Eleventh East Asia-Pacific Conference on Structural

Engineering & Construction (EASEC-11)

“Building a Sustainable Environment”
November 19-21, 2008, Taipei, TAIWAN

EVALUATION OF STRENGTH RECOVERY OF REPAIRED STEEL PIPE PILES

Y. Kitane 1 , N. Watanabe 2 , and Y. Itoh 3

ABSTRACT : Many coastal and offshore steel structures are in need of repair or strengthening due to
corrosion problems. The use of steel patch plates welded underwater over corrosion-damaged areas is
one of the most common repair procedures for corroded submerged steel pipe and sheet piles. A patch
plate thickness, a required weld length, a size of a patch plate, etc. are the design parameters of the
repair design. To determine these design parameters efficiently and achieve an effective repair design,
strength of a repaired structural member has to be accurately evaluated. This paper summarizes an
experimental study of steel pipes repaired with welded patch plates to evaluate the strength recovery
resulting from repair. In this study, the thickness of a portion of steel pipes was reduced by half to
simulate a reduction of the cross section from corrosion, and steel patch plates were fillet-welded to
the pipes with a reduced thickness area. One set of specimens were welded in the open air, and the
other welded in the underwater wet environment to understand effects of welding environments on
strength of repaired pipes. The repaired pipes were tested in flexure or compression until failure.
Based on the experimental results, stiffness and strength recoveries were evaluated, and the
effectiveness of patch plates was also examined.

KEYWORDS: Repair, Steel pipe pile, Underwater wet welding, Strength recovery, Corrosion

1. INTRODUCTION

There are a significant number of marine steel structures suffering severe corrosion damage, and they
are seriously in need of repair or strengthening to prevent structural failure. One of the most common
repair procedures for corrosion-damaged submerged structural steel components is the welding of steel
patch plates over the corrosion-damaged areas in the underwater wet environment [1,2].

Underwater welding is not a very new technique. It is becoming a more and more recognized and
important technique in recent years, especially in the oil and gas industry to maintain submerged steel
structures for energy explorations [3,4,5]. Underwater welding can be classified into two categories:
dry welding and wet welding [3,6]. In the underwater dry welding, a dry environment is established
around the welding area by constructing a dry chamber, and the welding is performed in the chamber.
On the other hand, in the underwater wet welding, the area to be welded is directly exposed to water,
and the welding is performed by a diver in a wet environment. Since the underwater wet welding does
not require construction of a chamber and any heavy equipments, it is much more efficient and
economical welding than the dry welding. Wet welding repairs can be completed at a significantly
lower cost than the dry welding repairs [5]. In addition, in the wet welding, the welder can reach
portions of structures where it may be difficult to construct a dry chamber.

However, there are a few disadvantages of the wet welding. First of all, the welder has to weld under
the influence of ocean currents and a low visibility condition. Secondly, the weld metal undergoes a
rapid cooling since it is directly exposed to the surrounding water. The quenching increases tensile

1
Assistant Professor, Department of Civil Engineering, Nagoya University, Japan.
2
Graduate Student, Department of Civil Engineering, Nagoya University, Japan.
3
Professor, Department of Civil Engineering, Nagoya University, Nagoya, Japan.

EASEC-11 1
 Eleventh East Asia-Pacific Conference on Structural Engineering & Construction (EASEC-11), Taipei, TAIWAN

strength of the weld, but it increases hardness and decreases ductility and impact strength. Thirdly,
hydrogen deassociated from water dissolves into the weld metal and heat affected zones. The high
hydrogen content in the weld metal can result in embrittlement of the material and microscopic cracks
[7,8]. Lastly, the inspection of welds deposited in the underwater wet environment is much more
difficult than that of those deposited in the open air. These factors can change the mechanical
behavior of welded connections made by the wet welding and may lower the welding quality [9,10].

Although there have been some studies [11] on the load carrying capacity of corrosion-damaged steel
pipes, the load carrying capacity of corrosion-damaged pipes repaired in the underwater wet
environment is not well understood. Therefore, an experimental study was performed to examine
strength recovery of steel pipes repaired with patch plates welded in the wet environment. One set of
test specimens were welded in the open air, and the other welded in the wet environment to understand
effects of welding environments on strength of repaired pipes. The repaired pipes were tested in
flexure or compression until failure. Based on the experimental results, stiffness and strength
recoveries were evaluated, and the effectiveness of patch plates was also examined.

