2016

MSc Projects 2016

MSc projects from 2016

Aina Nicolau Orell

Position Based Dynamics for Character Effects

Alexis Agrotis

Thesis A Fluid Implicit Particle (FLIP) Sovler Built in Houdini Abstract The following thesis presents the research and implementation of a Fluid Implicit Particle (FLUID). A literature review is included in order to highlight the most commonly used techniques for simulating fluids in the Visual Effects and Computer Graphics industries as well as critically analyse and compare them. A custom build solver was implemented in SideFX’s Houdini using primarily gas microsolvers and VEX Wrangles or VOP networks to simulate the different fields of the fluid as well as foam based on colour. The whole implementation process is broken down in detail justifying every step of the FLIP algorithm. Decisions were also made for the additional shading, lighting and rendering the simulation which required the transformation of particles into a surface. Various scenarios are described as well as possible future improvements that could be made.

Calum Devlin

An Investigation into an Assortment of Flocking Algorithms

Derik Gokstorp

Prototyping a Custom 3D Printing Slicer

Erika Camilleri

Path Planning for Natural Phenomena

Ewan Rice

OpenGL GPU Features and Fluid SPH

Ina M. Sorensen

Melting Simulation with the Material Point Method

Jennifer Moorehead

Thesis Smooth Particle Hydrodynamics Abstract This paper discusses the implementation of a fluid solver which implements a version of Smooth Particle Hydrodynamics approximations using the prediction relaxation scheme proposed by Clavet et al. In addition optimizations such as spacial hashing have been made and the point geometry has been exported to Houdini for meshing, lighting and rendering purposes.

Luke Bazalgette

Thesis Flocking system mimicking fish school behaviours Abstract In this thesis, a programmable flocking system was developed. The system is designed to emulate the movement patterns of fish in a body of water. Previous academic research into the field is reviewed and analysed through a critical lens, as a means of informing the design of the software. What follows is an extensive review of the implementation process, providing details on code structure, languages and libraries used to assist development. The resulting simulation is capable of interpreting flock behaviours through the use of a high level scripting environment which can be modified at run time. In order to demonstrate the functionality of the simulation, several scripts were produced. The default script emulates the behaviours of fish when confronted with danger through the formation of a defensive bait ball; the ball alters its shape based on its proximity to a moving point described as a predator. Upon analysing the results of the implementation, a reciprocal velocity based flocking system is proposed and the results of the implementation are presented and discussed.