Coupling the Multiple Mapping Conditioning Mixing Model with Reaction-diffusion Databases in LES of Methane/air Flames

ABSTRACT

The Multiple Mapping Conditioning (MMC) model for sparse Lagrangian Large-Eddy Simulations (LES) is coupled with tabulated chemistry based on reaction-diffusion manifolds (REDIM). In addition to assessing the reduction in computational cost compared to direct integration on the fly, a comparison with the previously implemented REDIM solver based on the Eulerian Stochastic Fields (ESF) is performed. The methane/air Sandia flames D and E are selected as a benchmark, which are amenable to a 2D REDIM database having a passive scalar and one progress variable as dimensions. A comparison between direct integration ESF and MMC shows that the latter gives better predictions of the conditional means and fluctuations, offering at the same time a significantly reduced computational cost. When coupled with REDIM, MMC and ESF models provide similar results, indicating the significant sensitivity of the predictions to the thermochemical states which are restricted to the same REDIM manifold. When the states are projected back onto the REDIM surface in the direction where the progress variables remain almost constant, accuracy is lost in the cold fuel-lean region with MMC. Furthermore, the computational performance gain in MMC compared to ESF is reduced when the REDIM tabulation is used, pointing to non-chemistry related computational bottlenecks in the MMC algorithms.

KEYWORDS:

• Sparse Monte–Carlo simulation
• LES
• multiple mapping conditioning
• reaction-diffusion manifolds
• Eulerian stochastic fields

Acknowledgments

Computing time was provided on the GCS Supercomputer SuperMUC at Leibnitz Supercomputing Centre (LRZ, https://www.lrz.de) under grant pn98ze. Paola Breda is gratefully thankful to Dr.-Ing. Chunkan Yu from the Institute of Technical Thermodynamics at the Karlsruhe Institute of Technology (KIT) for the fruitful discussions and the generation of the REDIM tables. Eshan Sharma would like to acknowledge the financial assistance provided by the Ministry of Human Resource Development (MHRD), Government of India during the PhD program at the Indian Institute of Technology, Kanpur.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ https://computing.llnl.gov/projects/sundials/cvode

Additional information

Funding

This work was supported by the Ministry of Human Resource Development.