Contribution à la caractérisation des moyens de mise en forme : application aux presses à vis

Abstract

The metal forming industry uses more and more numerical simulation, and especially the Finite Element software Forge®, with the aim to fulfill 3 main objectives: predict tooling lifetime, predict the amount of energy required to obtain the part desired and predict the thermomechanical path leading to the optimal final microstructure. Currently, numerical simulations allow good predictions regarding forging operations with currently used material, like steel. But to remain competitive forging industries cannot work only with steel, and have to innovate by developing new forging process for high performance material. But concerning the forging operations of these new high performing material, significant difference are observed between numerical and experimental results. In particular regarding the prediction of energy required to correctly forge a part. This study aims at understanding this disparity between numerical and experimental results, in the special case of screw presses. For that, the first step describes the functioning of screw presses but also the screw presses model implemented into the Forge software. Experimental results of upsetting tests in 2 different conditions have then been compared to results obtained numerically by simulating the corresponding forging tests. Significant differences are observed, and the necessity to define a stiffness coefficient for the press is highlighted. Two methods have been carried out to determine the stiffness of the press, a theoretical one and an experimental one. Concerning the experimental one, external tools of acquisition like a rapid camera and a 3D tracking points system have been used to analyze the press behavior during blows. Finally, stiffness values obtained with the two different methods are compared.