CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   general question about wall distance in ICEM + CFX (http://www.cfd-online.com/Forums/cfx/23247-general-question-about-wall-distance-icem-cfx.html)

Andy October 31, 2006 11:42

general question about wall distance in ICEM + CFX
 
Hello,

I am wondering about the wall distance that is achieved by using the prism-mesher in ICEM.

The manual is telling that when boundary layer seperation from a smooth surface is supposed to be simulated there has to be a wall distance of the first node of 'Y+'<2. Therefore an equation is given that gives an idea what initial height for the first prism layer (or the first variable node?) should be selected in order to achieve this Y+. (Eq.20 in 'Guidelines for Mesh Generation')

My question is: When using CFX the finite volume approach is used and hence the flow variables are finally calculated at the centres of the cells,right? If that is the case then when setting the initial HEIGHT of the first prism layer I am actually setting up DOUBLE the distance of the required height? Is this fact already incorporated into the Eq.20 that means does the ICEM-manual account for a finite volume approach that is taking the mesh nodes as the FV-vertices and not as the FV-center?

I hope that this has not been to confusing and some pro out there is able to answer those questions... ;-)

Thanks anyway, Andy

opaque October 31, 2006 11:55

Re: general question about wall distance in ICEM +
 
Dear Andy,

ANSYS CFX is a FV vertex based method, not a cell centered method.. That is, variables are computed at the ICEM mesh vertices. Will that explain your confusion?

Regards,

Opaque

Robin October 31, 2006 12:00

Re: general question about wall distance in ICEM +
 
Hi Andy,

The equation is only meant as a rough approximation and will get you in the right ball park. In terms of the actual location, CFX builds control volumes around the nodes of your mesh, which actually places a half control volume around the node at the wall, so the integration points are in fact located 1/4 of the distance to your first node.

In any case, don't worry about where the integration occurs. In the end, your Yplus is going to vary a lot over the surface. It won't hurt to have cells too close. At most, it may increase your mesh size by 1 or 2 percent (assuming you control your topology appropriately). It is better to have too much mesh in this case than to find you have too little later.

Regards, Robin


Andy November 2, 2006 03:39

Re: general question about wall distance in ICEM +
 
Thanks guys, but your answers raise two more questions:

Did I get it right that "opaque" is saying that CFX is vertex based (i.e. CV-vertex = Mesh-node) whereas "Robin" is saying that CFX is cell-centered (i.e. CV-center = Mesh-node)? So what is correct then?

@Robin: Does that mean when I am selecting a physical distance of 0.1mm for the initial height of my first prism layer the actual CV-center[=location for calculating the flow variables] is located at 0.1mm/4=0.025mm ?

Thanks and regards Andy

Robin November 2, 2006 08:59

Re: general question about wall distance in ICEM +
 
Hi Andy,

Opaqua and I are saying the same thing. When I said it builds control volumes around the nodes, I was referring to the corners of the elements.

Regards, Robin

Andy November 2, 2006 09:20

Re: general question about wall distance in ICEM +
 
Ah, right, my fault. Thanks for the help Robin, I really appreciate it. You are doing a great job!

tmajaykrishna February 28, 2013 02:44

I want to know how to calculate wall y plus (as in fluent) in CFX. which of the two variables ( y plus and wall distance) to be used to calculate

pls help

with regards,

Ajay Krishna


All times are GMT -4. The time now is 02:42.