Printer Friendly

The effects of frequency on fatigue threshold and crack propagation rate in modified and unmodified polyvinyl chloride.

INTRODUCTION

The increased application of pipes in gas and water supply systems has had a significant impact on materials development. Initially, metals were mainly used but with time, iron pipes have been replaced by polymers. Major benefits are derived from using plastics due to their excellent resistance to corrosion and chemicals such as acids and alkalis. Currently, unplasticized polyvinyl chloride (PVC-U) is one of the most widely used polymers owing to properties such as high durability, strength, and low cost. In spite of significant economic benefits, its acceptance in both static and dynamics applications has been influenced by the fracture characteristics of this material and the consequent need to apply comparatively high design factors. Under certain static loads or pressure transients and temperatures, cracks initiate and propagate, ultimately leading to pipe failure. In the 1970s, problems associated with the installation of piping systems were identified to cause service failures of PVC pressure pipes in the UK (1). As a result of these incidents, pipe quality was improved and a strict regulation of design and installation practices was imposed. Design guidelines were developed to address pipes used in applications involving cyclic mechanical loads.

One common technique that is currently being employed to enhance the toughness of PVC-U, and permit the use of higher design stresses, is to incorporate rubbery particles called impact modifiers into the matrix. These particles are capable of increasing the ductility of the PVC matrix by lowering the susceptibility to ductile-brittle transition (2). Nevertheless, the tensile yield stress is known to deteriorate with increase in the impact modifier content. PVC containing the impact modifier is known as modified-PVC (PVC-M) or alloyed-PVC (PVC-A). Chlorinated polyethylene (CPE) is one of the impact modifiers that has been successfully employed in PVC materials, because it gives the best balance of properties while simultaneously facilitating processing (3).

PVC and CPE form a microheterogeneous system in CPE-PVC blends and are thus only partially compatible. However, this structure is responsible for the observed higher toughness in modified-PVC. Two morphologies can be produced in PVC-CPE blends: a dispersion of discrete CPE particles and an interpenetrating network (IPN) structure which envelopes the primary PVC particles. Whittle (4), in a study of a system from the same source as this present study, found that the formation of the structure--IPN or discrete CPE particles or a mixture of both--is dependent on the processing conditions, barrel zone temperature, and screw speed of the extrusion or injection molding machine. Since the IPN structure reduces the yield strength to a greater extent, the fabrication of pipes with discrete structures has been suggested.

Nevertheless, no clear correlation between the impact-strength and fatigue behavior has been obtained from the modified-polymers. Previous studies have reported that such an improvement in toughness does not necessarily result in better fatigue resistance. Three different outcomes are reported by incorporating rubber particles into polymers: (i) improvements in both impact strength and fatigue behavior (5), (ii) improvement in only the impact strength but not in fatigue (4) and (iii) improvement in fatigue resistance but not impact strength (6). Many factors have been proposed in explaining the above results; however, the most accepted reason is that there is a distinction between the mechanisms of impact and fatigue fracture. Crazing and shear banding are the primary mechanisms behind fracture of polymers. Cavitation and particle-bridging phenomena have been observed in a rubber-polymer system (7). The optimization of toughening mechanisms depends on parameters such as carefully controlling the matrix material and size and concentration of the rubbery particles (8).

Numerous previous studies on PVC-CPE blends have focused primarily on the mechanical properties of the material, i.e., impact properties and processing parameters (9), (10). Little research has been undertaken on the fatigue properties of these materials.

Whittle reported that no improvement in PVC-M relative to PVC-U occurred under fatigue lifetime testing (4). The current article seeks to elucidate this finding by studying the effect of various frequencies of cyclic loading on the threshold value and fatigue crack growth behavior in modified- and unmodified-PVC. The relevant mechanisms that are associated with fatigue behavior are also studied to understand how the crack propagates, by giving emphasis to the process zone area at the crack tip. Finally, the crack propagation mechanism in PVC-M is compared to the PVC-U tested under similar conditions.

