Assistenzsystem für den wissensbasierten Aufbau konstruktionsbegleitender Finite-Elemente-Analysen

Description

Finite element simulations are crucial in today’s engineering design. They are increasingly applied to verify the strength of more and more complex products. The earliest possible use of finite element analyses saves high costs by avoiding engineering mistakes and by manufacturing less prototypes. However, extensive expert knowledge is required to set up efficient and reliable simulations and, due to capacity constraints, experienced simulation engineers cannot be consulted for every design step. Thus, the simulations are not applied early enough or have to be created by design engineers who often have less experience in simulations. The lower experience of the simulation users can therefore lead to inappropriate finite element models and wrong engineering decisions, which result in very costly and time-consuming iterations in product development.

In this thesis, a knowledge-based assistance system is developed to acquire the necessary simulation knowledge and to provide it to inexperienced simulation users. For the setup of the underlying knowledge base and the situational support of the simulation users, a novel ontology-based approach is presented. The innovation of this approach lies in the adaptation of Artificial Intelligence methods from the fields of Text Mining, Data Mining and Semantic Web. These methods are used to extract and purposefully query the required knowledge from existing simulation models and text-based documents created by experienced simulation users.