Testing for Environmental Stress Cracking

Smithers Rapra has extensive plastic testing experience & have gained a reputation as one of the leading test houses in the country for ESC testing.

Environmental Stress Cracking is the premature initiation of cracking & embrittlement of a plastic due to the simultaneous action of stress & strain & contact with specific chemical environments.

Smithers Rapra has developed rational test methods, which have proven themselves as invaluable methods for determining the compatibility for plastics & chemical environment pairs.

Smithers Rapra provide ESC consultation to both limit the occurrence & also offer testing services to find the route cause of the condition. Our services are often used by end users to discover failure & for a solution.

Table 1.0 Common Amorphous & Semi-crystalline Thermoplastics

Amorphous Plastics	Semi-crystalline
ABS	Acetal
ASA	Fluorocarbons
Cellulosics	Nylons
SAN	Polyethylene’s
MBS	Polypropylenes
Noryl	Polybutylene
PC	PETP
PEI	PBT
PMMA	PPS
Polysulphone	PEEK
PES	EVA
PS	Polymethylpentene

The high frequency of ESC failure suggests a widespread lack of awareness & understanding through out the polymer supply chain & end-user. In many instances people in the chain may make changes or introduce new steps in order to increase productivity or improve product performance, the consequences of which may have deleterious effect on product performance. Smithers Rapra have found that many ESC failures have been due to the introduction of a ‘secondary’ environment by a third party, which was not anticipated at the design stage. Examples of secondary environment include:

Adhesives	paints
lacquers	lubricants
cleaning agents	aerosol sprays
anti-rust agents	leak detection fluids
vegetable oils	fruit essences
inks	plasticisers

In contrast ESC failures due to contact with ‘primary’ environments i.e. those which will purposely make contact with product due to immersion or containment as in bottles, vessels, tubes & pipe, are comparatively rare.

Smithers Rapra have investigated hundreds of ESC failures. We have found that the source of stress is often that which is unavoidably processed in during moulding, & therefore applied stress is not a pre-requisite.

The effects of ESC can range from the merely cosmetic to the catastrophic & life-threatening. It is a failure mechanism, which contributes to many industrial & domestic accidents with substantial costs to industry. The following examples demonstrate the breadth of the problem & need for awareness, vigilance & control at all design & manufacturing stages.

• Injection moulded polycarbonate (PC) was used for the production of crash helmets since this super tough material promised excellent protection. However, in practice its performance fell well short of expectations. The underlying problem was traced to the fashion of adorning helmets with adhesive stickers & solvent based paint. In contact with these fluid environments the nominal low processed-in stress was sufficient to induce micro-cracking. This unanticipated secondary fluid contact compromised helmet structural integrity rendering them fragile under impact conditions.
• Rotational moulded polyethylene static oil tanks with an expected lifetime of 20 years (guaranteed for 10 years) were found to fail prematurely after only 2-3 years service. The contained kerosene environment was found to be an ESC agent for the polyethylene grade used.
• The clear polystyrene (PS) eyes of a soft toy became ‘milky’ as if the subject was suffering from cataracts. The eyes were held in place by metal circlips, which were coated with metal cutting fluid. The combination of assembly stress & fluid resulted in crazing, which interrupted light transmittance resulting in a milky appearance.
• A polyvinyl chloride (PVC) pressure pipe failed at the cement bonded joints because the solvent in the cement had not been given sufficient time to evaporate before the joints were pressure tested.
• Acrylonitrile butadiene styrene (ABS) consoles cracked due to plasticiser migration from the PVC wire insulation
• Heavy duty polyethylene packaging failed due to silicone oil on an ‘O’ ring seal
• Polycarbonate milk feeding bottles spontaneously cracked due to insect spray.
• Polycarbonate child pacifiers spontaneously cracked due to a combination of high stress at a heat weld & action of cleaning fluids.
• Machinery fabricated from a variety of metal & plastics parts was sealed in a container for transport. To protect the steel parts, a rust inhibitor containing organic amines was added. The vapour inhibited rust but stress cracked many plastic parts.
• Multiple cracking of PC sheet used in the fabrication of a bullet proof cashier window. Cracking was caused due to residual stress within the PC sheet, assembly stress on fixing & unauthorised use of commercially available cleaning agents.
• Multiple craze development within an acrylic (PMMA) medical component. Crazing was caused by residual stress within the moulding & the ESC nature of the adhesive bonding system.

Modest levels of stress applied over long periods of time induce purely ‘mechanical degradation’ in the form of crazes & cracks. This is the underlying cause of the long-term transition from ductile to brittle behaviour for ductile plastics as shown in Figure 1.0.

Figure 1.0 Time dependent creep rupture strength of an amorphous plastic showing ductile to brittle transition

This mechanism, known as creep rupture or static fatigue, is a common cause of product failure & is easily overlooked in favour of assumed environmental effects.

Many factors will accelerate this naturally occurring embrittlment process including:

• stress concentration (notch sensitivity),
• increase in temperature,
• cyclic loading (dynamic fatigue)
• contact with specific chemical environments (ESC).

The ESC mechanism in simple terms is a physical interaction involving highly localised plasticisation via stress enhanced fluid absorption at stress concentrating defects, which does not involve chemical change or molecular degradation of the plastic.

