Block-structured adaptive mesh refinement (AMR) provides the basis for the temporal and spatial discretization strategy for a number of Exascale Computing Project applications in the areas of accelerator design, additive manufacturing, astrophysics, combustion, cosmology, multiphase flow, and wind plant modeling. AMReX is a software framework that provides a unified infrastructure with the functionality needed for these and other AMR applications to be able to effectively and efficiently utilize machines from laptops to exascale architectures. AMR reduces the computational cost and memory footprint compared to a uniform mesh while preserving accurate descriptions of different physical processes in complex multiphysics algorithms. AMReX supports algorithms that solve systems of partial differential equations in simple or complex geometries and those that use particles and/or particle–mesh operations to represent component physical processes. In this article, we will discuss the core elements of the AMReX framework such as data containers and iterators as well as several specialized operations to meet the needs of the application projects. In addition, we will highlight the strategy that the AMReX team is pursuing to achieve highly performant code across a range of accelerator-based architectures for a variety of different applications.
Three-dimensional numerical simulations of canonical statistically steady, statistically planar turbulent flames have been used in an attempt to produce distributed burning in lean methane and hydrogen flames. Dilatation across the flame means that extremely large Karlovitz numbers are required; even at the extreme levels of turbulence studied (up to a Karlovitz number of 8767) distributed burning was only achieved in the hydrogen case. In this case, turbulence was found to broaden the reaction zone visually by around an order of magnitude, and thermodiffusive effects (typically present for lean hydrogen flames) were not observed. In the preheat zone, the species compositions differ considerably from those of one-dimensional flames based a number of different transport models (mixture averaged, unity Lewis number and a turbulent eddy viscosity model). The behaviour is a characteristic of turbulence dominating non-unity Lewis number species transport, and the distinct limit is again attributed to dilatation and its effect on the turbulence. Peak local reaction rates are found to be lower in the distributed case than in the lower Karlovitz cases but higher than in the laminar flame, which is attributed to effects that arise from the modified fuel-temperature distribution that results from turbulent mixing dominating low Lewis number thermodiffusive effects. Finally, approaches to achieve distributed burning at realisable conditions are discussed; factors that increase the likelihood of realising distributed burning are higher pressure, lower equivalence ratio, higher Lewis number and lower reactant temperature.
6 September 2022 in The International Journal of High Performance Computing Applications
Reacting flow simulations for combustion applications require extensive computing capabilities. Leveraging the AMReX library, the Pele suite of combustion simulation tools targets the largest supercomputers available and future exascale machines. We introduce PeleC, the compressible solver in the Pele suite, and detail its capabilities, including complex geometry representation, chemistry integration, and discretization. We present a comparison of development efforts using both OpenACC and AMReX’s C++ performance portability framework for execution on multiple GPU architectures. We discuss relevant details that have allowed PeleC to achieve high performance and scalability. PeleC’s performance characteristics are measured through relevant simulations on multiple supercomputers. The success of PeleC’s design for exascale is exhibited through demonstration of a 160 billion cell simulation and weak scaling onto 100% of Summit, an NVIDIA-based GPU supercomputer at Oak Ridge National Laboratory. Our results provide confidence that PeleC will enable future combustion science simulations with unprecedented fidelity.
1 January 2014 in Communications in Computational Physics
Efficient Variable-Coefficient Finite-Volume Stokes Solvers - Volume 16 Issue 5 - Mingchao Cai, Andy Nonaka, John B. Bell, Boyce E. Griffith, Aleksandar Donev
8 July 2005 in Proceedings of the National Academy of Sciences
We present a three-dimensional, time-dependent simulation of a laboratory-scale rod-stabilized premixed turbulent V-flame. The experimental parameters correspond to a turbulent Reynolds number, Re t = 40, and to a Damköhler number,
D
a = 6. The simulations are performed using an adaptive time-dependent low-Mach-number model with detailed chemical kinetics and a mixture model for differential species diffusion. The algorithm is based on a second-order projection formulation and does not require an explicit subgrid model for turbulence or turbulence/chemistry interaction. Adaptive mesh refinement is used to dynamically resolve the flame and turbulent structures. Here, we briefly discuss the numerical procedure and present detailed comparisons with experimental measurements showing that the computation is able to accurately capture the basic flame morphology and associated mean velocity field. Finally, we discuss key issues that arise in performing these types of simulations and the implications of these issues for using computation to form a bridge between turbulent flame experiments and basic combustion chemistry.