Flows inside the cylindrical container in Realflow

Discussions about the use of Realflow
jarina
Posts: 1
Joined: Fri Jan 19, 2018 8:33 am

Flows inside the cylindrical container in Realflow

Postby jarina » Fri Jan 19, 2018 8:46 am

Hi guys!

I ask you to help me, please!

1. There is a big, wide and not tall cylinder in the scene. I need to make a transparent liquid in it - I need to make small suspended particles of liquid, slowly moving in the cylinder. I need not only the liquid surface, but the motion of particles inside the liquid, it is important to me.

How can I make that? Is the liquid filling the tank, and ready to react actively? I ask you to help, I'm new to realflow and now I can not get for several days what I need.


2. "Flow from nowhere." The suspended particles of liquid in a large volume of the cylinder are placed uniformly.
0)There is no gradient of pronounced density transitions.
1)Next we see how several particles are captured by slow motion;
2) heavier particles are collected together and form high-density particles flow regions.
3)And these zones become visible as flow, which falls into the iron funnel.

How can I make that?

Many thanks to all who help the new girl and +1000 in your karma :-)


Forester
Posts: 99
Joined: Mon Mar 15, 2010 4:36 am
Contact:

Re: Flows inside the cylindrical container in Realflow

Postby Forester » Fri Jan 19, 2018 9:16 pm

There may not be enough information in your question for me to make a complete help, but I will start the process of helping you. A Realflow Version 10 (not Trial Edition) copy of the simulation files used for this demonstration can be downloaded from here.

http://www.expandingwave.com/rfif_help/ParticlesDiffWeights.zip


First point, you will want to use the Dyverso solver for this kind of project, but use the SPH Particles type. The Dyverso Solver has two particle types, PDB (default) and SPH. See the picture of the Node Editor below to learn where to change the Particle Type.

If you use the Dyverso solver with the default PDB particles, the "transparent" fluid will be too energetic for this purpose. PDB particles never come to rest - they are good for simulations of fountains and waterfalls, but not good for detailed studies of small, slow or "natural" particle movement. SPH particles lose their speed of movement in a normal way, and they will come to a rest position after many simulation steps. PDB particles never come to rest.

Second, for several important settings:

For the Cylinder Object, it is important that it have no interactions with the fluid you will make. For this purpose, change the type of Cylinder setting to "Shell" and make all the Object Interaction settings be 0.0. Also, change the Surface Offset size to match the Cell Size setting. Click on the picture to expand it and make it possible to see the Settings values.

Image

Set the Simulation Length Timeline to be about 1000 steps. You need a long set of simulation steps to allow the particles to exhibit all of their behavior.

Because you are simulating a fluid that has light weight particles and heavy weight particles, you need two different emitters. So, you will need two Dyverso Domains, one for light-weight particles, and one for heavy-weight particles.

To be able to see the particles and their behavior in a clear way, you need to reduce the number of particles put out by each emitter. The way to control the number of particles per area of volume is to reduce the "Resolution" of the particles in each Domain. For this demonstration, I have reduced the Resolution setting from a value of 1.0 to a value of 0.1. Reducing the number of particles per volume creates more space between them, and so it makes it easier to create a uniform distribution of the particles from two different emitters within one cylinder space.

Here, I have used "Fill Cylinders" for the emitters, but you could use two Circle Emitters if you like. For the Fill Cylinders, I have reduced the Speed setting to make the particles emit more slowly. I also have animated their Speed settings so that the emitters produce particles for only 9 simulation steps. They stop emitting particles after 9 steps (Speed = 0.0), but the particles continue to move around throughout the simulation, and react to various forces. I did this so that the emitters produce just enough particles to allow us to see them in a clear manner, and watch their behavior over time.

I changed the particle Display settings for each domain so that I could assign a White Color to the light-weight particles and a Red Color to the heavy-weight particles. Here is where, in the Domain Node Settings, you go to make a change of color for the particles.

Image

Finally, to make the particles have different "weights", I changed the Density Setting of the second Dyverso domain. In Realflow, the Resolution Setting controls what we often think of as "density" - the distance between particles within a volume. And the Density Setting controls forces that will make the overall fluid seem heavier or lighter. But you can use the Density Setting to make the particles act lighter or heavier in the presence of gravity.

So, you need to have just a little gravity in this scene. I added the Gravity Daemon, but changed its value from the default of 9.8 to a value of 1.0.

Now, if we run this simulation for 1000 steps, you can see the two kinds of particles mix and move around in the fluid within the cylinder, but the heavier red particles will settle to the floor of the cylinder over time, in the presence of a small amount of gravity.

Image

If you add a Vortex Daemon, or a Layered Vorex Daemon or a Coriolis Daemon, and give these just a small amount of force, you can see the particles rotate. Please note that the heavy-weight particles move to the outer part of the cylinder, while the lighter-weight particles are less affected by the circular forces, and remain in the center of the cylinder for a longer period of time.

Image
Windows 7 Pro 64 - Realflow 10.1.2.0162 (Nov 2017 patched) Standard Version

Forester
Posts: 99
Joined: Mon Mar 15, 2010 4:36 am
Contact:

Re: Flows inside the cylindrical container in Realflow

Postby Forester » Fri Jan 19, 2018 9:31 pm

You have asked for several things for this scene and I am not sure that I understand all that you want. The work above gives you the basic elements of the simulation you want to make.

To make this a better visualization, you could increase the scale of the cylinder (3x or 4x), and perhaps import a better, more finely made cylinder.

After you have the simulation working in a basic way, you could increase the scale of the Emitters, and then increase the Resolution of the two Domains so that many more particles are created. This will make the flows of the particles look good, but you will not be able to see the individual particles. (The ability to see the individual particles, versus the ability to see flows of particles masses is something I find confusing in your question.)

You can reduce the Speed settings of the Emitters to make all particle movement slow much further. But to make it all really slow, you would need to run the simulation all the way to the end and then use the Top Line Menu | Tools | Reset Timer to make the animation more slow. Do not do this until you have the simlulation running the way you want.

You can alter the Density of the heavy-weight particles to make them weigh much more. They will move more slowly, in comparison to the light-weight particles, and will sink much faster.

You can increase the force of the Gravity Daemon to make all of the fluid sink faster.

Please come back to this Forum and ask more questions and get more specific anwers to your request. We are all happy to help you.
Windows 7 Pro 64 - Realflow 10.1.2.0162 (Nov 2017 patched) Standard Version

Forester
Posts: 99
Joined: Mon Mar 15, 2010 4:36 am
Contact:

Re: Flows inside the cylindrical container in Realflow

Postby Forester » Sun Jan 21, 2018 9:13 am

The primary difficulty I see with this problem is the initial condition of the fluid, and how to achieve that state. If I understand the problem correctly, the initial condition of the fluid is supposed to be a transparent fluid made of light-weight and heavy-weight particles that are evenly distributed in the volume of the cylinder.

The only way that I can think of to create this condition is to mix the two kinds of particles as completely as possible, before making the observations of the heavier particles. I think this might be achieved by rotating the two emitters for a short period, during the production of the particles, and then stopping the rotation.

A second thing that needs to be done is to increase the external pressure value of the heavy-weight particles. That will cause them to clump together during the simulation.
Windows 7 Pro 64 - Realflow 10.1.2.0162 (Nov 2017 patched) Standard Version


Return to “Realflow”

Who is online

Users browsing this forum: No registered users and 2 guests