SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

Environmental stress failure of polymers (specifically glassy thermoplastics) is a major failure mode in application end use. The environmental stress failure discussion will primarily be limited to non-reactive environments. Crystalline polymers and even lightly crosslinked polymers can exhibit similar failure to glassy polymers but often require a much higher external applied stress. The mechanisms of environmental stress cracking proposed in the literature will be discussed in detail along with testing methods commonly employed. Approaches to resolve environmental stress failure have limited discussion in the open literature although a significant number of references can be found in the patent literature. Polymer blend approaches, fiberglass reinforcement, and impact modification are areas where industry has solved the environmental stress failure problems. An additional unique mode of failure with water exposure in glassy engineering polymers yielding microcavity formation will also be discussed. Environmental stress failure is an area of considerable concern to the utilization of polymers in a myriad of applications and the unsolved problems existing in the area will be detailed suggesting the need for additional research attention. POLYM. ENG. SCI., 2013. © 2012 Society of Plastics Engineers


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

One of the major shortcomings of polymeric materials involves the loss of inherent mechanical properties on exposure to adverse environments. This is a particular problem with glassy polymers where specific organic liquids can result in cracking or crazing upon very modest applied stress (often an order of magnitude lower than the actual tensile or flexural strength of the polymer). In some cases, the molded-in stress of the glassy polymer is sufficient to result in failure without any external applied stress. This failure for glassy polymers (Bisphenol A polycarbonate (PC), polysulfone (PSF), polystyrene, poly(methyl methacrylate) (PMMA), poly(2,6-dimethyl-1,4-phenylene oxide) (PPO), and PPO/polystyrene blends is of major concern in the choice of a particular polymer for any specific application. While a critical problem for glassy polymers, it is also a concern for crystalline polymers as has been well-documented for polyethylene in various applications. Crystalline polymers are inherently more resistant to these failures for reasons that will be noted later; however, stresses occurring in various applications are adequate to promote failure in specific environments. Crosslinked polymers are generally much more resistant to environmentally induced failures particularly if compared to their glassy polymer counterparts. A specific comparison would be epoxy compared to the polyhydroxyether of Bisphenol A (Phenoxy). The myriad of environments encountered in application end-use includes organic liquids (solvents), synthetic oils, naturally occurring oils, detergents/surfactants, mold release agents, plasticizers, sealants/caulking agents, adhesive formulations, inks, anti-corrosion additives, lubricants, metal cutting fluids and even in extreme cases naturally occurring oils from fingerprint contact.

One of first classic examples of environmental failure involved low density polyethylene (LDPE) utilized for wire and cable jacketing replacing crosslinked unsaturated elastomers. As wire and cable needed to be pulled through long conduits, soap was often applied to provide lubricity and lower the friction for the crosslinked elastomers. Although LDPE jacketing offered smoother surfaces and lower friction, the procedure of soap application continued. Cracking of the LDPE wire and cable coating was observed after specific time intervals where wire and cable bending was prevalent resulting in stress-induced failure. This was traced to soap application [1, 2] and resistance to surfactant-induced stress failure of polyethylene became a standardized test for polyethylene performance still employed today (ASTM: D1693).

Although environmentally induced failure was recognized for glassy polymers such as polystyrene and PMMA, the introduction of higher glass transition temperature and higher performance polymers such as PPO (and polystyrene modified versions), Bisphenol A PC, and PSF for more demanding applications led to increased attention to this problem. It became apparent that environmental stress failure was an “Achilles heel” for these otherwise high performance materials and exposure to application environments needed to be strongly considered. It has been estimated that 15% of all failures of plastic materials in commercial use are related to environmental stress cracking (ESC) [3]. A higher value (25%) was noted in a brief review by Jansen [4]). Estimates of 30–40% were noted by Arnold in a review of glassy polymer ESC [5]. For specific glassy polymers the failure rate is probably even higher.

While environmental failure has been one of the major concerns for choice of polymeric materials for specific applications and likewise one of the major failures of glassy polymers in application end-use, the academic community has paid decreasing attention to this problem since studies over three decades ago. There are a number of early references involving the observation of cracking/crazing during the unsteady state diffusion of penetrants into a glassy polymer (studies related to Case II sorption kinetics) that will be discussed in detail as these observations appear to be missing from the present explanation of the ESC mechanism. The most relevant mechanism of environmentally induced stress failure appears to be a seminal paper by Gent published over 40 years ago [6]. Structure–property studies by Bernier and Kambour also published over 40 years ago denoted the stress-cracking tendency of glassy polymers exposed to organic solvents and correlated the polymer solubility parameter with stress cracking magnitude [7]. In these discussions it should be noted that the failure in polymers can occur due to crack or craze formation. Crazing involves fine fibrils extending between the failure surfaces of a polymer as compared to true crack formation where no polymer reinforcement exists. Crazing was initially described in detail by Spurr and Niegisch [8] and has been intensively studied in the literature usually without environmental contributions. In this review, the environmentally induced failure mechanism may involve both crazing and cracking as the failure mechanism.

Stress cracking of biopolymers (e.g., corn, rice, soybeans, and pasta) is a problem observed during the drying process [9]. The surface transitions from a rubber to a glass state and the contraction of the surface due to loss of water yields stress sufficient to cause surface cracking.

LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

An early explanation noted that the reduction of the surface energy for crack formation occurs with organic liquid exposure allowing for lower stress for craze or crack formation compared to benign environments (air or vacuum) [10, 11]. This hypothesis notes that the stress required for crack formation is lowered by environmental exposure as the solid/liquid interfacial tension is lower than that expected for the solid/air interface. This reduction of the surface energy required for crack formation is embodied in Young's expression:

  • equation image(1)

where γSL is the solid/liquid interfacial tension, γS is the solid–air surface tension, and γL is the solid–liquid surface tension. This concept may contribute to environmental stress crazing/cracking but is generally believed to be a minor effect in consideration of the overall mechanism particularly where severe stress cracking occurs.

An early interesting crazing observation was noted with exposure of polystyrene immersed in methanol and then exposed to heptane vapor [12]. The resultant crazing was considered to be due to an apparent domain structure believed to exist in high molecular weight polymers.

