Forum: Reliability & Maintainability Questions and Answers
Posted by: Marina Karyagina (email@example.com )
Date posted: Thu Jun 15 1:35:56 US/Eastern 2000
Subject: availability modelling
I have a straightforward 3 pumps running in parallel system. The system is UP if at least 1 pump is working. I need to find Availability of the system.
There are formulas, I know that. Here is the complication: the only data I have is that the pumps last at least 500 hrs between repairs (from the vendor). So what we know is more or less useful life=500hrs.What we do not know is the hazard rate during useful life. So we bring a pump down every 500 hrs for an 8 hr (scheduled or un-scheduled) repair. I believe, 3*(1/500) is a good enough estimate of the rate with which the system goes into the '2-pumps UP' state. Now, during the repair we have a chance of the second pump failure. We can devise our maintenance strategies so that while the first pump is repaired the other two are not nearing their useful life. So, using 2*(1/500) as a transition rate from '2-pumps-UP' into 'one-pump-UP', I believe, will result in a gross under-estimation of system's availability. So, there is a question: is there a rule of thumb, estimate, guesstimate or else I can use to estimate the hazard rate during the useful life period from known useful life? MTIA
Posted by: Jim Neil
Date posted: Wed Jun 14 18:30:06 US/Eastern 2000
Subject: Failure rate
What're the typical FITS for Si FET, Bipolar, and Diodes?
Posted by: Matt Wiltshire (firstname.lastname@example.org )
Date posted: Tue Jun 13 8:42:07 US/Eastern 2000
Subject: Salt Spray Testing
I'm looking for a realistic test to qualify a supplier's powder coat paint process over cast aluminum. We currently require scratch, dent and ASTM B117 salt spray test (1000 hours), but the salt spray seems too much. Is there a more realistic qualification, perhaps SAE J2334?
Thanks, Matt Wiltshire
Posted by: Kang
Date posted: Mon Jun 12 19:40:25 US/Eastern 2000
Subject: Reliability Prediction of FET, Bipolar, Diodes, etc.
1. Are the reliability prediction results big different between using MIL-HDBK-217F notice2 and using Bellcore? I just try MIL. Using Bellcore can get "better" MTBF?
2. Except the above two, are there any other methods to predict the reliability of electronic components? Thanks in advance! Kang
Posted by: Johnson Wood (email@example.com )
Date posted: Sat Jun 10 2:54:04 US/Eastern 2000
Subject: Software Reliability Prediction
A software reliability prediction method is presented in the RAC publication NEW SYSTEM RELIABILITY ASSESSMENT METHOD.I have some questions about the table 2.2-21 in page 68:
1.What are the exact meanings of Fault Activation and Fault Latency in the box I and how to determind the values? 2.How to calculate the Inherent MTBF in box III?