Investigation of relationship between resistance spot welding condition and nugget shape by utilizing machine learning based technique
Houichi KITANO, Akira SATO, Muneyoshi IYOTA, Terumi NAKAMURA
pp. 53-59
DOI:
10.2207/qjjws.38.53Abstract
In this study, the effects of resistance spot welding conditions on the nugget diameter, which was one of the major influencing factors of resistance spot weld joint strength, was modeled by a machine learning method which had been proposed by the authors in recent years. Then, the applicability of constructed model and the effect of resistance spot welding conditions on the nugget diameters were discussed. The feature of the machine learning method used in this study was that the relationship between input and output could be derived as an easy-to-understand mathematical expression. A resistance spot welding condition-nugget diameter database was created through experiments using 590MPa class steel plates, and a nugget diameter prediction model was constructed to reproduce the database appropriately. As a result, it was indicated that the nugget diameter prediction model can predict the nugget diameter under welding conditions used for model construction and those not used precisely. Furthermore, it was found that the nugget diameter prediction model was composed of two terms that were presumed to reflect the spread of material melting due to heat input and the phenomenon at the beginning of energization.