[vodoley]

Hello,

Can I use and a dynamic mesh without using of UDF file?

The problem is simple. 2D, some rectangular region with air. Side walls - walls. Bottom - velocity inlet with U m/s, Upper boundary - pressure outlet. We have also 1 or 2 rigid bodyes (circules) placed in the air. Gravity (-9.81) is on.

Can I use and a dynamic mesh without using of UDF profile? Can the motion of rigid circules be determined by the flow calculations only (using Gravity g and U)?

[vinerm]

Older versions of Fluent always needed a UDF to provide properties of the object that has degrees of freedom. However, newer version do not need it. There are options to specify the properties within the GUI. There is not UDF needed for profile. Profile is required if user wants to specify the motion of the solid body. If user wants Fluent to determine the motion based on the flow, then 6DOF is to be used. If you want to specify the motion, you can use text based profile file. Here, all 6 motions (3 in 2D; two translations and one rotation) can be specified via a simple file.

[vodoley]

Thanks for reply.

I turned ON 6DOF option. But when I press Create button I receve message that there is no UDF profile. May be I should specify something else within the GUI?







The geometry is simple:

[vinerm]

Do you observe the option Six DOF UDF/Properties drop-down menu on the window? This is meant to provide properties of the solid object that is in motion. In newer versions, you can provide numbers via GUI but in older one, you have to use UDF.

[vodoley]

Yes, but "Six DOF UDF/Properties" drop-down menu is empty. May be I should specify something else within Fluent... I don't know where.

P.S. Of couse I can do unnecessary action in my problem described above (imho) and specify some property of rigid bodyes (density for example) whithout specifing any motion. But I cannot connect Fluent with Microsoft Visual Studio compiler...

[vinerm]

You have to select option Six DOF on the right. Then click on Settings and create a Six DOF object. Then, it will automatically appear in the list.

To connect VS, you can go to Command Prompt of VS; not the usual command prompt. Then start Fluent from this Visual Studio Command Prompt and compile the UDF.

[vodoley]

Thank you, you helped me a lot!

[vodoley]

Hello! I still keep performing the task described above.

2D, some rectangular region with air (160x160 mm). Side walls - walls. Bottom - velocity inlet with U m/s, Upper boundary - pressure outlet. We have also One rigid body (droplet 20 mm diam) placed in the air.

Gravity (-9.81) is ON. Multiphase is OFF.

I want to determine motion of the droplet (using Gravity g and air flow from the bottom U m/s).

In 'Dynamic mesh' I created 'droplet' 1e-06 kg:





I also created two Dynamic mesh zones: 'droplet' (SixDOF is on) and 'droplet_walls' (SixDOF is on; SixDOF is passive)



Velocity U is very big:



AND something wrong... My very light Droplet falls down in time. But with such a high flow rate (1000m/s) from below, it is obvious that the droplet (1E-6 kg) must moves up!

After that I turned gravity off. The Droplet froze and did not move in time. But with air flow from below, it is obvious that the droplet must moves up!

Can you tell me what the problem is? Such a simple task... Why the droplet is not affected by any force from below!

[vinerm]

You have not entered values for Moment of Inertia.

And why do you wish to use dynamic mesh for droplet? A 1 mg of droplet of water means a diameter of approximately 1 cm. That's quite large for a droplet to behave like a rigid body. You should use VOF.

[vodoley]

Originally Posted by vinerm

You have not entered values for Moment of Inertia.

And why do you wish to use dynamic mesh for droplet? A 1 mg of droplet of water means a diameter of approximately 1 cm. That's quite large for a droplet to behave like a rigid body. You should use VOF.

Thanks for reply!
I set some random mass and diameter of droplet so far... Just for debugging of the problem statement. I think this task is interesting for understanding how Fluent works.

In fact I want to determine the motion of real droplet (5.24E-10 kg; 0.1 mm). in air flow from the bottom with U m/s. I decided such droplet behave like a rigid body for simplicity.

You have not entered values for Moment of Inertia... You mean inertia tensor in 'Six DOF prorerties' or shear contition on droplet walls in 'Boundary conditions'?




I found some formula.

