Scholarlink

<p>Austenite grain size is a critical factor governing the mechanical properties of hot-rolled steel plates, yet the evolution of austenite grain size during non-isothermal heating has rarely been investigated. This work aimed to clarify the non-isothermal austenite grain growth behavior and develop accurate predictive models for industrial heating parameter optimization. The high-temperature laser scanning confocal microscope was used to simulate hot-rolling heating processes, with samples subjected to isothermal holding at 950 ~ 1250&#xa0;°C for 30&#xa0;min and dynamic heating rates at 0.5 ~ 30&#xa0;°C/s up to 1250&#xa0;°C, and grain size evolution was observed throughout. The results indicated that grain size increases with higher temperature, lower heating rate or longer holding time. Besides, when the heating rate exceeds 10&#xa0;°C/s, further increasing the heating rate exerted little influence over the grain size. Both physics-based and machine learning models were established for prediction. The machine learning model exhibits superior performance, with a mean squared error of 0.0098 and an average absolute relative error of 0.12%, and its results conform to physical and metallurgical laws. This model provides a reliable tool for optimizing heating parameters to achieve target grain sizes, thereby improving the mechanical properties of hot-rolled steels.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Investigation into Non-Isothermal Grain Growth of Austenite Grains of Medium Carbon Steel during Heating

  • 期刊论文

  • 发表日期：

    2026年2月6日

查看全文
  • Rong Ran,
  • Min Zhou,
  • Huiquan Han,
  • Gang Fang

摘要

Austenite grain size is a critical factor governing the mechanical properties of hot-rolled steel plates, yet the evolution of austenite grain size during non-isothermal heating has rarely been investigated. This work aimed to clarify the non-isothermal austenite grain growth behavior and develop accurate predictive models for industrial heating parameter optimization. The high-temperature laser scanning confocal microscope was used to simulate hot-rolling heating processes, with samples subjected to isothermal holding at 950 ~ 1250 °C for 30 min and dynamic heating rates at 0.5 ~ 30 °C/s up to 1250 °C, and grain size evolution was observed throughout. The results indicated that grain size increases with higher temperature, lower heating rate or longer holding time. Besides, when the heating rate exceeds 10 °C/s, further increasing the heating rate exerted little influence over the grain size. Both physics-based and machine learning models were established for prediction. The machine learning model exhibits superior performance, with a mean squared error of 0.0098 and an average absolute relative error of 0.12%, and its results conform to physical and metallurgical laws. This model provides a reliable tool for optimizing heating parameters to achieve target grain sizes, thereby improving the mechanical properties of hot-rolled steels.