height
region between Studs 1 and 2 at this time. The temperature profiles of Pb1-Ins-R and
Ins-Pb2-R were merged from the start of the test until the end of the test. The only
possible explanation for this would be that somehow, the two thermocouple wires
measuring the two interface temperatures must have got entangled with the hot
junctions of both touching each other resulting in identical readings. The Ins-Pb2-M
curve intersected the Pb1-Ins-M curve at about 106 minutes indicating the insulation
burn out at the mid-height of wall between Studs 2 and 3.
iii) Average temperature on the cavity facing surface of the exposed Plasterboard
2 (Pb2-Cav)
The temperature of this surface increased very slowly from the ambient and reached
approximately 1000C by 70 minutes from the start of the test (compared with 55
minutes observed in Specimen 6 using cellulose insulation in the cavity). The third
phase started beyond 70 minutes with an almost linear temperature growth rate and
reached 4000C by 108 minutes and 5000C by 120 minutes (compared with 68 minutes
and 72 minutes in Specimen 6). After 125 minutes the temperature difference across
the plasterboard thickness was approximately 4750C, which reduced sharply soon
after indicating it’s partial or full collapse.
iv) Average temperature on the cavity facing surface of the ambient Plasterboard
3 (Pb3-Cav)
As seen in Specimens 7 and 8 the temperature profile of Pb3-Cav surface traced the
Pb2-Cav profile with the lag not exceeding 750C during the test. The temperature of
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 323
the surface increased suddenly near the end of the test (125 minutes) indicating the
breaching of Plasterboard 2 (i.e. the exposed base layer plasterboard).
v) Average temperature of the interface surface between base layer Plasterboard
3 and insulation (Pb3-Ins)
The temperature of this interface remained under 1000C until the end of the test. The
temperature difference across the thickness of Plasterboard 3 was approximately
4000C towards the end of the test.
vi) Average temperature of the interface surface between the insulation and
Plasterboard 4 (Ins-Pb4)
The temperature of this interface remained under 900C for the entire duration of the
test (i.e. 125 minutes).
vii) Average temperature on the ambient side of unexposed Plasterboard 4
The surface recorded temperature values less than 700C until the end of the test.
0
100200
300
400500
600
700800
900
10001100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Tem
per
atu
re (
oC
)
AS 1530 Furnace FS Pb1-Ins Ins-Pb2Pb2-Cav Pb3-Cav Pb3-Ins Ins-Pb4 Amb
(a) Average Time-Temperature Profiles of Plasterboard Surfaces in Test Specimen 9
Figure 6-82: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 324
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Tem
per
atu
re (
oC
)
AS 1530 Furnace FS-L FS-M FS-R Pb1-Ins-L
Pb1-Ins-M Pb1-Ins-R Ins-Pb2-L Ins-Pb2-M Ins-Pb2-R Pb2-Cav-LPb2-Cav-M Pb2-Cav-R Pb3-Cav-L Pb3-Cav-M Pb3-Cav-R Pb3-Ins-L
Pb3-Ins-M Pb3-Ins-R Ins-Pb4-L Ins-Pb4-M Ins-Pb4-R
(b) Time-Temperature Profiles across the Left, Middle and Right Sections of Plasterboard Surfaces in Test Specimen 9
Figure 6-82: Time-Temperature Plots of Plasterboard Surfaces in Test Specimen 9
b) Steel Surfaces
Figures 6-83 (a), (b) and (c) give the temperature profiles for the hot flanges, webs
and cold flanges along the central studs with the second phase extending until 70, 75
and 80 minutes (compared to 50, 55 and 65 minutes in Test Specimen 6 using
cellulose fibre as cavity insulation). Temperature variations along the length were
more pronounced in the hot flanges than in the web and cold flange elements. The hot
flange temperature profiles very closely traced the Pb2-Cav temperature profile. A
maximum temperature difference of 1500C could be noted in the hot flanges along the
length of the central studs. Temperature variations along the length in the web and
cold flange elements appeared towards the end of the test. The maximum recorded hot
flange temperatures at 60, 80 and 100 minutes were 1000C, 2300C and 4200C,
respectively (compared to 2200C, 4600C and 6000C in the cellulose fibre cavity
insulated Test Specimen 6 at the same times as seen in Figure 6-60a). The maximum
temperature variation across the cross-section was observed to be about 2000C in Stud
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 325
2 at 100 minutes (compared to the temperature difference of 3300C in Stud 2 at 94
minutes for Test Specimen 6, Figure 6-61).
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Te
mp
era
ture
(oC
)
S2-T-HF S3-T-HF S2-M-HF S3-M-HF S2-L-HF S3-L-HF
(a) Time-Temperature Profiles of Hot Flange Surfaces of Central Studs in Test Specimen 9
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Tem
per
atu
re (
oC
)
S2-T-W S3-T-W S2-M-W S3-M-W S2-L-W S3-L-W
(b) Time-Temperature Profiles of Web Surfaces of Central Studs in Test Specimen 9
Figure 6-83: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 326
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 327
0
100
200
300
400
500
600
700
800
900
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Tem
per
atu
re (
oC
)
S2-T-CF S3-T-CF S2-M-CF S3-M-CF S2-L-CF S3-L-CF
(c) Time-Temperature Profiles of Cold Flange Surfaces of Central Studs in Test Specimen 9
Figure 6-83: Time-Temperature Plots of Flanges and Web Surfaces of Central Studs in Test Specimen 3
0
100
200
300
400
500
600
700
800
900
1000
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140Time (min)
Tem
per
atu
re (
oC
)
S2-M-HF S3-M-HF S2-M-W S3-M-W S2-M-CF S3-M-CF
Figure 6-84: Time-Temperature Profiles across Central Studs at Mid-Height in Test Specimen 9
6.5.9.3) Behaviour of Specimen
Each stud of the specimen was loaded at the ambient temperature conditions up to 15
kN without any signs of lateral deformations. The studs however suffered an axial
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 328
shortening of about 3.5 mm (Figure 6-85 a). On starting the fire test, the specimen
was seen to gradually start bowing towards the furnace. This lateral deformation
increased rapidly beyond 80 minutes. Both the central studs deformed more than the
end studs, with Stud 2 reaching a maximum lateral deformation of 22 mm by 110
minutes (Figure 6-86 b). At this time the temperature of the hot flange at the centre
was 5000C and the temperature difference across the stud depth was 1800C (Figure 6-
84). Beyond 110 minutes Stud 2 reversed its direction of lateral deflection and started
to straighten out. Stud 3 reached a maximum of 20 mm by the end of 120 minutes
after which it also reversed its direction of lateral deformation (Figure 6-86 (b)). At
120 minutes the central hot flange temperature in Stud 3 had reached 5300C and the
temperature difference across the cross-section was 1600C (Figure 6-84). After
reversing, the lateral deformations in both the studs progressed very rapidly, leading
to the failure of the frame at 124 minutes (Figure 6-87). The exposed Plasterboard 1
had partially fallen off about the central portion of the wall after 73 minutes as
indicated by the step in the temperature profile of Pb1-Ins-M (Figure 6-82 (b)).
However Plasterboard 2 was intact throughout the fire test offering protection to the
steel frame.
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
0 2 4 6 8 10 12 14 16
Load (kN)
Def
orm
atio
n (
mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(a) Axial Deformation -Load Profiles of Test Specimen 9 at Ambient
Temperature
9
Figure 6-85: Axial Deformation Plots for Studs of Test Specimen
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 329
-10
-5
0
5
10
15
20
25
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Def
orm
atio
n (
mm
)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Axial Deformation -Time Profiles of Test Specimen 4 at Elevated Temperatures
Figure 6-85: Axial Deformation Plots for Studs of Test Specimen 9
-30
-25
-20
-15
-10
-5
0
5
10
15
20
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Def
lect
ion
(m
m)
Stud 2 Stud 3
(a) Lateral Deflection -Time Profiles of Test Specimen 9 at Upper Level
Figure 6-86: Lateral Deflection-Time Plots of Test Specimen 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 330
-30
-25
-20
-15
-10
-5
0
5
10
15
20
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Def
lect
ion
(m
m)
Stud 1 Stud 2 Stud 3 Stud 4
(b) Lateral Deflection -Time Profiles of Test Specimen 9 at Middle Level
-30
-25
-20
-15
-10
-5
0
5
10
15
20
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Def
lect
ion
(m
m)
L1 L2 L3 L4
(c) Lateral Deflection -Time Profiles of Test Specimen 9 at Lower Level
Figure 6-86: Lateral Deflection-Time Plots of Test Specimen 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 331
10
11
12
13
14
15
16
17
18
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Lo
ad
(k
N)
Figure 6-87: Axial Load -Time Profile of Test Specimen 9 during Fire Test
6.5
st: 00%
re
the unexposed surface of the test specimen towards the 0 0
1 collapsed partially near the central portion of the wall at
.9.4) Specimen Report
a) Date of Test:
b) Severity of Te 1
c) Specimen Temperatu
The average temperature of
end of the test was 58 C indicating a rise of 42 C above the ambient temperature of
160C. The maximum temperature of the unexposed surface at that time was 610C.
d) Specimen Behaviour
The fire side Plasterboard
about 73 minutes from the start of the test, whereas Plasterboard 2 remained attached
to the steel frame until the failure of the test specimen. The lateral deflection of the
Test Specimen was initially towards the furnace and reversed in direction at the end of
110 minutes from the commencement of the test. The maximum deflection at mid-
height of the wall at that time was 22 mm. The structural failure of the frame resulted
within 15 minutes of the first stud reversing its lateral deformation.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 332
) Failure Criterion
s deemed to have failed at approximately 124 minutes from the
start of the test when the specimen could no longer sustain the applied load.
rved in respect of the following criteria:
s.
s
sio
nd plateau in the time-temperature graph of the plasterboards due to
the conversion into steam of the remaining portion (approximately 25%) of the
e a slight negative pressure
on the surface due to the rising hot gases favoring the moisture movement towards the
e
The test specimen wa
f) Performance
Performance obse
Structural adequacy - Failure at 124 minute
Integrity - No failure at 124 minute
Insulation - No failure at 124 minutes
6.6: Discus ns
The expected seco
chemically bound water present in the plasterboard was never observed, probably
because the conversion of the small quantity of bound water into steam was very
quick and the steam was able to escape across the plasterboard thickness very fast
without much of re-condensation or migration towards the ambient side (which was
possible during the formation of the first plateau) due to the already present numerous
shrinkage cracks over the entire body of the plasterboard.
