kth_logo.gif

Projekt

Higher-order finite difference methods for DNS/LES in complex geometries

Forskningsområde: Strömningsmekanik
Projektmedlemmar:
Brüger, A.
Gustafsson, B.
Henningson, D.S.H.
Johansson, A.V.J.
Lötstedt, P.
Nilsson, J.
Stålberg, E.
Finansiär: IVS,   NGSSC,   VR 

Projektbeskrivning

The aim of this project is to perform direct numerical simulations (DNS) of turb ulent flow phenomena in complex geometries. A so far unique discretisation method of the incompressible Navier-Stokes equations will be used. It is based on higher-order difference methods which allow for the use of curvilinear, staggered meshes. The compact higher-order stencils make it more efficient and less memory demanding than the standard second-order methods routinely used today. The final goal is to simulate turbulence phenomena on a wing profile geometry. This is an important step towards realistic applications of a DNS method. Another important issue is to simulate turbulent flow separation in an asymmetric diffuser, a case where ongoing experiments is performed at the department of Mechanics. In a further step the DNS code will form a basis for inclusion of large eddy simulation (LES) models. The research will build on the compact higher-order finite difference methods developed in an ongoing cooperation project between the department of Mechanics, KTH and TDB, Uppsala University.

Publikationer relaterade till projektet

ÅrTitel
2002A Compact Higher Order Finite Difference Method for the Incompressible Navier Stokes Equations
Journal of Scientific Computing   (Accepterad)

Interna rapporter relaterade till projektet

ÅrTitelDokumenttyp
2002Higher order methods suitable for direct numerical simulation of flows in complex geometries Licentiatavhandling