Fourth Asia-Pacific Conference on FRP in Structures (APFIS 2013) 11-13 December 2013, Melbourne, Australia

© 2013 International Institute for FRP in Construction

REPAIR WITH CARBON FIBRE REINFORCED POLYMER FOR STEEL PIPE HAVING PARTIAL LOSS OF AREA

T. Matsui 1 and Y. Matsumoto 2 1 Toray Industries Inc., Japan. Email: Takahiro_Matsui@nts.toray.co.jp

2 Toyohashi University of Technology, Japan. ABSTRACT

In this paper, the effects of reinforcement with CFRP for steel pipes having partially cut-out are investigated through axial loading tests. Firstly, the axial compression and tensile tests of steel pipes having partial cut-out are carried out. Secondly, the quantitative behaviour on axial stiffness, strength, yield stress and stress concentration are made clear. Finally, it is confirmed that axial stiffness and yield stress increases while using CFRP even if it is adhesively bonded to steel pipes on the outerside section. It has been suggested that the present reinforcing method provides the reasonable and useful procedure for reinforcing corroded steel pipe members under axial compression. KEYWORDS Steel pipe, Partial loss of area, Carbon fibre reinforced polymer, Repair 1. INTRODUCTION

CFRP has properties of light weight with high strength and high durability. So CFRP has been applied to many civil engineering structures for the purpose of strengthening. Also it has been noticed that the aging degradation and great earthquakes are affecting a lot of structures and the civil structures are seriously being damaged in Japan. For this reason the CFRP is gradually being applied to concrete structures as repair material, strengthening and seismic retrofitting purpose. In the case of steel structure, corrosion is the most important problem on the aging degradation which severely damage the civil engineering structures. In recent years, the studies on CFRPs application to steel structures have also been developing for opened sectional members such as H-shaped beams or L-shaped members. However, it is a remarkable attention that the mechanical performances of steel members are degraded due to corrosion inside the section along with the passage of time. Then the reinforcing of the corroded steel pipes on the outside section is particularly required and CFRP is expected to apply in reinforcing and strengthening procedure. While repairing of steel pipe member, it is needed to recover stiffness and strength. As stiffness, it assumes that the bending deformation occurs by eccentricity because neutral axis changes by loss of area. Thus, proposing the repair method to recover formed bending stiffness and confirms effect. That is, when recovering dominant bending stiffness for buckling strength, it is expected that the repair effect for compression strength occurs for long member indirectly, too. On the other hand, regarding as ultimate tensile strength, bonding strength will be dominant. But it will confirm in the basic bonding test and so on. Based on these, the effect for elastic stiffness and eccentricity in the range of allowable stress is mainly discussed in this study. 2. OUTLINE OF SPECIMEN AND METHOD OF REPAIR 2.1 Steel pipe and CFRP reinforcement

The diameter of the adopted steel pipes is 89.1mm, thickness of pipe is 3.2mmand the steel pipes are galvanized. In the study, corrosion geometry is adopted to be experimentally-modelled loose-hole as shown in Fig. 1. and in the case of compression test, steel pipes for specimen have 5%, 15% or 25% loss of area. Also in the case of tensile test, steel pipes have 15% or 25% loss of area as shown in Fig.1. Carbon fibre (CF) sheets bonded on the external surface of a steel pipe are “Torayca cloth” UM46-40G as axial CF sheet and “Torayca cloth” UT70-20G as circumferential CF sheet. The adopted epoxy resin, E2500 made by Konishi Co.,Ltd., is adopted as adhesive bonding and epoxy primer, E810L made by Konishi Co.,Ltd., is adopted to be surface treatment material for steel. The material properties of CF sheet and CFRP are shown in Table 1, 2 and 3.

2.2 Repair design method

2

In this study, the amount of CFRP for repair is calculated based on the bending stiffness of non-corroded steel pipe. The moment of inertia for the area of steel pipe with partial loss, ISL, is calculated by the following equation. and Fig.2.

