个性化硬脆材料零件高效精密磨削的基础问题研究

Precision grinding is the most popular machining method of parts made of hard-and-brittle materials such as crystal, glass and ceramic. However, due to the variation of material properties and the complexity of grinding process, there is no mature method for predicting the grinding-induced damage and optimizing the grinding parameters. In the future, diverse hard-and-brittle materials are expected to be used in the disciplines of optics, electronics, mechanics and so on, thus, a novel method which can be employed to make process strategy and optimize the processing parameters efficiently. In this research proposal, aiming to resolve the difficulty of high-efficient grinding process development for part made of various hard-and-brittle materials, it is proposed that combining the conventional modeling method and the artificial intelligence algorithm allows for attaining the advantages of these two methods, the clear interactions between variables and the universities. To realize this aim, firstly, material removal mechanism will be investigated for revealing the universal features in removal process. Then, a comprehensive model will be created followed by sensitivity analysis and variable interaction analysis for contraction of variables. Finally, artificial intelligence algorithm will be used to quantitatively predict the grinding damage, and based on which the grinding process parameters will be optimized. Additionally, to meet the requirements of intelligent manufacturing in the future, reverse-solution method for obtaining the value of some difficult-to-measure parameters and quick assessment method of surface quality will be proposed for less dependence on the expensive measurement instruments and additional manual operations. The work in this research proposal will provide a theoretical foundation for the high-efficient grinding of parts made of diverse hard-and-brittle materials.

精密磨削是晶体、玻璃和陶瓷等硬脆材料零件的主要加工方式，但由于材料特性差异和磨削过程的复杂性，目前仍无磨削损伤预测和磨削工艺优化的成熟方法。未来光学、电子学和机械等领域对硬脆材料的需求趋于多样化，需要一种能够快速进行新材料磨削工艺制定和优化的方法。本项目拟通过传统建模方法和人工智能算法的融合，发挥传统模型变量作用关系明确和人工智能算法普适性好的优势，解决个性化材料零件高效磨削工艺开发难题。为此，通过材料去除机理研究，掌握硬脆材料去除的一般规律，通过建立磨削过程全系统模型，进行敏感性分析和变量交互作用分析，实现变量缩并，最终通过人工智能算法实现磨削损伤量化预测，并在此基础上进行工艺优化。此外，为满足未来智能化制造需求，减少工艺开发过程中的高端测量设备和人力介入需求，开展参数反演诊断方法和表面质量评估快速方法研究。以上研究将为个性化硬脆材料零件的高效磨削加工奠定理论基础。

精密磨削是晶体、玻璃和陶瓷等硬脆材料零件的主要加工方式，但由于材料特性差异和磨削过程的复杂性，目前仍无磨削损伤预测和磨削工艺优化的成熟方法，大量重复研究和方法缺乏普适性是磨削技术研究的突出问题。传统依靠经验数据进行工艺优化成本高、速度慢，以及现有磨削工艺物理模型仍旧难以实现磨削过程定量预测，机器学习方法变量多及测量难度大，无法有效指导磨削组合工艺优化。未来光学、电子学和机械等领域对硬脆材料的需求趋于多样化，需要一种能够快速进行新材料磨削工艺制定和优化的普适方法。本项目主要针对这一需求，通过对硬脆材料去除机理、随机损伤测试方法、磨削过程变量分析、数据物理融合建模方法和磨削工艺优化方法等研究，完善了硬脆材料去除机理研究，提出了新的损伤测试方法和预测模型，明确了硬脆材料磨削表面质量的主要影响因素及其敏度，为解决磨削质量稳定性问题奠定了基础。此外，有效整合物理模型和机器学习方法，发展出一套能够指导硬脆材料磨削零件高效磨削的理论方法和测试技术，可满足未来科技发展对硬脆材料磨削工艺提出的新要求。.本项目研究发表 SCI/EI 收录论文 17 篇，其中 SCI 收录论文 15 篇，授权国家发明专利6项，美国专利1项，完成科技成果鉴定1项。培养博士研究生1名，培养硕士研究生 5名，1人学位论文获评辽宁省优秀硕士论文，在读博士生2名，硕士2名。联合英国和美国等该磨削加工领域国际知名学者，合作完成磨削表面质量预测方法的综述论文，发表一年多来他引30次，受到领域同行关注。