stress fatigue study for grp samanci tar akd (2)

7/26/2019 Stress Fatigue Study for GRP SAMANCI TAR AKD (2)

http://slidepdf.com/reader/full/stress-fatigue-study-for-grp-samanci-tar-akd-2 1/11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/239773623

Fatigue failure analysis of surface-cracked(±45°)3 filament-wound GRP pipes underinternal pressure

Article in Journal of Composite Materials · May 2012

Impact Factor: 1.17 · DOI: 10.1177/0021998311414945

CITATIONS

6

READS

352

3 authors, including:

Ahmet Samanci

Necmettin Erbakan Üniversitesi

5 PUBLICATIONS 51 CITATIONS

SEE PROFILE

Ahmet Akdemir

Necmettin Erbakan Üniversitesi

50 PUBLICATIONS 290 CITATIONS

SEE PROFILE

Available from: Ahmet Akdemir

Retrieved on: 17 May 2016

https://www.researchgate.net/profile/Ahmet_Samanci?enrichId=rgreq-d2102598-8079-4c90-98dc-1716d8a513a6&enrichSource=Y292ZXJQYWdlOzIzOTc3MzYyMztBUzoxMTk2ODQ1NDQzMzk5NjhAMTQwNTU0NjQyNzYyNg%3D%3D&el=1_x_4


https://www.researchgate.net/profile/Ahmet_Akdemir2?enrichId=rgreq-d2102598-8079-4c90-98dc-1716d8a513a6&enrichSource=Y292ZXJQYWdlOzIzOTc3MzYyMztBUzoxMTk2ODQ1NDQzMzk5NjhAMTQwNTU0NjQyNzYyNg%3D%3D&el=1_x_4


https://www.researchgate.net/?enrichId=rgreq-d2102598-8079-4c90-98dc-1716d8a513a6&enrichSource=Y292ZXJQYWdlOzIzOTc3MzYyMztBUzoxMTk2ODQ1NDQzMzk5NjhAMTQwNTU0NjQyNzYyNg%3D%3D&el=1_x_1


https://www.researchgate.net/institution/Necmettin_Erbakan_Ueniversitesi?enrichId=rgreq-d2102598-8079-4c90-98dc-1716d8a513a6&enrichSource=Y292ZXJQYWdlOzIzOTc3MzYyMztBUzoxMTk2ODQ1NDQzMzk5NjhAMTQwNTU0NjQyNzYyNg%3D%3D&el=1_x_6




https://www.researchgate.net/institution/Necmettin_Erbakan_Ueniversitesi?enrichId=rgreq-d2102598-8079-4c90-98dc-1716d8a513a6&enrichSource=Y292ZXJQYWdlOzIzOTc3MzYyMztBUzoxMTk2ODQ1NDQzMzk5NjhAMTQwNTU0NjQyNzYyNg%3D%3D&el=1_x_6



https://www.researchgate.net/?enrichId=rgreq-d2102598-8079-4c90-98dc-1716d8a513a6&enrichSource=Y292ZXJQYWdlOzIzOTc3MzYyMztBUzoxMTk2ODQ1NDQzMzk5NjhAMTQwNTU0NjQyNzYyNg%3D%3D&el=1_x_1

https://www.researchgate.net/publication/239773623_Fatigue_failure_analysis_of_surface-cracked_453_filament-wound_GRP_pipes_under_internal_pressure?enrichId=rgreq-d2102598-8079-4c90-98dc-1716d8a513a6&enrichSource=Y292ZXJQYWdlOzIzOTc3MzYyMztBUzoxMTk2ODQ1NDQzMzk5NjhAMTQwNTU0NjQyNzYyNg%3D%3D&el=1_x_3

https://www.researchgate.net/publication/239773623_Fatigue_failure_analysis_of_surface-cracked_453_filament-wound_GRP_pipes_under_internal_pressure?enrichId=rgreq-d2102598-8079-4c90-98dc-1716d8a513a6&enrichSource=Y292ZXJQYWdlOzIzOTc3MzYyMztBUzoxMTk2ODQ1NDQzMzk5NjhAMTQwNTU0NjQyNzYyNg%3D%3D&el=1_x_2



JOURNAL OF

C O MP O S I TE

MATER IALS Article

Fatigue failure analysis of surface-cracked(45)3 filament-wound GRP pipes under internal pressure

