1-s2.0-s0143974x06001465-main
TRANSCRIPT
-
7/29/2019 1-s2.0-S0143974X06001465-main
1/15
Journal of Constructional Steel Research 63 (2007) 460474www.elsevier.com/locate/jcsr
Inelastic performance of cold-formed steel strap braced walls
M. Al-Kharat, C.A. Rogers
Department of Civil Engineering and Applied Mechanics, McGill University, Montreal QC H3A 2K6, Can ada
Received 8 May 2006; accepted 27 June 2006
Abstract
The inelastic performance of sixteen 2.44 m 2.44 m cold-formed steel strap braced walls was evaluated experimentally. The performance
was affected by the holddown detail, which in many cases did not allow the test specimens to reach or maintain a yield capacity and severelydiminished the overall system ductility. Test-based Rd Ro values of 3.65, 2.11 and 1.72 indicate the low ductility levels, which were not
adequate to warrant the use of a seismic response modification coefficient of R = 4.0 in design. Capacity design of the SFRS elements must
account for the overstrength of the strap material.c 2006 Elsevier Ltd. All rights reserved.
Keywords: Cold-formed steel; Strap brace; Ductility; Inelastic; Seismic; Performance
1. Introduction
The use of cold-formed steel as the main framing element
in a structure is becoming more popular for the construction
of low- to mid-rise buildings across Canada, including areas
with a high seismic hazard. In order to maintain the integrity
of these structures when subjected to horizontal forces due
to an earthquake the use of diagonal flat steel strap cross
bracing may be a practical solution (Fig. 1). The straps act as a
vertical concentric bracing system, which transfers the lateral
forces from the roof and floor levels to the foundation. The
overall lateral strength, ductility and stiffness of this bracing
system may not be related solely to the steel straps; many
other elements in the lateral load carrying path can play a role,
such as the strap connections, the gusset plates (if needed), the
anchorage including holddown and anchor rod, etc.In Canada earthquake loading may often dictate the design
of the lateral force resisting system in a building in areasof high seismic hazard, such as found along the west coast
of the country as well as in the Saint Lawrence and Ottawa
River valleys. The 2005 National Building Code of Canada
(NBCC) [1] requires that seismic loading also be considered
in other areas of the country, where in the past it has not been
Corresponding address: Department of Civil Engineering and AppliedMechanics, McGill University, 817 Sherbrooke St. W., Montreal, QC H3A 2K6,Canada. Tel.: +1 514 398 6449; fax: +1 514 398 7361.
E-mail address: [email protected] (C.A. Rogers).
Fig. 1. Cold-formed steel strap braced walls under construction.
of significant concern for design engineers. This is due, in
part, to a change in the seismic hazard information used for
design. Seismic forces, which were previously based on a 10%
in 50 year probability of exceedance, i.e., corresponding to a
return period of 475 years, are now based on a uniform hazardspectrum having a 2% in 50 year probability of exceedance,
i.e., approximately a return period of 2500 years [2]. The 2005
NBCC also comprises a capacity based philosophy for seismic
design, where a fuse element in the seismic force resisting
system (SFRS) is selected to dissipate earthquake derived
energy. This energy dissipating element is expected to enter
into the inelastic range of behaviour, whereas the remaining
components of the SFRS are designed to carry the forces
associated with the probable capacity of the fuse element,
i.e., they should remain essentially elastic or experience only
0143-974X/$ - see front matter c 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jcsr.2006.06.040
http://www.elsevier.com/locate/jcsrmailto:[email protected]://dx.doi.org/10.1016/j.jcsr.2006.06.040http://dx.doi.org/10.1016/j.jcsr.2006.06.040mailto:[email protected]://www.elsevier.com/locate/jcsr -
7/29/2019 1-s2.0-S0143974X06001465-main
2/15
M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 46 0474 461
minor plastic damage. It is generally assumed that the straps
act as the fuse element in the SFRS of braced cold-formed steel
structures.
Guidelines that address the seismic design/inelastic perfor-
mance of cold-formed steel structures are not provided in the
2005 NBCC or in the Canadian Standards Association (CSA)
S136 Standard for the Design of Cold-Formed Steel StructuralMembers [3]. In contrast, seismic design information for cold-
formed steel structures is available in the US. ASCE 7-05 [4]
allows for the use of a seismic response modification coeffi-
cient of R = 4.0 for strap braced bearing wall systems, which
indicates a reliance on a moderate level of ductile/inelastic per-
formance of the SFRS as well as some overstrength. Use of
this R value necessitates that the material specific seismic de-
sign and detailing requirements of the American Iron and Steel
Institute (AISI) Lateral Design Standard [5] and the AISI Spec-
ification [6] be met. Even if not detailed for seismic resistance
ASCE 7-05 allows for an R of 3.0 to be used for the design
of strap braced walls. The AISI Lateral Design Standard states
that boundary members, chords, collectors and connections of abraced wall must be proportioned to transmit the induced forces
and the amplified seismic loads. Vertical chord members are
required to have the nominal strength to resist amplified seis-
mic loads, but not loads greater than what the system can de-
liver. The strength of brace connections need be the lesser of
the nominal tensile strength of the brace or the amplified seis-
mic load. Furthermore, strap bracing is to be designed in ac-
cordance with the AISI Specification or the AISI Standard on
General Provisions [7], which for the most part do not contain
any relevant seismic detailing information. Typically, the AISI
Standards and Specification are written in terms of strength
requirements for seismic design; however, no mention of ex-pected ductility requirements or recommended ductile connec-
tion/anchorage details is made. The US Army Corps of Engi-
neers has also published a document that addresses the seis-
mic design of cold-formed steel structures, TI 809-07 [8]. The
intent of this document is to ensure that ductile building sys-
tem performance is attained during large seismic events. Duc-
tile performance requires that the strap members of a braced
wall are first able to yield and then maintain this level of load
carrying capacity while being subjected to significant plastic
deformations. The failure of columns and connections must not
occur. The TI 809-07 provisions for seismic design are similar
to what is found in ASCE 7-05 and the AISI Lateral Design
Standard, except that additional prescriptive requirements for
material properties of the braces, as an example, exist.
2. Objectives and scope of research
Due to the lack of codified seismic design guidance in
Canada for cold-formed steel structures a research project
was undertaken to evaluate the inelastic performance of steel
framestrap braced walls that are not designed following a strict
capacity based design philosophy. The main objectives were to
determine the ductility of common strap braced walls by means
of physical testing and to assess the inelastic performance
with respect to the ASCE7-05 R-value of 4.0; that is the
ability of the flat straps to yield over extended displacements
without extensive damage to the other components in the
SFRS. Three typical wall configurations were tested; light,
medium and heavy in the context of cold-formed steel. Due
to this research being in its initial stages the investigation
involved only the assembly testing of representative strap
braced walls under lateral in-plane loading. A total of sixteen2.44 m 2.44 m walls with standard non-seismic details
were tested using monotonic and reversed cyclic loading
protocols. The performance of the walls was expected to
match that of a SFRS for which an appropriate capacity based
design approach had been implemented. That is, gross cross-
section yielding of the tension braces was the anticipated
failure mode, while the remaining elements in the SFRS were
expected to carry the brace force with no or only minor plastic
deformation. A comparison of the failure mode, ductility, shear
strength and shear stiffness characteristics of the strap walls is
presented.
3. Literature review of previous research on strap bracedwalls
Previous experimental and analytical research on the
performance of cold-formed steel strap braced walls was
reviewed to establish in-part the scope and methods of study for
the investigation described in this paper. Information from this
past research, summarized below, was used to select the wall
configurations and the test methods. In addition, the findings of
these studies were used to define the best case scenario of wall
performance in which inelastic deformations are limited to the
strap braces.
