Measuring impact using accelerometer.

[brashear]

I have a setup where a high velocity ball is directed towards a metal plate, which has an accelerometer mounted to the other side. When the ball collides with the plate (assume elastic deformation), there is a large oscillation observed in the signal from the accelerometer, which eventually damps out.

Is there a method to correlate the velocity of the ball to the oscillations seen by the accelerometer? What if I place a thin material on top of the plate that plastically deforms when the ball strikes it? Does this correlation change?
Basically what I'm trying to do is to measure the velocity of the ball using an accelerometer.

Thanks

 

[brashear]

Also note that the plate is clamped down.

 

[IRstuff]

Sorry, but that's not the way to do it. Google "ballistic pendulum" for the more appropriate way to do it. Or Google "chronograph" which is a more expensive, but more direct measurement.

TTFN

FAQ731-376

 

[FeX32]

You could try to estimate the impact velocity if you know the precise damping ratio of the plate. (assuming it is proportional damping) But this is likely to be not as accurate as what IRstuff suggests.
I would go with the pendulum.


Fe

 

[brashear]

Sorry, my question isn't on how to measure the velocity of a ball hitting a plate - it's whether it is possible to use the output from an accelerometer.

 

[IRstuff]

Fe gave you the answer to your question already

TTFN

FAQ731-376

 

[GregLocock]

Well, the answer to your basic question is yes, sort of, but not very well.

For a start the velocity of the ball after it bounces off will have a crucial effect on your estimates.

The way I'd do it is to work out the maximum energy seen in the plate, which might be very close to the bending elastic energy at maximum deflection, which you can get from the acc signal.

That is an upper bound to the KE of the ball, of sorts.

yes adding material to the plate will change the answer.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[PersonalProfile]

You could intergrate the measured acceleration with respect to time to calculate "a" velocity, though the calculated "a" velocity may not be "the" velocity you are actually seeking to measure.

Regards,
Lyle

 

[GregLocock]

Well i guess that'd be a lower bound.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[brashear]

Thanks for the responses,

Greg: Are you suggesting that I measure the maximum deflection using the accelerometer? I'm afraid this isn't possible as the plate's mass is very large compared to the energy of the ball, and deflection is not noticable.

Lyle: I don't think I can do that because the net acceleration is zero, since the plate is clamped. The whole thing just vibrates without net movement.

Basically what I'm trying to do is to correlate the vibratory response to the velocity of the impact.

 

[IRstuff]

As others have stated, if you cannot measure the initial velocity, then you can't measure the final velocity, so you cannot complete the energy balance equation to solve for the initial velocity, particularly, since it doesn't even seem like you can completely characterize the dynamical behavior of the plate.

TTFN

FAQ731-376

 

[GregLocock]

If you can measure the acceleration you are measuring the displacement. Not very well sometimes, but in practice double integrating an acceleration is quite useful.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?