What is sliding mesh? Is it structured or unstructured? Can it be used as structured?

Hi,

Your question sounds a little bit strange... As far as I know, a sliding mesh is a mesh which is in motion relatively to something else... usually to another mesh which is fixed, or more generally to other meshes which moves independently to each others.

The use of sliding meshes allows to deal with engineering problems such as interaction of several bodies moving relatively to each others within a fluid flow.

As a consequence, this meshes can be either structured or not. Moreover, structured and unstructured meshes can be encountered at the same time in this context.

Hoping it will answer your question...

Hi,

Thanks for answer my question. There must be relative movement between sliding mesh and the others. Are there any algorithms? Can you introduce some papers about it. I want to use structured sliding mesh to solve the flow field with relative movement bodies.

Dear Wang Tong,

There is an algorithmus named as CHIMERA grid approach to treat the discontinuity of mesh interfaces. I have used it several years for moving grids and find it satisfactory. I recommand you some papers:

Terry L. Holst: Numerical Solution of the Full Potential Equation Using a Chimera Grid Approach, AIAA Journal, Vol. 35, No. 9, September 1997

Kai-Hsiung Kao, Meng-Sing Liou: Grid Adaption Using Chimera Composite Overlapping Meshes, AIAA Journal, Vol. 32, No. 5, May 1994

There is a paper treating additional terms in transport equations because of the motion of grids:

Joseph L. Steger: Implicit Finite-Difference Simulation of Flow about Arbitrary Two-Dimensional Geometries, Vol. 16, No. 7, July 1978, AIAA Journal

Beat regards

X. Ye

Hi again,

I am sorry, but I don't know any paper concerning numerical methods for solving Navier-Stokes equations on sliding grids WITHOUT overlap. As for the Chimera technique (i.e. with overlapping grids) see the answer of Mr. Ye. Some of the features of the technique might be relevant for non-overlapping grids.

Yes, there are no papers to be found about the application of Chiemera approach for not overlaping grids. But in the view of CHIMERA approach, sliding mesh is only a special case of overlaping mesh. If two grids are not overlaping, you can do the smae as by overlaping: exchanging information between two grids through Chimera interpolation.

Best regards

Xiangyang Ye

Sorry, I forgot to write: If two grids are not overlapping but only contacting with each other, their dummy cells which are around each grid and specially for enchange of information between two grids, are overlapping with other grids, however.

Best regards

Xiangyang Ye

Thank you for your help.

I found that many algorithms for unsteady flow with relatively moving bodies are explicit, why?

There is no additional equations for pressure in incompressible flow. Is there any implicit algorithms for unsteady flow. I want to use the algorithms like SIMPLE to calculate it. Is it feasible?

Best wishes,

Wang Tong

To your first question: The reason may be that for sliding grids you have to applied the multiblock technique. For this type of grid structure the explicit method is easier.

To your second question, I am sorry, I can not give you answer, because I use the explicit method and do not have much expierence of implicit method.

Best regards

X. Ye

Thank you for so many useful messages. Let me try.

Thanks gain.

Wang Tong

The CFD Research Corporation in Huntsville Al have a compressible code with a chimera capability. It is used for modelling ejector-seats, canopy jetison, general stores release problems and any relative motion dynamics. It is fully coupled to a 6 DOF model. See their web site at www.cfdrc.com.

Thanks a lot.

a sliding grid is used when you have relative movement between several computational domains.

The Chimera method (Overlapping grids ) was mentioned in previous messages but this method is not a priori conservative. The solutions on the two grid levels are passed to each other in the overlapping region. However, for reasons of complexity, it is very difficult to ensure that the total sum of the metrics of each cell used for solution passing is equal on both sides (or grid levels)(the fluxes may then not be conserved). This has a negligible effect if the grids are quite fine and that no significant gradients of flow quantity occur in the overlapping region. Else, this non-conservative feature can be solved by resorting to complex interpolation methods which ensure global conservation.

However, in axial turbomachinery applications where the sliding interface is usually just a plane located at mid-distance between stationary and rotary blade rows, it is better to keep track of the correspondence between cell faces on both sides, so as to ensure conservation of fluxes.

Stephane

Thanks. All of you are so kind.

I am calculating the flow field in centrifugal fan. It has impellers and diffusers. I want to calculate it as a whole flied, not as different parts. I think the unstructured grid is complicated. If I can use the structured grid, implicit method and some algorithms to deal with the sliding interface, the problem will be simple.

Wang Tong