Multiscale thermomechanical modelling of additive friction stir deposition for magnesium and aluminium alloys: Analytical and finite element approaches

Background and Objective: Industrial adoption of Additive Friction Stir Deposition (AFSD) for lightweight alloys is hindered by the cost of trial-and-error parameter optimisation. This study develops a multiscale thermomechanical framework that integrates analytical modelling with three-dimensional finite element analysis to predict processing windows for AA6061 aluminium and AZ31B magnesium alloys, enabling a direct comparison of alloy-specific thermal sensitivities. Methods: A two-stage modelling strategy was adopted. First, an analytical model was used to estimate peak contact temperatures and to construct processing maps from frictional heat generation, using a homologous temperature criterion ( 0 . 6 … 0 . 9 ) T solidus as a stability window. Second, coupled three-dimensional thermomechanical simulations were performed in Ansys LS-DYNA using an Arbitrary Lagrangian–Eulerian formulation and a Johnson–Cook constitutive description for AZ31B and AA6061. Full-sticking contact conditions were assumed and benchmarked against friction stir welding studies. Results: The analytical predictions agreed well with the finite element results, with peak-temperature deviations below 8%. Parametric analysis across the full operational window (200–1000 rpm tool rotation speed, 2.5–7.5 mm tool radius, and 0.5–4.0 mm feed depth) identified thermally stable regions within the homologous temperature criterion for both alloys. The simulations further indicated that AZ31B exhibits higher thermal gradients and larger effective plastic strain (EPS) zones than AA6061, consistent with its lower thermal conductivity and hexagonal close-packed crystal structure. A representative regime (500 rpm and 2.0 mm depth) yielded EPS > 1 . 5 for dynamic recrystallisation while remaining below 0 . 85 T solidus . Conclusions: The proposed analytical–numerical framework supports industrial AFSD calibration by reducing the need for extensive experimental screening and by providing quantitative guidance for parameter selection in aerospace and automotive lightweight-component manufacturing.