Length and Velocity Scales

samwh · May 22, 2009, 02:03

Hi,

I am trying to simulate a submarine travelling just bellow a free surface using ANSYS CFX. I have been getting some unrealistic results and believe it is because of the time step I am using.

I would like to check it aginst the formula in the CFX users manual i.e. 1/3 of Lenght Scale on the Velocity Scale.

I have not come across a way to determine either to velocity or lenght scale could some one explain what they are and how to determine them?

I would appreciate any help

Cheers

triple_r · May 28, 2009, 17:33

Hi,

I hope you have solved this problem already, but just in case someone else has a similar question following is a brief description of length and velocity scale.

The condition for time steps (known as CFL condition) is for explicit transient solvers. As CFX treats steady problems as false transient problems, the time steps specified for steady problems must satisfy this condition.

CFL condition implies that if we consider a fluid chunk it will not travel more than a grid division during a time step. That fluid chunk will travel a distance equal to its velocity times the time-step during a time-step, so

(velocity * time step) must be less than (grid size)
or
(time step) must be less than (grid size / velocity)

If you are as conservative as I am, you can use the maximum velocity as your velocity scale, and the smallest grid size as your length scale. Considering that usually you need very fine grid near the walls, and velocity is very low there, so using the extremes can give a very small time step size.

Also the finer the mesh, the smaller the required time step would be.

I hope this information can be helpful.

May 22, 2009, 02:03	Length and Velocity Scales	#1
samwh New Member Sam W Join Date: May 2009 Posts: 4 Rep Power: 16	Hi, I am trying to simulate a submarine travelling just bellow a free surface using ANSYS CFX. I have been getting some unrealistic results and believe it is because of the time step I am using. I would like to check it aginst the formula in the CFX users manual i.e. 1/3 of Lenght Scale on the Velocity Scale. I have not come across a way to determine either to velocity or lenght scale could some one explain what they are and how to determine them? I would appreciate any help Cheers

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Terrible Mistake In Fluid Dynamics History	Abhi	Main CFD Forum	12	July 8, 2002 09:11
what the result is negatif pressure at inlet	chong chee nan	FLUENT	0	December 29, 2001 05:13
time averaged heat transfer in oscillating flow	Matthieu Ubas	Main CFD Forum	2	November 5, 1999 14:20
length scales at inlet for internal flows	Anne-Marie Giroux	Main CFD Forum	3	July 5, 1999 21:28

May 28, 2009, 17:33		#2
triple_r Senior Member Reza Join Date: Mar 2009 Location: Appleton, WI Posts: 116 Rep Power: 17	Hi, I hope you have solved this problem already, but just in case someone else has a similar question following is a brief description of length and velocity scale. The condition for time steps (known as CFL condition) is for explicit transient solvers. As CFX treats steady problems as false transient problems, the time steps specified for steady problems must satisfy this condition. CFL condition implies that if we consider a fluid chunk it will not travel more than a grid division during a time step. That fluid chunk will travel a distance equal to its velocity times the time-step during a time-step, so (velocity * time step) must be less than (grid size) or (time step) must be less than (grid size / velocity) If you are as conservative as I am, you can use the maximum velocity as your velocity scale, and the smallest grid size as your length scale. Considering that usually you need very fine grid near the walls, and velocity is very low there, so using the extremes can give a very small time step size. Also the finer the mesh, the smaller the required time step would be. I hope this information can be helpful.