Adham et al. [9] evaluated the lateral load versus deflectionbehaviour of six 2.44 m 2.44 m cold-formed steel planar
frames sheathed with steel straps and gypsum. Straps, 50.8 mm
and 76.2 mm in width with three different thicknesses (0.84,
1.09 and 1.37 mm) were screw connected to the framing
elements. Most walls were constructed with X straps as well as
gypsum panels on both sides. Holddowns were bolted to each
test specimen at the base to limit uplift of the cold-formed steel
frame. Adham et al. showed that stud buckling will lead to a
severe degradation in the shear load that can be applied to the
wall; however when this mode is properly addressed in design
strap braced systems are effective in dissipating energy under
reversed cyclic loading.
Serrette and Ogunfunmi [10] also investigated the perfor-mance of 2.44 m 2.44 m strap braced frames through experi-
ments of walls under lateral in-plane loading. Screw connected
walls constructed with 50.8 mm 0.88 mm straps on one face
were tested (3 specimens), in addition to walls with strap braces
on one face and gypsum sheathing board on the other (4 speci-
mens). A single test specimen with braces on both sides of the
wall was also included in the study. In all cases, it was neces-
sary to bolt an 11 mm thick steel clip angle to the chord studs
to act as a holddown device. Cold-formed steel gusset plates
were used to connect the strap braces to the studtrack cor-
ner locations. It was shown that walls with bracing on one side
alone failed by excessive out-of-plane deformation, which is
-
7/29/2019 1-s2.0-S0143974X06001465-main
3/15
462 M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 460474
not a favourable scenario in terms of maintaining lateral stabil-
ity of the braced frame, nor ductile performance under inelas-
tic shear deformations. Serrette and Ogunfunmi reported that
gypsum panels provide a substantial increase in shear capac-
ity compared with the 50.8 mm wide straps; however the use
of gypsum panels and strap braces together is not practical. It
was also noted that in the design of X-braced walls the engi-neer must be concerned with strap yield strengths in excess of
the minimum specified value, which may result in connection
or chord stud failure.
Barton [11] and Gad et al. [1214] investigated the
earthquake performance of strap braced cold-formed steel wall
structures as used in the Australian residential construction
industry. The impact of steel strap braces, as well as non-
structural components such as plasterboard and brick veneer, on
wall performance were evaluated through experimentation and
analyses. The research involved racking and dynamic (shake
table) testing of planar wall and 3D one room house specimens.
Relatively small strap braces were installed, 25 mm wide
1 mm thick, compared with previous studies by Adhamet al. [9] and Serrette and Ogunfunmi [10]. Ductility and
overstrength concerns were investigated given the possible
impact of non-structural components. Hysteretic load versus
deflection behaviour of the braced steel frame alone was
first obtained, followed by tests of sheathed walls and the
3D single-storey structures. In general, the steel frames were
able to perform well under seismic loading and the non-
structural components made a significant contribution to the
lateral bracing of the frames. It was reported that the pinched
force versus deformation behaviour was caused by elongation
of the straps, deformation of the connections, as well as initial
slack in the system. Screw failure was typically observed inracking tests. Stiffness of the bare steel frame was mainly
due to the strap braces and not the stud to track connection
detail even if welded. Dynamic shake table tests showed that
yielding of the braces could take place, in addition to slip
and in most cases failure of the brace connections. A 3D
finite element study was also completed, which included the
bare steel frame accounting for the brace, brace connection
and tensioner unit behaviour, as well as the effect of adding
a plasterboard lining. This allowed for different length walls
and boundary conditions for the non-structural components
to be evaluated. These studies showed that for single family
dwellings the non-structural components significantly increase
the strength and stiffness of strap braced wall systems. It is
possible that the use of relatively small brace members by
Barton and Gad et al. allowed for the plasterboard lining to
become dominant with respect to resisting lateral in-plane loads
and providing shear stiffness to the steel framing. A ductility
related response modification coefficient for seismic design
(R = 1.53.5) was recommended based on various yield
displacement models by Park [15] and subsequent nonlinear
time history dynamic analysis. A formulation was put forth to
predict the period of vibration for strap braced structures. An
evaluation of overstrength, which is highly dependent on the
non-structural components and their boundary conditions, was
also provided.
Fulop and Dubina [16] tested three X bracedscrew
connected wall specimens (3.6 m long 2.44 m high)
under in-plane lateral loading. Of the three wall specimens
one was tested monotonically and two cyclically. The walls
were constructed of a cold-formed steel frame connected to
110 mm wide 1.5 mm thick straps located on each side.
The screw connection configuration was selected to facilitateyielding along the length of the brace, i.e., to avoid net section
fracture of the strap through the screw holes. Chord members
were constructed of double stud members such that inelastic
deformations and ultimate failure of the walls would be limited
to the braces. U profiles were placed in the tracks at corner
locations to increase the holddown capacity and rigidity. Local
buckling of the lower track was observed during loading with
damage being concentrated in corner areas. Plastic elongation
of the strap did take place; however because of the unexpected
failure of the corners the results of the experiments may not
necessarily reflect the true ductility of a braced wall if yielding
(and failure) had been limited to the straps. Fulop and Dubina
suggested that the ideal configuration of the corners would besuch that the uplift force is directly transmitted from the brace
or corner stud to the anchoring bolt, without inducing bending
in the bottom track. Failure to strengthen the corners can have a
significant effect on the initial rigidity of the system and can be
the cause of larger than expected in-plane shear deformations
of the wall and premature failure of the braced frame.
Tian et al. [17] completed an experimental and theoretical
study on the racking strength and stiffness of cold-formed steel
walls, including frames with single and double X straps. A
total of five planar frames, 2.45 m in height 1.25 m in
length, composed of strap braces riveted to the steel framing
were tested. Brace size was either 60 mm 1.0 mm or60 mm 1.2 mm, and for all but one of the specimens
braces were installed on both sides of the wall. Monotonic
loading of all tests was carried out, which included single step
and three step protocols. Deformation behaviour and failure
modes were observed, and shear strength and stiffness of
the frames were measured. Tian et al. reported that frames
with straps on both sides have the best racking performance.
Compression failure of the chord stud members was observed
in the double sided specimens. Rivet failure at the brace to
frame connection was also observed. It appears that the walls
were not designed such that inelastic behaviour was limited
to the braces given the connection and chord stud failures that
were reported. Subsequent analyses of the test frames using an
elastic slope deflection method was completed to predict the
failure loads and initial shear stiffness. Tian et al. concluded
that it was possible to accurately predict the shear loads that
were measured during testing; however the in-plane shear
deformations of the walls could not be precisely determined
with their calculation method.
Pastor and Rodrguez-Ferran [18] presented the develop-
ment of an hysteretic model that can be used for the nonlin-
ear inelastic dynamic analysis of X-braced cold-formed steel
frames. The model captures the behavioural characteristics
of this framing type that have been observed during experi-
ments, including pinching and stiffness degradation of the force
-
7/29/2019 1-s2.0-S0143974X06001465-main
4/15
M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 46 0474 463
Fig. 2. Schematic drawing of displaced strap braced wall specimen in test frame.
Fig. 3. Schematic drawing of light strap braced test wall with corner detail.
versus deformation hysteresis as well as slack of the braces. The
use of this model assumes that the strap braced wall is able to
maintain its load carrying capacity over extensive and repeated
in-plane inelastic displacements. For the model to be valid the
walls also must be designed such that the strap enters into and
remains in the plastic range prior to buckling of the chord studs.