Stress-cracking of glassy polymers as a unique mechanism of dissolution in marginal solvents was noted by Ueberreiter [13] with PMMA exposed to dimethyl phthalate. During exposure to the solvent, it was noted that initial cracks appear perpendicular to the polymer surface due to the tensile stress resulting from surface swelling. Below the gel temperature, dissolution of the polymer by stress cracking into minute particles without the appearance of a gel layer was observed. This is illustrated in Fig. 1. It is interesting to note that stress cracking in this case occurs without application of an applied stress although frozen-in stress in the glassy polymer was suggested to be a factor. Above a temperature termed the gel temperature, solvent penetration leads to a polymeric gel and the initial cracks (noted in Fig. 2) are engulfed and disappear. These cracks were perpendicular to the surface and appeared to be very sharp cracks. The black particles present in Figs. 1 and 2 were carbon black particles incorporated to provide visual contrast between the polymer, the gel layer, and the solution. This was also noted with toluene dissolution of polystyrene where cracks occurred with some degree of regularity. A more recent review of polymer dissolution [14] discussed a number of references with observations similar to that of Ueberreiter noted above.

thumbnail image

Figure 1. Dissolution by stress cracking below the gel temperature (PMMA-dimethylphthalate) (reproduced from K. Ueberreiter, Diffusion in Polymers, 1968, 219–257, © Academic Press, New York).

Download figure to PowerPoint

thumbnail image

Figure 2. Dissolution above the gel temperature for PMMA-dimethylphthalate (reproduced from K. Ueberreiter, Diffusion in Polymers, 1968, 219–257, © Academic Press, New York).

Download figure to PowerPoint

The diffusion of a penetrant into a glassy polymer has the capability of creating large internal stresses (osmotic pressure) prior to the point that sufficient penetrant has reduced the glass transition to the exposure temperature. This was discussed by Crank [15] and noted to be a contributor to the anomalous diffusion behavior observed in glassy polymers. Anomalous diffusion in glassy polymers has been widely recognized and studied and often referred as Case II diffusion (or sorption) and considered a unique situation with glassy polymers where the diffusion process is much faster than the relaxation process allowing for large internal stress. Crazing was noted to be an inherent aspect of Case II transport in glassy polymers at high penetrant activity [16] as illustrated in Fig. 3. In some of these studies and reviews of the process, the swelling effects were noted to be sufficient to cause crazing and cracking at or near the advancing (and usually sharp) diffusion front [17–32]. This was also observed with water diffusion into poly(tetrahydrofurfuryl methacrylate-co-hydroxyethyl methacrylate) where microscopic cracks were noted based on environmental scanning electron microscopy (ESEM) and NMR analysis [33]. In essence, as the diffusion of a penetrant into the glassy polymer occurs, the surface penetrated layer must expand. At the diffusion front which could be quite sharp, the expanded surface creates a tensile stress on the underlying layer. While the expanded polymer is thus placed under a restraining compressive stress, the osmotic pressure on the expanded glassy polymer network probably well exceeds this compensation. With penetrant sorption, significant volume changes can occur prior to reduction of the glass transition temperature to the exposure temperature. For example, consider a penetrant with a glass transition temperature of 100 K, a polymer with a Tg of 373 K and an exposure temperature of 296 K. The Fox equation predicts 9.5 wt% penetrant sorption until the polymer/penetrant Tg reaches the exposure temperature. This predicts a high level of internal stress as the stress relaxation of the expanded glassy polymer chains will be much slower than the diffusion process. This concept has actually been noted in a number of early papers and noted to be a mechanism of swelling non-solvent induced stress cracking [34–38]. The concept reemerged with studies by Thomas and Windle [30] and discussed by Brown [39]. It is, however, rarely noted in recent ESC papers and not mentioned in several review articles [3, 5]. One reason why it has been discounted is due to the observation of stress cracking in fluids which show limited bulk sorption [40]. However, the evidence and logic appears that it should be considered a factor for the cases for ESC with swelling non-solvent exposure to glassy polymers.

thumbnail image

Figure 3. Generalized description of diffusion in glassy polymers (alkane transport in polystyrene) (adapted with permission from H. B. Hopfenberg and H. L. Frisch, J. Polym. Sci. Part B, Polym. Lett., 1969, 7, 405, © John Wiley & Sons).

Download figure to PowerPoint

The propensity of polymers to stress crack in the presence of organic liquids was noted to be related to the proximity of the solubility parameters of the liquid and the polymer [7]. Using the criteria of critical strain for crazing/cracking, it was observed that a minimum in critical strain was observed at the solubility parameter match for PPO and solvent environments. The data indicated that most organic liquids would have a reduction in critical strain for the polymer compared to the value in air and this reduction was related to the difference between the solubility parameter of the environment and the polymer. This approach was repeated with PSF [41], polystyrene [42], and PC [43] with general correlation with the solubility parameter but less definitive than observed with PPO particularly involving polystyrene partly due the low value of critical strain in air. Stress crack resistance of PSF and PC were noted to be similar based on similar solubility parameters. Additional data on this approach for polyetherimide (PEI) and poly(ether sulfone) [44] and a styrene-acrylonitrile copolymer and a dicyano bisphenol PC [45] have been published. Improved correlation was observed using a two-dimensional solubility parameter applied to PC [46] and a three-dimensional solubility parameter applied to PC and PSF [47]. A two-dimensional solubility parameter approach using the solubility parameter and a hydrogen bonding parameter was noted to be more predictive for environmental stress crack resistance (ESCR) than the solubility parameter alone [48]. The application of Hansen's solubility parameters to predict ESC has been noted in several other papers [49, 50]. Mai discussed the limitations of the solubility parameter approach for predicting ESC [51]. The ESC of mixtures of miscible solvent pairs was studied to determine whether the rule of mixtures would apply when solvents were chosen on either side of the polymer solubility parameter (thus a mixture should be a more aggressive ESC agent than either of the constituents [52]. The solubility parameter correlation was shown to be poor for mixtures. An improved correlation was noted with the combination on combining Hansen partial solubility parameters with an enthalpic parameter obtained from the Flory–Huggins theory.

