[Joe Wang]

Hi, FOAMers,
I'm working on a project requiring to use dynamic mesh technique. But it's quite a complicated problem for me and I'm not even sure if it is possible to conduct such a simulation. Hope to get some suggestions and helps here. The problem is described as following.
As shown in the attached sketch, I'm willing to simulate the release of compressed air blocked by an object. From left to right, the object would be lifted by the compressed air at the beginning. Once the pressure is reduced to a value that can not lift the object, the object will fall down (it may not fall back as shown in the sketch). From my point of view, I need to use 6DOF Motion Solver to simulate the motion of the object driven by air force and gravity. What's more, because the boundary detachment/attachment occurs and new blocks may need to be created, I'm not sure how to setup the dynamicMeshDict file, as well as other related files.
Does anyone ever work on a similar problem? Really appreciate if someone can help me.
Joe

P.S. To simplify the model, the object can be modified as a trapezoid (shown in the right bottom corner in the sketch).

[Tobi]

Dear Wang,

out if the box I would say, it is possible but not with the standard libraries. So you have to make some new library which accounts for removing, adding cells (or new meshing+mapping during the simulation based on some criteria - checkMesh).

ICON did something like that but not with the 6DoF library but with the normal motion one. As you are a PhD, I would say it is a half PhD project to get everything work.

Keep in mind to take the fluxes etc conservative.

Good luck.

Sent from my HTC One mini using CFD Online Forum mobile app

[johannes]

Hi Joe,

that's really an interesting case! I've been working on a similar one, at least when it comes to the problems (or challenges ).
First of all, as Tobi already mentioned, you'll need some kind of remeshing, because of the large motion of the FSI interfaces. Regarding the detachment/attachment aspect, I don't think that this will work in the intended way. Instead, you'll have to keep all the blocks connected through a small gap, which should be as narrow as cell skewness allows it. To stop the leakage through this gap, it may be necessary to apply a porous resistance in the gap and then reduce the porosity while the gap opens during the motion.
This is also the way I chose in my setup. I'm currently using Fluent coupled with ANSYS but I would prefer to do the same kind of simulations in OpenFOAM. So I would be happy if you could keep me updated on your progress with OpenFOAM.

Best regards,
Johannes

[Joe Wang]

Hi, Tobi,
Really appreciate your reply and showing me a way to go.
I'm almost done with my program so I may not try to create new libraries for it immediately. Instead, I will try to simplify the model and see if it works. But I'll continue to work on this problem then. Thank you.
Jue

Dear Wang,

out if the box I would say, it is possible but not with the standard libraries. So you have to make some new library which accounts for removing, adding cells (or new meshing+mapping during the simulation based on some criteria - checkMesh).

ICON did something like that but not with the 6DoF library but with the normal motion one. As you are a PhD, I would say it is a half PhD project to get everything work.

Keep in mind to take the fluxes etc conservative.

Good luck.

Sent from my HTC One mini using CFD Online Forum mobile app
Like
__________________
Keep foaming,
Tobias Holzmann

[Joe Wang]

Hi, Johannes,
You offered me a great alternative to solve this problem with porous media. Thank you so much.
I need to get it work as soon as possible so I may try some simplification first to see if it works. I'll continue to work on this project then.
Let's keep updating this topic or new ones if there's new progress.
Regards,
Joe

Originally Posted by johannes

Hi Joe,

that's really an interesting case! I've been working on a similar one, at least when it comes to the problems (or challenges ).
First of all, as Tobi already mentioned, you'll need some kind of remeshing, because of the large motion of the FSI interfaces. Regarding the detachment/attachment aspect, I don't think that this will work in the intended way. Instead, you'll have to keep all the blocks connected through a small gap, which should be as narrow as cell skewness allows it. To stop the leakage through this gap, it may be necessary to apply a porous resistance in the gap and then reduce the porosity while the gap opens during the motion.
This is also the way I chose in my setup. I'm currently using Fluent coupled with ANSYS but I would prefer to do the same kind of simulations in OpenFOAM. So I would be happy if you could keep me updated on your progress with OpenFOAM.

Best regards,
Johannes