Linear Friction Welding of Aerospace Materials: Modelling and Validation

Linear friction welding (LFW) is a solid state joining process in which a relative reciprocating linear motion between parts is used to generate heat at the weld interface. LFW is used in blisk (integrally bladed disc) manufacture to join blades and discs. Blisks can offer significant weight savings, increased performance, increased fuel economy and reduced CO₂ emissions. Much LFW research has focused on titanium alloys, however other materials have also been investigated (e.g. steels and superalloys) [1,2,3].

The linear friction welding process can be divided into three distinct phases: conditioning, equilibrium and deceleration. Initial heating until material starts to be extruded in the so called “flash” is known as conditioning. Equilibrium phase describes the steady state between heating at the weld interface and material extrusion. The final phase is deceleration when the oscillating movement is stopped and the two parts are aligned [4]. The main process parameters have been identified to be amplitude, load and pressure [5].

A finite element model has been used to analyse the influence of these parameters on the conditioning and equilibrium phase for LFW of the titanium alloy Ti6Al4V. This has enabled prediction of the evolution of temperature profile during conditioning. These predictions were applied as input conditions to predict upset rate and flash morphology during the equilibrium phase, allowing modelling the process from beginning to end [Figure 1]. Modelling results were validated by targeted experimentation using laboratory-scale apparatus. Thermocouples were applied to record temperatures during conditioning and equilibrium phase [Figure 2]. Temperature measurements were coupled with high-speed digital photography to link temperatures to the actual position of the workpieces for validation purposes and to visualize the flash formation [Figure 3]. The comparison of modelling and experimental results showed good agreement, confirming numerical predictions using finite element model are possible.