Further to this research, Casafont et al. [19] evaluated the seis-
mic performance of the screwed connections that are commonly
used for the straps of such braced walls. It was shown that the
straps were able to maintain their yield capacity over extended
inelastic displacements prior to failure of the brace by net sec-
tion fracture at the first line of screws and by tilting of the screw
fasteners. However, in cases where tilting, bearing and pull out
of the screws was observed then ductile yielding of the braces
was not obtained. A design criterion to induce a tiltingnet sec-
tion fracture failure mode, and hence ductile behaviour in the
strap braces, is provided by Casafont et al.
4. Test program
Assembly tests of sixteen strap braced stud wall specimens
(2.44 m 2.44 m) were carried out using a test frame
designed specifically for in-plane shear loading (Fig. 2). These
walls were not designed following a capacity based seismicdesign approach; rather the elements were selected given
typical wind loading levels where all of the components in the
lateral load carrying path were expected to remain elastic. The
predicted factored lateral in-plane resistance of the three wall
configurations in a wind loading situation was approximately
20 kN (light), 40 kN (medium) and 75 kN (heavy), respectively.
Schematic drawings of the three test wall configurations,
including an exterior and interior view of the corner connection
details and holddowns, are provided in Figs. 3 (light), 4
(medium) and 5 (heavy). A listing of the nominal design
(minimum specified) dimensions and material properties of
the test specimens with details of member components is
-
7/29/2019 1-s2.0-S0143974X06001465-main
5/15
464 M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 460474
Table 1
Matrix of strap braced wall tests (nominal design dimensions and material properties)
Specimen properties Test specimens
Light Medium Heavy
1A-M, 1B-M, 2A-C, 2B-C, 3A-M, 3B-M, 4A-C, 4B-C, 5A-M, 5B-M, 6B-C
1C-Ma 2C-Cb 3C-M 4C-C 5C-M
Strap bracing
Thickness (mm) 1.22 1.52 1.91
Dimensions (mm) 58.4 101 152
Grade - Fy (MPa) 230 230 230
Chord studs
Thickness (mm) 1.22 1.52 1.91
Dimensions (mm) 92 41 12.7 152 41 12.7 152 41 12.7
Grade - Fy (MPa) 230 345 345
Interior studs
Thickness (mm) 1.22 1.22 1.22
Dimensions (mm) 92 41 12.7 152 41 12.7 152 41 12.7
Grade - Fy
(MPa) 230 230 230
Tracks
Thickness (mm) 1.22 1.52 1.91
Dimensions (mm) 92 31.8 152 31.8 152 31.8
Grade - Fy (MPa) 230 345 345
Gusset plates
Thickness (mm) NA 1.52 1.91
Dimensions (mm) NA 250 250 300 300
Grade - Fy (MPa) NA 230 230
a Monotonic protocol.b CUREE reversed cyclic protocol.
Fig. 4. Schematic drawing of medium strap braced test wall with corner detail.
provided in Table 1. Note: the measured member dimensions
and material properties provided in Al-Kharat and Rogers [20]
may vary from these nominal values. The walls were braced
with diagonal flat straps installed in an X configuration on
both sides; a configuration that has been shown to have better
performance characteristics than single sided braced walls [10].
The braces specified for the medium walls were similar to those
used by Fulop and Dubina [16], whereas the heavy walls had a
gross cross-sectional area approximately 26 times that of the
straps found in any of the previous studies [914,1619]. Chord
stud members were composed of double C-section shapes stitch
welded front-to-front, while the remainder of the single interior
C-section studs were placed at a nominal spacing of 406 mm.
One row of 1.22 38 12.7 mm continuous bridging was
-
7/29/2019 1-s2.0-S0143974X06001465-main
6/15
M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 46 0474 465
Fig. 5. Schematic drawing of heavy strap braced test wall with corner detail.
welded in place through the web knockouts at the mid-height
of the walls. All steel framing was ASTM A653 [21] material,
either Grade 230 or 345 MPa (Table 1). Connections between
the studs and tracks were made with No. 1016 wafer-head self
drilling/self tapping screws.
The light walls were constructed of straps connected
directly to the stud framing by No. 1016 wafer-head self
drilling/self tapping screws, whereas the medium and heavy
walls comprised of straps that were fillet welded to the gusset
plates, which were in turn welded to the stud and track members
(Figs. 4 and 5). The welded connection detail had not been
used in any of the previous studies on strap braced walls [914,
1619]. L shaped holddowns with a factored uplift capacity of
35 kN were welded to the interior face of the chord studs of thelight walls and then connected to the test frame with a 15.9 mm
diameter ASTM A307 [22] equivalent threaded rod (Fig. 3).
Note that a 12.7 mm diameter threaded rod was instead used for
the first test to be carried out (1B-M). The L shaped holddowns
were fabricated of a vertical steel plate (364260 mm) which
was welded to a horizontal plate (20 64 70 mm). The load
path for the light walls traced from the straps to the chord studs
and then directly to the holddowns.In contrast, flat plate holddowns were placed within the
upper and lower tracks at the four corner locations of the
medium and heavy walls (Figs. 4 and 5). Walls 3A-M, 3B-M
and 5B-M had plates measuring 1990127 mm, whereas for
walls 3C-M, 4A-C, 4B-C, 4C-C, 5A-M and 6B-C plates 19
127 203 mm in size were installed in an attempt to increase
the holddown uplift capacity. Test specimen 5C-M was fitted
with modified holddowns that were fabricated from a C13010
channel section fillet welded to a 19 90 127 mm plate [20].
The holddown plates for the medium and heavy walls were
attached to the loading beam and reaction frame by means of
19 mm diameter ASTM A325 [23] equivalent threaded rods.
No direct connection was made from these holddown plates
to either the braces, gusset plates or the chord studs. Thus the
uplift forces in the medium and heavy walls were transferred
from the braces, through the gusset plates, to the track flanges
and web, and finally to the holddown plate and threaded rod.
Shear anchors (19 mm diameter ASTM A325 bolts [23])
were placed along the top and bottom tracks as indicated in
Figs. 35. All top tracks were drilled to accommodate the ten
shear anchors and two anchor rods, which connected the tracks
through an aluminium spacer to the loading beam. Similarly,
the bottom tracks contained four shear anchors and two anchor
rods, which connected the wall through an aluminium spacer
to the testing frame. The function of the top shear anchors was
to uniformly transfer the load from the loading beam to the top
track, whereas the function of the interior bottom shear anchors
was to connect the wall to the testing frame in a more realistic
fashion.
The testing frame was equipped with a 125 mm stroke
250 kN dynamic actuator. Displacement controlled monotonicand reversed cyclic protocols were used in testing. The testing
frame incorporated external beams to prevent out-of-plane
buckling of the wall specimen, such that only lateral in-
plane displacement would take place, as shown in Fig. 2.
Measurements consisted of strap width (Table 2), in-plane
wall displacements, strains in the steel straps, acceleration of
the loading beam assembly and the shear load at the wall
top. The LVDTs, strain gauges, load cell and accelerometer
were connected to Vishay Model 5100B scanners which were
used to record data using the Vishay System 5000 StrainSmart
software.
The monotonic loading procedure consisted of a steady
rate of displacement (2.5 mm/min) starting from the zero
load position. The CUREE ordinary ground motions reversed
cyclic loading protocol [24,25], run at 0.5 Hz, was chosen
for the testing of the strap braced walls. Previous research
at McGill University on cold-formed steel walls braced with
wood sheathing also incorporated this loading protocol [26].
