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

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

Posted by: Stefano Stanghellini ( )
Date posted: Thu May 23 9:51:57 US/Eastern 2002
Subject: Reliability assumption
I performed a safety study where only some particulare failure modes can cause the hazardous situation. In particualar I have to evaluate what is the probability to have railway signal blinking due toa failure. The situation is dangerous only if the signal blinks with a frequency i a range between 0.1 to 10 Hz. In my analysis I identified some integrated circuit which can self oscillate due to a failure and cause the dangerous situation. The considered components are designed to work till a frequency of about some Mhz. Because I have no information available about the failure mode distribuition, in my analysis I consider that the probability to fail at a certain frequency is homogemeous. In this way for the hazardous failure rate I considered a ratio of the total failure rate. The ratio I took into account was equal to: (dangerous frequency)/1Mhz. Somebody ask me to justify this assumption because seems to be not conservative. Can anybody help me? Is there some data i leterature about this kind of failure?


Subject: Poisson Process
Reply Posted by: S. Nath
Organization: MercMarine
Date Posted: Thu May 23 14:53:10 US/Eastern 2002
The assumption of homogeneous/stationary can be applied only if the distribution of the number of events which occur in any interval of time depends on the length of the time interval only. The failure mode in your analysis is more likely to follow a non-stationary poisson process wherein the IC may self-oscillate more during a set of conditions-A than set of conditions B. I will try and simulate the conditions and use monte-carlo method to estimate the parameters.

Back to message list Topics List About this forum
Reply to this message