The extreme heat on one side of the plasterboard can caus
hot surface, quickly removing the steam from the already gaping cracks caused due to
the plasterboard shrinkage. Also with the passage of time during the fire exposure the
total surface area of the plasterboard receiving heat from the furnace increases due to
the presence of numerous shrinkage cracks, which probably offsets the formation of
the second temperature plateau due to the conversion of the final portion of the
chemically bound water in the plasterboard.
Chapter 7: Discussions and Recommendations
In this research, the fire performance of non-load bearing wall test specimens was
studied with and without cavity insulation. Test specimens with insulation placed
outside the cavity and between external plasterboards were also tested and the fire
performance when subjected to cellulosic fire curve was studied and compared with
the conventional wall systems. Similar tests were carried out on large scale load
bearing wall test specimens subjected to a load ratio of 0.2. Effects of different types
of insulation material such as Glass fibre, Rock wool fibre and Cellulose fibre on the
thermal performance of non-load bearing and load bearing walls were studied and
compared in the earlier chapters of this thesis.
This chapter presents the outcomes of the tests performed on non-load bearing and
load bearing wall specimens with and without the use of insulating material.
7.1: Discussions
7.1.1 Comparison between conventional non-load bearing wall models and
proposed externally insulated wall models
The fire performance of the externally insulated wall specimens was considerably
better than the cavity insulated specimens as the steel stud frame in the former case
was well protected by the external layer of insulation. The temperature plateau in the
steel studs of externally insulated specimens lasted up to approximately 75 minutes
while it was 55 minutes for the cavity insulated specimens. The studs in the externally
insulated specimens not only enjoyed an extended period of almost constant
temperature, but also were benefitted by the lower temperature growth rates following
the plateau when compared with the rapid escalation of temperatures in the case of
cavity insulated specimens.
The absence of insulation in the cavity of externally insulated specimens helped to
equalize the temperatures across the stud cross-sections more rapidly through
radiation than in the cavity insulated specimens. The large temperature gradients
developed across the studs in the cavity insulated specimens forced the hot flanges to
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 333
expand much more than the cold flanges causing the walls to deflect laterally towards
the furnace. Externally insulated specimens on the other hand displayed minimum
lateral deflections during the tests.
Figures 7-1 to 7-3 compare the effects of external insulation to that of cavity
insulation. A clear separation of the time-temperature profiles of the studs in the two
kinds of wall specimens shows that externally insulated specimens took longer times
to reach the equivalent temperatures. This clearly demonstrates the resulting benefits
of placing the insulation outside the steel stud frame.
0100
200300
400500
600700
800900
10001100
12001300
0 20 40 60 80 100 120 140 160 180 200 220
Time (min)
Tem
per
atu
re (
oC
)
Sp4 HF Sp4 W Sp4 CF Sp7 HF Sp7 W Sp7 CF
Figure 7-1: Time-temperature Profiles for the Central Stud in Non-Load Bearing Wall Test Specimens 4 and 7 (Glass fibre insulation)
Note: Sp4/7 HF/W/CF: Time-temperature profile followed by the HF/W/CF of the central
stud in Test Specimen 4 (cavity insulations) and Test Specimen 7 (external insulation)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 334
0100200300400500600700800900
1000110012001300
0 20 40 60 80 100 120 140 160 180 200 220 240
Time (min)
Tem
per
atu
re (
oC
)
Sp 5 HF Sp 5 W Sp 5 CF Sp 8 HF Sp 8 W Sp 8 CF
Figure 7-2: Time-temperature Profiles for the Central Stud in Non-load Bearing Wall
Test Specimens 5 and 8 (Rock fibre insulation)
Note: Sp5/8 HF/W/CF: Time-temperature profile followed by the HF/W/CF of the central stud in Test Specimen 5 (cavity insulations) and Test Specimen 8 (external insulation)
0100200300400500600700800900
1000110012001300
0 20 40 60 80 100 120 140 160 180
Time (min)
Tem
per
atu
re (
oC
)
Sp 6 HF Sp 6 W Sp 6 CF Sp 9 HF Sp 9 W Sp 9 CF
Figure 7-3: Time-temperature Profiles for the Central Stud in Non-Load Bearing Wall
Test Specimens 6 and 9 (Cellulose fibre insulation)
Note: Sp6/9 HF/W/CF: Time-temperature profile followed by the HF/W/CF of the central stud in Test Specimen 6 (cavity insulations) and Test Specimen 9 (external insulation)
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 335
0100200300400500600700800900
1000110012001300
0 20 40 60 80 100 120 140 160 180 200 220Time (min)
Tem
per
atu
re (
oC
)
CI-GF S2-HF CI-RF S2-HF CI-CF S2-HFCP-GF S2-HF CP-RF S2-HF CP-CF S2-HF
Figure 7-4: Time-temperature Profiles for the Central Stud
Hot Flanges in Non-Load Bearing Wall Test Specimens 4 to 9
Note: CI-GF/RF/CF S2-HF: Time-temperature profile followed by the hot flange of the central stud in the specimen using glass fibres/rock fibres/cellulose fibres as cavity insulation
Amongst the three types of insulations tested, rock fibre gave the maximum
separation (see Figure 7-2). This was possibly due to its superior insulating properties
over the other two. The insulation gave the best results when placed on the outside of
studs. The lower levels of conductivity which had caused extensive damage when
placed in the wall cavity were seen to offer the best protection to the studs by
maintaining their temperatures at lower levels with minimum temperature gradients
across them. This ensures minimum lateral deflections and maximum periods of fire
endurance.
The wall specimens with cellulose fibre insulation on the other hand showed
minimum separation in the time-temperature profiles of their studs (see Figure 7-3).
This was considered to be due to the rapid disintegration of the cellulose fibre in the
wall cavity reducing it to the case of a wall specimen without the cavity insulation.
This actually helped by lowering the temperature gradients in the studs as more
energy could be dissipated into the cavity. Thus the disintegration of the cellulose
fibre in the cavity actually helped the studs to survive longer. The beneficial effect of
placing the insulation on the outside is also seen in Figure 7-3.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 336
The temperature escalations of the hot flanges in the cavity insulated and externally
insulated wall specimens are shown in Figure 7-4, which clearly brings out the
benefits of using the insulation on the outside rather than in the cavity of the wall
specimens. Among the externally insulated specimens rock fibre insulation was seen
to give the best results.
Figures 7-5 ((a) to (d)) give the hot flange temperature values of the central stud for
Specimens 3 to 9 at 60, 90, 120 and 150 minutes from the start of the test . The hot
flange temperatures of the studs in externally insulated specimens surpassed that of
the non-insulated specimen (Specimen 3) after 150 minutes. This was probably due to
the heat redirected towards the cavity by the external insulation on the ambient side.
By the end of 90 minutes the hot flange temperatures of the cavity insulated
specimens ranged from 1.5 to 3.5 times that of the externally insulated specimens.
(a) 60 minutes (b) 90 minutes
(c) 120 minutes (d) 150 minutes Figure 7-5: Hot Flange Temperatures of the Central Stud in NLB Wall Test
Specimens 3 to 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 337
Note:
CI-GF: Non-load bearing wall specimen using Glass Fibre (GF) as cavity insulation (CI)
CI-RF: Non-load bearing wall specimen using Rock Fibre (RF) as cavity insulation (CI)
CI-CF: Non-load bearing wall specimen using Cellulose Fibre (CF) as cavity insulation (CI)
CP-GF: Non-load bearing wall specimen using Glass Fibre (GF) as external insulation in
composite panels (CP)
CP-RF: Non-load bearing wall specimen using Rock Fibre (RF) as external insulation in
composite panels (CP)
CP-CF: Non-load bearing wall specimen using Cellulose Fibre (CF) as external insulation in
composite panels (CP)
Figures 7-6 ((a) to (d)) give the temperature difference across the central stud in Test
Specimens 3 to 9. For up to 120 minutes from the start of the test the temperature
difference across the externally insulated wall specimens can be seen to be much less
than the non-insulated and the cavity insulated specimens. At 90 minutes from the
start of the test the temperature difference across the cavity insulated specimens was
2.5 to 7.5 times more than the externally insulated wall specimens. The drop in the
temperature difference values of cavity insulated Test Specimen 5 (CI-RF) at about
150 minutes is probably due to the effect of the disintegration of insulation and fire
side plasterboards, thus reducing the temperature difference across the stud cross-
section.
(a) 60 minutes (b) 90 minutes
Figure 7-6: Temperature Difference across the Central Studs in NLB Wall Test Specimens 3 to 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 338
(c) 120 minutes (d) 150 minutes
Figure 7-6: Temperature Difference across the Central Studs C/S in NLB Wall Test Specimens 3 to 9
Figures 7-7 ((a) to (d)) show the temperature of the ambient surface of the external
plasterboard 2 for non-load bearing (NLB) wall Test Specimens 3 to 9. In the case of
cavity insulated Test Specimens 4 to 6 this was the temperature of the Pb2,Ins
interface and in the case of non-insulated and externally insulated Test Specimens 3
and 7 to 9, respectively, it was the Pb2-Cav temperature.
At the end of 90 minutes, the temperature of the Pb2,Ins interface is approximately 2
to 5.5 times greater than the Pb2-Cav temperature of the non-insulated or externally
insulated Test Specimens. At 150 minutes, the value in the case of Test Specimen 5 is
seen to cross 10000C due to a breach in the external platerboards.
(a) 60 minutes (b) 90 minutes
Figure 7-7: Ambient Side Temperature of External Plasterboard 2 in Test Specimens 3 to 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 339
(c) 120 minutes (d) 150 minutes
Figure 7-7: Ambient Side Temperature of External Plasterboard 2 in Test Specimens 3 to 9
(Pb2,Ins in case of cavity insulated Test Specimens and Pb2-Cav in case of non-insulated and externally insulated Test Specimens)
Figures 7-8 ((a) to (d)) show that the temperature on the ambient face of the externally
insulated test specimens is consistently lower than the cavity insulated test specimens
during the entire test.
