[Rodrigo_Eng_Mect]

Hi, I would like yours opinions! I am simulating the ANSYS CFX a rotor and centrifugal pump diffuser, the boundary conditions is an inlet pressure (0 Pa) and mass flow at the outlet, in fact the rotor has a certain roughness !! I enter this roughness parameter in "Wall Roughness"? For example: 0.1 [mm] - Steel! When I use the "Smooth Wall" My Y + Global option is at 123 when inserting roughness condition my Y + is in 1023! What do you think I enter a value in Wall Rougness? Thank you

[ghorrocks]

Please read the CFX documentation, Theory manual: Turbulence and Wall Function Theory/Flow Modelling near the wall/Mathematical Formulation/Treatment of rough walls.

It explains that rough walls lead to a shifting of the turbulent velocity profile near the wall, which leads to a shift in y+ value. So what you are seeing is the expected result.

[yanglu]

Originally Posted by Rodrigo_Eng_Mect

Hi, I would like yours opinions! I am simulating the ANSYS CFX a rotor and centrifugal pump diffuser, the boundary conditions is an inlet pressure (0 Pa) and mass flow at the outlet, in fact the rotor has a certain roughness !! I enter this roughness parameter in "Wall Roughness"? For example: 0.1 [mm] - Steel! When I use the "Smooth Wall" My Y + Global option is at 123 when inserting roughness condition my Y + is in 1023! What do you think I enter a value in Wall Rougness? Thank you

HI, How do you decide the wall roughness value with 0.1[mm]? Is this value too high? Thanks.

[ghorrocks]

0.1[mm] wall roughness is very small if the rotor is a 14m diameter hydro-power station turbine, but very big if the rotor is a 30mm diameter turbocharger. It all depends on how big the device is.

[Opaque]

If I recall correctly, the wall roughness physical height importance is relative (as Glenn suggested). What matters is its influence within the boundary layer, and it is usually done by using turbulent scales.

Similarly to y+, there is a roughness+ (k+).

You probably should read about how k+ is computed, and used within turbulence modeling.

[yanglu]

Thanks. I am calculating the oil drag power of the twin-screw compressors. The power seems very low with the smooth wall. So, I want to add the wall roughness of the rotor surface. By the way, when I enable the non-overlap condition to set the wall roughness = 1e-5, it gives me an error as following, do you know how can I solve this error? Thanks.

Domain Name : rotor
Mesh Coordinates

Details of error:-
----------------
Error detected by routine PEEKCS
CDANAM = HOW
CRESLT = NONE

Current Directory : /FLOW/BOUNDCON/ZN7/BCP19/NOVARIABLES/RGH

+--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Stopped in routine MEMERR |
| |
| |
| |
| |
| |
+--------------------------------------------------------------------+

+--------------------------------------------------------------------+
| An error has occurred in cfx5solve: |
| |
| The ANSYS CFX solver exited with return code 1. No results file |
| has been created. |
+--------------------------------------------------------------------+

[Opaque]

What version of the software are you using?

I recall seeing this error some time ago. You should probably contact ANSYS CFX support, and discuss this error with them. They can provide you a workaround