Printer Friendly

The effects of water and frequency on fatigue crack growth rate in modified and unmodified polyvinyl chloride.

INTRODUCTION

The success of plastic pipes in various applications such as water and gas distribution and sewer and irrigation systems has led to increasing demand for these materials. In applications which undergo stress fluctuations throughout service lifetime such as in pumped pipes, fatigue resistance is one of the key factors for consideration in design. Stress fluctuations have a potential to propagate cracks and induce premature fatigue failure. Knowledge of the fatigue threshold associated is another factor which needs to be analyzed along with crack growth rates.

Typically, fatigue lifetime or S-N curves (where S is the stress amplitude and N is the number of cycles to failure) show considerable scatter due to the inability of this test to separate fatigue crack initiation and growth, nor consider variations in initial defect size. This leads to challenges in accurately predicting operational lifetimes or the presence, or otherwise, of a fatigue threshold value. As an alternative, fracture mechanics concepts which refer to Paris's Law may be applied. This law defines crack growth to occur above a threshold stress state intensity factor range [DELTA]K = [DELTA][sigma]Y[square root of a], where [DELTA][sigma] is applied stress amplitude, a the defect size, and Y a crack geometry factor with crack growth per cycle (N)] and to following an empirically determined power-law relationship da/dN = A[DELTA][K.sup.m], where A and m are constants (1).

High strength and durability, low weight and cost-effectiveness are advantages offered by polyvinyl chloride (PVC-U) and these features have made it one of the most preferred materials for plastic pipes. However, the material is prone to brittle failure which limits its application potential. This problem has inspired many researchers to modify the current PVC systems to achieve greater ductility and toughness. The modification is performed by adding impact modifiers into PVC while maintaining the same base composition. Chlorinated polyethylene (CPE) is generally used since it has good compatibility with PVC. This product, PVC-M, has shown excellent service performance, similar to the more highly modified PVC which has been used for more than 30 years by mining industries in South Africa. In the current work, the specimen samples contain 6 pphr of CPE.

For thermoplastic materials, the presence of reactive liquids or gases in the environment may stimulate a condition that degrades their mechanical properties and leads to unpredictable brittle failures, as observed in previous studies (2), (3). This phenomenon is known as environmental stress cracking. Stress cracking is severely enhanced in the presence of environmental factors, whereas for polymers the required stress to initiate crazes may be lowered below that observed in ambient air. A study by Breen et al. (3), (4) found that the degradation of the mechanical properties of PVC and PVC-CPE in benzene vapor was associated with lower values of crazing stress. Kefalas and Argon (5) discovered that water is a mild crazing agent, even in hydrophobic polymers such as polystyrene.

The strong resistance of PVC-U to chemical environments such as acids and alkalis is well known. However, the effect of water environments have been largely overlooked and to date, little attention has been given to the possible effects of absorbed water on the fatigue properties of pipe materials, particularly with impact modifier, even though during service the pipes would have direct contact with water. Kim et al. (6) obtained better fatigue resistance of PVC-U in water at low stress ratios and associated these results with the crack closure concept. On the other hand, Maddox and Manteghi (7) in 1992 found a similar crack propagation rate for PVC-U pressure pipe in both air and water.

The objectives of the present work were to determine the effect of water immersion on the fatigue threshold and crack growth rates of PVC with CPE (PVC-M) and without CPE (PVC-U), and the effect of frequency of cyclic loading in comparison to an air environment. In addition, the investigation of water absorption on the fracture surface morphology was carried out by employing Field Emission Scanning Electron Microscopy (FESEM) and optical microscopy (OM). These techniques also assisted in revealing the mechanisms that occurred during fatigue crack propagation.

EXPERIMENTAL PROCEDURE

Materials

PVC-U and PVC-M materials were supplied by IPLEX Pipelines Australia Pty Ltd. Both pipes were fabricated in an industrial extruder using the same formulation, apart from the addition of chlorinated polyethylene (CPE) in PVC-M. The CPE is an impact modifier with a chlorine content of 36% by mass. In 100 parts by weight of PVC resin, the proportion of other additives is: 1.5 parts of rutile (titanium dioxide), 1.5 parts of calcium carbonate, three parts of lead based thermal stabilizer, and 6 parts of CPE. The K value of the resin was 67. The PVC-U and PVC-M compounds comply with the performance requirements of AS/NZS 1477 (8) and AS/NZS 4765 (9), respectively. Pipes with an outside diameter of 122 mm and wall thickness of approximately 9 mm for PVC-M and 8 mm PVC-U pipes were cut in a longitudinal direction and subsequently heated in an oven at 120[degrees]C prior to flattening between supporting plattens. The specimens were then cooled slowly to room temperature. Compact tension specimens with a width of 75 mm (W) were cut using a band saw. The specimens were notched with a band saw and then a sharp blade was slid into the root of the notch to form a sharp tip.

