CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   prismatic layers (http://www.cfd-online.com/Forums/ansys-meshing/78988-prismatic-layers.html)

vinuaero@yahoo.co.in August 9, 2010 07:02

prismatic layers
 
hi,
can anyone tell me how prismatic layers are related with aerodynamic conditions like pressure,temperature,velocity etc.

BrolY August 9, 2010 10:49

The prismatic layers are created to capture the boundary layer.
The boundary layer depends on your aerodynamic conditions (The number Y+ is usually used to calcul the height of the boundary layer). The most important one is the velocity.
You can google it for more details about the Y+ formula.

vinuaero@yahoo.co.in August 10, 2010 00:28

Hi Alex,
I have got the formula to solve it.but i didn't understand the term "REF. LENGTH".please guide me.
Thank you,

BrolY August 10, 2010 04:41

The reference length depends on your geometry.
For example, if the geometry is a pipe, you would use the diameter (or the radius) as the reference length. Got it?

vinuaero@yahoo.co.in August 10, 2010 05:12

yes i got it.Do u know about KARMAN and upsweep discretization in ICEM CFD TETRA?
Please give some suggestions.
thank you,

BrolY August 10, 2010 05:52

I'm sorry, I can't help you on that. Maybe someone else on the forum?
Good luck !

PSYMN August 10, 2010 15:48

vinuaero you may be putting too much pre-thought into it...

Basically, you just want your prism layer to be thicker than your boundary layer so you can more efficiently capture the boundary layer profile.

Run your Y+ Calcs, etc. (they are useful for the first cell height) but at some point you just need to put as many prism layers on there as you can and send it to the solver. Then look at the solution with the grid visible.

Is the majority of the boundary layer (slower flow) within the prisms? If so, is it relatively close or are your prisms much thicker than they needed to be (did you put excess work and sacrifice quality to get prisms you didn't need...? Learn for next time.) Does it look like your prisms are not thick enough to capture the viscous boundary layer profile?

Most CFD analysts are really repeat users (repeat offenders) who learn from previous runs and generate better meshes based on understanding a certain class of problem. You probably won't know exactly how thick your prism layer should be until you make a few runs on that type of configuration and get a feel for it... Just make sure to observe how your results relate to your mesh and learn as you go. Most mature customer sites have meshing guidelines for various classes of models which they guard closely as trade secrets.

You may also want to run some mesh refinement studies to see if your mesh is fine enough that the results are not dependent on them.


All times are GMT -4. The time now is 05:57.