EXPERIMENTAL

Materials

PVC-U and PVC-M were supplied by IPLEX Pipelines Pty (Australia). A similar formulation was used for both types of pipes as listed in Table 1. except that the PVC-M samples contained 6 pphr CPE. Pipes with outer diameter 122 mm and wall thickness of ~9 and 8 mm for PVC-M and PVC-U, respectively, were slit and warmed in an oven at 120[degrees]C for 25 min before being flattened by compressing between two metal plates. Previous studies have shown that this process has no observable effect upon mechanical properties (4).
TABLE 1. The formulation of PVC-M pipe.

Material Amount (pphr)

PVC resin 100
Ti[O.sub.2] 1.5
Ca[CO.sub.3] 1.5
Stabilizer 3
Pigment 0.05
CPE 6


Compact tension specimens with a specimen width (W) of 75 mm were machined from the plate in the extrusion direction. The specimens were notched with a band saw and then the tip sharpened by sliding a sharp blade across the root of the notch. The ratio of the initial notch length to specimen width, [a.sub.o]/W was [approximately equal to] 0.4.

Fatigue Crack Propagation

Fatigue crack growth testing was performed using an MTS closed-loop servohydraulic testing machine according to ASTM E647-91. Precracked specimens were cyclically loaded at a constant room temperature of 22[degrees]C. Crack lengths were monitored using a traveling microscope. All fatigue testing was undertaken at a constant R-ratio, where R is defined by the ratio of minimum to maximum stress intensity factors ([K.sub.min]/[K.sub.max]). The R-ratio was fixed at 0.1 to evaluate the effect of various frequencies on the fatigue crack growth rate of PVC-U and PVC-M. The frequencies used were 1, 7, and 20 Hz and the cyclic load waveform was sinusoidal. The stress intensity factor amplitude ([DELTA]K) for the compact-tension specimen was calculated using the following equation:

[DELTA]K = [[DELTA]P/([BW.sup.[1/2]])] x f(a/W) (1)

where [DELTA]P is load range. B is the specimen thickness, W is the specimen width, a is the initial notch depth and

f(a/W) = {[2 + (a/W)]/[1 - [(a/W).sup.[3/2]]]} X [chi]{0.886 + 4.64 (a/W) - 13.32 [(a/W).sup.2] + 14.72[(a/W).sup.3] - 5.6 [(a/W).sup.4]}

Scanning Electron Microscopy

Once fatigue testing was completed, specimens were fractured under liquid nitrogen, sections cut and their surfaces were coated with a layer of chromium. A Hitachi 4500II field emission scanning electron microscope (FESEM) was used to examine the fracture surface morphology at an accelerating voltage of 20 kV.

Optical Microscopy

The subsurface fatigue damage at the crack tip was examined using an optical microscope. In this investigation a Leica RM 2245 microtome was employed to cut the specimens into thin sections with thicknesses ranging from 16 to 20 [micro]m. The sections were placed on a glass slide prior to observation under a Nikon Eclipse ME600 microscope, under bright-field and cross-polarized viewing conditions. These same samples were also coating with chromium and the crack-tip region observed in the FESEM.

RESULTS

Fatigue Crack Propagation Behavior

All the fatigue propagation data were analyzed in accordance with the Paris power-law relationship.

[da/dN] = A[DELTA][K.sup.m] (2)

where da/dN is the crack growth per cycle. A and m are empirical constants. All graphs demonstrate a sigmoidal log-log shape typical for this analysis; the constants for the steady-state region, A and m, were determined using least-squares regression and are given in Table 2. Figures la-c, demonstrate the results of crack growth rate measurements of PVC-U and PVC-M at frequencies of 1, 7, and 20 Hz, respectively. In general, the crack growth rate of PVC-U is similar to PVC-M, particularly at a higher stress intensity factor range ([DELTA]K). At 1 and 7 Hz, a slight difference in crack growth rate between PVC-U and PVC-M can be observed in the low [DELTA]K range (Figs, la and b) of 0.3-1 MPa * [m.sup.[1/2]] where PVC-M shows lower crack growth rates than for PVC-U. In contrast, no differences in crack growth rate at 20 Hz are noticeable for both materials across all loads.

