CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   How can we effciently mesh for these type of geometries? (http://www.cfd-online.com/Forums/cfx/92198-how-can-we-effciently-mesh-these-type-geometries.html)

arunraj September 6, 2011 01:51

How can we effciently mesh for these type of geometries?
 
1 Attachment(s)
Hi all,

I have attached a picture of a screen geometry with this ..I want to know whats the efficient way for meshing this?...Also I want to know which type of meshing (Structured grid or Unstructured grid) I need to use ..

ghorrocks September 6, 2011 19:01

The type of mesh which is approriate also depends on the type of model you plan to do. If this is a 1mm wide liquid stirrer you would mesh it entirely differently to a 1km across wind turbine.

arunraj September 6, 2011 23:06

1 Attachment(s)
Hi I have attached the Image of geometry which I want to analyze....Have a look at it ..It is the distrtion screen which they will keep in-front of axial compressor to produce distortion

ghorrocks September 7, 2011 07:51

You still have not answered the critical question - how big is the object, or what Re number does it run at?

Also, why are you modelling this rather than imposing its effect as a momentum sink term?

arunraj September 7, 2011 08:25

how big is the object? 40 CM OUTER DIA AND 6 CM INNER DIA, 3 CM THICK

what Re number does it run at-- velocity -- 50 m/s
reynolds n0:1235500



Also, why are you modelling this rather than imposing its effect as a momentum sink term?
I don understand this question..

The main thing why am i simulating is i need to find the exit velocity components of this geometry and have to give that as entry velocity components for full rotor geometry?.. I have created the complete rotor geometry ..but i am struck with this ..I am not able to get result for this geometry..Since we don have enough computational facility we are simulating rotor and this distortion screen separately...

arunraj September 7, 2011 08:38

2 Attachment(s)
I have meshed this geometries in Icem cfd with volume mesh option and then i have imported in cfx...When i run this case solver shows it has isolated volumes because of domain interface not properly designed... I DON KNOW WHAT AM I DOING WRONG...

Inlet boundary condition is Cartesian velocity components

u=50 m/s
v=0m/s w=om/s

Outlet boundary condition is :: static pressure :: 101667 pa

inteface as :: GGI inteface

REMAINING FACES :: WALL :: FREE SHEAR WALL

I HAVE ATTACHED THE GEOMETRY OF THE ROTOR ALSO

arunraj September 7, 2011 08:47

i HAVE CREATED THIS SCREEN GEOMETRY IN ICEM CFD..i HAVE CREATED IT AS A PART ...i WANT TO KNOW WHETHER I HAVE TO CREATE IT AS A PART OR AN ASSEMBLY TO AVOID THESE KIND OF ISOLATED VOLUME PROBLEMS..yOUR HELP IS MUCH NEEDED HERE..

ghorrocks September 7, 2011 18:56

Quote:

Since we don have enough computational facility we are simulating rotor and this distortion screen separately...
In that case forget about meshing this feature and impose its pressure drop as a momentum source term. This will be much easier and just as accurate (maybe more so). The only reason to directly model this feature is if you want the detailed flow off it, but that does not seem to be what you want.

Quote:

reynolds n0:1235500
OK, so it is high Re number, so turbulent flow with very thin boundary layers. If you wish to model this you need to do a mesh to capture this very thin boundary layer. But as I said above, I would not bother.

arunraj September 7, 2011 23:23

I don no how to impose its pressure drop as a momentum source term?
Could you explain in detail...

Also if I wants to know if we do like that then whether we will get the exit velocity components as a file so that it will be easier for me to give it as an input for complete rotor

ghorrocks September 8, 2011 07:53

Work out a function for the drag of your body versus velocity. You should be able to do by assuming it is a cylinder and using the well known cylinder drag for this Re range. Then read the documentation about source terms and impose the drag on the flow as a momentum source term.


All times are GMT -4. The time now is 13:30.