A Samanci1, N Tarakcioglu2 and A Akdemir 3

Abstract

In this study, the fatigue behavior of (45)3 filament-wound composite pipes with a surface crack under alternatinginternal pressure was investigated. Glass-reinforced plastic (GRP) pipes were made of E-glass/epoxy and tested in theopen-ended condition. The pipes had a surface crack with a notch–aspect ratio of a/c ¼ 0.2 and notch-to-thickness ratiosof a/t ¼ 0.25, 0.38, or 0.50 in the axial direction. Tests were carried out in accordance with ASTM D2992. This standardoffers 25 cycles/min and a load ratio of R ¼ 0.05. Tests were performed at three different load levels: 50%, 40%, and 30%

of ultimate hoop stress. Whitening, leakage, and final failure of GRP pipes were observed, and fatigue test results werepresented by means of S–N curves.

Keywords

polymer–matrix composites (PMCs), fatigue, filament winding, surface crack, GRP pipe, internal pressure

Introduction

Glass fiber reinforced epoxy tubes are increasingly used

in gas and liquid transfer pipes, high-pressure containersin chemical plants, and in the aerospace and defense

industries. Filament-wound composite pipes made of

glass-reinforced plastic (GRP) have many potential

advantages over pipes made from conventional mate-

rials, such as high specific stiffness and strength, good

corrosion resistance, and thermal insulation. With

developments in manufacturing technology to produce

filament-wound pipes, there has been a growing interest

in the application of filament-wound fiber-reinforced

cylindrical composite structures. Polymeric composites

offer many cost advantages over metals due to a con-

siderably higher strength-to-weight ratio. In structural

filament-wound GRP pipes, cracks have been found in

different forms, locations, orientations, sizes, and types.

Surface crack problems are more complicated than

other problems. Stress fields and the crack growth

behavior of semielliptical or semicircular surface

cracks greatly depend on the crack shape and crack

size and inclination, as well as on the pipe dimensions.

Surface cracks that exist in pressure vessels, pipelines,

tanks, and rocket motor casings can lead to catastrophic

failures, especially under corrosive and cyclic loading

conditions.1

The most intensive research on filament-wound fiber–

glass/epoxy composite tubes was conducted by Perreux

and his coworkers2–6

who investigated the effect of fre-quency on the fatigue performance of composite pipes

under biaxial loading.6 The biaxial monotonic and

fatigue behavior of a multidirectional filament-wound

fiber–glass/epoxy pipe were investigated by Ellyin

and Martens.7,8 The biaxial fatigue and leakage charac-

teristics of fiber-reinforced composite tubes were studied

by Wolodgo.9 The uniaxial fatigue behavior of filament-

wound fiber–glass/epoxy tubes was researched by

Kaynak and Mat.10

Tarakciog ˘ lu et al. have also contributed greatly to

research on filament-wound fiber–glass/epoxy composite

1Cihanbeyli Vocational school, University of Selcuk, Konya, Turkey2Department of Mechanical Education, University of Selcuk, Konya,

Turkey3Department of Mechanical Engineering, University of Selcuk, Konya,

Turkey

Corresponding author:

A Samanci, Cihanbeyli Vocational school, University of Selcuk, Konya,

Turkey

Email: [email protected]

Journal of Composite Materials

46(9) 1041–1050

! The Author(s) 2011

Reprints and permissions:

sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/0021998311414945

jcm.sagepub.com



tubes11–16. They studied monotonic and fatigue behavior

of intact and surface-cracked specimens with different

winding angle composite pipes. The effect of surface

cracks on strength was investigated theoretically and

experimentally in GRP pipes that were exposed to open-

ended internal pressure.12 The fatigue failure behaviors of

glass/epoxy 55

and 75

filament-wound pipes underinternal pressure were investigated experimentally. The

whitening, leakage, and final failure stages of GRP

pipes have been observed, and the fatigue results were

presented by means of S–N curves.13,16 The fatigue behav-

iors of surface-cracked filament-wound pipes with high

tangential strength wereinvestigatedin corrosive environ-

ment.14 Fatigue crack growth of 55 and75 filament-

wound GRP pipes with a surface crack were investigated

under cyclic internal pressure.1,15 In these tests, the effects

of sizes of surface crack on fatigue failure behavior and

applied hoop stress levels were investigated. Arikan17

studied the failure analysis of filament-wound composite

pipes with an inclined surface crack under static internal

pressure. The failure of the GRP pipes was analyzed and

the dependence on crack behavior and burst strengths was

determined.

