UNIAXIAL COMPRESSION TEST OF STEEL PLATE

choruspillowUrban and Civil

Nov 29, 2013 (3 years and 7 months ago)

40 views

Page
1
of
8
UNIAXIAL COMPRESSION
TEST OF STEEL PLATE
BONDED VARIOUS FRP S
HEETS
Takeshi MIYASHITA
Specially appointed associate professor
Nagaoka University of Technology
1603
-
1 Kamitomioka, Nagaoka, Niigata, Japan
mtakeshi@vos.nagaokaut.ac.jp
*
Yusuke OKUYAMA
Ph.D.
Student
Nagaoka University of Technology
Dai WAKABAYASHI
Nippon Expressway Research Institute Company Ltd.
Norio KOIDE
Kawasaki Heavy Industries, Ltd.
Yuya HIDEKUMA
, Akira KOBAYASHI
Nippon Steel Composite Company, Ltd.
Wataru HORIMOTO
Kurabo Industri
es, Ltd.
Masatsugu NAGAI
Professor
Nagaoka University of Technology
Abstract
This paper is a fundamental study on rational repair and reinforcement of webs in corroded
steel girder bridges using Fiber Reinforced Plastic (FRP). Uniaxial compression test o
f steel
plates bonded various FRP sheets is carried out. The objective of this test is to select FRP
sheets that have reinforcing effect following large deformation induced by buckling.
Furthermore, a layer of polyurea putty is inserted between the steel p
late and the FRP sheet
and its effect is investigated
.
Lastly
, the method that predicts elastic buckling load of the steel
plate with FRP is developed.
Keywords
:
Fiber Reinforced Plastic, steel plate, uniaxial compression test, buckling
1.
Introduction
Mos
t of the deterioration for steel bridges is the corrosion. During their
period of
service,
progress of corrosion is inevi
table
due to the influence from surrounding circumstance. The
conventional methods repairing and reinforcing the damage are the replace
ment of corroded
members or the attachment of steel plate on them. However, these methods are regulated
in
service because of requiring heavy machineries. Therefore, efficient and rational method
repairing and reinforcing the damage is strongly needed.
In
this situation, Fiber Reinforced Plastics (FRP) has been paid to attentions due to light
weight and high stiffness, and many studies have been reported so far
[1], [2]
. Previous
studies mainly focus on the applications to members subjected to normal stress
; flanges in a
steel girder bridge or chord members in a steel truss bridge. However, corrosions in the steel
Page
2
of
8
bridges mostly occur at webs near the ends of supports where shear force with large
deformation at the ultimate state is dominant. Here
in
, althoug
h debonding of FRP under large
deformation becomes a problem, this type of investigation has been
few
reported.
Therefore, the objective of this research is to carry out a fundamental study on rational repair
and reinforcement of webs in corroded steel gir
der bridges using FRP. Uniaxial compression
test of steel plates bonded various FRP sheets is carried out. This test aims to select FRP
sheets that have reinforcing effect following large deformation induced by buckling.
Furthermore, a layer of polyurea pu
tty is inserted between the steel plate and the FRP sheet
and its effect is investigated.
Lastly
, the method that predicts elastic buckling load of the steel
plate with FRP is developed.
2.
Outline of experiment
2.1
Materials
The property used in this research i
s shown in
Table
1. In this study, six kinds of FRP sheets
are
selected
; high elastic carbon fiber (CE), high strength carbon fiber (CU), carbon fiber
strand sheet (CS), glass fiber (G), high strength polyethylene (P) and hybrid fiber (H,
C:G=1:1).
Table
2 shows the dimensions and the material characteristic of a steel plate used in
this study.
Table
3 shows the material properties of the polyurea putty and resin.
Table 1.
Properties of FRP sheets
.
Measured value
Thickness
Young's modulus
Young's modulus
mm
MPa
MPa
mm
MPa
t
cd
E
cd
E
cm
t
FRP
E
FRP
CE
High elastic carbon fiber
0.116
6.40E+05
7.80E+05
0.966
9.37E+04
CU
High strength carbon fiber
0.121
2.40E+05
2.79E+05
0.971
3.48E+04
CS
Carbon fiber strand sheet
0.286
6.40E+05
7.45E+05
1.976
1.08E+05
G
glass fiber
0.123
7.40E+04
1.05E+05
0.973
1.29E+04
P
High strength polyethylen
0.108
8.80E+04
9.30E+04
0.958
1.05E+04
H
Hybrid fiber
0.121
3.83E+05
5.01E+05
0.971
6.24E+04
FRP
Thickness
Young's modulus
Type
Fiber sheet
Design value
Sign
Table
2
.
Properties of steel plate
.
Length
Cross section
Young's modulus
Yield stress
mm
mm
MPa
MPa
SM490YB
400 / 800
60
×
9
2.00E+05
428
Table
3
.
Propertie
s of putty and resin
.
Polyurea putty
FU-Z
FR-E5P
FB-E7S