Fatigue Testing

An MTS closed-loop servohydraulic testing machine was used to perform the fatigue tests under tension-tension loading mode with a load ratio of R = 0.1. The fatigue crack propagation behavior of specimens was evaluated at different frequencies (1 and 7 Hz) and subjected to cyclic loading using a sinusoidal waveform. During testing, the sample was submerged in a transparent water tank containing tap water to stimulate actual service conditions and the water temperature was maintained between 21 and 23[degrees]C. The water pH was measured to be 7 prior to the testing. Crack lengths were measured using a compliance method which had been previously calibrated in air.

Microscopy

The fatigue fracture surface morphology was examined in a Hitachi 4500II, field emission scanning electron microscope (FE-SEM). All the specimens were coated with a chromium layer prior to observation and the instrument was operated at an accelerating voltage of 20 kV.

An optical microscopy technique was utilized to investigate the fatigue damage zone at the crack tip of the compact tension specimens. The thin sections (about 16 [micro]m) were obtained by slicing the specimens with a Leica, RM 2245 microtome before viewing in transmission mode under a Nikon Eclipse ME600 microscope. Both bright-field and cross-polarized conditions were employed.

Infrared transmission spectra were obtained from the fracture surfaces using a Nicolet Avatar 360 Fourier transform infrared (FTIR) spectrometer.

RESULTS

Fatigue Crack Propagation Behavior

Analysis of the fatigue crack propagation in water was undertaken as a function of stress intensity factor amplitude ([DELTA]K) and the crack growth rate (da/dN) as described by the Paris' law. The results for PVC-M and PVC-U at frequencies of 1 and 7 Hz in water arc presented in Fig. 1a-d and compared with results in air from previous work (10). It can be observed that the crack propagation rate was lower in water than in air at higher [DELTA]K and similar at lower [DELTA]K. The difference at high [DELTA]K was more marked at 1 Hz than at 7 Hz. The divergence occurs at higher [DELTA]K in air (0.30 MPa*[m.sup.[1/2]]) than in water (0.28 MPa*[m.sup.[1/2]]).

[FIGURE 1 OMITTED]

A knowledge of the fatigue threshold is crucial for design applications such as in pressure pipes. Table 1 shows fatigue threshold ([[DELTA]K.sub.th]) values for the samples tested in water and a comparison with those tested in air. It can be seen that the fatigue threshold stress intensity factor amplitude, [[DELTA]K.sub.th], generally increased with increasing frequency but less markedly in water than in air, and in PVC-U there is no frequency effect. Furthermore, the fatigue threshold of PVC-U at 1 Hz is higher than that of PVC-M. In a previous study in air, fatigue thresholds of PVC-M were lower than PVC-U at frequencies ranging from 1 to 20 Hz (10).
TABLE 1. The fatigue threshold values of PVC-U and PVC-M in water and
air medium.

                                Fatigue threshold
                               (MPa*[m.sup.[1/2]])

Material    Frequency (Hz)    In air (10)    In water

PVC-U            1               0.22          0.23
                 7               0.24          0.23

PVC-M            1               0.19          0.22
                 7               0.22          0.24


A comparison of frequency effect on the fatigue crack propagation rate of PVC-U and PVC-M in water is shown in Fig. 2. Similar to the previous work of Samat et al. (10), higher frequencies are shown to result in lower crack propagation rates. However, as observed in Fig. 2, a decrease in crack growth rate at 7 Hz occurred in low [DELTA]K regions; from 0.2 to 0.6 MPa*[m.sup.[1/2]]. As [DELTA]K increases (> 0.6 MPa*[m.sup.[1/2]]), the crack growth rates at 7 Hz in both specimens increased rapidly and at a very high [DELTA]K level the differences in the crack growth rates at frequencies of 7 and 1 Hz are distinguishable. In other words, the effect of frequency diminishes at higher stress intensity factor amplitudes ([DELTA]K).

