[valr]

hi everyone,

I am trying to simulate a movement of fluid inside a rotating drum (with baffles). I am using ansys 18 package .

When I change the domain option to "rotating" the liquid inside it starts to rotate but not the drum.I concluded this by looking at the baffles being stationary. I even tried changing the "wall rotating" option at the boundary.But it is not working.

Can anyone help me overcome this problem......

[ghorrocks]

What frame of reference are you looking at? If you are looking in the rotating frame of reference the drum will appear stationary.

[valr]

hi,

Thanks for your reply ghorrocks. I trying to view from boundary frame.

I just wanted the drum to rotate so that the I can analyse the fluid motion inside it. So, I changed the mesh deformation in default domain menu and kept mesh motion as "parallel to boundary " in boundary details menu. I also tried "surface revolution" as mesh motion option in boundary details . But it also didn't work.

[ghorrocks]

Are you using mesh motion to do this? If the motion is simply rotation then you should model this as a rotating frame of reference. It will be much quicker and more accurate.

[valr]

hi,

Thanks for the reply.

Yes,I am using mesh motion to do this. I am confused on which one to choose among the different choices available in Mesh Motion menu in boundary details option for this case.

Can you help me out in this. Can you also suggest if there are any other easier method to rotate the drum.

[urosgrivc]

1
If you chose mesh motion (you will rotate the whole mesh and your gravity shal stay fixed as everithing will rotate)
+If jou will make a video aftervards the drum with ewery wals will actualy rotate.
I am not sure but i think this will take some more computational time as all the nodes have diferent (xyz) values at each timestep.


2
if you use rotating frame of reference everithing will say put (nodes xyz fixed) and you will have to rotate gravity around ((sin(t),cos(t)) on two axis and 0 on third as you need to specify (xyz)values for gravity)
in this case wals will by default be rotating already
+If jou will make a video aftervards the drum will stay put and only the fluid will be slushing around.


I dont know much about your problem so this is just a first thought.
It can ocourse be done it is just a mater of seting it up corectly

the 2nd way of doing it is definitly more robust way of doing it, and espetialy as cfx is wery good with rotating frames of reference.
If you are a new to cfx I would sugest 1st aproach rotating mesh motion becouse there is not much to screw up in the proces and the wideo you will export at the end will be extremly awesome and jou will want to do more cfd stuf in the future.
Do you want to get some kind of results out of the simulation why are you doing it??

[blitzkg]

I am working on the same case. We have water(red) and air(blue) inside a cylindrical drum with baffle as shown (pls refer to the image).
when i rotate the cylinder i want to find out how the motion of the mixture is.

I rotated both Domain and boundary wall at same rpm (with mesh deformation and mesh motion "parallel to boundary") and solved, but in the animation the baffle is not rotating at all. I can see the fluids move, but the baffle is not moving.

Is there a way to make the baffle move?

PS: I am working on ANSYS CFX.

[ghorrocks]

Please attach your CCL or output file.

[valr]

sorry sir, i couldn't understand what do you mean by a CCL file.
Do you mean output video as the output file?

[urosgrivc]

CCL (Cfx Comand Language)is a text file which holds all the data or setings of the simulation, basicly instructions for the solver,
based on that it is posible to (an expert can) determine where the problem of your simulation is.

[blitzkg]

Please see the CCL file.

[ghorrocks]

Or the output file will do. It includes the CCL at the top of the file.

[ghorrocks]

Thanks, I understand what you have done now. I do not know why it does this but it appears that if it only has a rotating domain then it appears to be challenging to get the visualisation to show a rotating mesh. The simulation appears to be correct, but it is modelled in the rotating frame of reference which makes it difficult to visualise.

You might need to put a stationary frame of reference in as well and link the two together with a transient rotor stator GGI interface so you can view the geometry in a stationary frame of reference.

[blitzkg]

Where are we supposed to change the frame of reference to stationary? Is it the Domain motion?

[ghorrocks]

No, I did not say change the rotating frame to stationary. I said add another domain, this one being stationary and connected to the existing rotating one by a TRS GGI interface.