( )0

0

14

0220

20

2 sin4

24

R

y

SL

soRyRyRRyyI ⎥⎦

⎤⎢⎣

⎡+−−= − ( )

i

si

R

yi

iii R

yRyRRyy⎥⎦

⎤⎢⎣

⎡+−−− −1

42222 sin

42

4

2

2⎟⎠⎞

⎜⎝⎛ +

××− sisoss

yywt (1)

On the other hand, moment of inertia for the area of CFRP, IFR, is calculated by the following equation. ( )

42 2 2 2 12 Sin

4 4

F

Fo

R

FFR F F

F y

y R yI y R R yR

−⎡ ⎤= − − +⎢ ⎥⎣ ⎦

( )4

2 2 2 2 12 Sin4 4

o

Fi

R

oo o

o y

y R yy R R yR

−⎡ ⎤− − − +⎢ ⎥⎣ ⎦

(2)

Finally, CFRP amount of repair is determined from equation (3). Where, ES and EF are elastic moduli of steel, and longitudinal elastic modulus of CFRP respectively.

ESISL ≦ EFIFR (3)

Table.1 Material properties of steel pipe

Type Tensile strength [MPa]

Yield stress [MPa]

Elastic modulus [GPa]

STK400 Higher than 400

Higher than 235 205

Table.2 Material properties of carbon fiber sheet

Product name

Thickness [mm]

Tensile strength [MPa]

Elastic modulus[GPa]

UM46-40G 0.217 Higher than 2400 440 UT70-20G 0.111 Higher than 3400 245

Table.3 Material properties of CFRP Product name

Volume of fibre fraction [%]

Thickness [mm]

Elastic modulus [GPa]

UM46-40G 24.5 0.884 115

Figure.2 Size of section using method of repair design

2.3 Repair Method Repair method for steel pipe having partial loss of area is

adopted CF sheet bonding method and it is expected that the impregnating resin for CF sheet works as adhesive on steel. The repair procedure is as follows;

(1) sanding the external surface of steel pipe by sand blaster (2) primer is pasted on (3) CF sheets amount according to design are bonded with

impregnating resin In this study, the partial repair in which axial CF sheet is only

bonded at the loss of area is as shown in Fig.3 (a) and the all-around repair which axial CF sheet is bonded around steel pipe is as shown in Fig. 3 (b), are adopted. Also in the any repair method, contribution to stiffness and strength for circumferential CF sheet is

ws ws

ysi ysi yso yso yF yFo

yFi

Ri RoRi Ro

ts RF

(b)All-around repair (a)Partial repair

100

100

89.1

Partial loss of area (loose hole)

14

89.1 89.1

39 61

Partial loss of area(circular hole)

Figure.1 Outline of specimen of steel pipe

(a) Partial repair (b) All-around repair■:Section of steel pipe, ■: Axial CF sheet, □: Circumferential CF sheet

Figure.3 Section of specimens

3

ignored. The designed number of layer for axial CF sheet is shown in Table 4. Also Table 5 shows the axial stiffness and bending stiffness of the specimens and every improving ratio. Then in the case of specimen for 15% loss of area, number of layer calculated by partial repair method is adopted to all-around repair.

Table.4 Number of layer for axial CF sheet [ply]

Ratio of loss area 5% 15% 25% Partial repair 2 4 5

All around repair 1 4 3 3. COMPRESSIVE TEST 3.1 Specimen and test method

The adopted specimens of compressive test are shown in Table 6. Photo 1 shows the experimental setup and Fig. 4 shows the detail of the experiment.

