JC Chang, a PhD student at the Cheriton School of Computer Science, his supervisor Professor Christopher Batty and Vinicius Azevedo, a postdoctoral researcher at ETH Zurich, have received the best poster award at SCA 2019, the ACM SIGGRAPH / Eurographics Symposium on Computer Animation.
Held this year from July 26 to 28 in Los Angeles, California, SCA 2019 is the premier forum for innovations in software and technology of computer animation. Each year, SCA brings together researchers and practitioners working on all aspects of time-based computer animation.
JC Chang (left) and Christopher Batty run a new velocity interpolation method for grid-based simulation of fluids that closely mimics the flow of fluids around obstacles as found in nature.
Their research in poster form, titled Divergence-Free and Boundary-Respecting Velocity Interpolation Using Stream Functions, presented a new velocity interpolation method for grid-based simulation of fluids. When modelling the flow of a fluid around an object, the standard method has used direct velocity interpolation. However, when calculating the velocity of points by interpolating the motion of discrete samples, this method leads to divergence and clustering of particles, phenomena that do not mimic what happens in nature.
Simulation of particle trajectories in a horizontal flow. The top animation shows particle trajectories using direct velocity interpolation, the bottom animation shows more accurate particle trajectories using the stream function-based method that JC Chang, Christopher Batty and Vinicius Azevedo developed.
The research JC Chang, Vinicius Azevedo and Professor Batty conducted employed a new fluid simulation method using stream function-based interpolation, which doesn’t lead to particle divergence or to clustering, thereby more accurately modelling the motion of fluids around a solid object.
“Congratulations to JC, Vinicius and Christopher on winning the best poster ward at SCA 2019,” said Mark Giesbrecht, Director of the David R. Cheriton School of Computer Science. “The main advantages of their new stream function-based interpolation method are that the motion of particles they simulate does not lead to artificial gaps behind the solid object they flow around, and that the distribution of particles in the simulation is more uniform.”