moving tank
 

Hello,

I want to simulate a partly filled tank (free surface) moving with a constant velocity, that brakes after certain time.

Is it possible to define a course of the tank (time-dependent velocities in all 3 directions (maybe rotations, too))?

I only found the simple definition of constant velocities in 'define boundary conditions'->'define boundary regions'.

Re: moving tank
 

hi, try using the table to give the time dependent velocities.

Re: moving tank
 

hi,

Thanks. I defined the velocities of the walls, but the result just shows nothing but two zones (light and heavy fluid) not moving. There is no sloshing.

Maybe my approach is not appropiate:

The tank moves with a constant velocity v (in x). Then brakes and the heavy fluid sloshes in the tank.

So I built up a simple tank (see Tutorial 15, without obstacle and refined mesh) and I defined moving walls by creating a table for boundary region "0 Type Wall Default...".

After calculation I see only a still tank.

Has anybody an idea?

Re: moving tank
 

you should do this in star-ccm+, you can have a macro where v=f(t) or a rigid body motion where the translation velocity is function of time. There is even a tutorial I think. You can use star-cd and star-ccm+ with the same license and it is easy to install. It will be the right code for this application.

L

Re: moving tank
 

Hi Luck,

thanks for the hint. I will try.

Re: moving tank
 

Hi, me again.

Well, I now found an example of what I want to do:

http://www.adapco-online.com/adapco_...frontpage.html

Click on "Animations" (left-hand side) and select "Fuel Sloshing...".

I think, this model is produced with Star-CD (STAR-CCM+ is not available on my cluster). How do I have to define the boundary regions to get a moving/shaking tank like that?

What I tried until now: I assigned a velocity-table to the "0 wall default boundary region" with no result. And I created a new wall with an assigned velocity table. Again, no result.

The table: independent data is time dependent data is U.

I do another test...

Re: moving tank
 

Using the boundary conditions you can only specify velocities tangential to the wall.

To get a sloshing tank you can either use a moving mesh approach of define a time varying body force.

Re: moving tank
 

is there any tutorial about that? also can we see the stress of the tank in star cd v4???

thanks

Re: moving tank
 

Perhaps you could change g, both magnitude and direction, as a function of time.

Re: moving tank
 

Hi Paul,

thanks for the hint, but I think the sloshed liquid will not find its "normal" equilibrium position without gravity.

Re: moving tank
 

You can define a variable gravity acceleration with time. You always have 8.81 m/s in the vertical direction and you define a variabile acceleration in the horizontal. I've tried it using Fortran routines, but maybe you can do it also using a table.

Re: moving tank
 

Hi, c-dot, Jörn Beilke said, that using the boundary conditions you can only specify velocities tangential to the wall. To get a sloshing tank you can either use a moving mesh approach. This is like true. Use for example user subroutine NEWXYZ (Chapter 16 USER PROGRAMMING of User guide). See also Chapter 14 ROTATING AND MOVING MESHES of User Guide.

Good LUCK.

Re: moving tank
 

thank, you but i do not see a way to enter a table for gravity (by the way: i only use this navigation panel)

fortran subroutines...hmm, sounds complicated

Re: moving tank
 

hi KtoTo,

thank you. but there is nowhere explained how to define a table of translational velocities.

but i can present new results:

i defined a velocity table for the boundary wall and i typed

MVGRID, ON

after calculation i finally see a slightly moving liquid!!!

of course, i am a little unsure about the result. you and Jörn said, using boundary conditions i can only specify velocities tangential to the wall.

does that mean, typing u = 123 I specify the velocity in x of a wall in its own ccordinate system and not in the global coordinate system?