Large eddy simulation of a supercritical fuel Jet-In-Cross-flow using GPU acceleration

Kalyana C. Gottiparthi, Ramanan Sankaran, Anthony M. Ruiz, Guilhem Lacaze, Joseph C. Oefelein

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

12 Scopus citations

Abstract

Fuel is routinely injected into combustion chambers at elevated pressures resulting in formation of supercritical fluid flow. The dispersion of fuel and the combustion processes are significantly different in high pressure conditions as the traditional liquid atomization and evaporation are replaced by diffusion-dominated mixing. Modeling supercritical flow is challenging due to the complex thermodynamics and transport properties of the fluids and in most cases high fidelity numerical simulations are prohibitively expensive. However, numerical investigations must incorporate the flow properties appropriate for the regime for accurate description of the physical processes involved. In this paper, we consider real-fluid properties and simulate the turbulent mixing of fuel jet in cross flow at high ambient pressure. In order to make efficient use of the computer resources and reduce the time to solution, without compromising the resolution requirements, the real-fluid property routines are accelerated via Graphics Processing Unit (GPU) computation. The numerical simulations performed at different resolutions suggest that the high fidelity computations with refined mesh are essential to capture the non-linear dynamics of vortex evolution downstream of the fuel jet and signifies the necessity for an accelerated code. The code is accelerated using the Kokkos C++ library and is demonstrated to be about 2 times faster than the non-accelerated version. Considerable computational cost associated with the thermodynamics routines is reduced in the accelerated version. The results also suggest the resource requirements to compute turbulent closures and the need for acceleration of the routines associated with the sub-grid scale closures.

Original languageEnglish
Title of host publication54th AIAA Aerospace Sciences Meeting
PublisherAmerican Institute of Aeronautics and Astronautics Inc, AIAA
ISBN (Print)9781624103933
DOIs
StatePublished - 2016
Event54th AIAA Aerospace Sciences Meeting, 2016 - San Diego, United States
Duration: Jan 4 2016Jan 8 2016

Publication series

Name54th AIAA Aerospace Sciences Meeting
Volume0

Conference

Conference54th AIAA Aerospace Sciences Meeting, 2016
Country/TerritoryUnited States
CitySan Diego
Period01/4/1601/8/16

Funding

This research used resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Office of Science of the U.S. Department of Energy under contract DE-AC05-00OR22725. Support was also provided by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Biosciences. Sandia National Laboratories is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United. States Department of Energy under contract DE-AC04-94-AL85000. References.

Fingerprint

Dive into the research topics of 'Large eddy simulation of a supercritical fuel Jet-In-Cross-flow using GPU acceleration'. Together they form a unique fingerprint.

Cite this