A multi-scale approach for simulating hot tearing during the DC casting of aluminum alloys is presented. The novelty of this approach lies in the combination of a macro-scale finite element simulation of the DC casting process with direct prediction of hot tears via a meso-scale multi-physics granular model. This approach is capable of simulating hot tearing initiation, growth, and propagation within a representative volume element of the mushy zone. The change of cooling conditions experienced by the DC cast billet as a result of variations in casting speed as well as non-uniformity of heat extraction from different locations of the billet affect the deformation state, cooling rate, and thermal gradient, which further influence the strain rate, grain size, permeability, and feeding coefficient. Considering all the mentioned parameters, the multi-scale approach emphasizes the fact that hot tearing is a phenomenon resulting from the combination of the tensile deformation and restricted feeding of the mushy zone. The developed hot tearing formation maps identify the locations where hot tearing will occur as predicted by the multi-scale approach for two alloys—AA5182 and AA3104—thus demonstrating the approach’s sensitivity to both processing parameters and alloy composition.

提出了一种模拟铝合金直流铸造过程中热撕裂的多尺度方法。这种方法的新颖之处在于将直流铸造工艺的宏观尺度有限元模拟与通过中尺度多物理场颗粒模型直接预测热撕裂相结合。该方法能够模拟糊状区域的代表性体积元素内的热撕裂引发，生长和传播。DC铸造坯料经历的冷却条件的变化是铸造速度变化的结果，并且从坯料的不同位置抽出热量的不均匀性会影响变形状态，冷却速率和热梯度，从而进一步影响铸坯的形变。应变率，晶粒尺寸，磁导率和进料系数。考虑所有提到的参数，多尺度方法强调了这样一个事实，即热撕裂是由拉伸变形和糊状区域的进料受限共同造成的一种现象。所开发的热撕裂形成图确定了两种合金（AA5182和AA3104）的多尺度方法所预测的发生热撕裂的位置，从而证明了该方法对工艺参数和合金成分的敏感性。