Table.6 The adopted specimen

Name of Specimen

Loss of area

Repair

NG0-C01 N/A N/A NG0-C02

NG15-C01 15%

N/A NG15CL-C01 Partial NG15CA-C01 All around

NG25-C01 25%

N/A NG25CL-C01 Partial NG25CA-C01 All around

3.2 Results and Discussions

Results of compression test and axial stiffness are shown in Fig.5 and Table 7. Then, in the case of NG15CA-C01, un-loading is done after load reaches to 350kN. The elastic axial stiffness shown in Table 7 is calculated from 100kN to 203kN by the experimental results. Also, Table 7 shows improving ratio of axial stiffness compared with the specimen without loss of area. It appears that the axial stiffness is improved by CFRP repair. Axial stiffness in the any specimens is low due to characteristic of compressive test. Axial stiffness of non-repair specimens, NG15-C01 and NG25-C01, are much lower than theoretical value in Table 5. But it is confirmed that CFRP can recover the former axial stiffness. While comparing with improving ratio of theoretical value, specimens repaired by CFRP are almost matches, but NG15CA-C01 only differs largely. Then, it appears that local buckling occurs near loss of area due to yielding by stress concentration and it is supposed that de-bonding the CFRP substrate cannot contribute on CFRP repair effectively. Thus it is supposed that treating the countermeasure to prevent de-bonding will be able to reduce the strength reduction after yielding.

Apart from above, the allowable axial force 203kN can be calculated by using the usual allowable stress 235MPa, of steel in Japan. Therefore it is supposed that de-bonding does not occur until allowable force. Situation for bending deformation affected eccentricity is shown in Fig.6. It shows difference of displacement at open side and opposite side. It appears that eccentricity by aperture is prevented by CFRP repair and there is predominant effect of repair design method based on improving bending stiffness. Specimens after compressive test are shown in Photo.2. It appears that parts around aperture buckles by compression loading, and deforms such as bulging outside in Photo.2 (a) and (b). CFRP deformation is confirmed at the aperture of NG15CL-01 by this influence. In the case of NG25CL-01, CFRP around aperture ruptures by increasing deformation for

Figure.4 Detail of compressive testPhoto.1 Experimental setup

200

10

10

40

180

40

Specimen

Load cell

(CLP-

Table.7 Axial stiffness and its improving ratio of compressive test Name of specimen

Axial stiffness [×103kN]

Improving ratio [%]

NG0-C01 111 0 NG0-C02 110

NG15-C01 73 -36 NG15CL- 107 -3 NG15CA- 109 -1 NG25-C01 65 -41 NG25CL- 108 -2 NG25CA- 121 9

Table.5 Theoretical stiffness of specimen

Ratio of loss area

Axial stiffness[×103kN]

Bending stiffness[kN･mm2]

0% 5% 15% 25% 0% 5% 15% 25%Non-repair 177 168 151 133 164 146 122 112 Partial repair

(Improving ratio) - 176(0%)

179(1%)

177(0%) - 164

(0%) 174(6%)

189(15%)

All-around repair(Improving ratio) - 196

(11%)263

(48%)217

(23%) - 175 (7%)

249(52%)

205(25%)

Load cell

(CLP-200BWS,200tf)

Plate

Plate

Load cell

(CLP-200BWS 200tf)

4

increasing loss ratio and de-bonds. In the case of NG25CA-C01 of all-around repair, crack along with axial direction in CFRP occurs from aperture to the end of CFRP and de-bonds. Average strain and yield strain on inside at the both side of aperture are shown in Fig.7. The load at the yield strain effectively increases by the CFRP repair. However, the load at the yield strain cannot reach the load of normal specimen even if the CFRP repair is applied.

