[Burken]

Hello!

I am currently doing my master thesis project which consists of performing CFD analysis of a centrifugal compressor in a turbocharger.

One of the first parts of this project is to do a mesh study to determine how refined my mesh needs to be in order to get the requested simulation data in the least computational heavy simulation.

when it comes to the CFX setup, the company that I perform this project at has from previous work found out that the best suitable turbulence option is the SST Menter k − turbulence model with the Automatic near wall treatment as wall function.

I have read in the Ansys help that when using automatic near wall treatment, the solver based on both wall functions and Low-Reynold-Number methods overcomes the grid resolution requirements by gradually switching between wall function and LRN method.

From what I have understood, the automatic near wall treatment will switch to wall function treatments when , but I have also heard from colleagues that the switch actually happens for .

For that reason, I have tried to find answers to this online, where some say that the switch happens at . The fact that 11.06 seems to be the actual limit and not 2 makes me curious since none of the threads I have read about this refers to any documentation. So my question is, where is the line where the switch actually happens?

I am planning to make the mesh refined enough to avoid switching between wall functions as far as possible by having a refined mesh that puts me in the lower region. I would be very thankful if someone can help me with this question. Thanks in advance!

/Burken

[ghorrocks]

This is discussed in section 2.8.1.2 of the CFX theory manual, particularly equation 2.229 (CFX V19 help files). Eqn 2-229 is y+ = max(y*,11.06), so the y+ is limited to 11.06.

I do not know where you got y+=2 as the switch from.

[Burken]

Originally Posted by ghorrocks

This is discussed in section 2.8.1.2 of the CFX theory manual, particularly equation 2.229 (CFX V19 help files). Eqn 2-229 is y+ = max(y*,11.06), so the y+ is limited to 11.06.

I do not know where you got y+=2 as the switch from.

Thank you for your reply, I will check that out immediately.

I have read in section 4.2.3 from version 18.2:

"However, a strict low-Reynolds number implementation of the model would also require a near wall grid resolution of at least . This condition cannot be guaranteed in most applications at all walls. For this reason, a new near wall treatment was developed by CFX for the based models that allows for a smooth shift from a low-Reynolds number form to a wall function formulation. "

/Burken

[ghorrocks]

The quotation says that for good resolution when integrating to the wall you need y+<2, but for most cases this is difficult to achieve everywhere so the approach used by CFX can handle larger y+ than that by switching to wall functions.

The switch occurs at y+=11.06.

[Burken]

Originally Posted by ghorrocks

The quotation says that for good resolution when integrating to the wall you need y+<2, but for most cases this is difficult to achieve everywhere so the approach used by CFX can handle larger y+ than that by switching to wall functions.

The switch occurs at y+=11.06.

Ok, thank you so much for your help!