for
CS

Density
g/cm
3
1.25
1.17
1.48
Amount of coating
g/m
2
1000
1000
2500
Resin thickness
mm
0.80
0.85
1.69
Young's modulus
MPa
54.7
2533
2434
Resin
2.2
Specimens
Figure
1 shows the cross section of specimens. FRP sheets were bonded to both sides of the
steel plate. The parameters in this experiment are shown in
Table
4. The types of the specimen
are classified into two cases accor
ding to the number of laminated FRP sheets. In Case 1, the
number of FRP layers bonded to both sides of the steel plate is one. In Case 2, the number of
FRP layers is two.
Page
3
of
8
Resin
Fiber sheet
Resin
Polyurea putty
Primer
Resin
Fiber sheet
Resin
Primer
Steel
Steel
Figure
1
. Cross section of Specimen.
Table
4
.
Experimental parameter
(Unit: mm)
.
Specimens
FRP
FRP
Sheet length
Sign
1st layer
2nd layer
1st layer / 2nd layer
1-1
CEN
CE
-
400
800
×
1-2
CEU
CE
-
400
800

1-3
CUN
CU
-
400
800
×
1-4
CUU
CU
-
400
800

1-5
CSN
CS
-
400
800
×
1-6
CSU
CS
-
400
800

1-7
GN
G
-
400
800
×
1-8
GU
G
-
400
800

1-9
PN
P
-
400
800
×
1-10
PU
P
-
400
800

1-11
HN
H
-
400
800
×
1-12
HU
H
-
400
800

1-13
CEN
CE
-
300
400
×
1-14
CEU
CE
-
300
400

2-1
2CEN
CE
CE
450/400
800
×
2-2
2CEU
CE
CE
450/400
800

2-3
GCEN
G
CE
450/400
800
×
2-4
GCEU
G
CE
450/400
800

2-5
PCEN
P
CE
450/400
800
×
2-6
PCEU
P
CE
450/400
800

2-7
GCSU
G
CS
450/400
800

2-8
CECSU
CE
CS
750/700
800

Case
Polyurea
putty
Steel length
2.3
Experimental
method
In this experiment, horizontal displacements and strains on both sides of the specimen were
measured.
Figure
2 shows the measurement positions of strain gauges and displacement
sensors. In order to realize simple supported boundary
conditions, both ends of the steel plate
were sharply cut. As a preliminary experiment, uniaxial compression test for a steel plate
revealed its maximum load as 9.53kN. Since theoretical value (Euler buckling load) of the
plate is 9.68kN, it can be said th
at expected boundary condition is realized.
Page
4
of
8
400
800
60
9
192
192
8
8
188
188
100
112
112
100
300
400
60
9
138
138
37
37
25
25
42
42
8
8
50
50
(a)
Case
1
(Steel: 800mm, FRP: 400mm)
(b)
C
ase 1
(Steel: 400mm, FRP: 300mm)
400
800
60
9
167
167
8
8
25
25
100
100
87
87
188
188
25
25
300
700
800
60
9
8
8
25
25
17
17
25
25
17
17
22
388
388
22
300
(
c
)
Case
2
(Steel: 800mm, FRP: 4
5
0mm)
(
d
)
C
ase
2
(Steel:
8
00mm, FRP:
75
0mm)
Figu
re
2
.
Details of specimen
.
3.
Results and discussions
3.1
Reinforcing effect
Table
5 shows the results of experiment. In the
t
able
,
P
E
and
P
max
mean Euler buckling load
and the maximum load in experiment respectively. The reinforcing effect is defined as the
fol
lowing equation
.
Reinforcing effect (%)
=
max
100
E
E
P P
P


