Service life analysis from field data on age distributions

Abstract

In survey investigations of building stocks for example, data on age distributions of construction details are often collected. What interests the material scientist, though, is not the age distribution but rather the service life distribution. Presented herein will be a renewal process model describing the relationship between such an age distribution found and the corresponding service life distribution being sought. The model comprises a phenomenological parameter to accommodate correlations between objects and the service life’s determining factors. However, it is shown that, because of numerical ambiguity, the correlation parameter cannot be uniquely determined from an age distribution. Therefore, some additional information and data processing are required before the full advantage of this model can be realized. This somewhat difficult part has yet to be completed. Nevertheless, in its present form, the model does point out possible ranges of service life distribution parameters in terms of the correlation strength, thus providing uncertainties in the distribution parameters, which up until now have normally been overlooked when estimating the service lives of building materials. For wide service life distributions, it is shown that the possible parameter ranges will also the rather wide.