Research Agenda
All problems arising from natural phenomena are of multiphysics nature, whereas most of the engineered solutions available today are not. Therefore while all research is inter-related and it is not conceivable anymore to neglect this fundamental issue within the scientific and industrial communities, most industrial solutions are simplified down to extremes in order to reduce the complexities of design. This is of course what engineering is all about (solution after simplification or divide and conquer). However, it also has consequences on the quality and cost of the products, which have now become central issues for most of us including researchers and engineers.
During the last couple of centuries we have devised solutions to everyday problems via simplified mathematical models. Once the computers came around, we started building more and more complicated models, while still ignoring the “real-time” interactions between the different disciplines. As models became complicated, i.e., more mathematical and less engineering oriented - largely due to application of numerical methods such as finite elements- the physical complexities were more and more ignored or at best summarized for convenience by “boundary conditions” of a differential equation at the expense of “freezing” their impact on the quality of the solutions. Do these boundary conditions really exist in the nature? Where are the boundaries of a physical object and what happens if we extend those limits conditions a little bit further?
Some see the solution in ever increasing computing power and more detailed models (millions of “elements” or “cells” are everyday practice now). Recently, we are getting used to speaking of “robust” numerical models considering that the “dispersion” is a statistical observable quantity which can be identified like a model parameter, as being some sort of noise. In this vision all we need to do is to solve or launch many runs (hundreds or thousands) of a mathematical model with identified deviations from the nominal model.
The problems is that even in this more realistic perception of boundary limits, the issue of physical complexity is still ignored (and often confused with complicated models). Recall that nearly all differential equations have “constant” terms (or model parameters) and these in reality represent the effects (or interactions) of another simultaneously present and important physical phenomena which has been reduced to a simple, static effect.
There is obviously much ground here to be covered and we believe that multiphysics modeling is a particularly strong tool allowing for complex features of physical interaction to be taken into account. It is our belief that the key to treatment of sources of dispersion is in the field interactions which can only be taken into account via a true multiphysics treatment of the physical problems. This progress in the treatment of coupled field problems depends heavily on the success of solution methodologies requires further research in order to provide robust and reliable solution algorithms.
It is therefore essential to establish clear research directions within our future activities allowing for the above mentioned progress to be realized. The International Society of Multiphysics promotes collaborative research and know-how exchange throughout the world within scientific and industrial communities. In order to enhance its dissemination capacities the MPA promotes and facilitates the exchange of ideas and facilities via the creation of a network of excellence.
MULTIPHYSICS CENTRE (Research & Development)