Re: Face plate thickness Steve -
The 1/4 wavelength is only for continuous wave sound. Usually the thickness
is a little less for pulsed transducers and in some cases you are better
off with a long delay line. Ideal thickness is at best a compromise. One
way to improve on the compromise is to use two layers one of higher
impedance and one lower in which both are less than 1/4 wavelength. The
actual thickness is usually arrived at empirically in either case.
The case of an unknown impedance is usually handled by just picking
something that works well in general. The types of layers used for contact
transducers for steel usually are poor matches. Use of plastic for example
is popular but clearly doesn't meet the arithmetric or geometric mean
criteria at all. Another popular material is aluminum oxide which matches
the crystal better than either mean. Wear is the criteria being used to
justify aluminum oxide.
Not knowing the impedance is puzzling. Do you mean there is one and you
just don't know it? Or will it vary from part to part or within the part?
Even in these cases you should have some bounds?
In most cases you should be able to do well with any good quality contact
transducer. If much "scrubbing" is involved I would recommend the aluminum
oxide types because they wear much better. The only exception being
plastics where you may want to use a plastic faced probe because wear is
not a problem or the transducer face may wear the part.
: Sorry to come late, could I maybe sneak another
: question in?
: We all know that the best face plate for a transducer
: is 1/4 wave thick and its impedance is the mean (either
: arithmetic or geometric, depends on the paper) of the
: transducer and target impedances.
: My question is, what about when either you don't know
: what the impedance of the target is, or you cannot
: control the impedance of the face that well? Both
: apply to soft-faced contact NDT probes. How thick
: should the faceplate be then?