Study On The Identification Of Structural Inertial Parameters Based On Modal Testing

Rigid body inertial parameters,including mass,center of mass,moment of inertia and product of inertia,are essential for structural dynamic simulation,performance optimization,finite element analysis and suspension design of engines.It is of great engineering significance to study on an accurate and simple measurement method for rigid body inertia parameters.The main contents of this thesis are as the following.1.Based on the finite element model simulation,three commonly used methods for structural inertial parameter identification,i.e.Direct Physical Parameter Identification Method(DPPIM),Inertia Restraint Method(IRM)or named Mass Line Method(ML),and Modal Model Method(MM),are carefully compared.The results show that the three methods have respective advantages and disadvantages in identification accuracy and realization.A simulation analysis is carried out on a flat plate structure.The result shows that for each method the identification is influenced by structural support stiffness,noise,the coordinate error of measuring point,modal parameter identification results.And sometimes some parameters like mass may be asked to be known in advance to identify the other parameters.In addition,the number of measurement points and the test workload can be great reduced and the test process can be greatly accelerated.2.An improved identification method combining the DPPIM and IRM methods are presented to utilize their advantages.A practical experiment on the bogie structure of a high speed train is carried out to verify the presented method.The result shows that the presented method is still reliable even in an environment of high noise and high support stiffness.3.Based on the identification of inertial parameters,a calibration method for the impact hammer is proposed.Better result than the traditional calibration method based on the acceleration testing is obtained4.Based on MATLAB GUI technology,an inertia parameter simulation and identification toolbox is developed to provide friendly human-machine interface for research and engineering application.