1,2,* , phan viet nhut 1,3, daiki nakamoto 1

polymers

Article

Experimental Investigation on the Buckling Capacity of AngleSteel Strengthened at Both Legs Using VaRTM-ProcessedUnbonded CFRP Laminates

Fengky Satria Yoresta 1,2,* , Phan Viet Nhut 1,3, Daiki Nakamoto 1 and Yukihiro Matsumoto 1

��

Citation: Yoresta, F.S.; Nhut, P.V.;

Nakamoto, D.; Matsumoto, Y.

Experimental Investigation on the

Buckling Capacity of Angle Steel

Strengthened at Both Legs Using

VaRTM-Processed Unbonded CFRP

Laminates. Polymers 2021, 13, 2216.

https://doi.org/10.3390/

polym13132216

Academic Editor: Shazed Aziz

Received: 21 June 2021

Accepted: 1 July 2021

Published: 5 July 2021

Publisher’s Note: MDPI stays neutral

with regard to jurisdictional claims in

published maps and institutional affil-

iations.

Copyright: © 2021 by the authors.

Licensee MDPI, Basel, Switzerland.

This article is an open access article

distributed under the terms and

conditions of the Creative Commons

Attribution (CC BY) license (https://

creativecommons.org/licenses/by/

4.0/).

1 Department of Architecture and Civil Engineering, Toyohashi University of Technology,Toyohashi 441-8580, Aichi, Japan; [email protected] (P.V.N.); [email protected] (D.N.);[email protected] (Y.M.)

2 Engineering and Building Design Laboratory, Bogor Agricultural University, Bogor 16680, Indonesia3 Department of Civil Engineering, The University of Danang—University of Technology and Education,

Danang City 550000, Vietnam* Correspondence: [email protected]

Abstract: Strengthening steel structures by using carbon fiber reinforced polymer (CFRP) laminatesshowed a growth trend in the last several years. A similar strengthening technique, known as adhe-sive bonding, has also been adopted. This paper presented a promising alternative for strengtheningsteel members against buckling by using vacuum-assisted resin transfer molding (VaRTM)-processedunbonded CFRP laminates. A total of thirteen slender angle steel members (L65x6), including twocontrol specimens, were prepared and experimentally tested. The specimens were strengthened onlyat both legs and were allowed to buckle on their weak axes. The test showed that the unbonded CFRPstrengthening successfully increased the buckling capacity of the angle steel. The strengthening effectranged from 7.12% to 69.13%, depending on various parameters (i.e., number of CFRP layers, CFRPlength, and angle steel’s slenderness ratio). Flexural stiffness of the CFRP governed the failure modesin terms of location of plastic hinge and direction of buckling curvature.

Keywords: unbonded CFRP; angle steel; buckling; VaRTM; strengthening

1. Introduction

The use of CFRP as a strengthening material for concrete structures has gained wideacceptance worldwide due to its excellent properties [1–5], i.e., light weight, excellentfatigue behavior, resistance to corrosion, high strength-to-weight ratio, and easy installation.However, this has not entirely been the case for steel structures. Research and developmenton the use of CFRP for strengthening steel structures is still limited. A method that has beendeveloped is the same as that used for strengthening concrete structures, which involvesattaching the CFRP to the steel surface [6–10] and is well known as an externally bondedstrengthening technique.

Indeed, the externally bonded strengthening technique is effective for enhancing theperformance of steel structures. It has been convincingly proven that this technique isable to increase: the compression capacity of short steel members [11–13], flexural ca-pacity of steel beams [14–18], torsional capacity of steel members [19,20], load bearingcapacity of steel shear wall [21–23], tensile capacity of steel plates [24–28], and also buck-ling capacity of long steel columns [29–31]. However, this strengthening technique isconsidered to have drawbacks. The most critical issue is bond strength between CFRPand steel substrate [32,33]. Debonding failure between steel and CFRP has occurred inmany investigations [34–38], often before CFRP can provide its maximum contribution asa strengthening material. In order to obtain a better bond between CFRP and steel, thebonded strengthening technique requires proper steel surface treatments (for example,sandblasting, hand grinding, and/or grit blasting) prior to installation of CFRP. The surface

Polymers 2021, 13, 2216. https://doi.org/10.3390/polym13132216 https://www.mdpi.com/journal/polymers

https://www.mdpi.com/journal/polymers

https://www.mdpi.com

https://orcid.org/0000-0003-2909-5800

https://orcid.org/0000-0002-5935-6576

https://doi.org/10.3390/polym13132216



https://creativecommons.org/

https://creativecommons.org/licenses/by/4.0/

https://creativecommons.org/licenses/by/4.0/


https://www.mdpi.com/journal/polymers

https://www.mdpi.com/article/10.3390/polym13132216?type=check_update&version=1

Polymers 2021, 13, 2216 2 of 17

treatment activity becomes much more complicated and difficult when it is applied to exist-ing steel buildings because of the uncertainty and complexity of the real conditions in field.Different types of steel surface treatments will also effect to the bonding strength of CFRPto steel [39]. Moreover, the choice of one of the treatments cannot also completely ensurethat the bond strength of CFRP to steel will be uniform because there can be differencesin the quality of the steel surface after treatment. This is largely because the workers whoperform the treatment often have different levels of skill [40].

Even supposing that steel surface treatments can be performed very well, the bondstrength between steel and CFRP remains prone to deterioration due to inevitable environ-mental exposure. Experimental research has proven that the load-carrying capacity of theCFRP/steel adhesive-bonded joints is significantly reduced at temperatures around andabove the glass transition temperature (Tg) of the adhesive [41,42]. The CFRP/steel bondstrength also decreases in subzero temperatures (−40 ◦C, using MBrace Saturant) [43]. Inaddition, moisture absorption can affect the performance of CFRP composites [44,45].

In this paper, unbonded CFRP is proposed as an alternative to avoid the problemsarising in the bonding method. The unbonded CFRP is used to strengthen angle steelagainst buckling. The unbonded CFRP strengthening method eliminates the steel surfacepreparation process required by the adhesive-bonding method, because CFRP is not ad-hesively attached onto steel substrate [46]. An unbonded layer will exist between thesetwo materials as a separator. This means unbonded CFRP strengthening can be appliedfaster and easier on-site. Moreover, as it does not depend on the bond between steeland CFRP, the unbonded CFRP strengthening method is likelier to be used for a longerperiod of time compared to the adhesive-bonded strengthening. There will be no concernsabout the degradation of strengthening performance due to environmental exposure. Theperformance of unbonded CFRP strengthening will depend only on the flexural rigidity ofthe CFRP laminates.

2. Materials and Methods2.1. Materials

Angle steel L-65x6 is used as the main material in this experimental program. Thesteel has a grade SS400 according to JIS G3101:2020 (Japanese Industrial Standard).Figure 1 shows cross section of the steel with values of each symbol given in Table 1. Itshould be noted that measured wall thickness ta is 5.56 mm, which is slightly smaller thanthe nominal wall thickness t given in the table. Material properties of the steel angle, whichare also presented in the table, are obtained from the inspection certificate issued by themanufacturer. There was no tensile coupon test conducted in this study.