(a) 60 minutes (b) 90 minutes
Figure 7-8: Temperature on the Ambient Face of the Non-load Bearing Wall Test Specimens 3 to 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 340
(c) 120 minutes (d) 150 minutes
Figure 7-8: Temperature on the Ambient Face of the Non-load Bearing Wall Test Specimens 3 to 9
The fall off times of Plasterboard 2 in the case of externally insulated specimens 7 and
8 was seen to be around 200 minutes, almost an hour more than that observed in
cavity insulated specimens (see Figure 7-9). This could be due to the following three
factors.
1) Extra protection offered to them by the layer of insulation on the fire side.
2) Lower stress levels in the plasterboard due to smaller lateral and axial deformations
as compared to cavity insulated specimens.
3) Lower rate of temperature growth in the plasterboard on account of dissipation of
heat into the cavity as opposed to the accelerated temperature growth due to the
redirected heat from the insulation in the cavity insulated specimens.
Figure 7-9: Fall off Times of Plasterboard 2 in Non-Load Bearing Wall Test Specimens 4 to 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 341
The fall off times of Plasterboard 2 is very critical as its collapse would expose the
steel frame directly to fire and commence the failure of the wall. The thermal
insulation property of the externally insulated walls was also found to outperform the
cavity insulated specimens. The temperatures of the cavity facing surface of
Plasterboard 2 in the case of externally insulated specimens was seen to be one third
to one half of the temperatures of the corresponding face in the cavity insulated
specimens. This was the major factor responsible for delaying, the fall off time of
Plasterboard 2 in the case of externally insulated specimens. The ambient side
temperatures of the externally insulated specimens were also seen to be consistently
lower than those of cavity insulated specimens at any given time (see Figure 7-10).
0
20
40
60
80
100
120
140
160
180
0 20 40 60 80 100 120 140 160 180 200 220
Time (min)
Tem
per
atu
re (
oC
)
CI-GF CI-RF CI-CF CP-GF CP-RF CP-CF
Figure 7-10: Ambient Side Time-temperature Profiles of Test Specimens 4 to 9
Note: CI-GF/RF/CF: Test Specimen using glass fibres/rock fibres/cellulose fibres as cavity insulation
After 3 hours of test the ambient side temperatures of the cavity insulated specimens
had exceeded 1000C whereas the temperatures of the externally insulated specimens
were around 800C. The lowest temperatures were recorded by Specimen 8 using rock
fibre as external insulation. All of these observations imply that the new wall system
with external insulation is likely to provide improved performance under the three fire
rating criteria of stability, integrity and insulation.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 342
7.1.2 Comparison between conventional load bearing wall models and proposed
externally insulated wall models.
Figures 7-11 and 7-12 show the thermal responses of the central studs in the cavity
insulated load bearing wall specimens along with the externally insulated load bearing
wall specimens using rock fibre and cellulose fibre insulation, respectively. Graphs
comparing the performance of cavity insulated and externally insulated specimens
using glass fibre as insulating material (Specimens 4 and 7 respectively) could not be
made as the furnace malfunctioned during the testing of Specimen 7. The graphs were
drawn using the average temperature values of the central studs at mid-height of the
wall.
In the case of cavity insulated specimens, the temperature plateau (second phase) of
the hot flanges was seen to last only up to 40 minutes in comparison to 65 minutes in
the externally insulated specimens. The hot flange temperatures of cavity insulated
specimens were seen to rise very rapidly with large temperature differences across the
stud cross-sections due to the presence of insulation between the flanges in the wall
cavity. The hot flange temperatures of externally insulated wall specimens on the
other hand were seen to rise gradually with a small temperature difference across the
stud cross-sections due to the faster transfer of heat by radiation across the empty
cavity.
0
100
200
300
400
500
600
700
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140Time (min)
Tem
per
ature
(oC
)
Sp 5 HF Sp 5 W Sp 5 CFSp 8 HF Sp 8 W Sp 8 CF
Figure 7-11: Average Time-temperature Profiles for the Central Studs in Load
Bearing Wall Test Specimens 5 and 8
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 343
0
100
200
300
400
500
600
700
800
900
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Time (min)
Tem
per
atru
e (o
C)
Sp 6 HF Sp 6 W Sp 6 CFSp 9 HF Sp 9 W Sp 9 CF
Figure 7-12: Average Time-temperature Profiles for the Central Studs in Load Bearing Wall Test Specimens 6 and 9
Note:
Sp 5/6/8/9 HF: Average time-temperature profile followed by the hot flanges of the central studs in Test Specimens 5/6/8/9
Sp 5/6/8/9 W: Average time-temperature profile followed by the webs of the central studs in Test Specimens 5/6/8/9
Sp 5/6/8/9 CF: Average time-temperature profile followed by the cold flanges of the
central studs in Test Specimens 5/6/8/9
Figures 7-13 ((a) to (d)) give the average hot flange temperature values for the central
studs of load bearing wall Test Specimens 3 to 9 at 60 , 80, 100 and 120 minutes from
the start of the test. The hot flange temperatures of the externally insulated specimens
are noted to be substantially lower than the hot flange temperatures of non-insulated
and cavity insulated test specimens. At 80 minutes from the start of the test, the
externally insulated specimens had hot flange temperatures less than half of the
temperatures recorded by the cavity insulated specimens. As the non-insulated and the
cavity insulated test specimens had collapsed before 120 minutes, their values are not
displayed in Figure 7-13 (d).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 344
(a) 60 minutes (b) 80 minutes
(c) 100 minutes (d) 120 minutes
Figure 7- 13: Average Hot Flange Temperatures of the Central Studs of Load
Bearing Wall Test Specimens 3 to 9
Figures 7-14 ((a) to (d)) give the average temperature difference across the central
studs for load bearing wall Test Specimens 3 to 9. The temperature differences across
the studs in the externally insulated test specimens were seen to be very much lower
than the values observed in the cavity insulated test specimens during the entire fire
test. The non-insulated Test Specimen 3 had values intermediate to those of cavity
insulated and externally insulated test specimens at 80 minutes from the start of the
test. The temperature difference values for the cavity insulated specimens were 3 to 5
times higher than the values of the externally insulated specimens, whereas those
values for the non-insulated specimen were about 1.5 times higher.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 345
(a) 60 minutes (b) 80 minutes
(c) 100 minutes (d) 120 minutes
Figure 7-14 : Average Temperature Difference across the Central Studs for Load
Bearing Wall Test Specimens 3 to 9
The higher temperature differences across the stud cross-sections in the cavity
insulated specimens led to higher lateral deformations as compared to the externally
insulated specimens. Figures 7-15 and 7-16 show the temperature differences across
the stud cross-sections and the corresponding lateral deformations (LD) for the cavity
insulated and externally insulated specimens using rock fibre and cellulose fibre
insulations, respectively. The lateral deformations were seen to be proportional to the
temperature difference across the stud cross-sections.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 346
0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Time (min)
Tem
p D
iffe
ren
ce (
HF
-CF
) (
oC
)
0
5
10
15
20
25
30
35
40
Lat
eral
Def
orm
atio
n (
mm
)
(HF-CF) CI RF (HF-CF) CP RF LD-CI-RF LD-CP-RF
Figure 7-15: Average Temperature Difference across Central Studs and their Lateral Deformations (LD) versus Time for Test Specimens 5 and 8
0
50
100
150
200
250
300
350
400
450
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Tem
p D
iffe
ren
ce (
HF
-CF
) (
oC
)
0
5
10
15
20
25
30
35
40
45
La
tera
l D
efo
rma
tio
n (
mm
)
(HF-CF) CI CF (HF-CF) CP CF LD-CI-CF LD-CP-CF
Figure 7-16: Average Temperature Difference across Central Studs and their Lateral Deformations versus Time for Test Specimens 6 and 9
Figure 7-17 compares the average lateral deformation of the central studs at the end of
60, 80, 100 and 120 minutes from the start of the test for Test Specimens 3 to 9. At 60
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 347
minutes from the start of the test, the lateral deformations in the cavity insulated
specimens was close to 16 mm as compared to about 9 mm in the case of externally
insulated specimens and 10 mm in the non-insulated Test Specimen 3. By the end of
80 minutes, the lateral deformation in the cavity insulated specimens had reached 30
mm, whereas it was still less than 11 mm for the externally insulated specimens and
14 mm for the non-insulated test specimen The lower lateral deformations of Test
Specimens 3 and 5 at 100 minutes as compared to their values at 80 minutes is on
account of the reversal of lateral deformation leading to progressive failure of the test
specimens.
(a) 60 minutes (b) 80 minutes
(c) 100 minutes (d) 120 minutes
Figure 7- 17:Average Lateral Deformations of the Central Studs in Load Bearing Wall Test Specimens
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 348
The average hot flange and cold flange temperature profiles of the central studs for
Test Specimens 4 to 9 are shown in Figures 7-18 and 7-19, respectively. A distinct
separation in the time-temperature profiles of the studs in the two kinds of wall
specimens shows that the studs of the externally insulated wall specimens are well
protected and thus take longer time to reach the temperatures attained by the studs in
the cavity insulated specimens.
The hot flange and cold flange temperature profiles of the cavity insulated specimens
show that the thermal response of the studs is identical for rock fibre and cellulose
fibre insulation whereas for glass fibre insulation the stud temperatures are marginally
higher. This signifies a low influence on the stud temperatures by the material of
insulation used in the cavity. This is probably because, in the cavity insulated
specimens, the insulation is on the ambient side of the hot flange and thus incapable
of offering any protection to it.
In the case of externally insulated specimens, it is seen that the temperature profiles of
the studs are well separated implying the effect of the material of insulation on the
stud temperatures. From Figures 7-18 and 7-19 it is clear that rock fibre insulation
offers the maximum protection to the studs.