[FIGURE 1 OMITTED]
TABLE 2. Paris power-law constants, A and m at various frequency for
PVC-U and PVC-M.

Specimen Frequency (Hz) A ([mu]m/cycle) m

 PVC-U 1 3.18 3.14
 7 2.26 3.32
 20 1.26 2.98

 PVC-M 1 2.54 2.85
 7 1.79 3.17
 20 1.24 2.86


The effect of frequency on the fatigue threshold ([DELTA][K.sub.th]) value in PVC-U and PVC-M is also demonstrated in Figs, la-c and this result is summarized in Fig. 2. For PVC-M, [DELTA][K.sub.th] values increased with increasing frequency; 0.19, 0.22, and 0.24 MPa*[m.sup.[1/2]] at 1, 7, and 20 Hz, respectively (see Fig. 2). For PVC-U, [DELTA][K.sub.th] increased after the frequency was raised from 1 to 7 Hz (0.22 and 0.24 MPa*[m.sup.[1/2]]), but remained the same with further increase in frequency to 20 Hz. Overall, the [DELTA][K.sub.th] value of PVC-M is lower than that of PVC-U at 1 and 7 Hz, but similar at a frequency of 20 Hz. For PVC-U, the fatigue threshold values are consistent with previous work (11), (12). From this result it can be established that the addition of CPE particles lowers the [DELTA][K.sub.th] value of PVC-M pipe.

[FIGURE 2 OMITTED]

Fracture surface observations

The morphology of the fracture surface was examined with field emission scanning electron microscopy (FESEM) to ascertain possible crack growth mechanisms. At low magnifications, both materials exhibit two similar characteristics; (1) the absence of discontinuous growth bands or striations (Fig. 3) and (2) an irregular patch-structure on the fracture surface (sec Fig. 4). Both fracture characteristics have been described in previous fatigue reports (13), (14) and the patchy-structured surface pattern reported in the impact fracture surface of CPE-PVC materials (15).

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

A higher magnification examination revealed that the fracture surface of PVC-U and PVC-M pipes was dominated with highly drawn microfibrils and microvoids as seen in Fig. 5. Micrographs of the fatigue-fracture surfaces of PVC-U and PVC-M are compared in Fig. 5 at two magnification levels. These micrographs were taken at very near-threshold [DELTA]K < 0.26 MPa*[m.sup.[1/2]]. The fracture surface of PVC-U (Fig. 5a) shows inhomogeneous rough surfaces that comprise small deformed sections with discontinuous edges. Clear remnants of broken fibrils were also observed at the end of the discontinuous edge while the inner parts of small deformation sections were filled with microfibrils. Figure 5b shows the small deformation section at higher magnification. Here, the craze structure was identified by the features of broken fibrils and voids. It is observed that the drawn fibrils also diffused to the yielding fragment. Numerous round globules could be seen at fibril separation points. Hence, the formation of each deformed section could affect the crack advance rate through the initiation and growth of microfibrils in these regions.

[FIGURE 5 OMITTED]

In contrast to PVC-U, the fracture surface of PVC-M was covered with more homogenous short remnants of broken microfibrils and microvoids (Fig. 5c). The fractal dimension in PVC-M also appears larger than in PVC-U. When the magnification was increased, a higher density of craze fibrils with highly drawn or elongated fibrils was observed (Fig. 5d). The size of voids was also larger than PVC-U and there was less adherence between the microfibrils.

