Modularer Ansatz zur Simulation verfahrensübergreifender fertigungstechnischer Prozessketten

摘要

Several modern trends in product development have resulted in the material properties of a part having a greater influence than ever before on the entire life cycle of a product. The ongoing trend towards lightweight structures in the automobile industry, for example, is leading to a steady decrease in the wall thickness of workpieces, and thus to an increase in the influence that the rim zone has on the part as a whole. Another example is to be found in the aerospace industry which is facing ever more stringent safety requirements. These in turn directly influence the safety requirements demanded of the workpieces to be produced. Such trends in product development result in an immense number of specifications that require detailed knowledge of the properties of the finished part. The manufacturing history forms the basis for determining these properties. For some years now, the targeted development of manufacturing process simulations has introduced the benefits of numerical computation methods to the manufacturing domain. Simulation experiments are thereby intended to replace iterative optimisation cycles during production. In spite of the irrefutable advantages of numerical manufacturing simulations, there is also one significant drawback: when it comes to predicting the properties of a part, these simulations only allow one manufacturing process to be focussed on at any one time, meaning that the material history of the part, i.e. the previous manufacturing steps, are not taken into account. Within the scope of this thesis, a fundamental concept for coupling any number of different manufacturing simulations with a sequential process chain has been developed for the first time. Based on an analysis of the boundary conditions of the individual manufacturing simulations and on the requirements of the manufacturing partners, a concept has been developed that is divided into the man-machine aspect and the data processing aspect. The effectiveness of this concept has been proved within the framework of this thesis by applying it to the processes in the manufacturing chain, i.e. casting – heat treatment – machining. In particular, the high quality of the data processing chain has been substantiated. The considerable influence that material history has on the properties of a part has been demonstrated by comparing a machining simulation in which the part exhibited residual stresses and displacements from a heat treatment process with a machining simulation where the initial state of the part was neutral. Comparative and data comparative visualisation methods, developed as part of this thesis, were used to carry out the comparison. In accordance with the overall goal of this thesis, a modular approach to simulating a manufacturing process chain consisting of several different manufacturing steps has been developed for the first time. In this way, a basis for simulation-assisted analysis of entire manufacturing chains has been created.