0
100
200
300
400
500
600
700
800
900
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Time (min)
Tem
per
atru
e (o
C)
Sp 4 HF Sp 5 HF Sp 6 HF Sp 8 HF Sp 9 HF
Figure 7-18: Average Time-temperature Profiles of Hot Flanges for the Central Studs in Test Specimens 4 to 9
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 349
0
50
100
150
200
250
300
350
400
450
500
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140Time (min)
Tem
per
atu
re (
oC
)
Sp 4 CF Sp 5 CF Sp 6 CF Sp 8 CF Sp 9 CF
Figure 7-19: Average Time-temperature Profiles of Cold Flanges for the Central Studs in Test Specimens 4 to 9
Figure 7-20: Temperature Difference across the Central Studs in Cavity Insulated and Externally Insulated Specimens
The temperature difference between the hot and cold flanges of the central studs with
the passage of time during the fire test is shown in Figure 7-20. The graphs for all the
cavity insulated specimens using different types of insulation are seen to lie in a very
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 350
narrow band signifying the very low influence of the material of insulation on the
building up of cold flange temperatures. This is probably because any insulation by
virtue of its physical presence essentially serves the main function of eliminating the
transfer of heat across the wall cavity by radiation and convection which essentially
are the faster modes of heat transfer as compared to conduction. No cavity insulation
can reduce the transfer of heat towards the cold flange by conduction along the
metallic cross-section of the stud. Thus the cold flange picks up heat from the hot
flange by conduction along the web, which would be the fastest mode of heat transfer
in the case of cavity insulated specimens. Because of the very low conductivity of the
cavity insulating material as compared to steel, most of the heat gets directed and
channeled along and across the steel studs which act as the heat sink thus raising their
body temperatures much faster than in the case of non-cavity insulated specimens.
This makes the very presence of cavity insulation a threat to the survival of steel
during fire conditions.
Externally insulated specimens on the other hand can offer a much higher level of
protection to the studs as they are installed on the fire side of the studs thus
minimizing the transfer of heat by radiation (by virtue of their physical presence) and
conduction (on account of their low conductivity). Hence the quality of insulation
used externally would directly influence the level of fire protection offered to the
studs. Rock fibre insulation when used externally was seen to give the maximum
protection.
Figure 7-21 shows the average temperature profiles on the ambient surface of the
exposed base layer plasterboard 2 for the load bearing wall specimens. The profiles of
the cavity insulated specimens were seen to lie in a very narrow band with
temperature profiles well above the externally insulated and non-insulated wall
specimens. The cavity insulation caused the temperature profiles to rise sharply by
blocking and redirecting the heat flow back on to the cavity facing surface. The
temperature profiles of the externally insulated wall specimens were found to be the
most favorable. The externally insulated specimen using rock fibre as insulation gave
the best results.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 351
0
100
200
300
400
500
600
700
800
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150Time (min)
Tem
per
atu
re (
oC
)
LBW-CI-GF LBW-CI-RF LBW-CI-CFLBW-No Ins LBW-CP-RF LBW-CP-CF
Figure 7-21: Average Time-temperature Profiles of Pb2-Cav Surface in Test Specimens 3 to 9
Note: LBW: Load bearing wall CI: Cavity Insulated
CP: Composite Panel (Externally insulated)
Figures 7-22 (a) to (d) show the temperature profiles at 60, 80, 100 and 120 minutes,
respectively. At 80 minutes the temperature profiles of non-insulated and cavity
insulated test specimens ranged from 2 to 4 times that of externally insulated test
specimens. At 100 minutes (Figure 7-22 (c)) the plasterboard temperature of cavity
insulated specimens approached 7000C, whereas, it was below 3500C in the externally
insulated specimens. As the non-insulated and cavity insulated specimens collapsed
before 120 minutes, only the values of the externally insulated wall specimens is
shown in Figure 7-22 (d).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 352
(a) 60 minutes (b) 80 minutes
(c) 100 minutes (d) 120 minutes
Figure 7- 22: Average Time-Temperature Profiles on the Ambient Side of the
Exposed Base Layer Plasterboard 2 in LBW Test Specimens 3 to 9
The ambient side temperatures of all the wall specimens were observed to be under
700C and well below the insulation failure temperature (see Figures 7-23 (a) to (d)).
The failure of the specimens was always by the structural failure of the studs and
never by insulation or integrity failure. In some of the cavity insulated specimens, the
external plasterboards collapsed prior to stud failure thus hastening the collapse of the
wall by exposing the steel frame to direct furnace heat.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 353
(a) 60 minutes (b) 80 minutes
(c) 100 minutes (d) 120 minutes
Figure 7- 23: Ambient Side Temperatures of LBW Test Specimens 3 to 9
Table 7-1 gives the failure times of the load bearing wall specimens tested under a
constant load of 15 kN on each stud during the fire test. Only the failure times of Test
Specimens 3 to 9 can be compared as they have two plasterboards on either side of the
steel frame. The failure times of the cavity insulated specimens was noted to be less
than the failure time of Test Specimen 3 (without cavity insulation), however, the
failure times of externally insulated specimens was found to be the maximum. The
failure time of Test Specimen 7 cannot be considered as the furnace had
malfunctioned for some time during the test. The failure times of Test Specimens 8
and 9 could have been higher if the studs were allowed free thermal expansion near
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 354
the end of the test without additional load and stresses building up in the studs due to
the closure of the expansion gaps provided in the loading mechanism of the jacks.
Table 7-1: Failure Times of Test Specimens
Specimen No.
Configuration Test Insulation Ave. HF Temp. *
Failure Time (min)
01 Ambient None -
02 Fire None 48
03 Fire None 561 111
04 Fire Glass Fibre (Cavity
Insulation)
671 101
05
Fire Rock Fibre (Cavity
Insulation)
642 107
06
Fire Cellulose Fibre
(Cavity Insulation)
720 110
07 Fire Glass Fibre (External
Insulation)
522 181
08 Fire Rock Fibre (External
Insulation)
524 136
09 Fire Cellulose Fibre
(External Insulation)
610 124
*Average hot flange temperature at the centre for the middle studs at failure
For cavity insulated specimens the average hot flange temperature at the time of wall
failure was observed to be 6770C, whereas for the externally insulated specimens the
average hot flange temperature at wall failure was 5670C. Table 7-2 shows the stud
reversal times signifying the local buckling of hot flanges along with the
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 355
corresponding hot flange temperature. The average hot flange temperature for the
cavity insulated specimens initiating the local buckling of hot flange and reversal in
lateral deformation was noticed to be approximately 5700C. The higher temperature of
7060C in Stud 3 of Test Specimen 6 was not considered as it appears to be faulty. The
studs in Test Specimens 8 and 9 were seen to fail at an average hot flange temperature
of approximately 5000C slightly lower than the cavity insulated specimens, probably
because of the increased load on the individual studs after the closure of the free
thermal expansion gap in the loading mechanism.
Table 7-2: Stud Reversal Times for Cavity Insulated and Externally Insulated Specimens along with the Corresponding Temperatures
Sp.
No.
Specimen Insulation Reversal time
Stud 2
Reversal time
Stud 3
HF temp.
Stud 2
HF temp.
Stud 3
4
Glass Fibre
(Cavity Insulation)
85 95 582 566
5
Rock Fibre
(Cavity Insulation)
92 103 551 570
6
Cellulose Fibre
(Cavity Insulation)
96 107 582 706
7
Glass Fibre
(External Insulation)
174 179 - 500
8
Rock Fibre
(External Insulation)
123 127 468 457
9
Cellulose Fibre
(External Insulation)
110 120 529 523
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 356
7.2: Simplified Method for the Determination of Failure Times of Wall
Specimens
A graphical method is developed to yield approximate failure times of load bearing
wall systems. The wall systems considered are non-insulated, cavity insulated and
externally insulated wall specimens using glass fibre, rock fibre or cellulose fibre as
the insulating material.
The failure time of a load bearing wall system with the studs assumed to be
effectively restrained laterally about the minor axis depends upon the load ratio, the
type and placement of the insulation used in the wall system, the critical temperature
at which the studs undergo local buckling and lateral deflection. To help determine
the critical temperature the results from Chapter 3 are used in which the reduced
mechanical properties of cold-formed steel at elevated temperatures are presented.
Figures 7-24 shows the reduction in yield strength of 1.15 mm G500 cold-formed
steel at elevated temperatures, determined based on the 0.2% proof stress method
(Refer Eqs. 1(a) to (c) in Chapter 3). The reduction in yield strength occurs from a
temperature of about 3000C at a steady rate with a constant gradient and reaches about
12% of its original value by 6000C.
Figure 7-24: Variation of Yield Strength Reduction Factor of 1.15 mm G500
Steel with respect to Temperature
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 357
The load bearing wall specimens tested were subject to a load of 15 kN per stud
giving a load ratio of 0.2. Assuming the cross-sectional areas of the studs to remain
constant during the test, the load ratio at failure can be considered equivalent to a
strength reduction factor of 0.2 giving a corresponding critical temperature of
approximately 5650C from Figure 7-24. Load ratio at the time of failure is equal to the
ratio of yield load at elevated temperature to the yield load at ambient temperature,
which is equivalent to the strength reduction factor (ratio of the yield stress at elevated
temperature to the yield stress at ambient temperature) considering the cross-sectional
area to remain constant during the test (assuming full yielding).
Figure 7-25 shows the development of the maximum hot flange temperature at the
mid-height of the central studs of different wall specimens. The development of these
hot flange temperature profiles has been broken down into linear segments, which
closely approximate the temperature profiles observed in the experiments.
Figure 7-25: Idealized Hot Flange Temperatures of Load Bearing Test
Specimens 2 to 9
Note:
NI-1x1: Test Specimen 2 – Large scale non-insulated load bearing wall specimen
lined on both sides by a single layer of plasterboard.
NI-2x2: Test Specimen 3 – Large scale non-insulated load bearing wall specimen
lined on both sides by two layers of plasterboard.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 358
CI-GF: Test Specimen 4 – Large scale load bearing wall specimen lined on both sides
by two layers of plasterboard with Glass Fibre used as cavity insulation.
CI-RF: Test Specimen 5 – Large scale load bearing wall specimen lined on both sides
by two layers of plasterboard with Rock Fibre used as cavity insulation.
CI-CF: Test Specimen 6 – Large scale load bearing wall specimen lined on both sides
by two layers of plasterboard with Cellulose Fibre used as cavity insulation.
CP-RF: Test Specimen 8 – Large scale load bearing wall specimen lined on both
sides by two layers of plasterboard with Glass Fibre used as external insulation.