Sayman andhis coworkers studied fatigue behavior of

different filament winding angle composite pipes18–20.

The effect of winding angle on fatigue strength was inves-

tigated theoretically and experimentally in GRP pipes

that were exposed to close-ended internal pressure.

They reported that the optimum winding angle for

the composite pressure cylinders or vessels under

internal fatigue pressure load was obtained as 45

orientation.18

The present study shows the fatigue damage progres-

sion of (45)3 filament-wound E-glass/epoxy pipes with

a surface crack under pure hoop loading condition.

The whitening, leakage, and final failure stages of

GRP pipes with different crack depth ratios were

observed, and the results obtained were presented by

means of S–N curves for a crack depth ratio of only a/

t ¼ 0.5.

Experimental

Specimen preparation

Filament-wound GRP pipes with (45)3 winding

angles were manufactured by IZORELL Co. using a

computer numerical control (CNC) winding machine.

Vetrotex 1200 Tex E-glass fiber, a CIBA-GEIGY LY

556/HY 917/DY 070 bisphenol-A epoxy resin system

with 100:90:0.5 weight ratios and the CIBA-GEIGY

QZ-13 mold release agent were used to make the

GRP pipes. The mechanical properties of the matrix

and reinforcement materials are given in Table 1.

After winding process, GRP pipes were cured for

2hr at 135C on a mandrel in a slow motion rotary

oven. After pulling out the mandrel, the pipes were

post-cured for 2 hr at 150C. The pipes were cut into

test lengths of 300 mm using a diamond wheel-cutting

saw. The manufacturing properties of the filament-

wound composite pipes are given in Table 2. Axial

elliptical surface notches were cut using a diamond-

grinding disc (1 mm thick and 38 mm diameter) with a

horizontal machining center. Then the notches were

sharpened by a lancet. The pipes with a surface crack

had a notch–aspect ratio of a/c ¼ 0.2 and notch-to-thickness ratios of a/t ¼ 0.25, 0.38, and 0.50 in the

axial direction as shown in Figure 1. The parameter a

is the crack depth, 2c is the crack length, and t is the

wall thickness. The average wall thickness is 2.15 mm

and the standard deviation of the thickness of GRP

pipes were found as 0.1 mm.

Test setup

Fatigue tests were conducted using a programmable

logic controlled (PLC) servohydraulic testing machine.

Figure 2 shows the open-ended internal pressure fatigue

test setup for GRP pipe specimens. The procedure for

Table 2. Manufacturing and mechanical properties of GRP pipe specimens

Winding angle 45 Ultimate hoop strength (MPa) 337

Number of layer 3 Yield hoop strength (MPa) 80

Pipe length (mm) 300 Strains ratio (ehoop/eaxial) 0.94

Internal pipe diameter (mm) 72 Fiber volume fraction (%) 50

Pipe shell thickness (mm) 2.15 Elasticity modulus, Ehoop (GPa) 10.8

Table 1. Mechanical properties of the fiber and the resin

E

(GPa)

TS

(MPa) u

(g/cm3)

emax

(%)

E-glass (Vetrotex

1200 tex, fiber

diameter 17 mm)

73 2400 0.25 2.6 1.5–2

Epoxy resin

(Ciba Geigy LY 556)

3.4 50–60 0.38 1.2 4–5

1042 Journal of Composite Materials 46(9)



determining the long-term fatigue strength of a compos-

ite pipe was based on ASTM standard D2992.21 The

ASTM standard stipulates cycling the internal hydro-

static pressure at a rate of 25 cycles/min (0.42 Hz) over

the full pressure range and a load ratio of R ¼ 0.05. The

magnitude of the fatigue test stress levels was decided

based on the strength under static internal pressure.

Three different stress levels were applied. These

maximum stress levels were 30% (101 MPa), 40%

(135 MPa), and 50% (169 MPa) of the static strength

of the specimen.

