During hack week I will carry on the research I started to write my thesis. The work is called "Turbulence Driven Clustering in Nematic Active Particles" and will (hopefully) make it into a Physical Review Letters paper.

TASKS:

1) actually write the article, I assume this task will take (no more than 10% time)

2) add some features to the C code I used for the simulations.

2) In detail: the simulations generate turbulence through the use of the Computational Fluid Dynamics method devised by Robert Kraichnan (Kraichnan Method) using a parameter kappa. This method is computationally efficient but it's hard to relate to real physical systems. To do this, some more information is needed to translate the kappa parameter into general number describing turbulent flows, the Reynolds Number (wiki page at the bottom). The features I want to implement are the data analysis needed to get that additional information.

3) IF there's some spare time (highly unlikely) it would be extremely cool to refactor all the tools used to be a bit more reusable and general (by the UNIX principles) and to open source them.

RESOURCES

KRAICHNAN METHOD:

Kraichnan R H. "Diffusion by a Random Velocity Field". In: The Physics of Fluids (1970)

Re Number:

https://en.wikipedia.org/wiki/Reynolds_number

The tools used:

• a bunch of shell scripts to submit jobs to clusters
• the C simulation - needs some features
• a C script used to calculate clustering - highly inefficient but relatively unimportant
• a bunch (around 10+) matlab scripts to plot the data - the most interesting one uses Voronoi tessellation to calculate local density (and it takes hours and hours to do so)