Building a Shaker

[frankyg]

I was wondering if anyone has any experience building their own shaker tester. Commercial units are way out of our price range, and we would like to step up from our antiquated mechanical shaker. I have been investigating both electrodynamic and servo-hydraulic shakers. Servo-hydraulic seems to be the most logical for building our own, and I think we should be able to get the required frequency out of it. I am envisioning controlling the motion with Labview, and being able to take data that we measure with accelerometers in the real world application. Also, it would ideally be 3-axis testing, such as a platform that has at least one actuator in each axis applying vibration.

Has anybody here done anything like this, or would this end up being a huge job that wouldn't be cheaper than buying real commercial equipment? Unfortunately we have not been given the budget necessary to buy new commercial equimpent. I have seen some used electrodynamic shakers that may be in our price range, but I'm wondering I'd ideally like to go to 3 axis shaking.

Thanks for any feedback.

 

[GregLocock]

It is certainly possible to build your own 3 axis hydraulically controlled rig. My employer does precisely that.

I doubt that it is really cost effective unless you intend to build a few. If you have in-house hydraulic electrical mechanical and programming expertise then it is feasible. We only do it for one class of rig, usually we buy off the shelf.


Cheers

Greg Locock

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

 

[MikeHalloran]

Around 100Hz, the bulk modulus of the hydraulic fluid becomes a significant part of a hydraulic system. I.e., you can move a servovalve that fast, but whatever's attached to the other end of a column of hydraulic fluid is not necessarily going to follow the servovalve.





Mike Halloran
Pembroke Pines, FL, USA

 

[GregLocock]

Oh yes, high speed shakers are very exciting. I used to have a half inch Moog based (?) shaker that was theoretically good for 200 Hz.

Well, it may have been able to do that, but the THD was miserable once we got over 50 Hz.

Hydraulics don't really do sinewaves very well, no matter how many feedback channels you use.



Cheers

Greg Locock

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

 

[SomptingGuy]

Build or buy ... the common conundrum.

My employer has a combination of off-the-shelf and in-house tools. The main deciding factor is whether or not the off-the-shelf tool can do the job in the way that you want. Nobody in their right mind would build their own accelerometer or microphone, but our AI kit is home made, based on standard microphones/analysers/etc.

I think shakers fall into the accelerometer/microphone category. Buy one and spend the rest of the capital on software development to use it.

 

[frankyg]

Just a few more questions: what is the main impracticality in building a basic shaker? It seems like from the mechanical/hydraulics standpoint it wouldn't be unreasonable, but the programming would be the most difficult. I could see how costs would add up real fast on the hydraulics though. What are some recommended manufacturers of hydraulic shakers? What would a price range be for a setup capable of a payload of 500 pounds at 10G, and probably no higher than 50 hz (new or used)?

Thanks for your input.

 

[GregLocock]

I'd guess 50-100 thousand bucks, but that is purely a guess.

That is a very high frequency high power system.

MTS would be glad to help I'm sure.

Anthony Best Dynamics might be a good contact.

Cheers

Greg Locock

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

 

[IRstuff]

The main impracticality is that you don't know what you don't know. Since you've apparently never done such a design before, what are the odds that you've accounted for all the problems that anyone else has already worked through, like how to isolate, how to keep it from shaking itself apart, etc.?

I don't doubt that anyone with some amount of experience can build something that will shake a payload once. The issue is to design something that will reliably and repeatibly shake something with the accuracy and energy you desire. That will require some level of iteration, and cost, and time.

TTFN

FAQ731-376

 

[IRstuff]

Oh, and just as an example, a slip table shaker can't handle a top-heavy load, since the table itself will rock on the oil film creating a resonance that doesn't actually exist in the UUT. Took us nearly a month to figure that one out.

TTFN

FAQ731-376

 

[frankyg]

That is generally the conclusion I have reached as well. Obviously there are people that devote all their time to designing this equipment, so for us to design one in our spare time for less money seems unlikely.

Regarding the slip table, I was was not aware of this problem. Our products can weigh up to 200 pounds, and definitely need a method of testing in all axis. What other options would we have for 3 axis testing (costly multi-axis shakers probably out of the questions).

Thanks

 

[IRstuff]

We went to a testing house for our specific qual requirements, using something similar to:

http://www.lds-group.com/docs/display_product.php?product_item_id=456

The answer to your question really depends on what your overall testing needs are. We have a standard slip table for the occasional burn&vibe testing, which basically uses the table for horizontal vibration. Cross vibration requires rotating the UUT on the slip table, possibly with a separate mount adapter. Vertical is done by tilting the driver vertical, requiring possibly yet another fixture.

Your UUT weight isn't necessarily a problem, unless it has a high CG relative to its lateral dimensions.

TTFN

FAQ731-376