Experimental results and discussion

The fiber volume fraction (V f ) of the GRP pipes was

determined to be 0.5 by burn-off tests in accordance

with ASTM D2584. So theoretical modulus of the com-

posite in first and second principal material directions

as E 1¼ 38 MPa, E 1¼ 6.5 MPa, and theoretical ultimate

strength in tension as X t ¼ 1225MPa and Y t ¼50MPa

were calculated respectively by using the fiber and

matrix material properties in Table 1. Ultimate tensile

strength of the GRP pipe in fiber direction was found

theoretically as 1225 MPa by using well known the ruleof mixtures equation X t ¼ V f f þ (1 V f ) m ¼ 0.5

2400 MPaþ 0.5 50 MPa ¼ 1225 MPa where, X t is the

tensile strength of the composite in fiber direction, f and m are the tensile strength of the fiber and matrix,

respectively. If the assumptions as below were consid-

ered for simplicity, theoretical stress of the pipe in tan-

gential direction is approximately found as 866 MPa

according to netting analysis (1225 MPa sin 45 ¼

866 MPa). The pipe wall thickness is thin. When the

elemental piece of pipe was considered, the pipe can

be assumed as a flat plate. Netting analyses assume

that all loads are supported by the fibers only, neglect-

ing any contribution of the matrix and any interaction

between the fibers. The filaments also possess no bend-

ing or shearing stiffness and carry only the axial tensile

loads.

However, the tangential ultimate stress was found

experimentally as 337 MPa. Calculated ultimate stress

value is excesively higher than experimental result. The

difference between theoretical and experimental tangen-

tial stress was caused by manufacturing conditions such

as fiber wettability, porosity, winding and curing

parameters.

Figure 1. Schematic presentation of open-ended internal pressure test apparatus with surface-cracked pipe.

Figure 2. Internal pressure fatigue test setup.

Samanci et al. 1043



Statically internal pressure tests

Strain gauges were assembled on the unnotched GRP

pipes in the axial and hoop directions to obtain the

static burst strength. Stress–strain values were recorded

for determination of mechanical properties. These speci-mens were loaded with internal pressure up to burst

pressure at a 1 MPa/s loading rate. For determining the

stress level to be used in the fatigue tests, it was necessary

to know their static hoop strength values. Thus, the

mechanical properties of the specimens were found for

static internal pressure conditions. The average tangential

burst stress was found as 337 MPa with a standard devi-

ation of 15 MPa for three specimens. The mechanical

properties of the composite pipe are given in Table 2.

The first sign of damage is the formation of thin

white lines parallel to the fiber direction caused by the

matrix cracking and fiber–matrix separation. These

lines are formed by shear stress, which is parallel to

the fiber. Whitening starts and becomes concentrated

at nearly 80 MPa hoop stress value. The increase of

whitening with the pressure is the sign of debonding

and delamination. In addition, when the internal pres-

sure increases, the length of the pipe becomes shorter

and the diameter of the pipe grows. After the tests,

the fiber angle increases from 45 to 50 – 55 and

serious macro deformations are observed on the

specimens.

The stress–strain ( – e) values that were recorded

throughout the experiment are represented as a graph

in Figure 3.

Fatigue tests

In the fatigue experiments, three specimens were used

for each notch-to-thickness ratios (a/t ¼ 0.25, 0.38, or

0.50) and tests were performed at three different load

levels: 50%, 40%, and 30% of ultimate hoop stress.

The tests were repeated three times separately and the

experimental data that shows scattering were presented

in Figure 4. The previous studies1,13–16 on this topic

were conducted using selected applied load levels. As

a reference value, the applied load levels were chosen as

reported.

Fatigue test results are presented and interpreted by

means of S–N curves with scattering data for leakage

initiation and burst in Figure 4. It can be seen that

leakage initiation was not observed at high applied

stress ( applied ¼ 0.5 static) and pipe suddenly failed

catastrophically by bursting. On the contrary, at low

applied stresses ( applied ¼ 0.4 and 0.3 static) leakage

began firstly and then burst failure occurred after a

number of cycles. The tests were continued until burst

failure of the pipes. In case of oil leakage from sealing,

a metal clamp was applied to the region of sealing as

seen in Figure 6.

Figure 3. The stress–strain ( – e) graph of the (45)3 GRP pipe subjected to monotonic internal pressure under open-ended

conditions.




