Learning Forum

[Nikolaos Monokrousos]

Hello, I am currently trying to solve a three-dimensional, turbulent channel flow in a rough walls microfluidic channel. In the ANSYS Meshing I am creating inflation on the surrounding walls of the channel to be able to get a y+ value at around 1, using the "First layer thickness" option and thickness equal to double the value of the roughness height, as suggested by the ANSYS manual: "Note that it is not physically meaningful to have a mesh size such that the wall-adjacent cell is smaller than the roughness height. For best results, make sure that the distance from the wall to the centroid of the wall-adjacent cell is greater than Ks.".

 

However, I am not able to reach values of y+ at around 1. It is possible to achieve that with very high roughness values, such as the one I am encountering in my Additively Manufactured channel? I have not been able to find anything published so it would be great to get some opinions on this matter.

 

[Rob]

We're alittle restricted in what we (as Ansys staff) can comment on, so I can't give a full answer. 

In your case the channel is very small, so surface roughness may be big relative to the channel size: there's published data on micro channel shape issues but more so with laser & acid etching (well more that additive wasn't sensitive enough to consider using when I last looked into it). So, if you try and resolve down to y+ of around 1 you're resolving to the level where you may want to consider including the roughness in the geometry. However.... y+ is also a function used for turbulent boundary layers: is your system really turbulent? 

[Nikolaos Monokrousos]

Thank you for your reply. The system is turbulent, yes. I have been reading regarding the shift in the origin of the y-coordinate that roughness introduces, placing the origin inside the roughness sublayer instead of the wall of interest and it is understood. However, in the case we accept the ANSYS suggestion that the mesh size should be twice the roughness height (equivalent sand-grain roughness height), which other metric could be utilised to qualify the Boundary Layer mesh and the near wall resolution?

In section "7.4.15.3.1.1. Setting the Roughness Parameters" of the ANSYS User's Guide 2023R2 it is stated that: "The advantage of the rough wall formulation using a virtual shift of the wall (Equation 7.120 (p. 1426)) compared to reducing the roughness height as decreases (Equation 7.119 (p. 1425)) is that it eliminates all restrictions with respect to mesh resolution near the wall, and can therefore be used on arbitrarily fine meshes". Assuming that someone is employing the k-omega SST turbulence model which inherently adopts the rough wall formulation using a virtual shift of the wall, is it safe to say that the "restriction" of y+ around 1 is eliminated and it is only a matter of using a fine mesh?