Direct numerical simulation of turbulent flow in a square duct using a Graphics Processing Unit (GPU)

A. F. Shinn, S. P. Vanka, W. W. Hwu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A high-resolution Direct Numerical Simulation (DNS) of incompressible turbulent flow in a square duct has been conducted to demonstrate the capability of a Navier-Stokes solver implemented on a Graphics Processing Unit (GPU). The simulation used a frictional Reynolds number of 360 and a mesh of 26.2 million cells. The code was written for a GPU using CUDA (Compute Unified Device Architecture), a programming paradigm developed by NVIDIA. The GPU has permitted fast, high-resolution calculations on a rather inexpensive computer, opening up new opportunities to efficiently conduct Direct and Large Eddy Simulations of turbulent flows. Implementation details and code performance will be discussed.

Original languageEnglish (US)
Title of host publication40th AIAA Fluid Dynamics Conference
StatePublished - 2010
Event40th AIAA Fluid Dynamics Conference - Chicago, IL, United States
Duration: Jun 28 2010Jul 1 2010

Publication series

Name40th AIAA Fluid Dynamics Conference

Other

Other40th AIAA Fluid Dynamics Conference
Country/TerritoryUnited States
CityChicago, IL
Period6/28/107/1/10

ASJC Scopus subject areas

  • Fluid Flow and Transfer Processes

Fingerprint

Dive into the research topics of 'Direct numerical simulation of turbulent flow in a square duct using a Graphics Processing Unit (GPU)'. Together they form a unique fingerprint.

Cite this