In order to ensure an appropriate behaviour to materials submitted to mechanical, thermal or chemical solicitations, their microstructural properties have to be evaluated by a fast and nondestructive method. Among the non-invasive techniques, the Surface Acoustic Microscopy (SAM) allows an analysis of the material surface-characteristics. Since the penetration depth of the Rayleigh-waves depends on the acoustic-wavelength, a pulse technique is proposed to investigate the subsurface area of materials. A spherical (or cylindrical) shaped transducer is excited by a broad-band pulse in the 10 MHZ-20 MHZ range. The acoustic energy generated by the transducer in the coupling medium launches the sample surface and is partially reflected into an axial echo and converted into a Rayleigh wave if the condition of generation is fulfilled. These two echoes are resolved in time. A first analysis in the time domain enables to measure the time of flight of the Rayleigh wave and to compute the absolute value of the Rayleigh velocity with a resolution of 0.6%. Moreover, since the bandwidth of the signal is wide, the analysis of the evolution of the Rayleigh-wave velocity versus its wave penetration-depth is available. This is obtained by studying the V(z) curves which characterize the material, by sampling the spectrum of the reflected signal at different frequencies. Our study deals mainly with the stainless-steel characterization. A first approach demonstrates the high sensitivity of this method concerning the roughness of the surface. In this case, the Rayleigh velocity versus the frequency is shown to be constant on a polished surface and decreases on machined surface (Ra between 0.2 µ m and 2 µ m). The effects of microstructure are then analyzed. First, the texture is studied and the velocity variation due to rolling (reduction rate of 50%) is measured. This method enables also to detect grain size variations in the 10µm-100µm range. At last, we show a 0.4% velocity variation due to machining induced damage.
An impulse Modeling of the generation and the propagation phenomena is also performed for several materials. The different reflected waves and the attenuation are taken into account in the computations in order to describe accurately the wave-front shape and the amplitude of the measured acoustic signals.