PIC/FLIP Fluid Simulation Using Block-Optimized Grid Data Structure
Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
This thesis work will examin and present how to implement a Particle-In-Cell and a Fluid-Implicit-Particle (PIC / FLIP) fluid solver that takes advantage of the inherent parallelism of Digital Domain's sparse block optimized data structure, DB-Grid. The methods offer a hybrid approach between particle and grid based simulation.
This thesis will also discuss and go through different approaches for storing and accessing the data associated with each particle. For dynamically create and remove attributes from the particles, Disney's open source API, Partio is used. Which is also used for saving the particles to disk.
Finally how to expose C++ classes into Python by wrapping everything into a Python module using the Boost.Python API and discuss the benets of having a script language.
Place, publisher, year, edition, pages
2011. , 57 p.
PIC, FLIP, Boost.Python, fluid, Python, Blobbies, Improved Blobbies, Data structure, Partio
IdentifiersURN: urn:nbn:se:liu:diva-70794ISRN: LiU-ITN-TEK-A--11/055--SEOAI: oai:DiVA.org:liu-70794DiVA: diva2:441801
Subject / course
Master of Science in Media Technology and Engineering
2011-09-12, K25, Bredgatan 33, 601 74, Norrköping, 10:24 (English)
Unger, Jonas, Ph.D