(1)
Figure
3
shows the reinforcing effect in
C
ase1 and
C
ase2.
Table
5
.
Result of experiment
.
CEN-1
CEN-2
CEN-3
CEU-1
CEU-2
CEU-3
CUN-1
CUN-2
CUN-3
CUU-1
CUU-2
CUU-3
Euler buckling load
kN
9.57
9.52
9.45
9.69
9.47
9.52
9.57
9.57
9.63
9.57
9.80
9.48
Maximum load
kN
12.13
12.29
12.28
12.32
12.03
12.90
10.82
11.06
10.93
10.96
11.18
10.91
Central displacement
mm
43.65
52.60
49.65
33.05
38.90
33.05
97.80
111.95
91.50
80.30
94.10
94.35
CSN-1
CSN-2
CSN-3
CSU-1
CSU-2
CSU-3
GN-1
GN-2
GN-3
GU-1
GU-2
GU-3
Euler buckling load
kN
10.04
9.54
9.58
9.68
9.58
9.43
9.67
9.57
9.77
9.73
9.71
9.50
Maximum load
kN
16.44
16.47
16.31
15.34
15.15
14.81
10.77
10.62
10.93
9.86
10.36
10.50
Central displacement
mm
25.15
23.05
33.90
64.30
64.40
39.25
-
198.75
155.95
79.65
100.95
77.75
PN-1
PN-2
PN-3
PU-1
PU-2
PU-3
HN-1
HN-2
HN-3
HU-1
HU-2
HU-3
Euler buckling load
kN
9.57
9.60
9.46
9.70
9.46
9.55
9.58
9.47
9.64
9.68
9.75
9.54
Maximum load
kN
10.05
10.36
10.24
10.13
10.76
10.71
11.96
11.35
10.85
11.21
12.07
11.03
Central displacement
mm
212.05
219.55
125.45
109.65
229.05
179.40
45.45
49.80
74.65
25.00
41.15
26.25
CEN-1
CEN-2
CEN-3
CEU-1
CEU-2
CEU-3
Euler buckling load
kN
38.15
38.47
38.50
38.11
38.19
37.62
Maximum load
kN
49.64
47.41
50.99
53.60
53.20
51.57
Central displacement
mm
12.75
14.55
19.10
23.40
21.35
12.95
2CEN-1
2CEN-2
2CEN-3
2CEU-1
2CEU-2
2CEU-3
GCEN-1
GCEN-2
GCEN-3
GCEU-1
GCEU-2
GCEU-3
Euler buckling load
kN
9.61
9.57
9.60
9.20
9.99
9.43
10.00
9.53
9.49
9.96
10.00
9.54
Maximum load
kN
13.25
15.64
13.69
15.33
15.27
14.27
14.48
12.82
12.65
14.41
14.49
12.24
Central displacement
mm
35.15
34.40
39.60
75.95
62.00
62.40
43.40
46.05
52.05
44.80
46.60
54.90
PCEN-1
PCEN-2
PCEN-3
PCEU-1
PCEU-2
PCEU-3
GCSU-1
GCSU-2
GCSU-3
CECSU-1
CECSU-2
CECSU-3
Euler buckling load
kN
9.49
10.04
9.31
9.50
10.00
9.87
9.44
10.05
9.79
10.21
9.56
9.86
Maximum load
kN
11.15
13.80
13.82
12.86
14.07
13.99
16.82
17.32
17.31
26.55
25.48
27.24
Central displacement
mm
50.70
43.95
49.65
57.40
51.25
53.40
49.40
35.60
55.65
52.15
60.45
23.75
Specimen
Specimen
Specimen
Specimen
Specimen
Specimen
Page
5
of
8
0
20
40
60
80
100
120
140
160
180
200
CEN
CEU
CUN
CUU
CSN
CSU
GN
GU
PN
PU
HN
HU
CEN
(400)
CEU
(400)
(P
max
-
P
E
) / P
E
×
100 (%)
Average
0
20
40
60
80
100
120
140
160
180
200
2CEN
2CEU
GCEN
GCEU
PCEN
PCEU
GCSU
CECSU
(P
max
-
P
E
) / P
E
×
100 (%)
Average
(a)
Case
1
(
b
)
Case
2
Figure
3
.
Reinforcing effect
.
3.2
Deformation
In
Table
5,
central horizontal displacement at the ultimate states of FRP is also shown. Herein,
the ultimate state, which
were
debonding or breaking
,
was determined from measurements by
the strain gauges. It was confirmed from
the result of C
ase 1 that low modulus fi
ber sheets
such as P and G showed better performance on deformation.
3.3
Load
-
displacement curve
Representative examples of load
-
displacement curve in CE and 2CE are shown in
Figure
4
. In
the
f
igure
s, CEU and CEN mean the cases with and without putty. In the
case of CEN, the
load dropped suddenly when the central displacement reached beyond 50mm due to the
fracture of FRP sheets. On the other hand, the load in CEU did not show sudden drop by the
effect of putty.
Similarly, in the case of 2CEN, the load droppe
d suddenly when the central displacement
reached beyond 40mm. On the other hand, the load in 2CEU did not show sudden drop.
Therefore, it is found that the polyurea putty used in this study can prevent
the
debonding or
breaking of FRP, and improve the flex
ibility.
0
2
4
6
8
10
12
14
16
18
20
0
20
40
60
80
100
120
Load
[kN]
Displacement
[mm]
CEU
-
1
CEN
-
1
0
2
4
6
8
10
12
14
16
18
20
0
50
100
150
200
250
Load
[kN]
Displacement
[mm]
2CEU
-
1
2CEN
-
1
(a)
CE
(
Case1