2. EXPERIMENTAL PROGRAM

2.1 TEST SPECIMENS AND REPAIR DESIGN

Steel pipes with an outer diameter of 216.3 mm and a thickness of 12.7 mm, specified as STK400 in
Japanese Industrial Standard (JIS) G3444 were used in this study. The radius to wall-thickness ratio is
8.02. To simulate a thickness reduction due to corrosion, the pipe thickness was reduced by 6 mm
uniformly in the circumferential direction over a length of 150 mm.

The repair method used in this study is depicted in Figure 1. A thickness of a patch plate and lengths
of welds were determined so as to restore strength of a repaired pipe to the same level as that of an
intact pipe, where the allowable stress design method was used, and allowable stresses used are shown
in Table 1. The specified fillet weld size was 6 mm. Only side fillet welds were considered in the
design. Five different types of specimens were designed in this study. Type 0 is a new pipe. Type 1
has a 150-mm long reduced thickness section. Type 2, Type 3, and Type 4 all have a 150-mm long
reduced thickness section, and patch plates are welded to cover the reduced thickness portion as repair.
Type 2 has 6-mm thick patch plates welded over the reduced thickness portion, and a length of side
fillet welds are about 1/4 of a required length in the design. Type 3 also has 6-mm thick patch plates;
however, a required length of side fillet welds is provided. Type 4 has twice as many as the number of
side fillet welds than that of Type 2 or Type 3, and it has 12-mm thick patch plates. Test specimens
are summarized in Table 2. Chemical compositions of steels used in this study are shown in Table 3,
where yield and tensile strengths obtained from steel tensile coupon tests are also listed.

Front fillet weld Specified weld

size: 6 mm
Bending

Patch plate Thickness Side fillet

Front weld
reduced
Bending section Welding fillet
weld

150 mm Front fillet weld

200 mm Groove weld

Figure 1. Repair Method by Welded Patch Plates

EASEC-11 2
 Eleventh East Asia-Pacific Conference on Structural Engineering & Construction (EASEC-11), Taipei, TAIWAN

One set of specimens were welded in the open air, and the other set were in the underwater wet
environment so that differences in the load-carrying capacity resulting from the welding environment
could be examined. The underwater wet welding was carried out in the pool filled with natural
seawater. The pool was continuously supplied with new seawater from the sea, and the temperature
was 26˚C when the welding was carried out.

Table 1. Allowable Stress (MPa) [1]

STK400 SM400B
Field weld
(pipe) (patch)
Axial tension/compression 137 137 --
Flexural tension/compression 137 137 --
Shear 78.5 78.5 62.7

Table 2. Test Specimens

Type 0 Type 1 Type 2 Type 3 Type 4

Specimen type

Thickness
0 6 6 6 6
reduction (mm)
Repair No No Yes Yes Yes
Thickness of
-- -- 6 6 12
patch (mm)
Loading* C B C B C B C B C B
Length of each
side fillet weld -- -- 140 80 -- 300 380 370
(mm)
# of side fillet
-- -- 4 4 8
welds**
Specimen KA2C KA2B KA3B KA4C KA4B
KN0C KN0B KN1C KN1B --
designation*** KW2C KW2B KW3B KW4C KW4B
* C: compression, B: four-point bending
** Number of fillet welds for the half patch plate
*** The second letter of the specimen designation shows a welding environment:
N: no weld, A: open air welding, W: underwater wet welding

Table 3. Chemical Composition and Strength of Steels

Chemical composition (%) Yield Tensile

JIS
Material stress strength
designation C Si Mn P S (MPa) (MPa)
Pipe STK400 0.12 0.10 0.56 0.013 0.006 362 394
Patch (t=6 mm) SM400B 0.11 0.18 0.99 0.022 0.005 269 406
Patch (t=12 mm) SM400B 0.15 0.15 0.71 0.013 0.006 271 391
Electrode* (d=4 mm) D4301 0.10 0.10 0.43 0.015 0.007 410 460
* Data of electrode are from a product catalog.

EASEC-11 3
 Eleventh East Asia-Pacific Conference on Structural Engineering & Construction (EASEC-11), Taipei, TAIWAN

Test
specimen

700 600 700

1300
1500

Displacement
transducer
2000
2500
(units: mm)
(a) Compressive test (b) Four-point bending test
Figure 2. Test Setups

2.2 TEST SETUP

Figure 2 shows test setups for compressive and four-point bending tests. Compressive test specimens
were 1,500 mm long, and shortening was measured for a 1,300 mm long portion of a specimen. The
distance between the two supports of the four-point bending test was 2,000 mm, and the two loading
points were 600 mm apart. A flexural specimen was fixed to the axle supported by axle bearings, and
rotation was allowed to rotate at the supports and loading points. The load was applied monotonically
to failure.