[FIGURE 2 OMITTED]

Crack-Tip Profile Observations

Transmission optical microscopy was used to elucidate the deformation process occurring at the crack tip during cyclic fatigue growth in water. The process zone close to the crack tip was viewed under bright and cross-polarized light. All specimens in water exhibited a stress-whitening zone at the crack tip, indicative of crazing (see Fig. 3). For PVC-U, the plastic zone appearance was less pronounced than PVC-M as was observed in an earlier work undertaken in air.

[FIGURE 3 OMITTED]

Comparable with observations in air, no obvious occurrences of shear banding were seen on the cross-polarized TOM micrographs; as indicated by the absence of a birefringence zone during examination under a cross-polarized light. Therefore, it appears that the energy dissipation occurred only through a crazing mechanism. This observation emphasizes that the difference in crack growth rate in water versus an air environment is associated with the changes in the micromechanisms but not in the main mechanism (craze or shear bands).

Fracture Surface Inspections

Low [DELTA]K. When observed under an optical microscope the fatigue fracture surfaces of both PVC-U and PVC-M exhibited similar features to the specimens tested in air. These features include the absence of discontinuous growth bands or striations and an irregular patch-structure (11), (12). However, when the same fracture surfaces were investigated different surface morphologies were observed at near the threshold, [DELTA]K = 0.26 MPa*[m.sup.[1/2]] (see Fig. 4), and slightly above, [DELTA]K = 0.33 MPa*[m.sup.[1/2]] (see Fig. 5).

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

For PVC-U at low magnification, the level of surface deformation is considered low because the deformed surface was relatively smooth and only small indications of plastic deformation are observed. The discontinuous edges were terminated with large remnants of broken fibrils while the smooth regions were found to consist of short microfibrils which adhere to the matrix. Interestingly, at higher magnification (Fig. 4b), nodular structures are seen at the fibril-ends. These nodules have a partially round shape. The fracture surfaces of PVC-M show a slight difference when compared to PVC-U. As demonstrated in Fig. 4d, the deformation on the fracture surfaces of PVC-M was more homogenous with no evidence of the smooth regions (as seen in PVC-U) observed. An examination of PVC-M at higher magnification (Fig. 4e) showed slightly more elongated microfibrils compared with PVC-U. The microfibrils were also found to differ in size; being both large and fine. Microvoids also formed between the fibrils and on yielded fragments. Nodular structures were present on the fracture surface of PVC-M, similar to PVC-U.

When the frequency was increased to 7 Hz, at low stress intensity factor ([DELTA]K = 0.26 MPa*[m.sup.[1/2]]), the fracture surface was characterized by a nodular structure continuously seen on the surface in both samples (Figs. 4c and f). For PVC-U, at a frequency of 7 Hz (Fig. 4c), the amount of nodular structure is greater when compared to the fracture surface at 1 Hz at the same stress intensity factor amplitude (Fig. 4b). On the other hand, greater formation of microfibrils was observed in PVC-M (Figs. 4e and f). Hence, the change in the fracture surface characteristics corresponds to the higher fatigue threshold values in water than in air (see Fig. 2).

At a slightly higher stress intensity factor, [DELTA]K = 0.33 MPa*[m.sup.[1/2]] at 1 Hz (Figs. 5a-b), the level of deformation in PVC-U increased and simultaneously the formation of highly packed nodules was observed which became coagulated and fused to each other. In PVC-M, (Figs. 5d-e) the microfibrils were shorter and the nodules formed in clusters. A higher concentration of microvoids at the respective [DELTA]K could be due to the increased formation of thin microfibrils. Nevertheless, the fracture surface characteristics for PVC-U and PVC-M at a high frequency (7 Hz) for the same [DELTA]K level (0.33 MPa*[m.sup.[1/2]]) are unchanged; a cluster of nodule and shorter fibrils (Figs. 5c and f).

High [DELTA]K. From Figs, 1a and c, the effect of testing medium on the fatigue resistance in both specimens is greater at 1 Hz and this difference is more obvious at higher [DELTA]K levels. Therefore, the fracture surface morphologies in water and air at 1 Hz are compared in order to determine the cause of this difference. In this instance, the comparison was made at [DELTA]K = 0.7 MPa*[m.sup.[1/2]]. As depicted in Fig. 6, plasticized structures were observed on the fracture surfaces of specimens tested in water. From this structure, the fibril features were still discernible in PVC-U and PVC-M. In contrast, in air, the existence of fibrils was no longer evident on the fracture surface of both specimens. Therefore, the reason for the fatigue resistance improvement in water at. higher [DELTA]K level is believed to be attributed to the presence of fibrils on the plasticized structures. However, the advantage of the plasticized structure is only temporary, since these structures would disappear with further increase in stress amplitude values ([DELTA]K = 0.9 MPa*[m.sup.[1/2]]) and correspondingly, flat structures are visible in the microstructure (see Fig. 7).

