Abstract:
This research presents the applied acoustic emission signal (AE) in classification of nugget quality in resistance spot welding process
both directed current (DC) and alternated current (AC). The AE signal was denoised by Wavelet Transform and conducted to obtain
AE parameters which consisted of AE-RMS, Amplitude and Count to find the relation which used to specify nugget quality by comparison
with strength, nugget size and expulsion during the welding process. From analyzed results, AE parameters could be concluded that they
had characteristic which was able to classify nugget quality. This experiment was utilized by using Neural Network to learn and analyze
the accepted nugget quality through data sets of denoised AE parameter. The results were shown that there were accuracy in classification
by 85 percent for DC spot welding and by 89 percent for AC spot welding respectively. Therefore, the acoustic emission signals detected
from Non-Destructive Testing (NDT) have ability to use in order to classify nugget welded quality both DC and AC resistance spot welding
effectively.