[CFDST]

Hello all CFD's users,

I have a problem while meshing a imported 3d CAD model.The model is for HVAC simulation and the Inlets are at the top and the oulets are at the botom of the model. First I tried to imported is an iges file but when i choose a spatial import options and make all the options that I have read in the gambit(ansys) user guide, nothing happens I clicked accept and there is nothing on the screen. Next I tried to imported it again as an iges but a "native" import options this works but no volumes where recognised and the model is a little bit grey it is not colored well. Next I tried to import it as an acis file and it works best. I have one volume recognised, all the faces are also recognised. So I started to mesh it - first to mesh the volume and next the inlets,outlets and walls. Meshing the volume- the process starts but at the middle of it an error appears on the screen and it says: "That a fatal error has occurred.It is not possible for GAMBIT to continue from this point.Please contact your local fluent officer..... ". The only way to mesh the volume is with stairstep mesh,but this mesh is not complete and there are only 5 or 6 zones meshed in the whole model,so the only option for me was the stairstep despite of the mesh is not complete but it works somehow. Next meshing the outlets,inlets and walls- It works I meshed it with Tri-Pave 0.05. I made all the boundary types-the inlets with velocity inlets,the outlets with outflow and the walls. When I export the mesh and opened it in fluent and click grid-check- Errors appeared- Skipping all the outlets and inlets(not referenced by grid) and when I click display grid-the only grid that appears is the Volume-stairstep grid. My questions are: I have to clean the geometry,because it is too complex for gambit to mesh it?
2.Why fluent can not recognise the outlets and inlets grid despite of the fact they are meshed.
I have upload the cad file and the iges and sat file.Thank in advance.

[CFDST]

http://www.2shared.com/photo/33oHF3r...rt_scale_.html
http://www.2shared.com/file/hHiA5fWD...rt_scale_.html
http://www.2shared.com/file/_aSpXE9R...rt_scale_.html

[-mAx-]

If nobody helps you until friday, I will check it at the office (still in vacation )

[Marion]

Hi,
You can try cutting your volumes into smaller volumes in order to find out where the problem is located. IT is also a good way to control the mesh.
Marion.

[CFDST]

Thank you for your replies!
Marion- I have tried to clean up the geometry and split in some volumes,but it is still complex and the program give errors. I'm thinking to start using Ansys, but I don't have time now to learn it because I need the analysis of the hall in days.My task in the University is for the next week, so I'm running out of time.
Max- I will be very pleased if you can help me after your vacation.Enjoy it and all the best.

[-mAx-]

you have very thin entities in your model.
You will have to do some splits for meshing it with acceptable quality

[CFDST]

Thank you Max, I have made the volume again but I had some problems. I want to make it a single volume from all the faces with stitching and then to try meshing the volume. There is an error message: "Cannot create an Acis body. Is this an open shell?". I saw earlier thread in which you said that you have to click on shaded mode to see where are my missing faces,but I cannot see any holes in the model. Here is the model, I will be very pleased if you can see where are the missing faces for creating the volume.The model was made in Autocad 3d from surfaces and regions.

[CFDST]

http://www.2shared.com/file/cRB4WANR...ri_opit-1.html

[Far]

there are many small steps present in the geometry. Are these important?

[-mAx-]

Originally Posted by CFDST

Thank you Max, I have made the volume again but I had some problems. I want to make it a single volume from all the faces with stitching and then to try meshing the volume. There is an error message: "Cannot create an Acis body. Is this an open shell?". I saw earlier thread in which you said that you have to click on shaded mode to see where are my missing faces,but I cannot see any holes in the model. Here is the model, I will be very pleased if you can see where are the missing faces for creating the volume.The model was made in Autocad 3d from surfaces and regions.

From .sat geometry you get a closed volume

[CFDST]

Originally Posted by -mAx-

From .sat geometry you get a closed volume

You made a closed volume from the .sat file ?

[-mAx-]

the import generates the volume automatically

[CFDST]

Originally Posted by -mAx-

the import generates the volume automatically

http://www.2shared.com/file/cRB4WANR...ri_opit-1.html but when I import this file with file/import/acis gambit only recognise 114 faces but no volumes,and when I try to createa volume from wireframe or faces,there is a problem.