The fracture surface of PVC-U and PVC-M at [DELTA]K [approximately equal to] 0.27 MPa*[m.sup.[1/2]], just above [DELTA][K.sub.th], is shown in Fig. 6. At lower magnifications both pipes show a similar fracture surface to that observed at [DELTA]K < 0.26 MPa*[m.sup.[1/2]]. In PVC-U, at higher magnifications the broken fibrils can be seen to be shorter and there are fewer voids. A similar trend was also seen in the PVC-M (Fig. 6c-d); decreasing density of fibrils with thicker yielding fragments. Although the microfibrils in both of the PVC fracture surfaces were short, the microfibrils in PVC-M have elongated more than the microfibrils of PVC-U.

[FIGURE 6 OMITTED]

At higher [DELTA]K = 0.83 MPa*[m.sup.[1/2]], in the middle of the steady-state region, the morphology of the fracture surfaces is apparently different (see Fig. 7). For the PVC-M, the presence of large fibrils was observed, without the formation the microfibrils (Fig. 7b). Remnants of fractured craze fibrils with short and large fragments were visible on the fracture surface. Additionally, the voids and tiny fibrils connecting the large fibril fragments were still formed. In contrast, PVC-U has a surface morphology consistent with brittle fracture (Fig. 7a) at a given [DELTA]K.

[FIGURE 7 OMITTED]

Finally, close to unstable crack growth at [DELTA]K = 1.3 MPa*[m.sup.[1/2]], both pipes showed brittle fracture appearances with the surface being essentially flat with no voids or fibrils visible (see Fig. 8), although a patchy structure pattern still exists at low magnification. No new fibrils are formed at higher [DELTA]K. This observation correlates well with a similar fatigue resistance of both specimens at higher [DELTA]K. Hence, it is suggested that at higher stress intensity factor amplitude, the CPE particles do not affect the fracture mechanism, as reflected in the similar crack propagation rates.

[FIGURE 8 OMITTED]

In order to explain further the enhancement of fatigue resistance at a higher frequency, the fracture surface examination was undertaken following crack growth at a frequency of 20 Hz. Fracture surface observation was also performed near to the threshold region. [DELTA]K < 0.26 MPa*[m.sup.[1/2]] (see Fig. 9). At 20 Hz, both specimens have a similar appearance to the sample shown in Fig. 5. The short truncated fibrils are in all cases are about 0.5 [micro]m long and 100 nm in diameter and indeed are similar to those found in crazes. It appears from the lower magnification SEM of the PVC-U (Fig. 5a) that the surfaces are smoother and lack the spongy or network appearance more prominent in the other samples in Figs. 5c, 9a, and c.

[FIGURE 9 OMITTED]

SEM examination of the microtomed samples at the crack tip exhibited a craze structure as shown in Fig. 10a. From this micrograph the craze structure observed consisted of an array of fibrils and voids which connected the crack faces. These microfibrils were pulled from the matrix, which then results in their alignment in a direction parallel to the applied stress. As the craze has load bearing capability (16), it could assist to hinder crack propagation. However, the rupture of the fatigued microfibrils and the coalescence of voids that formed in front of the crack tip (as shown with the arrow in Fig. 10) results in crack extension. The remnants of fibrils on the broken surface have been shown in previous micrographs (Figs. 5 and 6). Our observation of the craze structure is in good agreement with previous studies (16). It is noticed that there is a relation between the microfibrils features on the fracture surface and the magnitude of crack-tip opening. As seen in Fig. 10a, the crack tip opening in PVC-M is larger than PVC-U which is consistent with a tougher material. A larger crack tip opening in PVC-M corresponds to longer microfibrils which bridge the interface before the crack propagates. Consequently, longer microfibrils are observed on the fracture surface of PVC-M than PVC-U (see Fig. 5).

