[martyn88]

Hello everyone,

I am attempting to mesh a converging diverging nozzle to be used for an LES simulation of a supersonic free jet. I have created the following in ICEM but have a few more questions.

nozzle.jpg

inlet_face.jpg

nozzle_exit_far.jpg

nozzle_exit_closeup.jpg


I require high resolution inside the nozzle and a boundary layer mesh along the nozzle wall. Also I require high resolution inside the jet core, particularly the shear layer.

So I would like to use a structured hex mesh for the regions outlined above however I don't need high resolution in the far field and was thinking of using tet cells for this region to reduce cell number.

I would also like to have smoother transitions between these areas of high resolution. Would Laplace smoothing be appropriate for this? How would I implement that?

Could anyone help me with this?

[Far]

Originally Posted by martyn88

Hello everyone,

I am attempting to mesh a converging diverging nozzle to be used for an LES simulation of a supersonic free jet. I have created the following in ICEM but have a few more questions.

Attachment 16024

Attachment 16025

Attachment 16026

Attachment 16027


I require high resolution inside the nozzle and a boundary layer mesh along the nozzle wall. Also I require high resolution inside the jet core, particularly the shear layer.

So I would like to use a structured hex mesh for the regions outlined above however I don't need high resolution in the far field and was thinking of using tet cells for this region to reduce cell number.

I would also like to have smoother transitions between these areas of high resolution. Would Laplace smoothing be appropriate for this? How would I implement that?

Could anyone help me with this?

These questions are more related to CFD solver then meshing. So Fluent forum would be nice place to post this one, otherwise we are always here to help out

[stuart23]

Hi Hugh,

You need to change your edge distributions (Pre-Mesh Parameters -> Edge Parameters). I tend to use exponential to obtain the bunching near the surface required to resolve the boundary layer. If you did not want to have exponential growth the whole way along, you could do a split near the boundary and apply an exponential (or other growth function) sizing near the wall and then uniform spacing away from the wall. This would be a good way of gaining resolution inside the jet.

To reduce the resolution at the jet exit in the farfield is more of a problem. If your solver allows it (I think it is ok in Foam), you could split the domain radially and refine only the inner block, therefore creating a 2-to-1 or 3-to-1 cell match up at the edge of the blocks. (There are also tools to refine the 3-to-1 hanging nodes if you don't like solver interpolation/you use CFX).

Another way of creating local refinement is by using an O-Grid at the nozzle exit. The O-Grid allows mesh to be bunched into a small area, however your cells will nolonger be parallel to the flow/geometry, and you will start getting a quite unthoginal mesh.


Stu

[martyn88]

Originally Posted by Far

These questions are more related to CFD solver then meshing. So Fluent forum would be nice place to post this one, otherwise we are always here to help out

Thankyou for the response, but I think it is more related to meshing.
I would like to know how to coarsen my mesh in the far field while still maintaining a structured hex mesh inside the nozzle and jet core region. Can a tet mesh be used that matches with high res hex mesh in the jet core region and then coarsens in the far field?

Also, I would like to know how to smooth my mesh and create cells that are more uniform (aspect ratio closer to 1) and with smoother cell size transitions. (Look at inlet face - hex cells could be more uniform)

Finally I would like to know the best way to create a boundary layer mesh.

I hope my questions are clear.

Thanks

[martyn88]

Thanks for the reply Stu,

I will look into some of those suggestions tomorrow and let you know how I go.

Cheers again