数智化正向设计方法及其在制造装备与过程中的应用

Digital Intelligent Forward Design Method and Its Application in Manufacturing Equipment and Process

针对制造装备与过程在精度和能效等方面的共性设计难题,提出具有泛化能力的数智化正向设计方法,从设计需求出发,根据制造装备的多模态信息对能耗、材料消耗和制造可靠性进行计算,以增强自感应、自适应、自学习和自决策能力,满足正向设计需求。将制造装备分解为控制系统、传动系统和辅助系统,并建立相应的能耗数学模型。以增材制造装备协同优化为例,从热力学能量转化角度,阐释增材制造装备的能耗组成,引入以热对流和热辐射为边界的非线性瞬态温度场模拟方法,基于模拟方法对增材制造过程所消耗的能量和材料进行精准推算。建立制造装备可靠性模型,对制造过程中的失效事件进行概率统计分析。采用平均无故障间隔时间对可靠性进行描述,建立综合工况下平均无故障间隔时间的基本概率函数。以汽车燃油颈管为例,使用高算力平台,基于非均匀有理B样条曲线的几何建模内核,对三维非网格的概念设计原型进行精准表达,实现数字孪生驱动的虚拟制造与打印。使用激光增材制造装备进行物理实验,对制造过程中消耗的能量和材料进行测量,能量消耗预测的平均绝对误差为11822.62 J,平均绝对百分比误差为0.0834,均方根误差为16845.69 J。材料消耗预测的平均绝对误差为0.0030 g,平均绝对百分比误差为0.0713,均方根误差为0.0041 g。实验表明,数智化正向设计方法在面向高效高精低碳的制造装备多工况服役中具有重要应用价值。

Aiming at the common design problems in precision and energy efficiency of manufacturing equipment and process,a digital intelligent forward design method with generalization ability is proposed.Based on the design requirements,the energy consumption,materials consumption and reliability are calculated according to the multi-modal information of manufacturing equipment.The design method can enhance the ability of self-sensing,self-adaptation,self-learning and self-decision to satisfy forward design requirement.The manufacturing equipment is divided into control system,drive system and auxiliary system.The corresponding energy consumption model is established.Taking collaborative optimization of additive manufacturing equipment as an example,the energy consumption model is analyzed in detail from the perspective of thermodynamic energy conversion.A nonlinear transient temperature field simulation method with thermal convection and thermal radiation as the boundary is established.The energy and materials consumed in the additive manufacturing process are accurately predicted based on the simulation method.The reliability model of manufacturing equipment is established.The model is used to analyze the data of failure events in manufacturing process based on probability statistics.The mean time between failures is used to describe the reliability of manufacturing equipment.The basic probability function of mean time between failures(MTBF)is established under comprehensive conditions.Based on the geometric modeling kernel of non-uniform rational B-spline curve,the 3D non-grid conceptual design prototype of automobile fuel neck pipe is accurately expressed using high computing capability.Virtual manufacturing driven by digital twins is realized.Physically manufacturing is carried out using laser additive manufacturing equipment.The energy and materials consumed during the manufacturing process are measured.The mean absolute error of energy consumption prediction is 11822.62 J.The mean absolute percentage error is 0.0834.The root mean square error is 16845.69 J.The mean absolute error of materials consumption prediction is 0.0030 g.The mean absolute percentage error is 0.0713.The root mean square error is 0.0041 g.The experiment results show that the digital intelligent forward design method has important application value for manufacturing equipment in multiple working conditions aimed at high efficiency,high precision and low carbon.