Cerisse

A research computational solver for Large Eddy Simulation and Direct Numerical Simulation of compressible turbulent reactive flows

Version 1 created by: Enson Un, Salvador Navarro-Martinez

Capabilities

• Fifth-order low-dissipation WENO¹ and TENO² schemes
• LES/PDF model solved using the Eulerian Stochastic Fields approach³
• Wall-modelled Stochastic Fields and wall-modelled LES with advanced SGS models
• Adaptive mesh refinement powered by AMReX⁴
• Realistic thermodynamic and chemical properties using PelePhysics⁵

Getting the code

1. Clone the repository

   git clone git@github.com:salvadornm/cerisse.git
   cd cerisse

2. Checkout to the cerisse1 branch

   git checkout cerisse1

3. Get all submodules

   git submodule init
   git submodule update

4. (Optional) Disable some PelePhysics constraints and aborts

   cd EB_CNS/Tools/scripts
   ./pp_modify.sh

Or download a release from here.

Compiling and running

We use GNU Make system for generating executables. The CMake system is not being maintained.

1. Go to an exec folder, for example,

   cd EB_CNS/Exec/ShockReflect

2. Compile the third-party SUNDIALS library first, which requires CMake. This only need to be done once per compiler settings combination.

   make TPL

3. Compile the executable. You may use the -j flag to build multiple jobs in parallel.

   make -j8

4. Run the executable, for example,

   mpirun -np 8 ./Cerisse2d.gnu.MPI.ex inputs

• See prob_parm.H, prob.H, and prob.cpp for the problem definition
• See inputs for runtime options. See default_parm.H for more options and their default values.
• See GNUmakefile for compile-time options
• More information about the solver can be found in docs. To generate the docs locally, do mkdocs serve in the parent directory and point the browser to 127.0.0.1.8000. You may also find cerrisse2 docs useful, but bear in mind that the two solvers are different.

Citation

This open-source software is distributed under the BSD3 license. If you use it in your work, please cite our Combustion and Flame article

@article{Un2025,   
   author = {Tin-Hang Un and Salvador Navarro-Martinez},
   doi = {10.1016/j.combustflame.2024.113897},
   journal = {Combustion and Flame},
   month = {2},
   pages = {113897},
   title = {Stochastic fields with adaptive mesh refinement for high-speed turbulent combustion},
   volume = {272},
   year = {2025}
}

We recommend to cite as well AMReX and PelePhysics as in the references.

References

Footnotes

1. Castro et al. (2011). High order weighted essentially non-oscillatory WENO-Z schemes for hyperbolic conservation laws. JCP, 230(5), 1766–1792.

   Balsara et al. (2024). Efficient Alternative Finite Difference WENO Schemes for Hyperbolic Conservation Laws. Comm. Appl. Math. Comput., 1–54. ↩

2. Fu et al. (2016) A family of high-order targeted ENO schemes for compressible-fluid simulations. JCP, 305, 333–359. ↩

3. Un & Navarro-Martinez (2025). Stochastic fields with adaptive mesh refinement for high-speed turbulent combustion. Combust. Flame, 272, 113897. ↩

4. Zhang et al. (2019). AMReX: a framework for block-structured adaptive mesh refinement. JOSS, 4(37), 1370. ↩

5. Henry de Frahan et al. (2024). The Pele Simulation Suite for Reacting Flows at Exascale. Proceedings of the 2024 SIAM Conference on Parallel Processing for Scientific Computing, 13–25. ↩