nekCRF: A next generation high-order reactive low Mach flow solver for direct numerical simulations

Stefan Kerkemeier,Christos E. Frouzakis,Ananias G. Tomboulides,Paul Fischer,Mathis Bode
2024-09-10
Abstract:Exascale computing enables high-fidelity simulations of chemically reactive flows in practical geometries and conditions, and paves the way for valuable insights that can optimize combustion processes, ultimately reducing emissions and improving fuel combustion efficiency. However, this requires software that can fully leverage the capabilities of current high performance computing systems. The paper introduces nekCRF, a high-order reactive low Mach flow solver specifically designed for this purpose. Its capabilities and efficiency are showcased on the pre-exascale system JUWELS Booster, a GPU-based supercomputer at the Jülich Supercomputing Centre including a validation across diverse cases of varying complexity.
Computational Physics,Performance,Fluid Dynamics
What problem does this paper attempt to address?
This paper aims to solve the problem of chemically - reactive flows in complex geometries and conditions for high - efficiency direct numerical simulation (DNS). Specifically, the paper introduces a new - generation high - order reactive low - Mach number flow solver named nekCRF, which is specifically designed to fully utilize the capabilities of current high - performance computing systems. The following are the main objectives of the paper and the problems it solves: 1. **Improve computational efficiency**: By leveraging the latest supercomputer architectures (especially GPU - accelerated systems), nekCRF can significantly improve computational efficiency while maintaining high precision. This enables large - scale, high - fidelity simulations, thus providing valuable insights for optimizing the combustion process. 2. **Support complex geometries**: nekCRF supports unstructured curvilinear boundary - conforming meshes and can accurately represent complex geometries. This is crucial for simulating combustion processes in practical applications, as combustion chambers and other devices in actual engineering often have complex geometric structures. 3. **Handle multi - scale problems**: Chemically - reactive flows involve a wide range of spatio - temporal scales, as well as a large number of thermochemical scalars and complex interactions between fluids and chemistry. nekCRF can effectively handle these multi - scale problems through efficient algorithms and parallel computing techniques. 4. **Open - source code**: Unlike most existing combustion research codes, nekCRF is open - source. This promotes cooperation, transparency, and benchmarking, and contributes to the common progress of the scientific community. 5. **Verification and performance evaluation**: The paper demonstrates the capabilities and efficiency of nekCRF by validating cases of different complexity levels. In addition, a detailed performance study was carried out on the pre - exascale system JUWELS Booster, including cross - code comparisons. ### Specific problems - **Efficient simulation of chemically - reactive flows**: Through optimized algorithms and parallel computing techniques, nekCRF can efficiently perform direct numerical simulations of chemically - reactive flows on GPU - accelerated supercomputers. - **Support for complex geometries**: nekCRF supports unstructured curvilinear boundary - conforming meshes, can accurately represent complex geometries, and is suitable for simulating combustion processes in practical engineering. - **Handling of multi - scale problems**: nekCRF can effectively handle the wide range of spatio - temporal scales and complex thermochemical scalars involved in chemically - reactive flows. - **Open - source promotes cooperation**: The open - source nature of nekCRF promotes cooperation, transparency, and benchmarking in the scientific community. - **Performance verification and evaluation**: By validating cases of different complexity levels and conducting detailed performance studies on pre - exascale systems, the high - efficiency and reliability of nekCRF are demonstrated. In summary, nekCRF provides a new solution for direct numerical simulation of chemically - reactive flows through its high - efficiency computational capabilities and support for complex geometries, which helps optimize the combustion process, reduce emissions, and improve fuel combustion efficiency.