[FIGURE 10 OMITTED]

Transmission Optical Microscopy of Crack-tip Process Zones

This microscopy technique was applied to ascertain the crack propagation mechanisms during cyclic loading. An examination was performed on thin sections that were taken close to the crack tip. Here, the plastic zone appears as a, so-called, "stress-whitened zone". Figure 11 shows the optical micrographs of thin sections of PVC-M and PVC-U. For the PVC-M (Figs. 11a and c) relatively large crazes were observed which indicate high levels of plastic damage within a large crack-tip process zone. A similar observation was found in PVC-U, except that its craze zone was smaller and less apparent (Figs, 11d and f). No evidence of birefringence zone was observed in either material under cross-polarized light as shown in Figs, 11b and e. This signifies that the energy dissipating process was only through crazing without a shear banding (7). This result appears to explain why PVC-U and PVC-M possess near similar fatigue crack growth rates.

[FIGURE 11 OMITTED]

Since the plastic zone of PVC-U is less apparent, further examination was only performed on the PVC-M. It is observed that the plastic zone size varies with stress intensity factor amplitude, [DELTA]K, and frequency. As shown in Fig. 12, the plastic/crazed zone becomes larger as the [DELTA]K was increased from 0.6 to 0.85 MPa*[m.sup.[1/2]]. However, the size of damage zone did not vary significantly with frequency as seen in Fig. 13, which compares crack-tip zones following cyclic loading at frequencies of 1 and 20 Hz for [DELTA] K ~ 0.7 MPa*[m.sup.[1/2]]. At a frequency of 20 Hz. the width and length of the zone are 150 and 550 [micro]m, respectively, while at 1 Hz the width and length are 125 and 500 [micro]m, respectively.

[FIGURE 12 OMITTED]

DISCUSSION

This investigation has demonstrated the effect of CPE particles on the fatigue crack propagation rates of modified-PVC along with elucidation of fatigue mechanisms through the use of scanning electron and optical microscopy. From Figs, la-c both material systems show a strong sensitivity to cyclic loading frequency, with fatigue crack growth decreasing with increasing frequency.

The variation of fatigue threshold value ([DELTA][K.sub.th]) of PVC-M and PVC-U as a function of frequency is shown in Fig. 2. The [DELTA][K.sub.th] value of PVC-M increased with frequency over the range of 1-20 Hz. On the other hand, the [DELTA][K.sub.th] value of PVC-U is constant for both 7 Hz and 20 Hz. The fatigue threshold value [DELTA][K.sub.th] of PVC-M is lower than PVC-U.

The existence of CPE particles in PVC is associated with a lower fatigue threshold value in the PVC-M. It is known that rubber particles may act as stress concentrators due to the large difference in elastic modulus between the rubber particles and the matrix, concentrating the stress in the surrounding matrix with the highest concentration along the equatorial zone of the rubber particles (17). At a certain stress level, crazes will be initiated earlier in the modified-PVC and a crack propagates perpendicular to the applied stress direction. But once this crack starts, the propagation rate slows as a result of the dispersed CPE particles being obstacles to crack propagation in the PVC matrix; thus, a slightly higher fatigue resistance was shown in certain [DELTA]K regions.

Observations from previous PVC fatigue studies showed discontinuous bands (striations) in contrast to the fracture surface of the present results. Previous studies have shown a granular structure of short fibrils and long curled fibrils with rounded ends (18), (19). On the other hand, plastic deformation features were observed on the fracture surface of PVC-U (20). However, an examination of the SEM micrographs in the present work exhibited the characteristic fibril and voids of the craze structures. A slightly better fatigue resistance of PVC-M than PVC-U at low stress intensity factor amplitude. [DELTA]K, is related to its characteristic craze structure. The longer micro-fibrils and larger voids of PVC-M might contribute to a more stable crack growth and also indicate that much energy was absorbed during fracture (Fig. 5d). Furthermore, the slightly lower fatigue crack propagation rate of PVC-M could also be due to the longer time required to break down the deformed process-zone regions because these regions were larger than in PVC-U (Figs. 5a and c). It is likely that the formation of these longer microfibrils resulted from the drawing of fresh microfibrils from the matrix rather than the creep of microfibrils because the size of microfibrils in both specimens was similar (Fig. 5b and d). Indeed this is observed in the SEM micrographs in Fig. 10a, in which it is clearly observed that the fibrils in PVC-M have been drawing from the crack face which is associated with longer microfibrils as shown in the previous micrographs (Figs. 5 and 6).