Polymers 2021, 13, x FOR PEER REVIEW 3 of 17

83

Figure 1. Angle steel cross-section. 84

Table 1. Parameter and properties of the steel angle. 85

Designation Cross-Section Dimension (mm) Yield Stress

(MPa) Tensile Strength

(MPa) Elastic Modulus

(MPa) L-65x6 A = B = 65; t = 6.0; r1 = 8.5; and r2 = 4.0 333 448 205,000

The CFRP laminates are made of bidirectional carbon fiber (termed as BT70-20) 86 which is a product of Toray Industries, Inc., Tokyo, Japan. In one sheet, this material has 87 two directions of fiber which are perpendicular to each other (Figure 2a). With this con-88 figuration, a laminate with the same properties in both directions of the weave (0° and 89 90°) can easily be produced. This is the main reason for choosing this material in this 90 study. Table 2 shows properties and characteristics of carbon fiber BT70-20. All the data 91 given are based on the manufacturer’s data sheet. In addition to carbon fiber, the adhesive 92 used is epoxy resin (a product of Konishi Co., Ltd., Osaka, Japan). Modulus of elasticity, 93 Poisson’s ratio, and tensile strength of this material are 1.285 GPa, 0.46, and 18.50 MPa, 94 respectively. These data are average values obtained from tensile tests of six specimens. 95 Figure 2b shows a representation of the stress–strain curve of the epoxy resin. 96

(a) (b)

Figure 2. Materials for CFRP: (a) carbon fiber BT70-20; (b) typical stress–strain curve of the epoxy adhesive. 97

98

0

6

12

18

24

0 3 6 9 12

Stre

ss,σ

(MPa

)

Axial strain, ε (%)

Figure 1. Angle steel cross-section.

Polymers 2021, 13, 2216 3 of 17

Table 1. Parameter and properties of the steel angle.

Designation Cross-SectionDimension (mm) Yield Stress (MPa) Tensile Strength

(MPa) Elastic Modulus(MPa)

L-65x6 A = B = 65; t = 6.0;r1 = 8.5; and r2 = 4.0 333 448 205,000

The CFRP laminates are made of bidirectional carbon fiber (termed as BT70-20) whichis a product of Toray Industries, Inc., Tokyo, Japan. In one sheet, this material has twodirections of fiber which are perpendicular to each other (Figure 2a). With this configuration,a laminate with the same properties in both directions of the weave (0◦ and 90◦) can easilybe produced. This is the main reason for choosing this material in this study. Table 2 showsproperties and characteristics of carbon fiber BT70-20. All the data given are based on themanufacturer’s data sheet. In addition to carbon fiber, the adhesive used is epoxy resin(a product of Konishi Co., Ltd., Osaka, Japan). Modulus of elasticity, Poisson’s ratio, andtensile strength of this material are 1.285 GPa, 0.46, and 18.50 MPa, respectively. Thesedata are average values obtained from tensile tests of six specimens. Figure 2b shows arepresentation of the stress–strain curve of the epoxy resin.


83

Figure 1. Angle steel cross-section. 84

Table 1. Parameter and properties of the steel angle. 85

Designation Cross-Section Dimension (mm) Yield Stress

(MPa) Tensile Strength


(MPa) L-65x6 A = B = 65; t = 6.0; r1 = 8.5; and r2 = 4.0 333 448 205,000

The CFRP laminates are made of bidirectional carbon fiber (termed as BT70-20) 86 which is a product of Toray Industries, Inc., Tokyo, Japan. In one sheet, this material has 87 two directions of fiber which are perpendicular to each other (Figure 2a). With this con-88 figuration, a laminate with the same properties in both directions of the weave (0° and 89 90°) can easily be produced. This is the main reason for choosing this material in this 90 study. Table 2 shows properties and characteristics of carbon fiber BT70-20. All the data 91 given are based on the manufacturer’s data sheet. In addition to carbon fiber, the adhesive 92 used is epoxy resin (a product of Konishi Co., Ltd., Osaka, Japan). Modulus of elasticity, 93 Poisson’s ratio, and tensile strength of this material are 1.285 GPa, 0.46, and 18.50 MPa, 94 respectively. These data are average values obtained from tensile tests of six specimens. 95 Figure 2b shows a representation of the stress–strain curve of the epoxy resin. 96

(a) (b)

Figure 2. Materials for CFRP: (a) carbon fiber BT70-20; (b) typical stress–strain curve of the epoxy adhesive. 97

98

0

6

12

18

24

0 3 6 9 12

Stre

ss,σ

(MPa

)

Axial strain, ε (%)

Figure 2. Materials for CFRP: (a) carbon fiber BT70-20; (b) typical stress–strain curve of the epoxy adhesive.

Table 2. Carbon fiber.

Type Orientation ofFiber

Tensile Strength(MPa)

Elastic Modulus(MPa)

Thickness(mm)

BT70-200o 2900 230,000 0.056

90o 2900 230,000 0.056

2.2. Description of Specimens

The specimens tested in this study consist of two groups of angle steel with differentslenderness ratios (λ). Each group has one control specimen that is not strengthened withCFRP. Specimens of Group A (λ = 128.4) are divided into two different CFRP lengths,i.e., 1000 mm and 500 mm, while specimens of Group B (λ = 97.02) have CFRP lengthof 500 mm. The number of CFRP layers in strengthened specimens is varied to furtherinvestigate the effect of strengthening. The CFRP is positioned in the middle of the steel,as shown in Figure 3. The total number of specimens tested is thirteen, and their detailedparameters are given in Table 3. As shown in the table, specimens are labeled for easyidentification. The first combination of ‘letter-number’ describes angle steel and its bucklinglength, the second combination is strengthening length, and the last combination indicatesnumber of CFRP layers. For instance, A16S10L25 identifies a specimen with 1636 mm

Polymers 2021, 13, 2216 4 of 17

buckling length, 1000 mm CFRP length, and 25 CFRP layers. Here, buckling length is atotal of specimen length L (Figure 3) and thickness of additional steel plates at both ends ofspecimen (Figure 4).


Table 2. Carbon fiber. 99

Type Orientation

of Fiber Tensile Strength


(MPa) Thickness

(mm)

BT70-20 0o 2900 230,000 0.056 90o 2900 230,000 0.056

2.2. Description of Specimens 100

The specimens tested in this study consist of two groups of angle steel with different 101 slenderness ratios (λ). Each group has one control specimen that is not strengthened with 102 CFRP. Specimens of Group A (λ = 128.4) are divided into two different CFRP lengths, i.e., 103 1000 mm and 500 mm, while specimens of Group B (λ = 97.02) have CFRP length of 500 104 mm. The number of CFRP layers in strengthened specimens is varied to further investi-105 gate the effect of strengthening. The CFRP is positioned in the middle of the steel, as 106 shown in Figure 3. The total number of specimens tested is thirteen, and their detailed 107 parameters are given in Table 3. As shown in the table, specimens are labeled for easy 108 identification. The first combination of ‘letter-number’ describes angle steel and its buck-109 ling length, the second combination is strengthening length, and the last combination in-110 dicates number of CFRP layers. For instance, A16S10L25 identifies a specimen with 1636 111 mm buckling length, 1000 mm CFRP length, and 25 CFRP layers. Here, buckling length is 112 a total of specimen length L (Figure 3) and thickness of additional steel plates at both ends 113 of specimen (Figure 4). 114

115

Figure 3. Strengthening scheme for the test specimen. 116

117

Figure 4. Test setup. 118

Figure 3. Strengthening scheme for the test specimen.

Table 3. Description of the test specimens.

