This paper presents the results of an experimental investigation into the Ultimate Elastic Wall Stress (UEWS) of ±55° filament wound composite pipes . The UEWS test appears to provide an attractive alternative to the current method. This method has proved to be one of the most effective to achieve the short term test due to the advantage of fast testing. Moreover, it has been found to be sensitive to changes in key manufacturing and raw material parameters.
The pipes were subjected to biaxial loading, which was achieved by combinations of hoop and axial stress. These loads were applied as groups of cycles which were gradually increased until the UEWS has been determined. Various ratios of hoop to axial stress were applied to the pipes ranging from pure axial to pure hoop loading at room temperature and at 65C°. These ratios were investigated by applying different pressures in both the main and small chambers that built inside the pipe; therefore it is unnecessary to add any external loads on the pipe wall. Tests were conducted to observe the leakage through the pipe wall.
The main failure mode observed was weepage through the pipe's wall, which was due to intensive matrix microcracking. The results from the UEWS tests were presented in the form of failure envelopes with the effects of testing at an elevated temperature. Finally, the degradation in elastic properties of the pipe's wall is also discussed and plotted against wall stress.
Introduction
Filament wound glass fibre reinforced epoxy (GRE) pipes are being increasingly used in wide range of applications, such as high pressure containers, petrochemical, and oil industries due to the superior corrosion resistance and high strength to weight ratio[1]. This has encouraged researchers to carry out and develop experimental and theoretical investigations in order to digest their mechanical and failure behaviour under variety of loading conditions[2].
With regarding to stress-strain relationship of filament wound composite material is generally non-linear up to failure, the non-linear is significantly enhanced by damage such as matrix cracking and yielding [3]. Nahas (1984), reviewed several methods for the analysis of non-linear behaviour for laminated reinforced composite materials[4]. Sun and Tao, (1996) modelled the stress-strain behaviour of composite laminate by considering both the material non-linearity and progressive matrix cracking[5].
The behaviour of composite fibre reinforced is quite complicated because of not only anisotropic but it is also having inhomogeneous actions. This leads to varity of failure mechanism. For example, GRP pipes subjected to pure hydrostatic pressure, matrix can develop cracks, plies could delaminate or fibres could debone from matrix and then fail. Whereas fibres will buckled and then broke under pure hoop loading [6]. Pipes tested under biaxial loading condition failed with different strengths; consequently a symmetrical biaxial failure envelope will be constructed. Although, biaxial failure envelope is a graph of the axial failure strength versus hoop failure strength, in this study the contribution of ultimate elastic wall stress envelope was fitted.
In current investigations, 200 mm inside diameter, 2 m length, and 6 mm wall thickness glass fibre reinforced epoxy pipes of ±55° layup were produced by filament winding, and the ply properties are determined using Halpin-Tsai equations[7]. These pipes were subjected to six stress ratios with positive internal pressure. The UEWS and stress-strain response up to failure were experimentally determined. These results are presented in term of failure modes as well as the UEWS envelope.
Biaxial Testing According to ASTM D2992
Currently, the ISO 14692 standard is used to predict the maximum service pressure for GRE pipes. The standard describes the method used to establish the regression line to predict long term allowable stress [8]. This method based on pressurising the GRE pipes statically or cyclically to get at least 18 points to establish an acceptable regression line, with at least one sample providing a point in excess of 10,000 hours. The graph with log-log format is plotted in term of stress versus time to failure. By extrapolating the lower Confidence Limit (LCL) from the regression line, the rating for a design lifetime of 20 years can be obtained. The expected failure mode of this test is Weepage which is generally due to the presence of a network of matrix cracks which form over time. The regression line obtained from this procedure is important information that qualifies the product to be manufactured and defines the pressure rating to be used in pipe's system design.
Figure 1: Example of regression line indicates the pressure at design lifetime of 20 years and the pressure at 1000 hours reconfirmation test [8]
The Hydrostatic Design Basis (HDB) is method to reconfirmation of the product using regression line that built from ASTM D2992. This is needed when there is change to materials, manufacturing processes, construction, liner thickness or fitting design. In the HDB test, the product is subjected to the 1000 hours hydrostatic pressure based on the 1,000 hour lower Prediction Limit (LPL) of the regression line obtained from ASTM D2992. If there is no weepage or burst at 1000 hours it means that the product is safe to conclude the same design life time of 20 years. Figures 1 illustrates the regression line of the pipe using ASTM D2992.
