An innovative inverse analysis based on the Bayesian inference for concrete material

Nondestructive tests and evaluations are robust techniques for inspecting different attributes of concrete configuration. However, most nondestructive techniques focused on an aspect of concrete configuration based on comparison to other samples. In this paper, an innovative inverse analysis technique was developed to inspect different attributes of concrete configuration simultaneously. The methodology was based on the scattering feature of the ultrasonic waves during propagation in heterogeneous media. The transition matrix method was employed to determine the scattered wavefield. This method considers the shape of objects, unlike most other numerical methods. Furthermore, a novel algorithm was presented to establish a realistic space in three-dimensional for concrete. The Voronoi diagram and shrinking process established the framework of the algorithm. The inverse model conducted observation data from media to concrete configuration through the direct model. The inverse procedure extracted vast information from the medium. Statistical theory provided statistical inference based on Bayesian statistics for this procedure. The introduced inverse analysis technique then scrutinized the concrete specimens. For this aim, geopolymer concrete with different configurations was nominated as a sample of concrete material. In the end, the precision, accuracy, and validity of the inverse model solution were assayed in the light of statistics. The assessments demonstrated that the proposed method provided a comprehensive description of the overall concrete configuration.