0

100

200

300

400

-4-3-2-10

[kN]

[mm]

NS15-C01NS15CL-C01NS15CA-C01NG15-C01NG15CL-C01NG15CA-C01

0

100

200

300

400

-4-3-2-10

[kN]

[mm]

NG25-C01NG25CL-C01NG25CA-C01

(a) 15% loss of area (b) 25% loss of area

Figure.5 Load - displacement relationships of compressive test

0

100

200

300

400

-0.5 0 0.5 1

[kN]

[mm]

欠損無 NG15-C01NG15CL-C01 NG15CA-C01

0

100

200

300

400

-0.5 0 0.5 1

[kN]

[mm]

欠損無 NG25-C01NG25CL-C01 NG25CA-C01

(a) 15% loss of area (b) 25% loss of area

Figure.6 Situation of eccentricity after compression test

(a)NG15-C01 (b)NG25-C01 (c)NG15CL-C01 (d)NG25CL-C01 (e)NG15CA-C01 (f)NG25CA-C01

Photo.2 Specimen after compression test

(a) 15% loss of area (b) 25% loss of area

Figure.7 Load - strain at both side of aperture 4. TENSILE TEST 4.1 Specimen and test method

The adopted specimens of tensile test are shown in Table 8. Photo 3 shows the experimental setup and Fig. 8 shows the detail of the experiment.

←Yeild strain

▲ strain gauge0

100

200

300

400

-10000-50000

[kN]

[μ]

欠損無 NG15-C01

NG15CL-C01 NG15CA-C01NG0-C02

5mm

0

100

200

300

400

-10000-50000

[kN]

[μ]

欠損無 NG25-C01

NG25CL-C01 NG25CA-C01

←Yeild strain

NG0-C01 5mm

▲ strain gauge

NG0-C01 NG0-C02

5

4.2 Results and Discussions The results of tensile test and axial stiffness are

shown in Fig.9 and Table 9. The elastic axial stiffness shown in Table 9 is calculated from 50kN to 203kN by the experimental results. Axial stiffness of non-repair specimens NG15-T01 and NG25-T01 is much lower as well as compressive test. But it is confirmed that CFRP can recover the former axial stiffness and the axial stiffness of all specimens almost matches the theoretical axial stiffness. Situation for bending deformation affected eccentricity is shown in Fig.10. Thus, it appears that partial and all around CFRP repair restrain eccentricity as well as compression test. Specimens after tensile test are shown in Photo.4. Regarding as specimen of 5% loss of area, a little de-bonding at the edge of CFRP repair occurs due to stretching axial direction but CFRP falls by crack or de-bonding and so on does not appear. While in the case of specimens of 15% and 25% loss of area, void appears between steel pipe and CFRP due to de-bonding. Regarding as partial repair specimen, steel pipe bends greatly by remarkable eccentricity after steel yielding, and circumferential CF cracks and raptures finally. De-bonding does not appear until even allowable stress in the tensile test, it reveals that this CFRP repair method contributes effectively under elastic design. Also, it appears that almost specimen except for NG25CL-T01 is approximately 330kN as de-bonding load, and de-bonding occurs at the 16MPa as average bonding shear stress, and it does not affect loss ratio and kind of repair method mostly, and average bonding shear stress evaluate highly comparing compression loading which outer plane force acts by local buckling.

Table.8 The adopted specimens Name of specimen Loss of area Repair

NG0-T01 N/A N/A NG15-T01

15% N/A

NG15CL-T01 Partial NG15CA-T01 All around

NG0-T02 N/A N/A NG5-T01

5% N/A

NG5CL-T01 Partial NG5CA-T01 All around NG25-T01

25% N/A

NG25CL-T01 Partial NG25CA-T01 All around

0

100

200

300

400

500

0 5 10 15 20

[kN]

[mm]

NS0-T01NG0-T01NG0-T02

0

100

200

300

400

0 2 4 6 8

[kN]

[mm]

NG5-T01NG5CL-T01NG5CA-T01

(a) Normal (b) 5% of loss area

0

100

200

300

400

0 1 2 3 4 5

[kN]

[mm]

NG15-T01NG15CL-T01NG15CA-T01

0

100

200

300

400

0 2 4 6 8

[kN]

[mm]

