Planning reliability demonstration test with performance requirements

Abstract

Carefully planning reliability testing is necessary for success of a product development. Sample size determination is an important step in test planning. One of most common questions any reliability engineer gets asked is “how many units do we need to test?” Of course the answer depends on a number of factors and information in hand. The design of reliability demonstration testing (RDT) has been widely studied by many researchers. Most planning methods are based on the time-to-failure properties of product under test, such as the Binomial equation, the Chi-Squared formula, and etc. However, quite often, reliability test requirements are defined by certain performance characteristic(s), or physical characteristic(s), or quality characteristic(s). Very few works have been published to plan a reliability test with performance measurements directly. Some works on degradation measurements analysis, including degradation test plan for Wiener degradation processes, have been carried out for maintenance optimization purpose. This paper develops a test planning method for reliability demonstration test when the reliability test requirement is constructed with its performance or degradation measurements. A simple formula for sample size determination is provided in the paper, and a couple of case studies will be presented to help reliability practitioners to plan the reliability testing using the proposed method.