We have two different piezoceramic patches made from the same material (first one is square in shape: 13 x 13 x 0.9; second one is rectangular: 18 x 9 x 1.76). All the dimensions are in mm.
As per the calculations, the only difference in the two is their capacitance and the resonance frequency of the patches in the polarization direction.
The capacitance of the square patch is almost double that of rectangular one. Given the dimensions in the polarization direction: square patch has resonance frequency of 109 kHz and rectangular one has resonance frequency of 75 kHz.
We are in the lower ultrasound range (30 kHz - 50 kHz). Pitch-catch approach over a fixed distance in our specimen using the rectangular patch works well as anticipated whereas the square one has barely any strength in the transmitted signal. I understand its because of the higher current requirements. But when I bump up the input voltage, it burns.
The square patch burnt when the input voltage was 2 Vp-p whereas nothing like this happens to rectangular patch even at 4 Vp-p. Is there any correlation here which I am missing. Both the patches are made from the same material.
Not easy to answer without more information such as material type, supplier etc but here goes. 2vp-p is very small driving voltage. Usually, UT probes are hit with 100's of volt spikes. When crystals are in the raw with nowhere for the energy to dissipate, you do have to be careful. Maybe try the same with them dipped in an insulating oil? 75 KHz and 109 kHz is a fair difference and the 75kHz one will be thicker. Also calculate the area's, I get them to be the same so it all comes down to the permittivity of the material and the thickness. The electrode material also has some influence. Piezo materials are often screen printed with a thick silver based material. Others are sputtered with a thin gold which does not support much current. Have you looked at just buying the transducer you need? These are very low frequencies for UT but there are plenty of suppliers out there making probes for all applications. You will need a front face (1/4 wave) material that acoustically bridges between your piezo and your sample and you will need a backing material that absorbs the wave and does not reflect it from its back face. Generally backing materials are crumb rubber and tungsten loaded epoxy to give a mechanical loading and sound absorption. Particle size should be a function of sound wavelength. There are some papers on transducer design. https://piezotechnologies.com/how-transducers-work/
https://warwick.ac.uk/fac/cross_fac/complexity/study/emmcs/outcomes/studentprojects/perez_diaz_m1.pdf
Good luck.
Thanks for the reply Chris.
Here is some more info to my question:
The patches that I want to compare are made from the same material PIC255 (page 3 in the document: https://fisica.cab.cnea.gov.ar/bt/images/d/d3/PICat.pdf).
I am using them for guided wave inspection of pipes i.e. multiple wave modes (longitudinal, circumferential, and flexural) may coexist at different frequencies, so fine tuning of the inspection frequency is crucial for us. We are currently interested only in longitudinal modes existing at these lower frequencies.
Patch 1 as the dimension of 18mm x 9mm x 1.76mm and it has polarization along the 18 mm length, frequency constant of 1420 in length direction gives a rasonance frequency of around 78 kHz. This one works as anticipated as at 30 kHz and 50 kHz signal and matches the theoretical predictions. Signal transmission is recorded at 200 mm length using similar patch.
Patch 2 (made from the same material) has a dimension of 13mm x 13mm x 0.9mm and also has polarization along the 13 mm length, so for same frequency constant of 1420, it has resonance frequency of 109 kHz. Signal transmission is recorded at 200 mm length using similar patch.
Problem that we are facing.
1) We barely see any signal getting transmitted using patch 2.
2) Patch 1 can easily take upto 4 Vp-p output from function generator (prior to amplifying by 55 dB) whereas patch 2 burn even at 2 Vp-p. It seems that there is no energy at all in the patch 2 excitation.
This is very different to a UT probe. I did some work on this in 2006. These are not easy transducers to get right. If I recall the mass on the back needs to be very large for the crystal to react against in the axial direction. Also the probe needs to be clamped hard against the surface. The driving voltage need not be high but would be a pulse train of a number of cycles. A front face is still needed that can pass the wave and protect the crystal.
Maybe something to lower the current on the pulse would be the solution to the immediate problem. On other probes we used a matching network of a coil and resistor(?). The coupling of the wave is complex. The movement is axial and launches a shear wave/love wave/reighley wave(?) which gets emitted along the surface and launches other waves as it travels. The only fluid which couples shear is molasses hence the clamping of the probe onto the surface. Sorry if this is a bit of a brain dump and some of it could be completely wrong. Good luck.
- Badges
- Science topic