Mounting accelerometers on DUTs during vibration durability tests?

[davidd31415]

Is anyone familiar with the practice of mounting accelerometers on DUTs during vibration durability tests?

I've been hearing conflicting (and rather passionate) views on if this is acceptable from a few different people. I'd like to find discussion of this subject haven't found one so far.

Thanks,

David

 

[jimkinney]

Check this thread.

http://www.eng-tips.com/viewthread.cfm?qid=224055&page=2

Jim Kinney
Kennedy Space Center, FL

 

[davidd31415]

Thank jimkinney,

Interesting thread but I am interested in where to place the control accelerometer, not how to mount it.

I found this:

http://archive.evaluationengineering.com/archive/articles/1006/1006control_accelerometer.asp

I'm interested in a similar discussion related to sinusoidal durability testing.

David

 

[btrueblood]

The control accel is usually placed at the customer's interface (i.e. the bolt face on your vibration fixture). But, as you found per MIL-STD-810, that is not defined in the test standard (you may, if you look closely at the vendors source control drawing and/or spec. that defines the vibration test PSD envelope, that the PSD is defined at the interface plane, this was usually the case for us...sometimes it got overlooked...). Usually, the location within the interface plane is chosen for easy access, but we have occasionally placed them so as to avoid exciting resonances (i.e. we did testing with dummy articles, and recording several different accel's in different locations, and chose the "best" spot from that survey).

 

[floattuber]

An interesting article. I've taken one of Tustin's newer classes, taught by Strether Smith but I don't recall talking much about averaging or extremal control strategies.

I'm not sure I agree with Tustin's recommendation of never using averaging or Extremal. One will always get undertesting and overtesting at various locations around the DUT, but if you don't have an accel there you'll never know it.

Anyway, back to the original question of location. We always try to mount the control accelerometer(s) as close to the DUT interface as possible, on a rigid location on the fixture. Rarely do we mount it on the DUT and only sometimes mount it to the slip table.

I think it's not mounted to the DUT because most of the time the customer doesn't want to drill and tap their product and a lot of times their product surface isn't beefy enough anyway.

The bottom line is the control accel just needs to be in a location that needs to see the desired vibration spectrum. That location is the only one where you can guarantee that it's seeing the spectrum. So the farther away you are from the DUT, and the more interfaces there are between the control and the DUT, the worse it is.

By the way, I highly recommend taking Tustin's (TTI) classes. It was probably one of the best classes I've ever taken.

 

[IRstuff]

The control accel is the one sensor that's supposed to reflect the environment being supplied, so it should be wherever the customer's interface to the DUT is.

TTFN

FAQ731-376