2
$\begingroup$

Background:
I have a piezoelectric force transducer that I am using to measure the force produced by a combustion reaction in a tube. The issue is that there is a lot of ringing that occurs during the reaction, leading to oscillatory "noise" going back and forth about the x-axis when we expect to see more of a big force curve. I want to eliminate this noise so that we can get a more proper idea of the force being generated.

My understanding:
The combustion reaction on the transducer can be modeled as a mass-spring-damper system with a forcing function. The mass (the tube + reactants, known), the spring (the transducer, known stiffness), the damper (transducer?, damping ratio TBD), and the forcing function (combustion, unknown).
If I can get the transducer's response to an impulse, I can use that to obtain an approximate transfer function for the impulse response using MATLAB. From here, I believe I should be able to deconvolve the force data and the impulse response in order to get the input force from the transducer.

Question:
What is the best way to generate an impulse signal that I may use for this task?

What I've tried:
I've done drop tests with two screws of two masses directly onto the transducer, but the data seems kind of inconsistent between the screws. I took the drop test signals, filtered out HF noise, subtracted the means, zeroed out everything before time t=0, and then normalized them according to the max force recorded. Then I used an impulse as the input (at t=0, input=1; 0 elsewhere) and the signal itself as the output for MATLAB's tfest function.

$\endgroup$
9
  • 2
    $\begingroup$ Cross posted here: electronics.stackexchange.com/q/601537/152903 $\endgroup$
    – Solar Mike
    Dec 23, 2021 at 16:50
  • $\begingroup$ Thanks, wasn't sure if I was supposed to mention it. $\endgroup$
    – Ben
    Dec 23, 2021 at 17:11
  • 1
    $\begingroup$ Idea is to save people writing duplicates. $\endgroup$
    – Solar Mike
    Dec 23, 2021 at 17:12
  • $\begingroup$ It might be easier to generate ramps or sinusoids. Impacting the transducer directly sounds iffy, make sure you're not creating peak load past its rating. You could also try impacting tube walls. Also consider other effects resulting in ringing, like waves bouncing back and forth along the system's long dimension. $\endgroup$
    – Pete W
    Dec 23, 2021 at 18:37
  • $\begingroup$ could you provide rates on the combustion rate, the frequency rate and the high pass filter settings? $\endgroup$
    – NMech
    Dec 24, 2021 at 9:38

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.