Inverse Problem Research in Upsetting Parameters Estimation

• Main Authors: Chen Chun Kung, 陳正堃
• Other Authors: Zone-Ching Lin
• Format: Others
• Language: zh-TW
• Published: 2001
• Online Access: http://ndltd.ncl.edu.tw/handle/19227417846972331877

碩士 === 國立臺灣科技大學 === 機械工程系 === 89 === A simulation framework that allows the simulation process to approximate the realistic metal formation process is presented in this paper. The framework gives us better control of the results of metal formation─the goal in this paper. The case of upsetting was discussed in this paper as the friction of the upsetting workpiece and die contact interface has always been an issue of considerable concern. This paper focused on how to determine the variation of the friction coefficient during the upsetting formation process, the concept of solving unknown parameters by means of inverse problem, and the selection of the Levenberg-Marquardt numerical optimization Method based on the experimental data. The variation of the friction coefficient was calculating the upsetting formation process inversely by means of a specific constraining function and convergence criterion, plus axial symmetrical thermo-elastic-plastic finite element method. The variation of friction coefficient was then explained in a micro perspective. As a result of the inverse calculation framework, physical properties that better approximate the realistic conditions were obtained, which presented a more comprehensive upsetting formation process. A case study featuring different geometric dimensions, different materials and different machining temperatures was presented. Specifically, mild steel cold upsetting (based on the experimental load), Molybdenum metal warm upsetting (based on the experimental load) and AISI1015 cold upsetting (based on the experimental temperature) were studied to observe the variation of friction coefficient during the upsetting process, the distribution of stress and temperature during workpiece simulation and appearance changes. The phenomena observed in this paper were found to be an exact match of those observed in previous studies. The findings prove that the use of load or temperature to the inverse calculation of the friction factor presented in this paper is a reasonable approach.