SpecimenName

Number of CFRPLayers (ply) CFRP Length (mm) Buckling Length

(mm)Angle Steel’s

Slenderness Ratio

Group AA16S00L00 - - 1636 128.4A16S10L18 18 1000 1636 128.4A16S10L25 25 1000 1636 128.4A16S10L30 30 1000 1636 128.4A16S10L35 35 1000 1636 128.4A16S05L18 18 500 1636 128.4A16S05L25 25 500 1636 128.4A16S05L30 30 500 1636 128.4

Group BA12S00L00 - - 1236 97.02A12S05L10 10 500 1236 97.02A12S05L18 18 500 1236 97.02A12S05L25 25 500 1236 97.02A12S05L30 30 500 1236 97.02


Table 2. Carbon fiber. 99

Type Orientation

of Fiber Tensile Strength


(MPa) Thickness

(mm)

BT70-20 0o 2900 230,000 0.056 90o 2900 230,000 0.056

2.2. Description of Specimens 100

The specimens tested in this study consist of two groups of angle steel with different 101 slenderness ratios (λ). Each group has one control specimen that is not strengthened with 102 CFRP. Specimens of Group A (λ = 128.4) are divided into two different CFRP lengths, i.e., 103 1000 mm and 500 mm, while specimens of Group B (λ = 97.02) have CFRP length of 500 104 mm. The number of CFRP layers in strengthened specimens is varied to further investi-105 gate the effect of strengthening. The CFRP is positioned in the middle of the steel, as 106 shown in Figure 3. The total number of specimens tested is thirteen, and their detailed 107 parameters are given in Table 3. As shown in the table, specimens are labeled for easy 108 identification. The first combination of ‘letter-number’ describes angle steel and its buck-109 ling length, the second combination is strengthening length, and the last combination in-110 dicates number of CFRP layers. For instance, A16S10L25 identifies a specimen with 1636 111 mm buckling length, 1000 mm CFRP length, and 25 CFRP layers. Here, buckling length is 112 a total of specimen length L (Figure 3) and thickness of additional steel plates at both ends 113 of specimen (Figure 4). 114

115


117

Figure 4. Test setup. 118 Figure 4. Test setup.

Polymers 2021, 13, 2216 5 of 17

2.3. Specimens Preparation

Both ends of steel angle are welded onto steel plates which are used for the installationof knife edges later on. Welding is carried out so that weak axis of the angle steel coincideswith center line of the steel plate. The unbonded CFRPs in all strengthened specimensare fabricated through a process of vacuum-assisted resin transfer molding (VaRTM). Theprocess begins by wrapping the steel angles in the mid-span position (strengthening zone)using one layer of peel ply. The wrapping process is not preceded by any treatment onthe steel surface. The carbon fiber with a predetermined number of layers is then installedfollowing the peel ply. The fiber, at 0◦, is positioned in the same direction as the longitudinalaxis of the steel angle. The carbon fiber is only assigned to both legs of the angle steel,not to cover the entire cross section (see Figure 5). Afterwards, an additional layer of peelply is used to cover the carbon fiber. Infusion meshes (resin media) are attached after thepeel ply to facilitate resin transfer during the impregnation process. A vacuum bag witha strong gum tape connection is applied at the same time as the installation hose in bothinjection and vacuuming sides. The next process is to impregnate resin by suction. Theresin impregnation process is carried out by tilting the specimens. This aims to decrease therisk of air bubbles trapped inside the mold. After CFRP is molded, specimens are cured forat least one week before being tested. Figure 6 shows the process of specimen preparation.


139


(a) (b) (c)

Figure 6. Specimen preparation: (a) angle steel before strengthening; (b) specimens ready for impregnation process; (c) 141 cross section of specimen during molding process. 142

2.4. Test Setup and Instrumentation 143

All the strengthened and control specimens are tested under compression load using 144 a 2000 kN capacity of Maekawa testing machine, as shown in Figure 4. The specimens are 145 in pinned-end conditions and designed to buckle on their weak axis (v–v) only (Figure 1). 146 To achieve these conditions, a knife edge is installed in both ends of specimens (Figure 4) 147 where it must also coincide with the weak axis. During the test, the applied load and axial 148 displacement are measured by a load cell within the machine. A total of 20 strain gauges 149

Figure 5. Strengthening scheme for the test specimen.

Polymers 2021, 13, 2216 6 of 17


139


(a) (b) (c)

Figure 6. Specimen preparation: (a) angle steel before strengthening; (b) specimens ready for impregnation process; (c) 141 cross section of specimen during molding process. 142

2.4. Test Setup and Instrumentation 143

All the strengthened and control specimens are tested under compression load using 144 a 2000 kN capacity of Maekawa testing machine, as shown in Figure 4. The specimens are 145 in pinned-end conditions and designed to buckle on their weak axis (v–v) only (Figure 1). 146 To achieve these conditions, a knife edge is installed in both ends of specimens (Figure 4) 147 where it must also coincide with the weak axis. During the test, the applied load and axial 148 displacement are measured by a load cell within the machine. A total of 20 strain gauges 149

Figure 6. Specimen preparation: (a) angle steel before strengthening; (b) specimens ready for impregnation process; (c) crosssection of specimen during molding process.

2.4. Test Setup and Instrumentation

All the strengthened and control specimens are tested under compression load usinga 2000 kN capacity of Maekawa testing machine, as shown in Figure 4. The specimens arein pinned-end conditions and designed to buckle on their weak axis (v–v) only (Figure 1).To achieve these conditions, a knife edge is installed in both ends of specimens (Figure 4)where it must also coincide with the weak axis. During the test, the applied load and axialdisplacement are measured by a load cell within the machine. A total of 20 strain gaugesare installed, scattered along the strengthened specimen to measure its longitudinal strainresponse. Meanwhile, only 12 strain gauges are used for control specimen. Out-of-planelateral displacement is measured by displacement transducers. The lateral displacementis measured at three different locations: around the bottom end of CFRP, mid-height ofspecimen, and around the top end of CFRP. Figure 7 shows the detailed position of straingauges and transducers on both control and strengthened specimens.

Polymers 2021, 13, 2216 7 of 17


are installed, scattered along the strengthened specimen to measure its longitudinal strain 150 response. Meanwhile, only 12 strain gauges are used for control specimen. Out-of-plane 151 lateral displacement is measured by displacement transducers. The lateral displacement 152 is measured at three different locations: around the bottom end of CFRP, mid-height of 153 specimen, and around the top end of CFRP. Figure 7 shows the detailed position of strain 154 gauges and transducers on both control and strengthened specimens. 155

(a)

(b)

Figure 7. Position of displacement transducers and strain gauges: (a) overall position; (b) detail in cross section. 156

3. Results and Discussion 157

Table 4 summarizes the results of the investigation in this study, including maximum 158 load capacity, CFRP thickness, and fiber content of the laminates. Also presented in Table 159 4 is the strengthening effect due to the application of unbonded CFRP. The strengthening 160 effect is given as the ratio of the difference of maximum load capacity of specimens (be-161 tween control and strengthened specimens) and maximum load capacity of the appropri-162 ate control specimen (A16S00L00 or A12S00L00). Based on Table 4, it is clear that un-163 bonded CFRP provides considerable positive strengthening effects. The greatest effect 164 was found in specimen A16S10L35 (69.13%), whereas specimen A16S05L18 experienced 165 the least strengthening effect (7.123%). Generally, the positive strengthening effect and its 166 variation in values are attributed to a delay in the overall buckling. These delays are due 167

Figure 7. Position of displacement transducers and strain gauges: (a) overall position; (b) detail in cross section.

3. Results and Discussion

Table 4 summarizes the results of the investigation in this study, including maximumload capacity, CFRP thickness, and fiber content of the laminates. Also presented in Table 4 isthe strengthening effect due to the application of unbonded CFRP. The strengthening effectis given as the ratio of the difference of maximum load capacity of specimens (between con-trol and strengthened specimens) and maximum load capacity of the appropriate controlspecimen (A16S00L00 or A12S00L00). Based on Table 4, it is clear that unbonded CFRP pro-vides considerable positive strengthening effects. The greatest effect was found in specimenA16S10L35 (69.13%), whereas specimen A16S05L18 experienced the least strengtheningeffect (7.123%). Generally, the positive strengthening effect and its variation in values areattributed to a delay in the overall buckling. These delays are due to the different values offlexural stiffness of the unbonded CFRP which exist in each specimen. The difference inCFRP flexural stiffness is defined by the application of different parameters, e.g., numberof CFRP layers and CFRP length. Another factor that exists is difference in fiber content ofCFRP in each specimen, as given in Table 4. The fiber content also contributes to flexuralstiffness of CFRP, although not to the same extent as number of CFRP layers and/or CFRPlength, by creating an elastic modulus of the laminates. Theoretically, the higher the fibercontent, the higher the elastic modulus of CFRP, and thus the greater the stiffness.