The Ultimate Elastic Wall Stress (UEWS)
While acknowledging the benefits of the current reconfirmation method HDB based on regression analysis, manufacturers, the need for a rapid and effective method of monitoring changes in product quality, becomes inescapable imperative. The UEWS test appears to provide an attractive alternative to this method. UEWS is the maximum circumferential wall stress resulting from internal hydrostatic pressure or axial loads that produces an elastic deformation in any direction [9]. rost and Cervenka1 observed that 'damage involving matrix cracking in composite materials is associated with observable non-linearities n elastic behaviour. rocedures have been described and reported to a limited extent in the public domain for use with vinyl ester-based pipe as well as PVC' [10] .
The purpose of the UEWS test is to study the stress-strain response, and defining the maximum stress level in elastic zone [9]. The main advantage of UEWS is that the test takes only a few hours per sample to perform, compared with the 1,000 hour HDB reconfirmation procedure. The UEWS has been found to be sensitive to changes in key manufacturing and raw material parameters such as the quality of the 'size' on the glass fibre, which influences the bond with the resin. These effects are not easily picked up in the ASTM2992 tests, indeed the HDB reconfirmation procedure provides no product information other than a yes/no outcome. The sensitivity of the test to material and quality parameters has also been found to be superior to tests such as interlaminar shear and through-thickness strength [11].
In practice the UEWS has been found to correlate well to the long term LCL value obtained from regression tests. The main criticism of the UEWS concept is that the exact significance of the measured quantity is not fully defined.
Comparison between procedures
The main advantage claimed for procedures based on cyclic and static fatigue is that they provide a realistic statistical approach to establishing a long-term pressure rating when there is a slow deterioration of properties. For this reason it makes them attractive in connection with statistically-based design. The ASTM2992 confirmation test appears to have significant drawback due to the time needed to achieve full qualification of new products (10,000 hours ~14 months).
For new piping products, where the regression line slope is not identifiable in advance, this requires significant trial and error to determine the pressures to be used, which often results in a qualification period that exceeds two years. Although it is generally agreed that proof of long term stability is desirable, long term static fatigue measurements may not be the best method of achieving this [11].
Table 1 discussed the differences between the regression-based ISO 14692 and the UEWS procedure.
Regression-based procedures
UEWS tests
Time, expense and convenience.
Expensive procedure requiring ~2 years to achieve qualification.
Simple procedure, which can be carried out in less than day.
Ability to define a long-term pressure rating.
Provides the basis for a 20 year design rating.
Identifies a stress level, below which the rate of damage progression is very low. This stress corresponds approximately to the stress determined by long term regression.
Ability to quantify changes due to process and materials.
HDB reconfirmation procedure is adequately sensitive to these effects, but takes a minimum of 1,000 hours to perform and only provides a yes/no answer.
Very sensitive to these effects.
Ability to quantify effects of chemical environment.
Limited.
Limited.
Determination of Ultimate Elastic Wall Stress
This procedure describes the determination of ultimate elastic wall stress of the Wavistrong thermosetting resin pipes in a short time period. This test method establishes the Ultimate Elastic Wall Stress (UEWS) of the pipes. Data obtained by this method are of use only in predicting the behaviour of these pipes under condition of temperature, time, method of loading and stress situation. It is appears to be similar to those used in the actual test and for comparison of the behaviour with the stated hydrostatic properties. They are generally not indicative of the long-term strength of the product used for the test. Either the internal pressure or the hoop stress may be listed in the requirements.
The UEWS test based on, loading a specimen to a prescribed Cycle Test Pressure (CTP) in the short time interval. The nominal group pressure rating was determined, which started with 10% of the expected UEWS ,these groups were gradually increased by 10% of the expected UEWS up to failure as shown in figure 2. The cycle pressure was applied to specimen from zero up to (CTP-), the number of cycles was 10, each cycle required to hold for 1 minute, and then 1 minute no pressure. The pressure increments from zero to CTP- preferred to be uniform with wall stress increment of about 5-10 MPa/min [9].
The test pressure and the strain that generated in the pipe wall were recorded at the end of each cycle. Next cycle test pressure (CTPi+1), was determined by increasing 0.1 of the expected UEWS to the previous CTP as the following:
Fig. 2, Definition of test cycle and Cyclic Test Pressure (CTP)[9].
The UEWS will achieved when the difference between maximum strain recorded in the end of tenth cycle is 5% greater than the maximum strain recorded in the end of first cycle of the same cycle group as shown in figure 3.
Two additional groups of ten cycles at least needed to be done beyond the UEWS point that will enable to clearly observe the ultimate elastic wall stress (UEWS) point.
Figure 3. The Ultimate Elastic Wall Stress (UEWS) is indicated by the intersection of line A and B
Experimental
Pipe specimen
The entire MDA epoxy/E-glass filament wound pipes are manufactured in Future Pipe Industries (FPI), with reinforced and tapered ends to be appropriate for the end caps. All the pipes were geometrically similar with inside diameter of 200 mm, 2 m pipe's length, 6 mm wall thickness, and ±55° winding angle. This angle is commonly encountered because Netting Analysis suggests that this is the best angle to use in piping system where the ratio of applied hoop to axial stress is 2:1 [7]. The following values of parameters are used to evaluate the elastic and shear modulus of the pipe wall.
