The first issue of the magazine “Benchmark”, the international magazine for engineering designers and analysts from NAFEMS, in 2018 is entitled “Bringing it all Together: Multiphysics, Multiscale & Co-Simulation”. NAFEMS is the international association for the engineering modelling, analysis and simulation community, a not-for-profit organization established in 1983. Dr.-Ing. Volker Gravemeier, chief executive officer of AdCo EngineeringGW, is an active member of the NAFEMS Multiphysics Working Group (NAFEMS MWG).
In his technical editorial, Prof. Henrik Nordborg, chairman of the NAFEMS MWG, addresses some of the trends driving the use of multiphysics. Among others, the driving forces are, on the one hand, the “constantly increasing demand for accuracy in simulations, making it impossible to continue to rely on over-simplified models”, and on the other hand, the fact that “computers are getting faster and cheaper, giving us the necessary computing power to handle complex models”. However, Prof. Nordborg emphasizes that the knowledge for successfully doing multiphysics simulations is particularly demanding: “Actually, the main limiting factor today might be the training of simulation engineers. Although modern simulation tools are easy to use, they are no replacement for knowledge. One cannot design a product or a process based on fluid flows without a basic understanding of fluid mechanics. Likewise, it is not possible to do mechanical, thermal, or electromagnetic design without knowledge of these fields. A major problem with multiphysics simulations is that most of us leave university with only one degree, or if more than one degree, the subsequent ones are usually in a similar area. Successful application of multiphysics simulations therefore depends on either having extremely erudite people or on well-functioning teams of simulation specialists.” Multiphysics is one of our specialist fields at AdCo EngineeringGW, and from our long-term experience in this field, we may add to this that not even adequate knowledge in all of the involved physical fields is sufficient, but particularly the coupling strategies between the fields and thus the knowledge thereof is key to accurate, robust, stable and reliable multiphysics simulations. AdCo EngineeringGW offers you the unified knowledge of the individual fields as well as the coupling of the fields for many industrially relevant multiphysics problems.
Following the technical editorial, a number of articles shed light on various multiphysics applications. Among others, Hubertus Tummescheit provides “an overview of co-simulation with a focus on best practices” entitled “Co-Simulation – Art or Science?”. This article includes a discussion of the drawbacks of co-simulation: “in co-simulation, the smooth approximations of variables at model interfaces are communicated in discrete time intervals. This coupling introduces discontinuities and a delay in the duration of the communication interval. These two flaws introduced by the fundamental setup of co-simulation are the root cause of all difficulties in using it. Even advanced co-simulation techniques with sophisticated master algorithms can reduce, but not eliminate the introduction of these two fundamental flaws.” In a subsequent and more detailed analysis, several issues are addressed, among other things, the stability of co-simulation: “The stability properties … of the coupled model are always worse than the stability properties of the system if it were modeled as a monolithic model. Co-simulation requires either (a) a delay of one time step in the propagation of results from the output of one model to the input of the next model or (b) some extrapolation from past results, to estimate the input to a model from past outputs upstream of that model. Coupling through co-simulation always deteriorates stability, as well known from the theory of feedback control with delays.”. These observations go well together with our own experience, based on which we preferably use and may offer you superior monolithic approaches for most multiphysics applications.