Introducing particle fluids parameters

Hi all 🙂

Here are basic functionalities of the particle fluids.
————————————————————-

Blender Particle Fluids integrate into the existent powerful Blender particle system a fluid simulation tool that allows a wide range of fluids types, from steam to water to goo, slime and giggly fluids.
The particle fluids inherit all the goodies of the base particle system, like baked simulations, scripting, etc.
In future releases it will also include meshing particles .
Blender Fluid Particles uses a boundless interpolation technique to sample the fluids body called SPH (Smoothed Particle Hydrodynamics)
Each particle can carry any fluid property: density, pressure, temperature, color, etc.

Smoothing radius is the most important parameter in the simulator, it controls the influence of each particle, and determines the performance and the type of the simulated fluid: for small fluid scales (droplets, tiny water) use a relatively big smoothing radius (1.0) and for big scales fluids (sand, sea, rivers) use a small one (0.2-0.1).

Mass controls the amount of fluid matter that each particle represents, it is important for multifluid interactions (denser and lighter fluids) and also for simulating different fluid scale with the same particle amount.

Viscosity controls the movement of the fluid, making it  more goo like and also more stable: it is a factor that dampens the relative velocity of each particle.

Particle damping is a stabilization factor and controsl the absolute particle velocity: when a simulation seems to be out of control try to increase it or play with viscosity. It differs from viscosity because the second damps the relative particle speed.

Stiffness controls the surface tension on an external level, but internally is an attraction force in each particle smoothing radius. For small scale fluids use high stiffness values (1.0) and for large scale fluids use low to none ones.

Repulsion factor has the opposite effect as stiffness factor, it repels each particle from its neighbours to maintain an Equilibrium.

Collider damping affects the bouncing of the particle in the collision event, the more collider damping the more energy the particle will loose the particle and consequently possibly get stick to the collider surface (not enough remaining energy to cancel the attraction force).

Collider particle friction affects the particle velocity as it travels over the collision object, increase its value to get realistic slipping liquid effects.

Rest density is an important parameter that sets the density at which particles will try to maintain under a zero force field or rest state. Visually it controls the relative distances where particles will be settled.

Tweak parameter is a time step tweak for the simulation, if a sim set up goes unstable first try to lower the time step by setting a fractional tweak. SPH simulators are not unconditionally stable: they are stable only for relative small time steps, that’s the reason why increasing the particle count without adjusting the rest of the parameters could lead to exploding simulations 🙂 The good news is that always will be a set of parameters that will make a desired simulation stable.
The particle integrator is the function that actually advances particles over time, they feature more fast and stable simulation, so correctly choosing the appropriate integrator is crucial. In general Midpoint is the most stable while Euler is the fastest and Verlet is more suited for fluid behaviors.

For Goo like fluids Spring is very important, this is a force among pairs of particles that try to maintain them to a fixed distance, the rest length.

The rest length is the distance particles will try to maintain with a force proportional to the spring factor.

Brownian adds a random movement to the particles, useful for small scale fluids or crazy FX’s.

Collider stickiness controls the force factor that try to stick particles to the collider surface, this is great for slipping fluids.
Square viscosity is a second order viscosity factor, used for advanced fluid for enhancing viscous behaviour.

————————————————————-
Particles fluids can be currently rendered as billboards, simple dots, halos, instanced objects and volumetric distorted spheres in the pointdensity texture type.

In the future they can be rendered as fluid isosurfaces.

There’re plenty of ways in particle fluids for interactions and advanced setups, and combinations of the parameters lead to non-linear behaviours so they are not the simply sum of
effects. If you want to master particle fluids a good amount of experimentation is advised as well as a basic understanding of particle fluids theory.

Here you could download this post as a pptx:
Blender Particle Fluids

Cheers

Farsthary

Advertisements
Introducing particle fluids parameters

25 thoughts on “Introducing particle fluids parameters

  1. Willem says:

    Superb! Thanks for the detailed explanation.

    When will we be able to render the particles as water/isosurface? Will external renderers like Yafaray and V-Ray be able to render the particles as isosurfaces?

    Like

  2. oslo says:

    Farsthary, thank you for this amazing piece of work, all the time and effort you have put in this!
    It makes Blender even more interesting to use 😉

    Thanks!

    Like

  3. Marcus says:

    Hi,

    great job!
    Easy to loose the overview of what does what, so presets or sample blend files would be very nice (waterfalls…)

    Like

  4. N.A. says:

    Could someone do a build for Graphicall.org (Win 32)? Or explain how to use patches with the Scons-System…?

    The Explanations are very good and the Pictures really help…it looks very good for Blenders Future 🙂

    Like

  5. Rickyblender says:

    very nice work i like this new feature it will allow to simualte different type of lfuids in blender

    1- first question is this available at graphicalorg version for 2.5

    or what version should i get to experiment with this?

    2 – fluid simulation
    anything in sight to simulate fluid like air or gas around objects like ariplanes or submarine may be

    i mean with all theses algos it should be possible to do that i think!

    3-
    i was not able to open the PPT file for particules at the end
    cause i only have Openoffice for the time being

    is it available as a PDF may be so i can read it

    keep up the good work
    and hope to see more features like that in blender

    Thnaks for your work on this and happy 2.5

    Like

  6. thanks farsthary! this is a very nice addition to already existing tools in blender.

    i agree with some of the other commenters, a few presets would be very nice and make a start easier. presets like a few simple gas basics and liquids.

    Like

  7. Xiban says:

    Astonisehd……. you coded a phisycs manual in Blender….
    Can’t wait to try it, when will us be able to render those fluids as surfaces?

    Like

  8. Great to find detailed examples of viscosity, damping and stickyness! Very helpfull to get ANY results in a decently enough quick time!
    Thanks a lot!!!
    😉

    Like

  9. Josh says:

    You implementation of this particle solver is going to open up so many possibilities. I think a few test .blend scenes would be really helpful as I (well I only spent an hour) spent a good amount of time trying to get water but it didn’t completely turn out how water should be.
    Possibly in the future some implementation of some presets would be nice.

    Like

  10. […] O autor do video noa explica o funcionamento de todos os parâmetros das partículas quando as mesmas se comportam como fluidos, mas se você quiser uma excelente referencia sobre esses parâmetros, resgatei um artigo que o Farsthary escreveu já faz um ano, em que ele demostra vários desses parâmetros com imagens. Para conhecer esses parâmetros das partículas no Blender 2.57, visite esse link. […]

    Like

  11. Chris Woodbury says:

    I have been looking for good information about SPH fluid particle settings for over a week now – this by far the most helpful I found yet. The 2.78 manual just reiterates the setting names and offers almost no information how the SPH sim works or the effect of each setting. Do you know which SPH algorithm variant Blender uses? I would like to delve deeper into the theory, but find there are many different versions out there.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s