Effects of experimental variables on the nonlinear harmonic generation technique.

Harmonic generation is a promising technique for measuring small changes in the microstructure of components. Its extreme sensitivity is a benefit for detection, but results in a high degree of variability in any measurements taken. This paper characterizes the effects of experimental variables throughout the measurement signal path, establishing their relative importance and making suggestions for the best way to take measurements using harmonic generation. A model is used to predict the harmonic amplitude as a function of position and thereby account for alignment inaccuracy, explaining the effects of this key experimental variable. Finally, the potential effect of all of the variables on damage detection is discussed.