Polymers 2021, 13, 2216 8 of 17

Table 4. Summary of test result.

SpecimenName

CFRP Thickness(mm)

Fiber Content(%)

Max. Load(kN)

Strengthening Effect(%)

Group AA16S00L00 - - 77.22 -A16S10L18 3.34 60.35 103.5 34.03A16S10L25 4.58 61.15 113.0 46.34A16S10L30 5.50 61.11 125.9 63.04A16S10L35 6.38 61.47 130.6 69.13A16S05L18 3.39 59.44 82.72 7.123A16S05L25 4.65 60.22 91.10 17.97A16S05L30 5.62 59.84 92.05 19.20

Group BA12S00L00 - - 122.7 -A12S05L10 1.97 56.76 132.2 7.742A12S05L18 3.30 61.17 140.8 14.75A12S05L25 4.78 58.59 154.2 25.67A12S05L30 5.60 59.96 150.2 22.41

The fiber content, as presented in Table 4, is a ratio between total thickness of carbonfiber used (number of layers × 0.112 mm) and thickness of CFRP. The thickness of CFRP(tCF) is calculated by involving the wall thickness of the specimen at the strengthened part(ttot) and the wall thickness of the angle steel (ta = 5.56 mm), namely tCF = 0.5 (ttot − ta). It isevident from Table 4 that CFRP has stable and higher fiber content (>50%, with a coefficientof variation of 2.30%). Therefore, in addition to being highly expected, this confirms theadvantages of the VaRTM process for producing CFRP laminates over the other techniques(e.g., hand-layup) [47].

3.1. Load–Displacement Response

Figure 8a,b shows the load versus axial-displacement response of all the specimensin Group A and B. It can be seen from the figure that the curves provide a steep initialresponse before reaching maximum load. After maximum load, the axial displacementincreases rapidly as the compressive load gradually decreases. Figure 8 graphically provesthat the presence of unbonded CFRP increases the maximum load, but it does not affectthe initial axial stiffness of the strengthened specimens. The slope of the curves remainsunchanged compared to the control specimen. This is reasonable as the CFRP is not bondedto the steel surface, which means that elastic behavior of the structure will not be affected.


(a) (b)

Figure 8. Load–axial displacement response of specimens: (a) Group A; (b) Group B. 200

(a) (b)

Figure 9. Load–lateral displacement response at mid-height of specimens: (a) Group A; (b) Group B. 201

3.2. Effect of Number of CFRP Layers 202

Increasing the number of CFRP layer gives a positive effect in enhancing the buckling 203 strength of the angle steel. This is confirmed by Figure 10 and Table 4, and can also be 204 seen in Figures 8 and 9. The strength increase in the specimens in Group A ranged from 205 7.123% to 69.13%, but in Group B the increases only varied between 7.742% and 25.67%. 206 The relationship between increasing strength and increasing the number of CFRP layers 207 is experienced by all strengthened specimens with appropriate CFRP length in both 208 Group A and B. However, the exception is specimen A12S05L30, which has 30 CFRP lay-209 ers. The strength increase in this specimen (22.45%) is not larger than that of specimen 210 A12S05L25 (25.71%), which only has 25 CFRP layers. The reason for this is most likely the 211 higher imperfection of specimen A12S05L30. The imperfection is indeed not measured in 212 this study, but it is reflected in the load–lateral displacement curve of this specimen (see 213 Figure 9a). The curve is less vertical (before reaching ultimate load) compared to that of 214 specimen A12S05L25. 215

0

50

100

150

0 2 4 6 8

A16S00L00A16S05L18A16S05L25A16S05L30A16S10L18A16S10L25A16S10L30A16S10L35

Com

pres

sive

load

, P[k

N]

Axial displacement (mm)

0

60

120

180

0 2 4 6 8

A12S00L00A12S05L10A12S05L18A12S05L25A12S05L30

Com

pres

sive

load

, P[k

N]


0

50

100

150

-40 -20 0 20 40

A16S00L00A16S05L18A16S05L25A16S05L30A16S10L18A16S10L25A16S10L30A16S10L35C

ompr

essi

ve lo

ad, P

[kN

]

Lateral displacement (mm)

0

60

120

180

0 10 20 30 40


Com

pres

sive

load

, P[k

N]


Figure 8. Load–axial displacement response of specimens: (a) Group A; (b) Group B.

Polymers 2021, 13, 2216 9 of 17

The load–lateral displacement responses at mid-height of the specimens are providedin Figure 9a,b. The curves show a steep response before overall buckling occurs. Afterward,lateral displacement increases rapidly. The displacement curve of A16S10L35 shows anegative value, indicating that this specimen buckles in the opposite direction to the others.It is clear from Figure 9 that unbonded CFRP reduces the lateral deflection.


(a) (b)

Figure 8. Load–axial displacement response of specimens: (a) Group A; (b) Group B. 200

(a) (b)

Figure 9. Load–lateral displacement response at mid-height of specimens: (a) Group A; (b) Group B. 201

3.2. Effect of Number of CFRP Layers 202

Increasing the number of CFRP layer gives a positive effect in enhancing the buckling 203 strength of the angle steel. This is confirmed by Figure 10 and Table 4, and can also be 204 seen in Figures 8 and 9. The strength increase in the specimens in Group A ranged from 205 7.123% to 69.13%, but in Group B the increases only varied between 7.742% and 25.67%. 206 The relationship between increasing strength and increasing the number of CFRP layers 207 is experienced by all strengthened specimens with appropriate CFRP length in both 208 Group A and B. However, the exception is specimen A12S05L30, which has 30 CFRP lay-209 ers. The strength increase in this specimen (22.45%) is not larger than that of specimen 210 A12S05L25 (25.71%), which only has 25 CFRP layers. The reason for this is most likely the 211 higher imperfection of specimen A12S05L30. The imperfection is indeed not measured in 212 this study, but it is reflected in the load–lateral displacement curve of this specimen (see 213 Figure 9a). The curve is less vertical (before reaching ultimate load) compared to that of 214 specimen A12S05L25. 215

0

50

100

150

0 2 4 6 8

A16S00L00A16S05L18A16S05L25A16S05L30A16S10L18A16S10L25A16S10L30A16S10L35

Com

pres

sive

load

, P[k

N]


0

60

120

180

0 2 4 6 8


Com

pres

sive

load

, P[k

N]


0

50

100

150

-40 -20 0 20 40

A16S00L00A16S05L18A16S05L25A16S05L30A16S10L18A16S10L25A16S10L30A16S10L35C

ompr

essi

ve lo

ad, P

[kN

]


0

60

120

180

0 10 20 30 40


Com

pres

sive

load

, P[k

N]


Figure 9. Load–lateral displacement response at mid-height of specimens: (a) Group A; (b) Group B.

3.2. Effect of Number of CFRP Layers

Increasing the number of CFRP layer gives a positive effect in enhancing the bucklingstrength of the angle steel. This is confirmed by Figure 10 and Table 4, and can also beseen in Figures 8 and 9. The strength increase in the specimens in Group A ranged from7.123% to 69.13%, but in Group B the increases only varied between 7.742% and 25.67%.The relationship between increasing strength and increasing the number of CFRP layers isexperienced by all strengthened specimens with appropriate CFRP length in both GroupA and B. However, the exception is specimen A12S05L30, which has 30 CFRP layers. Thestrength increase in this specimen (22.45%) is not larger than that of specimen A12S05L25(25.71%), which only has 25 CFRP layers. The reason for this is most likely the higherimperfection of specimen A12S05L30. The imperfection is indeed not measured in this study,but it is reflected in the load–lateral displacement curve of this specimen (see Figure 9a).The curve is less vertical (before reaching ultimate load) compared to that of specimenA12S05L25.