The generalized data relationship plotted in many of the above papers relating critical strain to solubility parameter is noted in Fig. 4. In several cases, a distinct minimum in critical strain versus solubility parameter was noted correlating with the polymer solubility parameter. However, with several other polymers (PC and PSF), minimum positions were noted in the range of the solubility parameter yielding comparative behavior as generalized in Fig. 4. In these studies and the unpublished observations of the author, it appears that the most aggressive stress cracking agents are often not the good solvents but swelling non-solvents. This hypothetically may be due to the good solvent creating a rapid gel layer that prevents propagation of the glassy crack. This is in agreement with observations of Ueberreiter [13] noted above. With PC, rapid swelling in good solvents can result in crystallization thus allowing a protective mechanism. The range of solubility parameter difference for environments causing crazing or cracking is at least 10 (cal/cc)1/2 and is often higher with polymers having higher critical strain required for crazing in air. The range of critical strain for crazing in air for the polymers noted goes from 0.35% for polystyrene to >2.0% for the tougher engineering polymers. As noted in Fig. 4, crazing is usually observed at higher critical strains with cracking noted for the more aggressive environments.

thumbnail image

Figure 4. Generalized behavior of critical strain versus solubility parameter difference between polymer and the environment liquid. (Solid line corresponds to polymers where a distinct minimum is observed; dashed line to where minima are observed on both sides of the zero value solubility parameter difference: see discussion above).

Download figure to PowerPoint

The effect of plasticization noted by Kambour and coworkers on the ESC of polymers has been well noted in the literature to play a key role in the failure mechanism. A more complete description of the mechanism was detailed by Gent [6] and presently embodies the principles generally considered as the physical process. The exposure of a surface flaw under stress will allow dilatant stress induced swelling resulting in the flaw tip to exhibit a greatly decreased Tg and propagate by viscous flow. Propagation of the crack tip depends on the crack tip radius and as the crack tip radius decreases the stress required for propagation decreases yielding stress concentration factors in the range of 10–50 were noted by Gent. At the tip of a sharp crack, the tensile stress (σc) can be significantly enhanced over the applied tensile stress (σ) expressed by the Inglis [53] expression noted by Gent.

  • equation image(2)

where lc is the length of the crack, r is the crack tip radius, and k is the stress concentration factor. The ability of significant increased solvent sorption under dilatant stress was noted by Gent to explain the stress cracking/crazing ability of liquids exhibiting low bulk solubility. This was predicted by modification of the Flory–Huggins swelling relationship to account for dilatant stress, σD. This expression

  • equation image(3)

where V1 is the molar volume of the solvent, ρ is the density of the polymer, Mc is the molecular weight between crosslinks (or physical entanglements for an uncrosslinked polymer), ν2 is the volume fraction of the polymer in the swollen polymer–solvent mixture, and χ is the Flory–Huggins interaction parameter. Although the theory is usually applied to crosslinked rubber, the stress induced sorption for amorphous and semi-crystalline poly(ether ether ketone) (PEEK) exposed to toluene has been experimentally observed [54].

Thus, the Kambour et al. approach and the mechanism proposed by Gent have a common significant variable as the Flory–Huggins interaction parameter is qualitatively related to the value of equation image as noted by the relationship:

  • equation image(4)

where νr is a reference volume, R is the gas constant, T is the temperature and 1 and 2 denote the polymer and the environment. It is noted that Bernier and Kambour [7] mentioned stress-induced sorption as a potential cause of ESC behavior. The observations of Kambour and the mechanism of Gent comprise the generally accepted explanation for ESC in glass polymers. The observations noted above for Case II transport need to be included in the ESC description for environments that cause significant swelling. The unsteady state diffusion of a penetrant into a glassy polymer can obviously result in crack formation without applied external stress. This provides the “flaw” which is an inherent assumption in the Gent model.

Polyethylene, although significantly more resistant to ESC than the glassy polymers discussed above, also exhibits similar failure when exposed to adverse environments. This behavior has been widely studied at least partly due to the large scale applications of wire and cable jacketing and container applications where surfactants, solvents, petroleum products, food products, and various consumer products are packaged. The most significant studies of ESC in polyethylene has been reported by Brown and coworkers [55–59] covering both low density and high density polyethylene (HDPE). Brown noted that the most prominent cracking agents for polyethylene were detergents with Igepal Co630 (ethoxylated nonyl phenol) being a prime example [55]. He proposed a mechanism for the ESC of polyethylene which had some similarity to that proposed by Gent for glassy polymers. The mechanism proposed involved stress-induced swelling and resultant plasticiziation of amorphous regions in polyethylene. The Flory–Huggins swelling relationship was employed to explain the stress-induced swelling and the importance of the proximity of the solubility parameter was noted to be an important variable. Thus, the basic elements of this mechanism have similarity to the discussion above on environmental cracking in glassy polymers. It was noted that the solubility parameter of Igepal Co630 was similar to polyethylene. Another aspect of this mechanism was that a large molar volume would favor the stress cracking tendency. With HDPE, it was observed that slow cooling or annealing of the specimens yielded improved ESC resistance [57]. Polypropylene is considered to be more resistant to ESC than polyethylene and studies in the literature are rare. One study noted ESCR results for impact polypropylene produced by the cascade reactor process by incorporation of ethylene during polypropylene polymerization [60]. The ESCR results (constant stress testing in isopropanol) showed increasing ethylene contents yielded lower ESCR.

The micromechanics/fracture mechanics of craze/crack growth in glassy polymers including craze nucleation, craze propagation kinetics, craze morphology, and fundamental molecular parameters has been studied in detail by many investigators without environmental exposure [61–67]. As noted by the references, several key investigators (Kramer, Andrews) have extensively studied this area. Extension of these studies to investigate the micromechanics of craze growth in glassy and semi-crystalline polymers with environmental exposure is noted in Refs 68–73. It is beyond the scope of this paper to discuss the details of craze/crack growth fracture mechanics and propagation kinetics.