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

Fracture surface examination at higher [DELTA]K levels was conducted for a frequency of 7 Hz for PVC-U at a [DELTA]K = 0.84 MP[a.m.sup.[1/2]]. Plasticized structures also exist at this frequency, similar to the observations at 1 Hz. Although the presence of fibrils is evident (see Fig. 8), the PVC matrix was severely deformed. At higher magnifications striation-like features were noticed in the deformed area, as indicated by arrows in Fig. 8b. Hence, the occurrence of severe plastic deformation might be the reasons for a decline in the fatigue resistance at 7 Hz once certain [DELTA]K levels were attained, i.e., [DELTA]K > 0.6 MP[a.m.sup.[1/2]] (see Fig. 2).

[FIGURE 8 OMITTED]

In this work, we found that the testing environment affects the fracture surface characteristics. Generally, fracture surfaces in water are characterized by the presence of short microfibrils (crazes), nodules, and plasticized structures. These characteristics are dissimilar to those observed in our previous study conducted in air (10) except for the craze structures. In most cases, the craze structures (which consist of microfibrils and voids) are dominant in both specimens, particularly at low stress intensity factor amplitudes. For comparison, the fracture surfaces of PVC-U and PVC-M in air at 1 Hz are shown in Fig. 9.

[FIGURE 9 OMITTED]

FTIR Spectrum. Figure 10 shows FTIR spectra indicating the presence of water molecules on the fracture surface of PVC-U and PVC-M in air and water environments at frequencies in the range 3000-3800 [cm.sup.-1]. The two main characteristic features of the water spectra observed in Fig. 10 are: (1) a broad peak band between 3100 and 3600 [cm.sup.-1] and (2) a relatively sharp peak at 3735 [cm.sup.-1]. However, the samples that have been tested in air exhibited different FTIR spectra to those samples that tested in water. Different spectra characteristics within the same frequency ranges were observed either in PVC-U or PVC-M. This analysis indicates that water molecules have been absorbed on the fracture surface of samples during fatigue testing in the water environment.

[FIGURE 10 OMITTED]

DISCUSSION

The Effect of Water Environment

Regardless of the testing environment, both materials showed fatigue crack growth rate sensitivity to the frequency of cycling as shown in Figs. 1a-d. This behavior is in accordance with many previous studies (13), (14) though these studies were executed in air. Some researchers (14), (15) associate this behavior with the hysteretic heating that occurred during fatigue testing. Conversely, Sauer and Richardson (16) have suggested that higher frequencies cause higher strain rates and subsequently lead to an increase in modulus and yield strength. Hence, the first suggestion may be applicable to explain the frequency sensitivity behavior in air whilst the second hypothesis is preferred to explicate behavior in water, where heat dissipation rates are greater and hence temperature rises slower than in air.

The results of our experiment on fatigue crack growth rate and analysis of FESEM and optical micrographs clearly indicate the effect of the testing medium. It is likely that the nodular structures are formed as a consequence of water molecules being absorbed onto the PVC matrix. To date, previous studies have not reported these nodular structures in PVC in a water environment. Most investigations of environmental conditions reported that this testing condition has caused deleterious rather than beneficial effects on tensile and fatigue properties. The deleterious effects are mainly associated with a faster craze formation (4) and a decline in craze lifetime (17). In contrast, blunting the flaws has given a beneficial effect (2). Presumably, therefore, the improvement in fatigue resistance and changes in fracture surface characteristics in this work were caused by the crack blunting effects.

The Effect of High Frequency

Unlike observations in the case of testing in air, for testing in water at certain high stress amplitude values, [DELTA]K, results in the plasticization of the PVC matrix and formation of a plasticized structure. The presence of these structures contributes to a lower fatigue crack growth rate in water at the [DELTA]K ranges used. As the stress amplitude increases to higher values, the fracture surface transforms again from a plasticized structure into a flat structure with a brittle fracture surface.

