Huamin Wang, Peter J' Mucha and Greg Turk - PowerPoint PPT Presentation

1 / 29
About This Presentation
Title:

Huamin Wang, Peter J' Mucha and Greg Turk

Description:

Huamin Wang, Peter J. Mucha and Greg Turk. Georgia Institute of Technology. From SIGGRAPH 2005 ... Q: Why small water behaves naturally different from large water? ... – PowerPoint PPT presentation

Number of Views:51
Avg rating:3.0/5.0
Slides: 30
Provided by: huami
Category:
Tags: greg | huamin | mucha | peter | turk | wang

less

Transcript and Presenter's Notes

Title: Huamin Wang, Peter J' Mucha and Greg Turk


1
Water Drops on Surfaces
  • Huamin Wang, Peter J. Mucha and Greg Turk
  • Georgia Institute of Technology
  • From SIGGRAPH 2005
  • Presented by Huamin Wang (whmin_at_cc.gatech.edu)

2
Motivation
  • Film special effects

3
Motivation
  • Scientific Visualization

4
Motivation
  • Game industry (eventually)

5
Motivation
  • Game industry (eventually)

6
Overview
  • Background
  • The virtual surface method
  • Dynamic contact angle model
  • Results and Analysis
  • Future work

7
Background
  • Small-scale liquid-solid Interactions

Q Why small water behaves naturally different
from large water? A1 Surface Tension (water
72 dynes/cm at 25º C). A2 Viscosity (water
1.0020 10-3 Ns/m2 at 20º C).
Lake view ( gt1 meter)
Water drops (in millimeters)
8
Background
  • To calculate surface tension force
  • Tension coefficient (always positive)
  • Mean curvature
  • Normal (always pointing outward)

(Laplaces Law)
Uniform curvature
Water sphere photo taken on the International
Space Station. Courtesy NASA
9
Background
Stable contact angle satisfies Youngs Relation
10
Background
  • Our work is based on fluid simulation using
    Computational Fluid Dynamics (CFD).
  • Solve the Navier-Stokes equations for the
    velocity field
  • Use the particle Level-Set method

11
Overview
  • Background
  • The virtual surface method
  • Dynamic contact angle model
  • Results and Analysis
  • Future work

12
The virtual surface method
  • Solution
  • Place a virtual surface beneath the solid plane
  • Estimate the surface tension using the new
    combined surface
  • Problem
  • Real World a stable contact front with contact
    angle
  • Simulation Curvature at contact front is
    positive, thus always pushes inward

Air
Liquid
Solid
Virtual Liquid
Virtual Surface
13
The virtual surface method
  • Create a virtual surface
  • Estimate curvature on the contact front
  • A kink cause the curvature to push the fluid
    front

Air
Air
Liquid
Virtual Liquid
Virtual Liquid
Solid
Solid
Virtual Surface
Virtual Surface
Advancing to right
Receding to left
14
The virtual surface method
  • Details? Please read our paper.

15
Overview
  • Background
  • The virtual surface method
  • Dynamic contact angle model
  • Results and Analysis
  • Future work

16
Dynamic contact angle model
Contact angle hysteresis Small water drops can
stay on a vertical plane, while large water drops
will flow down.
17
Dynamic contact angle model
  • Stable contact angles bounds
  • Advancing
  • Receding
  • a valid stable contact angle
  • Final pressure P is

Pa pressure calculated using Pr
pressure calculated using
18
Dynamic contact angle model
  • Dry/Wet conditions
  • wetting history map
  • Contact angles based on surface wetness
  • (wet advancing angle smaller than the dry
    advancing angle)

Dusted region is dry, transparent region is wet.
19
Overview
  • Background
  • The virtual surface method
  • Dynamic contact angle model
  • Results and Analysis
  • Future work

20
Results and Analysis
  • Physics phenomena

Capillary Action
21
Results and Analysis
  • Drop impacts, rivulets, dripping drops, and more

22
Results and Analysis
  • Drop impacts, rivulets, dripping drops, and more

23
Results and Analysis
  • Video

(playback speed 10 times slower than real world
speed)
24
Results and Analysis
  • Grid resolution 400400400
  • Simulation speed 20 minutes per frame
  • Each sequence has 500 frames
  • The total running time is
  • 20 500 / (60 24) 7 days

25
Results and Analysis
  • Why is it relatively computational expensive?
  • High grid solution (400400400)
  • Large viscosity effects
  • Implicit Euler method
  • The condition number of the linear system
    increased.

26
Overview
  • Background
  • The virtual surface method
  • Dynamic contact angle model
  • Results and Analysis
  • Future work

27
Future Work
  • Octree data structure
  • Virtual surface reconstruction based on particles
  • Distributed computing

28
Acknowledgements
  • We would like to thank
  • Mark Carlson, Chris Wojtan, Howard Zhou, Spencer
    Reynolds, Nathan Sisterson
  • Everyone supporting our work, including
    reviewers.
  • Gatech Computational Perception Laboratory,
    Geometry Group
  • CMU graphics lab
  • Funded
  • In part by NSF grant DMS 0204309.
  • Rendering
  • Physically Based Ray tracer (pbrt), Matt Pharr
    Greg Humpheys
  • Light Probe Image Gallery, Paul Debevec

29
Fin Any questions?
Water Drops on Surfaces
Huamin Wang, Peter J. Mucha and Greg Turk Georgia
Institute of Technology From SIGGRAPH
2005 Presented by Huamin Wang (whmin_at_cc.gatech.edu
)
Write a Comment
User Comments (0)
About PowerShow.com