TEST METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

One of the most utilized methods to ascertain resistance to ESC/crazing involves determination of the critical strain required for crack/craze development. The most common version of this test involves a three point bending apparatus with stress applied to the center of the specimen resulting in a flexural strain [7] (see Fig. 5a). The maximum fiber strain, εm, is determined from the expression:

  • equation image(5)

where d is the specimen thickness, l is the length of the specimen, and ym is the deflection of the center of the specimen from the unstressed position. A variation of this method employs an elliptical strain apparatus with the most common version referred to as the Bergen elliptical strain jig [74] (see Fig. 5b). The jig is in the shape of an elliptical quarter section expressed by the equation:

  • equation image(6)

where x, y represent semiaxis constants (in inches). The specimen is firmly conformed around the elliptical surface and allows for varying strain along the length of the specimen.

thumbnail image

Figure 5. (a–d) Illustration of various test methods to access environmental stress crazing/cracking.

Download figure to PowerPoint

Constant stress tests can simply involve lever arm loading of test specimens and either determining stress required for craze/crack formation or in many cases determining the time for specimen failure. This method is often referred to as tensile stress rupture or if the strain is measured tensile creep rupture. A dynamic variation of this method involves determination of the stress–strain behavior (at constant strain rate) in an environment compared to air [75]. This method can involve immersion of the sample in a specified environment or employing a localized environment at the center of the specimen to prevent clamp failure. The author utilized cotton as an adsorbent material wrapped around the specimen and saturated with the test environment. Plastic film and/or Al foil were employed to confine the environment and prevent evaporation (and evaporative cooling). Thin (20 mil thick, 0.125 inches wide) shear cut specimens were often employed with multiple specimens attached to lever arms and timers to determine failure allowing for stress versus time to failure determinations (as illustrated in Figs. 6 and 7) [76, 77]. The original plots were stress versus log time to rupture and exhibited curvature of the data. The log stress versus log time to rupture appears to have linearity in the range of variables tested. This procedure allowed for comparing various polymers for specific environments as well as investigating the ESC for polymer blends. This method was also employed to determine the stress failure behavior of polyethylene (HDPE) exposed to various environments. The methodology employed involved a low molecular weight HDPE exposed to various liquid environments (at 60°C and 1000 psi) compared to the undiluted Igepal Co630 (considered one of the most aggressive environments). This procedure was utilized (at Union Carbide in the 1970s and early 1980s) as a qualitative method to assess the potential of HDPE for packaging numerous materials in bottles/containers. One of the more aggressive environments involved cottonseed oil commonly present in various food products.

thumbnail image

Figure 6. Environmental stress failure results for glassy engineering polymers exposed to isooctane/toluene (70/30 by volume) environment (reproduced with permission from L. M. Maresca and L. M. Robeson, Engineering Thermoplastics: Properties and Applications, 1985, 255–281, © Taylor & Francis).

Download figure to PowerPoint

With polyethylene the choice of constant strain versus constant stress tests can lead to highly different results. The constant strain tests favor lower crystallinity (lower density) polyethylene as the resultant surface stress is much lower than HDPE expected from large differences in modulus. With constant stress tests the higher crystallinity HDPE has better ESCR as the strain is much lower at equal stress due to the higher modulus. The choice of test depends on the application as wire and cable would be best served by the constant strain test to reproduce bent wire and cable situations.

Another variation of the constant stress test involves a cantilever beam subjected to a load at the unsupported end of the beam [78] (see Fig. 5c). The application of a cracking/crazing environment will lead to craze/crack development from the higher stress supported end to a point along the beam where no crazes/cracks are observed. The critical stress, σc, is determined from the expression:

  • equation image(7)

where a is the applied load, w is the width of the specimen, d is the specimen thickness, and x is the distance from the end of the specimen to where crazes/cracks are first observed.

Microhardness is a testing method capable of observing polymeric surface changes due to environmental exposure (involving chemical or UV attack or just chemical exposure). This method involves a pyramidal diamond indentation impressed on a polymeric surface with varying load/time measurements. The impression area is microscopically measured after testing and compared to the environmental effect versus air. The method has been noted to ascertain ESCR for polymer environments where low sorption fluids are involved [79]. It is not clear whether this procedure may indicate surface plasticization due to stress-induced sorption as a precursor for ESC.

Fatigue crack growth involving precracked (notched) samples subjected to sinusoidal loading in specific environments compared to air demonstrated the ability to differentiate ESCR for polystyrene as a function of molecular weight [80]. This method noted impact polystyrene gave better ESCR than unmodified polystyrene with 2.0 μm rubber particles better than 5.0 μm particle size. Strain hardening measurements on different polyethylenes showed a correlation between the strain hardening modulus and the ESCR determined by full notch creep testing [81]. A number of standardized tests relevant to ESC exist from the American Society for Testing and Materials (ASTM) and the International Organization for Standardization (ISO) listed in Table 1.

Table 1. Standardized tests for environmental stress crack resistance
Test methodSubject of test method
ASTM D 1693Environmental stress cracking of ethylene plastics: bent strip ESCR (see Fig. 5d)
ASTM D 2561Environmental stress crack resistance of polyethylene blow molded containers
ASTM D5397Standard test method for evaluation of stress crack resistance of polyolefin geomembranes using notched constant tensile load test
ASTM D5571Standard test method for environmental stress crack resistance (ESCR) of plastic tighthead drums not exceeding 60 Gal (227 L) in rated capacity
ISO 16770Determination of environmental stress cracking (ESC) of polyethylene-full-notch creep test (FNCT)
ISO 22088-2Plastics—determination of resistance to environmental stress cracking (ESC)—Part 2: bent strip method
ISO 22088-3Plastics—determination of resistance to environmental stress cracking (ESC)—Part 3: constant load method
ISO 22088-4Plastics—determination of resistance to environmental stress cracking (ESC)—Part 4: ball or pin impression method
ISO 22088-5Plastics—determination of resistance to environmental stress cracking (ESC)—Part 5: constant tensile deformation method
ISO 22088-6Plastics—determination of resistance to environmental stress cracking (ESC)—Part 6: slow strain rate method

