Computational modelling of linear friction welding of Inconel 718

Abstract

Linear friction welding (LFW) is a solid-state joining technique that involves the rapid reciprocating motion of two workpieces under large compressive pressure, in order to generate friction heat and plastically deformed material at the contacting surfaces of workpieces. It is an advanced joining technology used for manufacturing and repairing complex assemblies like blade integrated disks (blisks) of aeroengines. LFW can avoid fusion welding defects such as hot cracking, pores, pin holes, solute segregation and solidification structures because it does not involve the remelting of weld material. In practical terms, it can be very difficult to use experimental methods to characterise some phenomena and processes such as the evolution of stress, strain, strain rate and microstructure of welds during the LFW. These challenges have motivated high demand for computational modelling of LFW, which can be used to not only predict these phenomena and processes but also explain or interpret the relationship between heat transfer, deformation of weld, and material microstructural evolution during the LFW. This research enabled the integrated computational modelling of the LFW process for the manufacture of Inconel 718 (IN718) alloy welds by sequentially (one-way) coupling a thermomechanical sub-model with two different microstructural sub-models. The thermomechanical sub-model of the integrated computational modelling involved two-dimensional and three-dimensional (2D and 3D) computations developed for two deformable IN718 workpieces. The thermomechanical sub-model employed Hooke’s law for material elasticity and the strain-compensated Arrhenius constitutive model for material plasticity. One microstructural sub-model formulated the process of dissolution of the δ phase of IN718 by using the time-temperature equivalence method. The other microstructural sub-model formulated the dynamic recrystallization of the primary γ grains during the LFW by using the Johnson-Mehl-Avrami-Kolmogorov (JMAK) model. The integrated computational modelling was implemented within the general-purpose finite element software package Abaqus in conjunction with related custom-written subroutines. On the scale of the overall weld, the integrated computational modelling predicted the macroscopic processes of flash formation, axial shortening, and the evolution of displacement, temperature, stress, strain and strain rate of IN718 during the LFW. The integrated computational modelling also predicted the volume fraction of the δ phase, the volume fraction of recrystallized γ grains, and the average size of γ grains of the weld. The integrated computational modelling was verified by comparing its modelling results of weld temperature, axial shortening, formation of flash, volume fraction of the δ phase and average size of γ grains of the weld to related experimental results of other researchers. The integrated modelling was subsequently used for optimising the LFW process parameters and determining the process windows of the LFW of IN718. By systematically analysing the influence of 10 to 20 different sets of LFW process parameters (using different combinations of friction pressure, oscillation frequency and oscillation amplitude), the friction pressure was identified as the most influential process parameter determining the weld temperature, axial shortening, dissolution of the δ phase and the DRX of γ grains during the LFW of IN718. It is the first time that such integrated computational modelling has been developed for LFW of an alloy. With regard to the contribution to the body of knowledge, the integrated computational modelling provided insight into the relationship and interaction between heat transfer, deformation of weld, and material microstructural evolution during the LFW of IN718. With regard to the contribution to the industry, this project’s integrated computational modelling developed an effective and efficient tool that can be directly used by the manufacturing industry to design and/or optimise its process parameters for LFW of IN718.