Workshop 1: Coupled flow: aspects of discretization, solvers, and high performance implementation" Workshop 2: Parameter Estimation and Optimum Experimental Design

Interdisciplinary Center for Scientific Computing, Heidelberg University in collaboration with Indrapratha Institute of Information Technology (IIIT Delhi) organised two concurrent workshops, titled, "Coupled flow: aspects of discretization, solvers, and high performance implementation" by Prof. Peter Bastian and Prof. Guido Kanschat and "Parameter Estimation and Optimum Experimental Design" by Prof. Ekaterina Kostina and Prof. Johannes Schloeder.  

 

The workshops aimed at highlighting the power of Scientific Computing and how it helps solve real world problems by applying Simulation, Optimisation and Visualisation techniques. The Interdisciplinary Center for Scientific Computing (IWR), Heidelberg University is one of Europe's biggest research units devoted to this challenging field of research. 

 

The purpose of the first  workshop was to discuss new developments in the simulation of flow and transport in porous media and its coupling with free flow regions. In the workshop it was discussed the discretization of single- and multiphase flow equations with finite volumes and mixed and discontinuous finite elements. The geometric and algebraic multigrid methods for the solution of the arising algebraic problems were presented and a special focus on high-performance implementation on modern parallel multi-core architectures scaling well in core count  and polynomial degree.

 

The second workshop focussed on mathematical methods of model-based simulation and optimization that are being increasingly used in science, engineering and industrial applications. Their successful application requires the development and quantitative validation of mathematical models. This difficult mission is supported by numerical methods for inverse problems of model validation and calibration. Here, fundamental tasks are the estimation of unknown model coefficients by means of experimental data and design of optimal experiments to choose e.g. measurement times and experimental conditions,which deliver sufficiently accurate estimates for the model parameters. The application of these methods in industrial practice shows, however,that in order to use their full potential we have to deal with a number of new mathematical challenges. The topic of this mini-symposium is on the new developments in efficient numerical methods for inverse problems of model calibration and validation  and on the transfer of these methods into challenging application areas.