Methoden zur Auslegung von robusten Prozessen in der Warmmassivumformung

摘要

In industrial production variations occur during processing and in the materials behavior. These variations are unavoidable. They can lead to products that do not meet the quality requirements. At worst, a stop of the production process can be caused. In today’s process design of hot forging processes these variations cannot be considered. This thesis focuses on the development of new methods and strategies for the identification and quantification of uncertainties and variations of all necessary input data to describe entire process chains in hot forging and further their implementation in numerical tools for process design, in particular in the FEM. Based on the well-established resampling methods a new method applicable for complex, non-linear, microstructure-based material models is presented. This method implements the variations in the material behavior by distributions of model parameters. Combining this new method with DoE-based simulation studies it is possible to predict all necessary product-relevant and process-relevant target values in statistical distributions. In this case, product-relevant values are e.g. grain sizes and recrystallized fractions. Process-relevant target values are e.g. forging loads and tool stresses. The results are validated by the production of near net shape bevel gears. Thus, an important basis for the future consideration of process and material variations is provided, which enables for estimating the robustness of processing conditions and product qualities during the design phase.