CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   grid generation problem (https://www.cfd-online.com/Forums/fluent/29578-grid-generation-problem.html)

vineet April 3, 2002 05:46

grid generation problem
 
i am solving fluid (water) flow past circular cylinder (2D). i am genertaing the grid in gambit. i want my grid to be logrithmic i.e dense near the body and spreadout as we move away from the body as i need better resolution near the body surface. Can anyone suggest how to do this. Thankyou.

Laika April 3, 2002 09:53

Re: grid generation problem
 
no problem: you first mesh the edge that defines your cilinder. Do it with a fine spacing. Then you use a pave scheme with a larger setting for the spacing for the entire surface. Your meshed edge will force the mesh to be finer near the cilinder. Play around with the settings to get a feel of it.

Other possibility: Use (Gambit2.0) the sizing functions.

You can also apply a boundary layer on the surface of your cilinder to obtain a good mesh near the surface to capture the high gradients in that region (recommended!)

greetings,

Laika, still orbiting

Chetan Kadakia April 3, 2002 16:04

Re: grid generation problem
 
Laika, is right. But if you haven't done this before it can be tricky. Seperate your domain around the cylinder into 4 areas. Mesh the edges (I believe there is a logarithmic biasing function availabe when meshing edges), Then mesh each face.

Can I ask why you want to use a logarithmic scheme?


vineet April 3, 2002 22:33

Re: grid generation problem
 
Chetan & Laika,thanks for your response.

Chetan, i wanted to use logarithmic scheme because i have myself wrote a programme for soving this problem there i transformed cartisian coordinates into logrithmic spherical coordinates to obtain rectangular domain with closer spacing near the cylinder surface something like x=theta and y=log(r), where theta is angle and r is radial distance. i just wanted to see if such grid was possible. In anycase it need not be logarithmic only it should dense near surface. i will try what u and Laika told. bye


All times are GMT -4. The time now is 16:18.