Reliability Estimation Including Redesign Following Future Test for an Integrated Thermal Protection System

1. Abstract Considering the effect of epistemic uncertainty on reliability at the design stage leads to a high probability of failure estimate. This probability is greatly reduced by post-design measures, such as redesign following tests. In this paper, we propose a methodology to accurately estimate the probability of failure at design stage by including the effects of future processes, such as tests and redesign. We expand a previous study applied to a single failure problem to multiple failure modes problem. An Integrated Thermal Protection System (ITPS) is one of concepts of the thermal protection systems for reusable vehicles, which is imposed to various failure modes simultaneously during atmospheric reentry. The proposed method is applied to an ITPS example, comparing with traditional deterministic approach. It is demonstrated that incorporating the effects of future redesign provides a significantly better accuracy of probability of failure estimate. Also, modeling probabilistic redesign brings a potential large mass reduction even after compensating for unconservative designs. These insights allow the designer to select a better initial design 2.