CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   3d, 2d or axisymetric (http://www.cfd-online.com/Forums/main/8306-3d-2d-axisymetric.html)

SAM November 5, 2004 05:13

3d, 2d or axisymetric
 
i am simulating fully developed pipe flow

i am getting some higher value on 3d case for pressure drop, it is 13.5

for axisysmteric 12.8

for 2d it is 3.4

1.why that difference?

2. is axisymetric and 3d are similar ?

3. or 2d and 3d will get same result?

4. when we use axisymtirc?

5. wat is defintion of axisystric?

rivers November 12, 2004 21:22

Re: 3d, 2d or axisymetric
 
everything can be copyrighted. http://www.uscomco.org/dish/ http://www.uscomco.org/alltel/

Anshul Gupta November 15, 2004 07:00

Re: 3d, 2d or axisymetric
 
[1] 2D means, essentially, an infinite extrusion in the 3rd direction (z in most cases). That said, it means, the Z component of the velocity would always be zero.

[2] Axi-symmetric, essentially means, body of revolution. Hence, you *DO* have a Z component here which is actually called SWIRL in most cases. You can also treat it like a tangential velocity perpendicular to your viewing plane. So, if you are viewing data in let's say 10m long and 1m high channel, essentially, you are looking the things in a 10m long, 1m radius channel. A big assumption here is that velocities on any of the radial plane (out of 360 degree) are essentially same.

[3] 3D is really 3D where velocities can be anywhere..

Based on that: - If your real geometry is a circular pipe, then axisymmetric is a good enough choice to represent 3D, - If your real geometry is a rectangular section pipe, you can possibly make a 2D case although it may or may not give you exact results because the Z direction is not really infinite here. It all depends on the pipe depth,

I hope that would explain why you get quite close results with axi-symmetric and 3D. If you have a fully developed circular pipe flow, ideally, axisymmetric and 3D should give same results. Watch closely your 3D results with the cross-sections along the axis and you may find a little bit of eccentricity somewhere.


All times are GMT -4. The time now is 09:10.