THE EVALUATION OF COEFFICIENT OF FRICTION FOR REPRESENTATIVE AND PREDICTIVE FINITE ELEMENT MODELLING OF THE INERTIA FRICTION WELDING

Inertia friction welding is the process in which stored kinetic energy in a flywheel is converted to heat by relative sliding movement between surfaces of axi-symmetric components to achieve a weld in the solid-state. The work in this paper relates to the production of dual-alloy shafts for aeroengines. Frictional characteristics determine the conditions at the weld interface and these are controlled by rotational velocity and applied axial pressure. So-called representative and predictive methods have been developed to evaluate friction conditions during the process and these are discussed in this paper. Weld data for the dissimilar weld between a high strength steel and a nickel-based super-alloy were provided by Rolls-Royce and MTU Acro Engines. The finite element software package DEFORM-2D is used to develop coupled thermo-mechanical axi-symmetric models. In previous work, methods employed to evaluate the efficiency of mechanical energy utilised during a weld, a parameter of great importance for numerical analysis, are not clear. Previous predictive approaches have employed test/weld data in one way or another to obtain the interface friction coefficient. This paper proposes a formula that incorporates the value of the mechanical energy efficiency of the welding machine into the calculation of coefficient of friction for representative modelling. It also introduces a predictive approach based on sub-layer flow theory to predict frictional behaviour during the welding process that is independent of test/weld data.