3. A practical guide to multidisciplinary analysis and optimization

3.1 Identifying the variables and functions involved

In order to create a multi-disciplinary process, it is necessary to couple different disciplinary models (or sub-systems) and thus define the different types of variables involved (e.g. design variables, interdisciplinary coupling variables) and the functions of interest (e.g. optimization criteria, constraints). This step is important because it aims to ensure the coherence of the multidisciplinary process, and requires system-level agreement on the exchanges between the different disciplines (or sub-systems) involved. Indeed, in order to build the multi-disciplinary process, inputs from each discipline must correspond either to design variables or to output coupling variables from other disciplines. This is why the preliminary...