Study On Modeling And Uncertainty Optimization Methods Of Dynamic Characteristics Of Automobile Drive Shaft Systems

The dynamic characteristics of automobile drive shaft systems mainly include the generated axial force(GAF)and the plunging force(PF),which are caused by the contact and friction characteristics of contact pairs inside drive shaft systems.Excessive GAF and PF will cause vibration and noise of the drive shaft system and the vehicle.At present,the researches on the dynamic characteristics of drive shaft systems focus on the deterministic research.The main problems in the researches are as follows: the computational accuracy of the analytical models of the GAF is low and the analytical model of the PF is lacking;the influence of rough surface characteristics(RSC)of contact pairs is not considered;and the influence of the uncertainty is not considered.These problems make it difficult to predict and optimize the dynamic characteristics effectively,and to study the influencing factors of the dynamic characteristics deeply and comprehensively.Aiming at these problems,the research on modeling,analysis and optimization methods of dynamic characteristics of drive shaft systems is carried out in this paper.The main research contents are as follows:(1)Based on the kinematic analysis of the drive shaft system and its internal rollers,deterministic analytical models of the GAF and PF of drive shaft systems are proposed.The validity of the proposed deterministic analytical models is verified by the measurement of the GAF and PF of a drive shaft system.It is shown that the proposed deterministic analytical models of the GAF and PF have high computational accuracy,which can provide an effective basis for the establishment of subsequent complex models and uncertainty models.(2)Based on fractal theory and the deterministic analytical models proposed above,deterministic models of the GAF and PF considering RSC are proposed.In the modeling process,a method for correcting the distribution of asperities between rollers and tracks inside the tripod joint is proposed to describe the micro contact and friction states more accurately.The rough surface profiles of rollers and tracks are measured to identify the fractal parameters,as well the validity of the deterministic models of the GAF and PF considering RSC is verified by the measurement of the dynamic characteristics.The proposed deterministic models considering RSC reveal the relationships between the dynamic characteristics and the fractal parameters,the material parameters and the correction coefficient,which can provide an effective basis for the establishment of subsequent uncertainty models.(3)Based on orthogonal polynomials and the deterministic models proposed above,interval uncertainty(IU)models of the dynamic characteristics considering RSC and without considering RSC are proposed,respectively.In IU models,all input parameters are regarded as interval variables.Combining orthogonal polynomials and vertex method,an effective method called orthogonal-vertex method is proposed to solve responses of IU models.Taking the upper bound of the dynamic characteristics as the optimization objective,IU optimization methods of the dynamic characteristics considering RSC and without considering RSC are proposed,respectively.Through numerical examples,the validity of the orthogonal-vertex method is verified,and the IU analysis and optimization of the dynamic characteristics are carried out.It is shown that compared with deterministic optimization,IU optimization ensures that the upper bound of the dynamic characteristics is within the design specification,which avoids the dynamic characteristics beyond the design specification due to the uncertainty.(4)Based on orthogonal polynomials and the deterministic models proposed above,hybrid random and interval uncertainties(HRIU)models of the dynamic characteristics considering RSC and without considering RSC are proposed,respectively.In HRIU models,some input parameters are regarded as interval variables,while others are regarded as random variables.Combining perturbation method and vertex method,an effective method called perturbation-vertex method is proposed to solve responses of HRIU models.The relationships between the reliability and the mean and variance of the dynamic characteristics are established.By taking the lower bound of the reliability as the optimization objective,HRIU optimization methods of the dynamic characteristics considering RSC and without considering RSC are proposed,respectively.Through numerical examples,the validity of the perturbation-vertex method is verified,and the HRIU analysis and optimization of the dynamic characteristics are carried out.It is shown that taking the lower bound of the reliability as the optimization objective,the upper bounds of the mean and variance are optimized,which ensures that the dynamic characteristics are within the design specification.