摘要
龙门混凝土浇灌机作为一种新型的混凝土浇灌设备,其设计多依赖于经验,设计的机身结构有较大的刚度、强度余量。为了充分发挥材料的承载性能,减少结构自重,提出了一种以龙门混凝土浇灌机各主要构件板厚为设计变量的优化方法。优化方法基于龙门混凝土浇灌机有限元模型,采用多学科优化软件ISIGHT中的拉丁超立方试验法对有限元模型进行灵敏度分析,筛选出对目标函数和约束函数影响较大的设计变量。利用筛选出的设计变量构建目标函数的响应面近似模型,最后采用多岛遗传算法对目标函数寻优。优化结果表明,在保证结构性能的前提下,龙门混凝土浇灌机结构质量减少了3979kg,与优化前质量相比降低了16.7%,优化效果显著,对龙门混凝土浇灌机的结构设计及改进有重要的实际意义。
As a new type of concrete pouring equipment,the design of the gantry concrete pouring machine relies on experience,the designed fuselage structure has large stiffness and strength allowance.In order to give full advantage to the bearing capacity of materials and reduce the structural weight,an optimization method for the design variables of plate thickness of the main components of the gantry concrete pouring machine is proposed.Based on the finite element model of the gantry concrete pouring machine,the sensitivity of the parameterized model was analyzed by the Latin hypercube test method in the multidisciplinary optimization software ISIGHT,and the design variables which had great influence on the objective function and constraint function were selected.The response surface approximation model of the objective function is constructed by using the selected design variables,multi-island genetic algorithm is used to optimize the objective function.The optimization results show that on the premise of guaranteeing the structural performance,the structure mass of the gantry concrete pouring machine decreased by 3 979 kg,and compared with the mass before optimization,the self-weight decreased by 16.7 %,and the optimization effect was remarkable,which is of great practical significance for the structural design and improvement of the gantry concrete pouring machine.
作者
韩露男
于兰峰
潘益明
Han Lunan;Yu Lanfeng;Pan Yimin(School of Mechanical Engineering,Southwest Jiaotong University,Chengdu 610031,China)
出处
《现代制造工程》
CSCD
北大核心
2019年第5期47-51,82,共6页
Modern Manufacturing Engineering
基金
国家自然科学基金资助项目(51675450)
中央高校基本科研业务费专项资金资助项目(2682016CX031)
关键词
龙门混凝土浇灌机
拉丁超立方试验法
优化设计
响应面近似模型
gantry concrete pouring machine
Latin hypercube test
optimization design
response surface approximation model