I need an accelerometer data anaylsis vi that will give displacement from transisent data.

[38curt]

Sample rate = 50 kS/s
no filter
event duration 20-40 msec
700 g's max
I want it to run inside of labview.
Any suggestion?

[markwysong]

Sure! Since displacement is just the double integral of acceleration, you can use the integral VI in labview under Analyze-> Signal Processing -> Time Domain -> Integral x(t) VI. I believe that this is in the Analysis Package, so you must have the Full Development Package of LabVIEW or purchased the package separately from LabVIEW.

By taking the values of your waveform, converting them to g values, wiring that to the input of the integral VI, and telling the integral VI the delta T of your measurements and taking the output of that into another integral vi with the same delta T, that should give you the displacement that you want.

Mark

[Gérard]

Curt,
I happened to be interested in this problem for some time to monitor
vibrations in rotating machinery with order track and a mix of eddy
probes, velocity sensors and..accelerometers. In order to compare
vibrations from eddy probes and those of seismic sensors, I had to
doubly integrate the acc responses--> your wish .
.. I guess you have only one channel and fairly fast card and not doing
it continuously in real-time but as one-shot transient analysis possibly
with pretriggering.
You have to doubly integrate the accelerometer signal with some AC
filtering and acquire the response above 100 KHz (2 X bandwidth of your
acc). That is easy to do: use integration vi from the library with an AC
filter (IIR takes much less time but FIR have a linear phase lag and do
n
ot distort the time signal but simply shifts it.
If you use 2 channels, beware that the integration vi is not reentrant,
but I guess it is no issue for you since you assume that the
acceleration is quasi zero before the shock transient. Of course you are
bound to miss the low-frequency contents of your acc response anfd thus
may miss some of the real displacement. That's physics and has nothing
to do either Labview or data processing.
If you are busy with Labview and S&V problems, contact me at
gdans@ulb.ac.be after Tuesday.
Gérard


38curt wrote:

> Sample rate = 50 kS/s
> no filter
> event duration 20-40 msec
> 700 g's max
> I want it to run inside of labview.
> Any suggestion?