Direct numerical simulations of reacting flows with detailed chemistry using many-core/GPU acceleration

Pérez, Francisco E. Hernández; Mukhadiyev, Nurzhan; Xu, Xiahong; Sow, Aliou; Lee, Bok Jik; Sankaran, Ramanan; Im, Hong G.

doi:10.1016/j.compfluid.2018.03.074

Cited by 57 publications

(19 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The KAUST Adaptive Reacting Flow Solver (KARFS) (Hernández Pérez et al 2018;Desai et al 2018) is used to solve the fully compressible Navier-Stokes, species, and energy equations for gaseous mixtures. The diffusive terms are discretized using an eighth-order finite-difference scheme, while the convective terms are discretized using a seventh-order mapped weighted essentially non-oscillatory scheme along with local Lax-Friedrich flux splitting to capture shocks and detonation waves.…”

Section: Numerical Methods and Initial Conditionsmentioning

confidence: 99%

Effects of Turbulence and Temperature Fluctuations on Knock Development in an Ethanol/Air Mixture

Luong

Pérez

Sankaran

et al. 2020

Flow Turbulence Combust

Self Cite

View full text Add to dashboard Cite

The effects of turbulence on knock development and intensity for a thermally inhomogeneous stoichiometric ethanol/air mixture at a representative end-gas autoignition condition in internal combustion engines are investigated using direct numerical simulations with a skeletal reaction mechanism. Two-and three-dimensional simulations are performed by varying the most energetic length scale of temperature, l T , and its relative ratio with the most energetic length scale of turbulence, l T ∕l e , together with two different levels of the turbulent velocity fluctuation, u ′ . It is found that l T /l e and the ratio of ignition delay time to eddy-turnover time, ig ∕ t , are the key parameters that control the detonation development. An increase in either l T or l e enhances the detonation propensity by allowing a longer run-up distance for the detonation development. The characteristic length scale of the temperature field, l T , is significantly modified by high turbulence intensity achieved by a large l e and u ′ . The intense turbulence mixing effectively distributes the initial temperature field to broader scales to support the developing detonation waves, thereby increasing the likelihood of the detonation formation. On the contrary, high turbulence intensity with a short mixing time scale, achieved by a small l e and a large u ′ , reduces the super-knock intensity attributed to the finer broken-up structures of detonation waves. Either ig ∕ t less than unity or l e = l T even with a large u ′ is found to have no significant effect on super-knock mitigation. Finally, high turbulent intensity may induce high-pressure spikes comparable to the von Neumann spike. Increased temperature and pressure by combustion heating, noticeably after the peak of heat release rate, significantly enhance the collision and interaction of multiple emerging autoignition fronts near the ending combustion process, resulting in localized high-pressure spikes.

show abstract

Section: Numerical Methods and Initial Conditionsmentioning

confidence: 99%

Effects of Turbulence and Temperature Fluctuations on Knock Development in an Ethanol/Air Mixture

Luong

Pérez

Sankaran

et al. 2020

Flow Turbulence Combust

Self Cite

View full text Add to dashboard Cite

show abstract

“…The KAUST Adaptive Reacting Flow Solver (KARFS) [2,3,41,49] is used to solves the fully compressible Navier-Stokes, species and energy equations for gaseous mixtures. An eighth-order finite-difference scheme was employed for the spatial discretization of the diffusive terms.…”

