Experimental estimation of the longitudinal-wave backscatter coefficients for ultrasonic interrogation of weak scattering materials

Abstract

The starting point for solving inverse scattering problems in acoustics is an estimate of the target's scattering function. A typical procedure for estimating a scattering function is to apply an inverse filter to the Fourier transform of the backscattered signal from the target. The inverse filter is based on an estimate of the measurement system response function that includes electronic, transducer, and wave propagation related effects. In this paper, we address the problem of estimating the backscatter coefficient for scattering from a large number of randomly distributed scatterers. A method is presented for estimating the backscatter coefficient for weak scattering polycrystalline metals for the case in which the backscatter coefficient is independent of depth into the material. Measurement system response function estimates were based on volume integrations over velocity fields calculated using GaussHermite polynomials. Ultrasonic measurements were made for direct backscatter in an immersion mode at normal incidence using both planar and focused probes operated broadband. Results are presented that show close agreement between the frequency dependence of the backscatter coefficient estimates and the frequency dependence of the classically estimated attenuation coefficient. The backscatter coefficient estimates are shown to be reasonably independent of the system configuration and depth into the material.

Ultrasonic attenuation in polycrystals using a self-consistent approach

“…In this scope, the SC model is expected to be a more suitable foundation from which more complicated models could be based, e.g., models accounting for grain elongation, macrotexture, grain size gradients, multiphases, defects, etc. Many of these more complicated microstructures have already been theoretically addressed using Voigt averages [38][39][40][41][42][43]. The present SC theory can be easily implemented by simply including the anisotropy coefficient γ and self-consistent elastic moduli C SC .…”

Testing for Nongaussian Fluctuations in Grain Noise

“…Consider in more detail the problem of estimating the number of significant scatters per resolution cell. For any chosen time within a grain noise A-scan and assuming single scattering, the incident pulse length and time-of-flight considerations dictate the material volume, called the resolution cell, in which all scattering must occur in order to contribute to the A-scan at the chosen time [11]. If we think of an idealized situation in which all of the randomly placed scatterers within the resolution cell are identical and the scatterers are bathed with a uniform sonifying field, then if the number of scatterers per unit volume and the size of the resolution cell are known, the number of contributing scatterers can be quantified in a meaningful way.…”

Creating Noise-Corrupted Flaw Signals for Ultrasonic Nondestructive Evaluation

“…The model can be stated as [12] f(t) = h(t)*r(t) + na(t) + ne(t) (1) where na(t) represents acoustic noise, and ne(t) represents electronic noise. The model can also be stated in the frequency domain as [12] …”

“…In Equations (1) and (2), the dependence on material and measurement system configuration parameters is implicit. The measurement system frequency response function, H(ro), is the response function for a single scatterer [1,12].…”