It was selected because it was anticipated that the dynamic
behaviour of the strap braced walls would resemble in some
ways that of the wood sheathed walls and because a direct
comparison of results would be possible. In a best case
scenario, where the braces are able to maintain their yield
capacity, and given the range of displacement available from the
actuator, no decrease in the wall resistance would be expected.
-
7/29/2019 1-s2.0-S0143974X06001465-main
7/15
466 M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 460474
Table 2
Average measured brace widths
Specimen Positive wall displacementa Negative wall displacementb
Front brace Back brace Front brace Back brace
(mm) (mm) (mm) (mm)
1A-M 58.6 58.5
1B-M 58.2 58.2 1C-M 58.3 58.7
2A-C 58.5 58.8 58.5 58.9
2B-C 58.5 58.4 58.4 58.4
2C-C 58.5 58.5 58.6 58.5
3A-M 101.2 101.3
3B-M 101.5 100.8
3C-M 101.3 101.0
4A-C 101.7 100.9 101.5 101.4
4B-C 104.1 104.3 104.3 104.2
4C-C 101.6 102.3 104.8 102.4
5A-M 152.2 152.6
5B-M 152.3 152.9
5C-M 152.4 152.2
6B-C 152.4 152.4 152.4 152.6
a Braces under tension during wall displacement in the positive direction.b Braces under tension during wall displacement in the negative direction.
Hence, it was not possible to rely on the 80% post peak-load
definition of the reference deformation [24,25]. Instead, the
yield displacement of the wall, y , was incorporated in the
calculation of the reference deformation for the determination
of the displacement amplitudes for the loading cycles. It
was assumed that = 2.667y , where y was obtained
from the nominally identical monotonic wall tests. Additional
information on the test program is provided by Al-Kharat and
Rogers [20].
4.1. Material tests
Material tests were carried out for the straps, chords and
tracks according to ASTM A370 [27] requirements. Coupon
tension tests were conducted at a cross-head rate of 0.5 mm
per minute in the elastic range, which was increased to a rate
of 4 mm per minute beyond the yield point. A 50 mm gauge
length extensometer was used to measure the extension of the
coupon and to calculate percentage of elongation, yield stress
and ultimate stress. Table 3 contains a listing of the minimum
specified (nominal design value) material yield stress, Fyn , and
thickness, as well as the measured yield stress, Fy , ultimate
stress, Fu , percent elongation and ratio of Fu /Fy , in addition
to the ratio of measured to nominal yield stress, Fy /Fyn . To
determine the base metal thickness of the material, the zinc
coating was removed with a 10% hydrochloric acid (HCL)
solution after testing. All of the steels used in the construction
of the test walls met the requirements of the North American
Specification for Cold-Formed Steel Members [3,6]. That is,the ratio of Fu /Fy was greater than 1.08, and the elongation
over a 50 mm gauge length exceeded 10%. It should be noted
that the 1.22 mm Grade 230 MPa steel was measured to have
a yield stress 54% greater than the minimum nominal specified
value, Fy /Fyn .
4.2. Modes of failure
In terms of ductile seismic performance, the desirable mode
of failure of a cold-formed steel braced wall system is generally
that of gross-cross section yielding of the straps, which form the
fuse element in the SFRS. The other elements and connectionsin the seismic force resisting system are expected to carry
the force associated with the strap yielding load level. The
strap braces should be able to enter into the inelastic range
of behaviour such that ground motion induced energy can be
dissipated. Ideally, the braces would be able to maintain their
yield capacity over extended lateral inelastic displacement of
the wall without failure of the connections, gusset plates, tracks,
chord studs or holddowns.
In general, the overall performance of the tested walls
under lateral loading was not governed by the yielding of the
straps, as indicated by the strain gauge measurements that were
taken [20]. Rather, failure of or extensive damage to the tracks,
chord studs, gusset plates, holddown threaded rods and straps
(due to net section fracture) was often observed depending on
the wall configuration being tested. These undesirable modes
of failure prevented the straps from maintaining their yield
load, or from yielding altogether. Thus the ductility and energy
absorption ability of the SFRS was reduced in comparison
to what could theoretically be expected given the material
properties of the strap braces and what inherently would be
assumed when a seismic response modification coefficient of
R = 4.0 is selected in design. A summary of the dominant
failure modes is provided below. More detailed information can
be found in Al-Kharat and Rogers [20].
Table 3
Material properties of strap and frame members
Member Nominal grade
(MPa)
Nominal thickness
(mm)
Base metal thickness
(mm)
Yield stress (Fy )
(MPa)
Ultimate stress (Fu )
(MPa)
Fu /Fy % Elng. Fy /Fyn
Strap 230 1.22 1.16 353 440 1.24 33 1.54
Strap 230 1.52 1.48 279 350 1.25 40 1.21
Strap 230 1.91 1.83 262 346 1.32 38 1.14
Track 230 1.22 1.22 320 380 1.19 31 1.39
Stud 230 1.22 1.23 336 398 1.19 35 1.46
Track 345 1.52 1.59 330 400 1.21 35 0.96
Stud 345 1.52 1.56 329 397 1.21 39 0.95
Track 345 1.91 1.94 348 474 1.36 37 1.01
Stud 345 1.91 1.91 352 489 1.39 35 1.02
-
7/29/2019 1-s2.0-S0143974X06001465-main
8/15
M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 46 0474 467
Fig. 6. Typicaltrackand connection failure modesat holddown location in light
walls.
Strain gauge measurements showed that yielding of the
straps occurred in the light walls; however, this was always
combined with the progressive compression failure of the
track and/or failure of the chord-to-track connection (Fig. 6).
This failure mode was similar to that reported by Fulop and
Dubina [16] for their strap braced wall tests. Fracture of a brace
occurred in only one wall (Test 1A-M) after approximately
30103 rad of shear deformation. In the first test to be carried
out (Test 1B-M) tension fracture of the 12.7 mm dia. anchorrod took place because, although adequate for the assumed 20
kN wind loading design level, the threaded rod was not able to
carry the force associated with the actual yield capacity of the
braces. The anchor rod size was increased to 15.9 mm dia. for
all subsequent tests of light walls to avoid this mode of failure.
Yielding of the straps occurred in the medium size walls
only when the larger holddown plates were installed. Even
so, the straps were not able to maintain their yield force level
due to extensive damage to the area adjacent to the holddown,
specifically in the track and gusset plates. Specimens 3A-M
and 3B-M, which were the first of this series to be tested,
were outfitted with 19 90 127 mm holddown plates. Thepunching shear capacity of the tracks around these plates was
not adequate. For the remainder of the medium strap braced
wall specimens a larger holddown plate, 19 127 203 mm,
was installed in an attempt to alleviate the punching shear
failure mode. This was successful to some degree; however
punching shear failure of the tracks, as well as permanent
deformation of the gusset plates and chord studs were still
observed (Fig. 7). Furthermore, the gusset plates created a
rigid corner element that would rotate in-plane due to the lack
of stiffness in the holddown/track area and the anchor rod
(Figs. 7 and 8). Local buckling of the chord studs on the uplift
side of the wall was caused by the extensive corner rotation
and the resulting applied moment on the framing member(Fig. 8). In one case (3C-M) punching shear failure of the
bottom track was observed along with fracture of the strap brace
(Fig. 7). The in-plane rotation of the bottom wall corner caused
excessive tensile stresses on the lower side of the strap brace
that ultimately resulted in its failure.
The heavy walls in the test study were not able to
demonstrate yielding of the strap braces along their length.