(a) (b)

Figure 10. Effect of several parameters: (a) effect of CFRP length (λ = 128.4); (b) effect of angle steel’s slenderness ratio 216 (CFRP length = 500mm). 217

3.3. Effect of CFRP Length 218

The effect of CFRP length is well displayed through comparison of specimens in 219 Group A (see Table 4) where two types of CFRP length are investigated, 500 mm and 1000 220 mm. For easy visualization, the comparison is also presented in Figure 10a. It is clear from 221 Figure 10a and Table 4 that strength increase is higher for specimens with longer CFRP. 222 The difference in strength increase between specimen A16S10L18 (34.03%) and specimen 223 A16S05L18 (7.12%), both with 18 layers of CFRP, is 34.04% − 7.12% = 26.92%. This value is 224 almost unchanged with the difference in strength increase in other specimens (A16S10L25 225 and A16S05L25), where the number of CFRP layers is increased to 25 (46.28% − 17.97% = 226 28.31%). However, for specimens with more than 25 layers and 1000 mm of CFRP length 227 (e.g., A16S10L30) there tends to be a greater strength increase compared to specimens with 228 500 mm CFRP length (e.g., A16S05L30). 229

3.4. Effect of Different Slenderness of Steel 230

Figure 10b shows the effect of different angle steel’s slenderness ratio (λ) on the 231 strength increase in the strengthened specimens. A detailed percentage of the increases is 232 presented in Table 4. Figure 10b and Table 4 show that a higher strength increase belongs 233 to specimens with a smaller angle steel’s slenderness ratio. This condition is clearly 234 demonstrated by all comparable specimens (same as number of CFRP layers) between 235 Group A and B having CFRP length of 500 mm. Compared to a specimen with a larger λ 236 (A16S05L18, 18 CFRP layers), a higher strength increase occurs in a specimen with a 237 smaller λ (A12S05L10), even though fewer layers of CFRP are used (10 layers). 238

3.5. Failure Modes 239

All control specimens undergo failure due to overall buckling of pinned-end col-240 umns, as shown in Figures 11a and 12a. Plastic hinge is confirmed at mid-height of the 241 specimens. For the unbonded CFRP strengthened angle steel, three different failure 242 modes can be observed. The first mode of failure is that the plastic hinge is located around 243 the edge of the CFRP (outside CFRP strengthening zone), not at mid-height of the speci-244 mens, as shown in Figure 11c,d,f–h) and Figure 12c–e. The buckling curvatures are similar 245 to each other, toward the inner side of the angle steel. In this case, the CFRP laminates can 246 perform their duty as a stiffener without suffering damage. In the second mode, the spec-247 imen fails when the plastic hinge lies within the strengthening zone, as shown in Figures 248 11b and 12b. The direction of the curvature remains the same as the former failure mode, 249 i.e., to the inner side of the angle steel. This mode of failure occurs when the number of 250 CFRP layers is lower (A16S10L18 and A12S05L10), meaning that the CFRP does not have 251

0

30

60

90

10 20 30 40

CFRP 0.5 mCFRP 1.0 m

Stre

ngth

incr

ease

(%)

CFRP layers (mm)

0

10

20

30

5 15 25 35

λ = 128.4λ = 97.0

Stre

ngth

inc

reas

e (%

)

CFRP layers (mm)

Figure 10. Effect of several parameters: (a) effect of CFRP length (λ = 128.4); (b) effect of angle steel’s slenderness ratio(CFRP length = 500 mm).

Polymers 2021, 13, 2216 10 of 17

3.3. Effect of CFRP Length

The effect of CFRP length is well displayed through comparison of specimens inGroup A (see Table 4) where two types of CFRP length are investigated, 500 mm and1000 mm. For easy visualization, the comparison is also presented in Figure 10a. It is clearfrom Figure 10a and Table 4 that strength increase is higher for specimens with longerCFRP. The difference in strength increase between specimen A16S10L18 (34.03%) andspecimen A16S05L18 (7.12%), both with 18 layers of CFRP, is 34.04% − 7.12% = 26.92%.This value is almost unchanged with the difference in strength increase in other spec-imens (A16S10L25 and A16S05L25), where the number of CFRP layers is increased to25 (46.28% − 17.97% = 28.31%). However, for specimens with more than 25 layers and1000 mm of CFRP length (e.g., A16S10L30) there tends to be a greater strength increasecompared to specimens with 500 mm CFRP length (e.g., A16S05L30).

3.4. Effect of Different Slenderness of Steel

Figure 10b shows the effect of different angle steel’s slenderness ratio (λ) on thestrength increase in the strengthened specimens. A detailed percentage of the increasesis presented in Table 4. Figure 10b and Table 4 show that a higher strength increasebelongs to specimens with a smaller angle steel’s slenderness ratio. This condition is clearlydemonstrated by all comparable specimens (same as number of CFRP layers) betweenGroup A and B having CFRP length of 500 mm. Compared to a specimen with a largerλ (A16S05L18, 18 CFRP layers), a higher strength increase occurs in a specimen with asmaller λ (A12S05L10), even though fewer layers of CFRP are used (10 layers).

3.5. Failure Modes

All control specimens undergo failure due to overall buckling of pinned-end columns, asshown in Figures 11a and 12a. Plastic hinge is confirmed at mid-height of the specimens. Forthe unbonded CFRP strengthened angle steel, three different failure modes can be observed.The first mode of failure is that the plastic hinge is located around the edge of the CFRP (outsideCFRP strengthening zone), not at mid-height of the specimens, as shown in Figure 11c,d,f–h)and Figure 12c–e. The buckling curvatures are similar to each other, toward the inner side ofthe angle steel. In this case, the CFRP laminates can perform their duty as a stiffener withoutsuffering damage. In the second mode, the specimen fails when the plastic hinge lies withinthe strengthening zone, as shown in Figures 11b and 12b. The direction of the curvatureremains the same as the former failure mode, i.e., to the inner side of the angle steel.This mode of failure occurs when the number of CFRP layers is lower (A16S10L18 andA12S05L10), meaning that the CFRP does not have the appropriate stiffness to overcomeoverall buckling of the angle steel. Thus, the CFRP at the outer side of the angle steel peelsoff and the CFRP at the inner side of the angle steel is under compression, which can bedamaged when excessive load is applied (see Figures 11b and 12b).

Polymers 2021, 13, 2216 11 of 17


the appropriate stiffness to overcome overall buckling of the angle steel. Thus, the CFRP 252 at the outer side of the angle steel peels off and the CFRP at the inner side of the angle 253 steel is under compression, which can be damaged when excessive load is applied (see 254 Figures 11b and 12b). 255

The last failure mode occurs when the plastic hinge is within the CFRP strengthening 256 zone but the buckling curvature occurs towards the outer side of the angle steel, as expe-257 rienced by specimen A16S10L35 (Figure 11e). This type of failure mode tends to occur as 258 the number of CFRP layers increases. This trend can be easily seen in specimens belonging 259 to Group A which have a CFRP length of 1000 mm (Figure 9a). It is clear from Figure 9a 260 that the lateral deflection curve (before reaching maximum load) becomes steeper as the 261 number of CFRP layers increases from 18 (A16S10L18) to 25 (A16S10L25). It is only within 262 the extreme condition of the use of 30 CFRP layers that the specimen curve (A16S10L30) 263 becomes perfectly vertical. Beyond this, a larger number of layers makes the curve re-264 sponse change to a negative value (specimen A16S10L35). This means that the buckling 265 curvature occurs in the opposite direction (Figure 11e). According to the authors, the rea-266 son behind this failure mode is that, in addition to the increase in the CFRP stiffness due 267 to the increase in the number of CFRP layers, loading eccentricity (in terms of the centroid 268 of the CFRP’s cross section) also increases. The larger number of CFRP layers leads to a 269 larger change in the centroid of the CFRP cross section, but the position of the applied 270 load remains unchanged on the weak axis of the angle steel cross section. 271

