Analyse expérimentale et modélisation thermomécaniques du soudage par friction de plaque de polymère

Abstract

This thesis focuses on the study of Friction Stir Welding (FSW), an innovative solid-state joining technique, applied here to high-density polyethylene (PEHD), a widely used thermoplastic polymer in the industrial sector. Unlike conventional welding processes that rely on the melting of materials, FSW produces solid joints through the frictional heat generated by a rotating tool, thus avoiding common defects such as porosity and thermal distortion. The main objective of this work is to investigate the influence of key process parameters — tool rotation speed, feed rate, and tilt angle — on the weld quality. An experimental approach was adopted, supported by a statistical modeling framework based on the Design of Experiments (DOE) methodology, with the aim of optimizing the welding conditions. The obtained results demonstrate the effectiveness of FSW in welding PEHD, producing joints with good mechanical performance. This study highlights the potential of this process for integration into industrial applications requiring high reliability, strength, and durability of welded assemblies. In addition, an artificial intelligence (AI)-based approach was used to analyze and model the experimental results. This method enabled us to effectively predict the quality of welded joints as a function of welding parameters, and to optimize process conditions.