Extensive damage to the frame and gusset area adjacent to
the holddown plate was typically observed (Fig. 9); modes of
failure that would not be expected under a capacity based design
approach. Punching shear failure of the track occurred in all
tests, which did not allow the braces to reach their yield capacity
Fig. 7. Typical punching shear failure mode at holddown location in medium
walls.
Fig. 8. Medium wall post-test deformations and flexural failure of chord studs.
Fig. 9. Typical punching shear failure mode at holddown location in heavy
walls.
in tension. It was also common to observe the chord studs
being pulled in towards the centre of the wall due to the loss
of compression resistance in the track and gusset plates after
punching shear failure had taken place. Pull-out of the screws
that connected the interior studs to the bottom tracks was alsowitnessed, mainly due to the large deformations experienced by
the walls. Similar to the medium walls, rotation of the corner
connections took place, which led to moment induced local
buckling of the chord studs (Fig. 8).
The failure modes that were observed were mainly due to the
fact that the selection of holddowns and anchor rods, as well as
other elements in the SFRS, was not based on an estimate of
the yield capacity of the strap braces. Instead, each of the SFRS
elements were chosen given the expected factored lateral force
on the wall, which was assumed to be equal to the factored
capacity; 20 kN (light), 40 kN (medium) and 75 kN (heavy).
The actual lateral capacity of a braced wall is typically higher
-
7/29/2019 1-s2.0-S0143974X06001465-main
9/15
468 M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 460474
Fig. 10. Measured and predicted wall strength and stiffness.
than the factored design level; hence, in order to ensure that the
braces yield in tension along their length and that the yield load
level is maintained into the inelastic range it is important that
a capacity based approach be implemented in design. Although
the braces in the light walls did reach and maintain their yield
plateau, the inelastic performance of these walls could have
been improved by reinforcing the tracks, selecting a thicker
track section, or with the use of a corner detail that transferred
the horizontal brace forces into the supporting foundation by
means of tension. The medium and heavy walls did not exhibit
the inelastic behaviour that would be expected if a hysteretic
model, such as the one developed by Pastor and Rodrguez-
Ferran [18], were utilized for the nonlinear inelastic dynamicanalysis of X-braced cold-formed steel frames. Nor did the
observed inelastic failure modes match that which would be
associated with the use of a seismic response modification
coefficient of R = 4.0 as recommended in ASCE 7-05 [4].
4.3. Measured and predicted performance parameters
The maximum load level reached by each braced wall
regardless of the failure mode was defined as the measured
yield strength, Sy . The measured initial elastic shear stiffness,
Ke, was defined as the secant stiffness from the zero load level
to the 40% of maximum load level, S0.4, as recommended
in ASTM E2126 [25] (Fig. 10). These measured values werecompared with predicted wall properties, which were calculated
using the minimum specified (nominal) brace dimensions and
material properties (Table 1), as well as the measured brace
dimensions (Table 2) and material properties (Table 3). None of
these prediction methods are code specific. Given that the topic
of this paper is the inelastic response of strap braced walls, it is
the measured and predicted yield strengths that are of primary
concern, not the initial elastic stiffness or the factored shear
capacity. A schematic drawing that illustrates the measured and
predicted wall properties is shown in Fig. 10 for monotonic
wall test 1-AM. In a similar fashion the outer envelope of the
hysteretic loops was used to obtain the measured properties of
the reversed cyclic tests. Note: the position of the predicted
strengths, Syn and Syp , with respect to Sy may vary from what
is illustrated depending on the particular wall being analysed.
The predicted nominal lateral yield strength, Syn , of the wall
was based on the tension yield strength of the braces determined
using their nominal area (width thickness) as well as the
minimum specified (nominal) yield stress (230 MPa) (Table 1).
The nominal tension yield capacity of the brace was adjusted
for the inclined position of the strap members with respect to
the horizontal. The predicted nominal lateral shear stiffness of
the wall, Kn , was calculated based on the axial stiffness of the
two tension brace members, which was also adjusted for their
inclined position with respect to the horizontal. In this case thenominal design width, thickness and length of the strap braces
(Table 1) and E = 203 000 MPa were utilized. The reduction
in shear stiffness of the wall assembly due to the flexibility of
the brace connections and holddown was not accounted for.
The predicted values Syn and Kn represent the nominal (not
factored) design parameters that an engineer would typically
be able to determine using minimum specified member sizes
and material properties without the aid of test results and
measurements.
The lateral shear strength and stiffness parameters of each
test wall were also predicted using the measured width (Table 2)
and base metal thickness (Table 3) of the strap braces, as wellas the measured yield stress (Table 3) and the CSA S136 [3]
specified Youngs modulus (E = 203 000 MPa). Syp is the
predicted lateral yield strength of the wall, which is typically
reached when the strap braces yield in tension. Kp is the
predicted lateral shear stiffness of the wall, again obtained from
the initial elastic axial stiffness of the strap braces alone.
The ductility, , defined as the ability of the strap braced
wall system to maintain its yield capacity while attaining
significant inelastic lateral deformations, was also determined
(Eq. (1)).
=0.8
syp(1)
-
7/29/2019 1-s2.0-S0143974X06001465-main
10/15
M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 46 0474 469
Table 4
Summary of monotonic test results
Specimen Sy Ke Ductility () 0.8 Energy Syp Kp Sy /Syp Sy /Syn Ke/Kp Ke/Kn(kN) (kN/mm) (mm/mm) (rad 103) (kN mm) (kN) (kN/mm) (%) (%) (%) (%)
1A-M 31.97 1.29 3.01 32.3 2086 33.93 4.00 94 126 32 28
1B-M 30.39 1.59 2.32 20.2 1123 33.73 3.98 90 120 40 35
1C-M 31.96 1.48 3.88 36.7 2810 34.03 4.01 94 126 37 32Avg 31.44 1.45 3.07 29.7 2006 33.90 4.00 93 124 36 32
SD 0.909 0.152 0.782 8.547 846 0.153 0.015
CoV 0.029 0.104 0.255 0.287 0.422 0.005 0.004
3A-M 55.37 2.34 1.57 16.2 1478 58.88 9.20 94 110 25 24
3B-M 48.29 2.99 1.28 10.3 1332 58.84 9.20 82 96 33 31
3C-M 55.12 2.56 3.10 29.2 3610 58.77 9.19 94 109 28 27
Avg 52.93 2.63 1.98 18.5 2140 58.83 9.20 90 105 29 27
SD 4.017 0.331 0.978 9.644 1275 0.056 0.006
CoV 0.076 0.126 0.493 0.520 0.596 0.001 0.001
5A-M 82.93 3.61 2.21 25.9 5622 103.4 17.20 80 88 21 20
5B-M 59.79 5.97 1.78 12.7 1943 103.6 17.22 58 63 35 33
5C-M 81.23 3.85 2.09 23.0 3537 103.4 17.18 79 86 22 22
Avg 74.65 4.48 2.03 20.5 3701 103.5 17.20 72 79 26 25
SD 12.90 1.299 0.222 6.944 1845 0.108 0.020
CoV 0.173 0.290 0.109 0.339 0.499 0.001 0.001
Table 5
Summary of reversed cyclic test results
Specimen Sy Ke Ductility () 0.8 Energy Syp Kp Sy /Syp Sy /Syn Ke /Kp Ke/Kn(kN) (kN/mm) (mm/mm) (rad 103) (kN mm) (kN) (kN/mm) (%) (%) (%) (%)
2A-C (+ve) 35.26 1.27 4.11 45.1 10 167 34.00 4.01 104 139 32 28
2A-C (ve) 35.29 1.08 3.10 40.1 104 139 27 23
2B-C (+ve) 34.50 1.18 3.83 45.1 10 571 33.83 3.99 102 136 30 26
2B-C (ve) 34.47 1.18 3.83 48.0 102 136 30 26
2C-C (+ve)a 38.97 2.26 6.33 38.9 5 967 33.91 4.00 115 153 56 49
2C-C (ve) 35.49 1.22 4.22 46.7 . 105 140 31 27
Avg 35.00 1.19 3.82 44.0 8 902 33.91 4.00 103 138 30 26SD 0.48 0.07 0.44 3.637 2 550 0.08 0.01
CoV 0.014 0.059 0.114 0.083 0.286 0.002 0.002
4A-C (+ve) 59.47 2.36 1.98 20.3 19 006 58.98 9.22 101 118 26 25
4A-C (ve) 60.07 2.09 1.89 21.9 102 119 23 22
4B-C (+ve) 62.31 2.27 2.23 24.4 18 663 60.64 9.48 103 124 24 24
4B-C (ve) 60.59 2.05 1.87 22.6 100 120 22 21
4C-C (+ve) 55.69 2.21 2.00 22.3 18 513 59.80 9.35 93 110 24 23
4C-C (ve) 56.40 2.43 2.26 22.9 94 112 26 25
Avg 59.09 2.24 2.04 22.4 18 727 59.81 9.35 99 117 24 23
SD 2.55 0.15 0.17 1.355 253 0.74 0.12
CoV 0.043 0.067 0.082 0.060 0.013 0.012 0.012
6B-C (+ve) 87.13 3.79 2.01 22.6 26 051 103.5 17.2 84 92 22 21
6B-C (ve) 83.56 3.48 1.88 22.9 81 88 20 19
Avg 85.35 3.64 1.95 22.7 26 051 103.5 17.2 83 90 21 20
a 2C-C (+ve) not included in calculation of statistical parameters.
