Community
CFD Forum
Welcome to Autodesk’s CFD Forums. Share your knowledge, ask questions, and explore popular CFD topics.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Free surface in orbiting glass

9 REPLIES 9
SOLVED
Reply
Message 1 of 10
bulnes
1031 Views, 9 Replies

Free surface in orbiting glass

I have been trying to set up my simulation withouth much succes. How can I simulate a glass "half filled with water" that is moving in an orbit. Just like earth around sun but without the rotation only translation.

The setup I have tried  is as follows:

Loaded geometry from Inventor. It consists of the plastic vessel, and 2 volumes of fluid inside one on top of the other. The lower one is the height of the fluid at starting point.

I assign material to the walls as solid,  ABS plastic. Fluid water to both internal volumes.

Glass with waterI assign a HoF to the lower volume as initial condition.

To avoid loosing to much time I leave the mesh with all the automatic settings. It will be coarse but can refine it later when this works.

 

I apply a combined orbital rotational motion to the solid walls of the glass.

 

When I run the simulation the solid moves perfectly but the fluid stays in its place just trespassing the glass walls (obviously un-realistic) and then simulation diverges. I want the walls to move the fluid. I have also tried applying symmetry/slip boundary conditions to the surfaces of the water volume in contact with the solid but still no success.

Fluid stays in place while the solid moves around

 

Is this simulation possible in Autodesk CFD 2015? SimTV has really been no help at all and I can not find any tutorial or manual regarding flow driven by the solid (Setup for an impeller is totally different), all tutorials are about flow driven motion. I have run this simulation in another CFD software but can´t seem to make it work in autodesk CFD. Any advice?

Thanks in advance.

Best Regards,

 

 

 

 

9 REPLIES 9
Message 2 of 10
OmkarJ
in reply to: bulnes

Okay, this is interesting. Let me give it a go...

 

Imagine that your empty glass is tied to a rope and you are swinging it horizontally in a circle. Is this what you want to simulate?

Beware that the glass WILL rotate around itself in the same time it takes to do one revolution around you. Kind of Earth-Moon situation.

 

If that were the case, you will need to include body force in addition to gravity force on the water in momentum equation, essentially by specifying accelerations. 

While gravity is easier to specify, the other acceleration water is facing is centrifugal acceleration away from you. And since periods of rotation and revolution of glass are same, in the frame of reference of the glass, for an ant swimming in the glass, it will always be in one direction.

 

The centrifugal acceleration is omega X V, with V= omega X r

where

omega= angular velocity

V = tangential velocity

r= radius vector

 

Since all these vectors are perpendicular, centrifugal acceleration is r*omega^2.

 

Apply this in one of the directions perpendicular to gravity, and you can see what our swimmer ant is seeing INSIDE THE GLASS.

May not work though if you want to see a fancy animation with glass rotating and water spilling out!

Message 3 of 10
bulnes
in reply to: bulnes

Thanks for the answer but unfortunately it is not what I need to solve for. I might not have explained myself correctly. I am posting copy from the Autodesk knowledge Network defining the kind of motion I need to simulate in this case.

"Orbital motion is the circular displacement of an object about an axis. The orientation of an object in pure orbit (with no rotational component) does not change. This is shown below:

 

The Autodesk info can be found here.

So if I simulate the glass tied to a rope it will be always facing the same side, like you said earth-moon movement. So if I spin this glass the water will start to go to the further side and stay there due to centrifugal centripetal forces.

In the case I want to simulate, how would the boundary conditions be defined. As far as I know in Autodesk CFD the only option for a wall condition is slip/symmetry. When I run the simulation without orbit just rotation over its own axis and applying boundary condition of rotation to all surfaces of the fluid, and suppressing the solid, everything runs good and I get a nice HoF line and curvature due to rpm effect. I dont seem to find how to give an orbital movement to these surfaces. If I could do this then maybe I dont need to include the solids at all in the meshing and simulation. Any idea?

Best regards and thanks again for your time.

 

 

 

 

Message 4 of 10
bulnes
in reply to: bulnes

Any ideas on this one, anyone?

Message 5 of 10
apolo_vanderberg
in reply to: bulnes

If you watch the SimTV on Free Surface, I posted, you will see a moving / tilting glass (in 2D) as an example of Free Surface with motion.

As I mentioned in the video - an Automatic sized mesh may not be enough, we will need a number of nodes through the thickness of the solid (3-4).