A microscopic analysis by Anderson (21), mentioned that the aligned structure of fibrils could bear higher stresses than in undeformed amorphous state owing to the existence of stronger and stiffer covalent bonds compared with secondary bonds. Crack growth is related to the craze growth and the stability of fibrils. The shape of the craze in front of the crack tip is determined by the stress distribution at the craze interface and the load carried by the fibrils (22). According to Yu et al. (23), the PVC matrix will control energy absorption during crack initiation. Meanwhile, the impact modifier will affect the rate at which the material responds through the process of craze growth, i.e., the processes of crazed zone widening and lengthening. Therefore, the PVC-M may sustain higher loads than the PVC-U before the fibrils break and form critically sized voids which eventually lead to crack propagation.

Transmission optical microscopy indicated that crazing was the sole toughening mechanism under cyclic loadings both for PVC-M and PVC-U (see Fig. 11). Other toughening mechanisms such as shear banding and cavitation are not able to occur during fracture in the presence of rubber particles, and so lead to similar fatigue resistance. However, the toughness was significant. It is postulated that, because of the extremely small size of the CPE domains, the transformation of mechanism from crazing to shear banding would not be promoted via internal cavitation. The mechanism of cavitation is important for the enhancement of toughness because it relieves hydrostatic tension. In addition, it also induce shear yielding that will stop the crazes from propagation (24). Using an interparticle distance model. Dompas et al. (25), (26) have suggested the critical particle size for cavitation toughening to occur. This model is related to the rubber particles diameter ([d.sub.o]) and the relative volume strain ([DELTA]):

[d.sub.o] = [[12([[gamma].sub.r] + [[GAMMA].sub.sc])]/[[K.sub.r][[DELTA].sup.[4/3]]]] (3)

where [[gamma].sub.r] is the van der Waals surface tension, [[GAMMA].sub.sc] is the chain scission energy per unit surface and [K.sub.r] is the rubber bulk modulus.

TEM microscopy from Whittle (4) has demonstrated that the size of the CPE in the PVC-M specimen was about 100 nm. These particles are much smaller than the minimum size required for the cavitation which is calculated using Eq. 3 to be 200 nm. Very small particles will resist cavitation because the surface energy to create the void is higher than the release of the strain energy from cavitation (26). Hence, in this study, the formation of shear banding is suppressed and voids/crazing were formed instead.

The existence of CPE particles increased the yielding of PVC matrix which indicated that the adherence of these particles to the matrix is strong. Siegmann and Hiltner (27) suggested that the CPE must adhere well to the matrix to increase the stability of voids so that their coalescence during tensile loading is restricted. In fact, the effect of this adherence is still observed in fatigue testing where the CPE enhanced the plastic flow of the matrix through the formation of longer craze-induced fibrils.

Different blend systems may generate different toughening routes; studies by Sue and Yee, 1989 show that the transformation from crazing into shear banding enhanced the toughness in PA/PPO alloys, while Pearson and Yee (7) reported that the toughening mechanisms started with the cavitation of rubber particles and ended with shear deformation of matrix. However, a single fracture mechanism could also increase fatigue resistance. Morelli and Takamori (28) demonstrated that for rubber-PXE blends, the toughening effect was achieved with the formation of massive microcrazing. Discrete CPE particles are obtained by controlling the processing parameters; barrel zone temperature and screw speed. Hence, it is believed that the size of CPE particles might differ according to the temperature and screw speed. However, further studies are needed to determine the optimum particle size for the transformation of toughening mechanism.

