Plastic Design & Material Selection

Accurate material selection is a critical factor in eliminating warranty & recall risk from delayed cracking in plastic products & components

More than 5,000 plastic product failures have been the subject of study at Smithers Rapra. These have been classified on the basis of primary failure mode as shown in Figure 1.0. A further breakdown of plastics product failure due to human causes is given in Figure 2.0 of which 45% are due to material mis-selection & poor material specification.

Figure 1.0 Material / phenomenological causes of plastics failure

Figure 2.0 Human causes of plastic product failure

Material Selection

Failures due to poor material selection are very common in the plastics industry & appear to be related to a lack of awareness & understanding of plastic properties. When, considering the design & development of a plastic component it is imperative that designers & engineers fully understand the fundamental nature of plastics & that:

• Plastics are non-linear, visco-elastic materials
• Plastics have temperature & time dependant properties
• Plastics age physically & chemically
• Plastics are susceptible to chemical attack & environmental stress cracking (ESC)
• Plastics are susceptible to weathering
• Plastics will, under the action of a tensile stress, eventually fail
• The time to failure will diminish:
  • as the stress increases
  • as the temperature increases
  • in the presence of certain environments
  • under the action of cyclic loading
• The moulding process can result in significant levels of moulded-in (residual) stress
• Weld lines due to converging melt fronts result in planes of weakness, particularly in fibre reinforced materials
• Plastics exhibit notch sensitivity
• The addition of any form of filler will always have some form of detrimental effect on a material
• Plastics can show mechanical anisotropy due to the alignment of fibre reinforcement

The problem for designers & engineers is further exacerbated due to the diverse range of plastics available. There are over 90 generic classes of plastic, which can be broken down into approximately 400 sub-generic modifications & finally tens of thousands of commercial grades available from hundreds of suppliers.

When approaching the task of material selection polymer structure & polymeric properties must be considered. Polymers can be broken down into three main groups;

• Thermosetting Plastics
• Thermoplastic Plastics
• Elastomers (Rubbers).

For plastics products, there are two basic choices, Thermoplastic or Thermoset. Thermoplastics are broken down into two groups – amorphous & semi-crystalline.

Amorphous plastics have no ordered structure; semi-crystalline plastics have areas of order, which form crystallites. Generally, amorphous plastics exhibit improved creep resistance - constant static load over time; semi-crystalline materials provide better fatigue, chemical, & environmental stress cracking (ESC) resistance.

Amorphous plastics are likely to be the best option when the following properties are required:

• Good material transparency
• Low mechanical abuse
• Limited or no chemical contact

Semi-crystalline materials should be selected when resistance to the following is required:

• Chemical ESC resistance is required
• High Mechanical abuse is anticipated
• Cyclic loading

Plastic product failure

A significant number of plastic products fail before their design life expectancy. Some of these failures can be traced to designers & engineers who historically been accustomed to working with materials that exhibit a predictable, linear elastic stress-strain relationship. Polymers behave differently as they are visco-elastic materials & respond to stress as if they were a combination of elastic solids & viscous fluids. Plastics exhibit a non-linear stress-strain relationship & their properties depend on factors such as the time under load, temperature, environment & the stress or strain level applied.

When using Finite Element Analysis (FEA) for elastic materials the software requests the modulus (stiffness) of the material. For metals the tensile modulus of the material is simply entered (depending on the loading scenario) & the FEA software calculates the result. When designing with elastic materials, designers & engineers have the option to rely on instantaneous stress-strain properties, & for most applications can disregard the effect of temperature, environment & the long-term effect of load. In most applications plastic products are designed to last for longer than the 20 seconds it takes to generate a stress/ strain curve with a tensile machine. A product may have to operate for 20 years, under load at 40C in a chemical environment!

For plastics simply inputting tensile modulus into a FEA programme at 20C will ensure the product will be under-designed, the modulus at 20 years is likely to be significantly lower than the instantaneous modulus value. In addition, the stiffness at 40°C may be significantly lower than that at 20C. Reliance on short-term data will likely result in premature failure & the costs associated with, product recall, product replacement, loss of brand credibility, litigation & re-tooling.

In order to avoid under designing, properties such as creep rupture strength & creep modulus must be determined experimentally under worst-case scenario operational conditions. Testing should include maximum service temperature, maximum stress & environmental effects.

Material Creep

Creep is recognized as the continued deformation of a material over time under a static load & is a function of the viscous nature of the polymer. All unreinforced thermoplastics exhibit significant creep properties demonstrated in the creep curves for polypropylene (PP) in the figure below. Polymers are visco-elastic & time dependency needs to be taken into account. For PP, the creep modulus decays rapidly under load giving values of approximately 50% of that calculated from a tensile modulus short-term curve.

