Learning Forum

[KSD]

[Rob]

Can you check to see how many edges/faces you have in that area and that they're in the inflation selection.

[Keyur Kanade]

Please check if there is any small edge or some extra vertices.
Regards Keyur
How to access Ansys Online Help Document
How to show full resolution image
Guidelines on the Student Community
How to use Google to search within Ansys Student Community

[KSD]

Thanks a lot for answering Sir/Madam Yes you are right, there is a very small curvature (Of radius 0.254 mm) in diverging part of nozzle compared to converging part of nozzle whose radius is (25.4 mm), But I don't know why while importing the geometry from solid works the small curvature is not shown in Ansys meshing.
I think, I should ignore that small curvature part and make it conical divergent section shape geometry without any curvature.
And I have another doubt in this is : As inflation was not working so I tried biasing option, it worked well despite of the small curvature part.
As you can see from the images that I only want boundary layer mesh around the "red marked " wall. (Blue marked is the line of truncation of nozzle) .
If I use inflation type or biasing type, it would automatically connect the domain outside if nozzle wall, which I don't want to happen

Are there any way to make boundary layer mesh only in the red marked region without linking this with entire domain of flow field and still getting structured mesh?
Are there any special ways to partition this geometry to get the desired results.
I cannot think of a way to stop the boundary layer mesh till wall and also have good structured mesh for interior flow domain.
I have tried the following grids with different partition line but this grids are not satisfactory as it links the boundary layer mesh with outside region of nozzle wall.

I am really struggling to get a structured mesh without BL mesh at nozzle exterior domain. (I want Type-5 type of BL mesh but without it being unstructured)
Please Help to solve this issue.

[Rob]

If you want a structured mesh you'll need to be very careful with cell sizes and growth rates to avoid excessive jumps in cell size. If you don't manage this the solver may be unstable and the results poor. I'd be using unstructured hex and a lot of adaption to model this.

[KSD]

@Rob Sir, Sir seems like I wouldn't get my desired mesh using Ansys meshing, will I be able to get what I want using Ansys ICEM CFD as I am found in many posts that ICEM CFD is better for structured meshing and has better control? But before turning to ICEM I want to first ensure this because even getting the desired first cell height using ANSYS meshing needs lot of trial and error if I use biasing method.

[KSD]

Sir, will I be able to make such mesh in ANSYS ICEM CFD because as it uses projection or blocking type meshing? Because I don't think Ansys Meshing, Can create such mesh as once you use boundary layer meshing, the boundary layer biasing get connected to entire domain of the flow feild?

[Rob]

You can use local inflation in Workbench Meshing, the global settings are a "default" and don't use automatic inflation.