Accelerated Life test (random)

[bcg444]

Hi All,
I'll try to make my question short.
The random profile for truck application I use has its highest levels up to 10 hz. When I calculate what an accelerated level should be for a reasonable amount of time of actual test, it brings me up too high (1.5 - 2 G^2/hZ).
The question is: is it valid to replace this portion of the profile by some representative shock tests? Otherwise, how can I work this issue out except increasing the duration?

Thanks a lot for any input.

 

[mechengdude]

sorry to answer your question with a question but...
when you say "has its highest levels up to 10Hz" what do you really mean? Up to 10hz has the highest g2/hz value or GRMS? In many cases you may have lower g2/hz values at a higher frequency which really gives you a higher GRMS value which is not intuitive.

I have seen test plans ignore testing in the 5 -10 hz range for the issue you bring up. (ie) many shakers can't handle the large displacement required.

You may also want to consider a low input sweep to determine natural frequencies. If there are no resonant frequencies in this range you give yourself a good feeling that a shock test would suffice for that frequency range while you consider fatigue in the higher frequency range you are able to acheive by test.

good luck

 

[IRstuff]

The basic answer is no; while shock and vibe both use g's, the frequency content and durations are substantially different.

Are you concerned about the vibe level for some reason? Your "accelerated" levels appear to "normal" for some military applications.

TTFN

FAQ731-376

 

[mechengdude]

I don't know that I would say that the answer is no. It depends on what your definition of valid is. It is true a shock test and the random are certainly not the same things. But remember, the tests are simply an attempt to obtain a level of confidence that the part/system will survive in the real world.

Why are you doing the test? To meet a customer requirement or to satisfy internal company concerns. Who has final buy off on the test plan? These are all important questions that play a part. There are usually lots of compromises when setting up a test.

Using MIL-STD-810 as a guide, the tracked vehicle section gives about .05 g^2/hz at 12 Hz as the random vibration profile in the vertical direction. If you use 2 g^2/hz at that freq. that translate to about 21 GRMS! That’s quite a bit.

Back to my earlier point, who do you have to satisfy with the test? If it were me, I'd give you some latitude. Sounds like IRstuff would have you go "back to the drawing board".

 

[GregLocock]

20 g rms is about what you get on a 4 cylinder engine at 5-6000 rpm.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[IRstuff]

You need to use MIL-STD-810F carefully. The guidance in Method 514.5A stipulates as an example, that the exposure levels in Table 514.5C-1 and 514.5C-VII represent 240 km of travel for each 20 minutes of test duration. This is the basic integrity test, which, as mechengdude states, is only a confidence test.

But a life test for something like 240,000 km would require 20,000 minutes, or 333 hrs of continuous exposure.

The purpose of the accelerated life test is to trade a higher level for a shorter test duration. I don't know what the acceleration factor is for applying a higher level, but any test levels about 10-g will be challenging for most test facilities.

TTFN

FAQ731-376

 

[bcg444]

Thanks folks,

Mechengdude captured exactly the issue I am trying to achieve. Finally, what I did, just keeping the same Grms level, increased an amplitude on higher frequencies. Given the fact that vibration facilities often have low frequency limit, combination of this profile with shock test should get me a good results, I hope.