Transient Analysis

krrishcfd · July 4, 2011, 07:52

Hi every one,

I would like to carry out transient analysis for a centrifugal pump consisting of Elbow, Impeller and Volute can any body help me out in defining the time step and various other parameters that are to be considered for this analysis.

Thanks in Advance.

ghorrocks · July 4, 2011, 08:04

In concept the approach is simple. Guess a time step and get a solution. Then halve it and compare the results to the previous simulation. If the results are the same to within a tolerance you are happy with then your time step is OK. If not accurate enough then halve the time step again and continue.

A faster way to find the timestep is to use adaptive timestepping, converging to 3-5 coeff loops per time step (or 5-10 if a complex model such as multiphase). This generally gives you a pretty good time step size.

krrishcfd · July 4, 2011, 08:08

@ghorrocks: thanks 4r ur reply .. i vll try thz way ...

July 4, 2011, 07:52	Transient Analysis	#1
krrishcfd New Member KrishnaMohan.Siddu Join Date: Mar 2009 Location: Pune Posts: 8 Rep Power: 17	Hi every one, I would like to carry out transient analysis for a centrifugal pump consisting of Elbow, Impeller and Volute can any body help me out in defining the time step and various other parameters that are to be considered for this analysis. Thanks in Advance.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
2-D airfoil transient analysis	mike	FLUENT	1	July 15, 2017 04:33
FSI Transient Thermal Analysis	Abduri	CFX	1	January 1, 2017 22:54
Transient analysis with turbFoam very high Courant nr	hardy	OpenFOAM Running, Solving & CFD	13	November 19, 2008 13:36
Transient analysis	Andy F	CFX	8	January 6, 2006 09:55
transient analysis of silencer using cfx.	mehul	CFX	0	October 16, 2005 04:48

July 4, 2011, 08:04		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	In concept the approach is simple. Guess a time step and get a solution. Then halve it and compare the results to the previous simulation. If the results are the same to within a tolerance you are happy with then your time step is OK. If not accurate enough then halve the time step again and continue. A faster way to find the timestep is to use adaptive timestepping, converging to 3-5 coeff loops per time step (or 5-10 if a complex model such as multiphase). This generally gives you a pretty good time step size.

July 4, 2011, 08:08		#3
krrishcfd New Member KrishnaMohan.Siddu Join Date: Mar 2009 Location: Pune Posts: 8 Rep Power: 17	@ghorrocks: thanks 4r ur reply .. i vll try thz way ...