Forces and Particle Controls

Futuremark Corp., used with permission

Particle Systems
General Particle Systems
 Particles are NOT independent
Force among them!

Cloth Crowd Fluid

Image Witkin & Baraff Auklet flock, Shumagins 1986.

Imagge from public domain: https://digitalmedia.fws.gov/ Image by Müller et al. 2005
Generalizations - Fluid Müller et al. 2005

 It’s not
all hacks:
Smoothed Particle Hydrodynamics (SPH)
A family of “real” particle-based
fluid simulation techniques.

 Fluid flow is described by the

Navier-Stokes Equations, a nonlinear
partial differential equation (PDE)
These Stanford folks use SPH for
resolving the small-scale spray
and mist.

Losasso, F., Talton, J., Kwatra, N. and Fedkiw, R., "Two-way Coupled SPH and
Particle Level Set Fluid Simulation", IEEE TV CG 14, 797-804 (2008). 6
What is a Force ?
A force changes the motion of the system
Newton says: When there are no forces,
motion continues uniformly in a straight line
(good enough for us)
(force)
 Forces can depend on location, time, velocity
(mass)
Gravity, spring, viscosity, wind, etc.

 For point masses, forces are vectors (gravity)

Wikipedia
What is a Force ?
A force changes the motion of the system
Newton says: When there are no forces,
motion continues uniformly in a straight line
(good enough for us)

Force Fk
 Forces can depend on location, time, velocity
Gravity, spring, viscosity, wind, etc. Mass m

Force G
 For point masses, forces are vectors

Wikipedia
I.e., to get total force, take vector
sum of everything
What is a Force ?
A force changes the motion of the system
Newton says: When there are no forces,
motion continues uniformly in a straight line
(good enough for us)

 Forces can depend on location, time, velocity

Gravity, spring, viscosity, wind, etc. Mass m

 For point masses, forces are vectors Total Force

Wikipedia
I.e., to get total force, take vector
sum of everything
Reference: https://www.school-for-
champions.com/science/force_vectors.
htm#.XVuh-OgzaUk
Forces: Gravity on Earth
 Depends only on particle mass
 f(X,t) = constant
 Hack for smoke, etc: make gravity point up!
Well, you can call this buoyancy, too.

v0
mi g

Wikipedia
Forces: Gravity (N-body problem)
 Gravity depends on all other particles
 Opposite for pairs of particles
 Forcein the direction of pi-pj with magnitude
inversely proportional to square distance

where G=6.67×10-11 Nm2/kg2

Pi (𝑚𝑖 )

 Testing all pairs is time consuming!

𝐺𝑚𝑖 𝑚𝑗
distance = 𝑟 𝐹𝑖𝑗 =
𝑟2
Particles are not
independent! Pj (𝑚 ) 𝑗

Pk
Real-Time Gravity Demo

http://www.youtube.com/watch?v=uhTuJZiAG64
Forces: Viscous Damping

 Damping force on particle i determined its

velocity
Opposes motion
E.g. wind resistance
 Terminal Velocity
https://youtu.be/EabUUrZFnFE
Forces: Viscous Damping

 Damping force on particle i determined its

velocity
Opposes motion
E.g. wind resistance
 Terminal Velocity
Liquids with different viscosities.
 Too much damping makes motion like in glue The liquid on the right has higher viscosity than
the liquid on the left.
Image form Wikipedia.
Forces: Spatial Fields
 Externally specified force (or velocity) fields
in space
 Force on particle i depends only on its position

Reference: https://ukabuer.github.io/curl-noise-fluid
 Arbitrary functions
wind
attractors, repulsors
vortices
 Can depend on time
Example: Procedural Spatial Field
 Curlnoise for procedural fluid flow, R. Bridson,
J. Hourihan, and M. Nordenstam, Proc. ACM
SIGGRAPH 2007.

Plausible, conrollable
force fields – just
advecting particles
along the flow gives
cool results!

And it’s simple, too!

Forces: Other Spatial Interaction

 E.g.,
approximate fluid using
Lennard-Jones force:
force

 Repulsive + attractive force

distance
Forces: Other Spatial Interaction

 E.g.,
approximate fluid using
Lennard-Jones force:
force

 Repulsive + attractive force

distance
 Again, time consuming to test all pairs
usually only local
https://youtu.be/nl7maklgYnI
3D ModeI Crack Formation

http://www.youtube.com/watch?v=dHWCT7RPjPo
More Eyecandy from NVIDIA

https://youtu.be/RuZQpWo9Qhs
Where do particles come from?
 Often created
by generators or emitters
Can be attached to objects in the model

https://youtu.be/XOKsgU7fiaI
Where do particles come from?
 Often created
by generators or emitters
Can be attached to objects in the model

 Designappropriate emitters
Rate of creation, randomness, etc.
Black art

https://youtu.be/NDCCzOqsPro
Particle Controls
 In production
tools, all these variables are time-
varying and controllable by the user (artist)
Emission rate, color, velocity distribution,
direction spread, textures, etc. etc.
All as a function of time!
 Example: Unity3D (Unity Particle Pack)

Images: Unity particle effects (from Unity Particle Pack)

 Emitter Controls
Velocity over lifetime Size over lifetime Color over lifetime/speed

/ Speed
Property:
size

Lifetime

Unity Particle System Curve

Markers control alpha and

colors(red, green, blue)

speed

Lifetime

Reference:
https://docs.huihoo.com/unity/4.3/Documentation/Manual/ParticleSystemCurv eEditor.html
https://w ww.raywenderlich.com/138-introduction-to-unity-particle-systems
Emitter Controls – Color By Speed / Size Over Lifetime

https://www.youtube.com/watch?v=XMSpIVRLdT8
Rendering and Motion Blur

Metal Gear Solid by Konami

Image from Sameboat
Star Trek 2 – The Wrath of Khan
 One of the earliest particle systems and
fractal landscapes (from 1982)
Images from Star Trek 2 – The Wrath of Khan, Paramount Pictures
 Described in [Reeves, 1983]

30
https://youtu.be/dyxpmAU9I8M