Improvement of hammering inspection system to detect defects for joint gap of concrete bridge

Abstract

In recent years, the deterioration of concrete structures has become a serious issue in Japan. The hammering test examined by human inspectors has been widely used for inspection of these concrete structures. However, joint gaps of concrete bridge, which are less than about 100mm, are so narrow that inspectors cannot get in and investigate. We have studied and developed automatic hammering machines to deal with the issue. The purpose of this study is to apply the hammering mechanism that can examine the hammering test for these joint gaps. We firstly developed the prototype to strike the surface of the joint gaps and record reflection sounds by microphone. Then an unsupervised learning method so called Self-Organizing Map (SOM) was applied to judge the soundness of their state. However, this prototype was found to have a problem that the defect detection performance was deteriorated by the echo sound generated by the gap space. Therefore, a new device was developed to measure the surface vibration by utilizing accelerometer and detection performance becomes higher without being affected by the echo sound.