ml-evs / vfmcpp

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

vfmcpp

vfmcpp is a C++ implementation of the vortex filament model of superfluids, used to study the dynamics of ring/ring and ring/line scattering.

A ring-line collision

To compile in Linux, you can use the makefile make; it is advised you use GCC 4.7+, as that is all that has been tested. Help with initial conditions can be found by running run.sh with the flag -h. The OpenMP version can be found here.

Visualisation is currently performed by a 3D matplotlib script (requires latest matplotlib version), which leaves a lot to be desired. Below are some dated animations (with old version of reconnection) to give an example visualisations only. This must be run with the data folder as an argument, e.g.

python animate.py ~/vfmcpp/data/example_run

run.sh

DESCRIPTION:

Run the vfmcpp code.

USAGE:

run.sh [OPTIONS] [FILENAME]

OPTIONS:

-c
	Recompile the code.
-d
	Run the code inside the gdb debugger.
-b 
	Run benchmarking and profiling versions.
-h
	Show usage information.

Flags cannot be combined.

FILENAME:

The initial condition file can be specified; defaults to "run.in".

INITIAL CONDITIONS:

path 
	data folder path relative to vfmcpp root folder (will be created if doesn't exist or emptied if exists)

time	
	total length of time to simulate

resl	
	specify the resolution of the simulation in metres, filaments will be created, default=6.28e-8 (1um radius ring w/ 100 pts) 

ring [radius, x, y, z, alignment]
	make a closed filament, requires 5 arguments as above, where the 5th is one of [0,1,2] for axis to align to.

line [length, x, y, z, alignment]
	make an open filament, requires 5 arguments as above.

delayed_ring [radius, x, y, z, alignment, t]
	additional argument for time t in ms to delay the addition of the ring.

lfil [path]
	reads positions and velocities of a premade (normally distorted) line from path/pos.dat and path/vel.dat respectively.

Eext [strength, duration, direction]
	include an external electric field, requires 3 arguments where direction is either 0, 1 or 2 for x, y or z.

q_pt [filament, point, magnitude]
	add a charge to a filament, requires filaments to be specified eariler in file, 
	filament is the relative position of the filament in the .in file, (first filament = 0), 
	point is the index of the desired charged point on the filament, and magnitude is the size of the charge.

#
	indicates a comment, which will be ignored by vfmcpp

EXAMPLE FILE:

path data/init_example
time 1e-3
resl 6.28e-8
ring 1e-6 0 0 5e-6 0 
ring 9e-7 0 0.025e-6 0 0
Eext 10000 1e-3 0 
q_pt 0 50 1.6e-19

About


Languages

Language:C++ 57.5%Language:Python 35.6%Language:Shell 6.1%Language:Makefile 0.8%