(a) A16S00L00 (b) A16S10L18 (c) A16S10L25 (d) A16S10L30


(e) A16S10L35 (f) A16S05L18 (g) A16S05L25 (h) A16S05L30

Figure 11. Failure mode of specimens: (a) A16S00L00; (b) A16S10L18; (c) A16S10L25; (d) A16S10L30; (e) A16S10L35; (f) 272 A16S05L18; (g) A16S05L25; (h) A16S15L30. 273

(a) A12S00L00 (b) A12S05L10 (c) A12S05L18 (d) A12S05L25 (e) A12S05L30

Figure 12. Failure mode of specimens: (a) A12S00L00; (b) A12S05L10; (c) A12S05L18; (d) A12S05L25; (e) A12S05L30. 274

3.6. Strain Response 275

Figures 13 and 14 show the load versus longitudinal strain response of all specimens 276 tested in this study. It is confirmed by Figures 13a and 14a (control specimens) that only 277 compressive strain at mid-height of the specimens continues to develop and reaches the 278 angle steel yield strain at about 0.16%. This finding corresponds to the failure modes pre-279 viously discussed. For strengthened specimens, the strain on CFRP also increases, but the 280 rate of increase at the end of CFRP is smaller than that at the middle, and it only occurs 281

Figure 11. Failure mode of specimens: (a) A16S00L00; (b) A16S10L18; (c) A16S10L25; (d) A16S10L30; (e) A16S10L35;(f) A16S05L18; (g) A16S05L25; (h) A16S15L30.

Polymers 2021, 13, 2216 12 of 17


(e) A16S10L35 (f) A16S05L18 (g) A16S05L25 (h) A16S05L30


(a) A12S00L00 (b) A12S05L10 (c) A12S05L18 (d) A12S05L25 (e) A12S05L30

Figure 12. Failure mode of specimens: (a) A12S00L00; (b) A12S05L10; (c) A12S05L18; (d) A12S05L25; (e) A12S05L30. 274

3.6. Strain Response 275

Figures 13 and 14 show the load versus longitudinal strain response of all specimens 276 tested in this study. It is confirmed by Figures 13a and 14a (control specimens) that only 277 compressive strain at mid-height of the specimens continues to develop and reaches the 278 angle steel yield strain at about 0.16%. This finding corresponds to the failure modes pre-279 viously discussed. For strengthened specimens, the strain on CFRP also increases, but the 280 rate of increase at the end of CFRP is smaller than that at the middle, and it only occurs 281

Figure 12. Failure mode of specimens: (a) A12S00L00; (b) A12S05L10; (c) A12S05L18; (d) A12S05L25; (e) A12S05L30.

The last failure mode occurs when the plastic hinge is within the CFRP strengtheningzone but the buckling curvature occurs towards the outer side of the angle steel, as experi-enced by specimen A16S10L35 (Figure 11e). This type of failure mode tends to occur as thenumber of CFRP layers increases. This trend can be easily seen in specimens belongingto Group A which have a CFRP length of 1000 mm (Figure 9a). It is clear from Figure 9athat the lateral deflection curve (before reaching maximum load) becomes steeper as thenumber of CFRP layers increases from 18 (A16S10L18) to 25 (A16S10L25). It is only withinthe extreme condition of the use of 30 CFRP layers that the specimen curve (A16S10L30)becomes perfectly vertical. Beyond this, a larger number of layers makes the curve responsechange to a negative value (specimen A16S10L35). This means that the buckling curvatureoccurs in the opposite direction (Figure 11e). According to the authors, the reason behindthis failure mode is that, in addition to the increase in the CFRP stiffness due to the increasein the number of CFRP layers, loading eccentricity (in terms of the centroid of the CFRP’scross section) also increases. The larger number of CFRP layers leads to a larger changein the centroid of the CFRP cross section, but the position of the applied load remainsunchanged on the weak axis of the angle steel cross section.

3.6. Strain Response

Figures 13 and 14 show the load versus longitudinal strain response of all specimenstested in this study. It is confirmed by Figures 13a and 14a (control specimens) that onlycompressive strain at mid-height of the specimens continues to develop and reaches theangle steel yield strain at about 0.16%. This finding corresponds to the failure modespreviously discussed. For strengthened specimens, the strain on CFRP also increases, butthe rate of increase at the end of CFRP is smaller than that at the middle, and it only occursup to a certain load (the load where strain stops developing or returns to zero), e.g., inspecimen A12S05L18 (see Figure 14c for easy visualization, the load is around 60 kN). Inother specimens, the strains in CFRP may not develop at all or develop only up to a verysmall load, for example, specimens A12S05L25 or A12S05L30 (Figure 14d,e). These twospecimens have higher CFRP stiffness as a result of many layers of CFRP and shorter CFRPlength. In some other specimens (mainly in Group A), the load at which the strain stopsdeveloping may not be easily identified. The increase in compressive strains on CFRP inall of the strengthened specimens indicates the existence of a minor bond of CFRP to steel.This is highly possible because during the VaRTM process, the resin will also impregnateinto the peel ply.

Polymers 2021, 13, 2216 13 of 17


up to a certain load (the load where strain stops developing or returns to zero), e.g., in 282 specimen A12S05L18 (see Figure 14c for easy visualization, the load is around 60 kN). In 283 other specimens, the strains in CFRP may not develop at all or develop only up to a very 284 small load, for example, specimens A12S05L25 or A12S05L30 (Figure 14d,e). These two 285 specimens have higher CFRP stiffness as a result of many layers of CFRP and shorter 286 CFRP length. In some other specimens (mainly in Group A), the load at which the strain 287 stops developing may not be easily identified. The increase in compressive strains on 288 CFRP in all of the strengthened specimens indicates the existence of a minor bond of CFRP 289 to steel. This is highly possible because during the VaRTM process, the resin will also 290 impregnate into the peel ply. 291

(a) (b)

(c) (d)

(e) (f)

0

30

60

90

120

-2500 -2000 -1500 -1000 -500 0 500 1000 1500

Str. 1Str. 2Str. 3Str. 4Str. 5Str. 6Str. 7Str. 8Str. 9Str. 10Str. 11Str. 12

Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500

Str. 1Str. 2Str. 3Str. 4Str. 5Str. 6Str. 7Str. 8Str. 9Str. 10Str. 11Str. 12Str. 13Str. 14Str. 15Str. 16Str. 17Str. 18Str. 19Str. 20

Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2000 -1500 -1000 -500 0 500 1000 1500 2000


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

30

60

90

120

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)Polymers 2021, 13, x FOR PEER REVIEW 14 of 17

(g) (h)


(a)

(b) (c)

(d) (e)

0

30

60

90

120

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

30

60

90

120

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500

Str. 1

Str. 2

Str. 3

Str. 4

Str. 5

Str. 6

Str. 7

Str. 8

Str. 9Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

Figure 13. Failure mode of specimens: (a) A16S00L00; (b) A16S10L18; (c) A16S10L25; (d) A16S10L30; (e) A16S10L35;(f) A16S05L18; (g) A16S05L25; (h) A16S05L30.

