First thing I am going to say is, I strongly suggest some more study. You can consider the flaw to be a lamination. 23mm x 13mm and 3mm deep, remember I said it was parallel to the surface. Transverse waves never come into the picture, only longitudinal waves will travel in water. I fail to see where you got the 5mhz comment. Velocity is independent of frequency. There is a reason the wavelength formula is wavelength = velocity/frequency. For a 10mhz transducer in water that's a small wavelength. Attenuation is another matter, velocity-frequency and associated wavelength plays a large roll in that. Your system would have to be broken not to pick up a planar flaw parallel to the part surface in the dimensions listed.
Quote "Now you could possibly have longitudinal waves coming back faster than your transverse waves." At no time will a transverse wave outrun a longitudinal for equal distance and materials. That is assuming a transverse wave can even propagate in the material.
I've tried to help you here, and it seams you've taken exception to it. Therefore after this I'll let it die. It's not my money being paid on the test, But I do very strongly advise you to reconsider and look the matter up for yourself.
Heres a couple of good sites to get you started, but if you already have all the answers, why did you ask your initial question?
http://www.kettering.edu/~drussell/Demos.htmlhttp://www.ndt.net/ndtaz/ndtaz.phpRegards,
Gerald