An overset-grid method for 3D unsteady incompressible flows

H. S. Tang, S. Casey Jones, Fotis Sotiropoulos

Research output: Contribution to journalArticlepeer-review

104 Scopus citations

Abstract

A composite-grid numerical method is developed for simulating unsteady, three-dimensional (3D), incompressible flows in complex geometries. The governing equations are solved using a second-order accurate, finite-volume method based on the dual time-stepping artificial compressibility approach. Overset grids are employed to discretize arbitrarily complex geometries, and a new interface algorithm is developed to facilitate communication between neighboring grids. The algorithm is inspired by the necessary and sufficient conditions for satisfying global mass conservation in a composite domain and is simple to implement in 3D. Numerical experiments show that the new interpolation scheme is superior to straightforward, trilinear interpolation of all flow variables as it minimizes non-physical spurious oscillations in the overlap region, is less sensitive to grid refinement, and greatly enhances the computational efficiency of the iterative algorithm. The advantages of the new method are especially pronounced when adjacent overset subdomains are discretized with different spatial resolutions. The potential of the method as a powerful technique for simulating complex engineering flows is demonstrated by applying it to calculate vortex shedding from a circular cylinder mounted between two endplates and flow in a rectangular channel with two wall-mounted obstacles.

Original languageEnglish (US)
Pages (from-to)567-600
Number of pages34
JournalJournal of Computational Physics
Volume191
Issue number2
DOIs
StatePublished - Nov 1 2003

Keywords

  • Artificial compressibility method
  • Chimera grids
  • Dual time-stepping
  • Grid interfaces
  • Incompressible Navier-Stokes equations
  • Multigrid acceleration
  • Overset grids

Fingerprint Dive into the research topics of 'An overset-grid method for 3D unsteady incompressible flows'. Together they form a unique fingerprint.

Cite this