@article{M. Yusof_Ishak_Ghazali_2019, title={L-Statistical Analysis of Sound Signal Acquired from Pulse Mode Laser Welding for Characterising Weld Geometry}, volume={16}, url={https://journal.ump.edu.my/ijame/article/view/832}, DOI={10.15282/ijame.16.3.2019.12.0524}, abstractNote={<p>Many ongoing studies have proven that statistical features extracted from the acquired sound during a laser welding process significantly yield some pictures on the weld condition, including weld geometry. However, a considerable amount of studies has underlined the use of common statistical features in which they are restricted to some limitations when dealing with non-stationary random sound signal. In the present study, the main aim is to study the correlation between the L-statistical features trend of the sound amplitude distribution with respect to the change in weld geometry during pulse mode laser welding compared to common statistical features. In an attempt to achieve this goal, a pulse mode laser welding was subjected to 22MnB5 boron steel with variations in the peak power, pulse width, and focal length. Meanwhile, the sound signal was acquired during the process, with standard deviation, interdecile range, mean absolute deviation, L-Cv (scale), and L-kurtosis extracted from the analysis. The degree of correlation between these statistical features and weld geometry was compared from the R-square value. According to the reported results, L-kurtosis yielded the strongest correlation with both weld penetration depth and bead width compared to the remaining five statistical features. This showed that the use of L-statistical features was significant to improve the correlation between sound signals and weld geometry.</p>}, number={3}, journal={International Journal of Automotive and Mechanical Engineering}, author={M. Yusof, M. F. and Ishak, M. and Ghazali, M. F.}, year={2019}, month={Oct.}, pages={6987–7006} }