RANS Modeling- Estimating K and Epsilon
 

Hi everyone,

I'm working on a summer internshp program and need to estimate k and epsilon from two sources of data.

The first source is a hotwire which has been placed in a channel.

The second is a pipe in which I know the velocity throughout.

I've estimated the first by using the Bousinesq hypothesis to relate the turbulent dissipation energy to the turbulent kinetic energy. The relationships are

k=1/2(u'^2+v'^2'w'^2), estimating w' to be the average of u' and v'.
epsilon=du/dy*cmu*k^2/(u'v')

For the second, I used empirical relationships from the following site;
http://www.sharcnet.ca/Software/Flue...ug/node238.htm

This got me k and epsilon values at the inlet of the pipe flow.

If anyone more familiar with CFD theory could have a look at my derivation/results for k and epsilon, it would be incredibly helpful. I'm afraid my solutions aren't converging with reasonable results for k and epsilon. k tends to drop to machine zero, which obviously isn't valid.

Attached is a link to the spreadsheet where I worked with the data
https://dl.dropboxusercontent.com/u/...elp/sample.ods

Any help would be appreciated,
Thanks,
Matt