[fluid23]

I am currently working my way through the tutorials contained in the Star-CCM+ help material (and saved under \doc\startutorialsdata in the installation directory). I am working the 'Steady Flow-Laminar and Turbulent in an S-Bend' example which illustrates the Generalized Cylinder Mesh tool. However, the mesh that is generated does not resemble the mesh shown in the tutorial (pg 8400 of the pdf). The screenshot in the tutorial shows that the curved regions of the pipe have higher element density than the straight sections. My generated mesh is the exact opposite and as far as I can tell I have followed the directions exactly. My questions are:

1. Given that the mesh tool is optimized for circular pipe flows, do I need to worry about decreased fidelity in curved regions as compared to straight? The scalar plots look identical and the only discrepancy in the vector plots appear to be density/locations of the vectors which seem to be linked to cell location.

2. Is there a way to set the clustering of these elments so that they occur in regions with curvature rather than straight regions? I played around with changing the Extrusion Law to Symmetric Hyperbolic Tangent with some limited success but I was unable to get anywhere close to what is shown in the tutorial. I also managed to crash the software doing this, FYI.

I am running v9.02, if that makes a difference. See attached image for further details.


Thanks in advance.

[ping]

it looks to me like they used another mesher for the image and i agree that the cylinder mesher cant do that at present
you can get close to that by spliting your pipe into 3 separate wall boundaries by using imprinting in 3d-cad then apply different constant pipe extrusion parameters to each wall

[ViktorKL]

The post under this link shows how to apply the generalized cylinder mesh and extrude from the interface.