基于生物分枝热/流路径的机电装备散热冷却结构设计方法研究

Cooling problem has become a major obstacle to pursue the high-speed, high-precision and high-reliability performances for high-end NC machine tools, and also is the main obstacle for wide application of high-power electronic devices. The heat and mass transfer performances as well as cooling characteristics of the exquisite natural branching structures, which are the survivors after natural evolution based on survival of the fittest, are always optimal. Thus natural branching structures provide new idea for the innovative design of cooling structures of mechanical and electrical equipment, such as machine tool spindles and electronic devices. Bionic design criteria deriving from the growth processes of biological branching structures, which tend to realize optimization, will be explored in this project. The forming, transferring and coupling behaviors of heat and fluid flow inside the cooling structures will be analyzed. The influence laws of the distribution patterns of heat/fluid transfer paths on cooling performances will be investigated. The governing function for the selection of the optimal heat/fluid transfer paths will be built. Then the optimal designs of cooling structures are converted into the selection of the optimal heat/fluid transfer paths. And then the corresponding bionic optimal design method for cooling structures controlling by heat/fluid transfer paths will be developed, which may support the theory for the innovative design of the cooling system of mechanical and electrical equipment, such as machine tool spindles and electronic devices.

散热冷却问题已成为制约高档数控机床追求高速、高精度和高可靠性及大功率电子器件大规模应用的主要障碍。经优胜劣汰自然演变和进化胜出的精巧自然分枝结构往往具有非常优异的传热传质、散热冷却性能，为机床主轴、电子器件等机电装备的散热冷却结构的创新设计提供了新思路。本项目拟探索蕴藏于自然生物分枝结构趋优生长过程的仿生设计准则，研究热量和工质流在散热冷却结构内部的形成、传递和耦合行为，剖析热/流传递路径分布特性影响结构散热冷却性能的规律；构建最优热/流传递路径选择控制函数，将散热冷却结构的优化设计转化为最优热/流传递路径的选择问题，研究基于热/流传递路径驱动的散热冷却结构仿生优化设计方法，以期为创新设计机床主轴、电子器件等机电装备的高效散热冷却系统提供新的方法。

散热冷却问题已成为制约高档数控机床追求高速、高精度和高可靠性及大功率电子器件大规模应用的主要障碍。经优胜劣汰自然演变和进化胜出的精巧自然分枝结构往往具有非常优异的传热传质、散热冷却性能，为机床主轴、电子器件等机电装备的散热冷却结构的创新设计提供了新思路。本项目研究了生物分枝型散热通道构型规则，分析了生长和分歧规则，研究了热流载荷路径及其可视化描述，分析了载荷路径疏密程度等分布特性的影响规律；为了开发散热通道仿生生成式设计方法，研究了基于无网格伽辽金（EFG）法进行散热分析的数值实现，基于生物分枝生长和构型规则，构建了热流路径驱动的散热通道脉序生长算法和仿生分级生长设计方法；分析了传统SIMP拓扑优化方法的求解流程，探究了设计过程中的耗时规律，通过引入人工智能深度学习算法，开发了基于深度自编码网络的导热拓扑结构加速设计方法和基于条件变分自编码网络的导热拓扑结构智能预测方法。本项目的研究成果可为创新设计飞机雷达T/R组件、机床主轴、电子器件等机电装备的高效散热冷却系统提供新的方法。