CFD Online Discussion Forums

CFD Online Discussion Forums (
-   ANSYS Meshing & Geometry (
-   -   [ICEM] Circular 2D Flow Field Setup - (

Bourke December 7, 2015 18:41

Circular 2D Flow Field Setup -
5 Attachment(s)

I'm having issue setting up a geometry for a flow field around what essentially is a circular cylinder.

The center of the field contains a cylinder with a synthetic jet cavity and orifice. For testing purposes, I need to rotate the synthetic jet through a number of degrees so that the incoming fluid flow impacts it at different angles.

I plan on setting up a hybrid mesh, with an unstructured mesh in the far field and a structured mesh close to the cylinder using ICEM once I've imported the geometry.

My issue is however that I don't know the best way to go about changing the angle at which the flow is coming in at. The range of angle change at maximum will be 180 degrees.

Initially, I've created the geometry with a circular boundary, thinking that I could possibly change the incoming fluid velocity direction at the boundary and I'd split the outer perimeter in half, with one section being the inlet and the other being the outlet.
But then the issue is encountered that I need the inlets and outlets to rotate around the synthetic jet in the center and I don't know how to easily do that. There is also an issue encountered with associating the curves of the original initialized block seeing as its a four sided shape and the external circular boundary is only split into inlet and outlet.

The other method I've thought of is rotating the synthetic jet geometry itself and creating multiple different geometries and hence test files, but that would require a lot of work since I would have to setup meshing for each different file across a range of angles. I would like to avoid doing this.

A final method I've thought of is creating a circular cylinder and synthetic jet geometry with an inner domain surrounding it which I create a structured mesh for. Then maybe I can export it and merge with different outer domain geometries each of which will be unstructured and have the necessary boundary conditions.

I'm attaching some images just to hopefully clarify what I'm talking about.
If you have any recommendations on how best go about changing the incoming fluid flow direction in relation to the synthetic jet, please do share!

Thanks a lot for your time and consideration!

Bourke December 10, 2015 14:00

I've decided to do something similar to the discussion in this thread that I've linked, whereby I load a circular mesh with the synthetic jet cylinder already meshed, into a large rectangular field and then rotate it for different tests.

I'm not sure how to go about generating the mesh around in the far field though or how to merge the subsequent two meshes so that the nodes from the inner section and outer are conformal.
When I attempt to generate the tri surface patch dependent mesh in the outer field area, it doesn't seem to line up the nodes with the nodes of the inner field boundary.
I've turned on respect line boundaries etc. I've also reduced the max element size to a really small value to see if that'll help i.e. the outer field might be too coarse to align nodes with the inner field, but as a result of doing so, the entire outerfield mesh becomes too fine.
Any advice that can be afforded would be appreciated.

Bourke December 10, 2015 22:51

3 Attachment(s)
Just attaching some images to make it a bit clearer what I'm talking about.

As a note, in the third image, the tri mesh seems to generate some unusual excess lines which I don't really understand what they are. They don't appear to form an element so I'm not really sure what's their cause, but either way I thought it might be good to include it.

I've tried merging the two meshes by combining them as parts, but then when I try and use the mesh in fluent, there is a wall between the two zones.
I've made sure that all dividing lines like the interface between the two are set as internal but still no luck.
Looking at it, I think that the two meshes just aren't merged correctly and stop at their respective boundaries, I'm not sure how to solve this, my method was probably incorrect at some point.

All times are GMT -4. The time now is 13:54.