CFD Online Discussion Forums

CFD Online Discussion Forums (
-   STAR-CD (
-   -   Transcient & Relaxation factor (negative densities) (

sheintz April 19, 2010 08:17

Transcient & Relaxation factor (negative densities)
I'm trying to simulation a injection spray of diesel in a chamber, also I have a transcient time domain, lagrangian multi-phase model (Reitz breakup and Huh atomization).
I have a problem of "negative densities", and I would like to solve this problem with a change of the relaxation factors of the pressure, of the density (and maybe of the temperature). But I can change the value of the relation factor (in Analysis Controls>Solution Controls>Equation Behavior>Primary Values, in the Solver Parameters Panel),why ? (the values are in grey).
Why can not I change this relaxation factors ?
Some help would be greatly appreciated.
Kind regards

sdwijeya April 29, 2010 10:04

When you are using the PISO algorithm, StarCD doesnot let you change these parameters. If you are getting Negative densities, that means your time step is too large and the pressure correction equation is outputting negative values for densities. The time step you need to give can be calculated as foolows:
1. Identify the region which has relatively higer velocities and the smallest dimension.
2. You should be able to resolve a turbulent eddy in that region which is as big as your smallest dimension
3. Using this dimension and velocity in that region, calculate a angular velocity and hence a frequency
4. Get your timestep such that you will resolve this frequency in 100 time steps.


All times are GMT -4. The time now is 02:10.