Convergent Science

Detailed Chemistry Modeling

Fast detailed chemistry modeling for CFD simulations

SAGE, with no mechanism size limit, is equipped with a multi-zone model, Dynamic Mechanism Reduction (DMR) and adaptive preconditioning to increase the efficiency of linking detailed chemistry with CFD. With these techniques, mechanisms with up to 1,000 species are now being simulated with CONVERGE™. In many cases the best combustion model is solving the elementary reactions directly.

CONVERGE™ includes this efficient parallel detailed chemistry solver.

SAGE Features

  • Part of the standard version of CONVERGE™
  • Reads data in CHEMKIN format
  • Can be run in parallel to significantly reduce run-times
  • Several numerical improvements have been made to reduce run-times

The image below shows the temperature field on a cut-plane through the cylinder of a spark-ignited engine simulation. The simulation was done in CONVERGE™ using the SAGE detailed chemistry solver. Adaptive Mesh Refinement was used in combination with SAGE to accurately resolve the location of the flame.

 

flame_sage_grid

 

 

Through a partnership with Lawrence Livermore National Laboratory, the multi-zone chemistry solver has been incorporated with SAGE to decrease run times with minimal accuracy penalty. The combined functionality, SAGE-Multizone, groups cells with similar temperature and equivalence ratio into zones and solves chemistry on the zones collectively. As the number of zones is much lower than the number of CFD cells, SAGE-Multizone greatly reduces the computational requirements for detailed chemistry. The user prescribes the desired tolerance for zoning (more zones increases accuracy and run-time).

 

sage2

 

For some cases such as dual-fuel applications, two dimensional zoning (i.e. temperature and phi) is insufficient to accurately model the combustion. For these cases, SAGE-Multizone allows an arbitrary number of bins to be used.

SAGE-Multizone is available as a standard feature of CONVERGE™ and allows for detailed chemistry simulations to be run in a reasonable time frame.

More information on the partnership with Lawrence Livermore Laboratory can be found here: https://ipo.llnl.gov/?q=about_ipo-success_stories-convergent_science

With SAGE, the user can specify any chemical mechanism they choose (web mechanisms can simply be cut and pasted into the SAGE input files). There isn’t any theoretical limitation on the number of species or reactions. However, there is a practical limitation since the run-time will scale with the mechanism size.

With Dynamic Mechanism Reduction run-times are getting short and shorter. The below graph shows the evolution of the SAGE solver and run-time speeds.

 

SageEvolution

 

DetailedChemGraph

 

For more information on the SAGE chemistry solver, please contact us.

Share This Page