Polymers 2021, 13, 2216 14 of 17


(g) (h)


(a)

(b) (c)

(d) (e)

0

30

60

90

120

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

30

60

90

120

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500

Str. 1

Str. 2

Str. 3

Str. 4

Str. 5

Str. 6

Str. 7

Str. 8

Str. 9Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

0

40

80

120

160

-2500 -2000 -1500 -1000 -500 0 500 1000 1500


Com

pres

sion

load

, P (k

N)

Strain, ε (x10-6)

Figure 14. Failure mode of specimens: (a) A12S00L00; (b) A12S05L10; (c) A12S05L18; (d) A12S05L25; (e) A12S05L30.

4. Conclusions

This paper investigates the use of unbonded CFRP laminates, as an alternative toadhesive-bonding CFRP, for strengthening angle steel members against buckling. The CFRPis fabricated directly onto the surface of the steel members, using the VaRTM technique,and is adjusted to cover both legs of the angel steel only. Based on the experimentalinvestigation, the following findings can be drawn:

1. Unbonded CFRP strengthening can successfully increase the buckling capacity of theangle steel. The strengthening effect varies, and the maximum load increase is 69.13%;

2. The buckling capacity of angle steel increases with increasing number of CFRP layersand CFRP length. In addition, to strengthen without changing CFRP length, theincrease in buckling capacity is greater for specimens with shorter angle steel (smallerangle steel’s slenderness ratio);

Polymers 2021, 13, 2216 15 of 17

3. Based on the position of the plastic hinge, failure modes of the unbonded CFRPstrengthened specimens can be divided into two: plastic hinge at around the end ofCFRP (outside CFRP strengthening zone), and plastic hinge within the CFRP strength-ening zone. Both failure modes are affected by flexural stiffness of the CFRP laminates;

4. The VaRTM process is the best alternative for making CFRP. The CFRP fabricatedthrough the VaRTM process has a higher fiber volume content of around 60%.

It should be noted that angle steel is often used as a lateral resisting element inbuilding structures, so it is very susceptible to buckling failure and needs to be strengthened.Therefore, in conjunction with this research program, the authors also demonstrate thepotential of unbonded CFRP strengthening for real-world applications.

Author Contributions: Conceptualization, Y.M.; methodology, F.S.Y.; validation, Y.M.; investigation,F.S.Y., P.V.N. and D.N.; writing—original draft preparation, F.S.Y.; writing—review and editing, F.S.Y.and Y.M.; supervision, Y.M.; project administration, Y.M. All authors have read and agreed to thepublished version of the manuscript.

Funding: This research was funded by JSPS KAKENHI, grant number 20K04788.

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.

Data Availability Statement: The data required to reproduce these findings cannot be shared at thistime, as they also form part of an ongoing study.

Acknowledgments: The authors wish to acknowledge all members of Structural Engineering Labo-ratory, Department of Architecture and Civil Engineering, Toyohashi University of Technology, forsupporting the experimental work.

Conflicts of Interest: The authors declare no conflict of interest.

References1. Trung, K.L.; Lee, K.; Lee, J.; Lee, D.H.; Woo, S. Experimental study of RC beam–column joints strengthened using CFRP

composites. Compos. Part B 2010, 41, 76–85. [CrossRef]2. Salama, A.S.D.; Hawileh, R.A.; Abdalla, J.A. Performance of externally strengthened RC beams with side-bonded CFRP sheets.

Compos. Struct. 2019, 212, 281–290. [CrossRef]3. Rahai, A.; Akbarpour, H. Experimental investigation on rectangular RC columns strengthened with CFRP composites under axial

load and biaxial bending. Compos. Struct. 2014, 108, 538–546. [CrossRef]4. Farahbod, F.; Mostofinejad, D. Experimental study of moment redistribution in RC frames strengthened with CFRP sheets.

Compos. Struct. 2011, 93, 1168–1177. [CrossRef]5. Chen, W.; Shou, W.; Qiao, Z.; Cui, S. Seismic performance of non-ductile RC frames strengthened with CFRP. Compos. Struct.

2019, 221, 110870. [CrossRef]6. Siwowski, T.W.; Siwowska, P. Experimental study on CFRP-strengthened steel beams. Compos. Part B 2018, 149, 12–21. [CrossRef]7. Weerasinghe, K.A.B.; Gamage, J.C.P.H.; Fawzia, S.; Thambiratnam, D.P. Experimental investigation on flexural behaviour of

vertically-curved circular-hollow steel sections strengthened with externally bonded carbon fibre reinforced polymer. Eng. Struct.2021, 236, 112040. [CrossRef]

8. Doroudi, Y.; Fernando, D.; Hosseini, A.; Ghafoori, E. Behavior of cracked steel plates strengthened with adhesively bonded CFRPlaminates under fatigue Loading: Experimental and analytical study. Compos. Struct. 2021, 266, 113816. [CrossRef]

9. Chataigner, S.; Benzarti, K.; Foret, G.; Caron, J.F.; Gemignani, G.; Brugiolo, M.; Calderon, I.; Pinero, I.; Birtel, V.; Lehmann, F.Design and testing of an adhesively bonded CFRP strengthening system for steel structures. Eng. Struct. 2018, 177, 556–565.[CrossRef]

10. Chataigner, S.; Wahbeh, M.; Sanchez, D.G.; Benzarti, K. Fatigue strengthening of steel bridges with adhesively bonded CFRPlaminates: Case study. J. Compos. Constr. 2020, 24, 05020002. [CrossRef]

11. Imran, M.; Mahendran, M.; Keerthan, P. Experimental and numerical investigations of CFRP strengthened short SHS steelcolumns. Eng. Struct. 2018, 175, 879–894. [CrossRef]

12. Karimian, M.; Narmashiri, K.; Shahraki, M.; Yousefi, O. Structural behaviors of deficient steel CHS short columns strengthenedusing CFRP. J. Constr. Steel Res. 2017, 138, 555–564. [CrossRef]

13. Ghaemdoust, M.R.; Narmashiri, K.; Yousefi, O. Structural behaviors of deficient steel SHS short columns strengthened usingCFRP. Constr. Build. Mat. 2016, 126, 1002–1011. [CrossRef]

14. Linghoff, D.; Al-Emrani, M.; Kliger, R. Performance of steel beams strengthened with CFRP laminate—Part 1: Laboratory tests.Compos. Part B 2010, 41, 509–515. [CrossRef]

http://doi.org/10.1016/j.compositesb.2009.06.005

http://doi.org/10.1016/j.compstruct.2019.01.045





http://doi.org/10.1016/j.engstruct.2021.112040

http://doi.org/10.1016/j.compstruct.2021.113816

http://doi.org/10.1016/j.engstruct.2018.10.004

http://doi.org/10.1061/(ASCE)CC.1943-5614.0001014


http://doi.org/10.1016/j.jcsr.2017.07.021

http://doi.org/10.1016/j.conbuildmat.2016.09.099


Polymers 2021, 13, 2216 16 of 17

15. Kadhim, M.M.A.; Wu, Z.; Cunningham, L.S. Experimental and numerical investigation of CFRP-strengthened steel beams underimpact load. J. Struct. Eng. 2019, 145, 04019004. [CrossRef]

16. Tafsirojjaman, T.; Fawzia, S.; Thambiratnam, D.; Zhao, X.L. Behaviour of CFRP strengthened CHS members under monotonicand cyclic loading. Compos. Struct. 2019, 220, 592–601. [CrossRef]

17. Chen, T.; Qi, M.; Gu, X.L.; Yu, Q.Q. Flexural strength of carbon fiber reinforced polymer repaired cracked rectangular hollowsection steel beams. Int. J. Polym. Sci. 2015, 2015, 204861. [CrossRef]

18. Altaee, M.J.; Cunningham, L.S.; Gillie, M. Experimental investigation of CFRP strengthened steel beams with web openings. J.Constr. Steel Res. 2017, 138, 750–760. [CrossRef]

19. Wu, C.; He, L.; Ghafoori, E.; Zhao, X.L. Torsional strengthening of steel circular hollow sections (CHS) using CFRP composites.Eng. Struct. 2018, 171, 806–816. [CrossRef]