Methods for determination of stress crack resistance have been noted (critical strain, critical stress, stress-time to failure) for broad comparison of polymeric materials. Often practical testing related to specific applications are utilized to quality control production with several examples noted below. One test employed for thermoformed low rubber content impact polystyrene used for container lids involved warm chicken soup. The natural oils in chicken soup could result in stress cracking under stress/temperature. The test involved three stacked containers of chicken soup held at a specific temperature. The reason three containers was chosen was that a person could carry a maximum of three. Another example of very practical testing involved PSF used for microwave trivets for bacon. Bacon contains aggressive fats/oils for specific polymers but PSF was noted to be resistant to stress cracking attack if molded properly. Quality control testing involved multiple microwave cooking of bacon. The person conducting the testing did this continuously and became very popular. Although eating in a laboratory was verboten and the cooked bacon was to be properly destroyed, the smell permeating the laboratory and the adjacent areas was too much of a temptation and the obvious occurred.

APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

One of the obvious approaches to reduce ESC is in the polymer choice for a specific environment. Crystalline or crosslinked polymers can provide significant ESCR over most glassy polymers. In consideration of glassy polymers, increasing the value of equation image will increase the ESCR. This is illustrated in Fig. 6 where the isooctane/toluene environment (70/30 by vol) to simulate gasoline shows improved ESCR as the improvement parallels increasing values of equation image. The solubility parameter of isooctane/toluene (70/30) is 15.3 MPa1/2 and the polymers have a range of ∼ 19 MPa1/2 for modified PPO to ∼ 22.1 MPa1/2 for polyarylethersulfone. (Gasoline was not allowed in the laboratory without special permission due to the low level of benzene typical of unleaded gasoline). Other methods commonly employed commercially include polymer blend approaches, fiberglass addition, and impact modification. These methods will be briefly discussed. Although these methods are commonly employed to resolve ESC problems with specific polymers, the literature documentation is rather limited.

POLYMER BLEND TECHNOLOGY

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

With the advent of amorphous engineering polymers in the 1960s (PPO, PC, PSF), problems with ESC surfaced as these polymers were introduced to a myriad of diverse applications. In order to expand the potential applications, it was noted that resolution of the environmental stress crack issues would be necessary. A significant number of polymer blends resulted at least partly due to solving the ESC issue while maintaining much of the performance value of the specific engineering polymer.

The addition of a crystalline polymer (with proper compatibilization) to the amorphous engineering polymers can lead to significant improvements in ESCR as noted in Fig. 7 for the miscible blend of amorphous PEI with the crystalline PEEK. The improvement of PC ESCR was shown with the addition of a miscible crystallizable cyclohexane dimethanol tere/iso phthalate polyester [82]. Improvement in the ESCR for PMMA with the addition of crystalline and miscible poly(vinylidene fluoride) was noted for both critical strain and critical stress measurements [83]. An early example for phase separated blends involved the addition of a nylon-6/PSF block copolymer to PSF [84]. With sufficient addition of the block copolymer such that it contributed to the continuous phase morphology, substantial improvements in ESCR were noted. The development of engineering polymer blends for automotive applications often involved high Tg polymers (PC, PPO) with crystalline polymers (poly(butylene terephthalate) (PBT), nylon 6,6) [85]. The crystalline polymer addition dramatically improved the ESCR to gasoline, oil, lubricant, and detergent exposure. Commercial systems included Noryl® GTX (PPO/nylon 6,6) and Xenoy® (PC/PBT). Other blends proposed for similar application involved Gemax® (PPO/PBT), and acrylonitrile-butadiene-styrene (ABS)/nylon 6 or nylon 6,6 (Triax, Elemid). Fiberglass-reinforced PSF/poly(ethylene terephthalate) (PET) blends were commercialized under the tradename Mindel B for electrical/electronic connectors where cleaning solvents are typically encountered. PPO/poly(phenylene sulfide) (PPS) blends (under the Noryl variant trademark) and PEI/PPS blends (Supec™) have also been commercial offering improved ESCR for the amorphous polymer as PPS is a crystalline engineering polymer. PC/PET blends (Macroblend, Sabre) have also been offered commercially with ESCR improvements over PC. The addition of polyethylene to impact polystyrene at modest levels has been utilized to improve ESCR for food packaging applications. In one study involving high impact polystyrene (HIPS) blends with HDPE, it was noted that a compatibilizing agent was required to yield the desired improvement in ESCR [86]. A styrene/ethylene-butylene/styrene (SEBS) block copolymer was employed for compatibilization.

thumbnail image

Figure 7. Environmental stress rupture of PEI/PEEK blends exposed to a toluene environment (adapted with permission from J. E. Harris and L. M. Robeson, J. Appl. Polym. Sci., 1988, 35, 1877, © John Wiley & Sons).

Download figure to PowerPoint

There are examples of amorphous polymer blends where the ESCR improvement of one of the polymers has allowed for additional applications. The prime example is PC/ABS (commercial for over 40 years with Cycoloy, Bayblend, and Pulse tradenames). ABS addition to PC offers some ESCR improvement for specific applications. ABS/PSF blends offer similar ESCR attributes (Mindel A). The patent literature notes many examples where the polymer blend approach was employed to improve the ESCR of a specific engineering polymer [87–98] as well as a PhD thesis on PC blends to improve ESCR [99].

Many of the blends noted above exhibit phase separation thus sufficient addition of the ESCR improving polymer must be present in a sufficient concentration such that it contributes to the continuous phase morphology (generally above 30 vol%) before significant improvement is observed. The modulus and strength of phase separated blends can be predicted using various models from the unblended values. A comparison of these models for predicting the stress to failure (with time to failure held constant) will be made. The parallel model is considered the upper bound and approximates the situation where the higher strength polymer is the continuous phase. The series model is considered the lower bound and approximates the condition where the lower strength polymer is the continuous phase. With polymer mixtures, continuous phase behavior is observed at the low and high ends of the concentration range whereas at intermediate concentrations both polymers contribute to the continuous and the dispersed phases. The contribution to the continuous phase can be predicted by the equivalent box model (EBM) [100, 101] based on the DeGennes percolation parameters [102]. These models are noted below:

  • equation image(8)
  • equation image(9)
  • equation image(10)

where σb, σ1, and σ2 represent the stress to failure for the blend, Component 1, and Component 2. ϕ1 and ϕ2 are the volume fractions of Components 1 and 2 in the blend.