Where, is the elastic modulus of the fibre, is the elastic modulus of the matrix, is the fibre volume fraction, are the Poisson ratios of fibre and matrix respectively.
The engineering elastic constants were calculated from the rule of Halpin-Tsai equations [7] by:
(3), (4)
(5), (6)
(7), (8)
(9)
The above calculation was implemented in Excel. The stiffness matrix in the directional sate was determined and then transformed to obtain the corresponding values in coordinate system of the pipe wall. The resulted axial, hoop and shear modulus were obtained as:
Biaxial loading test procedure
Figure 4 shows a schematic diagram of biaxial loading test rig. It consist of two pistons screwed onto a hollow shaft, two tapered end caps were fitted onto pipe end by means of an epoxy bond, air driven pump, and intensifier. This rig will enable to control the ratios of axial and hoop stresses in the pipe wall.
In this design the wall stress ratios were investigated by pressure difference between the main pressure applying in the pipe and small chamber pressure, due to an intensifier (no need to apply any external load to the pipe).This would enable a multi-axial tests to be carried out in verity ratios of hoop and axial stress.
DAQ
Comp.
Pressure release tank
Pump
Intensifier
Figure 4. Schematic diagram of the test rig for carrying out biaxial loading.
The extent stress ratios in this investigation was, ranging from pure axial to pure hoop loading at room temperature and at 65C°, using an environmental chamber. Figure 5 shows the UEWS equipments used in the experimental test.
Figure 5. Test equipment for UEWS testing: (1) environmental chamber, (2) data capture, and (3) pipe spool with bonded end fittings.
Two transducers were detected the pressure reading from the main and small chambers, as well as two strain gauges (60 mm length), were fixed in the pipe wall in axial and hoop direction in order to detect the strains in these direction.
Stiffness reduction
The stress-strain curve of filament wound composite pipes is predominantly linear, and then strongly non-linear behaviour takes place for the final failure. This comportment causes degradation in elastic modulus that could be due to matrix degradation. This degradation considerably noted when axial loading was dominated. However, slightly changes in wall stiffness occurred under hoop loading. The stress-strain response and elastic modulus decline of filament wound composite pipe under design load configuration 2:1 hoop/axial stress will be discussed later.
Results and discussion
Stress strain response
Figure 6 shows the stress-strain diagram of the pipe under 2:1 hoop/axial stress ratio (pure internal pressure) and the decline of elastic modulus. This is the ideal loading arrangement for the ±55° filament wound pipes, so the load is taken up by fibres and the matrix played a little part in the behaviour of the pipes at low strains. The tensile test was conducted to the pipe in the conditions of virgin, 50% of UEWS, 100% UEWS and weepage. The elastic modulus was determined from the linear regression of stress strain curve.
Figure 6. Stress-Strain response up to failure. Elastic modulus decline as function of wall stress
The initial reduction in axial elastic modulus was noted which stabilized at 50% of UEWS (90MPa). A notable reduction in elastic modulus occurred at 100% of UEWS (180MPa), when possibly cracks start to develop parallel to the fibres. Pipe at weepage has slightly regression in elastic modulus from that at the point of UEWS, as shown in figure 7.
Figure 7. Reduction in elastic modulus of the filament wound pipe under 2H:1A stress ratio
Biaxial loading test
The UEWS test was carried out to 200 mm MDA fibreglass epoxy pipes under biaxial loading with hoop/axial tress ratios of 0:1, 0.5:1, 1:1, 2:1, 4:1, and 1:0, at room temperature and at 65C°. Figure 8, appears to show the stress-strain response and the UEWS evaluation of the tested pipe with stress ratio of 2: 1 hoop/axial.
At room temperature the UEWS has been achieved with axial stress of about 90 MPa. This was reduced to 52 MPa under pure axial loading, the strength reduction was approximately 40 % of that in the design loading configuration (2:1). On the other hand, the strength increment was about 47 % at 4:1 hoop/axial stress ratio. This ratio appears to be the highest capability of the pipe under these ratios. Pipe under pure hoop loading has UEWS of about (250MPa), 22% greater than that in the design loading case. However, the rest ratios were followed the smooth curve that connected between the 2:1 point and pure axial point.
