I am trying to capture a damage based nonlinear phenomenon (generation of higher harmonics) in a system but instruments induced nonlinearity (specially from high voltage amplifier) is so dominant that damage induced nonlinear phenomenon is insignificant compared to it.
Apply (1) stabilized voltage amplifiers, (2) filters to eliminate unwanted intermittencies of electrical origin, and so on.
In reality, many experiments have shown that superharmonics may be generated in piezoelectric transducers, liquid couplers, electronics etc.
Unlike superharmonics, subharmonics are generated only in the studied nonlinear medium.
Try to use ONLY subharmonics for structural health monitoring.
However, the subharmonic generation mechanism is not yet crystal clear.
It needs to be studied more deeply.
Regards
1- Use "linear" high power amplifiers in generation side. RITEC system has linear high power amplifier that means it does not inject extra higher harmonics to primary signals. You can also use high power filters to remove second harmonics. This is more affordable for group who does not have funding to buy expensive linear amplifiers. So, always make sure your wave generation method does not generate higher harmonics.
2- Try to not use any amplifier or filter on reception side. In general, PZT sensors do not need to be empowered by amplifiers. MST, and PVDF sensors are also very good in SNR, while MST can generate some sort of higher harmonics inside.
3- Do not use computer board oscilloscopes. Boards inside the computer cases are exposed to higher noise level and hard to suppress them. Make sure your Osc. has very low noise level when it is grounded.
4- Use stable grounding to make sure the noise level is minimum and electormagnetic perturbations are damped.
5- Use high number of averaging. Usually 500 averaging is a reasonable number, while some studies have used up to 2000 averaging to get rid of stochastic fluctuations.
6- Waves with different polarizations in first and second harmonics are interesting since it is possible to separate the receptions by different sensors (Please check our JAP 2018 paper). The idea is extendable to bulk waves, as some groups have been working on it (Referenced in our JAP 2017 and 2018 papers).
7- Using wave mixing, it is possible to effectively isolate majority of electronic nonlinear perturbations. We have worked on guided wave mixing in plates and you can check our papers. Plus some other papers are highlighted referring to bulk waves and surface waves.
hope my answer can be helpful!
In your case, I am afraid there is not any post processing (signal prosessing) method to distinguish material and electronic nonlinearity, since they are mixed during the experiment. If you can repeat the tests, the simplest way is to use high power filter right after your high power amplifier in order to filter higher harmonics.
Jay,
see this paper for a wavelet-based decomposition algorithm for nonlinear ultrasonics which reduces the signal processing error and is not influenced by signal processing variables, so you can make sure there is no nonlinearity coming from the post processing.
https://www-sciencedirect-com.proxy.cc.uic.edu/science/article/pii/S0263224117302282
Bests,
Negar
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Waves