image: The "AlgoHex" project is developing algorithms for the automatic breakdown of shapes into hexahedral elements. The elastically deformable elements (bottom image) allow any shape to be modeled. This is a huge advantage over rigid "Lego models" which can only be modified in stages (top image). view more
Credit: Image: © David Bommes
The European Research Council (ERC) announced the recipients of its prestigious starting grants, awarded as part of the European Union research and innovation program "Horizon 2020". They include two researchers from the University of Bern.
David Bommes, Professor at the Institute of Computer Science, is receiving around 1.5 million euros for his five-year "AlgoHex" (Algorithmic Hexahedral Mesh Generation) project. Pierre Lanari at the Institute of Geological Sciences has been awarded around two million euros for "PROMOTING" (PROgrade metamorphism MOdeling: a new petrochronological and compuTING framework), also a five-year project.
AlgoHex: New algorithms to improve digital 3D models
Digital 3D models play a key role in many areas of today's world, including weather forecasting, the film industry, biomedical applications and the development of safe and efficient vehicles. In simulation in particular - the computer-assisted prediction of reality - structures are often split into small, cube-like elements known as hexahedra. Each element is elastically deformable, allowing one to model any type of shape. This is hugely important for the accuracy of simulation and offers an enormous advantage over "Lego models" in fixed, rigid sections.
Despite extensive work, researchers have so far failed to find suitable algorithms for automatically dividing shapes into hexahedral components. That failure equals huge expense in practice: generating 3D models often accounts for the lion's share of simulation project costs. David Bommes and his "AlgoHex" project are seeking a breakthrough with a new mathematical approach to the problem and looking to develop algorithms for automatic division into hexahedra. "Automatic algorithms are immensely important if we are to simulate complex phenomena in new applications in the future and also to reduce costs for existing applications," explains David Bommes.
He intends to use the funds to hire two postdocs, three doctoral students and a number of student assistants. "The Starting Grant enables us to put together a team that will focus in depth on all aspects of AlgoHex," says Bommes, who believes the University of Bern offers the ideal conditions for that work. "Switzerland - and the University of Bern in particular - is an attractive location for excellent researchers from around the world, and we need those researchers for an ambitious project like AlgoHex."
PROMOTING: Modeling fluid pathways in rock
The progressive transformation of crustal rocks during burial and heating releases fluids, which have an important role for earthquake generation, magmatism, crustal growth and for global geochemical cycles. The scientific community has made several hypotheses about fluid-rock interactions in the crust, but it still remains highly challenging to recognize and quantify fluid fluxes in crustal rocks and to model fluid pathways.
Pierre Lanari's "PROMOTING" project aims to understand how fluids affect the rock transformations in the Earth's interior between depths of 5 and 100 km using computer simulation. To this end, Pierre and his team will develop new high-resolution imaging techniques for rock analysis and produce the first computer model of metamorphism (changes in the mineralogical composition and rock texture) that integrates element mobility in fluids from rock scale to crustal sections. The results of computer simulation can then be compared with geochemical data obtained on rocks from all around the world.
"The Institute of Geological Sciences has the high-tech facilities required for this project," explains Pierre Lanari, adding that "the funds will enable me to set up a research team, acquire new instruments and organize field campaigns to collect important rock samples."
###