摘要
薄层单元可有效地表征螺栓连接结合部,但要进行动力学分析及优化必须辨识出薄层单元参数。为此,在建立薄层单元参数与结合部动态特性关系的基础上,提出一种准确、高效辨识薄层单元参数的方法。通过分析薄层单元的刚度矩阵及本构方程,确定待辨识参数;采用拉丁超立方抽样方法选取参数样本,结合试验与有限元计算的模态参数建立目标函数,用响应面模型表征薄层单元参数与目标函数的关系,并评价模型精度;采用遗传算法在响应面模型基础上实现参数优化辨识。分别辨识均一材料和考虑结合面应力分布差异化材料的薄层单元参数,参数优化辨识后,试件的试验与有限元计算固有频率误差均不超过4%,证明了该辨识方法的有效性。
Bolted joints could be characterized by thin-layer element effectively,and it’s necessary to identify thin-layer element parameters for dynamic analysis and optimization of bolted joints. An accurate and efficient method is proposed to identify parameters based on thin-layer element method. Through analyzing the stiffness matrix and constitutive equation of thin-layer element,parameters to be identified are determined. Latin hypercube sampling method is adopted to select samples of parameters. Then,the relationship between thin-layer element parameters and objective function is characterized by response surface model. Finally,genetic algorithm is used to optimize and identify parameters based on the response surface model. Parameters of homogeneous material element and differential material element are identified respectively. The errors of natural frequencies are no more than 4% which demonstrated the validity of the identification method.
作者
孙志勇
孙伟
孙清超
黄伟强
SUN Zhi-yong;SUN Wei;SUN Qing-chao;HUANG Wei-qiang(School of Mechanical Engineering,Dalian University of Technology,Liaoning Dalian 116024,China)
出处
《机械设计与制造》
北大核心
2019年第1期50-54,共5页
Machinery Design & Manufacture
基金
辽宁省科技攻关计划(201301002-01)
关键词
结合部
薄层单元
参数辨识
响应面法
遗传算法
Bolted Joints
Thin-Layer Element
Identification for Parameters
Response Surface
Genetic Algorithm
