Convergence plot

[fabiano.mannara]

Hi everyone,

I am running the model of a 7-blades axial fan using the "rotating region approach" according to the Autodesk best practice guide: 

phase1: run 103 timesteps of 0.003896 sec (in order to estinguish the transient which lasts for 0.2 sec)

phase2: run 140 timesteps of 0.003896 sec (in order to complete 20 revolution at full speed)

phase3: run 120 timesteps setting 3 degrees/timestep (in order to have one full revolution)

Here the link which explains how to set the timestep dimension and number: 

 https://knowledge.autodesk.com/support/cfd/learn-explore/caas/CloudHelp/cloudhelp/2021/ENU/SimCFD-Us...

Depending on the mesh, I got two dfferents convergence plot:

 

Does anyone know if one of them is right and why? Or are these two both wrong?

Thank you in advance.

Here is the link of the model in case any of you wants to have a look:

https://drive.google.com/file/d/1UombdWVN50GR7OHdDEsy_8C-8C4MfRkm/view?usp=sharing

[frederic.gaillard.7]

Hello, 

Based on your image, your rotationnal speed seems to change over the time. 

What i recommand is a ramp up time  (50 iterations) where the fan start to rest and gradually reach his full speed, to avoid impulsive start.  and finally a constant speed for a long period of time.  Make sure your time is selected to ensure that converge can occurs. 

about the TimeStepSize, 3deg/timestep is a good place to start. 

Hope it helps 

Fred

[fabiano.mannara]

Thank you for your answer,

I set the transient for the rotational speed but maybe it is not set correctly

 

However what do you mean by  "Make sure your time is selected to ensure that converge can occurs"?

[frederic.gaillard.7 · Accepted solution]

Hello Fabiano, 

Your last picture seems to be correct.

What i meant is to define a long time interval where your rotational speed is define to ensure that all the region of interest have time to fully develop (velocities, pressure, temperature, scalar...) 

In your case is 100 second, seems good. 
Good Work 
Fred