Hello, first thank for your site, and for your forum it is very interresting.Question:PLC(Automate Programmable) is composed of many Input/Ouput Card and CPU.Concerning the card, the Failure rate was calculated using model of IEC 62308 by taking into account the external T° = 23°C and internal T° ( inside cubicle that containe the IO card and the CPU) =30°c and the ( thermal amplitude of the cycle) Delta T=15°C.Now the PLC is functionning in an environnement that the external T° is variable that is to say in a day the external T° can vary from 22° to 28°C and some times it reach 34°C. the Internal T° ( T° inside cubicle that contains the I/O cardPLC)1- I need to found out if the type of the variation of temperature ( internal and external) as described above should be taken into account in the life profil?2- In this case, how can I use FIDES model or how to define the life profil ( according to FIDES GUIDE) knowing that there is more than 1 delta T°( so that delta T ( external cubicle) = can be > = 2°C, 3°C... in a day I can have delta T 1 = 3° , delta T 2 , 4°C ....etc)Delta T ( internal cubicle) = can be > = 2°C,4°C, 6°C...etc) each delta T can last 1 hours n same times 2 hours , it is very variable. Thank you in adavance for your help
I need clarification in order to understand how I define the life profile, which are the correct parameter needed to take them into accont using FIDES MODEL.Je vous remercie d'avance de votre réponse et de votre aide.
Dear,
There if a very detailed chapter in the FIDES Guide on how to build life profiles. The section for thermal cycles is especially detailed and it explains how to deal with superposed cycles. I recommend you to have an extensive reading of this chapter and to have a close look on the life profile example.
If you don't find the answer to your questions, come again to this forum for more explanations (please be a little bit more precise, your question is a bit confused). But I'm sure the answer is already writen in the Guide.
Nevertheless:
- As you come from the IEC 62308 "RDF 2000" standard, my first advice will be to forget every cycles and life profiles estimations done within the context of this standard and start a new analysis for FIDES. There is no real method to build life profile in this standard and its way to deal with thermal cycle lead to major misunderstandings (and wrong results).
- If your question is that you don't know what will be the real cycles borne by your product, the answer is not strictly within FIDES methodology. FIDES says that if you cannot predict the life of your product you cannot predict its reliability. Various approaches are possible to deal with that very common and indeed permanent question. The first step is to build the life profile with some hypothesis (it may requires to add some specific phases in the life profile) and to do several sensitivity analysis to see how much the parameter has an effect on the result.
Also I recall that a FIDES failure rate shall be interpreted as a calendar value, for a population of product and for their whole life. Therefore you always have to consider typical representative life profile that will never be perfect. For a given parameter you may: choose a unique representative value (not necessarily the mean value, see FIDES Guide), have several phases in the profile each weighted according to its probability... If you have powerful calculation tools you can also try Monte-Carlo simulations!
I hope it helps.
Regards,
Lambda