In order to fully understand ESC mechanism, consider that on a micro-scale polymer surfaces contain a high density of stress concentrating defects. There is no consensus as to the nature of these defects other than surface flaws, chance defects, or local or perhaps characteristic fluctuation in the materials microstructure. Typically these defects range in stress concentration factor (SCF) from 1 to 50. Local yielding will occur at these defects if the product of the applied stress & the SCF exceeds the yield strength of the material. When subjected to high stress a high density of locally yielded sites will form, which will then grow & multiply with time under stress due to the time dependent reduction in yield strength. This results in a coalescence of yielded sites & eventual microscopic yield failure of the material. However, at low applied stresses only small number of severe sites will be micro-yielded. These sites will grow slowly in the plane normal to the principal applied stress direction with negligible chance of coalescence with other sites in close proximity. These planar yielded zones cavitate & fibrillate to become crazes, which eventually reach a critical length for unstable crack growth. Consequently high stress promotes early ductile failure & low stress promotes delayed brittle failure as shown in Figure 1.0.

Absorption of any fluid plasticises the polymer & reduces its yield strength. The greater the concentration of absorbed fluid, the lower the yield strength. The tips of flaws under stress absorb more fluid than the surrounding material. Consequently the yield strength of the material is the lowest at points of highest stress. This geometrical coincidence tends to reduce the diameter of the yielded zone with little blunting of the stress concentration. The stress concentration in front of the initial yielded zone remains high, fluid is locally absorbed & the zone grows over a thin plane normal to the applied stress until crazing & fracture occur.

As most fluids have a greater affinity for plastics than air, then most fluids (including water) will accelerate the embrittlement process. Thus it may be assumed that in terms of their effect on a particular plastic all fluids may be described as stress cracking agents. The chances of tracking down any experimental data on ESC for a given fluid/plastic pair are generally remote. This is because there are tens of thousands of fluid chemicals & mixes & many generic plastics. Consequently given the common query “what plastic is safe to use in contact with particular fluids?” the only practical options are prediction or testing. From the available published data & the incidences of ESC, the following generalizations can be made:

• Amorphous thermoplastics are considerably more prone to ESC than either semi-crystalline thermoplastics or thermosets. The ESC resistance of semi-crystalline polymers is attributed to regions of closely packed crystals, which act as barriers to fluid penetration & as crack stoppers. Significant research has concentrated on the ESC of polyethylene, a semi-crystalline plastic. The molecular architecture, density, length & degree of entanglement of intercrystalline (tie) molecules, have been found to have a profound influence of ESC resistance. For polyethylene’s this research has allowed molecular architecture to be optimised to give enhanced ESC performance. Unfortunately in the case of glassy amorphous thermoplastics there are no tie molecules, & therefore options to improve ESC resistance are restricted.
• Fluids with modest hydrogen bonding are most likely to be severe or moderate stress cracking agents. Typically these include organic fluids classed as aromatic hydrocarbons, halogenated hydrocarbons, ethers, ketones, aldehydes, esters & nitrogen / sulphur containing compounds. Non-hydrogen bonded fluids such as aliphatic hydrocarbons & highly hydrogen bonded fluids such as water & alcohols tend to be milder in this respect.
• Fluids with a high molar volume, which tend to have high viscosity & high boiling points are less likely to be severe stress cracking agents.
• Fluids are most aggressive near to their boiling point.
• Amorphous plastics are most susceptible to ESC at temperatures that approach their glass transition temperature (Tg).
• Polymers with low molecular weight (high melt flow index (MFI)) suffer from reduced ESC resistance.

Types of testing available which are provided by Smithers Rapra to assess ESC compatibility include:

1. Bending beam tests (single cantilever, 3-point bending & Bell Telephone test) – Faced with the situation one or more candidate plastics & a number of fluids, various inexpensive tests can be applied to eliminate plastic / fluid pairs.
2. Tensile creep rupture – this is the most direct method of quantifying the effect of a fluid on the durability of a plastic material. It involves the application of a tensile stress & recording the time to rupture. As demonstrated in Figure 2.0, in air the time to rupture increases gently as the applied stress decreases. A safe allowable stress would be one half the stress required to cause rupture at the expected service life. A stress cracking fluid may have little or no effect at high stresses & short rupture times. However, in the medium to longer term there will be a massive reduction in strength & therefore a massive impairment of durability.

Figure 2.0

1. Tensile creep – If the tensile strain of a material in air & contact with a stress cracking fluid are monitored prior to creep rupture the influence of the fluid can be detected at shorter times as shown in Figure 3.0. The departure in creep response occurs typically at one tenth of the time for departure in the creep rupture response. The increase in creep rate coincides with crack / craze initiation & as such can be taken as a non-subjective criterion for safe application (allowable stress & allowable service life). In many applications, for example the crash helmet, the initiation damage is more rational criterion than actual rupture.

Figure 3.0

Summary

From our experience of ESC failure at Smithers Rapra we advise designers to consider most carefully the potential for ESC at all stages of the supply chain & to implement measures to ensure tight control & appropriate end-user warnings. Furthermore, we strongly support material & product exposure testing in all potential environments as part of any product development programme. Testing of this nature is a small price to pay when the costs of product recall, litigation & loss of reputation are taken into account.

Dr Chris O’Connor is the Technology Manager at Smithers Rapra. The company provides a complete range of services including materials selection, FEA, mould flow, long term property generation & lifetime prediction services to assist with all stages of product development through to failure diagnosis. Contact Dr O’Connor directly on 01939 252423 or via email at: coconnor @rapra.net