During the fatigue tests, three damage mechanism

stages were observed. The first damage mechanism

was whitening. At this stage, matrix cracking, debond-

ing, and delamination occurred. The white zones grew

wider and deeper along the fiber-winding direction.

Microcracks started along this direction. It is observed

that all of the specimens have whitened approximately

at first cycles (Figures 5–8). The second damage stage is

the pinhole formation stage, followed by a leakage

stage. This situation is depicted in Figure 5(b). The pin-

hole occurred due to delamination in this region. It is

reported that the pinhole progressed from the innersurface of the pipe to outward due to the pressurized

fluid effect at each loading cycle. The leakage began

at the pinhole as a small droplet and after a while

intense leakage began.16 Finally, burst failure occurred

by fiber breakage.

In the fatigue test, when the internal pressure was

applied to the pipes, the diameter of the pipes enlarged

and the length of the pipes shortened. At this time,

there was no deformation out of the sealing equipment

of the pipe. The bottleneck formation occurred in the

vicinity of the sealing region (Figures 5(a) and 6). Some

specimens showed an oil leakage from sealing. To avoid

this situation, metal clamping was applied to this region

as in Figure 6. This metal clamping allows shortening

of the pipe in the axial direction by sliding easily over

the sealing.

The pipe specimens with a surface crack (a/t ¼ 0.5)

final failure damage occurs at the surface crack (Figures

6 and 8(b)). However, a majority of pipe specimens

with shallow surface cracks (fully in a/t ¼ 0.25 and

partly in a/t ¼ 0.38) were not damaged in the crack

region. Failure occurred in the region near the sealing

equipment because of the bottleneck effect (Figures 5(a)

and 8(a)). In damage formation, the effect of bottleneck-

ing was greater than the crack effect in these specimens.

This situation was also reported by Tarakciog ˘ lu.12 Thus,

the fatigue lives of these pipes do not represent the effect

of the surface crack. For this reason, the S–N curves of

these specimens could not be drawn in Figure 4 because

of this situation.

The pipe burst from sealing region as shown in

Figure 5(a) was sliced along the axial direction on the

surface crack. Hence, the damage zone underneath the

surface crack could be observed in Figure 5(b). As seen

in this image, the delaminated zone has formed underthe surface crack and has been associated with pinhole

formation leading to the formation of a leakage path

for oil to reach the surface of the pipe. However, before

the completion of the pinhole formation, the pipe has

failed by a burst at the sealing region due to bottleneck-

ing effect at the 11946th cycle. It can be seen that the

pinhole could not reach the surface crack. If this had

happened, the pipe would have burst at this region.

Fatigue damage behavior

Fatigue damage can be classified into fiber debonding,

matrix cracking, delamination, and fiber fracture.

Fiber debonding

The diameter of the pipes increased due to hoop stress,

similar to a compressed helical spring. The diametrical

expansion of the pipe caused a shear stress between the

fiber bands. As a result of this situation, shear stress

among the fibers caused fiber debonding. The debond-

ing damage can be related to the whitening along the

fiber direction as shown in Figures 5, 6(a), 7, and 8.

0

100

200

000.01000.100101

H o o p s t r e s s , σ h o o p

( M P a )

Number of cycle, (N)

Burst failure

leakage initiation

Figure 4. Stress number of cycles graph of (45)3 GRP pipes with a surface crack for different damage propagation stage

(a/t ¼ 0.50).

Samanci et al. 1045



Figure 5. Fatigue failure of the pipe (a/t ¼ 0.25, applied ¼ 0.4 static , N ¼ 11,946). (a) Damage zone at the end of the pipe. (There isno explode damage in the surface crack region). (b) Damage zone underneath the surface crack on the pipe (Penetrant dye: Methylene

blue).

Figure 6. Damage zone at the center of the (45)3 filament wound GRP pipe (a/t ¼ 0.5, applied ¼ 0.4 static , N ¼ 1,150).




Figure 7. Damage zones of the (45)3 filament-wound GRP pipe (a/t ¼ 0.5, applied ¼ 0.4 static, N ¼ 380). (a) Matrix cracking.

(b) Oil droplets formation as a result of pinhole (N¼ 750).

