A STUDY OF INTERNAL STRUCTURE DIAGNOSIS OF CONCRETE STRUCTURES USING DIGITAL HAMMERING INSPECTION AND MACHINE LEARNING

Abstract

In a previous study, the authors clarified the physical phenomena of concrete defects and digital hammering inspection results through experiments and FEM analysis, and developed a forward analysis model that outputs digital hammering inspection results from the defects state. In this study, an in-verse analysis model was developed to quantitatively evaluate the defects state of concrete from the digital hammering inspection results by machine learning the database created by the forward analysis model, and the accuracy of the evaluation and the effectiveness of the approach using the database were summarized.