Ageing of cable insulators made of polyethylene in nuclear environment

This thesis deals with lifetime prediction for polyethylene in nuclear environment. It is mainly characterised by the search for a non empirical solution. We have tried to elaborate a model describing the polymer evolution (including the skin-core heterogeneity due to the kinetic control of oxidation by oxygen diffusion) at any temperature between ambient and the melting point of the polymer, and at any dose rate between 0 and 1 kGy.h-1, with lifetimes extending to several decades of years. The main difficulty was to take into account the dual character of initiation which results from the combination of polymer radiolysis and hydroperoxide thermal decomposition. The problem was resolved considering first the asymptotic regimes where one initiation process can be neglected relatively to the other one. The kinetic parameters specific to those regimes being identified, we have built a numerical model including all the processes. In the case of un-stabilised polyethylene, this model displays excellent predictive qualities, as well in lifetime, as in thickness distribution of oxidation products. The study of stabilised polyethylenes was then initiated with samples respectively stabilised by a hindered phenol, and a thio-diester. In the case of phenol, the study of consumption kinetics reveals the existence of complex mechanisms, especially the existence of a 'reservoir effect' linked to the presence of a separated phase of phenol in excess. A kinetic model has been proposed, but we still do not know how the results can be generalised to other phenolic stabilizers. (author)