Finite Element Model Updating Method Based On Frequency Response Function

The biggest problem of finite element model updating method in engineering application is that the calculation efficiency is too low,because the finite element model needs to be repeatedly called to carry out a large number of complex data calculation in the process of model updating.Especially for large mechanical equipment which structure is complex and involves various parameters.The finite element model is more complex and large,so the method of frequency response function and surrogate model is considered.Firstly,the most widely used model updating method based on modal parameters can not ensure sufficient accuracy in the process of finite element model updating which by the measured modal frequencies and modes of the previous order.There is the problem of mode analysis and mode mismatch.Therefore,introducing the frequency response function into the model updating not only avoid the problem of deviation caused by the mismatch between modal analysis and mode,but also provide more structural vibration information and more available data information.So the model updating process has high accuracy.Secondly,in order to improve the accuracy and efficiency of the response surface surrogate model updating.The modal participation criterion and the effective independent method are selected to the excitation point and the measuring point position.Then select the sample points by Latin supercubic sampling method,and solve the corresponding frequency response function by finite element software.Then the surface surrogate model is improved by the ratio weighting method.The weight is assigned according to the distance from the sample point to the objective function according to the ratio weighting method.Then,in order to avoid the coarse error and improve the precision of the response surface surrogate model,the outlier is eliminated according to the triple variance criterion.Then the updated value of the constructed objective function is solved by the cuckoo intelligent optimization algorithm.The method is verified by a truss structure and gearbox box model.The results show that the improved method not only improves the computational efficiency while ensuring the accuracy,but also fits the model's imaginary and real part curves more closely with the experimental model.Finally,in order to avoid the safety accident caused by crack damage,health monitoring is carried out for the gear box,which has been working under high speed and heavy load and serious pollution.Therefore,the proposed improved response surface model updating method combined with the updated gearbox box model is used to study the damage identification of the cracked box.The results show that the the accuracy is lower when the crack is smaller.However,considering that the gear box structure is not symmetrical and uniform,and the working environment is complex.But most of the errors after recognition are below 5%,it shows that this method has a good application prospect.