Adaptive mesh refinement sensors for vortex flow simulations
2004 (English)Conference paper (Refereed)
This paper presents results obtained with different sensors used to automatically refine a CFD mesh for vortex-dominated flow applications. The paper focuses on the sensors used to locate a vortex in a flow field in order to refine the mesh in its vicinity. Several approaches are evaluated: the first type of methods is based on flow quantities that are commonly used to visualize vortices in a flow field e.g. ratio of total pressures and creation of entropy (in this case all the cells with values above or below a certain threshold are refined); the second approach identifies the vortex core in order to refine the cells close to it. The identification in this latter case is done by analysing the eigenvalues of the tensor of the velocity gradients. The three sensors are compared through the h-refinement of a mesh around a 70° -swept delta wing. The calculations, for inviscid conditions at an angle of attack of 20° and a Mach number of 0.2, are performed with the unstructured flow solver EDGE. All sensors capture the vortex very well and lead to improved resolution in the vortex. This produces a solution with a stronger and sharper vortex as well as an improved pressure distribution on the wing. The sensor based on the eigenvalue analysis of the velocity-gradient tensor concentrates on the vortex core and the shear layer coming from the leading-edge. An analysis of the influence of the threshold chosen to start the refinement reveals that this latter sensor is more local than the two others and less sensitive to the choice of the threshold, which makes it the most promising candidate for a fully automatic refinement process without user intervention.
Place, publisher, year, edition, pages
2004. 1-20 p.
Adaptation, Delta wing, Eigenvalue analysis, Mesh refinement, Velocity gradient, Vortex
Engineering and Technology
IdentifiersURN: urn:nbn:se:kth:diva-144461ScopusID: 2-s2.0-84893512455OAI: oai:DiVA.org:kth-144461DiVA: diva2:713545
European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2004; Jyväskylä; Finland; 24-28 July 2004
QC 201412102014-04-232014-04-232014-12-10Bibliographically approved