where 0.8 is the failure displacement in the post-peak range
that corresponds to a wall resistance of 80% of the maximum
level measured (Fig. 10). In an ideal case where failure is
limited to yielding of the brace members, no reduction in shear
strength should take place; and hence, the failure displacement
would be defined as the maximum in-plane displacement
reached by the wall as limited by the stroke of the actuator.
It was however typical for the walls in this study to experience
some decrease in shear resistance due to failure modes other
than yielding of the strap braces. The elastic yield deformation,
syp, was calculated using the measured elastic stiffness, Ke,
and the predicted lateral yield strength of the wall, Syp (Fig. 10).
A summary of the test results and predicted wall properties is
found in Tables 4 and 5.
4.4. Comparison of measured and predicted performance
4.4.1. Light walls
An average yield strength of Sy = 31.44 kN, equal to 93%
of the predicted value based on the measured properties of the
brace members (Sy /Syp ), was attained for the light walls tested
-
7/29/2019 1-s2.0-S0143974X06001465-main
11/15
470 M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 460474
Fig. 11. Cyclic resistance versus displacement curve of 58.4 mm (light) strap
braced wall 2A-C.
monotonically (Table 4). The full predicted capacity was not
reached most likely because of the increase in Fy due to the
strain rate effect that occurred during testing of the coupons.
The estimated strain rate for the coupons was 0.0667 /min,
which was 92 times greater than the approximate strain rate
determined for the monotonically tested strap walls (0.000725
/min). The average measured yield strength of the cyclic tests
was 35.0 kN, which was 3% higher than the predicted Sypvalue (Table 5). A higher yield capacity was measured because
the walls, which were tested at a frequency of 0.5 Hz, would
have reached a much higher strain rate than experienced by the
monotonic tests. This would likely have elevated the yield stress
of the strap braces.
A comparison of the measured yield strength reached, Sy ,
by each of the light walls with the factored design resistance
(20 kN) including both monotonic results as well as the results
of positive and negative excursions for the reversed cyclic
tests gave a ratio of 1.66. The test-to-nominal predicted shear
capacity, Sy /Syn , listed in Tables 4 and 5 was 1.24 and 1.38
for the monotonic and reversed cyclic tests, respectively; which
indicates that the test walls were able to reach the expected
nominal design shear strength. However, using the nominalshear strength of the wall based on the brace yield capacity
did not provide for an accurate estimate of the actual force in
the SFRS when the straps yield. An increase in the nominal
prediction of between 1.2 and 1.4 is necessary to determine a
more realistic force level for the brace connections, holddowns,
chord studs, etc. in a capacity based design context because
the actual yield stress of the strap material is higher than the
minimum specified 230 MPa.
The light walls were able to perform in a ductile manner;
that is, they were able reach and maintain their yield capacity
throughout most of the reversed cyclic loading protocols;
however, often the resistance decreased in the latter stages of
Fig. 12. Cyclic resistance versus displacement curve of 58.4 mm (light) strap
braced wall 2B-C.
each cycle. This reduction in load was caused by the damage
that occurred at the holddown/track-to-chord stud connection
locations (Fig. 6). An example of this can be seen in Fig. 11,
where for the negative load/displacement region of test 2A-C
there is a significant decrease in load above the 30 103
rad displacement level compared with test 2B-C (Fig. 12).
On average the light walls reached a 0.8 (Fig. 10) shear
deformation of 34.5 103 rad and 44.0 103 rad for
the monotonic and reversed cyclic tests. Note; the monotonicaverage was calculated without using the result from test wall
1B-M because it failed prematurely due to the fracture of the
12.7 mm dia. threaded anchor rod. These shear deformation
measurements correspond with an average ductility of 3.07 and
3.82 for the monotonic and reversed cyclic tests, respectively.
A test-based estimate of the ductility related seismic force
modification factor, Rd, can be defined as shown in Eq. (2) [28].
An average Rd of 2.48 was obtained for the light strap braced
walls. Furthermore, Mitchell et al. [29] recommended that the
overstrength related seismic force modification factor, Ro, can
be estimated by considering the product of the average Sy /Syn
ratio and the inverse of the resistance factor, 1/. CSA S136 [3]specifies that = 0.9 for gross cross section yielding design of
a tension member. For comparison purposes the ASCE 7-05 [ 4]
defined R-value can be thought of as the product of the ductility
related, Rd, and overstrength related, Ro, force modification
factors as found in the 2005 NBCC [1]. The resulting overall
average Ro factor of 1.47 and the Rd factor noted above provide
for an R value of 3.65 for the light walls. This test-based
estimate of the seismic response modification coefficient is
between the R = 3.0 and 4.0 recommended by ASCE 7-05.
Reinforcement of the track member of the wall, such that it is
capable of carrying the strap yield forces, would likely improve
the ductility such that the test-based R is at a level of 4.0
-
7/29/2019 1-s2.0-S0143974X06001465-main
12/15
M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 46 0474 471
or higher.
Rd =
2 1. (2)
In terms of predicted stiffness levels, none of the walls
were able to reach the expected 4.00 kN/mm. As is illustrated
in Fig. 11, Ke was substantially lower than Kp. In a similar
fashion the nominal stiffness, Kn , was also not reached byany of the light walls. This can also be seen in the Ke/Kpand Ke/Kn ratios provided in Tables 4 and 5. These predicted
stiffness values were based solely on the dimensions and
material properties of the straps. From observations of the large
deformations and damage at the holddown locations, as well as
the measured stiffness values, it is apparent that in this case the
flexibility of the holddowns and brace connections has caused a
decrease in the stiffness of the test walls. The predicted lateral
in-plane elastic stiffness of the braced wall cannot be based
solely on the axial stiffness of the straps.