In order for a droplet hung in a stream of vapor (wet air) with the flow velocity U m/s, the weight of the droplet should be compensated by the force which air flow acts on a droplet:


D = 0.1mm
md - droplet mass = 5.24E-10 kg
g = 9.81

Cd - a drag coefficient of a spherical droplet, depending on the Reynolds number;
Sd - cross-sectional area = 3.14*D*D/4=0.00785 mm2;
ρv — saturated air density;
U ~ 0.27 m/s.

[vinerm]

The equations for rigid body are simpler than those for multiphase flow. Whether a bubble or droplet deforms or not depends on We and not on diameter but for water-air system, 0.1 mm droplet may maintain spherical shape. So, you can certainly use rigid body (six-DOF) solver.

Yes, you have to provide inertia tensor. However, only three values (and all equal to each other) exist for a sphere. So, you only need to provide Ixx, Iyy, and Izz and all will be equal.

The expression you mentioned is nothing but comparison of drag and weight and is usually used to determine terminal velocity. Fluent has got nothing to do with it nor do you need to worry about it. Just provide droplet mass, its inertia, and it would work alright provided you have a good quality mesh.

[vodoley]

Thank you, now I will try to provide inertia tensor. First of all I need to find the correct values of inertia tensor. It's not so obvious for me yet. I'll write teh result.

[vinerm]

For a sphere of radius r and mass m, it is just



And all three values are same. Ixy, Iyz, and Izx are 0.

[vodoley]

Yes, you are absolutely right that Ixx=Iyy=Izz=2*m*r*r/5 for solid sphere.

Since I have a flat geometry I tried to not fill in Izz. But Fluent does not allow you to do this so I put Ixx=Iyy=Izz=2*m*r*r/5 and Ixy, Iyz, and Izx are 0...

2D, some rectangular region with air (160x160 mm). We have also One rigid body (droplet 20 mm diam) placed in the air. Bottom - very big velocity inlet with U m/s (5000 m/s!), Upper boundary - pressure outlet.

Gravity (-9.81) is OFF. I put droplet mass to 1E-05, therefore Ixx=Iyy=Izz=4E-10...



And the droplet is not affected by air flow from below. The droplet stands still.

But there is another problem. After several time steps I saw that Ixx and Iyy becomes empty automatically! Only Izz=2*m*r*r/5. But I have a flat geometry...

[vinerm]

What you are modeling is not a droplet, but a cylinder. So, you have to provide moment of inertia values accordingly. The equation is different for each axis.

As far as the motion is concerned, that depends on rigid body setup. You need to have only one Dynamic Mesh Object, the droplet. Secondly, if there is not enough force due to the flow, then the object won't move.

[vodoley]

I simplified the problem: 2D, rectangular region with WATER (160x160 mm, 998 kg/m3). We have also One rigid body (20 mm diam, 100kg/m3) placed in water. Upper boundary - pressure outlet, other boundaries - walls.

According to Archimedes' principle the rigid body should moves up in water. But it does not.

May be the problem is... When I provide moment of inertia (any values of Ixx, Iyy) and press 'Calculate' after several time steps Ixx and Iyy become 0.0 automatically!

[vinerm]

Since you are dealing with 2D, planar problem, Ixx and Iyy have no significance. The rotation is allowed only around z-axis, hence, Ixx and Iyy are being reset to 0. As far as the motion is concerned, buoyancy works due to gravity. If gravity is not enabled for Six-DOF setup, it won't work.

[vodoley]

Yes, I understood. I turned on gravity and rigid body (100 kg/m3) started to falls down in water (998 kg/m3). I setted arbitrary values (Izz) from 1E-08 to 1.0.

May be I do something wrong. I'll check...

[vinerm]

MoI affects only rotation. As long as there is no rotation, which is not expected for your case, it won't matter.

[vodoley]

I continue to struggle with this simple task

I found macro for Six DOF. To specify custom external forces to a rigid moving object I use the following macro:

SDOF_LOAD_F_X, /* external force */
SDOF_LOAD_F_Y, /* external force */
SDOF_LOAD_F_Z, /* external force */

As far as I understand it is available only in UDF. I guess that for my task I need a UDF to provide properties of the object (mass and external force SDOF_LOAD_F_Y). But what is the dimension of the external forse in SDOF_LOAD_F_Y...