CP-CF: Test Specimen 9 – Large scale load bearing wall specimen lined on both
sides by two layers of plasterboard with Cellulose Fibre used as external insulation.
Following equations represent the portion of the temperature-time graph for
temperature values ranging from 1000C to 8000C for the seven test specimens listed
above excluding Test Specimen 8. (‘T’ is the temperature in degree Celsius of the hot
flange at mid-height of the central studs reached in time ‘t’ measured in seconds)
NI-1x1: T = 13.46 t – 75.0 Eq. 1
NI-2x2: T = 8.75 t – 320.0 Eq. 2
CI-GF: T = 12.743 t - 511.5 Eq. 3
CI-RF: T = 10.48 t - 401.1 Eq. 4
CI-CF: T = 10.48 t - 401.1 Eq. 5
CP-RF: T = 6.25 t - 306.2 Eq. 6
CP-CF: T = 8.85 t - 475.7 Eq. 7
From Figure 7-25, approximate failure times for the hot flange of each type of wall
specimen can be obtained using the critical temperature corresponding to the load
ratio. The development of hot flange failure times is shown in Figure 7-26 for a load
ratio of 0.2.
Table 7-3 compares the predicted hot flange failure times with the actual hot flange
failure times. The graph gives the times at which the hot flange of the stud begins to
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 359
buckle locally leading to reversal in lateral deformation. The reversal in lateral
deformation subsequently leads to failure of the stud cross-section.
(a) Determination of Critical Temperature for a given Load Ratio or Strength
Reduction Factor
(b) Determination of Stud Failure Times for Various Wall Specimens for a given
Critical Temperature
Figure 7-26: Development of Hot Flange Failure Times for a Given Load Ratio
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 360
A close agreement is noticed between the predicted and actual values. The failure
time of the entire stud is slightly higher (approximately 15 minutes) than the hot
flange buckling failure of the individual stud due to the post-buckling strength of the
stud and the redistribution of forces in the studs within the frame.
The failure time of Test Specimen 7 has not been considered as there were difficulties
encountered in maintaining the heating profile in the furnace due to some mechanical
problems. Also in the case of Test Specimens 2 and 3, the end conditions of the studs
at the top were different from those of Test Specimens 4 to 9, and hence the actual hot
flange failure times could not be compared. The actual local buckling of studs in Test
Specimens 7 and 8 was found to be earlier than the predicted values as the studs
towards the end of the test could not expand freely as assumed in the predicted values
due to the closure of the expansion gaps leading to increased thermal strains and
consequently increased load ratio.
Table 7-3: Comparison of Predicted Hot Flange (HF) failure Times of Load Bearing Wall Specimens with Actual Local Buckling of HF (minutes) at a Load
Ratio of 0.2
Test Specimen
No.
Wall Specimen
Predicted Hot Flange
Failure Time (minutes)
Actual Local Buckling of HF (minutes)
indicated by reversal in lateral deformation
of stud
Wall Failure Time
(minutes)
2 NI-1x1 48 - 53
3 NI-2x2 101 - 111
4 CI-GF 85 85 101
5 CI-RF 92 92 107
6 CI-CF 92 96 110
7 CP-GF - - -
8 CP-RF 141 123 136
9 CP-CF 119 110 124
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 361
Figure 7-27: Determination of Hot Flange Failure Times using Load Ratio
Figure 7-27 shows a relationship between the load ratio and the HF failure times. The
graph was developed by combining Figures 7-24 and 7-25. The failure time can be
obtained as the corresponding co-ordinate for the required load ratio. For example, a
load ratio of 0.2 yields the HF failure times for the different specimens as displayed in
column 3 of Table 7-3. The intercept on the ‘X’ axis gives the failure times for non-
load bearing walls. The change in gradient below the load ratio of 0.1 in Figure 7-27
is on account of the reduced rate of reduction in the yield strength of steel as seen in
Figure 7-24. For a load ratio of 0.03 the corresponding critical temperature as
obtained from Figure 7-24 is 8000C. The critical temperature for non-load bearing
walls is assumed to lie in the range of 800 to 8500C as the walls although treated as
non-load bearing will still be carrying their own self weight, which comprises of
plasterboards (weighing 13 kg/m2), insulation of varying density and the steel frame
yielding a load ratio between 0.003 and 0.03. Considering 8000C as the critical
temperature the failure times for non-load bearing wall specimens is as obtained by
the ‘X’ intercepts of the graphs in Figure 7-27. Similarly the gradient of the graph in
Figure 7-27 is very small from a load ratio of 1 to 0.9 as the yield strength of steel
hardly changes up to 3000C (refer Figure 7-24). The wall collapses instantly (i.e. at t =
0) at ambient temperature when the load reaches the ultimate load bearing capacity of
wall (i.e. LR = 1).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 362
Table 7-4: Comparison of Predicted Failure Times of Non-load Bearing Wall Specimens with their Actual Failure Times.
Test Specimen
No.
Wall Specimen
Predicted stud failure
time (minutes)
Time taken for HF
temperature to reach 8000C
Wall failure Remarks
4 CI-GF 103 125 125 LB of HF
5 CI-RF 115 130 145 LB of HF
6 CI-CF 115 143 145 LB of HF
7 CP-GF - 185 198 Pb2 fall off
8 CP-RF 177 195 200 Pb2 fall off
9 CP-CF 144 158 163 Pb2 fall off
Table 7-4 gives a comparison of the predicted failure times of non-load bearing wall
specimens with the time taken for the hot flange of the central stud to reach a
temperature of 8000C and the actual failure time of the wall specimen. Test
Specimens 2 and 3 have not been included as conclusive stud failure times could not
be established in the experimental work. For Test Specimens 4, 5 and 6 the local
buckling of the central stud was observed when the temperature of the hot flange was
in the vicinity of 8000C leading to the failure of the wall. The local buckling of the hot
flange is characterized by the reversal in lateral deformation of the central stud. In the
case of Test Specimens 7, 8 and 9, Plasterboard 2 was observed to fall off when the
temperature of the hot flange was in the vicinity of 8000C. This was probably because,
even though there was no observed reversal in lateral deformation of the studs (as the
temperature difference across the stud cross-section was almost uniform and there was
very little lateral buckling) the local deformation of the cross section would easily
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 363
lead to the tearing of the weakened plasterboard. Hence in the case of both cavity
insulated specimens and externally insulated specimens the failure of non-load
bearing wall specimens occurred with the temperature of the hot flange reaching a
temperature of approximately 8000C.
The effects of lateral deformation are considered to be very small in the case of
externally insulated test specimens as their lateral deformations are minimal due to a
more or less uniform temperature across the stud cross-section. However, in the case
of cavity insulated test specimens the lateral deformations would result in additional
stresses due to the moments generated by the developing eccentricity along with the
secondary moments due to P-delta effect. Effect of lateral deformation has not been
considered in the development of hot flange failure times.
To account for all these effects ABAQUS finite element program was run using as
input: The temperatures of the hot flange, web and cold flange at mid-height of the
central studs along with their lateral deformations with respect to time when subjected
to the cellulosic fire curve. The change in the modulus of elasticity of steel along with
its yield strength across the depth of the cross-section due to temperature variation
was also accounted. The stud was assumed to be laterally restrained by plasterboards
on either side with a screw spacing of 300 mm along the length of the stud. The end
conditions were assumed to be pinned.
To study the stud failure, the temperature distribution across the mid-height of the
central studs was used as input at intervals of 30 minutes along with the
corresponding lateral deformation. For each input the program was run to yield a
failure load for that particular temperature variation, lateral deformation and time.
Figure 7-28 shows the variation of load ratio with respect to hot flange temperature
obtained by several runs of the program at specified intervals of time (Gunalan 2010).
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 364
Figure 7-28: Load Ratio vs Critical Hot Flange Temperatures at Stud Failure
Figures 7-29 to 7-35 show the comparison of the graph as obtained from the program
with the one plotted using the material strength factor alone ignoring the effect of
eccentricities induced by lateral deformation and varying material properties across
the depth of the cross-section. The intersection of zero eccentricity graphs with the
time axis is considered to correspond with the failure of the non-load bearing wall
specimens.
Simple linear equations (see Table 7.5) have been proposed to predict the stud failure
times based upon the experimental results and the graphs drawn using the Abacus
program, and are graphed as dotted lines in Figures 7-29 to 7-35. Figure 7-36 shows
the comparison between the graphical representations of the proposed linear equations
for all the tested wall models, to predict the stud failure times.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 365
Figure 7-29: Load Ratio Vs Stud Failure Times for Test Specimen 2
Figure 7-30: Load Ratio Vs Stud Failure Times for Test Specimen 3
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 366
Figure 7-31: Load Ratio Vs Stud Failure Times for Test Specimen 4
Figure 7-32: Load Ratio Vs Stud Failure Times for Test Specimen 5
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 367
Figure 7-33: Load Ratio Vs Stud Failure Times for Test Specimen 6
Figure 7-34: Load Ratio Vs Stud Failure Times for Test Specimen 8
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 368
Figure 7-35: Load Ratio Vs Stud Failure Times for Test Specimen 9
The Abacus graphs in Figures 7-29 to 7-35 are seen to intersect the zero eccentricity
graphs between a load ratio of 0.5 and 0.4, implying the reducing influence of
eccentricity with the lowering of failure load. Beyond the intersection the gap in the
graph probably denotes the interval of time between the time at reversal in lateral
deformation of the critical stud and the complete failure of the stud. The hot flange
buckling times and the complete stud failure times are compared with experimental
results in Table 7.6
Figure 7-36: Load Ratio Vs Stud Failure Times for all Test Specimens using
Predictive Equations
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 369
The failure times for the Test Specimens can be represented by the following
equations:
Table 7-5: Predictive Equations for obtaining Stud Failure Times for Different Wall Models
Test Specimen Predictive Equation Range Eq. No.
