Abstract
The hammering test is widely used as a non-destructive testing method for inspection of internal defects in concrete structures. In this paper, we use a rotary hammer instead of a conventional inspection hammer to improve the inspection efficiency and the hammering force variation. We propose a method based on the self-organizing map (SOM) for automatically distinguishing the defective parts from the healthy parts of concrete structures. The hammering test experiments of concrete specimens with artificial defects are conducted. The hammering sounds are measured with a microphone moving with the rotary hammering device. The frequency spectra of hammering sounds measured at the impact locations are used as input data for the SOM and the impact locations are partitioned into two groups of the defective and the healthy parts from the SOM results. The results show that for the frequency spectrum, a dominant peak frequency varies depending on the defect depth at the defective parts. After the SOM map is trained, the unit with the smallest average overall value has almost all the impact locations of the healthy parts. By removing the impact locations belonging to its unit, the impact locations corresponding to the defective parts are extracted. The extracted impact locations identify the exact locations and widths of the defects.
