基于响应面方法的叶栅摆动装置有限元模型修正

摘要

为提高叶栅摆动装置模态分析与模态试验的相关性，使用响应面方法对有限元模型多个参数进行优化，实现模型修正。首先使用有限元模型进行模态分析，提取整体模态；其次，采用移动传感器方法对大型装配体进行整体模态试验，通过模态判据准则检验模态试验结果、模态参与因子确定主要整体模态；再次，基于有限元模型误差分析，确定对主要部件分别采用不同材料修正参数，通过中心复合试验设计确定样本空间，使用多目标响应面方法对样本进行回归分析，在回归分析的响应面内对待修正参数进行非线性约束优化，得到最优解；最后，使用修正参数重新进行模态分析实现模型确认，并进行动力学计算，与实际测试结果对比。结果表明，修正后的模态分析与模态试验结果相关性提高，前三阶整体模态频率误差均值由9．71％减小至0．73％，振型相关性由0．74提升至0．89。%In order to improve the correlation between the results of modal analysis and modal test for a blade swing mechanism,a multiple parameters optimization was brought into updating the finite element model by using the response surface method.The overall modal parameters of the structure were extracted by the numerical modal analysis.And for this large assembly,a modal test was also developed by moving the sensor other than the hammer.The assurance criterion and the mode participation factor were made in use respectively to validate the test results and to confirm the main modes. Based on the finite element model error analysis,the model updating was carried out,using different material modification parameter at different part of the structure.The sample space was calculated and determined by the central composite test design,and the regression analysis on the samples was achieved by use of the multiple objectives response surface method,where the model updating targets were optimized.Finally,the modified parameters were applied in a new modal analysis to validate the accuracy of the updated model.On this basis,the dynamic simulation results were compared with the actual test results.The results show that the correlation between the results of modal test and updated modal analysis is improved,the average frequency error of the first three modes is reduced from 9.71% to 0.73%,and the modal shape MAC is improved from 0.74 to 0.89.