Hello,

Are you sure your question is sufficiently developed? From this very general point of view, it will be difficult to help you. You can expect a signal processing solution however, provided you can analyse the effect of the Lamb waves alone from your sensors.

 Hello Patrice Simard,

Thanks for your answer. I am conducting lamb wave experiments on an aluminum disk. I excite lamb wave at point A and capture the response at point B. The sensor signal is very complicated due to reflections from the boundaries of the disk. So, one way is to make a specimen large enough so that the boundary is very far from the sensing point. However, as you mentioned, It would be much more efficient to get a signal processing solution to this. I would appreciate if you have any advice for me..

Thanks

You can use some signal processing methods for seismic data, like f-k (frequency-wavenumber) filtering, to extract the head waves and to remove the reflections with definite velocities. You may refer to Rector et al. 1995, Geophysics, 60(3): 692-701.

I hope it would help.

For experimental work,  you can try pasting sufficient honey bee wax or raw bread dough or mud paste along the boudary of aluminum plate.

What kind of the Lamb wave do you mean? For the lowest antisymmetric mode A0 you could try to make the edge sharp. Under certain conditions it supresses reflection of A0 mode. This is called acoustic black hole effect. Works best together with absorber, honey for example. See works by V. Krylov for details.

Hi, You could probably keep an absorbing material on the surface of the boundaries, such as silver on Stainless steel plate surface. I remember to have gone through a report wherein it is mentioned that a silver brazing material on a stainless steel plate absorbed the Lamb waves.

Hi Mehdi

I had to deal with this problem during my researches. there is always some types of signal processing method but also you can use absorbing materials as the most reachable one. try different kind of them and on-line check your signals to see what happens there. you can rely on this method if you are completely sure that there is no defect in your plate. because sometimes defects reflection may be consider as the back walls signals

If you have an experimental B-scan containing direct and reflected waves, the best way is to apply 2D-FFT and 2D-IFFT to regenerate 2 different B-scans containing only direct waves and reflected waves.

Hi

In experimental, you can put some playdough on the edges. So lamb wave reflections from boundary will reduce easily. I did it, and it was helpful.