4.4.2. Medium wallsA performance ratio of Sy /Syp = 90%, which corresponds
to an average monotonic lateral resistance of 52.93 kN
was measured for the medium walls (Table 4). The 10%
shortcoming in the ratio of Sy /Syp is again likely due to the
strain rate used for the monotonic wall testing compared with
that used for the coupons. The early onset of punching shear
failure in the tracks may also have limited the capacity of the
wall. Similar findings were obtained for the reversed cyclic tests
except that a higher average resistance of 59.09 kN was reached
(Table 5). This resulted in a performance ratio of Sy /Syp =
99%. The increased load levels can be attributed to the strain
rate effect experienced by the strap braces. Nonetheless, as isshown in Figs. 13 and 14, these walls were unable to maintain
their load carrying capacity due to punching shear failure of
the tracks (Fig. 7). No yield plateau was observed; instead
a sharp peak resistance was recorded, followed by a sudden
degradation in load carrying ability. The fuse element in the
seismic force resisting system ultimately was the holddown
plate/anchor rod/track connection in combination with the strap
braces.
A comparison of Sy for the medium walls with the factored
design resistance (40 kN) for the monotonic and reversed
cyclic tests gave a ratio of 1.40. The test-to-nominal predicted
shear capacity, Sy /Syn , listed in Tables 4 and 5 was 1.05 and
1.17 for the monotonic and reversed cyclic tests, respectively;which shows that the wall was able to reach the expected
nominal design shear strength. However, the nominal shear
strength of the walls based on the brace yield capacity did not
provide for a precise estimate of the actual force in the SFRS
when the straps yield, although the estimate was improved
compared with the light walls. An increase in the nominal
prediction of approximately 1.2 is necessary to determine a
more realistic force level for the brace connections, holddowns,
chord studs, etc. in a capacity based design context. Note that
this increase is somewhat lower than that recommended for the
light walls, mainly because of the significantly higher material
yield strength, Fy , that was measured for the braces of the
Fig. 13. Monotonic resistance versus displacement curve of 101 mm (medium)
strap braced wall 3C-M.
Fig. 14. Cyclic resistance versus displacement curve of 101 mm (medium)
strap braced wall 4B-C.
light walls compared with the nominal (minimum specified)
230 MPa (Table 3).
The medium walls attained a 0.8 (Fig. 10) shear deforma-
tion of 18.5 103 rad and 22.4 103 rad for the monotonic
and reversed cyclic tests. The corresponding average ductility
for the monotonic and reversed cyclic tests was 1.98 and 2.04,
respectively. These ductility measurements provided an average
Rd value of 1.71 for the medium walls. An overstrength related
Ro value of 1.23 was also obtained, which when combined with
the Rd value results in a test-based R value of 2.11, far below
-
7/29/2019 1-s2.0-S0143974X06001465-main
13/15
472 M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 460474
Fig. 15. Monotonic resistance versus displacement curve of 152 mm (heavy)
strap braced wall 5A-M.
that specified in ASCE 7-05 [4]. The punching shear mode of
failure did not allow for the yield capacity of the braces to be
maintained; thus, the flat plate holddown detail used for the
medium walls did not provide for a sufficient level of ductility
to warrant a seismic response modification coefficient equal to
4.0 or even 3.0. It is recommended that the flat plate holddown
detail not be used for braced walls with unlipped channel track
sections because of the lack of a direct connection between the
straps and anchor device.The average Ke of 2.63 kN/mm for the monotonic tests was
well below the expected stiffness, Kp = 9.20 kN/mm, due to
the flexibility of the flat plate holddown detail, as well as the
extreme damage that occurred (Fig. 7). Likewise, the nominal
elastic stiffness, Kn , was also not reached by any of the test
walls.
4.4.3. Heavy walls
The monotonic tests 5A-M, 5B-M and 5C-M (heavy walls)
had the lowest performance ratio of all the strap braced walls
that were included in the study. An average capacity of 74.65
kN was measured, which corresponds to an Sy /Syp ratio of72% (Table 4). Yielding was seen in some areas of the braces,
based on strain gauge measurements; however the overall yield
capacity of the brace was not reached at any time (Fig. 15). As
was observed for the medium walls, punching shear failure of
the track controlled the wall resistance, stiffness and ductility
(Fig. 9). Given the poor results of the monotonic tests only
one reversed cyclic test was completed (6B-C) (Fig. 16).
The average maximum resistance of the negative and positive
displacement cycles was 85.35 kN (Table 5). This provided a
performance ratio ofSy /Syp = 83%, somewhat higher than the
monotonic tests, but not adequate when compared to the shear
load associated with brace yielding.
Fig. 16. Cyclic resistance versus displacement curve of 152 mm (heavy) strap
braced wall 6B-C.
A comparison of Sy for the heavy walls with the factored
design resistance (75 kN) for the monotonic and reversed
cyclic tests gave a ratio of 1.07. The test-to-nominal predicted
shear capacity, Sy /Syn , listed in Tables 4 and 5 was 0.79 and
0.9 for the monotonic and reversed cyclic tests, respectively,
which shows that the walls were not even able to attain their
expected nominal design shear strength. Punching shear failure
of the track members limited the amount of force that could be
transferred to the brace members; hence, the strap braces did
not at any time reach their potential tensile yield capacity whilebeing tested.
Punching shear failure of the track once again controlled the
behaviour of the wall. As found for the medium strap braced
walls, the flat plate holddown detail was inadequate to allow
for the wall to maintain its yield capacity, and hence to act
in a ductile fashion. This can be seen in the measured 0.8(Fig. 10) shear deformation of 20.5 103 rad and 22.7
103 rad reached by the monotonic and reversed cyclic tests.
The average ductility calculated using these shear deformation
values for the monotonic and reversed cyclic tests was 2.03 and
1.95, respectively; which was similar to the medium walls. An
average Rd value of 1.72 was obtained; however, because theSy /Syn ratio was below 1.0, no overstrength existed and as such
Ro = 1.0. The product of Rd and Ro gives a test-based R
value of 1.72, again significantly lower than that specified in
ASCE 7-05 [4] for strap braced bearing wall systems. It is again
recommended that the flat plate holddown detail not be used for
braced walls with unlipped channel track sections because of
the lack of a direct connection between the straps and anchor
device.
The measured initial elastic stiffness, Ke, was in the range
of 20%26% of the expected Kp and Kn values (Tables 4 and
5). This finding can again be attributed to the flexibility of the
flat plate holddown detail.
-
7/29/2019 1-s2.0-S0143974X06001465-main
14/15
M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 46 0474 473
5. Conclusions and recommendations
In general, the strap braced test walls, as constructed, were
not able to maintain a yield level load carrying capacity over
extended displacements, with the exception of the light walls.
Moreover, the heavy walls were not able to even reach the
load level associated with gross cross-section yielding of thebraces. The extensive damage to the holddown/gusset/chord
stud/track location in almost all test walls showed that the
inelastic deformations were not limited to the brace elements of
the lateral force resisting system. Furthermore, punching shear
failure of the tracks severely reduced the inelastic performance
of the medium and heavy walls.
Given the results of testing, it is not possible to consider the
walls to have performed in the ductile fashion that would have
been associated with a response modification coefficient ofR =
4.0 and assumed if a capacity based design approach had been
followed. The medium and heavy walls did not even possess
the ductility and overstrength to validate the use of R = 3.0, as
allowed in ASCE7-05 when seismic details are not incorporatedin the design of a structure. However, the fact that the test
walls were designed and constructed without capacity based
concepts in mind does indicate that the inelastic performance
could possibly be improved if additional design steps were
taken. This is most evident for the light walls in which a test-
based RdRo value of 3.65 was attained. In contrast, however,
the medium and heavy walls were only able to exhibit a test-
based Rd Ro value of 2.11 and 1.72, respectively; which is
approximately half of the ASCE7-05 upper R-value specified
for cold-formed steel strap braced bearing wall systems.