As far as the motion in general, typically you will need to have a domain that is large enough to encompass the 'potential' areas that contain the solid and fluid.

So for your example you will want an external domain that can capture where the glass would move with the liquid within it, then ensure the mesh is appropriate for capturing the motion as well as the translating fluid. The rest of the domain can have a more coarse mesh
Message 6 of 10
bulnes
in reply to: apolo_vanderberg

Thanks for you answer, I will try setting up the problem as you advice. I will create an outside volume that will cover the full motion range. Do you think the problem of the fluid staying in place and not following the solid will be solved by this new setup? Is the 2D example available as a tutorial file or something similar?

 

Best regards,

David

Message 7 of 10
bulnes
in reply to: apolo_vanderberg

Hallo Apolo,

Well I run the simulation again, this time with a round vessel, which includes 2 volumes of water, the lower one is the HoF, also a external volume was created big enough to include the translation orbit the vessel will follow. I Know the mesh is really ugly, huge and un-refined, but my problem here is what you can see on the attached picture. The water inside the vessel remains in the same place and the solid vessel is already moving around the defined path.

Why are these "solid" walls not retaining the liquid inside.

You can see the only zones with  velocity is where the solid is going through the liquid or around the vessel walls. If you have any idea on how to fix this or what I am doing wrong it would be greatly appreciated. I already found another way of performing the simulations without the solids but I do not like it as the wall roughness can not be defined if I do not have the solids wall (as far as I know.) and the solutions do not really agree with the real experiments until now.

vessel-orbitaltranslation-solids-withwatervolume.jpg

eeeee.jpg

Message 8 of 10
apolo_vanderberg
in reply to: bulnes

Yes that mesh is not great and that will be the culprit.

If you watched my SimTV video you would have noticed I specifically mention that the mesh HAS to have 3-4elements through the thickness of the solid to 'hold water'

This also means that the fluid domain (where the solid will be moving and masking) will also require to have a small enough mesh to maintain this.

 

I would highly recommend trying all of this in 2D to learn before stepping to 3D.

 

Message 9 of 10
bulnes
in reply to: apolo_vanderberg

Hallo Apolo,

Thanks for your quick answers it is really helpful as no one had any idea about it.

Only one thing, I am sorry but I do not seem to find the video where you state that you need 3 to 4 elements on the solids, I have watched two of the videos referencing free surface several times, one is "Free Surface Flow in CFD" and the other is "Results Available for Free Surface Flow" if I am missing one of them maybe that is the whole reason.

From those two videos I understood that the mesh where the liquid is going to move needs to be very fine to capture detail and continuity, time 03:28 and 4:07 from the Free surface Flow in CFD video, but  it is never mentioned that the mesh has to have 3-4 elements through the thickness of the SOLID to 'hold water'.  I was expecting setting a volume as solid would allow it to hold water. For example in the tutorial of the valve the poppet is meshed only to a "adequate" size for motion analysis maybe due to the form and size it has more than 3 to 4 elements but how am I supposed to know this.  I have gone through all the material in the help files, and can not find reference to needing the walls or solid parts to have more than 3 to 4 elements accross to be water tight.

Point 5 of setting up Free surface simulation tutorial only states this: "Mesh all fluid parts and parts that the liquid will occupy, even partially. The mesh requirement for free surface simulations is high. A fine mesh is required in regions where the liquid and air interact. Be sure to define a fine mesh distribution at the liquid-gas interface as well.".

Maybe I am missing a part of the manuals, tutorials that are available? If so I would very much appreciate being pointed out where to obtain them so I quit posting so simple problems. If it not stated It would be also very nice to add it to the help files and tutorials in a more easy to find way.

This is the link to the autodesk simulation TV I have access, if it is not the original or it is outdated please let me know.

http://www.imaginit.com/software-solutions/solutions-simulation/autodesk-simulation-tv

 

Thanks again for helping me on this problem,

 

Best Regards,

 

David

 

Message 10 of 10
apolo_vanderberg
in reply to: bulnes

In the Help - http://help.autodesk.com/view/SCDSE/2015/ENU/?guid=GUID-C8FB4A11-2A9D-4DB8-80E5-1D0F050ED4B8

We talk about how our motion masking works.
Too coarse and we get bleed through of the flow solution (and this is for typical flow models not Free Surface).

Free surface will require more than this

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report