CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Flow over a sphere, wich model should i use (http://www.cfd-online.com/Forums/cfx/77095-flow-over-sphere-wich-model-should-i-use.html)

dfmona June 12, 2010 22:09

Flow over a sphere, wich model should i use
 
greetings i need some help
im confuse, i dont know wich turbulance model should i use, mi reynolds number is 12500, de diameter of the sphere is 0.065 m its enclose by a variable area duct, i have the idea that if i wanted to find accurate forces (y direction) on the sphere i should use k-e, but the resulst do not compare to the real model, you see im tryng to model a postion control of pneumatic levitation, is a steady state analysis, y got the real forces, at differnt heighs for different flow velocitys, plus im considering gravity on -y direction.
any ideas?

note: i put the gravity in the bouyancy option, 0 in x and z direction, and -9.8 in y direction is it right?

velocitys vary from 1.6m/s to 4.1m/s

Attesz June 13, 2010 10:20

I think, SST k-w model is well suitable in your case. If you want to get the forces accurate, you should use a fine mesh around the sphere.
gravity settings seem to be OK, but does the gravity have significant effect?

dfmona June 13, 2010 12:39

thanx for the fast reply, refering to the model im gonna show you how is my mesh, the y+ parameter is under 3 i guess thats enough, is it? plz help me decide if the mesh is ok, remember mi only goal is to get acurate forces on y direction, its kinda like drag force.http://yfrog.com/mwmeshp its a 3d model this is a mesh cut plane view, thanks in adanvance

http://yfrog.com/mwmeshp

http://a.yfrog.com/img824/2462/mesh.png
http://yfrog.com/mwmeshp

Attesz June 13, 2010 16:50

hi, yplus is recommended to be under two, or mostly under 1. if you can do, use yplus under 1. otherwise, you should do mesh sensitivity studies to check your mesh-dependency. if you use more finer mesh and the forces doesn not change significantly your mesh is fine enough, otherwise you should use more cells.
anyway, the mesh seems to be good enough to me.

dfmona June 13, 2010 17:32

thanx, yup i know, i have to do a mesh independence study. ok then, im gonna try moving down y+ parameter thanx alot :)


All times are GMT -4. The time now is 16:09.