20. Chahkand, N.A.; Jumaat, M.Z.; Sulong, N.H.R.; Zhao, X.L.; Mohammadizadeh, M.R. Experimental and theoretical investigationon torsional behaviour of CFRP strengthened square hollow steel section. Thin-Walled Struct. 2013, 68, 135–140. [CrossRef]

21. Hatami, F.; Ghamari, A.; Rahai, A. Investigating the properties of steel shear walls reinforced with carbon fiber polymers (CFRP).J. Constr. Steel Res. 2012, 70, 36–42. [CrossRef]

22. Poul, M.K.; Alahi, F.N.; Zhao, X.L. Experimental testing on CFRP strengthened thin steel plates under shear loading. Thin-WalledStruct. 2016, 109, 217–226. [CrossRef]

23. Sahebjam, A.; Showkati, H. Experimental study on the cyclic behavior of perforated CFRP strengthened steel shear walls. Arch.Civ. Mech. Eng. 2016, 16, 365–379. [CrossRef]

24. Lu, Y.; Li, W.; Li, S.; Li, X.; Zhu, T. Study of the tensile properties of CFRP strengthened steel plates. Polymers 2015, 7, 2595–2610.[CrossRef]

25. Wang, Y.; Zheng, Y.; Li, J.; Zhang, L.; Deng, J. Experimental study on tensile behaviour of steel plates with centre hole strengthenedby CFRP plates under marine environment. Int. J. Adhes. Adhes. 2018, 84, 18–26. [CrossRef]

26. Wang, Z.Y.; Zhang, N.; Wang, Q.Y. Tensile behaviour of open-hole and bolted steel plates reinforced by CFRP strips. Compos. PartB 2016, 100, 101–113. [CrossRef]

27. Colombi, P.; Poggi, C. Strengthening of tensile steel members and bolted joints using adhesively bonded CFRP plates. Constr.Build. Mat. 2006, 20, 22–33. [CrossRef]

28. Yiyan, L.; Haojun, Z.; Suli, L. Experimental study on tensile properties of steel plate bonded by CFRP. J. Wuhan Univ. Technol. Mat.Sci. Ed. 2008, 23, 727–732.

29. Shaat, A.; Fam, A.Z. Slender steel columns strengthened using high-modulus CFRP plates for buckling control. J. Compos. Constr.2009, 13, 2–12. [CrossRef]

30. Ahmed, E.Y.S.; Shaat, A.A.; Abdallah, E.A. CFRP-strengthened HSS columns subject to eccentric loading. J. Compos. Constr. 2018,22, 04018025. [CrossRef]

31. Ritchie, A.; Fam, A.; MacDougall, C. Strengthening long steel columns of S-sections against global buckling around weak axisusing CFRP plates of various moduli. J. Compos. Constr. 2015, 19, 04014074. [CrossRef]

32. Yu, T.; Fernando, D.; Teng, J.G.; Zhao, X.L. Experimental study on CFRP-to-steel bonded interfaces. Compos. Part B 2012, 43,2279–2289. [CrossRef]

33. Teng, J.G.; Yu, T.; Fernando, D. Strengthening of steel structures with fiber-reinforced polymer composites. J. Constr. Steel Res.2012, 78, 131–143. [CrossRef]

34. Katrizadeh, E.; Narmashiri, K. Experimental study on failure modes of MF-CFRP strengthened steel beams. J. Constr. Steel Res.2019, 158, 120–129. [CrossRef]

35. Narmashiri, K.; Jumaat, M.Z.; Sulong, N.H.R. Strengthening of steel I-beams using CFRP strips: An investigation on CFRP bondlength. Adv. Struct. Eng. 2012, 15, 2191–2204. [CrossRef]

36. Zeng, J.J.; Gao, W.Y.; Liu, F. Interfacial behavior and debonding failures of full-scale CFRP-strengthened H-section steel beams.Compos. Struct. 2018, 201, 540–552. [CrossRef]

37. Li, C.; Ke, L.; He, J.; Chen, Z.; Jiao, Y. Effects of mechanical properties of adhesive and CFRP on the bond behavior in CFRP-strengthened steel structures. Compos. Struct. 2019, 211, 163–174. [CrossRef]

38. Ritchie, A.; MacDougall, C.; Fam, A. Enhancing buckling capacity of slender s-section steel columns around strong axis usingbonded carbon fibre plates. J. Reinf. Plast. Compos. 2015, 34, 771–781. [CrossRef]

39. Teng, J.G.; Fernando, D.; Yu, T.; Zhao, X.L. Treatment of steel surfaces for effective adhesive bonding. In Proceedings of the 5thInternational Conference on FRP Composites in Civil Engineering (CICE), Beijing, China, 27–29 September 2010.

40. Yoresta, F.S.; Nhut, P.V.; Matsumoto, Y. Finite element analysis of axial compression steel members strengthened with unbondedCFRP laminates. Materials 2020, 13, 3540. [CrossRef] [PubMed]

41. Nguyen, T.C.; Bai, Y.; Zhao, X.L.; Al-Mahaidi, R. Mechanical characterization of steel/CFRP double strap joints at elevatedtemperatures. Compos. Mat. 2011, 93, 1604–1612. [CrossRef]

42. Li, S.; Zhu, T.; Lu, Y.; Li, X. Effect of temperature variation on bond characteristics between CFRP and steel plate. Int. J. Polym. Sci.2016, 2016, 5674572. [CrossRef]

43. Al-Shawaf, A.; Zhao, X.L. Adhesive rheology impact on wet lay-up CFRP/steel joints’ behaviour under infrastructural zubzeroexposures. Compos. Part B 2013, 47, 207–219. [CrossRef]

http://doi.org/10.1061/(ASCE)ST.1943-541X.0002288


http://doi.org/10.1155/2015/204861



http://doi.org/10.1016/j.tws.2013.03.008


http://doi.org/10.1016/j.tws.2016.09.026

http://doi.org/10.1016/j.acme.2016.01.009

http://doi.org/10.3390/polym7121537

http://doi.org/10.1016/j.ijadhadh.2018.01.017


http://doi.org/10.1016/j.conbuildmat.2005.06.042

http://doi.org/10.1061/(ASCE)1090-0268(2009)13:1(2)






http://doi.org/10.1260/1369-4332.15.12.2191



http://doi.org/10.1177/0731684415580331

http://doi.org/10.3390/ma13163540

http://www.ncbi.nlm.nih.gov/pubmed/32796595


http://doi.org/10.1155/2016/5674572


Polymers 2021, 13, 2216 17 of 17

44. Abanilla, M.A.; Li, Y.; Karbhari, V.M. Durability characterization of wet layup graphite/epoxy composites used in externalstrengthening. Compos. Part B 2006, 37, 200–212. [CrossRef]

45. Heshmati, M.; Haghani, R.; Al-Emrani, M. Durability of bonded FRP-to-steel joints: Effects of moisture, de-icing salt solution,temperature and FRP type. Compos. Part B 2017, 119, 153–167. [CrossRef]

46. Yoresta, F.S.; Maruta, R.; Mieda, G.; Matsumoto, Y. Unbonded CFRP strengthening method for buckling control of steel members.Constr. Build. Mat. 2020, 241, 118050. [CrossRef]

47. Uddin, N.; Vaidya, U.; Shohel, M.; Perez, J.C.S. Cost-effective bridge-girder strengthening using vacuum-assisted resin transfermolding (VARTM). Adv. Compos. Mat. 2004, 13, 255–281. [CrossRef]



http://doi.org/10.1016/j.conbuildmat.2020.118050

http://doi.org/10.1163/1568551042580163

1,2,* , phan viet nhut 1,3, daiki nakamoto 1

Documents