3. RESULTS AND DISCUSSIONS

3.1 COMPRESSIVE TEST

Figure 3 shows load-shortening curves from the compressive tests. Both load and shortening in Figure
3 were normalized by the theoretical yield load, Py0, and the theoretical yield shortening, δy0, of Type 0
specimen (KN0C), respectively. Table 4 summarizes the initial stiffness, KI, the obtained maximum
load, Pmax, and the shortening at the maximum load, δmax, for each specimen. The initial stiffness was
determined from the load-shortening curve up to 60% of the theoretical yield load.

Failure modes of Type 0 and Type 4 were the same, and they were plastic local buckling near the ends
of specimen. Type 4 did not show any local buckling or weld failure in the repaired region. Type 1
showed plastic local buckling in the reduced thickness portion. Type 2 showed plastic local buckling
of the repaired portion, resulting in failure of 1.4
front and side fillet welds.
1.2

Table 4 shows that the initial stiffnesses from 1.0

Applied Load (P/Py0)

the tests are in good agreement with

0.8
theoretical values, where the maximum
relative difference is only 3%. 0.6 KN0C
KN1C
0.4 KA2C
The ultimate strengths of Type 4 specimens KW2C
were close to that of Type 0 because their 0.2 KA4C
KW4C
failure modes were similar. For Type 2 0.0
specimens, the ultimate strengths were about 0 5 10 15 20
Shortening (δ /δ y0)
10% smaller than that of Type 0, and δmax
values were about 1/3 of that of Type 0. Figure 3. Load-Shortening Curves
(Compressive test)

EASEC-11 4
 Eleventh East Asia-Pacific Conference on Structural Engineering & Construction (EASEC-11), Taipei, TAIWAN

Table 4. Stiffness, Ultimate Strength, and Shortening at Ultimate Strength (Compressive test)

Type 0 Type 1 Type 2 Type 4

Specimen KA2C KW2C KA4C KW4C
KN0C KN1C
(Air) (Wet) (Air) (Wet)
KI (MN/mm) 1.23 1.09 1.37 1.35 1.83 1.79
(% error from theory) (+1.0) (-0.78) (+1.5) (+1.2) (-0.54) (-3.4)
Pmax (MN) 3.29 1.61 2.98 2.94 3.42 3.36
δmax (mm) 26.1 3.11 8.61 8.25 13.1 14.5
Pmax/Py0 1.17 0.57 1.06 1.04 1.21 1.20
δmax/δy0 11.3 1.34 3.72 3.56 5.65 6.27

There was not much difference in the ultimate strength or shortening at the ultimate strength between
specimens welded in the open air and those in the underwater wet environment.

3.2 FOUR-POINT BENDING TEST

2.5

Figure 4 shows load-deflection curves from

2.0
the four-point bending tests. Deflection was
Applied Load (P/Py0)

measured at midspan. Both load and

1.5
deflection in Figure 4 were normalized by the
theoretical yield load, Py0, and the theoretical KN0B KN1B
1.0
yield deflection, δy0, of Type 0 specimen KA2B KW2B
(KN0B), respectively. Table 5 summarizes KA3B KW3B1
0.5
the initial stiffness, KI, the maximum load, KW3B2 KA4B

Pmax, and the deflection at the maximum load, KW4B

0.0
δmax, for each specimen. 0 5 10 15 20
Midspan Deflection (δ/δy0)
All specimens formed a plastic hinge at Figure 4. Load-Deflection Curves
midspan. Type 1 showed plastic local (Four-point bending test)
buckling on the compression side in the
reduced thickness portion. Types 2 and 3 showed bulging of patch pates near front fillet welds that
were close to the reduced thickness portion, and those front fillet welds failed when the load was
around the maximum. In addition, for KW2B, KW3B1, and KW3B2, the groove weld on the
compression side also failed at about the maximum load.

Table 5 shows that the initial stiffnesses from the tests are in good agreement with theoretical values,
where the maximum relative difference is only 5%.

