Minimization of defects generation in laser welding process of steel alloy for automotive application

Laser welding (LW) thanks to its flexibility, limited energy consumption and simple realization has a prominent role in several industrial sectors. LW process requires careful parameters' tuning to avoid generating internal defects in the microstructure or a poor weld depth, which reduce the joining mechanical strength and result in waste. This work exploits a supervised machine learning algorithm to optimize the process parameters to minimize the generated defects, while catering for design specifications and tolerances to predict defect generation probability. The work outputs a predictive quality control model to reduce non-destructive controls in the LW of aluminum for automotive applications.