Multi-objective optimization of laser metal-polymer joints for automotive and aerospace applications with evolutionary fuzzy systems

The increasing demand for lightweight, high-performance materials in automotive and aerospace sectors has driven the development of metal-polymer hybrid structures without adhesives or fastenings. Laser joining is a promising technique for fabricating such structures with high precision and efficiency. However, optimizing shear bond strength is challenging due to multiple interacting process parameters and material combinations. This study proposes a multi-objective evolutionary fuzzy system that integrates genetic algorithms and fuzzy logic to model, predict, and optimize shear strength efficiency of laser-jointed metal-polymer hybrids under uncertainty. Experimental data from six metal-polymer combinations, including stainless steel, aluminium, titanium, and various polymers, were used for training and validation. Results show significant variation in joint performance, with the titanium-PEEK combination achieving the highest shear strength efficiency (34.6%) and ultimate shear strength (20.6 MPa), surpassing typical structural requirements. Enhanced joint strength is attributed to surface texturization promoting mechanical interlocking and chemical bonding at the interface. Process parameters such as laser power, interaction time, and metal surface texture spacing strongly influence the joint strength. Reducing groove spacing increases textured area and adhesion but requires careful tuning to avoid polymer degradation. The fuzzy model effectively captures nonlinear behaviors and uncertainty, offering uncertainty bands that support decision-making. Inverse analysis suggests optimal conditions near 30 s interaction time, laser power above 155 W, textured area greater than 72 mm2, and a joint index around 11 for maximum joint reliability. This approach can reduce costly trial-and-error, improve process control, and has strong industrial relevance for lightweight metal-polymer hybrid manufacturing.