Figure 8. Shows the evaluation of UEWS for ±55° filament wound composite pipe
These UEWS points, failure stress and the equivalent internal pressure in both the main and small chambers at failure region are listed below:
Stress ratio Hoop/ Axial
Failure pressure (main chamber) bar
Failure pressure (small chamber) bar
UEWS (Hoop) MPa
UEWS (Axial) MPa
Hoop failure stress MPa
Axial failure stress MPa
Failure mode
1
0 :1
0
92
0
52
0
77
Complete matrix failure
2
0.5:1
24
96
32
64
40
80
Complete matrix failure
3
1:1
55
110
76
76
92
92
Complete matrix failure
4
2:1
140
140
180
90
220
110
Complete matrix failure
5
4:1
220
110
360
90
420
105
Burst
6
1:0
160
0
250
0
300
0
Burst
Table 2. Summary of the test results of ±55° filament wound glass-fibre /epoxy pipes under biaxial loading
At 65C° the strength is generally reduced with the greatest reduction occurred in axial load sector, which was about 10%. Nevertheless, the strength at 65 C° was 5% greater than that detected at room temperature in the case of design loading configuration. This increment was possibly due to the residual stresses that could generated during pipe manufacturing and curing. In the case of hoop load sector the strength failed by about 8%. This reduction has shrinking the failure envelope towards the origin figure 9.
The UEWS Envelopes
All the envelopes that provided from the researchers in field of filament wound composite pipes refer to failure state. However, the collected UEWS points resulted from the experimental work achieved for the GRE pipes under verity of stress ratio were contributed as UEWS envelope. The UEWS envelope is a graph of axial UEWS strength versus hoop UEWS strength for each loading ratios. Figure 9, shows the UEWS envelope for the ±55° filament wound composite pipe for six different loading ratios. This envelope can provide the product strength under different loading conditions and it appears to be sensitive to discern the changes of the parameters such as temperature, manufacturing, fibre and resin types.
The general form of the UEWS envelops is might be expected, with high strength over the full range from pure axial to pure hoop loading at room temperature ,with falling strength at every load ratios (apart from 2:1 hoop/axial Stress ratio) as temperature is increased. The narrowing of the envelopes with increasing temperature could be due to high of the resin component to the water absorption effects compared to the fibre component. At elevated temperature the observed water acts a plasticiser which eventually lowers the glass transition temperature of the resin so that could change the matrix material to amorphous state. This conversion will influence the material stiffness by decreasing the elastic modulus of the pipe and then that could allow the passage of liquid between the fibre even though it remaining intact in most cases to cause failure by weeping.
Figure 9. Shows the UEWS envelope of ± 55° filament wound fibreglass epoxy pipe at room temperature and at 65C°
Macro observations
Observations of macro failure mechanisms are based on the naked eye inspections. Figure 10 shows pictures of specimens during and after testing under their respect loading ratios. Three types of macro failures were observed in the range of testing related to the load conditions.
Tests with 2:1, 0.5:1, 1:1 hoop/axial stress have a positive axial strain that enables the matrix to develop cracks parallel to the fibre direction, and then eventually weepage occurred as a result of the cumulative matrix microcracking as shown in figure 10 a. However, helical delimitations were noted followed fibre direction in condition of pure axial loading with a maximum hoop stress of 52 MPa, figure 10 b. This could be occurred as a result of high stress applied in axial direction in order to observe the failure; since there is no pressure applied to the pipe (no leakage was expected). Pipes under hoop dominated loading caused high strain in hoop direction and negative axial strains. This possibly affected the whitening that monitored under pure hoop loading with maximum hoop stress of 300 MPa. The negative axial strain might prevent any visible microcracking to develop parallel to fibres and then that could led to rupture. The rupture was due to a massive buckling occurred to the pipe, which could initiate fibres to fail near to the end fitting where possibly stresses concentrated at this particular area figure 10 c.
a
b
c
Figure 10 . Shows the failure mode for the pipes under, (a) 2:1 Hoop/Axial stress, (b) pure axial loading, (c) pure hoop loading
Conclusion
A biaxial environmental testing rig has been constructed; capable of independently loading 200 mm diameter, 2 m pipe's length, enabling to apply loads ranging from pure axial to pure hoop stress, at room and at elevated temperature up to 95C°.
UEWS test have been conducted for these pipes at room temperature, and at 65C°, with a six different ratios of hoop/ axial stress.
The results from the UEWS tests were presented in the form of failure envelopes with the effects of testing at an elevated temperature. It was found that the strength of the epoxy pipes reduced at elevated temperature, (apart from 2:1 hoop/axial stress), where the strength at 65C° was 5% higher than that at room temperature which that possibly refer to residual stresses.
The decline of elastic modulus in hoop and axial direction has been discussed. A significant decline were noted at axial load dominated, whereas, slightly changes in wall stiffness occurred under hoop loading.
The failure modes were related to load conditions and strain behaviour. For positive strains the failure occurred due to intensive matrix microcracking which leads to weepage. However, violent failure was observed in larger hoop strains which could be as a result of delamination and fibre breakage.
