Bayesian Design and Analysis of Accelerated Life Testing with Step Stress

Abstract

Abstract : An accelerated life testing problem in which the stress s can take only a fixed, finite number of values is known as accelerated life test with step stress. In general, the item under test is started under the lowest stress first and if an item has not failed until a certain specified time, then it is moved into the next level of stress. This process is continued until the item fails. Thus the life test duration of the item is shortened. A Bayesian formulation of the problem is given in this paper. It is assumed that there only two stress values under consideration (1) standard use environmental condition (ii) a higher level of stress that is fixed in advance and is the same for all items to be tested. However, the time at which an item on test is taken out of use environment and put under higher stress environment can be chosen by the experimenter subject to a cost structure. We consider the inference and the optimal design problem of when the change the stress as the test progress. Keywords: Accelerated life test, Step stress, Optimal design, Bayesian approach.