[victoryv]

I have recently joined a new company where people just started to use solidworks flow for CFD analysis . I haven't' used it yet. I don't know how good it is, but I have some questions.

1. People here were obtaining results without even doing mesh independent study. I was startled. Does solidworks take care of mesh itself?

2. Do we have greater control over mesh ? Can we create boundary layers?
Can we create local meshes?

3. How good are the results?

[Boris_M]

Hi Victoryv,

Congratulation to become a FS or EFD user ;-)

1. Kind of... usually we suggest to do some mesh studies in order to get some experience in the beginning when starting to use the software. There are some rule of thumbs such as at least 3 cells over a small gap in which you have some important flow, you don't have to resolve all small gaps with that rule where there might be some weak secondary flow but more the more important gaps. If you want high accuracy you should resolve a nozzle for example with ~8-10 cells etc.
There is some meshing and solver technology that resolves some of the typical issues you spend days on mesh convergent studies for certain mesh regions.

2. Yes, you can create local meshes and go into more detail of the mesh settings than automatic setting Level x. But why do you need boundary mesh if the solver handles it for you. If your mesh is too coars for the boundary layer is not resolved with the mesh so we apply automatically a boundary layer apprach which you can find in the documentation and if the mesh resolves the boundary layer the boundary layer is calculated by the solver directly.
You have to consider to think different compared to the typical CFD tools.

3. That really depends on the application you are trying to calculate as with every solver. Some are ideal for turbomachineries but not so good as others for external aerodynamics etc. What are you applications?

Regards,
Boris

[victoryv]

Thank for the info Boris. Here people are using it for more design optimization. Let me play with the software more. If it is able to handle mesh and boundary layers itself then that is great.

Originally Posted by Boris_M

Hi Victoryv,

Congratulation to become a FS or EFD user ;-)

1. Kind of... usually we suggest to do some mesh studies in order to get some experience in the beginning when starting to use the software. There are some rule of thumbs such as at least 3 cells over a small gap in which you have some important flow, you don't have to resolve all small gaps with that rule where there might be some weak secondary flow but more the more important gaps. If you want high accuracy you should resolve a nozzle for example with ~8-10 cells etc.
There is some meshing and solver technology that resolves some of the typical issues you spend days on mesh convergent studies for certain mesh regions.

2. Yes, you can create local meshes and go into more detail of the mesh settings than automatic setting Level x. But why do you need boundary mesh if the solver handles it for you. If your mesh is too coars for the boundary layer is not resolved with the mesh so we apply automatically a boundary layer apprach which you can find in the documentation and if the mesh resolves the boundary layer the boundary layer is calculated by the solver directly.
You have to consider to think different compared to the typical CFD tools.

3. That really depends on the application you are trying to calculate as with every solver. Some are ideal for turbomachineries but not so good as others for external aerodynamics etc. What are you applications?

Regards,
Boris