Parallel GPU-Based Fluid Animation

Examensarbete för masterexamen

Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.12380/242964
Download file(s):
File Description SizeFormat 
242964.pdfFulltext24.64 MBAdobe PDFView/Open
Type: Examensarbete för masterexamen
Master Thesis
Title: Parallel GPU-Based Fluid Animation
Authors: Svensson, Jakob
Abstract: In fluid animation today a choice has to be made between accuracy and speed. Accuracy is chosen when animating fluids for special effects to achieve a believable animation. Speed on the other hand is chosen for interactive applications where a fast simulation is a must to keep the animation running in real-time. However, with increases in computational power in computer hardware and as more efficient simulation methods are developed a question arises. Is it possible to have both an accurate and fast fluid animation? In this thesis this will be investigated by looking at how the Jacobi method, which is used in real-time fluid simulation due to its ability to run in parallel, can be accelerated to achieve a greater accuracy in a real-time simulation. To try this an application for real-time fluid animation was developed. The computation in the simulation was run in parallel on the GPU. The results showed that the acceleration of the Jacobi method does increase the accuracy, but not enough to achieve both accurate and fast animations. A method for simulating viscoelastic fluids was also found when investigating techniques for accelerating the Jacobi method.
Keywords: Människa-datorinteraktion (interaktionsdesign);Human Computer Interaction
Issue Date: 2016
Publisher: Chalmers tekniska högskola / Institutionen för tillämpad informationsteknologi (Chalmers)
Chalmers University of Technology / Department of Applied Information Technology (Chalmers)
Series/Report no.: Master thesis - Department of Applied Information Technology, Chalmers University of Technology, Göteborg, Sweden : 2016:122
URI: https://hdl.handle.net/20.500.12380/242964
Collection:Examensarbeten för masterexamen // Master Theses



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.