CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Full Multigrid Initialization (https://www.cfd-online.com/Forums/fluent/168199-full-multigrid-initialization.html)

Mr.Goodcat March 17, 2016 06:43

Full Multigrid Initialization
 
Good morning, everyone.:)

I've recently started to learn how to use Fluent, and i'm currently trying to simulate for my thesis the flow field for a helicopter rotor in hover.
(Steady-state, pressure based coupled solver, kwSST turbulence model, with multiple moving reference frames)

I've received opposing suggestions regarding the method to use to initialize the solution, in particular concerning the use of hybrid initialization versus a combination of standard(or hybrid) initialization and FMG initialization.


I've read the User's Guide and understoond more or less how the full multigrid initialization works and how to set it. Also, i've been told that setting residuals of 10e-3 is more than sufficient for FMG init and that even if they aren't exactly reached in the end, the initialization can be considered successful, as long as the values were decreasing more or less steadily.


What i don't quite grasp is how i should behave when, after having done the hybrid initialization (or the standard one: i tried with both), the FMG doesn't converge to 10e-3 on all the grid levels, and moreover for the levels non converged i get high oscillations and no noticeable decreasing trend.

Does that mean that there is something wrong with the way i set the model and the solution parameters, and that i should re-check my model or start "tinkering" with my settings? :confused:

Should i leave behind the FMG and just try a few simulations with only the hybrid initialization?

Many thanks in advance for any help you may provide!:)


-------------------Just thought about it:

I'm starting my simulation with a rotational speed of 4rad/s and planning to increase it every "few" iterations, until i reach the nominal speed for my simulation of 40rad/s.
My mesh was created (the first cell height was calculated) to obtain a y+=1 at the nominal speed of 40rad/s. Could that be the reason for which the FMG initialization won't converge?
If so, then how should i proceed? Should i simply use the hybrid one, since i'm not working anyway with my final intended speed?


All times are GMT -4. The time now is 22:08.