SRC Forum - Message Replies


Forum: Reliability & Maintainability Questions and Answers

Topic: Reliability & Maintainability Questions and Answers

Topic Posted by: Reliability & Maintainability Forum (src_forum@alionscience.com )
Organization: System Reliability Center
Date Posted: Mon Aug 31 12:47:36 US/Eastern 1998

Back to message list Show all replies Topics List About this forum
Original Message:

Posted by: Elizabeth M. Chappel (elizabeth_M._chappel@belcan.com )
Organization:Belcan Engineering
Date posted: Tue May 11 16:31:55 US/Eastern 1999
Subject: FMEA (CA) Definitions and Guidelines
Message:
I am establishing the guidelines and the definitions for severity (criticality), failure mode probability and detection for a System Level FMEA. The product has an aerospace application in which it is active at all times. Some vendors will be providing data (failure rates) for electronic components, while others may find it easier to provide probabilities of occurrences. However, there's the rub, how to define and correlate the two and make sense. I need to provide def'ns that make sense. The present military definition for Qualitative Probability isn't helpful, nor is the automotive. Help. What type of scale do you suggest for a product that is always active? (I have completed the Severity (Criticality) Definition, Detection won't be a problem, it's just this probability). The various sources I've viewed don't clarify it any. (old Mil-Std-1629, SAE-ARP-5580 advanced review copy). Thanks Beth


Reply:

Bruce Dudley Subject: FMEA Probabilities
Reply Posted by: (bdudley@alionscience.com )
Organization: Reliability Analysis Center
Date Posted: Tue May 18 14:18:26 US/Eastern 1999
Message:
Your question regarding a conversion factor or definition for quantitative to qualitative or reverse, is a significant problem and I don't have an instant conversion definition. The best way of course, is to require compliance to one way or the other by the vendors so that a complete FMEA can be performed. Given that this is not possible, you should consider the use of the Reliability Analysis Center's Non-Electronic Reliability Data Book as a quantitative failure rate source for those qualitatively expressed units/components. Also, our Reliability Toolkit has some equipment level quantitative figures of merit that could be used as a measure. If this is not sufficient or some items are missing form the data, you must assign a period of performance so that the unit qualitative measures can be converted on a failure mode by failure mode basis via a probability function to quatitative values. If you determine the time, the failure rates can be developed using a Weibull model for whatever failure distribution is expected.


Reply to this message
Next reply