FRP: 400mm
)
(
b
)
2CE
(
Case2

FRP: 450mm
)
Figure
4
.
Load
-
d
isplacement curve
.
3.4
Failure modes
3.4.1
Case1
The most of failure mode in this case was debonding or breaking at the center of the
spe
cimens in tensile side. How
ever
,
the case using C
S only shown other failure mode
that
was
debonding at the end of the specimens
in
Figure
5
. This is the reason why the change of cross
section at the end of the specimen is larger since CS is thicker than other FRP sheets.
In the case
of CSN, FRP sheets were delaminated at the end of the specimen in small
deformation. On the other hand, in the case of CSU, the polyurea putty suppressed the
Page
6
of
8
progress of delamination although FRP sheets were delaminated at the end of the specimen.
(a)
CSN
(
b
)
CSU
Figure
5
.
CS(Ca
s
e1

FRP:400mm)
.
3.4.2
Case2
In this case, the combination of FRP sheets lead to different failure modes. In the case of 2CE,
which is the case using two CE l
ayers, failure mode was the delamination at the end of the
specimen in spite of the existence of the polyurea putty.
In the case using low elastic FRP sheets such as G or P, the most of failure mode was the
breaking at the center of FRP sheet although the
re was a little difference by the existence of
the polyurea putty.
In the case of GCS that G and CS were used in
the first and second layer, failure modes were
the delamination at the end of
the all specimen. However, GCSU
shown different failure
mode comp
aring to other two specimen as shown in
Figure
6
. The specimen bent at the end of
FRP sheet. On the ot
her hand, in the case of CECSU
, the specimen bent at its center in spite
of using CS sheet as shown in
Figure
7
. Thus, when
the length of
FRP sheet having
high
stiffness such as CS is short on the steel plate, it is necessary to consider
the effect of
reinforcing
length
.
Figure
6
.
GCSU(Ca
s
e2

FRP:400mm)
.
Figure
7
.
CECSU(Ca
s
e2

FRP:750mm)
.
Page
7
of
8
4.
Prediction of elastic buckling load
4.1
Formulation
In t
he case not considering FRP bonding length
,
Euler buckling load
for
composite section of
the steel and FRP
is the following.




2
V
V
P EI L


(
2
)
w
here
(EI)
V
is a flexural rigidity of composite cross section,
L
is a length of the steel plate. In
this method, it is assumed that the bonding length of FRP sheet is equal to the length of the
steel plate.
Next, in t
he cas
e considering FRP bonding length
,
the method predicting elastic buckling load
is needed to
develop
. At first, as shown in Figure
8
, differential equations of
w
i
, which are
displacements in beam
i
(
i
= 1 ~
N
), are expressed as follows.
4 2
2
4 2
0
i i
i
d w d w
dx dx

 
(
3
)
where


i
i
P EI


(
4
)
N
is the total number of beam
segments
,
L
i
is a length of each beam,
(EI)
i
is a composite
flexural rigidity,
P
is a
n
applied
axial load.
General solutions of equation (3) are given as the followings.
1 2 3 4
sin cos
i i i i i i i
w C x C x C x C
 
   
(
5
)
where
C
ji
(
j
=
1, 2, 3, 4) are unknown coefficients determined from boundary, continuity and
symmetrical conditions.
Then, equations determining unknown coefficients can be summarized as follows.