ϕ1p, ϕ2p, ϕ1s, ϕ2s, and ϕs are defined by the expressions:

  • equation image(11)
  • equation image(12)

ϕ1cr and ϕ2cr are the critical threshold percolation values of Components 1 and 2 and T1 and T2 are the critical universal exponents for the components. ϕ1cr, ϕ2cr, T1, and T2 can be considered adjustable parameters. For discrete spherical domains; ϕ1cr = ϕ2cr = 0.156 and T1 = T2 = 1.833 as predicted from percolation theory [11] (3-dimensional array). In the regions of low concentration where 0<ϕ11cr or (0<ϕ22cr);ϕ1p = 0 and ϕ1s = ϕ1 or (ϕ2p = 0 and ϕ2s = ϕ2) and equation image. A is considered the interfacial adhesion from 0 (no adhesion) to 1 (perfect adhesion) and is assumed to be 1 in this comparison. The stress required for failure in one minute for polystyrene/styrene-acrylonitrile (SAN) copolymer blends exposed to a heptane environment were utilized [103] and the experimental results compared to model predictions are shown in Fig. 8. While the qualitative trend is predicted by the EBM, the experimental results show a higher ESCR than the prediction with some agreement with the parallel model where SAN is expected to be the continuous phase.

thumbnail image

Figure 8. Various mechanical models applied to the environmental stress rupture of styrene/acrylonitrile (SAN)/polystyrene blends exposed to heptane (adapted with permission from L. M. Robeson, Polymer Blends: A Comprehensive Review, 2007, © Hanser Publishers and L. M. Robeson, Problem Solving with Plastics, 1971, 87–90, © National Association of Corrosion Engineers).

Download figure to PowerPoint

The application of polyolefin blends to optimize the ESCR for wire and cable as well as packaging utility has been practiced commercially for over five decades. Most of the information is either as trade secrets or in the patent literature [104–106]. Several open literature references do discuss this approach to improve ESCR [107–109]. The addition of ethylene vinyl acetate copolymers to LDPE was shown to improve the ESCR using the notched bent strip Bell-telephone test [107]. The addition of linear low density polyethylene (LLDPE) to HDPE showed improved ESCR in ethylene glycol in a notched constant stress testing protocol [108]. Addition of higher molecular weight HDPE to lower molecular weight HDPE also showed improved ESCR of the lower molecular weight HDPE. The improvement in ESCR was attributed to the tie-molecule density. The addition of LLDPE to HDPE was found to be significantly more beneficial in the ESCR than LDPE in the ASTM D1693 test [109].

FIBER REINFORCEMENT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

Fiberglass (and other fibrous reinforcements) offers a facile method to significantly enhance the resistance to environmental stress failure. In essence, fibers can bridge the cracks or crazes which may develop at the surface and prevent easy propagation through the material. This was illustrated in a study where the stress rupture of polystyrene in acetone required much higher stresses with fiberglass addition whereas glass bead addition at higher levels showed little to no improvement [103]. It was shown that the stress carried by the fiberglass needed to be surpassed in order for crack propagation to proceed to sample failure. This is illustrated in Fig. 9. This approach to alleviate environmental failure has been utilized in many commercial examples. One case involved the choice of PSF for an automotive application as a spring loaded safety interlock device in the 1970s. PSF was chosen due to the ability to withstand paint oven temperatures. Solvent resistance was one of the material selection tests and it was observed that certain solvents would result in part failure with the spring ejecting from the device. A “frantic” call was received from the company reporting that the personnel testing the device were amusing themselves by “shooting” the springs with solvent exposure. The solution to the problem was simply fiberglass reinforcement which passed the battery of environmental tests and resulted in a very successful automotive application for PSF. It is of interest to note that although fiber reinforcement is a very effective way to alleviate ESC problems the literature (except for suppliers literature) is virtually devoid of studies related to this potential solution. Unpublished studies by the author on mineral fiber reinforcement showed modest ESC resistance improvement but not to the level of fiberglass [110]. This was believed to be due to the much smaller fiber lengths and decreased ability (compared to fiberglass) to bridge the resultant cracks developing from environmental exposure. Particulate reinforcement gave limited to no improvement. A search on nanoparticle reinforcement (such as exfoliated clay) did not reveal any relevant references. One patent is noted claiming the use of nanoparticles to improve the ESCR [111].

thumbnail image

Figure 9. Environmental stress rupture data for polystyrene composites exposed to an acetone environment (adapted with permission from L. M. Robeson, Polymer Blends: A Comprehensive Review, 2007, © Hanser Publishers and L. M. Robeson, Problem Solving with Plastics, 1971, 87–90, © National Association of Corrosion Engineers).

