about MIL-STD-1796 . Have you seen it? Opinions?

[DHambley]

MIL-STD-1796 claims to be the new standard for aircraft, military and other mil-aerospace industries. Has anyone here had a chance to review it? Does it address modern electronic parts? (not interpolation on reliability of tubes)

Many of us know the deficiencies of MIL-STD-217 and others (for example, the incorrect assumption that an electronic part which is subject to constant hot/cold cycling has better reliability than the same part which is kept at a constant hot temperature; simply incorrect.) For decades, I have been hoping for a new standard which is more realistic.

The authors of 1796 encourage firms to put it as a contractual requirement for aircraft equipment. This document can become a major influence on the expense of new equipment. I would hope it has valid equations for modern equipment.

 

[IRstuff]

It sounds like you haven't read it either. 1796 is simply a condification of reliability beast practice from a design perspective. It's essentially what you would do for a design, given a customer specification that stipulates the environments and operating conditions applicable to your product.

" constant hot/cold cycling has better reliability than the same part which is kept at a constant hot temperature; simply incorrect."

I think you're misunderstanding something here. Every physics of failure scenario and analysis shows reliability degradation with increasing operating temperatures; it's the foundation of the applicability of the Arhenius equation to temperature activated failures. There is no assumption here; there's 50 yrs of data behind this.

TTFN
faq731-376

 

[DHambley]

No, I haven't read it. That was my question. Has anyone reviewed it yet? And about reliability; No, I'm not "misunderstanding" anything. Sounds like you misread my post and thought I was saying that temperature isn't related to failures. Failures due to temperature alone may use the Arhenius equation but that is not the only temperature-related issue. Physical stresses during hot and cold cycling eventually weaken parts, due to everything from micro cracks, differential expansion, to separation at boundaries. A device cycling from -55C to +125C constantly would not have the same failure modes as that same device at a constant 125C.

A big problem I have seen in aerospace an several different firms is that Reliability Engineers I have worked often do not take thermal cycling into account. Their prediction of 20yrs MTBF is shown to be incorrect when the products start coming back, then under the microscope we see the actual cracks.

 

[IRstuff]

I have found that reliability engineers usually depend on systems engineers to give them the correct environments, but projects rarely have sufficient time or budget to do a thorough and proper analysis. Additionally, it's often the case that the presumed environment actually looks nothing like the actual.

Generally, if you are seeing lots of failure returns, there was a gross underdesign involved.

TTFN
faq731-376