Creep modulus is the value that should be used in FEA computations. An advantage of designing with metals is that, with the exception of high temperatures this value will not change over time. With polymers the creep value should be calculated at service temperatures, in the appropriate environment & under a specific load.

A problem highlighted by FEA might be that the level of stress may appear low compared to the short-term tensile strength of the material. This could be misleading & give a false sense of security. Designers should be aware that dynamic fatigue (cyclic loading) of polymers can result in a ductile to brittle transition, resulting in failure stresses significantly lower than the short-term tensile strength.

With amorphous polymers this is even more important as shown by the S-N curve (stress versus cycles to failure) generated for polycarbonate (PC) below.

The S-N curve clearly demonstrates that in fatigue, the strength of PC after 1 million cycles is reduced massively from 60MPa (reported from supplier short-term tensile strength data) to only ~10 MPa, a reduction of approximately 83%. Using a design safety factor of two in a stress calculation, an operating stress of 5 MPa for this material would be required, only one twelfth of its short term tensile strength.

In contrast, a lower strength material such as semi-crystalline PP does not experience a dramatic ductile – brittle transition. It can be seen from the S-N curve that after 1 million cycles PP's failure stress has fallen by ~50% to 12MPa. Applying a design safety factor of two gives an operating stress of 6MPa. Consequently, the lower strength PP demonstrates better fatigue endurance in the long-term compared with PC.

When looking at strain, many polymers can endure levels of 200% or more, for long-term performance the window for design strain is massively smaller. Recommended design strains are as follows:

Static stress conditions

• Amorphous plastics ≤ 0.5% strain
• Semi-crystalline plastics ≤ 0.8% strain

Cyclic stress conditions

• Amorphous plastics ≤ 0.3% strain
• Semi-crystalline plastics  0.5% strain

Concern might exist at the reduction in mechanical properties of polymers in the long-term. If good, reliable long-term performance data is generated & material behaviour within a specific service time frame is understood, designers can take appropriate action & compensate accordingly.

Many designers & engineers only use short-term test data provided by the plastic manufacturer's datasheets & are unaware of the long-term properties of plastics. Utilisation of this short term data results in the thousands of product failures seen every year at Smithers Rapra, products have been designed without consideration for load, time, temperature & chemical environment.

It is recommended that designers & engineers only use data sheets for comparing property values of different plastic materials & should be considered only as screening tools. Data sheets are not intended to provide information for design & final or ultimate material selection. Information from short-term tests provide single point measurements without considering the effect of time, temperature, environment & chemicals. It should also be noted that test pieces are also simple in shape & are moulded & possibly tested under ideal conditions. This rarely applies to products which are likely to be of complex geometry & have potential structural weaknesses from the design & manufacturing process.

Product designers should have a working understanding of Environmental Stress Cracking (ESC). ESC is the premature initiation of failure & apparent embrittlement of a polymer under the simultaneous actions of stress & the environment. The presence of both a stress & the chemical environment can lead to dramatic effects, mainly catastrophic brittle fracture of even highly ductile materials such as polyethylene (PE). These failures are unlikely to be observed in standard chemical resistance testing where unstressed specimens are immersed in a chemical for a time period with a measurement of properties before & after exposure (typically tensile & impact properties).

Amorphous plastics are, in general, more susceptible to ESC than semi-crystalline materials but Smithers Rapra recommend testing in the appropriate chemical environment when possible.

Many plastic products are stressed in service from moulded-in stresses created in the manufacturing process alone - injection moulding, extrusion, & thermoforming, in addition to thermal stresses & applied loads. ESC should be a consideration for any designer working with plastics.

Smithers Rapra regularly assist with product design & material development to ensure a product performs as anticipated in its given application. The Consultancy Centre has a knowledge bank of how plastics materials perform in practice & is based on many years & many thousands of actual industrial examples. This expertise provides impartial independent assessments of the suitability of polymeric materials for a given application & to highlight where conflicts of requirements might affect long term performance. Design services include FEA, material selection, mould flow, long-term property generation & lifetime prediction services.

Smithers Rapra has worked with clients to characterise materials for long term design - up to fifty years using time-temperature superposition. Smithers Rapra has fully equipped long term creep & fatigue laboratories to operate at extremes of temperature & in aggressive chemical environments. When considering the cost of mould tooling or tool re-designs, product recall & litigation due to premature failure this type of support is invaluable.