SimScale CAE Forum

Wedge BC meshing issue - Help Please


#1

Hi, I am trying to simulate flow through an orifice plate. My aim is to determine discharge coefficients at various reynolds numbers. However before i get there i am having a few issues meshing the model which can be found here… https://www.simscale.com/workbench/?pid=3126067294385075099&mi=spec%3A98d29b85-5848-4ce3-bdfd-272842703310%2Cservice%3ASIMULATION%2Cstrategy%3A59

The geometry is axi-symmetric so i am trying to make use of that by using the wedge BC. There are no issues with using either a standard or tet dominant mesh … however i am unable to mesh using HEX mesh, i am wondering that this is because of the geometry as a result of it converging to a point at the axis of the cylinder which means you need a wedge element in this location. Is there a way of doing this in simscale ? Or maybe a better question would be… what is the best practice for meshing this problem?

Appreciate any help given,. Thank you, Ross


#2

Hey Ross!

Have you tried the other mesh option, namely the standard mesher and see if it worked out with that one? It is more robust in most of the cases, give it a spon and let us know how it went.

Best,

Jousef


#3

Hi Jousef,

Thanks for responding. I have tried other meshers’s both standard and tet, which have meshed however when running the solver convergence cannot be achieved which i believe is due to the mesh. Reading the wedge boundary condition:

  1. The axisymmetric ‘wedge’ boundary is specified by two planes that must be selected on separate sides of the domain on surfaces running along the axis. My geometry is setup in this way as described

  2. In case of a 2-dimensional cylinder, the geometry would be specified as a wedge with a small angle (less than 5 deg) . My geometry is setup in this way as described - 1 deg angle

  3. 1 cell thickness. _This is not achieved during after meshing, i seem to get at least 2 cells in the through thickness of the geometry. Is there a way to force the number of cells in the through thickness to be 1? I would have though a brick, and wedge element would be desirable for this. I also notice that the faces with the applied wedge BC do not appear to be mapped i.e. symmetry of the node points either side of the domain plane. _

  4. I also have issues creating an inflated boundary layer, but hopefully if the global mesh can be fixed then this will take care of itself!

thanks for all the help,
Ross


#4

Thanks for the update Ross, I will forward that to my colleagues from the CAD team, maybe there is a reason behind this behavior.

Cheers!

Jousef


#5

Hi Ross,

Some comments from my colleague @jprobst in this regard: “The model must be symmetric around a coordinate system plane, it must only be a single layer of cells, and the points must be colocated (see docs).” Also he thinks the only way anyone can successfully use the wedge BC is if you upload the mesh.

Hope that helps!

Jousef


#6

Hi Jousef,

Thanks for getting back to me. Its a shame you need an external mesher to use this feature, i think it would be a good edition to SimScale to imbed this meshing feature into the software. Having the ability to run quick 2d simulations for the purpose of sensitivity analysis is very useful.

As a bit of feedback, i think you should state on the wedge boundary condition that an external mesher needs to be used as i have used CPU hrs trying to get this working with no success because its not actually possible within the SimScale environment.

Thank you
Ross


#7

Will forward this to my colleagues Ross, thanks a lot for the feedback :slight_smile:

Let us know if you need anything else!

Jousef