Control oriented prediction for FSW based on experimental data and conventional software tools

Abstract

This document presents a control-oriented strategy to predict temperature distribution during the friction stir welding (FSW) process. The proposal uses a programming platform that solves heat transfer problems based on material thermal properties and geometry, internal and external heat sources, and boundary conditions for transient problems. Finally, the strategy provides a single equation for predict the heat dissipation from operative FSW parameters. The results are evaluated by comparing experimental and simulated temperature results and coherence between predicted and experimental temperature distributions. A model such as the presented allows designing a model-based control structure for the FSW process. In that sense, the weld quality is the controlled variable and the options for manipulated variables are the travel and rotational speeds of the tool.