[EnriqueP]

Hello, I'm trying to simulate a fixed velocity fall of a body into a fluid, I am setting up the domain and introducing this UDF.

DEFINE_CG_MOTION(OW_fall, dt, vel, omega, time, dtime)
{
vel[1] = -5;
}

It is as simple as that but it does not work, when I start the simulation the body does not move at all and remains like that for any time.

Thank you in advance.

Enrique P.

__

[obscureed]

Hi Enrique,

Please could you sketch out how you have set up the Dynamic Mesh Zones? (I think you need a deforming cell zone plus a Rigid Body for the moving face zones. The Rigid Body is where you hook in the UDF.) Also describe the type of Dynamic Mesh that you have set up -- layering, smoothing, or whatever -- and summarise the settings there.

Thanks,
Ed

[EnriqueP]

Hi Obscureed, thanks for the answer.

I have my rigid body which is a cylinder, surrounded by another zone made by an inflation which moves with the cylinder. I have layering and remeshing that work nice because I have already used them to study the free fall under gravity of the same cylinder.

I attach some images with the setting of the dynamic mesh.

Cylinder_6DOF.PNG

Cylinder_inflation.PNG

Remeshing.PNG

Smoothing.PNG

[obscureed]

Hi Enrique,

Thanks for those images. A couple of thoughts come to mind:

-- Is 6DOF what you want in this model? I think if you want to impose a fixed motion, 6DOF should be off. I am actually confused about how the motion is currently defined: the UDF, the specified velocity in the Dynamic Mesh Zones window, or 6DOF. I think you need a setup where the UDF is more clearly in control. You also want a setup where a UDF of type DEFINE_CG_MOTION is called for -- currently, your UDF of this type is hooked in where DEFINE_SDOF_PROPERTIES would be appropriate.

-- You seem to have given the same motion settings to the cylinder and to the fluid. (Although, as in my previous point, I am not entirely sure what those motion settings are.) So perhaps you would not expect to see relative motion. I think a correct setup would have the cylinder (boundary face zone) being given rigid-body motion, and the moving_fluid (surrounding cell zone) deforming without motion.

By the way, I would recommend following the example in the help files to set the other velocities to zero:

  /* reset velocities */
  NV_S(vel, =, 0.0);
  NV_S(omega, =, 0.0);
  if (!Data_Valid_P())
    return;

As usual, it is a good idea to save case and data files before you start simulating with new settings. Please tell us how it goes -- when you reach a working setup, please share it in images like the current ones.

Good luck!
Ed

[EnriqueP]

Thank you again Obscureed

You are right, I don't reallt want 6DOF but I didn't know where to introduce the UDF function. I tried in the same way as the free fall simulation (I did the free fall case with DEFINE_SDOF_PROPERTIES and it works properly).

The "moving_fluid" zone that is in the simulation is only the portion of fluid with fixed mesh wich moves along with the body not the whole domain. Sorry about that, I should have uploaded an image of the mesh setup.

To be clear, the only motion that I want is the one of the cylinder falling at constant velocity (and the surrounding zone aswell). Where shall I hook up the DEFINE_CG_MOTION UDF?

[EnriqueP]

Thansk for the help.
My error was to enable the 6DOF solver as obscureed pointed out I'm using the UDF in the help manual as you suggested and it is working fine. The setup is the one uploaded before but disabling the 6DOF solver.

Now I get to the point of obtaining results of pressures and forces. I'll update you or ask for more help.

Enrique P.
__

[EnriqueP]

Im having some trouble again trying to impose the movement in two directions.
My UDF is the following:

DEFINE_CG_MOTION(OW_fall, dt, vel, omega, time, dtime)
{
/* reset velocities */
NV_S(vel[0], =, 40.0);
NV_S(vel[1], =, -1.5);
NV_S(omega, =, 0.0);
if (!Data_Valid_P())
return;

}

I get an error of the type "error: subscripted value is neither array nor pointer". Apparently it is related with the definition of the array vel which I think is defined by default as a 3-d array.

Any suggestions?
Thnaks.

[EnriqueP]

Im having some trouble again trying to impose the movement in two directions.
My UDF is the following:

DEFINE_CG_MOTION(OW_fall, dt, vel, omega, time, dtime)
{
/* reset velocities */
NV_S(vel[0], =, 40.0);
NV_S(vel[1], =, -1.5);
NV_S(omega, =, 0.0);
if (!Data_Valid_P())
return;

}

I get an error of the type "error: subscripted value is neither array nor pointer". Apparently it is related with the definition of the array vel which I think is defined by default as a 3-d array.

Any suggestions?
Thnaks.