Problem with moving boundary
Hi,
I am trying to model a simple test case: 2D lid driven cavity flow with no gravity.
The boundary on top is modeled as a moving boundary with constant velocity, while there is a periodic open boundary on the x-axis. For more details there is a picture of the first simulations step:
http://www.directupload.net/file/d/4344/7rjstx5k_png.htm
Now there is a problem. The interaction between the moving boundary an the fluid doesn't work very well. In the beginning the fluid starts to get a velocity, but as the simulation keeps going a gap is formed between boundary and fluid. As a result there is no interaction in velocity between boundary and fluid anymore:
http://www.directupload.net/file/d/4344/d7qnbjnu_png.htm
I tried to fix this behavior with the different options:
-reduce dp
-viscosity treatment (artificial or laminar); different values
-increase smoothing length
-shifting options
Nothing helped so far... Are there any options left to influence the interaction between a moving wall and a fluid ?
Is there a other way to assign a given velocity to a boundary?
Is there an option/keyword in xml which allows to assign a velocity to a boundary particle, but the boundary particle stays in the same place? Or is this something which requires a change to the code?
Regards,
fko
I am trying to model a simple test case: 2D lid driven cavity flow with no gravity.
The boundary on top is modeled as a moving boundary with constant velocity, while there is a periodic open boundary on the x-axis. For more details there is a picture of the first simulations step:
http://www.directupload.net/file/d/4344/7rjstx5k_png.htm
Now there is a problem. The interaction between the moving boundary an the fluid doesn't work very well. In the beginning the fluid starts to get a velocity, but as the simulation keeps going a gap is formed between boundary and fluid. As a result there is no interaction in velocity between boundary and fluid anymore:
http://www.directupload.net/file/d/4344/d7qnbjnu_png.htm
I tried to fix this behavior with the different options:
-reduce dp
-viscosity treatment (artificial or laminar); different values
-increase smoothing length
-shifting options
Nothing helped so far... Are there any options left to influence the interaction between a moving wall and a fluid ?
Is there a other way to assign a given velocity to a boundary?
Is there an option/keyword in xml which allows to assign a velocity to a boundary particle, but the boundary particle stays in the same place? Or is this something which requires a change to the code?
Regards,
fko
Comments
However in your case you may be creating a very large tank to move the top part... Please can you send to us your XML (dualsphysics@gmail.com) and we will take a look to find a better solution to reduce that strange gap?
Regards
does anybody have a solution for this problem yet? I just ran into exactly the same situation as Fkornmei.
Thank you and best regards