An experimental evaluation of two effective medium theories for ultrasonic wave propagation in concrete.

This study compares ultrasonic wave propagation modeling and experimental data in concrete. As a consequence of its composition and manufacturing process, this material has a high elastic scattering (sand and aggregates) and air (microcracks and porosities) content. The behavior of the "Waterman-Truell" and "Generalized Self Consistent Method" dynamic homogenization models are analyzed in the context of an application for strong heterogeneous solid materials, in which the scatterers are of various concentrations and types. The experimental validations of results predicted by the models are carried out by making use of the phase velocity and the attenuation of longitudinal waves, as measured by an immersed transmission setup. The test specimen material has a cement-like matrix containing spherical inclusions of air or glass, with radius close to the ultrasonic wavelength. The models are adapted to the case of materials presenting several types of scattering particle, and allow the propagation of longitudinal waves to be described at the scale of materials such as concrete. The validity limits for frequency and for particle volume ratio can be approached through a comparison with experimental data. The potential of these homogenization models for the prediction of phase velocity and attenuation in strongly heterogeneous solids is demonstrated.