Events Calendar

PhD Defense - Yili Zhao

PhD Candidate: Yili Zhao

Comittee member:
Jernej Barbic (Chair)
Ulrich Neumann
Igor Kukavica (Math Department)

Location:
SAL 322

Time:
10:00AM - 12:00PM


Title:
Plant Substructuring and Real-time Simulation Using Model Reduction

Abstract:

This research is focusing on real-time, physically-based simulation of plants undergoing large deformations. To achieve this goal, we first propose a novel algorithm based on model
reduction and domain decomposition. It extends 3D nonlinear elasticity model reduction to open-loop multi-level reduced deformable structures. We decompose the input mesh into
several domains, build a reduced deformable model for every domain, simulate each one separately, and connect domains using proper inertia coupling. This makes model reduction deformable simulations much more versatile: localized deformations can be supported without prohibitive computational costs, parts can be re-used and precomputation time can be shortened. Our method does not use constraints, and can handle large domain rigid body
motion in addition to large deformations, due to our derivation of the gradient and Hessian of the rotation matrix in polar decomposition. We show real-time examples with multi-level
domain hierarchies and thousands of reduced degrees of freedom.

Then we design a pre-processor which takes a plant “polygon soup” triangle mesh as the only input and quickly pre-compute necessary data for the subsequent simulation. This tool breaks the ice for adoption of our multidomain dynamics simulator in practice. Our
pre-processor is robust to non-manifold input geometry, gaps between branches or leaves, free-flying leaves not connected to any branch, small unimportant geometry (“debris”) left in
the model, and plant self-collisions in the input configuration. Repeated copies (instances) of plant subparts such as leaves, petals or fruits can be automatically detected by our preprocessor. We enhanced our multidomain dynamics simulator to provide plant fracture, and inverse kinematics to easily pose plants. It can simulate complex plants at interactive rates, subjected to user forces, gravity or randomized wind. We simulated over 100 plants from diverse climates and geographic regions, including broadleaf (deciduous) trees and conifers,
bushes and flowers. Our largest simulations involve anatomically realistic adult trees with hundreds of branches and over 100,000 leaves.

Finally, we propose our future research in several directions including adding hierarchical instancing, collision detection and handling, etc.

Location: Henry Salvatori Computer Science Center (SAL) - 322

Audiences: Everyone Is Invited

Contact: Lizsl De Leon