[-mAx-]

With your last sat file, yes I also have 114 faces.
But with your first sat file, I get a volume (Gambit 2.4.6)

Originally Posted by CFDST

http://www.2shared.com/photo/33oHF3r...rt_scale_.html
http://www.2shared.com/file/hHiA5fWD...rt_scale_.html
http://www.2shared.com/file/_aSpXE9R...rt_scale_.html

[CFDST]

Originally Posted by -mAx-

With your last sat file, yes I also have 114 faces.
But with your first sat file, I get a volume (Gambit 2.4.6)

Thank you for the replies. I made a new geometry I have meshed the volume,but now the problems are in Fluent. I meshed the volume with Tet/Hybrid TGrid 0.6,all the air terminals with Tri Pave 0.2, The walls connected with the Air terminals with Tri Pave 0.6 and it works.I set the boundary conditions to the air terminals on top and those on the wall to "velocity inlet" and the floor air terminals to "outflow",all the walls to boundary condition" wall ", the volume to fluid. I have exported the mesh into fluent.But there are problems with the iterations and there is a mistake "outflows reversed flow in 74 faces on outflow 8", I have tried to make the outflow terminals from "Outflow" boundary condition to "Outflow-Pressure" and the erorr message is the same.
My steps in fluent are:
1.File/Read/Case-Hall.msh
2.Grid/Check
3.Display/Grid
4.Define/models/solver-Pressure based and Implicit
5.Define/models/viscous- K- epsilon
6.Define/models/energy-energy equation
7.Define/materials-air
8.Define Boundary Conditions- Velocity inlets Top- 5 m/s and 306 K
Velocity inlets wall-7 m/s 293 K
Walls- Temperature 293 K
Outlets-Flow rate weigthing - 1
Hall- Volume-fluid
9.Solve-Controls-Solutions
Flow and Turbulunce , Second order upwind
10.Solve/monitor/residiuals - plot
11.Solve/Initialize/init
12.file/write/case
13.Solve/Iterate


1.Is this Outflow error from the reason that we haven't put all the air terminals in the hall? because we want to try and then to draw them all,because they are about 350 air t.,
2.May be somewhere in the boundary conditions we are missing something?
3.May be in the mesh it has to be with more precise?
Here are some pictures and the .msh and .dbs files.I will be very pleased if somebody helps me.

[CFDST]

http://www.2shared.com/photo/LXPhn26I/Before-mesh.html
http://www.2shared.com/file/-xRALM_O/Hall.html
http://www.2shared.com/file/B5XooOmd/Hall.html
http://www.2shared.com/file/OPnrkpnY/Hall.html
http://www.2shared.com/file/5rk8u-NK/Hall.html
http://www.2shared.com/file/ZOsPAqdg/Hall.html
http://www.2shared.com/photo/l9zSoce...ations-20.html
http://www.2shared.com/photo/6safZpA...ations-50.html
http://www.2shared.com/photo/YJhsNDDz/Mesh-faces.html
http://www.2shared.com/photo/iaGvVxE7/Pathlines.html
http://www.2shared.com/photo/1FFnMtm5/Volume-mesh.html

[-mAx-]

reversed flow isn't a huge problem.
You may let iterate, and it can disappear.
You could also use pressure-outlet instead of outflow.
Important is to check the convergence. You can also monitor a value like the pressure at your inlet

[CFDST]

Originally Posted by -mAx-

reversed flow isn't a huge problem.
You may let iterate, and it can disappear.
You could also use pressure-outlet instead of outflow.
Important is to check the convergence. You can also monitor a value like the pressure at your inlet

I have made 200 iterations and it ends at the 198 iteration with erorr " Floating point error: invalid number"
Erorr object: (). I checked the convergence
criterion and made it 0.01. I change the outflows to presure outlets and the error is the same. Is the problem from the continuity line, or may be if I put all the air terminals in the hall may be it will go down or ?

[-mAx-]

*any errors or warning in the check Mesh?
*is mass conservation respected (massflow in = mass flow out)?
*before iterations break, display velocity contour, and check if it well flows
*check also if you don't have other boundaries as yours, which appear (wall-shadoww, or whatever)

[Aadhavan]

Hi every one,
I am facing strange problem in Gambit, while creating the geometry in gambit. The sudden crash is happening often middle of my work, I dont know what is the reason.

I have attached the error message please have a look and give me some idea what to do?

Thanks,
Aadhavan