SRC Forum - Message Replies
Forum: Reliability & Maintainability Questions and Answers
Topic: Reliability & Maintainability Questions and Answers
Topic Posted by: Reliability & Maintainability Forum
Organization: System Reliability Center
Date Posted: Mon Aug 31 12:47:36 US/Eastern 1998
Posted by: Dwayne
Date posted: Sat Aug 31 0:48:23 US/Eastern 2002
Subject: Accelerated Reliability Demonstration
Is it a good idea to perform an accelerated RD? Originally our plan is to perform it over a span of 5 years but i intend to compressed it into 2 yrs instead so that failures can be surfaced quickly. Any comments from RD point of view? thanks
Subject: Accelerated Testing
Reply Posted by: B.W.Dudley
Date Posted: Tue Sep 3 14:18:48 US/Eastern 2002
There are no magic analytical models that accurately estimate the life of complex assemblies. Each life analytical model describes physical change mechanisms associated with specific material characteristics. The first step in constructing accelerated tests is to define the failure mechanisms in terms of the materials used in the product to be tested. The next step is to determine the environmental parameters the product will be exposed when operating and when not operating or stored. Based on the failure mechanisms most likely to limit the life of the product, one can choose a test or tests that will accelerate that failure mechanism. Some of the models that can be considered are: Miner's Rule for accumulated fatigue, Arrhenius Temperature test for accelerated temperature and chemical aging effects, Peck's Model for temperature and humidity combined effects, Coffin-Manson non-linear mechanical fatigue damage model and Eyring/Black/Kenney models for temperature and voltage acceleration.
In the 2001 Reliability and Maintainability Symposium, Pantelis Vassiliou and Adamantios Mettas presented a complete tutorial on "Understanding Accelerated Life-Testing Analysis" that is worth reviewing as they developed a presentation on the problems and pitfalls of accelerated testing. Inaddition to the tutorial, J.D. Seager of the IBM corporation published a paper on " A Method to Predict an Average Activation Energy for Subassemblies" (IEEE, Transactions on Reliability, Vol. #37, No.5, 1988) that could be considered in the search for a single test value. This approach uses a combination process to average the expected component activation energies. In any event, the only tried and true method is to perform multiple high stress experimental tests to determine the real activation energy for the product using plots of the data.