However, the transformation of the plasticized structures to a flat structure is accelerated at higher frequency, which is associated with the degradation in fatigue resistance properties. As shown in Fig. 2, the crack propagation rate at 7 Hz increased rapidly above [DELTA]K = 0.6 MP[a.m.sup.[1/2]]. Fracture surface examination at these [DELTA]K ranges revealed that the plasticized structures would actually first undergo severe plastic deformations before transforming into a flat structure. Hoa et al. (18), (19) also made a similar observation associated with fast crack propagation rates in samples toward the end of their fatigue life. Furthermore, the plasticization would decrease the glass transition temperature ([T.sub.g]) of craze fibrils which would then lead to craze fibril rupture (20). Therefore, it can be concluded that a high frequency can severely decrease the fatigue resistance; particularly at higher [DELTA]K.

Mechanism

A schematic for the fatigue fracture/toughening mechanism in air and water environments is illustrated in Fig. 11. In most cases, for testing in air fatigue fracture occurs after the breakdown of microfibrils and coalescence of voids (crazes). The fatigue fracture mechanisms of PVC-U and PVC-M in air were described in our earlier work (10).

[FIGURE 11 OMITTED]

For testing in water, the following fatigue fracture mechanisms are suggested. When the load is applied cyclically, the craze structures will initiate at the crack tip. As the testing environment is water, it is expected that the water molecules may diffuse into the craze structures. The absorption of water may retard the microfibril (craze) extension and simultaneously cause swelling of some parts of fibrils and the matrix surface in the craze region. As a result, this phenomenon will produce short microfibrils and nodules on the fracture surface. Owing to the fact that the crazes which bridged the crack are shortened, the magnitude of crack tip opening for the specimen in water is smaller than that for the specimen in air. Consequently, after the microfibrils fail, the crack would advance a short distance which is associated with the decreased crack propagation rate. Further evidence of these effects was seen with a higher fatigue threshold observed in water for both PVC-M and PVC-U than in air. The diffusion of water is also evidenced by the formation of nodules at the fibril ends, even after the fibrils are separated.

On further loading, the time of exposure of the crack surface to an aqueous condition would increase thereby increasing [DELTA]K values which correspond to a wider crack mouth. Consequently, more water molecules would diffuse into the matrix. As the fraction of water that is absorbed is relatively high, the high-level of [DELTA]K at these stages would induce the plasticization of PVC matrix and transformation of the appearance of the fracture surfaces, as noted here. Accordingly, the existence of fibrils in water (Figs. 6(b,d)) is still observed even though the stress intensity factor was high ([DELTA]K = 0.7 MPa * [m.sup.[1/2]]). On the other hand, from Figs. 6(a,c), it is seen that the specimens tested in air had shown different fracture surface characteristics for the same [DELTA]K level (0.7 MPa * [m.sup.[1/2]) and flat structures were formed. This observation indicates that the plasticization phenomenon only occurred in the specimens tested in water. Furthermore, when the [DELTA]K becomes high, the deformation in the plasticized structures is greater and is associated with the presence of the striation-like features on PVC matrix (Fig. 8b). Finally, the specimens will undergo fracture and brittle fracture characteristics such as flat structures are seen on the fracture surface.

It is well known that PVC is a hydrophobic material due to the presence of chlorine atoms (CI) and [CH.sup.-] groups (21). However, at saturation, PVC still has a capability to adsorb water at lower amounts than hydrophilic polymers (22). van Oss (23) stated that if polar moleculesor particles are immersed in water, a hydrophobic attraction exists between them. The driving force of this attraction is the hydrogen-bonding (Lewis AB) free energy of cohesion of the water molecules that surround these molecules or particles. A study by Yarwood et al. (24), found evidence of water molecules absorbed onto PVC thin films from FTIR ATR spectra. Since PVC is an amorphous polymer, the water molecules can enter the amorphous parts and bind to the functional group (chloride) via hydrogen bonding (25).

Furthermore, Vincent and Raha (26) mentioned that hydrogen bonds in PVC are associated with the hydrogen atoms next to the chlorine atoms, such that these hydrogen atoms act as proton donors. They also suggested that the polymers could interact with a liquid containing complementary proton donating or accepting molecules. Water is such a substance with that unique chemical property; it can behave both as an acid (a proton donor) and a base (a proton acceptor). Therefore, in the present work, it is believed that the interaction between water molecules and PVC has occurred because the water acts as a proton acceptor when interacting with PVC. Consequently, this supports our initial assumption that the diffusion of water molecules causes changes to the fracture behavior.