NI-1x1: T = -150(LR) + 150
T = -55.5(LR) + 65
0.9 ≤ LR ≤ 1.0
0 ≤ LR ≤ 0.9
Eq. 8a
Eq. 8b
NI-2x2: T = -869.5(LR) + 869.5 0.954 ≤ LR ≤ 1.0 Eq. 9a
T = -92.24(LR) + 128 0 ≤ LR ≤ 0.954 Eq. 9b
CI-GF: T = -403.2(LR) + 403.2 0.876 ≤ LR ≤ 1.0 Eq. 10a
T = -60.5(LR) + 103 0 ≤ LR ≤ 0.876 Eq. 10b
CI-RF: T = -454.54(LR) + 454.54 0.89 ≤ LR ≤ 1.0 Eq. 11a
T = -73(LR) + 115 0 ≤ LR ≤ 0.89 Eq. 11b
CI-CF: T = -449.4(LR) + 449.4 0.9 11≤ LR ≤ 1.0 Eq. 12a
T = -82.32(LR) + 115 0 ≤ LR ≤ 0.911 Eq. 12b
CP-RF: T = -1111(LR) + 1111 0.946 ≤ LR ≤ 1.0 Eq.13a
T = -123.67(LR) + 177 0 ≤ LR ≤ 0.946 Eq.13b
CP-CF: T = -576.9(LR) + 576.9 0.896 ≤ LR ≤ 1 Eq. 14a
T = -93.75(LR) + 144 0 ≤ LR ≤ 0.896 Eq. 14b
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 370
Table 7.6 : Table Comparing the Actual Stud Reversal Times and Wall Failure Times with the Predicted Values
Test Specimen
No.
Wall Specimen
Predicted Hot Flange
Failure Time
(minutes)
Reversal in Lateral
Deformation of Stud (minutes)
Predicted Stud
Failure Time
(minutes)
Wall Failure Time
(minutes)
2 NI-1x1 48 - 52 53
3 NI-2x2 101 - 107 111
4 CI-GF 85 85 91 101
5 CI-RF 92 92 101 107
6 CI-CF 92 96 106 110
7 CP-GF - - -
8 CP-RF 141 123 157 136
9 CP-CF 119 110 125 124
The predicted stud failure times correlate well with the actual wall failure times for all
the tested wall models. The reversal in lateral deformations is also predicted very
accurately by the predicted hot flange failure times.
7.3 Essential Points to Consider for Thermal Modeling
The numerous fire tests carried out on plasterboards, composite panels, non-load
bearing walls and load bearing wall specimens has helped in formulating certain
important assumptions or essential factors to be considered in the thermal modeling of
the stud wall systems.
1) The time of exposure to the cellulosic fire curve determines the approximate
depth up to which the free and chemically bound water present in the gypsum
plasterboard gets expelled. On average, 1 minute of fire exposure is required to
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 371
expel water from 1 mm thickness of plasterboard. Hence in the case of 13 mm thick
plasterboard exposed to standard time-temperature curve from one side, the
temperature on the ambient surface would be maintained at about 1000C up to 13
minutes and in the case of 16 mm plasterboard it would be maintained for up to 16
minutes.
2) The paper on the exposed face of the plasterboard lasts only for 3 to 4 minutes.
3) After the calcination of the plasterboard, the temperature can be assumed to drop
linearly across the thickness of the plasterboard from the exposed face to the
unexposed face.
4) Interfaces between plasterboards do not influence the linearity of the temperature
variation across the layers of the plasterboards, however, when 16 mm plasterboards
are used, the duration of the second phase in the temperature profile of the ambient
side is seen to extend by approximately 30 minutes per interface.
5) A temperature gradient of 40 degrees per mm thickness can be assumed for the
linear variation across a single layer of plasterboard. However, if two layers are used
the gradient can be assumed to be 26 (degrees/mm) and for three layers a
temperature gradient of 19 (degrees/mm) can be assumed after the complete
calcination of the plasterboards.
6) When three layers of plasterboard are used, the exposed layer (Pb1) should be
assumed as ineffective from 150 minutes onwards in a thermal model.
7) The thermal performance of glass fibre insulated composite panels can be
assumed to remain unchanged regardless of the thickness and density of insulation
used.
8) Regardless of insulation thickness and density, Glass fibre insulation used in
composite panels can be dropped from the thermal model at approximately 90
minutes from the start of the test as at around this time its temperature reaches 700oC
and starts to disintegrate rapidly.
9) Thermal performance of Rock fibre insulations of varying thickness in the
thermal modeling of composite panels can be assumed to be practically same.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 372
10) Plasterboard 1 in the composite panels using rock fibre as insulation should be
removed from the thermal model when the temperature of the interface (Pb1-Ins)
reaches 900oC.
11) In the case of non-load bearing wall specimens with a single layer of plasterboard
on either side of the steel frame the influence of plasterboard joints on the insulation
failure of the wall specimens can be ignored in the thermal modeling as the joints on
the exposed plasterboards have little or no effect on the temperature profiles of the
ambient side plasterboards (see Figure 5-21).
12) The failure of 1x1 NLB wall assemblies is entirely due to the inadequacy in
insulation and not due to loss of integrity or structural stability, hence collapse or
failure of exposed plasterboards need not be considered in the thermal modeling for
such wall assemblies (see Figures 5-16 & 5-17).
13) A difference of less than 250C in the hot flange temperatures of corresponding
end studs in Specimens 1 and 2 (see chapter 5) at the time of failure implies that the
extra heating of the central stud in Specimen 2 due to the joint does not much
influence the heating of the end studs (see Figures 5-18 &5-20), hence the lateral
transmission of heat in the plane of the cavity from central stud to the end studs can
be neglected in thermal modeling.
14) Central studs should be considered for modeling as they show higher
temperatures at any time than the end studs.
15) Effect of plasterboard joints on the lateral deflection of the studs can be ignored
in the thermal models (see Figure 5-23).
16) As joints in plasterboard do not influence the failure of 1x1 non-load bearing wall
specimens, it is reasonable to ignore the effect of such joints in the thermal modeling
of 2x2 wall specimens especially when the joints are staggered.
17) In the case of 2x2 non-load bearing wall specimens without cavity insulation, the
exposed Plasterboards 1 & 2 can be assumed to remain intact and effective until the
end of the test.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 373
18) In the case of 2x2 non-load bearing wall specimens without cavity insulation, the
studs can be assumed to have uniform temperature across the cross-section.
19) The exposed Plasterboards 1 & 2 in the case of 2x2 cavity insulated non-load
bearing specimens need to be removed in the thermal model at a certain stage of the
test. Plasterboard 1 can be assumed to be ineffective when the temperature on the
ambient side of Plasterboard 1 (Pb1-Pb2) crosses 9000C.
20) The formation of the second plateau in the time-temperature profile of the
plasterboards need not be assumed in the thermal modeling, as it is not evident in the
temperature profiles of the plasterboards.
21) In the case of 2x2 non-load bearing wall specimens using Glass Fibre or
Cellulose Fibre as cavity insulation, the insulation can be removed from the model
when the temperature of the Pb2-Ins interface crosses 7000C.
22) In the case of 2x2 cavity insulated non-load bearing wall specimens the specimen
can be considered to have failed structurally when the temperature of the Pb2-Ins
interface reaches 7000C. At this temperature the hot flange becomes sufficiently soft
to initiate a reversal in lateral deformation of the studs. Also the screws holding the
plasterboards to the steel frame begin to rotate downwards under the plasterboards
weight as the soft hot flange is unable to offer any degree of fixity. This leads to the
collapse of the plasterboard exposing the entire frame to direct fire.
Thus when the temperature of the interface Pb2-Ins is in the range of 7000C to
7500C, four things happen very quickly;
a) Cavity insulation (Glass Fibre or Cellulose Fibre) starts to disintegrate rapidly.
b) Studs undergo reversal in lateral deformation.
c) The screws connecting the Plasterboards to the steel frame start rotating
downwards as they lose their fixity.
d) Plasterboard 2 falls off as the sudden jerk introduced by the reversal in lateral
deformation of the studs coupled with the rotation of the screws used for fixing
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 374
the plasterboard to the frame are sufficient to induce a partial or complete
collapse in the already weakened exposed plasterboard.
Only Rock Fibre when used as cavity insulation resists burnout well beyond 7000C.
This delays the reversal in lateral deformation of the studs and allows Plasterboard 2
to remain in position till the Pb2-Ins interface temperature reaches 9000C leading to
the collapse of Pb2 and reversal in lateral deformation of the studs resulting in the
failure of the wall.
23) 2x2 non-load bearing wall specimens (both cavity insulated and externally
insulated) fail by stud buckling before insulation failure can occur.
24) Plasterboard 1, Insulation between the exposed plasterboards and Plasterboard 2
from the externally insulated wall models need to be removed after certain time in
the thermal modeling. The critical temperatures identifying the removal for these
elements are identical with the cavity insulated specimens. The insulation if it is
Glass Fibre or Cellulose Fibre can be removed from the model when the Pb1-Ins
interface temperature reaches 7000C and Plasterboard 1 can be removed from the
model regardless of the type of insulation used when the interface temperature
reaches 9000C. When Rock Fibre is used as external insulation it can be assumed to
become ineffective when the interface temperature reaches 9000C.
Similar to the cavity insulated specimens, Plasterboard 2 can be removed from the
model when the temperature of the cavity facing surface of the plasterboard (Pb2-
Cav) is in the range of 7000C – 7500C. The wall can be assumed to have failed by
structural inadequacy on the removal of Plasterboard 2.
25) Vertical plasterboard joints along the stud length affect the thermal performance
of 1x1 LBW and leads to a structural failure.
26) 2x2 load-bearing walls fail by structural inadequacy and not by insulation or
integrity failure.
27) The type of insulation used in the cavity of load-bearing walls has a low influence
on the stud temperatures.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 375
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 376
28) Hot flange temperatures of externally insulated load-bearing walls is influenced
by the type of insulation used.
7.4: Conclusion:
In the comparative study of different wall models, the fire performance of externally
insulated wall specimens both load bearing and non-load bearing was found to be
considerably better than the non-insulated and cavity insulated wall specimens. The
thermal insulation property of the externally insulated wall specimens was seen to be
much superior when compared with the insulation characteristics of cavity insulated
specimens. The ambient side temperatures and the lateral deformations of the
externally insulated specimens were also seen to be consistently lower than what was
observed in the case of cavity insulated specimens.
In the case of externally insulated wall specimens, the quality of insulation used was
observed to directly influence the fire performance of the specimen with the Rock
fibre giving the best results, whereas, in the case of cavity insulated specimens, the
type of insulation used did not much affect the fire performance of the wall models.
