Cyber-physical systems (CPS) are characterized by the interaction of mechanical, electronic and information technology subsystems. Model-Based Systems Engineering (MBSE) is an approach for the cross-domain development of CPS and requires compatible methods and models for a function-oriented collaboration of the domains. However, the mechanics operate mainly component-oriented and thus cannot participate in the function-oriented development process. We propose a new modeling method that allows mechanics to develop a consistent and function-oriented system model from requirements and functions to principle solutions. The principle solution formally specifies the physical effect, active surfaces and material through which a function is fulfilled. Since principle solutions are consistently parameter-based, they can be calculated by linked external models and checked against requirements. This enables to examine possible solutions for functions at an early stage without having to develop components. Since principle solutions consist of recurring elements, we also propose a modeling concept for a solution libraryso that proven models can be efficiently reused and the modeling effort is reduced. Modeling method, test and solution library are explained using the example of an electric water pump of an automotive cooling circuit.

Cyber-Physical Systems (CPS) connect mechanics, electrics and electronics as well as software. Considering the interactions between these domains is a major challenge in product development. Model-Based Systems Engineering (MBSE) enables cross-domain system development, based on commonly understood system architectures. However, there is still a gap in using MBSE for development, especially in mechanics. Systematic evaluation of design decisions based on system architecture models is not done. The strong simulation methods predicting physical product behaviour are not well connected to the evolving system architecture. We propose a new approach for connecting simulation and design models in a System Modelling Language (SysML) system model using the example of a combustion engine’s cooling circuit. The electric-mechanical coolant pump is chosen for design. The design processes of the hydrodynamic pump wheel and the volume flow controller are modelled as workflows in SysML. Here, we integrate design models for the pump wheel and the controller into the workflows. Thereby, parameters of the pump wheel and controller are calculated. Design results are automatically transferred to CAD. A workflow is created, testing the cooling circuit’s behaviour against requirements, combining a Simulink control model and a Simcenter Amesim heat transfer model. The approach results in a framework for automated design and test processes.

No abstract