TRIAL INSPECTION OF EXTERIOR-TILE WALL SPECIMENS WITH A PROTOTYPED TILE DEBONDING DIAGNOSTIC DEVICE

Abstract

 Exterior tile cladding can suffer debonding or adhesive failures as it ages. Failures that result in cladding tiles falling from the building are highly dangerous for people passing beneath, and consequently it is important to be able to identify debonding defects at an early stage to prevent tiles falling. This study aimed to develop a high-precision device for efficient detection of exterior tile debonding. This paper details the results of verification of a tile debonding diagnostic device and the detection parameter proposed in the previous paper by testing twelve exterior-tile wall specimens with artificially debonded areas. The device is characterized by having an impact force sensor and four microphones, and is calibrated using a glass-made standard specimen. The detecting parameter R_AF is measured at each microphone. The outline and the results of this paper are as follows.

 

 1) Twelve exterior-tile wall specimens with artificially debonded areas were prepared. The types of tiles, bonding materials, debonded interface, debonded depth and debonded area were different for each specimen.

 2) Among all the impact points on each specimen of X0, X10, X20 and X30 having different debonded depth respectively, three characteristic points were selected, that is, point “a” as a representative point in the non-debonded area, point “b” as one at the center of debonded area and point “c” as one in the vicinity of the debonded boundary. The R_AF value at each microphone was displayed as the size of diameter of circle at each microphone’s position on each figure of the specimens. In the case of impact at point “b”, the center of the debonded area, the R_AF values of four microphones were much larger than those of point “a” in the non-debonded area, and the four R_AF values were almost same as each other. In the case of point “c”, in the vicinity of the debonded boundary, the R_AF value of the microphone positioned inside the debonded area was larger than the others. Therefore it is easier to detect debonding where at least one of the microphones is located inside the debonded area. In the case of point “a”, in the non-debonded area, all of the R_AF values were very small, even if one microphone was in the debonded area.

 3) The maximum value of R_AF of four microphones was defined as R_AF-max, and the results of R_AF-max were displayed as diameters of circles at the impact points on figures of the specimens. A threshold value of R_AF-max was set based on the distribution tendency of R_AF-max, and the results of detection of debonding were displayed on figures of the specimens. In the case of specimens with shallow debonded areas, it was possible to accurately detect debonding in the vicinity of the debonded boundaries. In the case of specimens with deep debonded areas, it became difficult to detect debonding clearly in the vicinity of the debonded boundaries, but the existence of debonding could be detected clearly further inside the debonded area. In the case of the specimens manufactured in this paper, it was possible to detect the existence of debonding to a depth of 39 mm.

 4) As stated above, in this paper, the effectiveness of the prototyped device and the proposed detection parameter was verified.

 

 Further research will be continued in order to develop reasonable methods for zoning the detected debonding points and determining debonding depth. In addition, on-site experiments will be carried out for further verification and improvement of the device and the system.