Features

PARIS has the following features:

- Free code under the GPL.
- Written in Fortran 90/95/2003
- Use a regular structured grid and finite volume discretisation.
- Uses MPI and a regular array of subdomains for parallelisation.
- Momentum : QUICK, ENO, Superbee, Verstappen either with a velocity or a momentum-conserving formulation.
- Implicit viscous terms.
- Either Volume-Of-Fluid or Front-Tracking may be used to follow interfaces.
- VOF uses either CIAM or Weymouth-Yue scheme for advection.
- Height-Function method for curvature estimation and surface tension.
- Parallelized Lagrangian Point Particles (LPP) with VOF to LPP and LPP to VOF conversion
- Second order in time and space with projection method.
- Uses either hypre or internal Poisson solver.
- Immersed solid boundaries
- Has either Interfaces or Free surfaces
- Thermal effects (energy equation).

Speed

Paris has good speed properties compared to generalist codes such as OpenFoam or Gerris as shown in the figure below. However this comparison is not entirely fair since regular structured grid codes are much easier to optimise than unstructured or oct-tree grid codes. We are currently working on comparisons on other regular structured grid codes.

Speed plot from Arrufate et al. 2014

Speed of Paris Simulator in numbers of cells/core/second compared to OpenFoam and Gerris on a porous medium invasion problem. Results obtained on the supercomputer Rostand (Intel six-core X5660 Westmere at 2.8 GHz) from Total's CSTJF in Pau, France on the porous medium invasion problem. From reference Arrufat et al. 2014

speed plot on Curie

Speed of Paris Simulator in numbers of cells/core/second compared on the Tier0 supercomputer CURIE from the TGCC (Intel Sandy Bridge E5-2680 at 2.7 GHz). Configuration close to that of Ling et al 2015b. Simulation on 4096 x 2048 x 2048 = 17 Billion cells for 20 steps. The HYPRE-PFMG solver is used and a first-order integration in time is performed.

Parallel speed-up

Paralell speed up

Ideal and actual parallel speedup for ParisSimulator on the BG/Q machine Turing at IDRIS.

Paralell speed up on Curie

Peak and average speed up for ParisSimulator on CURIE at TGCC. Configuration close to that of Ling et al 2015b. Simulation on 4096 x 2048 x 2048 = 17 Billion cells for 20 steps. The HYPRE-PFMG solver is used and a first-order integration in time is performed.