Samanci et al. 1047



Matrix cracking

The hoop stress caused by the diametric expansion of

the pipe increased over the yield stress of the matrix

material. The matrix cracking occurred along the

pipe axis on the pipe surfaces. This matrix micro

cracking propagates with increase in the fatigue

cycle (Figure 7(a)). Figure 7(b) shows the oil droplet

on the surface of the pipe as a result of pinhole

formation.

Delamination

The delamination damage mechanism occurred when

the fatigue cycle began. The diameter of the pipe

increased and the length of the pipe decreased with

increasing fatigue cycle under internal pressure. The

maximum level of this expansion-shortening of the

pipe caused a bottleneck effect near the piston seal at

both ends of the pipe. This bottleneck effect caused

bending stress due to the experimental conditions.

Figure 8. The effect of stress ratio on where the fracture occurred for pipes with the same crack depth ratio (45, a/t ¼ 0.25). (a)

Final failure near the sealing region ( applied ¼ 0.5 static , N ¼ 1430). (b) Final failure of the crack region ( applied ¼ 0.3 static , N ¼

78.000).




The pipes consist of six layers, which have 45 fil-

ament winding. These layers are similar to coaxial heli-

cal springs that have different diameters. The 45

filament winding angle can be considered a right–left

helix. The shear stress as defined by MOD III occurred

between the different winding angle layers. The other

mechanism is bending stress as defined by MOD II. Thebending stress occurred due to the effect of bottleneck-

ing. The final failure in the GRP pipe with a low crack/

depth ratio occurred due to the bottleneck effect at the

end of the pipe.

Fiber fracture

Due to fiber debonding, matrix cracking, and delami-

nation failures, the pipes showed a sudden increase in

successive and instantaneous fiber breakage. Figure 8

shows the final damage of the pipes loaded at maximum

and minimum (50% and 30% of static hoop stress,

respectively) stress levels of test loads. The final failure

began with sweating of the pipe shell surrounding

the pinhole. Catastrophic failure occurred with fiber

pullout and fracture. The macrocrack propagation

occurred parallel to the fiber direction in which the

minimum amount of energy had been dissipated to

achieve fiber pullout and fracture. As seen in Figure

8(b), the pipes showed triangular leaf-shaped openings

when they underwent sudden bursting. A similar failure

mechanism was reported by Gemi et al.16 The stress

ratio affected where the fracture occurred, whether in

the crack region or sealing region of the pipe for the

pipes with the same crack/depth ratio. For instance, thedamage occurred in the crack region when a low stress

ratio (0.3 static) was applied. On the contrary, the

damage occurred in the sealing region with a high

stress ratio (0.5 static). Both situations are shown in

Figure 8.

Conclusion

Experimental investigations of (45)3 filament-wound

GRP pipes with a surface crack were conducted. The

GRP pipes with a surface crack were exposed to open-

ended internal pressure. The fatigue failure behavior of

composite pipes was investigated. The conclusions are

as follows:

. The damage starts with whitening and matrix crack-

ing. With the help of matrix cracking, a leakage path

starts to form from the inner side toward the bottom

of the surface crack. Once the leakage path is com-

pleted, the leakage begins. This damage stage is fol-

lowed by the catastrophic failure of the pipe.

. It is observed that when the applied load is high,

the leakage and final failure coincide, whereas

when the applied load is low, the leakage is followed

by the final failure. In other words, when the applied

load is high, the pipe suddenly bursts without any

leakage, and when the applied load is low, the pipe

shows leakage first and then fails.

. Fiber debonding, matrix cracking, delamination,

and fiber fracture affect the damage formation andpropagation. The first three damage stages occur

simultaneously, and then final failure (fiber fracture)

occurs.

. The damage occurred in the crack region when the

low stress ratio ( applied ¼ 0.3 static) was applied. On

the contrary, the damage occurred near the sealing

region with a high stress ratio ( applied ¼ 0.5 static).

. The maximum fatigue life is obtained in the crack

region with N ¼ 78.000 for the pipes with a/t ¼ 0.25,

applied ¼ 0.3 static. The minimum fatigue life is

obtained in the crack region with N ¼ 175 for the

pipes with a/t ¼ 0.5, applied ¼ 0.5 static.

. Authors believe that bottlenecking effect can be

eliminated by increasing the pipe thickness gradually

at the sealing region.

Acknowledgment

The authors acknowledge the support of the Coordination

Committee of Scientific Research projects of Selcuk

University, Project no: 2003-45.

