How to reproduce very low frequency vibration on shaker table

[CaptainCrunch]

Hello All,

I have a tamper application that experiences very high amplitude, very low frequency vibration (e.g. 20g @ 0-3 Hz) when tamping a hard surface. Normally to reproduce field data on the shaker I exclude the 0-10 Hz vibration data and run 10-800 Hz, the range of structurally damaging vibration for our applications. If I exclude the 0-10 Hz vibration range of measured data on the shaker then vibration signal is actually less for tamping concrete than for tamping soft grass.

In the field these tampers fail quickly when tamping very hard surfaces. During operation you can feel the machine violently jumping around. Problem is when you look at the spectrum you get alot of gross acceleration (at 0-5Hz) of the device bouncing around.

The makers of the shaker controller recommend not running in the 0-10 Hz range due to control issues, the natural frequencies of the table are in the 2-5 Hz region.

I prefer to accelerate the shaker tests by amplifying the whole PSD curve, and from that I feel I can make more accurate field life estimates. But in these cases if I can replicate the very low frequency vibr., which is dominant in terms of device damge, the field life perdiction is not very accurate.

I was wondering whether any body has come across this issue before and come up with a novel workaround. Thanks in advance for any help and thanks for reading the very long post.

Regards,

George V.
Briggs & Stratton

 

[GregLocock]

Two random thoughts:

Use a hydraulic actuator rather than an electromagnetic one, with a mechanical suspension.

The failure mode is probably not due to resonant behaviour of the structure, so perhaps some proof load might suffice.




Cheers

Greg Locock

 

[MikeHalloran]

A few motion simulators provide extra capability in one axis by the simple expedient of mounting what would otherwise be the complete simulator on a movable carriage, and accelerating the carriage at the lowest frequencies.

Similarly, you could mount your shaker table on a carriage driven by a hydraulic ram, and use a crossover network to apportion the forcing spectrum by frequency.

However, I'm suspicious that your 20g number may be too low. Given that the tamper launches itself into the air on each hard- surface cycle, I'd guess that the impact from landing is what's killing them, and your shaker table doesn't have enough power to simulate that.

-Mike-