
[Sponsors] 
April 25, 2009, 19:53 
Meshing space capsule

#1 
New Member
Talvinder Singh
Join Date: Apr 2009
Posts: 5
Rep Power: 9 
I am a beginner to gambit and CFD. I have to mesh the geometry of a space capsule whose wire frame has been given to me. I have to find the surface properties at hypersonic conditions. I will like to give the steps I followed, because I am not sure if I on the right track. The wire frame is given below:
I created the geometry in the following steps: 1) Creating an edge by joining the vertices at the nose and the corners. 2) Giving a 360 degree rotation to the edge thus formed, to get the following geometry 3) Next I meshed the edges(the 3 circles) and then the face to get the following The domain is: I am facing problems in the direction of Boundary layer. The Boundary layer should be outside the space capsule since i need to find the surface results, but the direction of Boundary layer is always inside. I reffered to the gambit modeling guide and searched internet for relevant examples, but failed. The direction at the nose is as shown: and if i try creating boundary layer of a face i get this: I am unable to find a solution to this. Any kind of help is most appreciated. Thanks 

April 26, 2009, 02:43 
Meshing Space Capsule

#2 
New Member
Adrian
Join Date: Apr 2009
Posts: 2
Rep Power: 0 
I have not used GAMBIT before, so my advice is rather generic in nature but should be of assistance. It seems you have blocked the internal region of your model rather than the flow domain. My advice in this case would be as follows:
1. Your flow type appears axisymmetric. If you developed a 2D axisymmetric grid instead, the results would be comparable to the 3D case and it would significantly reduce both the meshing and compute time. 2. In order to accurately capture the boundary layer it would be best to use a fully structured mesh. An example of this approach can be seen at http://fun3d.larc.nasa.gov/mesh_tn.png The axisymmetric approach will yield good results and use significantly less computational resources than a fully three dimensional mesh. If you still want a 3D run than just perform a rotation of the 2D axisymmetric case and this should suffice. Hope this helps 

April 26, 2009, 09:54 

#3 
New Member
Talvinder Singh
Join Date: Apr 2009
Posts: 5
Rep Power: 9 
Yes the geometry is axis symmetric but I am not sure if a 2D simulation will generate the results i need. Correct me if I am wrong, wont 3D effects matter a lot in this case since its a hypersonic simulation.
I will try the 2d simulation as well, but I am not able to figure out the boundary layer diirection solution as yet Thanks in advance 

April 26, 2009, 10:37 

#4 
New Member
Talvinder Singh
Join Date: Apr 2009
Posts: 5
Rep Power: 9 
Yes the geometry is axis symmetric but I am not sure if a 2D simulation will generate the results i need. Correct me if I am wrong, wont 3D effects matter a lot in this case since its a hypersonic simulation.
I will try the 2d simulation as well, but I am not able to figure out the boundary layer diirection solution as yet Thanks in advance 

April 26, 2009, 11:03 

#5 
New Member
Talvinder Singh
Join Date: Apr 2009
Posts: 5
Rep Power: 9 
Yes the geometry is axis symmetric but I am not sure if a 2D simulation will generate the results i need. Correct me if I am wrong, wont 3D effects matter a lot in this case since its a hypersonic simulation.
I will try the 2d simulation as well, but I am not able to figure out the boundary layer diirection solution as yet Thanks in advance 

April 26, 2009, 11:20 

#6 
New Member
Talvinder Singh
Join Date: Apr 2009
Posts: 5
Rep Power: 9 
Yes the geometry is axis symmetric but I am not sure if a 2D simulation will generate the results i need. Correct me if I am wrong, wont 3D effects matter a lot in this case since its a hypersonic simulation.
I will try the 2d simulation as well, but I am not able to figure out the boundary layer diirection solution as yet Thanks in advance 

April 26, 2009, 20:03 
Axisymmetric Mesh

#7 
New Member
Adrian
Join Date: Apr 2009
Posts: 2
Rep Power: 0 
Provided your freestream flow field is also symmetric about the same axis as the model a 2D axisymmetric you will be able to capture most of the flow features. I assume you have access to a cluster facility of some sort to perform your calculations as you are looking at least for 23 million cells for a three dimensional run.
A freely accessable report that may be of assistance in your approach is available at http://www.rtechengineering.com/docs/AA3200829.pdf. This document is looking at comparison with an experiment in which a sting is used to suspend the model. This is not really a valid use of the 2D axisymmetric model but is just a comparison. I think you will find the the primary 3D effects are the 3D releiving effect as the flow expands radially. Once you hit the base flow region of course you are into another region of currect state of the art research. If you are after surface pressure or drag measurements than I would recommend that the 2D axisymetric model will yield good results. If you still do the 3D model than it would be interesting to compare the results from the two models and see how close they are. Choice of turbulence model will also have a significant effect on the results. 

April 27, 2009, 02:28 

#8 
New Member
Daniel Langmayr
Join Date: Mar 2009
Posts: 15
Rep Power: 9 
It seems that you have a volume inside the body which should not be there. Due to your meshing approach (no blocking), you should proabably have just one volume inside your model. So you can easily check in gambit if this is the case.
In Gambit just go to geometry, volumes and summarize volumes. There you can select volumes and you should just be able to select one. In addition, from the figure it looks as if you have two edges at the circle with the largest zvalue (i may be mistaken here). In Gambit one has to be very careful to avoid those double edges and surfaces. Good luck! 

April 27, 2009, 06:07 

#9 
Senior Member
Ahmed
Join Date: Mar 2009
Location: NY
Posts: 251
Rep Power: 10 
I have a question for you, Did you subtract the capsul volume from the flow domain volume?


February 11, 2010, 10:37 

#10 
New Member
Margarita DUFRESNE
Join Date: Jan 2010
Posts: 13
Rep Power: 8 
Hi,
Don't use Fluent for hypersonic flow. The special code are TINA, FASTRAN KR, MDZ. 

Tags 
boundary laye, fluent, gambit, mesh, space capsule 
Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Best Meshing scheme for Cylinder  Nutrex  Main CFD Forum  4  July 29, 2008 11:03 
Meshing locks workbench window.  andy2o  CFX  0  February 1, 2008 06:01 
Meshing  Kevin  FLUENT  1  June 28, 2004 04:04 
Singularity of grid?Volume meshing vs face meshing  Ken  Main CFD Forum  0  September 4, 2003 11:09 
Volume Meshing & Face Meshing? singularity of grid  ken  FLUENT  0  September 4, 2003 11:08 