CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   deciding time duration in transient simultion (http://www.cfd-online.com/Forums/cfx/98113-deciding-time-duration-transient-simultion.html)

 moniker March 3, 2012 09:43

deciding time duration in transient simultion

I am doing a transient simulation for my model ,
I have decided to use time steps as 0.0632 seconds,how do I decide the time duration ?I first gave time duration as 65 and time steps as 0.0632 but the
solver ran continously ..has'nt converged at all

 ghorrocks March 4, 2012 05:05

How have you decided 0.0632s is the right time step size? If it is not converging it probably is not the right size.

And as for the time duration - what are you modelling? An inkjet printer ejection (total time about 10 microseconds), or magma motion in a magma chamber (turn over time 140 million years). I have done work in both and guess what? They both use different total simulation times.

 moniker March 4, 2012 07:17

I am modelling the flow of air through the intake manifold of a 4cyl diesel engine ,naturally aspirated ,upto before valve..now I calculated timestep by taking the whole length of manifold as distance and inlet velocity of air , that value I took multiplied by (1/20) to get 0.0632 sec ..my length is 0.349 m and intake velocity is 27.61m/s ..1/20 because that is the time required for a particle of fluid to traverse the length of object

 ghorrocks March 4, 2012 17:41

Hmmm, I don't think you have much idea of how to set a time step :)

My PhD thesis was on IC engine modelling. Have a look at the time step size I used:
http://hdl.handle.net/2100/248

 moniker March 5, 2012 08:47

Thanks for that thesis..i have got new doubts...how is the CFL stability criterion related to timestep..and I am not able
to comprehend what a timestep is and what are the parameters that affect it.

 ghorrocks March 5, 2012 18:16

Quote:
 how is the CFL stability criterion related to timestep
CFX is an implicit solver so it is not relevant.

Quote:
 and I am not able to comprehend what a timestep is and what are the parameters that affect it.
If you do not understand the basics you do not have much hope of doing a useful analysis. You need to do some studies or reading into numerical methods and CFD before you can start doing simulations.

 All times are GMT -4. The time now is 14:04.