Description of the infrastructure
TransAT© is a multi-physics, finite-volume code based on solving single and multi-fluid Navier-Stokes equations on structured multi-block meshes. The grid arrangement is collocated and can thus handle more easily Body-Fitted Coordinates (BFC) grids. The solver is pressure based (Projection Type), corrected using the Karki-Patankar technique for subsonic to supersonic compressible flows. High-order time marching and convection schemes can be employed; up to 3rd order Monotone and TVD-bounded schemes in space and 5thorder RK in time. An algebraic multigrid algorithm is employed for the pressure equation, involving relaxation, restriction and prolongation up to 6 levels, to achieve high rates of convergence and error estimation. Multiphase flows can be tackled using (i) interface tracking techniques (Level Set, VOF with interface reconstruction, and Phase Field), (ii) N-phase, phase-averaged mixture with Algebraic Slip, and (iii) Lagrangian particle tracking (one-to-four way coupling).Re-distancing of level set is ensured via 3rd order WENO scheme in Cartesian meshes and via fast marching BFC grids. Mass conservation is enforced using global and local mass andvolume conserving schemes.The objective is to offer beyond current licensing model of CFD code TransAT, which is presently based on a workstation-based approach, where the user pays per installation on a workstation (or on a cluster).
From the user point of view, the optimal situation is that where the solution can be run from a browser. SME’s being most often connected to the Internet via classical means (ADSL, public fibre optic) working should be user friendly with a bandwidth of 200 Mb/s.
Detailed description of the demonstrator
A cloud-based solution will allow running cases that cannot be handled on local PC’s. It will be more cost effective since the power will be adapted to the cases and all the indirect costs (support and maintenance for example) will be scaled.
It is expected that the cloud behaves like a supercomputer and allows the user to see a single computer with specified number of cores, memory per node and processor. Concerning the User interface, it’s needed a web based solution that allows the user to build at once its configuration: number of nodes, software licenses, etc. In addtion this experiment will need occasionally access to a very large number of cores (105 to millions). Moreover new business scenarios will follow through a “Simulation as a service” model. This will enlarge at moderate cost the capacity of EUROBIOS to serve its clients. This also will enlarge the licenses sold by ASCOMP.
Computational Fluid Dynamics, fluid flow, heat transfer