Visualizing Defects of Irregular Weld Seams Based on MultiMode TFM Detection

Abstract

An image reconstruction method based on the multimode total focusing method (MTFM) is proposed to overcome the limitations of traditional total focusing method (TFM) imaging in detecting tiny discontinuities at complex locations. We conducted MTFM detection and TFM image reconstruction experiments for irregular welds containing multiple discontinuities. In an experiment using four 1 mm diameter manufactured defects fabricated on two aluminum alloy welded test blocks, we achieved two significant contributions. First, we accurately detected small discontinuities by combining CIVA simulation with robotic arm assistance. Second, we proposed fusion factor coefficients for TFM image processing, which considered different modal weights for image fusion and de-noising, thereby preserving the integrity of the fused images. Our experimental results demonstrate that the reconstructed TFM images effectively represented all defect information. Compared with other modal TFM images with the highest signal-to-noise ratio (SNR), the amplitude-corrected optimized TFM image exhibits an improved SNR of 51.95% without losing defect information.