Accurate flaw sizing is essential for evaluating weld quality. While ultrasonic testing is widely used, the precision of common manual sizing techniques varies significantly. This study establishes controlled benchmark data for three established methods: distance amplitude correction (DAC), distance gain size (DGS), and the 20-dB drop technique, tested under strictly controlled laboratory conditions. Artificial defects were embedded in shielded metal arc welds and verified by radiography before ultrasonic inspection with 45° and 70° probes. DAC signals often exceeded reference levels, making sizing impossible. DGS was limited to flaws smaller than 10 mm and consistently underestimated size. The 20-dB drop method achieved the highest accuracy, measuring a 4.00 mm defect as 4.02 ± 0.02 mm, with stable results across probe angles. The findings demonstrate that a simple geometric approach can yield reliable, repeatable sizing under laboratory conditions. The study’s scope was limited to lateral sizing by a single operator; future work should expand to multi-operator trials, height measurements, and field conditions for full validation.
Eljaafari et al. (Thu,) studied this question.