There appears, to be a relationship between fatigue properties and craze structure. Comparison of micrographs in Figs. 5 and 9, shows that the craze/microfibril length is shorter with increasing frequency in accord with previous work by Lang et al. (29). They found that the measured craze length of PMMA decreased at higher frequency, in which an optical interferometry technique was applied to measure the craze length. In their work, they suspected that the decrease in craze length of PMMA was caused by the change in the stress for crazing rather than a hysteretic effect except at very high frequency, i.e., beyond 100 Hz. In addition, the higher the applied frequency, the higher the strain rates leading to an increase in the modulus and yield strength (30). Therefore, these observations support our assumption that the [DELTA][K.sub.th] value of the PVC, especially PVC-U, is less frequency dependent at high frequency.

As has been shown in Figs. 11 and 12, the process zone was formed around the crack tip with the deformation lines nearly parallel to the crack propagation direction. The presence of the CPE particles has caused an increase in the process zone size in PVC-M. Comparison of micrographs in Figs. 12a and b demonstrated that the size of plastic zone is proportional to the increase in [DELTA]K. This observation is correlated to Dugdale's model that is commonly used in estimating the plastic zone size (31),

[r.sub.p] = [[[pi][K.sub.max.sup.2]]/8[[sigma].sub.y.sup.2]] (4)

where [r.sub.p] is the radius of the plastic zone, [K.sub.max] and [[sigma].sub.y] are the maximum stress intensity factor and the yield stress, respectively.

However, the plastic zone size is observed to be insensitive to the cyclic frequency (see Fig. 13). The existence of crazes parallel to the crack, which are in fact remnants of the plastic zone preceding a slower crack, can be used as an additional indicator of the size of the plastic zone itself. Therefore, it is clear from Fig. 11 that the plastic zone in PVC-M is larger than in PVC-U due to the wider craze region parallel to the main crack.

[FIGURE 13 OMITTED]

A schematic of the fatigue toughening mechanism proposed in this study is illustrated in Fig. 14. There are two mechanisms involved; an intrinsic and extrinsic one. The intrinsic mechanism involves toughening being governed by extension of fibrils near the crack tip. This toughening effect was experienced more in PVC-M than PVC-U owing to the longer fibrils present and dominates at the beginning of cyclic loading or at low [DELTA]K levels.

[FIGURE 14 OMITTED]

As [DELTA]K, and hence [DELTA][K.sub.max], increases, the toughening mechanism will transform from the intrinsic to an extrinsic toughening. Here the extension of fibrils at the crack tip is less significant and increases in toughening are dependent more on the size of the craze process zone. This zone acts as a shield and hinders the propagation of cracks. Similar to intrinsic toughening, the extrinsic toughening is greater in PVC-M in which its plastic zone larger than in PVC-U, scaling with [DELTA]K. However, under cyclic loading the fibrils in PVC-U degenerate relatively more than PVC-M in the toughening zone.

CONCLUSIONS

The study shows that the addition of rubber particles does not produce a significant improvement in fatigue resistance of modified-PVC pipes. However, the addition of CPE particles has some beneficial effect on the fatigue resistance in a low stress intensity factor range, and lead to a lower fatigue threshold. This enhancement is associated with increased formation of craze fibrils. The absence of craze fibrils gives a similar fatigue crack propagation rate in both materials at high stress intensity factors. The cyclic fatigue threshold of PVC-M is more frequency dependent than in PVC-U, especially at higher frequencies. Crazing is the only fracture mechanism observed to occur in both PVC-M and PVC-U samples under cyclic loading.

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.) A.J. Whittle and A. Teo, Plast. Rubber Compos., 34, 434 (2005).

(2.) G.P. Marshall and D.P.S. MacKellar, Plast. Rubber Compos. Process Appl., 25, 276 (1996).

(3.) M.A. Osry. Plast. Rubber Compos.. 34. 9 (2005).

