This has been the first edition of this call with a highly competitive success rate near 6%. The awarded project is based on the open-source scientific library Gridap.jl, developed by Francesc Verdugo. Gridap.jl is a general-purpose finite element package used to solve partial differential equations in different scientific domains. This software has been used by researchers from renowned academic institutions such as TU Delft, VU Amsterdam, RWTH Aachen, Imperial College, Chalmers, MIT, Caltech, and Monash.
Check the website of E science centre for more information on the call.
Check the github repository. for more details about the Gridap.jl library.
This project aims at enhancing the multi-purpose finite element (FE) library Gridap to meet the demands of the research communities interested in the solution of partial differential equations (PDEs), namely domain scientists, numerical analysts, and HPC experts. To this end, we will implement three key features missing in the code:
1. a well-documented low-level interface that provides easy access to the library internals
2. a parallel implementation able to solve multi-physics problems that couple physics on domains of different topological dimensions
3. support for graphic processing units (GPUs).
This project's objective is not only implementing these features but also engaging the targeted communities to use and contribute to the development of the code. This will include:
1. the implementation of concrete examples showing how the library is able to handle specific cases of interest in the targeted research areas
2. the creation of a developer’s manual and other actions to make it easier to contribute to the code
3. the organization of two workshops.
The proposed work is backed up by a multidisciplinary research team composed of significant members of the communities we want to make the code accessible to. This includes two applied mathematicians which are respective experts in the development of so-called structure-preserving discretization methods and isogeometric analysis, a domain expert in the field of multi-physics problems for off shore renewable energies, a leading expert in high-performance distributed computing, and the creator and main developer of the Gridap library.