No abstract

The conceptual design is a decisive phase where the simulation teams would like to quickly pre-validate spatial architectures from the physical architecture proposed by the system architects. In order to support them to efficiently achieve this task while meeting various industrial requirements, three approaches were proposed and compared, and finally the last and innovative approach called SAMOS (Spatial Architecture based on Multi-physics and Organization of Systems) is presented and described. The corresponding platform will allow to validate the spatial allocation of components under geometrical and multi-physical constraints, while facilitating the collaboration between the different design actors during the conceptual design.

Today, automotive design has to face numerous exciting challenges. The growing globalization causes an intensified competition amongst car manufacturers and forces them to reduce the required development time in order to shorten time to market, to appear first with attractive new products. Efficient and flexible processes and tools are necessary to handle the arising complexity efficiently. Parametric-associative 3D-CAD systems offer ideal conditions to face this challenge in virtual development. The present paper focusses on a special issue in automotive concept phasethe vehicle architecture layout process and required parameterization strategies. In most cases, parametric-associative relations defined within 3D-CAD models are of rigid kind. This implies that a formula, which is defined within a 3D-CAD model in order to evaluate a specific parameter, cannot change the input/output situation of involved parameters. In most application cases, this disadvantage can be neglected, but not in case of vehicle layouting in the early concept phase. Since geometric boundary conditions which define the geometric base of a vehicle concept can vary significantly, a rigid model parameterization is not the proper solution and prevents efficient reuse of 3D-CAD models. Additionally, rigid parameterization concepts lack of the required flexibility when having to manage multiple design variants in a single model. Therefore, the present paper outlines a possible strategy, which enables the use of advantages of parametric-associative design, while allowing changes of relations-evaluation behavior in context of respective technical issues and simultaneously preserving necessary geometrical model consistency.

This paper develops a mathematical formulation of a margin problem in an automotive battery sizing use case. This formulation is done thanks to theoretical models of margin. This enables to use an approach with explicit margins, which is compared to a worstcase analysis and a probabilistic modeling. The models of margin are then adapted to a numerical implementation through the definition of patterns and presets adapted to the case study.

In a highly competitive industrial environment, reducing development time while maintaining a low level of risk makes simulation‐based design an effective way to cope with the complexity of multidisciplinary systems. To accelerate the design cycle, capitalization and reuse processes can support the search for similar models in an existing database. Therefore, we propose to prototype a search engine for simulation models from system modeling. This paper focuses on the presentation of our initial works around an ontology‐based search engine of simulation models representing the behavior of a given system function. Algorithms and similarity metrics developed are based on both thesaurus reference functions associated with semantic relations, and on various port attributes. Finally, an implementation of the corresponding inference engine has been demonstrated in an autonomous vehicle case study.