code_saturne User's Forum

Hello,

Just a quick question. Below is an output in my listing file for a simple simulation where I have a domain entirely consisting of 0.08m cubes and an average velocity of around 0.27m/s. Can someone explain to me how the "Rough dimensionless distance yplus" is calculated below? I have read the best practice guide which, whilst being very helpful, hasn't answered my question. What I am after is the exact definition of "Rough dimensionless distance yplus" as listed below because no matter how I play with the numbers I do not get yplus = 0.8 out!

thanks in advance!

Susan

** BOUNDARY CONDITIONS FOR ROUGH WALLS
--------------------------------------
------------------------------------------------------------
Minimum Maximum
------------------------------------------------------------
Rel velocity at the wall uiptn : 0.14807E-01 0.76368E-01
Friction velocity uet : 0.46803E-01 0.70534E-01
Friction velocity uk : 0.21112E-01 0.42061E-01
Rough dimensionless dist yplus : 0.80000E+00 0.80000E+00
------------------------------------------------------
Nb of reversal of the velocity at the wall : 0


------------------------------------------------------------

Hello,

Is no one able to shed some light on this? As far as I can tell the value listed simply corresponds to 10x whatever grid spacing I used at the wall. I am struggling to see how this is useful? Is it just there as a quick reminder of what my deltaY is so that I can manually work out Yplus? In which case why is it x10?!

cheers

Susan

Hello,
For rough walls is computed from the roughtness (and not uk).

Regards,

Hello Jacques,

thank you very much for your reply! That instantly makes it clear. Out of interest is that information recorded anywhere and I have missed it? Though I had done a pretty extensive search before asking on here.

So on each of these runs I had been using a roughness of 0.1m (very high I know but trying to represent a surface covered in breeze blocks!) which explains why I thought it was x10.

So, do the rules regarding y+ for k-epsilon (in the range 30 -100 etc) apply to this wall roughness definition or to the normal one involving friction velocity etc?

Also I though uet was the friction velocity and uk was the friction velocity attributable to turbulence. If this is the case should I not be using uet to work out y+?

Finally in order to output y+ with its standard friction velocity definition do I need to write a user routine involving yplpar or is this possible from somewhere in the GUI?

thanks for your help and apologies for following it with a lot more questions!

Susan

I just want to update this topic. Because I have the same problem.

I need to calculate flow over very rough seabed (y0 = 0.4).I guess it is necessary to use cell size > y0, but I want to use fine mesh to study the phenomena that affect objects of size 0.1. What the practice should be applied in that case?
Is it possible to use roughness wall in that case? If yes, I have a few questions. Should I calculate y+ manually from shear stress? At what height should I estimate this value? What is the target y+ value in this case?

Thank you.