Fundamental investigation on the mechanical behavior of weld seams

The solid state welding processes allow joints free from defects with reduced residual stresses and low distortion. However, such processes are generally characterized by a complex mechanical junction because of the particular material flow. Therefore, the design and optimization of the processes in which the solid bonding phenomenon occurs are difficult and requires a large number of tests. The purpose of this research is the description on the mechanical behavior of welded seams obtained through the friction stir welding (FSW) and linear friction welding (LFW) processes by analyzing the same phenomenon that occurs in both metallurgical processes, named the solid bonding phenomenon. The aim of this research is the development of a technique for the FSW and LFW processes, able to quantitative determine the weldability degree of a material as a function of the technological parameters of the processes. The description of this phenomenon has been done through the solid bonding criteria. The development and the applicability of the analytical bonding criteria, originally developed for different manufacturing processes, has been studied on the FSW and LFW processes in order to obtain an integrated numerical tool able to predict the occurrence of solid bonding starting from the process input parameters. LFW and FSW experimental campaigns AA6082 and AA6061 aluminum alloys were carried out with the aim to study the effects of the process parameters on the quality of the welded joints. Appropriate process windows have been defined for both the processes, to acquire the data needed to implement the integrated numerical tool of the processes. At the same time, a numerical model of the processes have been developed and set up, obtaining information on the field variables needed to implement the bonding criterion proposed. The experimental and numerical campaign led to an exhaustive knowledge of the material flow phenomena as function of the process operative parameters, thus obtaining effective design tools for both the solid state welding processes. The applicability of the pressure-time criterion has been tested on the FSW process and LFW process using the two different aluminum alloys. Additionally, the Neural Network was developed able to predict, starting from the evolution of temperature, strain, strain rate and pressure on the bonding line, the occurrence of the solid bonding as well as quantitative parameter Q providing information on the “quality” of the weld.