Accelerometer gives wrong signals

[StathPol]

Dear all, 

I have a problem with the signal I got from an accelerometer. I am doing a drop test and I have mounted an accelerometer on the bullet dropped in free fall. The objective of the test is to measure the impact force.

The accelerometer is a single axis accelerometer with capacity of +/- 10000 g. The accelerometer is connected with an amplifier and then to the NI6361. The module is then connected with a PXI 1073. 

The bullet (steel) weight of 10.6 kg was dropped from 100 mm in a concrete base and the response I got was 3500g. I think this is unreasonable as the theoretical calculations in conjunction with the image analysis using a high speed camera provided me with the acceleration of 40g.

According to the supplier, each 1 Volt corresponds to 1000 g. Any help will be much appreciated. 

Thank you, 

Stath

[johnsold]

What does the waveform look like? Is it possible that you have very high acceleration for a very short time, corresponding to a point contact impact? What is the resonant frequency of the weight?

Lynn

[Gregory]

I am assuming the "g" stands for grams and not 9.81m/s^2. So what does it mean to have an acceleration of 40 grams?

[StathPol]

Lynn and Gregory

Thanks for your replies. I have attached how the graph looks like in a pdf file. g is the Earth's gravitational acceleration of around 9.81 m/s^2. Please find below how the scale was configured using MAX. 

Acceleration values of around 4000g for a falling object dropped from 100 mm are unreasonable, I think.

Thanks, 

Stath

[Gregory]

Do you have a link to the product page?

[johnsold]

Can  you post the data sheet (or a link to it) for the accelerometer?

Lynn

[altenbach]

---

@StathPol wrote:

Acceleration values of around 4000g for a falling object dropped from 100 mm are unreasonable, I think.

---

The acceleration during dropping will be ~1g. The decceleration on impact can easily be 4000g because it happens over a very, very short distance and time if the floor and object are relatively rigid. Numbers seem reasonable.

At impact, you object will be going at a speed of about 5km/hour and will deccelerate to zero over a very small fraction of a millimeter. Do the Math!

[StathPol]

Lynn, Altenbach and Gregory,

Thanks for your replies. I have attached the certificate for the one single axis accelerometer and the power supply settings. The signal is acquired using a voltage card and then is converted to g. 

Altenbach -- During the impact event, I measured the acceleration using a high speed camera which gives almost the same result with the theoretical calculation based on the steel-concrete pulse width (impact force-time). Today, I have tried to hit a load cell from a distance of 50 mm using a weight of 10.6 kg. Because it is steel-steel contact (elastic collision), we expect the higher forces to occur. So the softer the material is, the longer the pulse width will be. 

Again, I have a feeling that something is not correct with the settings using LabVIEW. I have checked many published papers that they used 600 kg from 1 meter free falling and the acceleration was not that high. 

Regards, 

Stath

[Gregory]

I am getting a drop time of .143s, and velocity 1.4m/s before hitting the ground. To have an acceleration of 4000g it has to lose this velocity in ~36us. Your graph separations are 1000us, so those spikes could be anywhere from 10-100us long, it does seem to make sense.

[johnsold]

0.5 mV/g * 10000 g = 5 V.

It appears from the fuzzy image you posted earlier that your scaling is reversed: 5000 g => 10 V.

Lynn