In-Process Detection of Faying Surface Sealant Application Flaws in Friction Stir Welding

Abstract

In this study, a process for detecting faying surface sealant application flaws in friction stir welded (FSW) lap joints is developed. The process uses a technique shown previously to enable the detection of machined gaps in the same joint type. This technique involves computing the frequency spectra of the process forces and reducing the dimensions of the data using well-known methods for discrimination purposes. Aluminum alloys 2024-T3 and 7075-T6 in 0.063-in.- (1.6-mm-) thick sheets were welded with a variety of PR-1432-GP sealant configurations, including in both the cured and uncured state, and applied in the tool path and adjacent to the tool path. It is shown that sealant flaws such as gaps or thin spots can indeed be discriminated from control welds with proper sealant application, and the success of this technique depends directly on the input force signal, the sealant configuration, and the dimensional reduction method. Factors affecting the real-time implementation of this technology in aerospace manufacturing are also examined.