Register Blogs Members List Search Today's Posts Mark Forums Read

 February 23, 2009, 23:43 advection time scale #1 Tanveer ahmad Rubani Guest   Posts: n/a 1.what fraction of Length/Velocity scale should be used for iving value of Physical Time scale in Solver control of ANSYS CFX Pre. 2.Why Physical Time scale value should be less than advection time scale as is mentioned in help for divergence.The advection time scale appears in solver out file under average time scale information.How is advection time calculated. 3.How sensitive is problem convergence to time step. Is is possible that a problem never converges with auto time scale value but will converge with Physical Time scale value.

 February 24, 2009, 16:16 Re: advection time scale #2 Glenn Horrocks Guest   Posts: n/a Hi, I assume you are talking about a steady state simulation. 1. One is a starting point but many simulations can use much higher than that. It is problem dependant. 2. Smaller timesteps are less prone to divergence. However once the flow has started and the basic flow pattern is established you can generally use a much bigger timestep. 2a. Advection time is a length scale/velocity scale. The length scale is (I think) the cube root of the total volume and the velocity scale is some sort of average velocity. 3. Problem dependent. Simple simulations are not very sensitive, but numerically unstable simulations can be very sensitive. 3a. Yes. Auto time scale is just a guess, but with physical time scales you can set it to be as high or low as is necessary. Glenn Horrocks