impact measurement

[Nolimit]

Hello Guys,

First I want to thank you all for the amazing help in getting me this far in my project.Please i have this issue,when an impact occurs and the system triggers, the output signal goes negative after the impact and settles.Please find the attached file for better understanding of my question.Thanks.

[Henrik_Volkers]

I do not open docx  from forums...  please post a png or a vi

Usually after an impact you have a swing into the 'negative' because of the high pass characteristic of the sensor-conditioner chain.  An indicator is the lowest frequency you can measure.

[Nolimit]

Thanks Henrik.Please find the attached file.it is in png.The circle points on the file are my concern.Is there a way i can condition it?

[Henrik_Volkers]

Well these curve looks reasonable at a fist glance.

Three possible reasons for the undershoot:

1. It's real: You have a neg. acceleration (also solid steel behave like pudding, just a question of scale 😉 )

2. Your sensor see a neg. acc. maybe because it's a seismic mass type sensor and you have to account the mass/spring system of the sensor

3. The signal conditioner has a high pass (maybe a two wire, const current exitation sensor and a cap for decoupling the bias) 

That all I can say without a crystal ball 😉

If you have more questions please name the DAQ, conditionerr, sensor (with links to datasheet?) and explain or picture the application.

Is the amplitude in the diagram  acceleration,speed or distance?  The shock is about 500µs, just from a view I expect power content up to 15kHz 

PS: In this forum you can say thank you by klicking the star (Kudos) an the left of a post 😉

[Nolimit]

The amplitude is in voltage.I am using a Kistler piezoelectric sensor type 9011A and a kistler  charge amplifier.The DAQ is NI-PXIe 6124.The sensor is preloaded and a ball is dropped from a height on it.The link to the datasheet is ( http://www.kistler.com/de_en-de/13_Productfinder/App.90x1A_BP.17/Product-Finder.html).Once agan Thanks.

[Henrik_Volkers]

If I understand right, your scale is voltage and represents force.

You haven't named the charge amplifier, which might have a low pass , easy test: apply a static load with a mass and monitor the output....

But as a more contrete guess the undershoot is due to the spring*/mass of the preload construction. Your system rings (@1-2kHz) after a shock exitation.

If you have access to a small impact hammer and a accelerometer? 

*) spring is build by the sensor and the screw in the mounting hole

[Nolimit]

The voltage is from the charge amplifier 5073,the link to the datasheet is (http://www.kistler.com/de_en-de/13_Productfinder/App.5073A_BP.17/Product-Finder.html).I performed a test without preloading the sensor i got same result.But i am thinking it could be as a result of the triggering.The charge amplifier low pass is set to 3030Hz.Please find the attached file.It is the result of static loading.The spikes are as a result of resetting the amplifier.

[Henrik_Volkers]

The preload might not have a big influence on the resonant frequency of your construction, as long you are in the linear  region (Hook)

Think of your construction as a buch of small mass points conneted by springs and dampers

Do you have a picture/drawing of your mounting?

A mechanic resonance at about 1-2 kHz is currently my fav. theory.

A measurement with an accelerometer and a hammer would answer that question.

[Nolimit]

I dont have accelerometer and hammer.The mounting picture is attached.Thanks.

[Henrik_Volkers]

Nice pic 🙂

I'm quite shure that you have a mec. resonace,

From the dimensions more a bending resonace. (the ball won't hit a plane surface and you bend the screw... or

How does the signal look like if you make a hammer/ball impact near the sensor on the base plate? I expect you can even hear a 2kHz ping 😉

BTW: Are you shure that the surfaces of your mounting meets the spec (flatness)?