A Nondestructive Testing Method Of Mechanical Elastic Parameters And Application In Personalized Fault Diagnosis

Beams, shafts and shells are essential parts of mechanical equipments. When the machine processing in High-speed, high efficiency, high-precision and the mechanical equipments worked in overload, fatigue, corrosion, high temperature and other complex severe conditions, the components of mechanical systems (such as drive shafts,beams, shells and so on) will produce varying degrees of damage inevitably. This paper develops a method to detect mechanical parameters of structures non-destructively and a method for characteristic parameters detection of joint surface based on impulse excitation technology, and builds a relatively accurate numerical analysis model of structural damage of the drive system, presents the methods of mechanical fault personalized diagnosis based on intelligent classification algorithms (such as support vector machine, neural networks and so on).Numerical simulation has become a developed design and manufacturing analysis technique. However, the result of simulations is not reliable because the precise mechanical parameters are not known.The dynamic Young's modulus and shear modulus are the key parameters for numerical simulation. It is meaningful that a method can be developed to detect the reliable mechanical parameters for simulations. A method combined finite element method (FEM) and response surface method (RSM) developed to detect the mechanical parameters of I-beams, hollow cylinder, stepped shafts and layer composites based on impulse excitation technique (IET) in this paper and the experimental formulas used to calculate the parameters also obtained. The accuracy of this method can be promised through the experiment investigation.Model updating method is essential for finite element model setup.The vibration signal of model is obtained and compared with the vibration information of experimental investigation. The consistency between simulation signal and experimental signal is determined using the accuracy evaluation parameters (such as mean absolute percent error and Theil's inequality coefficient). Modify some parameters (mechanical parameters, stiffness coefficient and damping coefficient) repeatedly until the finite element models meet the requirements.For the fault diagnosis of drive systems, Non-destructive testing of mechanical parameters provided reliable values parameter of structures and the numerical simulation model of the shafts are constructed using finite element method (FEM), the vibration signal of the faulty shafts obtained and wavelet packet transform (WPT) is employed to decompose the vibration signal into several signal components. Specific time-domain feature parameters of all the signal components are calculated to generate the training samples. Finally, the measured vibration signal and its components decomposed by WPT are served as a test sample to the trained SVM and the fault type of faulty shafts are finally determined.In this paper, the method combined numerical simulation and experimental investigation is used to detect the dynamic mechanical parameters of the mechanical structures, the character analysis of the mechanical joint surface and the prediction and diagnosis of the shafts faults. The results of simulations and experiments show the method is reliable and can be applied in the engineer.