Download figure to PowerPoint

IMPACT MODIFICATION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

Many of the available commercial polymers are provided with impact modification. Impact polystyrene and ABS are among the major commercially available products in this classification. It has been well-recognized in the literature that rubber impact modification (particularly in the case of polystyrene) leads to improved ESCR [47, 80, 103, 112–118]. This was illustrated in a comparison of the case of a blend of polystyrene with a high rubber content polystyrene where the stress required for failure with exposure to heptane increased by more than an order of magnitude with increasing high rubber content PS concentration during constant stress loading [103]. The influence of rubber particle size was noted to be important in impact polystyrene ESCR with larger particles having improved ESCR compared to smaller (<1–2 μm particles) [113]. It was noted that a limit exists to the particle size where >6 μm particles (rubber content constant) were less effective. The addition of mineral oil was found to be detrimental to the ESCR of impact polystyrene [114]. In a study of the orientation effects of ESC, the critical strains for cracking were significantly higher for the ABS samples than unmodified SAN [115]. This study tested injection molded specimens in the direction of flow (oriented) and transverse to the direction of flow. The effect of orientation (produced by injection molding) was more pronounced, however, for the ABS samples compared to SAN. Significant increases in the critical strain for craze/crack formation was observed with increasing rubber content of ABS for methanol, ethanol, and glacial acetic acid [47] with a similarity noted with the observations of Ref.103. Rubber modification of polystyrene was shown to yield a higher critical strain for a series of alcohol environments tested between −20 and 20°C [116]. The addition of core-shell acrylic rubber particles to polystyrene or SAN was noted to improve the ESCR to Freon vapor [119]. An earlier patent reference noted similar ESCR improvement with the addition of core-shell ABS rubber particles to impact polystyrene to improve Freon resistance [120]. The rubber particle size was shown to be an important variable with 1–2 μm preferred for polystyrene and 0.30 μm preferred for SAN. It should be noted that the rubber modification results in lower modulus thus lower surface stress at equal strain compared to unmodified polystyrene. Also the surface crazing that occurred may be stabilized in a rubber toughened system compared to the results noted for constant stress time to failure experiments.

Impact modification of PC offering improved ESCR has been noted in the patent literature [121]. Grafted ethylene-propylene rubber and grafted acrylic rubber modifications were noted. Improved ESCR for PC was noted with addition of a hydrogenated styrene-butadiene-styrene block copolymer [122]. The addition of thermoplastic polyurethanes to PC improved both impact strength and solvent resistance [123]. Rubber modification of polyethylene is well recognized to improve the ESCR. A study of the variables showed that major improvements can be obtained with high molecular weight butyl rubber offering the best increase in ESCR [124].

A variation of toughened polystyrene involves the polystyrene-polybutadiene-polystyrene ABA block copolymers where compositions in the range of 30 wt% polystyrene show elastomeric properties. These polymers are referred to as thermoplastic elastomers. These polymers also exhibit ESC although only a few studies are noted in the literature [103, 125]. It was noted that cracking is most prevalent with solvents that prefer the polystyrene block. Measurement of the stress–strain curves for the unmodified block copolymer versus increasing levels of center block crosslinking showed major improvements in the ESCR (acetone environment) approaching stress–strain behavior in air at the highest crosslinking employed [103].

UNDERSTAND THE ENVIRONMENT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

With amorphous glassy polymers, many of the application failures involve adverse environmental contact. Often these environments are found after the material has entered production and in many cases after the product has been in commercial use. Technical personnel have to exhibit detective skills in elucidating the environmental problem. Several examples will be noted to demonstrate this. First, glass coffee containers had a serious breakage problem that presented both safety and economic concerns. A successful solution to this problem involved the use of PSF as a transparent top part of the container crimped around a stainless steel bottom to contact the heater plate. Initial prototyping and testing were successful and large scale production was started at a company in Hong Kong. The initial shipment of thousands of the molded composite containers was checked in Hawaii prior to shipment to the mainland US. Cracks were observed in the polymer around the crimped interface between two components which had not been observed in initial testing. The PSF technical service person was immediately sent to Hong Kong and quickly found the problem. The initial molding was done in a smaller (and more environmentally pristine) area without problems. Transfer to production involved proximity to metal cutting equipment and an atmosphere of aerosoled metal cutting fluids containing aggressive solvents for PSF. Production was transferred to a less aggressive environment and the application employing PSF was successful. The second example involved molded parts for dairy equipment. PSF offered excellent hydrolytic stability and the ability to resist the harsh cleaning environments for the application. It was noted by the customer that cracking failures were noted for some of the samples. Some had major cracks and others had minor cracks but most of the molded articles were fine. On a visit to the customer to observe the molding operation, it was realized that the intermittent use of mold release caused the problem. The first part after mold release application developed major cracks and the following part had minor cracks. Testing confirmed the stress cracking ability of the mold release and a change to a benign mold release resulted in a successful application.

There are a myriad of similar examples most of which have never been documented in the literature. However, many diverse environments have been mentioned in the trade literature, patents, and the open literature as illustrated in Table 2.

Table 2. Diverse environment ESC
PolymerEnvironmentCommentsRef.
PolystyreneHigh pressure CO2ESC not observed at low pressure CO2 (1 atm.)126
Nylon 6 and 6,6Inorganic salt solutionsAqueous solutions of zinc chloride and lithium halides127
PolycarbonateWindshield wiper fluidsSurfactant, monoethanolamine128
PolycarbonateButter and related chemicalsTriglycerides found to be more aggressive than butter129
PolycarbonateSunscreen (oils)Potential problem in automotive applications130
PolycarbonateSilicone oilESC observed with silicone oil of <410 g/mol molecular weight131
SAN copolymersFreon vaporCritical stress for craze initiation increased 4-fold for AN increase from 0 to 33 wt%132
PVCBenzene enriched natural gasSignificant decrease in crazing stress with >25,00 ppm benzene133
Polysulfone Poly(ether sulfone)Boiler additives (e.g., morpholine)Encountered in steam sterilization in hospital steam sources134
PolycarbonateCyanoacrylate adhesiveStress cracking at ≥0.4% strain135
PolytetraflurorethylenekeroseneNo measurable sorption136
PolycarbonateFat emulsion for parenteral nutritionProblem with intravenous infusion employing polycarbonate80

WATER CRACKING/CRAZING

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

A specific situation of crazing/cracking in specific glassy polymers has been noted where hot water exposure is encountered. The crazes/cracks which develop are usually disc shaped, are internal in the specimen, and can be described as microvoids [137–143]. For injection molded specimens, these microvoids tend to be more concentrated in regions of highest molded-in stress (as determined from polarized light observations). The studies showed that cyclic exposure (hot water/cold water) yielded much more pronounced microvoid formation. The microscopic observation of these microvoids showed a clamshell failure pattern with distinct areas of cavitation as illustrated in Fig. 10 for a single microcavity and in Fig. 11 for a higher magnification of the microcavity near the nucleation site. This behavior is particularly relevant for PC and poly(ester-carbonate) although has been observed for PEI samples.

thumbnail image

Figure 10. Microcavity from a fracture surface of poly(ester-carbonate) exposed to cyclic hot and cold water exposure (reprinted with permission from L. M. Robeson, B. L. Dickinson, and S. T. Crisafulli, J. Appl. Polym. Sci., 1986, 32, 5965, © John Wiley & Sons).

