CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Physical timestep Calculation (http://www.cfd-online.com/Forums/cfx/22802-physical-timestep-calculation.html)

Joe July 4, 2006 11:09

Physical timestep Calculation
 
I am making a simulation of a wigley hull, involving a free surface between air and water. While trying to make the simulation, I defined the physical timestep as in the CFX manual but it doesnt seem to work.

A wall is positioned at the outlet and an irregular flow is also seen there. By inputting a smaller timescale the simulation progresses kind of better but still there is a problem.

The size of the domain is big enough for the simulation to take place without any problem but i think the timescale is not correct.

Can someone tell me a way to calculate the timescale?

Regards Joe


Joe July 4, 2006 11:16

Re: Physical timestep Calculation
 
Read the Best Practices Guide for Multiphase flow in the manual appendix.

James Date July 4, 2006 13:31

Re: Physical timestep Calculation
 
Have you tried running a steady state simulation yet? You should get a converged solution using this method. If you do run the transient case, make sure the time step satisfies the Courant requirement being < 1, based on the wave speed. Info about this can be found in the CFX manual. If you are using CFX-10, download the latest CFX-11 preview from www.ansys.com because it has improved numerics which help to reduce the spurious waves which can occur in the far field when using CFX-10. Are you validating your Wigley results against those obtained by Tokyo University in their tank tests?

James


Joe July 4, 2006 15:07

Re: Physical timestep Calculation
 
I am running steady state simulation with CFX 5.7. Some results showed me good approximations with the experimental even though the simulation did not completely converge.



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