NG25-T01NG25CL-T01NG25CA-T01

(c) 15% of loss area (d) 25% of loss area

Figure.9 Load - displacement relationships of tensile test

2760400

20

200

20 40

Welding

89.1

40 M39

Photo.3 Experimental setup Figure.8 Detail of tensile test

Table.9 Axial stiffness and increasing ratio of tensile test Name of specimen

Axial stiffness [×103kN]

Improving ratio [%]

NG0-T01 175 0 NG0-T02 172 0 NG5-T01 168 -3

NG5CL-T01 200 15 NG5CA-T01 201 16 NG15-T01 122 -30

NG15CL-T01 184 6 NG15CA-T01 221 27

NG25-T01 37 -79 NG25CL-T01 188 8 NG25CA-T01 233 34

6

0

100

200

300

400

0 0.2 0.4 0.6 0.8 1

[kN]

[mm]

欠損無 NG15-T01NG15CL-T01 NG15CA-T01

0

100

200

300

400

-1 -0.5 0 0.5 1

[kN]

[mm]

欠損無 NG25-T01NG25CL-T01 NG25CA-T01

(a) 15% loss of area (b) 25% loss of area

Figure.10 Influence of eccentricity of tensile test

(a)NG15-T01 (b)NG15CL-T01 (c)NG15CA-T01 (d)NG25-T01 (e)NG25CL-T01 (f)NG25CA-T01

Photo.4 Specimen after tensile test 5. CONCLUSION

In this study, to confirm the effect of CFRP repair method for steel pipe having partial loss of area, compressive and tensile tests for steel pipe repaired by CFRP are carried out. The results obtained from this study are as below. 1) It is made clear that CFRP de-bonding does not occurs until allowable stress of steel, CFRP repair improves

axial stiffness and reduce bending deformation by eccentricity through the compressive and tensile test. 2) It is confirmed that axial stiffness for steel pipe repaired by CFRP almost match a theoretical axial stiffness

under allowable stress of steel though steel yields near aperture. 3) In the case of compressive test, CFRP de-bonds from steel pipe substrate due to local buckling around

loss of area, and it appears that the ultimate state is ruptures after de-bonding. Based on these results, it appears that CFRP repair method is very effective for steel pipe having loss of

area under elastic behavior until CFRP de-bonding. On the other hand, regarding as de-bonding load, bonding length and preventing outer plane de-bonding by local buckling under compression loading are future problems. ACKNOWLEDGMENTS

The authors greatly acknowledge to Prof. Seishi Yamada who coached us politely. Also, this study was supported by Mikimoto Co.,Ltd., Toray Construction Co.,Ltd. and The Kansai Electric Power Co.,Inc. REFERENCES Yuya Hidekuma, Akira Kobayashi, Masatsugu Nagai, Takeshi Miyashita and Yutaka Wakui (2009). ”Repair

Method for the Steel Member by FRP Sheets”, Proc. of the 3rd Symposium on FRP Hybrid Structures and Bridges, pp.91-96 (in japanese)

Dai Wakabayashi, Takeshi Miyashita, Yusuke Okuyama, Masatsugu NAGAI, Norio Koide, Akira Kobayashi, Yuya Hidekuma, Wataru Horimoto (2011). ”Exprimental Study on Repair and Reinforcement Method for Web in Steel Girder Bridge using FRP”, Proc. of The 9th Symposium on Research and Application of Hybrid and Composite Structures (in japanese)

Japan society of civil engineers (2009). ”Manual for Durability Verification of Corroded Steel Structures”, (in japanese)

Xiao-Ling Zhao and Lei Zhang (2007). ”State-of-the-art review on FRP strengthened steel structures”, Engineering Structures 29, pp. 1808-1823

Yukihiro Matsumoto, Nguyen Duc Long, Seishi Yamada and Takahiro Matsui (2012). ”Mechanical Characteristics of CFRP Reinforcement for Corroded Steel under Axial Tension”, CD-ROM Proceeding of The 3rd Asia-Pacific Conference on FRP in Structures

NG0-T01 NG0-T02