面向智能制造的复杂多向模锻过程空间多轴联动协调控制

The multi-directional die forging technology has obvious advantages in the integrated high performance manufacturing of complex hetero-shaped hollow load-bearing components. In the manufacturing process of multi-directional die forging, the defects of unreasonable forging flow lines distribution and uneven deformation in mold cavity caused by poor multi-axial coordination, affect the mechanical properties of forgings drastically. In addition, poor quality stability, weak correlation between process mechanism and system control and low precision of multi-axial position control are also difficult to be addressed in multi-directional die forging process..The nuclear power joint components are regarded as research objects, and the correlation of multi-directional die forging technology, metal microstructural evolution, forging flow lines distribution and mechanical properties of forgings were investigated by combining the methods of numerical simulation and process experiments. The feasible domains of forging process parameters were established in order to provide process control windows in the intelligent operation of multi-directional die forging process. The model of estimation of mechanical properties of forgings were developed based on multi-source signal acquisition of forging process and rheological state prediction. Furthermore, the approaches of self-learning and model evolution of forging process parameters were proposed. The methods of prediction of macroscopic deformation resistance of multi-directional die forging process and deflection torque of moving beam were developed using the method of integrated mathematical model and data compensation. Also, the multi-axial high precision instantaneous displacement coordinated control strategy were proposed based on hierarchical modeling and multiple controller switching. This project will provide theoretical basis for improving the intellectualization degree of multi-directional die forging process and realizing the high quality manufacturing of forgings.

多向模锻成形技术在复杂异形中空类承力构件的整体性化高性能制造中具有明显优势。多向模锻加工过程中多轴运动不协调将导致锻件流线分布不合理、变形不均匀，直接影响锻件的力学性能。复杂工况下加工质量稳定性差、工艺机理与系统控制的关联性不强、以及全锻造过程多轴位移控制精度较低也是多向模锻制造过程控制面临的难题。.本项目以核电接头件为结合对象，采用数值模拟和工艺试验相结合的方法，揭示多向模锻工艺-组织演变-流线-构件性能之间的关联规律，建立多向锻造工艺参数可行域，为多向模锻过程智能化运行提供工艺控制窗口。建立基于多源信息感知和流变状态预测的锻件性能预报模型，提出工艺参数自学习与进化方法。建立集成数理模型与数据补偿方法的多向模锻加工过程宏观变形抗力和偏转力矩预测方法，研究基于分区间建模和多控制器切换的多轴瞬时位移协调控制策略。本课题的研究将为提高多向模锻过程智能化程度，实现锻件形性协同高品质制造奠定基础。

项目以T型三通多向模锻成形过程为研究对象，针对多向流变成形关联规律、工艺可行域构建、工艺知识进化机理、多向成形载荷预测、高品质多向协同控制策略等方面开展研究。. 项目的主要研究内容包括：. 1）研究了不同变形温度、应变速率、变形量对316LN核级不锈钢高温流变行为的影响规律，建立了材料的高温流变应力本构模型。. 2）揭示了变形温度、应变速率、变形量对316LN微观组织演变的影响规律，建立了316LN动态再结晶动力学模型和晶粒尺寸模型，获得了应变速率与成形温度匹配的最优工艺参数范围。. 3）研究了不同工艺参数对三通件多向模锻成形典型工艺过程型腔充填、流线演变、温度场分布的影响规律，确定了多向成形工艺参数可行域；建立了不同工艺参数下多向成形锻件的映射模型多向模锻工艺-流变特性-成形性能映射模型。. 4）提出了一种多向多级渐进加载的模锻成形工艺，揭示了工艺知识进化机理，研究了多级渐进加载三通成形件变形规律，有效抑制了坯料金属回流造成的折叠、空穴等缺陷。. 5）研制了基于NoSQL的多向模锻成形材料和零件工艺知识数据库，为智能成形制造过程工艺数据高通量流转、工艺知识迭代、模锻装备驱动参数调控提供了载体。. 6）推导了T型三通多向模锻全过程空间变形抗力的计算表达式，建立了多向锻造全过程成形载荷预测模型和液压驱动力流传递非线性模型，验证了所建立模型的准确性。. 7）设计了一种基于干扰补偿的多向成形液压驱动速度跟踪控制器，减少了非线性环节对系统跟踪性能的影响；针对多向模锻成形多轴联动高效高精度的制造需求，提出了一种基于滑模控制和变形力分区间实时预测的多向模锻过程多轴联动控制策略，保证了驱动轴之间的联动控制精度。