非完全试验信息的悬架橡胶-钢衬套本构模型多参数识别

Multi-parameters Identification of Constitutive Model of Suspension Rubber-steel Bearing Based on Incomplete Test Information

以轿车扭力梁后悬架的橡胶-钢衬套为对象,通过应变能密度函数确定Yeoh多项式为衬套的本构模型。进行结构试验并获取X,Y,Z轴向"位移-力"历程曲线的非完全试验信息,利用HyperStudy与ABAQUS进行初始试验仿真,分析试验与仿真曲线的一致性;设定Yeoh多项式的多参数C10,C20,C30、Di为优化变量并赋初始值,优化目标是"面积差"或"位移差"平方为零的6类函数。采用自适应响应面法,经过14次迭代,6类函数的总加权值收敛逼近于零,识别出一组能较准确模拟衬套力学行为的本构模型参数值。试验仿真结果发现Y向曲线拟合精度高,预紧模拟及网格歧异等导致X与Z向曲线拟合差,但瞬态刚度值误差均在8%内,在工程中属于可接受范围。

In view of bearings＇ greater impact on suspension and vehicle performances,selecting a rubber-steel bearing of one car＇ s torsion beam rear suspension as a research object,Yeoh polynomial was determined as the bearing＇s constitutive model by deriving from a strain energy density function. The bearing＇s structure test was done to get some incomplete information of ‘displacement-force＇history test curves in X /Y /Z axial,the initial test was simulated to get ‘displacement-force ＇ history simulation curves in X /Y /Z axial by using HyperStudy and ABAQUS,the simulation curves were compared with the test curves to evaluate consistency between them. The multi-parameters C10,C20,C30、Diof Yeoh polynomial were set as optimization variables and their initial values were given,6 types of optimization objective functions were designated to ‘square of area difference＇or ‘square of displacement difference＇equaling to zero,14 interactive processes were performed by using the adaptive response surface（ ARS） method,then the total weight value of 6 types of optimization objective functions converged and approached to zero,and a set of values of the multi-parameters,which can more accurately simulate the bearing＇s mechanical behavior,were identified. The bearing＇ s test was simulated again; the results show that the fitting precision is high between test and simulation curves in Y axial. Because of pre-tightening simulation and mesh divergence etc.,the fitting precision is relative low between them in X and Z axial,but both of the error of transient stiffness values is all less than 8%. If the average error is considered,this is can be accepted in engineering.