Service Life Prediction from Accelerated Aging Test Results Using Reliability Theory and Life Testing Analysis

Abstract

The prediction of the service life of a building product is needed for making life cycle cost decisions, for determining the risks and liabilities in marketing a product, and for helping to identify the causes of a product’s failure. Accurate estimates of the service life for building products, however, have traditionally been difficult to obtain. The most timely source for obtaining service life estimates for a building product is through the use of accelerated aging tests. Unfortunately, most accelerated aging tests for building materials are not capable of providing quantitative estimates of a product’s service life since they were only designed to screen out bad products or to make qualitative comparisons between the service lives of different products. To rectify this situation, accelerated aging test procedures used in high technology industries were applied to building products. The emphasis in these procedures is in the mathematical analysis of the life data. One of the more successful mathematical procedures is reliability theory and life testing analysis. In this paper, the basic format of this analysis procedure is outlined and several applications are presented in which these techniques were applied to predicting the service lives of different building materials and products. The results of these analyses are very promising and it is concluded that reliability theory and life testing analysis techniques should be readily adaptable for predicting the service lives of a wide range of building products and materials.