Informatics Publishing Limited

M. Sasidhar ¹, A. M. K. Prasad ², K. Jayanendra Babu ¹

Affiliations
1 Dept. of Industrial and Production Engg., K L College of Engg., Vaddeswaram, IN
2 Dept. of Mechanical Engg., College of Engg., Osmania University, Hyderabad, IN

Abstract

In late 1991, The Welding Institute (TWI), a British research and technology organization, invented and patented a welding process named Friction Stir Welding (FSW). Since its introduction, FSW development has progressed rapidly and is currently being used in a number of aerospace and industrial applications. Friction stir welding uses the high rotational speed of a tool and the resulting frictional heat created from contact to crush, stir together, and forge a bond between two metal alloys. A welding tool moves along the area to be joined while rotating at a high speed. The action between the tool and the aluminum creates frictional heat, which softens the aluminum but does not melt it. The plasticized material is then, in essence, consolidated to create one piece of metal where there were originally two. In this paper, a FSW process with varying pin geometries (cylindrical and conical) and advancing speeds is numerically modeled, and a thermo-mechanically coupled, rigid-viscoplastic, fully 3D FEM analysis able to predict the process variables as well as the temperature distribution. The obtained results allow finding optimal tool geometry for improving nugget integrity of aluminum alloys.

Full Text

     

K. Rama Kotaiah ¹, J. Srinivas ², K. Jayanendra Babu ¹

Affiliations
1 Department of Industrial and Production Engg., K.L. College of Engineering, Vaddeswaram, Guntur (Dist.), IN
2 Department of the Mechanical Engineering, Chaitanya Engineering College, Visakhapatnam, IN

Abstract

This paper presents, a control methodology based on experimental data of the tool wear as a function of cutting variables. In automatic machine tools, there is strong need to control the tool wear by adjustment of the cutting parameters. In this connection, a control system, which can adjust the cutting parameters for a desired wear rate, is necessary. A regression relation is also established between the flank-wear, and the cutting parameters. An inversely trained neural network model, which supplies the modified values of the cutting parameters, is used as a controller. The results are shown in the form of tables and graphs.

Full Text