Discussions
Categories
Groups
Documentation
Knowledge base
Developer portal
App catalog
The Hub
GitHub
Home
Newly Imported
Restricted
ZF
Dangerous undetected failure rates propagation from FMEDA to FTA
medini_1028038066
Status : feature requested with Project Owner Priority : Major, #1 Schedule : (according to Project Owner) Current state : not available Critisism on current state : all the data is available in medini and this could be supported in the native tool implementaiton Objective of the change : faster and efficient workflow Proposed change : We wish for a semiautomatic approach to propagate each HW effect with its assigned safety mechanisms and diagnostic coverage to the FTA. Original author: Peter Ploß: "NEW: How to efficiently propagate the dangerous undetected failure rate from FMEDA to FTA? We wish for a semiautomatic approach to propagate each HW effect with its assigned safety mechanisms and diagnostic coverage to the FTA. Hint: This shall be done independently from safety goals as we want do derive the classification from the FTA. We have multilple FTAs, one for each safety goal. OLD: failure rates of a hardware part are not propagated to another "collecting" failure mode?! (example: "INTERLOCK: No function" that inherits "short circuit" from C1000 and "reduced value down to 0,5" from C1001 with 0.018 and 0.451 FIT respectively); it should be the sum of the allocated causes' failure rates, line 23"
Find more posts tagged with
medini
medini legacy
Accepted answers
medini_1028038066
Dear Peter, due to your request we have opened an internal ticket for review from the Product Management Team. The requested functionality will not be provided, we will not allow a direct connection between FTA to FMEDA, but need to go through the model.When we move to Component Fault Trees (CFT), they will essentially contain faults/events together with the SM/diagnostic, where the value you are looking for will become part of the model and not only of the table as of now. We will keep you informed when the CFTs will be released.
All comments
medini_1036016653
Hint: I'll present our (ZF's) current solution/approach in the workshop taking place on 2020-02-05 in Schweinfurt. Teaser (NDA restrictions apply):
medini_1036016653
Today, we still would love to tap into a failure mode's "residual fault rate", as is visible from a DC Worksheet where safe-fault fraction and Diagnostic Coverage are assigned. Proposal: Extend "Use Rate From" functionality to allow to alternatively select a "DCFailureModeEntry" (currently only "FailureMode"). In that case allow the user to select "base failure rate", lambda_residual,SPF or lambda_resdiual,MPF (independent, if the DCFailureMode is actually marked as SPF or MPF via checkbox). Handling of evaluation groups needs some thought. This would allow to get rid of a considerable amount of AND gates that are merely used for mathematical modelling of the multiplication "lambda_base * (1-DC)". Thank you Peter
medini_1028038066
Dear Peter, due to your request we have opened an internal ticket for review from the Product Management Team. The requested functionality will not be provided, we will not allow a direct connection between FTA to FMEDA, but need to go through the model.When we move to Component Fault Trees (CFT), they will essentially contain faults/events together with the SM/diagnostic, where the value you are looking for will become part of the model and not only of the table as of now. We will keep you informed when the CFTs will be released.
medini_1036016653
Dear Krasimira, thank you for giving feedback on your thoughts on this topic. As you might remember, we make use of the Safe Fault Fraction (SFF) to model for example "lower EMI performance" HW Effects or actual tested safe fault fractions. We do propagate this factor as well, so the formula actually is "p=(1-DC)*(1-SFF)" and this probability is ANDed with the base failure rate from the "HW Effect" failure mode (that uses "whether to compute failure rate based on its causes (or not)" functionality to accumulate the failure rate of the causing HW Part Failure Modes). Consequently, for the CFT approach, please consider means to also propagate the aspect of SFF from FMEDA to FTA. Best regards Peter
Quick Links
All Categories
Recent Posts
Activity
Unanswered
Groups
Help
Best Of