The failure of all the wall specimens was noted to occur primarily due to the structural
failure of the studs and never by insulation or integrity. Cavity insulated specimens
were seen to fail earlier than similarly built non-insulated specimens, whereas, the
failure times of the externally insulated Test Specimens were seen to be maximum.
Simple linear equations and graphs based upon experimental work have been
proposed to predict the growth in hot flange temperatures of load bearing non-
insulated, cavity insulated and externally insulated wall specimens. This temperature
growth model is used to develop predictive equations to estimate the failure times of
all the tested wall models. A close agreement has been observed between the
predicted failure times and the actual failure times of load bearing and non-load
bearing wall models with and without insulation.
Chapter 8: Conclusions and Recommendations
This thesis has described a detailed investigation into the structural and thermal
performance of cold-formed LSF stud wall systems lined with gypsum plasterboards
under fire conditions. It included both the conventional steel stud wall systems with
and without cavity insulation as well as a new steel stud wall system based on a
composite panel in which a layer of external insulation was used between the two
plasterboards. Both non-load bearing walls and load bearing walls were tested in a
detailed experimental study. This research has thus developed comprehensive
experimental thermal and structural performance data for both the conventional and
the new non-load bearing and load bearing cold-formed steel stud wall systems under
fire conditions including simple and accurate methods to predict their fire resistance
rating. It has improved the knowledge and understanding of the fire performance of
cold-formed LSF stud wall systems under fire conditions, and has led to the
development of safer design methods for fire conditions and new LSF stud wall
systems with increased fire rating.
A detailed literature review of the current body of knowledge in this field was
undertaken first (Chapter 2). High grade cold-formed steels are increasingly used not
only in LSF stud wall systems, but also in other building systems. The lack of reliable
mechanical property data of these high grade cold-formed steels at elevated
temperatures was addressed in Chapter 3, which describes the steady state tensile
coupon tests undertaken in this research at ambient and elevated temperatures. The
use of non-contact Laser Speckle Extensometer (LSE) was found to be highly
successful in providing the required strain measurements at elevated temperatures
instead of the resistance type strain gauges. This experimental study led to the
development of predictive equations for the determination of yield strength and elastic
modulus of high strength steels at elevated temperatures. The developed equations
compared well with the test results, and are considered to eliminate the conservative
predictions given by the current Australian and European standards.
In Chapter 4 the thermal performance of the gypsum plasterboards was investigated
using 15 small scale fire tests of Type X gypsum plasterboards supplied by Boral
Plasterboards under the product name FireSTOP. Thermal performance of single,
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 377
double and triple layers of plasterboards was investigated in detail. It was found that
the discontinuity or interface between the layers of plasterboard increased the fire
performance of the wall system. Thermal performance of composite panels using
different types of insulating material of varying thickness and density was also
investigated, which allowed the assessment of the fire performance of insulations such
as glass fibre, rock fibre and cellulose fibre and also the determination of the
temperature at which the fall off of external plasterboards occurred.
Chapter 5 presents the details of nine small scale wall models built and fire tested to
investigate the thermal performance of conventional steel stud wall systems with and
without the use of cavity insulation and the innovative steel stud wall systems using
composite panels. The composite panels were seen to offer greater thermal protection
to the studs as compared to the conventionally built non-load bearing wall models.
The use of cavity insulation regardless of the type and density of insulation has been
shown to lower the fire rating of the walls. Rock fibre was identified to have the
maximum detrimental effect on the fire performance of non-load bearing walls when
used as cavity insulation. This chapter identifies and discusses the deficiencies in the
conventional stud wall systems. Time-temperature measurements from the tests
clearly demonstrated the superior thermal and fire performance achieved by the use of
composite panels. The benefits of adopting this new system over the conventional
stud wall systems are discussed in this chapter.
Chapter 6 presents the details of nine full scale load bearing wall models built and fire
tested to study the thermal and structural performance of the load bearing wall
assemblies lined with single or dual layers of plasterboards with and without cavity
insulation and compares the results with the thermal and structural performance of
load bearing wall models built using composite panels on either side of the steel
frame. Details of the results, including the temperature and deflection profiles
measured during the tests are presented along with the stud failure modes. The
analysis showed that the proposed cold-formed steel stud wall systems with external
insulation provided considerably increased fire resistance rating with smaller lateral
deformations than the conventional cavity insulated stud wall systems.
Chapter 7 presents the outcomes of the tests performed on non-load bearing and load
bearing conventional steel stud wall systems with and without cavity insulation and
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 378
the innovative steel stud wall systems using composite panels. Idealized hot flange
temperature profiles of non-insulated, cavity insulated and externally insulated load
bearing wall models using composite panels have been developed and presented in
this chapter along with suitable equations for predicting their failure times (fire
resistance rating). The chapter also presents the development of a simple graphical
method to predict the failure times of non-load bearing and load bearing wall models
under different load ratios.
8.1. Main Research Outcomes
The most valuable outcomes from this research are as follows:
Significantly improved the knowledge and understanding of the structural and
thermal performance of cold-formed LSF stud wall systems under fire
conditions. Both non-load bearing and load bearing walls with varying
arrangements of plasterboard and insulation were included.
Developed an innovative cold-formed LSF wall system with increased fire
resistance rating through the use of a composite panel system in which a layer
of insulation is placed externally between the two plasterboards. This thesis is
the first one to propose and investigate the use of such an innovative
composite panel system, and demonstrate its superiority and benefits over
conventional panels.
Developed comprehensive structural and thermal performance data for both
the conventional and the new LSF stud wall systems, which can be used for
accurate numerical modelling and design of LSF stud walls by fire researchers
and designers.
Developed simple predictive models for the mechanical properties of high
grade steels at elevated temperatures for use by researchers and engineers as
the values given in the current steel design standards are either too
conservative or unsafe.
Developed idealised time-temperature profiles of studs in LSF walls under fire
conditions and their fire resistance rating as a function of varying
arrangements of plasterboards and insulation, and load ratios. Engineers,
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 379
designers and researchers in this field can use them without the need for
further expensive and time consuming full scale fire tests.
This research has produced many useful outcomes for the designers of LSF
stud wall systems; for example, this research has shown that the common
industry belief that the use of cavity insulation improves the fire rating of
walls is not true. The use of cavity insulation was found to reduce the fire
rating of load bearing walls regardless of the type and density of insulation.
This research has paved the way for Australian building industries to develop
new LSF stud wall systems based on the new composite panels proposed in
this research with increased fire rating for commercial applications worldwide.
Developed an excellent fire testing facility at the Queensland University of
Technology that has the capacity to simulate both standard and real fire curves
on LSF stud walls and to obtain high quality temperature and deformation data
using the latest technologies. This is currently being used by other researchers.
8.2 Recommendations to the Construction Industry
Based on the research reported in this thesis, the following recommendations are
made to the building construction industry.
Implement the use of the new composite panel proposed in this research in the
standard wall panel systems. For this purpose develop improved cost-efficient
methods of building and installing composite panels.
Develop thinner sheets of insulation with higher fire performance and thermal
insulation characteristics to facilitate easy construction of composite panels.
Develop appropriately sized composite panel units with interlocking joints to
promote speedy construction of the new wall systems with minimum labour.
Develop improved methods for fixing multiple plasterboards and composite
panels to the steel studs of the walls in order to improve the fire performance
of wall systems with reduced thermal bridging problems.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 380
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 381
The technique of spraying cellulose fibre insulation onto the plasterboard
needs to be standardized to achieve a uniform distribution and density of the
sprayed insulation layer. This will avoid the formation of weak areas within
the composite panel that leads to a weaker fire performance of the wall
systems.
8.3 Future Research
This research has addressed the structural and thermal performance of cold-formed
LSF stud wall systems under fire conditions using an extensive experimental study
program. However, further research is needed in some areas as shown next.
Effect of varying stud sizes on the fire performance of load bearing wall
systems built using composite panels.
Extension of external insulation concept to ceiling elements in order to
improve their fire performance.
Numerical modeling of both conventional and proposed LSF stud wall
systems to simulate both their thermal and structural behaviour. Such
numerical thermal and structural models can be validated using the vast
amount of experimental results presented in this thesis.
Additional experimental investigations on the fire performance of both
conventional and externally insulated load bearing wall specimens at different
load ratios to verify the predictive equations presented in the thesis.
Experimental investigations on the fire performance of both conventional and
externally insulated load bearing wall specimens using different types of
popularly used plasterboards used in the construction industry.
Effect of screw length and screw spacing on the fire performance of stud walls
using multiple plasterboards. This is because the screws used in the test
specimens with multiple plasterboards on the fire side or composite panels
were found to be severely bent at elevated temperatures, promoting the
collapse of the outermost external plasterboards
References
Ala-Outinen and Myllymaki, J. (1995) The Local Buckling of RHS Members at Elevated
Temperatures, Technical Research Centre of Finland, Espoo.
Alfawakhiri, F. (2001), Behaviour of Cold-formed-Steel-framed Walls and Floors in
Standard Fire Resistance Tests, PhD, Thesis, Department of Civil and Environmental
Engineering, Carleton University, Ottawa, Ontario, Canada.
American Society for Testing and Materials (ASTM E119, 1995), Standard Test Methods
for Fire Tests of Building Construction and Materials, West Conshohocken, PA, USA.
Andersson, L. and Jansson, B. (1987), Analytical Fire Design With Gypsum – A
Theoretical and Experimental Study, Institute of Fire Safety Design, Lund, Sweden.
Baux, C. Melinge, Y. Lanos, C. And Jauberthie, R. (2008) Enhanced Gypsum Panels for
Fire Protection, Journal of Materials in Civil Engineering, Vol. 20, pp. 71-77.
Benichou, N. and Sultan, M.A. (2005), Thermal Properties of Light Weight Framed
Construction Components at Elevated Temperatures, Fire and Materials, Vol. 29, pp.
165-179.
British Standards Institute-(BSI) (1990), British Standard, BS 5950, Structural Use of
Steelwork in Building, Part 8, Code of Practice for Fire Resistance Design, London, UK.
Buchanan, A. H. (2001), Structural Design for Fire Safety, John Wiley and Sons, HK.
The Steel Construction Institute (1993), Building Design Using Cold Formed Steel
Sections: Fire Protection, SCI Publication P129, Berkshire, UK.
