LIFETIME PREDICTION IN THE HYDROLYTIC AGING OF POLYESTERS

Linear and cross-linked polyesters are subject to hydrolytic ageing. Ester hydrolysis leads to chain scission and thus to embrittlement. We attempt to review some problems of lifetime prediction in this field. We define first an ideal case of hydrolysis in which the chain scission is non-autocatalytic, non-diffusion controlled and purely random. For linear polymers, we can use non-arbitrary end-life criteria based on the well-known relationship between the ductile-brittle transition and the entanglement limit of molecular weight. For crosslinked polymers, such criteria cannot be used owing to the initially brittle character of the material. Crosslink density measurements can be used to evaluate the degradation rate, but spectrometric measurements of the ester concentration are recommended in the range of high conversions. Non-ideal cases are reviewed. Many reasons can explain the departure from ideality: diffusion control of hydrolysis kinetics, heterogeneity of semi-crystalline polymers, variation of the hydrophilicity with the hydrolysis conversion, etc. Some questions about the morphological aspects remain unanswered. Other chemical reasons can be involved: autocatalysis, which is still controversial, and non-random chain scission, which seems especially important for unsaturated polyesters.