Étude expérimentale et numérique de la réparation des polymères fissurés par friction malaxage

Abstract

This work falls within the scope of repairing polymer structures damaged by cracking, proposing Friction Stir Welding (FSW) as an innovative alternative to conventional fusion welding processes. Following a literature review dedicated to the properties of High-Density Polyethylene (HDPE), the cracking mechanisms of semi-crystalline polymers, and the limitations of classic welding techniques, the study utilizes the Response Surface Methodology (RSM) / Design of Experiments (DoE) to quantify the influence of the FSW operating parameters—rotational speed, traverse speed, and tool tilt angle—on the maximum temperature developed within the weld zone. A full factorial design was employed to establish a predictive regression model, complemented by an artificial intelligence approach based on an Artificial Neural Network (ANN) to refine the thermal response modeling. The obtained results confirm that the process takes place entirely in a steady visco-plastic state, below the crystalline melting point of HDPE, and highlight the predominant role of the rotational speed and its interaction with the tilt angle. The identified operating conditions open promising perspectives for the industrial application of FSW in repairing thermoplastic polymer components, while emphasizing the need for additional testing to consolidate the statistical robustness of the proposed model.