Download figure to PowerPoint

thumbnail image

Figure 11. Higher magnification from microcavity of Fig. 10 of a poly(ester-carbonate) microcavity near the nucleation center (reprinted with permission from L. M. Robeson, B. L. Dickinson, and S. T. Crisafulli, J. Appl. Polym. Sci., 1986, 32, 5965, © John Wiley & Sons).

Download figure to PowerPoint

This microcavity formation due to cyclic exposure (hot water/cold water) is hypothesized to be due to phase separation of water during cooling leading to localized (nucleation sites) of water separation [137–139]. This creates an internal stress which can promote stress-induced hydrolysis of polymers subject to hydrolytic chain scission. While this is much more prevalent during cyclic exposure, a similar explanation may be hypothesized during continuous exposure where the water clustering at a nucleation site leads to localized cracking resulting from stress-induced hydrolysis which relieves stress but continues as further water diffuses to the point of nucleation. Although PSF and polyether sulfone exhibit higher water sorption values than PC, microcavity formation does not occur as these polymers are very hydrolytically stable. It is important to note that “spiking” the PSF samples with 0.1 wt% NaCl can yield the observed microcavity formation [137]. Similar microcavity formation was noted with inclusion of water soluble salts (potassium acetate, cobaltous chloride, calcium nitrate) in crosslinked epoxy samples exposed to ambient temperature water immersion [142]. Internal pressure calculations indicted triaxial stresses of a magnitude capable to initiate crack formation to relieve the stress.

INJECTION MOLDING PARAMETERS: EFFECT ON ESCR

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

Although limited references are in the open literature, the ability of injection molded parts to withstand ESC in application end-use requires careful consideration. The part design, the elimination of sharp (high stress) edges, weld-line position, gate position, and design as well as the molding variables are all important variables affecting the performance of the part even without environmental exposure. The gate and the weld line can be positions where environmental attack may initially be observed. Exposure of molded parts to very aggressive environments is a rapid method to assess areas of concern. The author has observed situations where contacting the gate area (of a tensile bar configuration) with an aggressive solvent results in a crack rapidly spreading throughout the specimen carving out an internal tensile bar. The fracture path corresponded to the position of highest molded-in stress in the tensile bar. Molded-in stresses can often be quite significant yielding cracking/crazing without externally applied stress. These stresses can be reduced by annealing at temperatures approaching the glass transition temperature. Injection molding cycle time is important in the overall economics, and the use of rapid injection into cold molds leads to rapid production but high levels of molded-in stress and thus can be counterproductive when ESC failures are possible. The birefringence (photoelasticity) pattern for transparent glassy polymers (viewing under cross-polarized light) is a useful tool for assessing the level of stress, the areas of stress concentration in the molded part, and the effectiveness of annealing in reducing molded-in stress.

UNSOLVED PROBLEMS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES

The microscopic stress field associated with penetrant unsteady-state diffusion into a glassy polymer has not been appropriately determined. The diffusion of a penetrant into the glassy matrix will yield a volume expansion. As the polymer chain relaxation is considered much slower than the rate of diffusion, stress will be placed on the network analogous to an externally applied triaxial tensile stress although in this case the imposed stress is an internally induced stress. As volume expansion can be in the range of 10% or higher, it is not surprising that craze/crack formation can occur at the microscopic level (see Figs. 1 and 2). Once the penetrant concentration lowers the Tg to the exposure temperature, the relaxation process rapidly increases and the resultant gel layer (while greatly expanded) has an extremely low modulus relative to the glassy state. The gel layer will expand perpendicular to the surface without exerting any significant stress on the underlying layer. The sharp diffusion front of the glassy layer, however, exhibits a complex (and potentially large) and rapidly changing stress field with film position and time. Several cases are noted in the literature involving crosslinked polymers where the stress in the gel layer is sufficient to cause specimen cracking during the unsteady-state diffusion process [34, 144]. Extension of Case II transport studies to the higher Tg engineering polymers with experimental observation of crazing/cracking at the advancing diffusion front would be useful as the early studies were primarily conducted on polystyrene and PMMA.

The literature in the past decade has countless references to polymeric nanocomposites of which many involve exfoliated clay reinforcement. These references cover a multitude of structure–property relationships but studies discussing the ESC/crazing do not appear to be in the open literature. The ability of exfoliated clay to provide reinforcement and gas barrier properties to various polymers suggest that there may be benefits in reducing the environmental stress failure.

Another area virtually ignored in the open literature involves fiberglass reinforcement of thermoplastic polymers as an effective method for ESCR improvement. Definition of structure–property variables (fiberglass type, aspect ratio, fiber orientation, coupling agents, polymer matrix, and environment) would be useful. Studies of other fiber reinforcements (e.g., carbon fiber, natural polymer fibers, carbon nanotubes, and various nanofibers) would be relevant.

Impact modification has also been noted to be an effective method to improve ESCR. Although countless studies exist relating the structure–property variables to performance, very few discuss these variables relative to ESCR. Is the mechanism of rubber phase toughening analogous to the observed improvement in ESCR? When toughness is optimized, is ESCR also optimized?

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. LITERATURE REVIEW OF THE MECHANISM OF ENVIRONMENTAL STRESS FAILURE
  5. TEST METHODS
  6. APPROACHES TO REDUCE ENVIRONMENTAL STRESS CRACKING/CRAZING
  7. POLYMER BLEND TECHNOLOGY
  8. FIBER REINFORCEMENT
  9. IMPACT MODIFICATION
  10. UNDERSTAND THE ENVIRONMENT
  11. WATER CRACKING/CRAZING
  12. INJECTION MOLDING PARAMETERS: EFFECT ON ESCR
  13. UNSOLVED PROBLEMS
  14. REFERENCES