This paper was produced for the 2019 NAFEMS World Congress in Quebec Canada

Resource Abstract

After decades of using CAE so to reproduce the physical behaviour of physical components to predict performances and support the design process, the notion of digital twin has been created reflecting the growing maturity of CAE technology, but also the fact simulation is the virtual and accurate copy of a real asset. Digital twin integrates all data, models (engineering, data-based, simulation), and otherwise structured information of a product. However, what has been changing the most since the initiation of the CAE methodology, is how much of the different aspects linked to product design (like plant, infrastructure system, or production process) have started to be integrated in the “Digital Twin”. Indeed, depending on the moment it’s used in the design workflow, the digital twin may represent a product, a production system or a production process. In all cases however, the objective is to have a digital representation suited to the purpose in terms of level of detail, completeness, accuracy, and execution speed.



In this paper, we will consider different examples involving mechatronic systems, composite structure or the use of additive manufacturing and see how simulation extends from “classical” performance simulation to a much more global “Digital Twin” to be used by CAE Engineers so to optimize the product, and in best case scenario, design it right at the first time.



The challenges -and opportunities- however are multifold. Interoperability of models, model transformation and co-simulation are key to realize system level simulation. Computing performance alone cannot address the requirements posed by ever more complex systems and applications like massive design space exploration or interactive simulation, expressing the need for disruptive and novel solver technology. But also modelling in itself needs to be extended. For example, modelling the effect of assembly and manufacturing processes should be included in industry workflow as they have an impact on product performances. Not only at the level of strength or durability, but also on cost and delays.



Finally, integrating the digital twin with data measured on the physical assets, hereby reusing all previously established engineering knowledge allows new applications such as virtual sensing, model-based control and hybrid system modelling including hardware, software and numerical components.