基于分子与多物理场模拟的高炉炉缸内形在线监测与优化方法

The working conditions of blast furnace are severe, especially for the blast hearth where the lining refractory is easily corroded and destroyed. This has been an important factor that limits the service life of modern blast furnaces. Focusing on the online monitoring of the lining erosion of the blast hearth, the project aims to solve problems including sensors failure, inaccurate sampling information and reconfiguration of physical fields at boundaries, by indentifying and fusing sampling data integrated with erosion mechanism analysis and statistical approach. Meanwhile, taking into account influences of the operational process on the thermo-physical performance of lining materials in the blast furnace, the molecular modeling will be included and models will be developed for both the diffusion process of molten slag/iron in the linings and performance predictions of the refractory linings. Instead of specifying temperatures of walls as generally used in numerical simulations, a two-step correction approach will be adopted for solution of inverse heat transfer problems and determination of the actual profiles of the hot surface. Moreover, a knowledge discovery method will be developed to extract and establish rules for judgement of hot surface profile from simulation results, and approaches on knowledge basis will be studied for online monitoring and optimization of the lining working status, therefore the time-consuming 3D numerical studies can be greatly reduced. The outcome of the project will be of great help to improve the system reliability of the online monitoring technology of the blast hearth inner profile, thus control abnormal lining erosion in time and prolong the service life of blast hearths.

在役炼铁高炉的炉缸工作环境恶劣，内衬材料易被侵蚀破坏，是制约现代高炉寿命的主要因素之一。本项目围绕炉缸内衬侵蚀状况在线监测方法，融合机理解析与统计方法对检测信息进行辩识和信息融合，解决传感器失效、检测信号失真以及边界物理场重构等问题；考虑到高炉运行过程状况对材料热物理性能参数的影响，引入分子模拟技术，建立渣铁物质在炉衬材料中的扩散模型以及渣铁与炉衬材料热物理性能参数预报模型；针对传统方法将炉衬热面温度设为固定值的局限性，采用渣铁-炉衬流固多物理场耦合数值模拟技术确定炉衬热面温度，采用两步修正法求解传热逆解问题，获得炉衬热面形状；针对三维数值模拟耗时长的问题，研究从数值模拟结果中提取和构建热面形状判别规则的知识发现方法，建立基于知识库的内衬状态在线监测与优化方法。项目研究成果将为提高炉缸内衬侵蚀状况在线监测的可靠性、及时发现并抑制炉缸内衬的异常侵蚀，进而延长炉缸寿命提供重要的技术支撑。

在役炼铁高炉的炉缸工作环境恶劣，内衬材料易被侵蚀破坏，是制约现代高炉寿命的主要因素之一。为避免炼铁高炉炉缸内衬侵蚀穿孔而发生铁水渗漏的严重事故、提高高炉寿命，本项目对高炉炉缸内衬侵蚀状况的在线监测方法开展研究。以柳钢2#高炉为研究对象，对数百个点热电偶在线采集的数据、布料参数、焦炭与铁矿石组成等现场参数进行了收集与整理。基于Lammps软件与EAM势函数，计算出了不同温度下铁水的密度、比热、粘度、热导率、自扩散系数、热扩散系数等物性参数，拟合出了铁水热物理性能参数随温度变化的预报模型；搭建了炉缸内熔体流动传热的试验装置，通过对不同熔体浮力模型计算结果与实验数据的比较，明确了适用于炉缸熔体流动传热计算的浮力模型；采用渣铁-炉衬流固多物理场耦合数值仿真技术确定炉衬热面温度，采用传热逆解问题的两步修正法计算炉衬热面形状；通过大量计算，获得了不同炉缸工作状况下的炉衬热面形状，建立了相应的数据库，并用支持向量机的方法，对热面形状的在线监测进行了有益的尝试；同时针对侧壁内衬损伤检测的传统方法存在的问题，提出了一种基于角热阻分段计算的损伤在线诊断方法。项目研究成果为提高炉缸内衬侵蚀状况在线监测的可靠性、延长炉缸寿命提供了重要的技术支撑，且为高炉的信息化提供了一定的技术基础。