An estimate of the force in the SFRS due to brace yielding
needs to account for the possible overstrength of the strapmaterial, such that failure or plastic deformation of other
elements in the lateral load path is avoided. The nominal
capacity of the strap members (Ag Fy ) does not indicate
the true force level that may be reached in the system. This is
due to the actual yield strength of the cold-formed steel strap
members, which in the case of this study reached as high as
1.54 times the minimum specified 230 MPa.
It is recommended that supplementary tests of similar size
strap braced walls be carried out, for which the elements
in the seismic force resisting system are selected based on
the probable yield capacity of the strap braces. An accurate
estimate of the yield stress of the brace material is needed,
which accounts for both the effects of the higher than minimum
nominal yield stress due to the manufacturing processes and the
strain rate under seismic loading. Furthermore, the holddown
detail needs to be improved, such that the probable brace loads
can be carried with minimal rotation and inelastic damage to
the track, chord studs, gusset plate, anchor rod and holddown
itself. In terms of recommendations for designers, at the very
least it is necessary that a capacity based design approach be
implemented for the selection of SFRS elements. The use of
corner holddown plates placed in the bottom and top tracks of
a strap braced wall does not provide for an adequate transfer of
brace induced forces due to the possibility of punching shear
failure. Moreover this holddown failure mode is not ductile
in nature, and hence does not allow for the strap brace yield
capacity to be maintained. Additional research that includes
more specific detailing requirements for strap braced walls is
ongoing.
Acknowledgements
The authors would like to acknowledge the support provided
by the Canada Foundation for Innovation and the Canadian
Sheet Steel Building Institute. Test specimens were generously
supplied by Genesis by KML Ltd. of Cambridge, ON, Canada.
A thank you is also extended to the students K.E. Hikita,
A. Frattini, T.L.W. Lim and Z. Fu for their assistance in carrying
out the braced wall tests.
References
[1] National Research Council of Canada (NRCC). National Building Code
of Canada. Ottawa (ON, Canada); 2005.
[2] Heidebrecht AC. Overview of seismic provisions of the proposed 2005
edition of the National Building Code of Canada. Canadian Journal of
Civil Engineering 2003;30(2):24154.
[3] Canadian Standards Association S136. North American specification for
the design of cold-formed steel structural members. Mississauga (ON,
Canada); 2001.
[4] American Society of Civil Engineers ASCE 7. Minimum design loads for
buildings and other structures. Reston (VA, USA); 2005.
[5] American Iron and Steel Institute. Standard for cold-formed steel framing
lateral design. Washington (DC, USA); 2004.
[6] American Iron and Steel Institute. North American specification for the
design of cold-formed steel structural members. Washington (DC, USA);
2001.
[7] American Iron and Steel Institute. Standard for cold-formed steel framing
General provisions. Washington (DC, USA); 2001.
[8] TI 809-07 . Technical instructions: Design of cold-formed loadbearingsteel systems and masonry veneer/steel stud walls. Washington (DC,
USA): US Army Corps of Engineers. Engineering and Construction
Division. Directorate of Civil Works; 2003.
[9] Adham SA, Avanessian V, Hart GC, Anderson RW, Elmlinger J,
Gregory J. Shear wall resistance of lightgage steel stud wall systems.
Earthquake Spectra 1990;6(1):114.
[10] Serrette R, Ogunfunmi K. Shear resistance of gypsum-sheathed light-
gauge steel stud walls. Journal of Structural Engineering ASCE 1996;
122(4):3839.
[11] Barton AD. Performance of steel framed domestic structures subject to
earthquake loads. Ph.D. thesis. Melbourne (Australia): Department of
Civil and Environmental Engineering, University of Melbourne; 1997.
[12] Gad EF, Chandler AM, Duffield CF, Hutchinson GL. Earthquake
ductility and overstrength in residential structures. Journal of Structural
Engineering and Mechanics 1999;8(4):36182.[13] Gad EF, Duffield CF, Hutchinson GL, Mansell DS, Stark G. Lateral
performance of cold-formed steel-framed domestic structures. Journal of
Engineering Structures 1999;21:8395.
[14] Gad EF, Chandler AM, Duffield CF, Stark G. Lateral behaviour
of plasterboard-clad residential steel frames. Journal of Structural
Engineering ASCE 1999;125(1):329.
[15] Park R. Evaluation of ductility of structures and structural assemblages
from laboratory testing. Bulletin of the New Zealand National Society for
Earthquake Engineering 1989;22(3).
[16] Fulop LA, Dubina D. Performance of wall-stud cold-formed shear panels
under monotonic and cyclic loading. Part I: Experimental research. Thin
Walled Structures 2004;42:32138.
[17] Tian YS, Wang J, Lu TJ. Racking strength and stiffness of cold-formed
steel wall frames. Journal of Constructional Steel Research 2004;60:
106993.
-
7/29/2019 1-s2.0-S0143974X06001465-main
15/15
474 M. Al-Kharat, C.A. Rogers / Journal of Constructional Steel Research 63 (2007) 460474
[18] Pastor N, Rodrguez-Ferran A. Hysteretic modelling of x-braced shear
walls. Thin Walled Structures 2005;43:156788.
[19] Casafont M, Arnedo A, Roure F, Rodrguez-Ferran A. Experimental
testing of joints for seismic design of lightweight structures. Part 1.
Screwed joints in straps. Thin Walled Structures 2006;44:197210.
[20] Al-Kharat M, Rogers CA. Testing of light-gauge steel strap braced walls.
Research report. Montreal (QC, Canada): Dept. of Civil Engineering,
McGill University; 2005.
[21] ASTM A653. Standard specification for steel sheet, zinc-coated
(galvanized) or zinc-iron alloy-coated (galvannealed) by the hot-dip
process. West Conshohocken (PA, USA); 2002.
[22] ASTM A307. Standard specification for carbon steel bolts and studs, 60
000 psi tensile strength. West Conshohocken (PA, USA); 2003.
[23] ASTM A325. Standard specification for structural bolts, steel, heat treated
120/105 ksi minimum tensile strength. West Conshohocken (PA, USA);
2002.
[24] Krawinkler H, Parisi F, Ibarra L, Ayoub A, Medina R. Development of a
testing protocol for woodframe structures. Report W-02, CUREE/Caltech
woodframe project. Richmond (CA, USA); 2000.
[25] ASTM E2126. Standard test methods for Cyclic (reversed) load test for
shear resistance of framed walls for buildings, West Conshohocken (PA,
USA); 2005.
[26] Branston AE, Chen CY, Boudreault FA, Rogers CA. Testing of light-
gauge steel-frame wood structural panel shear walls. Canadian Journal
of Civil Engineering 2006;33(5):56172.
[27] ASTM A370. Standard test methods and definitions for mechanical
testing of steel products. West Conshohocken (PA, USA); 2002.
[28] Newmark NM, Hall WJ. Earthquake spectra and design. Engineering
monograph, Berkeley (CA, USA): Earthquake Engineering Research
Institute; 1982.
[29] Mitchell D, Tremblay R, Karacabeyli E, Paultre P, Saatcioglu M,
Anderson DL. Seismic force modification factors for the proposed 2005
edition of the National Building Code of Canada. Canadian Journal of
Civil Engineering 2003;30(2):30827.