Road test (g vs. time) vs. PSD table (g-rms)

[dag113]

In the past, we performed a random vibration test on a particular component (PSD from MIL-STD-209). The component has now been installed on a vehicle and that vehicle driven over test courses. The component was instrumented with accelerometers for the test.

I've been given the test results in the form of an accel. vs. time plot. Someone else has said that because the peak acceleration on the accel. vs. time plot is greater than the g-rms value from the PSD table, the component is seeing accelerations in excess of its original testing. I'm not sure one can compare those two items that directly (g-peak vs. g-rms).

Can anyone set me straight? Many thanks for your time.

 

[BobM3]

It is not necessarily true that the time history has a higher peak acceleration than what your PSD test produced. PSDs only show the average energy at (actually "around") a certain frequency. If your PSD was generated with a band limited random signal where many frequencies where playing out at one time then the peak acceleration would be higher than the PSD of that signal would indicate.

 

[GregLocock]

Agreed. The only way you can really compare them is to look at the same measuremnt. In this case it is easy for you to create rms g, by repeating the original analysis.

Incidentally that is misleading , even comparing psds is not absolutely reliable, you really need to be comparing fatigue lives for chassis components, depending on which brand of voodoo you subscribe to that might mean psds, but is more likely to be level counts or peak values. For many components you can use acceleration as surrogate load (stress) data, there's a paper somewhere on the web by DCX outlining a reasonable approach.


Cheers

Greg Locock

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

 

[dag113]

Thank you both very much for the responses; that's very helpful. I'm glad to see I was at least on the right track.

I'm going to go digest all that and try to go from there. (By the way, that PSD was from MIL-STD-810; I apparently had a small aneurysm when typing and listed the wrong spec). Thanks again!