References

1. Tarakciog ˘ lu N, Samanci A, Arikan H and Akdemir A.The fatigue behavior of (55)3 filament wound GRP

pipes with a surface crack under internal pressure.

Composite Structures 2007; 80: 207–211.

2. Rousseau J, Perreux D and Verdiere N. Influence of wind-

ing patterns on the damage behavior of filament-wound

pipes. Composite Science and Technology 1999; 59:

1439–1449.

3. Ferry L, Perreux D, Rousseau J and Richard F.

Interaction between plasticity and damage in the behavior

of [þ,]n fiber reinforced composite pipes in biaxial load-

ing (internal pressure and tension). Composite Part B Eng

1998; 29: 715–723.

4. Richard F and Perreux D. A reliability method for opti-

mization of f fiber reinforced composite pipes. Reliability

Engineering and System Safety 2000; 68: 53–59.

5. Joseph E and Perreux D. Fatigue behavior of glass–fiber/

epoxy matrix filament-wound pipes. Composite Science

and Technology 1994; 52: 469–480.

6. Perreux D and Joseph D. The effect of frequency on the

fatigue performance of filament wound pipes under biaxial

fatigue experimental results and damage model. Composite

Science and Technology 1997; 57: 353–364.

7. Ellyin F and Martens M. Biaxial fatigue behavior of a

multidirectional filament-wound glass–fiber/epoxy pipe.

Composite Science and Technology 2000; 61: 491–502.

Samanci et al. 1049



8. Martens M and Ellyin F. Biaxial monotonic behavior of

a multidirectional filament-wound glass–fiber/epoxy

pipe. Composites Part A 2000; 31: 1001–1014.

9. Wolodgo J, Biaxial fatigue and leakage characteristics of

fiber reinforced composite tubes. PhD. Thesis, University

of Alberta, Canada, 1999.

10. Kaynak C and Mat O. Uniaxial fatigue behavior of fila-

ment-wound glass–fiber/epoxy tubes. Composite Scienceand Technology 2001; 61: 1833–1840.

11. Akdemir A, Tarakciog ˘ lu N and Avci A. Stress corrosion

crack growth in glass–polyester composites with surface

crack. Composites Part B 2000; 32: 123–129.

12. Tarakciog ˘ lu N, Akdemir A and Avci A. Strength of fil-

ament wound GRP pipes with surface crack. Composites

Part B 2001; 32: 131–138.

13. Tarakciog ˘ lu N, Gemi L and Yapici A. Fatigue failure

behavior of glass/epoxy (55) filament wound pipes

under internal pressure. Composite Science and Technology

2005; 65: 703–708.

14. Avci A, Sahin OS and Tarakciog ˘ lu N. Fatigue behavior

of surface cracked filament wound pipes with high tan-gential strength in corrosive environment. Composites

Part A 2006; 38: 1192–1199.

15. Samanci A, Avci A, Tarakciog ˘ lu N and Sahin OS.

Fatigue crack growth of filament wound GRP pipes

with a surface crack under cyclic internal pressure.

Journal of Material Science 2008; 48: 5569–5573.

16. Gemi L, Tarakciog ˘ lu N, Akdemir A and Sahin OS.

Progressive fatigue failure behavior of glass/epoxy

(75)2 filament-wound pipes under pure internal pres-

sure. Materials and Design 2009; 30: 4293–4298.

17. Arikan H. Failure analysis of (55)3 filament wound

composite pipes with an inclined surface crack understatic internal pressure. Composite Structures 2010; 92:

182–187.

18. Erkal S, Sayman O and Benli S. Fatigue damage in

composite cylinders. Polymer Composites 2010; 31:

707–713.

19. Onder A, Sayman O and Dogan T. Burst failure load of

composite pressure vessels. Composite Structures 2009;

89: 159–166.

20. Sayman O. Analysis of multi-layered composite cyl-

inders under hydrothermal loading. Composites Part

A Applied Science and Manufacturing 2005; 36:

923–933.

21. Standard practice for obtaining hydrostatic or pressuredesign basis for ‘‘fiberglass’’ (glass-fiber-reinforced

thermosetting resin) pipe and fittings. American

Society for Testing Materials (ASTM) designation:

D2992-91.


stress fatigue study for grp samanci tar akd (2)

Documents