Lifetime Support Calculation - Thermal Operating Profile

[I hate math]

Hi,

When it comes to lifetime calculation I am always confuse.

I have this thermal operating profile as below:-
Tj = 110C (operating hours = 400hrs)
Tj = 105C (operating hours = 1200hrs)
Tj = 100C (operating hours = 1650hrs)
Tj = 95C (operating hours = 2200hrs)
Tj = 90C (operating hours = 1520hrs)
Tj = 85C (operating hours = 73hrs)

Assuming available stress condition is at 125C at 1000hrs with typical voltage at 1V
Can the above profile exceed the lifetime of 10 years?
In terms of lifetime calculation, should we be calculating both temperature & voltage acceleration or just temperature acceleration alone as we only have the thermal operating profile?

Hope to get the answer ASAP.
Thanks in advance.

 

[Miner]

Several things are unclear, so we are confused also:

You provide what you call a thermal operating profile. An operating profile defines actual use conditions. Was that your intent, or is Tj the life result at various temperatures? Or, is this a test profile where Tj is the test duration at various temperatures?

What is the available stress condition? Is this the expected use condition, or an accelerating test condition?

What is the duty cycle? Does it run 24/7/365?

Are you using voltage to accelerate? Is 1V the operating voltage, or an accelerating voltage?

 

[I hate math]

Yes, is an actual use conditions with various of temperature operated at the specified operating hours.
Available stress condition is the accelerating test condition.
The duty cycle based on the profile is that it operated at total 7043hours per year (added all the operating hours = 7043 hours).
1V is operating voltage, if accelerating voltage is about 1.2V.

 

[I hate math]

Any idea whether the above profile can exceed lifetime of 10 years?
How to calculate that?
Would appreciate if can share the calculation method.
Thank you.

 

[Miner]

There is no way to say without test results. Do you have an accelerated life test plan or test results? What is the reliability requirement for this product? What distribution (e.g., Weibull, etc.) and parameters (e.g., shape) are you using? Number of samples?

Does the product go through one complete operating cycle per year then sit idle for the remaining time, or are these a summation of multiple events at those temperatures?

Answering an earlier question, you can create acceleration factors for both temperature and voltage. Temperature usually follows the Arrhenius equation, but voltage equations can vary depending on the product. Depending on your answer to operating cycle question, you may also be able to accelerate through the duty cycle.

 

[Ronen E]

Temperature usually follows the Arrhenius equation

Only where failure involves a chemical reaction of order 0 or 1, or where otherwise rationalized.

 

[Miner]

Only where failure involves a chemical reaction of order 0 or 1, or where otherwise rationalized.

Yes, that's why I used "usually". There are always exceptions.

 

[Ronen E]

Yes, that's why I used "usually". There are always exceptions.

Sorry to be pedantic, but I think "sometimes" is more appropriate than "usually" here. The point is that Arrhenius equation (and its derivatives) has become so popular that seemingly people tend to apply it without questioning whether it's applicable to the case at hand, even without realising it needs substantiation. I'd dare a guess that it's misused more often than not. Hence my insistence on reminding.