Automotive Hub Unit Bearing Riveting Process Parameters Optimization Base On FEM Analysis

Riveting assembly process is an efficient hub bearing unit assembly process proposed tomeet the lightweight, integrity and high reliability requirements for hub bearings. Its mainidea is to provide axial preload for hub bearing unit by the plastic deformation of the shaft endinstead of the original lock nut, so that the axial preload can be more precisely controlled,ensuring the quality and reliability of the hub unit bearings. Therefore, the optimization of theriveting assembly process is of significant importance, it will be a guide for the design anddevelopment of the hub bearing.Riveting assembly process is a complex forming process under the coupling effects ofdifferent parameters with three-dimensional deformation, asymmetric, unsteady, highly non-linear characteristics. During the manufacturing process, the shape of rivet head can be con-trolled more easily than the other parameters, it will greatly influence the forming load, stressdistribution during and after the process, as well as the shape and the quality of the finalproduct. Therefore, in order to reflect the authenticity, based on FEA and support vector re-gression surrogate model, this paper aims to conduct systemic research to optimize the surfaceof the rivet head, the main contents and results are as follows:(1) By tensile tests, ring upsetting tests, measurement of the feed speed and axial load,accurate derivation of the rivet head trajectory in the riveting process, the parameters of thesimulation model are determined.(2) By addressing key FEM techniques such as meshing,contact settings, boundary con-ditions, as well as choosing the proper scaling factor, a3D elastic-plastic simulation model forthe riveting process is built. And the validity of the model is verified by physical test.(3) The optimization goal for parameters in riveting process is decided, and parameteri-zation equation of the rivet head surface is established, based on Latin hypercube samplingstrategy and reasonable3D explicit dynamic finite element modeling technique, how theforming surface impacts the forming shape, axial preload, diameter variation of the inner ring,load distribution and so on, is concluded.(4) According to the analysis result of the finite element model, a mathematical surrogatemodel based on support vector regression is established, and the forming surface of the rivethead is optimized by using genetic algorithms. The corresponding FEA model is establishedto verify the surrogate model, the mathematical model is corrected in order to improve predic-tion accuracy of the surrogate model based on optimization and simulation results. Experi-ments are taken to verify the optimization results. In this paper，a systemic riveting process optimization solutions for hub bearing unitfocused on the forming surface of the rivet head, is proposed, it may be a guide for the designand development of hub bearing units. With of the optimization program proposed in this pa-per, other parameters in riveting process, like feed rate, rotation speed and tilt angle of therivet head, etc. can also be optimized...