[Bf-committers] Blender - GSOC 2012

[Tom M]

hi stanley please do put your proposal in the tracker google-melange
proposal tracker, you can still edit it after it is in the tracker, it
just makes it easier for other mentors to see the comments and
followups that the proposal has recieved from other mentors.

LetterRip

On Thu, Apr 5, 2012 at 10:07 AM, Lewis, Stanley R. <stan.r.lewis at ou.edu> wrote:
> Hi all,
>
> My name is Stanley Lewis, and I am considering applying for 2012's Google Summer of Code with Blender with the aim of improving the fluid dynamics library to support thin fluid features and rigid body to fluid interaction. I realize that this is very late in the application process (seeing as the applications are due tomorrow), but I was only turned on to GSOC within the last few days (namely, tuesday 4/3/2012).
>
> Anyways, I was wondering if any of you would be willing to take the time to provide feedback to me about my application, which i have pasted at the end of this email, below my signature.
>
> Any feedback you can provide would be greatly appreciated!
>
> Thanks,
>
> Stanley Lewis
> Electronics, Data Acquisition, and Vehicle Dynamics/Aerodynamics
> Sooner Racing Team - sae.ou.edu
> cell: (832)721-9664
> e-mail: stan.r.lewis at ou.edu
>
> Begin application:
>
>
> Name
>
> Stanley Lewis
>
>
> Email / IRC / WWW
>
> Email: stan.r.lewis at ou.edu
>
> IRC:StanL (I don't usually use IRC, but when I do, I use this nick)
>
> webpage/blog: sae.ou.edu (not a personal website, but it shows what I work on during the school year and the end product)
>
>
> Additional Contact Info
>
> Phone:(832)721-9664
>
> Physical Address:
>
> (school)
>
> 1800 Beaumont Drive Apt 413C
>
> Norman, Ok 73072
>
>
> (home)
>
> 14111 Radford Park Circle
>
> Houston, Tx 77062
>
>
> Synopsis
>
> Blender's current fluid simulation code is based off the Lattice Boltzmann methods for computational fluid dynamics. As opposed to the more commonly used (for engineering CFD at any rate) Navier-Stokes based methods, which seek to solve a set of non-linear PDE's through either a finite volume or finite element method, the Lattice Boltzmann equations instead treat the fluid as a collection of very small (sub micrometer, but larger than molecular) particles. This allows for meshless simulation of fluid dynamics.
>
> I would plan to increase Blender's fluid simulation capabilities in several key ways, notably:
>
> a) addition of thin fluid features to elbeem
>
> b) extending elbeem to allow interaction with rigid body objects
>
> c) seeking to improve general fluid simulation performance (either algorithmically, or via GPU acceleration, or possible a combination thereof)
>
>
> These additional features/capabilities will allow Blender's users to create fluid animations that are more realistic in both appearance and behavior, while also seeking to improve the speed at which the animations can be computed/generated, thereby decreasing the time necessary to assemble the work in its entirety.
>
>
> Deliverables
>
> An extended elbeem fluid simulation library will support
>
> -thin fluid features
>
> -rigid body/fluid collisions
>
> As well, an extended user-interface will be designed and implemented, in order for users to implement the additional simulation options into their work.
>
> Then, last but not least, tutorials/training documents will be created, both in textual and video form, documenting the basic workflow and features of the improved simulation library.
>
>
> Project Details
>
> A good portion of the details are listed in the synopsis section.
>
> This project would need to be split into 4 main phases
>
> 1) Further research into necessary fluid simulation concepts necessary for the project, including Lattice-Boltzmann methods, thin fluid feature implementation details, and rigid-body to fluid collision details. I have read the papers on thin fluid features written by elbeem's creators, and it should be sufficient to start work on the product, however, more research into papers written by other developers could prove useful.
>
> 2) Development of thin fluid extension for elbeem
>
> 3) Addition of rigid-body collision code to elbeem
>
> 4) Creation of user interface for above 2 phases
>
> 5) Creation of documentation, for both users and future developers, in both textual and video/image form regarding phases 2-4.
>
>
>
> Project Schedule
>
> I am available to start heavy work on the project almost immediately following the conclusion of this semester (should be able to start sometime around May 14th). I do expect to have to take some time off (3-4 days) for my two race team competitions, the first of which runs from May 24th - 27th, and the second runs from June 20th - 23rd.
>
> I plan to use the first week (starting may 14th) to do more research into other developer's methods of implementing thin fluid features and rigid body collisions (as well as brushing up on my knowledge of Lattice Boltzmann). The next four weeks will be spent working on implementing thin fluid features (ending sometime around June 15th ish), and the following four weeks will be used implementing rigid body collisions (ending sometime around July 13th ish). This will give me around 3-4 weeks to test the code, find + squash bugs, and develop a clean user interface with documentation to match.
>
>
> Bio
>
> I am a current junior in Computer Engineering at the University of Oklahoma. I particularly enjoy video gaming (who doesn't?), photography, and motorsports (namely SCCA autocross and karting - road tracks are too expensive for my budget). I am heavily involved in my school's FSAE race team, where we design and build a small 70-ish horsepower formula race car from the ground up each school year to compete with during the summer. On the race team, I am currently the lead designer of 2 systems, primarily the electronics/data acquisition team and the aerodynamics team. On the electronics team, I am responsible for designing, building, and programming (in C++ and some Labview) the car's data acquisition and control systems (including the electronic shifting, driver interface, active aerodynamics controller, and others). As the aerodynamics team lead, I am responsible for designing, building, and implementing a full aero package (consisting of front and rear wings with the possibil
>  ity of a diffuser) for the car. This work involves lots of CFD and vehicle dynamics simulations for the initial design, with lots of engineering validation with collected data following the car being built.
>
>
> My experience with computer graphics programs is mostly on the engineering side of things. I have extensive experience with Autodesk Inventor, PTC Pro Engineer, and Autodesk AutoCAD for modeling, and have lots of experience in CD-Adapco Star-CCM+ and some experience with FLUENT for fluid simulation. I do have some (rather minor) experience with Blender in my past internships (NASA/JSC and Disney), but it almost entirely consists of just being able to convert files from engineering cad packages into a format the animators/producers can use in their software.
>
> In short, I am the best person to work on this project because of my computer engineering knowledge and experience from school and internships, as well as my mechanical engineering/simulation experience from the race team. I believe I have a skill set which would allow me to quickly get up and running towards creating a quality end product for the Blender project.
>
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> http://lists.blender.org/mailman/listinfo/bf-committers