推进技术 ›› 2020, Vol. 41 ›› Issue (5): 1121-1129.DOI: 10.13675/j.cnki.tjjs.190423

• 燃烧 传热 传质 • 上一篇    下一篇

基于变分法的冷却结构优化设计与分析方法

李轩1,2,陆阳1,吴坤1,范学军1,2   

  1. 1.中国科学院力学研究所 高温气体动力学国家重点实验室,北京 100190;2.中国科学院大学 工程科学学院,北京 100049
  • 发布日期:2021-08-15
  • 作者简介:李 轩,博士生,研究领域为冷却结构优化设计。E-mail:lixuan@imech.ac.cn

Optimization Design and Analysis of Cooling Structure Based on Calculus of Variations

  1. 1.State Key Laboratory of High Temperature Gas Dynamics,Institute of Mechanics, Chinese Academy of Sciences,Beijing 100190,China;2.School of Engineering Science,University of Chinese Academy of Sciences,Beijing 100049,China
  • Published:2021-08-15

摘要: 为了改善冷却结构的综合换热性能,需要对冷却结构的优化设计进行研究。以冷却结构的平均温度、温度不均匀度和冷却剂流动压力损失为目标函数,基于变分法生成优化分布的冷却通道,该通道可以根据当地的边界条件而自适应地调整。随后对冷却结构进行流固耦合传热的数值计算,得到冷却结构的温度分布、冷却剂的压力损失、冷却剂出口温度等参数,多次迭代进而得到满足目标要求的冷却结构优化设计。计算结果表明,基于变分法的冷却结构优化设计方法可以根据当地边界条件生成优化的冷却通道;对于不同的优化方案,存在各自对应的最佳通道个数和冷却通道分布使目标函数最优;对于优化方案三,优化设计的冷却通道和常规设计的冷却通道相比:冷却结构平均温度从816K下降到807K,温度不均匀度从211K降低到172K,代价是压力损失从8kPa上升到17kPa。

关键词: 变分;冷却;优化;自适应;多目标

Abstract: It is essential to study the optimal design of the cooling system for improving the comprehensive heat transfer performance of the cooling structure. The average temperature, temperature unevenness of the cooling structure and coolant flow pressure loss were chosen as objective functions. First step, the optimized cooling channel distribution was generated based on the calculus of variations, which can be adjusted adaptively according to local boundary conditions. Second step, the fluid-solid coupling heat transfer numerical calculation was performed to obtain the temperature distribution of the cooling structure, the coolant flow pressure loss and the coolant outlet temperature, etc. The optimization design of cooling structure which meets the target requirements can be obtained by multiple iterations. The calculation results show that the optimization of cooling structure based on the calculus of variations can generate cooling channels according to the local boundary conditions. For different optimization cases, there are corresponding optimization of cooling channel number and cooling channel distribution for each case to optimize the objective function. For the optimization case 3, compared with the regular cooling channel, the average temperature of the optimized cooling channel decreases from 816K to 807K and the temperature unevenness decreases from 211K to 172K, with the cost that the pressure loss rises from 8kPa to 17kPa.

Key words: Calculus of variation;Cooling;Optimization;Adaptive;Multi-objective