CAN/ULC-S101-M89, Standard Methods of Fire Endurance Tests of Building
Construction and Materials, Underwriters Laboratories of Canada, Scarborough, Ontario,
1989.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 382
Chen, J. and Young, B. (2004), Mechanical Properties of Cold-formed Steel at Elevated
Temperatures, Proceedings of 17th International Speciality Conference on Cold-formed
Steel Structures, University of Missouri-Rolla, USA, pp. 437-466.
Chen, J. and Young, B. (2007), Experimental Investigation of Cold-formed Steel Material
at Elevated Temperatures, Thin-walled Structures, Vol. 45, pp. 96-110.
Cooke, G. M. E. (1987),Thermal Bowing and How it Affects the Design of Fire
Separating Construction, Fire Research Station, Building Research Establishment, Herts,
UK.
Davies, J. M. (2000), Recent Research Advances in Cold-formed Steel Structures,
Journal of Constructional Steel Research, Vol. 55, No. 1-3, pp. 267-288.
EN 520 (2004) Gypsum Plasterboards, Definitions, Requirements and Test Methods,
CEN, Brussels.
EN 15283-2 (2008) Gypsum Boards with Fibrous Reinforcement – Definations,
Requirements and Test Methods - part 2: Gypsum fibre boards, CEN, Brussels.
European Prestandard ENV 1993-1-2, Eurocode 3 – Design of Steel Structures – Part 1 –
2: General Rules – Structural Fire Design, European Committee for Standardization (CE
N), Brussels, Belgium, September 1995.
Feng, M. Wang, Y. C. and Davies, J. M., (2003), Thermal Performance of Cold-formed
Thin-walled Steel Panel Systems in Fire, Fire Safety Journal, Vol. 38, pp. 365-394.
Feng, M. Wang, Y. C. (2005), An Experimental Study of Loaded Full-Scale Cold-
Formed Thin-Walled Steel Structural Panels Under Fire Conditions, Fire Safety Journal,
Vol. 40, pp. 43-63.
Franssen, J. M. (1999), Thermal Properties of Gypsum Board Walls Submitted to the
Fire, Literature Survey, UNIVERSITE DE LIEGE, Belgium.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 383
Frangi, A. Schleifer, V. Fontana, M. and Hugi, E. (2010), Experimental and Numerical
Analysis of Gypsum Plasterboards in Fire, Fire Technology, Vol. 46, pp. 149-167.
Gerlich, J.T. (1995), Design of Load-bearing Light Steel Frame Walls for Fire
Resistance, Fire Engineering Research Report 95/3. School of Engineering, University of
Canterbury, New Zealand.
Gerlich, J.T. Collier, P. C. R. and Buchanan, A. H. (1996), Design of Steel-framed Walls
for Fire Resistance, Fire and Materials, Vol. 20, No. 2, pp. 79-96.
Ghazi, W. and Hugi, E. (2009), Four Types of Gypsum Plaster Boards and their
Thermophysical Properties under Fire Condition, Journal of Fire Sciences, Vol. 27, No.1,
pp. 27-43.
Gonclaves, T. Clancy, P. and Poynter, W. (1996), Mechanical Properties of Fire Rated
Gypsum Board, Victoria University of Technology, Australia.
Harmathy, T. Z. (1998), Properties of Building Material, The SFPE Handbook of Fire
Protection Engineering, Society of Fire Protection Engineering/National Fire Protection
Association, Bostan, United States.
ISO 834 -1, (1999), Fire Resistance Tests – Elements of Building Construction – Part 1:
General Requirements, International Organization for Standardization, Geneve,
Switzerland.
Klippstein, K.H. (1978), Strength of Cold-formed Steel Studs Exposed to Fire,
Proceedings of the 4th International Speciality Conference, Missouri-Rolla, St.Louis,
Rolla, Mo, USA, pp. 513-555.
Kodur, V. K. R. Sultan, M. A. Latour, J. C. Leroux, P. and Monette, R. C. (1999),
Experimental Studies on the Fire Resistance of Load-Bearing Steel Stud Walls, Internal
Report, IRC, NRC, Ottawa, Canada.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 384
Kwon, I. K. Choi, K. and Jee, N. Y. (1998), Fire Resistance of Bearing Wall Using Steel
and Gypsum, Proceedings of the Fourteenth International Speciality Conference on Cold
Formed Steel Structures, St. Louis, MO, pp. 379-391.
Lee, J. Mahendran, M. and Mäkeläinen, P. (2003), Prediction of Mechanical Properties of
Light Gauge Steels at Elevated Temperatures, Journal of Construction Steel Research, pp.
1517-1532.
Lie, T.T. (Ed.), (1992), Structural Fire Protection, American Society of Civil Engineers,
New York, NY.
Mäkeläinen, P. and Miller, K. (1983), Mechanical Properties of Cold-formed Galvanized
Sheet Steel Z32 at Elevated Temperatures, Helsinki University of Technology, Finland.
Manzello, S.L. Gann, R.G. Kukuch, S.R. Prasad, K. And Jones, W.W. (2005), Real Fire
Performance of Partition Assemblies, National Institute of Standards and Technology
(NIST), USA.
Manzello, S.L. Gann, R.G. Kukuch, S.R. Lenhert, D.B. (2006), Influence of Gypsum
Board Type (X or C) on Real Fire Performance of Partition Assemblies, National
Institute of Standards and Technology (NIST), USA.
Mehaffey, J.R. Cuerrier, P. and Carisse, G. (1994), A Model for Predicting Heat Transfer
through Gypsum-Board/Wood-Stud Walls Exposed to Fire, Fire and Materials, Vol. 18,
No. 5, pp 297-305.
Milke, J. A. (1999), Analytical Methods to Evaluate Fire Resistance of Structural
Members, ASCE Journal of Structural Engineering, Vol. 125, No. 10, pp. 1179-1187.
Noureddine, B. and Sultan, M. A. (2005), Thermal Properties of Lightweight-framed
Construction Components at Elevated Temperatures, Fire and Materials, Vol. 29, pp.
165-179.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 385
Outinen, J. Kaitila, O. Makelainen, P. (2000), A Study for the Development of the
Design of Steel Structures in Fire Conditions, Proceedings of 1st International Workshop
of Structures in Fire, Copenhagen, Denmark.
Purkis, J. A. (1996), Fire Safety Engineering Design of Structures, Reed Educational and
Professional Publishing Ltd. Oxford, UK.
Ranby, A. (1999), Structural Fire Design of Thin Walled Steel Sections, Licentiate
Thesis, Department of Civil and Mining Engineering, Lulea University of Technology,
Stockholm, Sweden.
Ranawaka, T. and Mahendran, M. (2009), Experimental Study of the Mechanical
Properties of Light Guage Cold-formed Steels at Elevated Temperatures, Fire Safety
Journal 44, 219-229.
SCI Publication P 129 (1993), Building Design using Cold-formed Steel Sections: Fire
Protection.
Sidey, M. P. and Teague, D. B. (1988), Elevated Temperature Data for Structural Grades
of Galvanized Steels, Technical Note No. WL/PM/TN/251/88/D, Welsh Laboratory,
British Steel Corporation, Port Talbot, UK.
Son, B. C. and Shoub, H. (1973), Fire Endurance Tests of Double Module Walls of
Gypsum Board and Steel Studs, NBSIR 73-173, Centre for Building Technology,
National Bureau of Standards, Washington, DC.
Standards Australia (SA) (1998), AS 4100, Steel structures, Sydney, Australia.
Standards Australia (SA) (2005), AS 4600, Cold-formed steel structures, Sydney,
Australia.
Standards Australia (SA) (2005), AS 1530.4, Methods for fire tests on building materials,
components and structures, Part 4: Fire-resistance tests of elements of building
construction, Sydney, Australia.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 386
Standards Australia (SA) (2005), AS 1391, Metallic materials-Tensile testing at ambient
temperature, Sydney, Australia.
Standards Australia (SA) (1979), AS 2291, Methods For the Tensile Testing of Metals at
Elevated Temperatures, Sydney, Australia.
Sterner, E. and Wickstrom, U. (1990), TASEF- Temperature Analysis of Structures
Exposed to Fire, Fire Technology SP Report, Swedish National Testing Institute, Boras,
Sweden.
Sultan, M. A. (1995), Effect of Insulation in the Wall Cavity on the Fire Resistance
Rating of Full-Scale Asymmetrical (1 x 2) Gypsum Board Protected Wall Assemblies,
Proceedings of the International Conference on Fire Research and Engineering, Orlando,
FL, Lund D. P. (Ed.), Society of Fire Protection Engineers, Bostan, MA, pp. 545-550.
Sultan, M.A. (1996), A Model for Predicting Heat through Non-insulated Unloaded
Steel-Stud Gypsum Board Wall Assemblies Exposed to Fire, Fire Technology, Vol. 32,
No. 3, pp. 239-259.
Sultan, M.A. (2010), Comparison of Gypsum Board Fall-off in Wall and Floor
Assemblies, 12th International Conference on Fire Science and Engineering Conference,
Nottingham, UK, pp. 1-6.
Sultan, M.A. Alfawakhiri, F. Benichou, N. (2001), A Model for Predicting Heat Transfer
through Insulated Steel-Stud Wall Assemblies Exposed to Fire, Fire and Materials,
International Conference, San Francisco, pp. 495-506.
Sultan, M.A. and Kodur, V. R. (2000), Light-Weight Frame Wall Assemblies: Parameters
for Consideration in Fire Resistance Performance-Based Design, Fire Technology,
Vol.36, No. 2, pp. 75-88.
Sultan, M.A. and Lougheed, G. D. (1994), The Effect of Insulation on the Fire Resistance
of Small-Scale Gypsum Board Wall Assemblies, Proceedings of the Fire and Materials
Third International Conference and Exhibition, Washington, DC, Interscience
Communications Limited, London, UK, pp. 11-20.
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 387
P.N.Kolarkar: Structural and Thermal Performance of Cold-formed Steel Stud Wall Systems under Fire Conditions 388
Thomas, G. (2002), Thermal Properties of Gypsum Plasterboard at High Temperatures,
Fire and Materials, Vol. 26, pp. 37-45.