Impact test in concrete beams with accelerometers

[billko]

Hey, I was just thinking... maybe your setup was good all along.  Maybe you're not used to thinking of g-forces in the real world, so this threw you off the track.  You can read up on it here: https://en.wikipedia.org/wiki/G-force, but basically you weigh what you do because the ground is pushing against you at a force of 1g (and not because you eat too much, like you thought).  The accelerometer should read 1g.

Objects subjected to mechanical shock experience a high amount of g's for a fraction of a second, so I could see tapping the accelerometer with a screwdriver or something hard producing ~10k g's for a split second, just as you showed.

[StathPol]

Hey Bill. Thanks for your comments. I thought about of what you said, the weird thing is that the initial accelerations were not showing -1g, even when the accelerometers were in rest position or with the tests as suggested by Henrik. However, the analog signal is very noisy. Thanks for the link. 

[StathPol]

Henrik: Thanks for the test suggestions. I aim to do the ruler test shortly to check the accelerations, natural frequencies and displacements.  

Regards. 

[billko]

Starting from scratch is a good thing.  Maybe there was a loose wire or you did one step slightly different.  Maybe all that confusion was all because something was connected quite right.  It's happened too many time to me.  😉

[StathPol]

Hi, 

Do gravity needs to be considered as well?

Regards, 

StathiPol

[Henrik_Volkers]

On this type of test and with these sensors/couplers ... no, not really. DC values are filtered by the high pass filter.

But you always have to consider gravity ! Otherwise things will drop 😄

[StathPol]

I found from the amplifier that it has a short time constant where it is always hovering around zero at rest. It shows the dynamic signal.

Mike and I did the experiment that you suggested with the cantilever, as I think this question was uploaded in a different board from Mike as we had some issues of getting the displacement from acceleration in DIAdem. 

A ruller was fixed on the table and then the free end was allowed in free vibration. Accelerometer was measured in g force, then multiplied by 9.81 to give m/s^2. The results for velocity and position are very very weird. 

I also considered a low pass filter and offset. Please find the attached pdf file showing the results. 

Thank you , 

StathPol

[Henrik_Volkers]

As noted before: Why don't you post some (raw) data in a format others can help you to interpret? ( Like TDMS or the vi with saved values )

Have you tried the sound and vibration vis? They have a integration vi that includes a highpass and migth avoid the drift you see ......

The fixure of your additional mass can be improved 😉  If I interpret you gragh correctly the ripple on the first peak migth be due to a lift off of the taped mass 🙂 (So acceleration was sligthly higher than 1g;) )

[StathPol]

I have posted the vi, pdf and excel file with the raw data from 3 trial tests using 50 mm initial vertical displacement. It doesn't recognise the TDX file of the TDMS, so I cannot use it that’s why the raw data were transferred in the excel file. Unfortunately, I don't have access on the sounds and vibration vis. 

I have done some tests today but again, there is a lot of noise in the signal, despite the fact that the signal range was devreased to +/- 1000 g for +/- 10 Volts as you can notice in the vi. Indeed, a high pass can be applied to to avoid the drift on the signal. In the pdf document, the procedure followed was development of fft, filter, offset and double integration. 

Thanks a lot, 

StathPol