Finite Element Model Updating Of Gearbox Based On Modal Correlation

As the development of social production and the improvement of science and technology, more mechanical products has been put into production and people’s life. In order to meet the production requirements and ensure to be used safely, mechanical products need not only to be designed strictly, but also to be tested and analyzed comprehensively and rigorously. A kind of testing method called the finite element method have the advantage of low cost and high efficiency,the finite element model of the structure is simulated and analyzed by this method, but whether the results have significance depends on the accuracy of finite element models,that means the information of the finite model and the actual structure are all the same, otherwise the results of finite element calculation will lose its practical significance.Therefore,it is important to find a method for establishing a precise finite element model.The modal properties reflect the dynamic characteristics of the structure itself,it is the inherent response of vibration.Modal parameters can be obtained by calculation or testing,for the same structure, the results of the modal should be the same.But generally the results of experimental modal analysis is reliable, and the results of calculation modal analysis may be unreliable for the error of finite element models. This puts forward an idea about the finite element model updating:make the experimental modal data as the standard, correcting the finite element model to make the calculation and experimental modal data consistent.The modified finite element model can be considered more accurately to reflect the dynamic characteristics of the actual structure.Consequently, the method of correcting the finite element model of gear box based on the modal correlation is proposed. The specific approach is: first,the model of gearbox established by Pro/E software using it’s design drawing, import it into the ANSYS software and get the modal parameters by calculation;Then the experiment platform for the modal test was setted to get experimental modal parameters of gear box;Finally using the experimental modal parameters as standard, modify finite element model to make the parameters of calculative modal and experimental modal consistent. Genetic algorithm was used for finding the minimum value of the objective function of the relative error between calculative and experimental modal.The mathematical model corresponding the relationship between the calculative modal and gear box structure parameters was essential.In order to establish mathematical model to quickly and exactly representing relationship of calculative modal and structure parameters,a innovative gray neuron model was builded.The model have two inputs,which are quadratic polynomial fitting of data in a least-squares criterion, vector neurons integrate the two input to the output by using custom summation operation.9groups of original data were used to fit the corresponding relation between the gear box calculative modal and structure parameters, fitting and forecasting accuracy were under1%, meet the need of high precision calculation. Through the modification, the prior four order modal parameters of errors between calculation and experiment are below0.7%,while relative error between3.17%-9.19%before corrected,proved the correction method success.The research object is gear box, although the structure is not complicated,the box is the supporting structure of the system,which design needs more stringent standards, the characteristics of structure and research method have very strong representative, therefore the research method and conclusion have some practical valve for the model updating of the other structure.