EUROPEAN CONGRESS AND EXHIBITION ON ADVANCED MATERIALS AND PROCESSES - EUROMAT 2021

In alloy casting, primary dendrite arm spacings (PDAS) affect the mechanical properties of individual grains, solute (micro)segregation, defects (e.g. freckles), but also electrochemical properties (e.g. corrosion resistance). Moreover, in the same alloy, the PDAS may vary by orders of magnitudes when processed under different conditions. Usual models rely on power laws linking PDAS to processing conditions (e.g. temperature gradient, cooling rate) and alloy phase diagram (e.g. partition coefficient). Macroscopic volume-averaged models may incorporate the effect of the PDAS, but they do not directly predict them. Meanwhile, microscopic-scale physics-based models, such as phase-field, suffer from a high computational cost, such that they can only be compared to reduced-scale experiments [1].

Here, we combine phase-field (PF) and dendritic needle network (DNN) models to predict PDAS in Al-Cu alloys [2]. We compare our results to measurements from an extensive literature review and from instrumented lab-scale casting experiments. Using a dendrite tip selection constant calculated with PF in our DNN simulations, first we show that both models lead to similar results for a dilute Al-1wt%Cu alloy, and then we upscale our simulations to a Al-4wt%Cu alloy (too computationally demanding for PF due to the scale separation between tip radius and diffusion length). Our simulations show that PDAS can be calculated by directly combining physics-based models — with barely any adjustable parameters. They also highlight novel fundamental observations on PDAS selection, such as a widening of the PDAS stability range with a decrease of the temperature gradient.

[1] A.J. Clarke, et al., Acta Materialia, 129 (2017) 203-216. https://doi.org/10.1016/j.actamat.2017.02.047

[2] B. Bellon et al., Acta Materialia, in press (2021). https://doi.org/10.1016/j.actamat.2021.116686