0
B

C
(
6
)
where
C
is a vector consisting of unknown coefficients. Condition having nontrivial solution
in equation (6) is


det 0
B

(
7
)
The minimum value of
P
(

0) satisfying equation (7) affords the elastic buckling load in
Figure 8. The ca
lculation of equation (7) is carried out numerically since it is difficult to
calculate analytically as the number
N
increases.
Figure
8
.
Analytical model for formulation
.
Page
8
of
8
4.2
Result
Figures
9
and 1
0
show predicted result from equations (2) and (7) in Case
1 and Case2
respectively. Here, vertical axes in the figures are evaluated by the following equation.
max
100
V
V
P P
P


(
12
)
where,
P
max
means the maximum load in experiment and
P
V
means the predicted loads from
equations (2) and (7).
It is found from
F
igures
9 and 10
that the accuracy of prediction becomes better in equation
(7) than in equation (2) since the exact FRP bonding length is considered in equation (7).
-
2.9
-
1.9
1.3
1.7
-
14.3
-
20.1
7.2
2.1
2.8
5.7
-
1.0
-
1.5
-
2.8
4.9
-
30
-
20
-
10
0
10
20
30
CEN
CEU
CUN
CUU
CSN
CSU
GN
GU
PN
PU
HN
HU
CEN
(400)
CEU
(400)
(P
max
-
P
V
) / P
V
(%)
Average
3.0
4.0
3.6
4.1
1.8
-
5.0
8.0
2.9
3.6
6.5
2.7
2.2
-
2.0
5.8
-
30
-
20
-
10
0
10
20
30
CEN
CEU
CUN
CUU
CSN
CSU
GN
GU
PN
PU
HN
HU
CEN
(400)
CEU
(400)
(P
max
-
P) / P (%)
Avarage
(a)
eq. (2)
(
b
)
eq. (7)
Figure
9
.
Prediction of elastic bu
ckling load (Case1).
-
17.4
-
12.6
-
8.7
-
7.4
-
10.9
-
7.4
-
26.7
0.4
-
30
-
20
-
10
0
10
20
30
2CEN
2CEU
GCEN
GCEU
PCEN
PCEU
GCSU
CECSU
(P
max
-
P
V
) / P
V
(%)
Average
-
7.1
-
1.6
-
0.3
1.1
-
2.6
1.1
-
6.9
0.8
-
30
-
20
-
10
0
10
20
30
2CEN
2CEU
GCEN
GCEU
PCEN
PCEU
GCSU
CECSU
(P
max
-
P) / P (%)
Avarage
(a)
eq. (2)
(
b
)
eq. (7)
Figure
10
.
Prediction of elastic buckling load (Case2).
5.
Conclusions
In this research, a fundamental study on rational repair and reinforcement of webs in corroded
steel girder bridges using Fiber R
einforced Plastic (FRP) was carried out. Uniaxial
compression test of steel plates bonded various FRP sheets was conducted in order to select
FRP sheets that have reinforcing effect following large d
eformation induced by buckling.
Lastly
, the method that p
redicts elastic buckling lo
ad of the steel plate with FRP
wa
s
developed.
References
[1]
Okura I., Fukuui T., Nakamura K. and Matsugami T.,

Decrease in Stress in Steel Plates
by Carbon Fiber Sheets and Debonding Shearing Stress

,
Journal of Structure
Mechanics and Earthquake Engineering
,
No.689, pp
.
239
-
249
, 2001
(in Japanese).
[2]
Kishi N., Mikami H. and Zhang G,

Numerical Analysis of Debonding Behavior of FRP
Sheet for Flexural Strengthening RC Beams

,
Journal of Materials, Concrete
Structures
and Pavements
,
N
o.725,
pp
.
255
-
272
, 2003
(in Japanese).