FTIR analysis (see Fig. 10) has confirmed the absorption of water molecules on the fracture surface of PVC-U and PVC-M samples. In this analysis, we have compared our results to the spectra from a previous study which involved bulk water and hydrophobic materials (27). Their result shows the absorption of water molecules onto hydrophobic polymers. We also compared our results with the FTIR spectra of water clusters at low and high concentrations. In this case, the sharp absorption peak was seen near 3710 [cm.sup.-1] (28). However, Page et al. (29) have reported an absorption peak at 3730 [cm.sup.-1] at low water concentrations which is close to the peak seen in our study. Thus, these comparisons support our result that the absorption of water molecules into the PVC matrix during the testing is responsible for differences in fracture surface morphologies and fatigue resistance behavior.

On the basis of the transformation of the fracture surface morphology from the threshold region to the upper [DELTA]K region, it appears that the water absorption rate is dissimilar throughout the testing. Sauer and Smith (22) reported that the fractional amount of water absorbed onto PMMA was proportional to [t.sup.[1/2]] in the initial period, where t refers to the immersion time. It is expected that the rate of water absorption onto craze or PVC matrix could also be relatively proportional to [t.sup.[1/2]] mainly at the threshold region. As mentioned earlier, the observation from FESEM micrographs showed that the absorption of water had started at a region close to [DELTA][K.sub.th]. Initially, the absorption of water molecules during this period had only caused swelling of the microfibrils through the formation of nodular structures but was insufficient to plasticize the PVC matrix. However, the plasticization of PVC occurs at high [DELTA]K ranges, in which the water absorption at these stages is presumably high.

The Effect of CPE

As demonstrated in Fig. 6, the difference in fracture surface features between PVC-M and PVC-U in both testing media is related to the existence of CPE particles. Comparable with observations in air (Figs. 9b and d), the microfibrils of PVC-M in water were also elongated slightly as compared with the microfibrils of PVC-U (see Fig. 4); moreover, the amount of nodular structures was smaller. Nevertheless, the microfibrils in PVC-M in water are shorter than those in air which is a consequence of the interaction between PVC matrix and water molecules impeding the elongation of the microfibrils. However, as the frequency is increased a higher density of short microfibrils is formed (Fig. 4f); in which these fibrils are able to sustain higher loads. Therefore, PVC-M sample has shown a better fatigue resistance when compared with PVC-U in low [DELTA]K region at a frequency of 7 Hz.

The Cooling Effect

Some authors have suggested that the decrease in fatigue crack propagation rates was due to cooling effects from the surrounding water (30). Warty et al. (2) studied the fatigue of polystyrene in alcohol but found no effect of cooling on fatigue crack growth rates. Furthermore, in the present work, a significant improvement of fatigue resistance in water as compared to in air was obtained at 1 Hz. At this particular frequency, the increase in temperature is considered insignificant even if the tests were done in air (13), (31). Dissimilar fracture surface morphology in air and water medium is given as evidence the water affects the fracture micromechanism leading to a reduction in the crack propagation rate via blunting effects and not by the cooling effects.

Comparison With an In-Service Fatigue Failure of a PVC-U Water Pipe

Interestingly the FESEM micrographs in this work have exhibited a similar fracture surface morphology to a PVC-U pipe that had undergone cyclic pressure testing. A pipe with an outside diameter of 177.7 mm and a 6.81 mm wall thickness was tested at a temperature of 20[degrees]C at a frequency of 0.3 Hz. The internal pressure was applied by varying the water pressure between 142 psi (0.979 MPa) and 203 psi (1.40 MPa). Failure occurred after 5.96 million cycles was reached. However, the failure cycles are greater than predicted. Inspection of the fracture surface by FESEM found that the crack propagated outward radially from the inner wall to the outer pipe wall (see Fig. 12), as also observed by Joseph and Leevers (32).

[FIGURE 11 OMITTED]

Figure 12 shows micrographs that were taken at four different points on the fraction surface; from the slow to fast crack propagation rate regions. The stress intensity factor (K) for each semi-elliptical was calculated (33) as a function of crack and pipe thickness:

K = [F.sub.e]([a/b], [a/t]) [sigma][([pi]a).sup.[1/2]] (1)

where [F.sub.e] is the correction factor, [sigma] is the hoop stress, t is the pipe thickness, a is the crack depth, and b is the crack width.

