Current status of the coffe solver within hpcmp create™-av kestrel

Kevin R. Holst, Ryan S. Glasby, J. Taylor Erwin, Douglas L. Stefanski, Daniel Prosser, W. Kyle Anderson, Stephen L. Wood

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

8 Scopus citations

Abstract

Recent development for United States Department of Defense HPCMP CREATE™-AV Kestrel component COFFE has led to significant speed, robustness, and work-flow improvements. Significant speed-up is achieved through vectorization pragmas and compiler options; robustness is improved through the adoption of a new pressure-based shock switch and slight modification to the negative side of the SA-neg turbulence model; and post-processing work-flow is enhanced with the addition of dimensional outputs and Kestrel taps. These improvements are demonstrated through efficient and accurate simulations across a variety of geometries and flow conditions, including a high-speed cylinder, a double cone, the Common Research Model, and the JAXA Standard Model. Furthermore, linear CPU strong scaling is demonstrated for up to 7,128 cores for a grid made up of 49,073,713 nodes.

Original languageEnglish
Title of host publicationAIAA Scitech 2020 Forum
PublisherAmerican Institute of Aeronautics and Astronautics Inc, AIAA
ISBN (Print)9781624105951
DOIs
StatePublished - 2020
Externally publishedYes
EventAIAA Scitech Forum, 2020 - Orlando, United States
Duration: Jan 6 2020Jan 10 2020

Publication series

NameAIAA Scitech 2020 Forum
Volume1 PartF

Conference

ConferenceAIAA Scitech Forum, 2020
Country/TerritoryUnited States
CityOrlando
Period01/6/2001/10/20

Funding

Material presented in this paper is a product of the Air Vehicles element of the HPCMP CREATE™ program. The authors would like to thank the HPCMP for the many hours of computing time that are so critical for the successful development and validation of COFFE. The authors would also like to thank their colleagues on the CREATE senior management and Kestrel teams for their support and hard work in producing the exceptional HPCMP CREATE™-AV Kestrel product. Specifically, a special thanks goes to Dr. Robert Meakin and Dr. Nathan Hariharan as well as Dr. Scott Morton, Dr. David McDaniel, Dr. Robert Nichols, and Mr. Todd Tuckey. Furthermore, the authors would like to thank Dr. James Coder, Dr. Steven Allmaras, Dr. David Williams, Dr. Ryan Bond, Elizabeth Lee-rausch, Dr. Philippe Spalart, Dr. Timothy Eymann, Dr. James Newman, Dr. Steve Karman, Dr. Michael Brazell, Mr. Kevin Doetsch, and Dr. Greg Denny for providing meaningful technical guidance and relevant test cases for COFFE.

FundersFunder number
HPCMP

    Fingerprint

    Dive into the research topics of 'Current status of the coffe solver within hpcmp create™-av kestrel'. Together they form a unique fingerprint.

    Cite this