基于元胞自动机模型的铝合金晶粒尺寸预测

Prediction of Grain Size of Aluminum Alloy in Casting Process Based on Cellular Automaton Model

铸造过程期望获得细小均匀的等轴晶粒,但晶粒尺寸难以直接测量。因此,建立可靠的数学模型来预测晶粒尺寸是必要的。基于元胞自动机(CA)模型模拟枝晶生长,该模型在基于枝晶生长动力学基础上模拟铝合金凝固过程中的微观组织演变,以此获得平均晶粒尺寸。CA模型模拟结果与Lipton Glicksman Kurz(LGK)模型结果基本吻合,对AA5182铝合金搭建重力铸造实验,验证了模型的准确性。铸造过程中的晶粒形态根据温度梯度和冷却速度变化,进行一定规模计算总结平均晶粒尺寸与温度梯度、冷却速率的关系以预测铝合金铸件的平均晶粒尺寸。计算结果显示,降低温度梯度和增加冷却速度有助于细小晶粒尺寸的形成。

Fine and uniform equiaxed grains are expected to be obtained during the casting process,but the average grain size is difficult to measure directly.Therefore,establishing a reliable mathematic model to predict the grain size is necessary.In this research,a cellular automaton(CA)model was established.This model based on dendrite growth kinetics can simulate the microstructure evolution during solidification of aluminum alloy and obtain the average grain size.The simulation results were in agreement with Lipton Glicksman Kurz(LGK)model,and the growth of single dendritic,multi-dendritic and columnar crystals can be acquired by this model.A gravity casting experiment of AA5182 aluminum alloy was conducted to verify the accuracy of the model.The grain morphology during the casting process changes on the basis of the temperature gradient and cooling rate.An average grain size map was created as a function of temperature gradient and cooling rate to predict the average grain size of the aluminum alloy castings.This map showed that decreasing the temperature gradient and increasing the cooling rate were contributed to the refinement of the grain size.