(4.) A.J. Whittle, The optimisation of rubber modified poly (vinyl chloride) pipes for pressure applications. Ph.d. Thesis, University of New South Wales (UNSW), Sydney, (2002).

(5.) H.R. Azimi, R.A. Pearson, and R.W. Hertzberg, J. Appl. Polym. Sci., 58. 449 (1995).

(6.) M.G. Wyzgoski and G.E. Novak, .J. Appl. Polym. Sci., 51, 873 (1994).

(7.) R.A. Pearson and A.F. Yee, J. Mater. Sci., 26. 3828 (1991).

(8.) A.M. Donald and E.J. Kramer, J. Appl. Polym. Sci., 27. 2729 (1982).

(9.) M.T. Berard and S.M. William, J. Vinyl Additive Tech., 2, 117 (1996).

(10.) X. Xi, M. Xiande, and C. Keqiang, Polym. Eng. Sci., 27. 391 (1987).

(11.) H.S. Kim. Y.W. Mai, and B. Cotterell, Polymer, 29. 277 (1988).

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

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

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

(15.) A.H. Siegmann and A. Hiltner. Polym. Eng. Sci., 24, 877 (1984).

(16.) A.M. Donald, "Crazing." in The Physics of Glassy Polymers, 2nd ed., R.N. Haward and R.J. Young, Eds., Chapman & Hall. London, UK, 295 (1997).

(17.) G.H. Mitehler and J.-U. Starke, "Investigations of Micromechanical and Failure Mechanisms of Toughened Thermoplastics by Electron Microscopy." in Toughened Plastic 11: Novel Approaches in Science and Engineering, C.K. Riew and A.J. Kinloch. Eds., American Chemical Society. Washington. DC. 251 (1996).

(18.) L. Konczol, M.G. Schinker. and W. Doll, J. Mater. Sci., 19, 1605 (1984).

(19.) M.G. Schinker, L. Konczol, and W. Doll, Colloid Polym. Sci., 262. 230 (1984).

(20.) W. Balika and R.W. Lang, Macromol Symp, 181. 341 (2002).

(21.) T.L. Anderson. Fracture Mechanics: Fundamentals and Applications. Taylor & Francis Group, United State of America, (2005).

(22.) J. Breen,.J. Mater. Sci., 28, 5833 (1993).

(23.) J. Yu, J. Summers, A. Hiltner, and E. Baer, J. Vinyl Additive Tech., 10. 11. (2004).

(24.) H.J. Sue and A.F. Yee, J. Mat. Sci., 24, 1447 (1989).

(25.) D. Dompas and G. Groeninckx. Polymer, 35, 4743 (1994).

(26.) D. Dompas, G. Groeninckx, M. Isogawa, T. Hasegawa, and M. Kadokura, Polymer. 35. 4750 (1994).

(27.) A.H. Siegmann and A. Hilner, Polym. Eng. Sci., 24, 869 (1984).

(28.) T.A. Morelli and M.T. Takemori, J. Mat. Sci., 18. 1836 (1983).

(29.) R.W. Lang. J.A. Manson. R.W. Hertzberg, and R. Schirrer, Polym. Eng. Sci., 24, 833 (1984).

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

(31.) L. Mascia, Thermoplastics Materials Engineering, Elsevier, London, (1989).

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

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

(2) Department of Manufacturing and Materials, HUM, Malaysia

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

(4) Iplex Pipelines Australia Pty Ltd, 35 Alfred Road, Chipping Norton, NSW 2170 Australia

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

DOI 10.l002/pen.2l369

Published online in Wiley InterScience (www.interscience.wiley.com).

[C] 2009 Society of Plastics Engineers
COPYRIGHT 2009 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 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:Technical report
Date:Jul 1, 2009
Words:5417
Previous Article:Anisotropic mechanical properties of hot-pressed PVDF membranes with higher fiber alignments via electrospinning.
Next Article:Dual-Type electrochromic device with single wall carbon nano tube employment in gel electrolyte.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters