The initial solidification process of microalloyed steels was simulated using a confocal scanning laser microscope, and the growth behavior of austenite grain was observed in situ. The method for measuring the initial austenite grain size was studied, and the M ^*₀ (the parameter to describe the grain boundary migration) values at different cooling rates were then calculated using the initial austenite grain size and the final grain size. Next, a newly modified model for predicting the austenite grain size was established by introducing the relationship between M ^*₀ and the cooling rate, and the value calculated from the modified model closely corresponds to the measured value, with average relative error being less than 5%. Further, the relationship between T^γ (the starting temperature for austenite grain growth) and equivalent carbon content C_P (C_P > 0.18%) was obtained by in situ observation, and it was introduced into the modified model, which expanded the application scope of the model. Taking the continuous casting slab produced by a steel plant as the experimental object, the modified austenite grain size prediction model was used to predict the austenite grain size at different depths of oscillation mark on the surface of slab, and the predicted value was in good agreement with the actual measured value.

用共聚焦扫描激光显微镜模拟了微合金钢的初始凝固过程，并在原位观察了奥氏体晶粒的生长行为。研究了测量初始奥氏体晶粒尺寸的方法，然后使用初始奥氏体晶粒尺寸和最终晶粒尺寸计算了不同冷却速率下的M ^* ₀（描述晶界迁移的参数）值。接下来，通过引入M ^* ₀之间的关系，建立了新的奥氏体晶粒尺寸预测模型 冷却速率和根据修改后的模型计算出的值与测量值非常接近，平均相对误差小于5％。此外，关系Ť ^γ（起始温度为奥氏体晶粒生长）和当量碳含量Ç _P（C ^ _P > 0.18％）通过原位观察获得，并将其引入修改后的模型中，扩大了模型的应用范围。以某钢厂生产的连铸板坯为实验对象，采用改进的奥氏体晶粒尺寸预测模型对板坯表面不同振痕深度处的奥氏体晶粒尺寸进行预测，预测值吻合良好。与实际测量值。