The circumferential stress or hoop stress ([DELTA][sigma]) was calculated from internal pressure, as for a thin-walled tube

[DELTA][sigma] = [DELTA]p [(D - 2t)/2t] (2)

where P is the pressure, D is the pipe diameter, and t is the pipe thickness

Point 1 corresponds to a low rate of crack propagation, with a stress intensity factor of 0.22 MP[a.m.sup.[1/2]]. Here, the morphology of nodular structures can be easily seen. Although the nodules were coagulated with each other, their circular shapes were apparent. As the stress intensity factor increased from 0.22 to 0.37 MP[a.m.sup.[1/2]], the crack propagates to the second point region. In this region, the fracture surface has started to form plasticized structures but the nodule structures were still discernible. The observation in the third semielliptical region showed that the fracture surface has completely transformed into a plasticized structure and the stress intensity factor was estimated to be 0.56 MP[a.m.sup.[1/2]]. A severe plasticization of PVC-U matrix is noticed with the formation of the fibrils pointing upward. Eventually, a brittle fracture appearance without any fibrils was evident in the final semielliptical region ([DELTA]K = 0.78 MP[a.m.sup.[1/2]]). These observations show that the fracture surface characteristics in the current study are similar to that observed in-service--nodules structures, plasticized structures, and flat structures, and their formations are dependent on the stress intensity factor value and the fraction of water absorbed. Moreover, from this analysis, we can conclude that water environments affect the fatigue crack propagation rates and the fracture surface characteristics.

CONCLUSIONS

Cyclic fatigue crack growth measurements were undertaken on modified and unmodified PVC in water environments and have revealed that

1. A water environment increases the fatigue resistance of PVC pipes, more significantly for PVC-U at higher stress intensity factor amplitudes ([DELTA]K) and lower frequencies.

2. The fatigue resistance of PVC-M is generally similar in air and water, except under certain testing conditions.

3. Both materials showed a similar behavior at higher frequencies with the enhancement of fatigue resistance and thresholds and these results are more evident in water environment.

4. The fatigue fracture in water is considered to be sensitive to frequency and the diffusion rate of water absorption, being lower in PVC-M than PVC-U.

5. Nodular structures form at the ends of fibrils on the fracture surface, most evidently at lower [DELTA]K.

ACKNOWLEDGMENTS

The authors thank the International Islamic University Malaysia (IIUM), Ministry of Higher Education, Malaysia, and IPLEX Pipelines for the support of this research.

REFERENCES

(1.) P.C. Paris, R.J. Bucci. E.T. Wessel, W.G. Clark Jr., and T.R. Mager, "Extensive Study of Low Fatigue Crack Growth Rates in A533 and A508 Steels," in Stress Analysis and Growth of Crack: Proceedings of the 1971 National Symposium on Fracture Mechanics, Part I, ASTM STP 513, Am. Soc. for Testing and Materials, West Conshohocken, PA, 141 (1972).

(2.) S. Warty, D.R. Morrow, and J.A. Sauer, Polymer, 19, 1465 (1978).

(3.) J. Breen and D.J. Van Dijk, J. Mater. Sci., 26, 5212 (1991).

(4.) J. Breen, J. Mater. Sci., 28, 3769 (1993).

(5.) V.A. Kefalas and A.S. Argon, J. Mater. Sci., 23, 253 (1988).

(6.) H.S. Kim, Y.W. Mai, and B. Cotterell, "Effect of Water on Fatigue Crack Growth Rate in uPVC," in Fracture Mechanics in Engineering Practice: Proceeding (Melbourne University), 54 (1988).

(7.) S.J. Maddox and S. Manteghi, Plast. Rubber. Compos. Appl., 17, 5 (1992).

(8.) Australia/New Zealand Standard AS/NZS 1477, Poly Vinyl Chloride (PVC) Pipes and Fittings for Pressure Applications, Standards Australia International, Sydney (2000).

(9.) Australia/New Zealand Standard AS/NZS 4765 (Int), Modified PVC (PVC-M) Pipes for Pressure Applications, Standards Australia International, Sydney (2000).

(10.) N. Samat, R. Burford. A. Whittle, and M. Hoffman, Polym. Eng. Sci., 49, 1299 (2009).

