4.1 The use of this practice presupposes that the failure criteria selected to evaluate materials (that is, the property or properties being measured as a function of exposure time) and the duration of the exposure can be shown to relate to the intended use of the materials.

4.2 Plastic materials exposed to heat are subject to many types of physical and chemical changes. The severity of the exposures in both time and temperature determines the extent and type of change that takes place. A plastic material is not necessarily degraded by exposure to elevated temperatures. However, extended periods of exposure of plastics to elevated temperatures will generally cause some degradation, with progressive changes in physical properties. Specific properties and failure (or lifetime) criteria for these properties are typically chosen for the evaluation of thermal endurance.

4.3 Generally, short exposures at elevated temperatures drive out volatiles such as moisture, solvents, or plasticizers, relieve molding stresses, advance the cure of thermosets, and may cause some change in color of the plastic or coloring agent, or both. Normally, additional shrinkage should be expected with loss of volatiles or advance in polymerization.

4.4 Some plastic materials become brittle due to loss of plasticizers after exposure at elevated temperatures. Other types of plastics become soft and sticky, either due to sorption of volatilized plasticizer or due to breakdown of the polymer.

4.5 The degree of change observed will depend on the property measured. Different properties, mechanical or electrical, may not change at the same rate. For instance, the arc resistance of thermosetting compounds improves up to the carbonization point of the material. Mechanical properties, such as flexural properties, are sensitive to heat degradation and may change at a more rapid rate. Ultimate properties such as strength or elongation are more sensitive to degradation than bulk properties such as modulus, in most cases.

4.6 The material studied can change inherent behavior with change in temperature as for example when crossing α, β, and γ transitions. These transitions should be avoided both in the range of aging temperatures used, as well as in extrapolation of the lifeline. Arrhenius principles may only be used to accelerate a chemical mechanism if there are no fundamental changes in the material properties. With semi-crystalline and highly crystalline polymers, elevated temperatures may cause significant changes to the morphology of the material, invalidating or compromising that assumption.

Note 2: Caution should be exercised in using the Arrhenius relation and knowledge of physical changes in the material at elevated temperatures is important. Guidance given in ISO 9080 for characterizing lifetime of plastic materials in pipe form by extrapolation suggests that the highest oven aging temperature should be at least 15°C lower than the Vicat softening temperature for glassy amorphous polymers, and at least 15°C lower than the melting point for semi-crystalline polymers.

4.7 Effects of exposure can be quite variable, especially when specimens are exposed for long intervals of time. Factors that affect the reproducibility of data are the degree of temperature control of the enclosure, humidity of the oven, air velocity over the specimen, and period of exposure. Errors in exposure are cumulative with time. Certain materials are susceptible to the influence of humidity.

4.8 It is not to be inferred that comparative material ranking is undesirable or unworkable. On the contrary, this practice is designed to provide data which can be used for such comparative purposes. However, the data obtained from this practice, since it does not account for the influence of stress or environment that is involved in most real life applications, must be used cautiously by the designer, who must inevitably make material choices using additional data such as creep and creep rupture that are consistent with the requirements of the specific application.

4.9 It is possible for many CUT and TI values to exist. Therefore, for any application of the CUT or the TI (temperature index) to be valid, either the thermal aging program must duplicate the intended thermal exposure conditions of the end product, or the Arrhenius relation must apply.

4.10 There can be very large errors when Arrhenius plots or equations based on data from experiments at a series of temperatures are used to estimate time to produce a defined property change at some lower temperature. This estimate of time to produce the property change or "failure" at the lower temperature is often called the "service life;" however, using this term should be avoided as this implies the tester has information on specific failure criteria in end-use, while numerous factors are not under the scope of this test. It is preferable to use terms such as "end point," "thermal endurance time," and such. Because of the errors associated with these calculations, this endurance time should be considered as "maximum expected" rather than "typical."