[worldsheep]

Hello everyone,

I am running a simulation using pimpleFoam with a rotating oscillating mesh. I just found out that the rotating function is hardcoded in "oscillatingRotatingMotion.C" and is:

vector eulerAngles = ampltude_*sin(omega_*t)

Which means that the angular velocity goes as a cos function. This implies that my rotating element starts at full speed which is noti ideal for my application.

Conclusion: I would like to find a solution to start the simulation at rest and increase speed gently. Ideally I wiuld like a cos instead of a sin fot the angular position.

I found the following post:

AMI speed performance

But modifying the source code is not really an option for me.

My final question is... Is it possible to start a simulation at a time different than t = 0s? By properly choosing the initial time I could transform the sin function into a cos (phase shift).

[hxaxtma]

Hi,

1st option: Hard coded: it is possible but you need to adapt the solver. (e.g.: as an example define scalar t_offset -> (sin(t+t_offset))) and so on...

2nd option: Run "moveDynamicMesh" and stop at your wanted endTime of your initial stage. Next, run pimplefoam

Have fun

[worldsheep]

Thank you for your reply, I tried a third option which is a direct consequence of number (2):

3) Create a geometry with the rotor at the position which satisfies angular velocity equal to 0. Then change the name of the 0 folder into the time at which you want the rotor to have zero angular velocity (position is going to be -cos). Finally, change the controlDict and set:

startFrom       latestTime;

This has the advantage of keeping the position at wich angular velocity equals zero unchanged and have the amplitude as a parameter without needing to modify the geometry each time.

For the moment my simulation explodes,... Still trying to understand why