An Adaptive Hybrid (Cartesian-Prism) Grid and Flow Solver for Navier-Stokes Equations

Information

  • NSF Award
  • 9460372
Owner
  • Award Id
    9460372
  • Award Effective Date
    1/1/1995 - 29 years ago
  • Award Expiration Date
    5/31/1996 - 28 years ago
  • Award Amount
    $ 75,000.00
  • Award Instrument
    Standard Grant

An Adaptive Hybrid (Cartesian-Prism) Grid and Flow Solver for Navier-Stokes Equations

This Small Business Innovation Research Phase I project proposes to use overlapped Cartesian grid and Body-Fitted (BF) prism grids for solving viscous flow problems. Given a geometry definition in the form of triangularized surfaces, a BF prism mesh will be generated using the advancing layers technology with grid clustering near body surfaces. This prism mesh is then overlapped with a coarse uniform Cartesian background grid. Initially the Cartesian grid will be refined to have similar resolution near the prism grid. A high-order flow solver based on linear reconstruction and Roe's flux-difference splitting will be developed. Grid adaptation according to the physical nature of the flow field will be carried out to place control volumes where required the most, achieving maximum accuracy with minimum resources. An explicit multi-grid procedure will be employed to speed up the simulation process. The interface between the Cartesian and prism grid will be treated by conserving mass, momentum and energy flux on the outer boundary of the prism grid. The overall algorithm integrates grid generation and flow solver into one complete system to obtain maximum solution accuracy, efficiency and user friendliness. In Phase I, the above methodology will be implemented and demonstrated for 2-D viscous flow problems. Extension to 3-D and moving body problems and implementation in a commercial code will be made in Phase II.

  • Program Officer
    G. Patrick Johnson
  • Min Amd Letter Date
    1/11/1995 - 29 years ago
  • Max Amd Letter Date
    2/28/1996 - 28 years ago
  • ARRA Amount

Institutions

  • Name
    CFD RESEARCH CORPORATION
  • City
    HUNTSVILLE
  • State
    AL
  • Country
    United States
  • Address
    701 McMillian Way NW, Suite D
  • Postal Code
    358062923
  • Phone Number
    2567264800

Investigators

  • First Name
    Z.J.
  • Last Name
    Wang
  • Email Address
    zjw@ku.edu
  • Start Date
    1/11/1995 12:00:00 AM

FOA Information

  • Name
    Other Applications NEC
  • Code
    99
  • Name
    Mathematics
  • Code
    21

Program Element

  • Text
    SMALL BUSINESS PHASE I
  • Code
    5371

Program Reference

  • Text
    INFRASTRUCTURE PROGRAM
  • Code
    1260
  • Text
    ADVANCED SOFTWARE TECH & ALGOR
  • Code
    9216
  • Text
    HIGH PERFORMANCE COMPUTING & COMM