When the ultimate strength of each specimen is compared with that of Type 0 (KN0B), KA2B, KW2B,
and KW3B1 were lower than KN0B by only a few percent, KW3B2 was lower by 10 percent, and
KA3B was higher than KN0B by only a few percent. Therefore, for Type 2 and Type 3, the ultimate

Table 5. Stiffness, Ultimate Strength, and Deflection at Ultimate Strength (Four-point bending test)

Type 0 Type 1 Type 2 Type 3 Type 4

Specimen KA2B KW2B KA3B KW3B1 KW3B2 KA4B KW4B
KN0B KN1B
(Air) (Wet) (Air) (Wet) (Wet) (Air) (Wet)
KI (kN/mm) 58.5 45.1 60.4 62.8 68.5 69.9 68.7 94.8 94.3
(% error from theory) (+3.2) (-5.4) (-2.3) (+1.0) (-3.3) (-3.0) (-2.2) (-2.0) (-1.2)
Pmax (kN) 614 303 585 595 633 583 552 890 873
δmax (mm) 103 25.6 39.0 44.0 38.9 29.5 28.5 66.5 67.4
Pmax/Py0 1.58 0.78 1.51 1.53 1.63 1.50 1.42 2.29 2.25
δmax/δy0 15.1 3.73 5.69 6.41 5.67 4.31 4.16 9.70 9.82

EASEC-11 5
 Eleventh East Asia-Pacific Conference on Structural Engineering & Construction (EASEC-11), Taipei, TAIWAN

strength was recovered to the level of Type 0, except for KW3B2. The ultimate strength of Type 4,
KA4B and KW4B, was about 1.4 times the ultimate strength of Type 0, implying that it was not an
efficient design. It was also found that there was no significant difference between Type 2 and Type 3
because side fillet welds in Type 2, which were shorter than those required in the design, did not fail in
the test. Type 2 and Type 4 specimens did not show significant differences in the load carrying
capacity between the open air welding and the wet welding. However, Type 3 showed a 10%
reduction in the ultimate strength of specimens repaired underwater. It is likely that this reduction was
caused by the groove weld failure that occurred in KW3B1 and KW3B2.

4. CONCLUSIONS

In this study, stiffness and strength recoveries of steel pipes repaired with patch plates welded in the
underwater wet environment were examined experimentally in compression and flexure. Test results
showed that the initial stiffness was recovered as designed, and that the ultimate strength could be
about 10% less than that of a repair carried out in the open air. These results are only applicable to the
specific cases examined in this study. It is necessary to perform finite element analyses of repaired
steel pipes with a wide range of different design parameters by strategically including characteristics
of underwater wet welds.

ACKNOWLEDGEMENTS

The research reported herein was partly supported by The Japan Iron and Steel Federation. The
authors gratefully acknowledge the support.

REFERENCES

1. Coastal Development of Institute of Technology, Port Steel Structure Corrosion-Prevention

and Repair Manual, Coastal Development of Institute of Technology, Japan, 1997 (in
Japanese).
2. Kiyomiya, O., Noguchi, T., and Yokota, H., “Mechanical properties of repaired steel sheet
piles,” Report of the Port and Harbor Research Institute, Vol. 28, No. 3, 1989, pp. 147-198
(in Japanese).
3. Nixon, J. and Billingham, J., “A survey of underwater welding techniques,” Endeavour, New
Series, Vol. 11, No. 3, 1987, pp. 143-148.
4. Wernicke, R. and Pohl, R., “Underwater wet repair welding and strength testing on pipe-patch
joints,” Journal of Offshore Mechanics and Arctic Engineering, Vol. 120, 1998, pp. 237-242.
5. Perez-Guerrero, F. and Liu, S, “Maintenance and repair welding in the open sea,” Welding
Journal, Vol. 84, No. 11, 2005, pp. 54-59.
6. American Welding Society, Specification for Underwater Welding, AWS D3.6M: 1999,
American Welding Society, Miami, FL, 1999.
7. Ibarra, S., Grubbs, C.E., and Olson, D.L., “Metallurgical aspects of underwater welding,”
Journal of Metals, Vol. 40, No. 12, 1988, pp. 8-10.
8. de Medeiros, R.C. and Liu, S., “A predictive electrochemical model for weld metal hydrogen
pickup in underwater wet welds,” Journal of Offshore Mechanics and Arctic Engineering, Vol.
120, 1998, pp. 243-248.
9. Brown, R.T., and Masubuchi, K., “Fundamental research on underwater welding,” Welding
Journal, Vol. 54, No. 6, 1975, pp. 178-188.
10. Fukute, T., Abe, M., Hasegawa, H., and Matsuda, S., “A study on fractural mechanism and its
improvement for underwater-welded steel sheet pile structures,” Report of the Port and
Harbor Research Institute, Vol. 36, No. 4, 1997, pp. 43-68 (in Japanese).
11. Fujii, K., Kondoh, T., Tamura, I., Watanabe, E., Itoh, Y., Sugiura, K., Nogami, K., and Nagata,
K., “Ultimate strength of corroded circular steel tube in marine environment,” Journal of
Structural Engineering, JSCE, Vol. 52A, 2006, pp. 721-730 (in Japanese).

EASEC-11 6