A Study on Microstructure of AA2219 Friction Stir Welded Joint

Abstract

Abstract. Friction Stir Welding (FSW) is an environment friendly solid-state joining technology that uses a rotating tool to generate frictional heat and plastic deformation with mixing of the soft material by the pin along the welding line. The advantages are low distortion, lack of traditional defects commonly found in fusion welds and high joint efficiency. The paper focuses on microstructure evolved after welding and the effect of welding process is discussed for aluminum alloy 2219. This alloy is selected because of high strength and good resistance to stress corrosion cracking at cryogenic temperature. The optimum process condition for FSW is determined with microstructural and mechanical properties. It is shown that the FSW provides higher performance for this alloy than the conventional welding technology.

Conditions for void formation in friction stir welding from machine learning

“…From the previous experimental and modeling research, [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][23][24][25][26][27][28][29] it is known that the welding parameters affect important variables such as temperature, strain rate, torque, and maximum shear stress on the tool pin that affect properties of the plasticized alloy and its smooth flow. Since the disruption of the smooth flow is thought to cause the void formation, the values of these four variables are important for the void formation.…”

“…Efforts have been made to experimentally determine the effects of welding parameters on void formation in commonly used aluminum alloys. [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] Tracers have been used to examine experimentally how the flow of materials affect the void formation. [20][21][22] The time lapse determination of the tracer's position in some cases indicated that the voids occurred near the bottom of the pin where the flow of the plasticized material was interrupted.…”

“…Here we examine the effectiveness of supervised machine learning (ML) algorithms to forecast the void formation during FSW. One hundred and eight sets of data for the FSW of three aluminum alloys, AA2024, AA2219 and AA6061 obtained from the peer-reviewed literature [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] have been analyzed using neural network (NN) and decision tree (DT) to examine the effectiveness of ML to mitigate void formation. Vibration and poor fixtures may affect the quality of welds, at least in principle.…”