PRISM Forum - Message Replies
Topic: PRISM Questions and Answers
Topic Posted by: SRC
Date Posted: Wed Jan 12 8:33:33 US/Eastern 2000
Topic Description: Welcome to the PRISM forum! Please feel free to post your questions and comments about the PRISM assessment software here.
Posted by: Dave Ellis
Organization:Ellis Quality Engineering Services
Date posted: Tue Oct 10 11:50:29 US/Eastern 2000
Subject: System/Assembly level failure rate roll ups
I am in the process of generating a medium sized prediction and have noticed that the system or assembly level failure rate summarizations don't agree with manually calculating the total failure rates. For example, one assembly has 2.379, another 0.094 and another 0.065, but the overall system failure rate is only at 0.589 and the system level hardware component is less than the total sum. Comments?
Subject: System/Assembly Level failure rate roll ups
Reply Posted by: Norman B. Fuqua
Date Posted: Tue Oct 10 13:36:54 US/Eastern 2000
Subject: PRISM System Level Model
Message: I think that what you are referring to is an error in a previous version of PRISM that has since been corrected.
The PRISM system level model uses a process grade factor to account for the factors contributing to system reliability. This process grade factor is applied at each level of assembly during a system analysis. Currently, if an assembly is part of a system or is a subassembly, a process grade factor is reapplied to the assembly modifying its failure rate. The intent of the PRISM process grade is to assess the manufacturer’s process associated with each part of the system. Therefore, once a process grade has been established for an assembly or system, the resultant failure rate should not be modified further by other process grade factors. When multiple levels of assemblies are incorporated in PRISM, assembly failure rates are not being assessed properly. With this in mind, PRISM is being modified as follows:
The failure rate of an assembly, with no subassemblies, is determined by summing the failure rates of the component hardware and applying the associated process grade factor to this value. The software failure rate is then added, which results in the total failure rate for the assembly. If this assembly is integrated into another assembly or into a system, the process grade of the higher assembly or system is not reapplied to this “subassembly” failure rate. The process grade factor is only applied to hardware components that directly reside beneath the current assembly.
The values associated with the PRISM process grade factor have not changed. The PRISM defaults for the process grade factor and environment were normalized to approximately unity for an environment that has an operating temperature of 30oC, a dormant temperature of 14oC and a vibration level of 1. Therefore, selecting these values will approximate a PiE of 1 and a process grade factor of 1 in PRISM.
This PRISM modification will be released in Version 1.2 scheduled for the Fall of 2000.