Section: Numerical Methods and Initial Conditionsmentioning

confidence: 99%

Knock propensity in a thermally inhomogeneous DME/air mixture: a DNS study

Luong

2022

AIAA SCITECH 2022 Forum

Self Cite

View full text Add to dashboard Cite

Superknock propensity in a stoichiometric dimethyl-ether (DME)/air mixture with temperature inhomogeneities under realistic IC engine conditions is investigated using two-dimensional direct numerical simulations (DNS). The developing detonation regime at different conditions is identified by varying the initial mean temperature lying in the low-, intermediate-, and high-temperature chemistry regimes, the level of temperature fluctuations, and its characteristic length scale. We found that the cool flame from the first-stage ignition induces synergistic effects on promoting knock tendency. First, it significantly decreases a minimum run-up distance requirement for developing detonation due to the low-temperature chemistry. Second, analyzing the temporal evolution of the spatial distribution of ignition delay field reveals that the heat release rate from the first-stage ignition effectively modifies the initial field of the ignition delay time, thereby shifting the mixture towards the developing detonation regime. The interaction of multiple ignition kernels is also found to play an important role in enhancing the onset of detonation.

show abstract

“…In general, GPUs operate differently from CPUs, requiring the algorithmic implementations for expensive routines, such as the aforementioned source term computations, to be altered in order to leverage their specific hardware architecture to extract as much computational gain as possible. To this end, approaches for GPU-offloading for chemical kinetics have been explored in detail in recent years to success [20][21][22], and their implementation into high-fidelity parallel solvers has also been demonstrated [23]. These approaches traditionally rely on translation of the exact equations for kinetics and time-integration methods into the GPU environment.…”

Section: Introductionmentioning

confidence: 99%

“…These approaches traditionally rely on translation of the exact equations for kinetics and time-integration methods into the GPU environment. However, much of the literature in the context of kinetics offloading has been geared towards either the time-integration aspect of the kinetics problem [19,22] or the role of the kinetics offloading in the context of a full reacting flow solver [23], and not on the computationally intensive evaluation of the source terms in isolation.…”

Section: Introductionmentioning

confidence: 99%

A Neural Network-Inspired Matrix Formulation of Chemical Kinetics for Acceleration on GPUs

Barwey

Raman

2021

Energies

View full text Add to dashboard Cite

High-fidelity simulations of turbulent flames are computationally expensive when using detailed chemical kinetics. For practical fuels and flow configurations, chemical kinetics can account for the vast majority of the computational time due to the highly non-linear nature of multi-step chemistry mechanisms and the inherent stiffness of combustion chemistry. While reducing this cost has been a key focus area in combustion modeling, the recent growth in graphics processing units (GPUs) that offer very fast arithmetic processing, combined with the development of highly optimized libraries for artificial neural networks used in machine learning, provides a unique pathway for acceleration. The goal of this paper is to recast Arrhenius kinetics as a neural network using matrix-based formulations. Unlike ANNs that rely on data, this formulation does not require training and exactly represents the chemistry mechanism. More specifically, connections between the exact matrix equations for kinetics and traditional artificial neural network layers are used to enable the usage of GPU-optimized linear algebra libraries without the need for modeling. Regarding GPU performance, speedup and saturation behaviors are assessed for several chemical mechanisms of varying complexity. The performance analysis is based on trends for absolute compute times and throughput for the various arithmetic operations encountered during the source term computation. The goals are ultimately to provide insights into how the source term calculations scale with the reaction mechanism complexity, which types of reactions benefit the GPU formulations most, and how to exploit the matrix-based formulations to provide optimal speedup for large mechanisms by using sparsity properties. Overall, the GPU performance for the species source term evaluations reveals many informative trends with regards to the effect of cell number on device saturation and speedup. Most importantly, it is shown that the matrix-based method enables highly efficient GPU performance across the board, achieving near-peak performance in saturated regimes.

show abstract

Direct numerical simulations of reacting flows with detailed chemistry using many-core/GPU acceleration

Cited by 57 publications

References 15 publications

Effects of Turbulence and Temperature Fluctuations on Knock Development in an Ethanol/Air Mixture

Effects of Turbulence and Temperature Fluctuations on Knock Development in an Ethanol/Air Mixture

Knock propensity in a thermally inhomogeneous DME/air mixture: a DNS study

A Neural Network-Inspired Matrix Formulation of Chemical Kinetics for Acceleration on GPUs

Contact Info

Product

Resources

About