CFD Online Discussion Forums - LES evaluation help needed

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- Main CFD Forum (https://www.cfd-online.com/Forums/main/)

- - LES evaluation help needed (https://www.cfd-online.com/Forums/main/113173-les-evaluation-help-needed.html)

Ok... I get them from ICEM. What do you use for meshing?

I use StarCCM+... do you have any chat software such as google talk, or skype or MSN so that I can talk to you ?

Quote:

Originally Posted by RodriguezFatz (Post 479310)

Ok... I get them from ICEM. What do you use for meshing?

Sorry, I don't know anything about Starccm, maybe people in the sub-forum can help you...

But can you tell me how to get it from ICEM please? I use ICEM and CFX too... but currently working with Starccm+..

Thank you !

Quote:

Originally Posted by RodriguezFatz (Post 479313)

Sorry, I don't know anything about Starccm, maybe people in the sub-forum can help you...

Thank you for your kind reply. But regardless, to calculate d+, the width or length of the cell is needed, so : how to get that information from the software? I am using Starccm+, do you have any clue? I just asked RodriguezFatz how to get it from ICEM..

Thank you !

Aqua

Quote:

Originally Posted by sbaffini (Post 479300)

I guess what RodriguezFatz was trying to say is that, as those parameters actually depend from your flow direction, there is no obvious way to compute them without knowing in some detail your flow.

And even in that case, your grid might not be aligned with the flow direction, then how would you say what is dx^+ (stream-wise grid spacing in viscous units) and what is dz^+ (span-wise grid spacing in viscous units)?

However, for a given cell near the wall, you usually know the instantaneous y+ or, if you already have your statistics, you can compute the average y+ trough the mean wall shear stress. Then, if you know the grid (in Fluent you have access to several cell-related geometrical parameters) you can compute whatever d+ you want.

I remember doing something like this for a conference, where i had to plot the delta+ in a swirler for different swirl numbers (= different mean flow directions). I did it quite easily in Fluent but there was some hard-coding related to the geometry in order to obtain the correct values.

Do you use blocking/hexa for meshing?

Yes, I do use blocking/hexa for structural mesh..

Quote:

Originally Posted by RodriguezFatz (Post 479318)

Do you use blocking/hexa for meshing?

Then you just need to go to the relevant edge (parallel to the surface) and read the sizing of the cells.

And then? I mean, each cell has its size, and its own local friction velocity. Then the next step should be to calculate the d+ by the size of the cell and its local friction velocity, which should be gained from CFX or Fluent. How do you combine the two factor: size and friction velocity? (my geometry is complex, so the size and local friction velocity vary a lot..)

Quote:

Originally Posted by RodriguezFatz (Post 479320)

Then you just need to go to the relevant edge (parallel to the surface) and read the sizing of the cells.

Ok, if they vary strongly then you are doomed. Maybe you can create a custom function in Fluent? Is dx and dz available in Fluent?

Wow... I don't know Fluent, what I know is CFX and starccm+... I am doomed, I guess...

Thank you

Aqua

Quote:

Originally Posted by RodriguezFatz (Post 479322)

Ok, if they vary strongly then you are doomed. Maybe you can create a custom function in Fluent? Is dx and dz available in Fluent?

Ok let's clarify the issue. You are using starccm+. We (me and Philip, i guess) usually use Fluent. You should give first a look at what are the plotting options in your solver. Are you sure there is no section related to the geometry of the cell? In Fluent, for example, you have:

x,y,z centroid coordinates

face area magnitudes

cell volumes

etc.

In some cases, with proper insight, you can manage a proper combination of cell parameters to finally obtain the formula proposed by Philip.

Now, let us suppose you don't have such information. Then you necessarily have to go in the mesh generator (ICEM, as i understood) and retrieve the local information you are looking for.

However, note that:

1) If your grid has such a variability that a representative grid parameter cannot be evaluated for a given zone then probably it has also little sense to try to obtain it.

2) You still have the y+ value. I can hardly imagine a wall parallel spacing which has so large variations to make even a guess completely meaningless.

3) More often than not you just need approximate maximum and minimum values, which means guessing maximum and minimum grid spacings in the wall-parallel directions. Also, you usually have a guess of the required spacing (say, dx+<50-60, dz+ < 30). If a rough check cannot give you insight on correctness of the spacing, then probably you are already far from the correct one.

4) If there is such variability to make this evaluation impossible, there is possibly much more to check than just these values. Indeed, these values are just a little part of the picture for selected kind of flows (wall bounded, possibly in equilibrium). In particular, choosing the wall shear stress as relevant quantity to nondimensionalize the grid spacing implies that it actually is a relevant quantity to do the job. If the flow is so complex that a relevant quantity is not identifiable then the effort to produce such d+ values is not worth it.

This sounds perfect. Sorry aqua I am too buisy and had such short comments only!