Getting Started
Just run
npx create-splitfxm-app
This creates an example boundary-value problem application that uses splitfxm
package for solution.
If you are just interested in the package, that can be installed using
pip install splitfxm
There is an examples folder that contains a test model - Advection-Diffusion
You can define your own equations by simply creating a derived class from Model
and adding to the _equations
using existing or custom equations!
A basic driver program is as follows
from splitfxm.domain import Domain
from splitfxm.simulation import Simulation
from splitfxm.models.advection_diffusion import AdvectionDiffusion
from splitfxm.schemes import default_scheme
from splitfxm.visualize import draw
import matplotlib.pyplot as plt
# Define the problem
method = 'FVM'
m = AdvectionDiffusion(c=0.2, nu=0.001, method=method)
# nx, nb_left, nb_right, variables
d = Domain.from_size(20, 1, 1, ["u", "v", "w"])
ics = {"u": "gaussian", "v": "rarefaction", "w": "tophat"}
bcs = {
"u": {
"left": "periodic",
"right": "periodic"
},
"v": {
"left": {"dirichlet": 3},
"right": {"dirichlet": 4}
},
"w": {
"left": {"dirichlet": 2},
"right": "periodic"
}
}
s = Simulation(d, m, ics, bcs, default_scheme(method))
# Advance in time or to steady state
s.evolve(t_diff=0.1)
bounds = [[-1., -2., -2.], [5., 4., 3.]]
iter = s.steady_state(split=True, split_locs=[1, 2], bounds=bounds)
# Visualize
draw(d, "label")
plt.show()