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: Terry McCrossan
Date posted: Mon Jul 5 17:24:19 US/Eastern 1999
Subject: Component Derating
Does anyone have any new quantifiable data on the use of component parameter derating (thermal, power etc). I am used to the usual "margin of ignorance" type arguments, package glass transition temperatures and the junction/channel type Arrhenius relationships (generally with assumed activation energies). However, is there any more recent quantifiable data available to convince a sceptical design engineer and to what extent is derating used?
Reply Posted by: Seymour F. Morris
Organization: Reliability Analysis Center
Date Posted: Mon Jul 12 9:45:44 US/Eastern 1999
Based on a search of current literature, I was unable to find any "new" data on component derating neatly summarized for a wide array of electronic components. Much of the current guidance is based on studies performed by the Air Force during the 1980's, such as RADC-TR-84-254 (ADA153744), Reliability Derating Procedures. These studies often found there way into later guidelines such as MIL-STD-975C(NASA)-Appendix A: Standard Parts Derating Guidelines, probably with some modification. However, for the most part, the basic guidelines have not changed much over the years for many components. The underlying basis for the guidelines is from sources such as Reliability Physics Symposium (http://www.irps.org/). While technical papers in these symposiums often do not use the term "derating", their findings often can be used to estimate levels of risk associated with different levels of applied component stress. If you are interested in failure rate versus stress levels for a specific class of components, scouring proceedings of symposia such as IRPS can often be a valuable source of data that leads to answers. As you imply, the appropriate stress level to use is not a black and white issue. It really comes down to the critically of the system and the tolerance for failure one can accept.