(11.) H.S. Kim and X.M. Wang, J. Mater. Sci., 29, 3209 (1994).

(12.) Y.W. Mai and P.R. Kerr, J. Vinyl. Tech., 7, 130 (1985).

(13.) R.W. Hertzberg, J.A. Manson, and M. Skibo, Polym. Eng. Sci., 15, 252 (1975).

(14.) G. Pinter, M. Haager, W. Balika, and R.W. Lang, Plast. Rubber Compos., 34, 25 (2005).

(15.) N. Merah, F. Saghir, Z. Khan, and A. Bazoune, Eng. Fract. Mech., 72, 1691 (2005).

(16.) J.A. Sauer and G.C. Richardson, Int. J. Fract., 16, 499 (1980).

(17.) L. Josserand, R. Schirrer, and P. Davies, J. Mater. Sci., 30, 1772 (1995).

(18.) S.V. Hoa, A.D. Ngo, and T.S. Sankar, Polym. Compos., 2, 162 (1981).

(19.) S.V. Hoa, A.D. Ngo, and T.S. Sankar, Polym. Compos., 3, 44 (1982).

(20.) E.J. Kramer, "Environmental Cracking of Polymers," in Development in Polymer Fracture-1, E.H. Andrews, Ed., Applied Science London Ltd, London, 55 (1979).

(21.) A.V. Tvardovskiy, "Sorbenl Deformation," in Interface Science and Technology Series, A. Bubbard, Ed., Elsevier Ltd, The Netherlands, 13 (2007).

(22.) J.A. Sauer and L.S.A. Smith, "Water Absorption in Glassy Polymers and its Effects," in Polymers in a Marine Environment, Trans ImarE(C): Proceeding, Marine Management (Holdings), London, 97, 95 (1985).

(23.) C.J. van Oss, Interfacial Forces in Aqueous Media, Taylor & Francis Group, United States (2006).

(24.) J. Yarwood, C. Sammon, C. Mura, and M. Pereira, J. Mol. Liq.,80, 93 (1999).

(25.) D. Balkose, F. Ozkan, S. Ulutan, and S. Ulku, J. Therm. Anal. Calorim., 71, 89 (2003).

(26.) P.I. Vincent and S. Raha, Polymer, 13, 283 (1972).

(27.) H. Kunagawa and S. Yukawa, Polymer, 35(26), 5637 (1994).

(28.) D.F. Coker, R.E. Miller, and R.O. Watts, J. Chem. Phys., 82(8), 3554(1985).

(29.) R.H. Page, J.G. Frey, Y.R. Shen, and Y.T. Lee, Chem. Phys. Lett., 106, 373 (1984).

(30.) C.C. Chen, J. Shen, and J.A. Sauer. Polymer, 26, 89 (1985).

(31.) W.M. Cheng, G.A. Miller, J.A. Manson, R.W. Hertzberg, and L.H. Sperling, J. Mater. Sci., 25, 1924 (1990).

(32.) S.H. Joseph and P.S. Leevers. J. Mater. Sci., 20, 237 (1985).

(33.) D.P. Rooke and D.J. Cartwright, Compendium of Stress Intensity Factors, Her Majesty's Stationery Office, London (1974).

Noorasikin Samat, (1), (2) Robert Burford, (3) Alan Whittle, (4) Mark Hoffman (2)

(1) School of Materials Science and Engineering, The University of New South Wales, New South Wales, 2052 Australia

(2) Department of Manufacturing and Materials, International Islamic University Malaysia, 53100 Gombak, Selangor, Malaysia

(3) School of Chemical Sciences and Engineering, The University of New South Wales, New South Wales, 2052 Australia

(4) Iplex Pipelines Australia Pty Ltd, 35 Alfred Road, Chipping Norton, New South Wales, 2170 Australia

Correspondence to: Noorasikin Samat; e-mail: z3145562@student.unsw.edu.au

DOI 10.1002/pen.21547
COPYRIGHT 2010 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Samat, Noorasikin; Burford, Robert; Whittle, Alan; Hoffman, Mark
Publication:Polymer Engineering and Science
Article Type:Report
Geographic Code:8AUST
Date:Feb 1, 2010
Words:5982
Previous Article:Rheology and melt strength of long chain branching polypropylene prepared by reactive extrusion with various peroxides.
Next Article:Modeling the phase behavior in binary mixtures involving blowing agents and thermoplastic resins.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters