hi, i am trying to mash my module for two weeks, but it dont work, i try the tgrid scheme and i cant do it. the only option is stairsteps. what should i do? i realy need help my dead line is realy soon

thanks

Describe your geometry and the mesh you're looking for (Tet or Hex, Structured or Unstructured) and what you mean by "it don't work"... what exactly is the error or problem you are running into?

Without any other info, I would say use Gambit and create a Tet mesh using sizing functions to control the mesh.

Good luck, Jason

from your mail i think you do not have much i dea about meshing. i wish to say few things (will be useful):

1. this is basic rule and applies to almost all the software you are going to use. " in any engineering software, on the GUI, things are given in logical order in which they are needed"

so this means if you open fluent, from left to right things are given as you are going to use them, as first reading , then grid manipulations then defining the case then solving etc.

the same rule applies in case of gambit too. things are given in the order you are going to need them. So in another words, you knw the software if you know what order you are going to need them. (this is important point i might coem to this again)

2. visit the software (in this case gambit), from left to right to see what are the fuction you do not know. in other words, know what these things exactly do, or make sure of their functionality; example there is substarct with retain and there is split. in this case the major difference is connectivity of faces, in split they will be connected.

i would advise you to look at the behaviour of faces connectivity when you perform any operation (this is major cause of problems with gambit users) for example you can not split a real face with virtual face and get two real faces.

3. now you know what they do, so next thign is do some simpler things which are similar to your actual problem.

4. think and devise a plan how you are going to mesh the geometry based on the knowledge you have gathered. this is important, lot of people just start meshing it, and finally found themselves in mess. spend your time till you are convinced that this way you will not face any problem till the end.

5. execute the plan, i usually follow a principle i call it "One shot kill", that means when i do things for the first time, i leave no room for error, and imagine that this is final work. and do mostly finish my work in one go. (it saves lot of time) i am not saying that you shall implement such hard rules in life (though it helps), but plan everything so nicely that leave no room for error.

6. if you fail, try searching documentation and keep thinking about another way of doing it. its like a chess game, once you start enjoying it, you will never play startegy games on computer.

I like the last comment about it being like a chess game. I never thought of it like that, but I agree totally.

Another thing I will do sometimes is to create a relatively simple tet mesh (not a coarse mesh though... this usually ends up being a large mesh, but still leaves a few things to be desired) using sizing functions where appropriate. Then I'll start that model running, and while its running, I'll completely remesh the model using a structured mesh. Gives me the chance to always work out new ways of decomposing the volumes to get what I want (keeps me on my toes I guess you could say). Also, I have no luck using BL mesh with an unstructured grid, but I can pull it off if I'm decomposing the volume for a structured mesh. The other bonus is that I have a rough solution that I can look at and see if there's any effects I wasn't expecting and I can modify my new mesh to capture those.

Jason