Hi, I am working with block structured grid generation for fluid-structure interaction problems, but I have just started. I have been searching for some good bibliografic references, but I have had no success. Can anyone help me? My first idea to generate the multi block scheme is to use advancing front techniques in order to obtain an O-type grid around the structure. Is it a good idea?

Márcio Ricardo Pivello Doctoral Student School of Mechanical Engineering Federal University of Uberlândia - MG - Brazil

Do you really attempt to progrm your own hex mesher ?

If possible try a commercial mesher like ICEM Hexa. Cheap licenses for universities should be available. Look at www.icemcfd.com

I guess ICEM would be a very good starting point to me, to learn more about block structured grids. But my Doctoral research is about automating the block topology generation, because it seems that the comercial softwares avaiable are not completely developped in this specific item. After a bibliografic research, I found some schemes based on geometric similarities or primitives which generate some types of pre-defined topologies (mainly developed by Dannenhoffer), but I want to TRY to fully automate this task.

I read a very interesting paper relating CFD and this kind of grid, "Elements of computational fluid dynamics on block structured grids using implicit solvers", by K. J. Badcock and M. A. Woodgate. The authors talk about the state-of-the-art of this kind of grid, and what you can see is that it is a very hard task to automate the block topology generation, and that´s why I decided to do my research in this area. My idea is to use some kind of advancing front technique to generate offsets of the original structural geometry, and the space between two offsets would be divided into blocks.

I still think you should contact ICEM because they are working a lot on automatic topology generation at the moment. The way is something like automatically generating a rough tet mesh and get a proposal for a topology from that.

Regards Helge

Yes, you´re right. Besides that, I can go to the next step of my research, which is solve fluid-structure problems with this kind of grid. Do you know how to apply boundary conditions in such grids? I read something about using global unstructured grid connectivity, but I think that by doing this we´ll lose the main advantage of this kind of grid, that is the local structured connectivity. But how to apply the BCs to the internal faces of the blocks?

Thanks

Márcio

No, but that is not my topic

In the block structured mesh, the information on the internal faces of the blocks are exchanged with each other by MPL. Please search the concerning literatures about the "dummy cell" and "Virtual cell". They are generated around the block for communication