Commit b04caa75 authored by Markus Holzer's avatar Markus Holzer Committed by Jan Hönig
Browse files

Dataclasses for method creation

parent 4e47c1eb
......@@ -143,7 +143,7 @@ ubuntu:
before_script:
- apt-get -y remove python3-sympy
- ln -s /usr/include/locale.h /usr/include/xlocale.h
- pip3 install `grep -Eo 'sympy[>=]+[0-9\.]+' setup.py | sed 's/>/=/g'`
# - pip3 install `grep -Eo 'sympy[>=]+[0-9\.]+' setup.py | sed 's/>/=/g'`
- pip3 install git+https://gitlab-ci-token:${CI_JOB_TOKEN}@i10git.cs.fau.de/pycodegen/pystencils.git@master#egg=pystencils
script:
- export NUM_CORES=$(nproc --all)
......
......@@ -16,11 +16,12 @@ It even comes with an integrated Chapman Enskog analysis based on sympy!
Common test scenarios can be set up quickly:
```python
from lbmpy.scenarios import create_channel
from pystencils import Target
from lbmpy.session import *
ch = create_channel(domain_size=(300,100, 100), force=1e-7, method="trt",
ch = create_channel(domain_size=(300, 100, 100), force=1e-7, method=Method.TRT,
equilibrium_order=2, compressible=True,
relaxation_rates=[1.97, 1.6], optimization={'target': 'gpu'})
relaxation_rates=[1.97, 1.6], optimization={'target': Target.GPU})
```
To find out more, check out the interactive [tutorial notebooks online with binder](https://mybinder.org/v2/gh/mabau/lbmpy/master?filepath=doc%2Fnotebooks).
......
......@@ -29,8 +29,8 @@ templates_path = ['_templates']
source_suffix = '.rst'
master_doc = 'index'
copyright = f'{datetime.datetime.now().year}, Martin Bauer'
author = 'Martin Bauer'
copyright = f'{datetime.datetime.now().year}, Martin Bauer, Markus Holzer'
author = 'Martin Bauer, Markus Holzer'
# The short X.Y version (including .devXXXX, rcX, b1 suffixes if present)
version = re.sub(r'(\d+\.\d+)\.\d+(.*)', r'\1\2', lbmpy.__version__)
version = re.sub(r'(\.dev\d+).*?$', r'\1', version)
......
This source diff could not be displayed because it is too large. You can view the blob instead.
This diff is collapsed.
%% Cell type:markdown id: tags:
# Shan-Chen Two-Phase Single-Component Lattice Boltzmann
%% Cell type:code id: tags:
``` python
from lbmpy.session import *
from lbmpy.updatekernels import create_stream_pull_with_output_kernel
from lbmpy.macroscopic_value_kernels import macroscopic_values_getter, macroscopic_values_setter
from lbmpy.maxwellian_equilibrium import get_weights
```
%% Cell type:markdown id: tags:
This is based on section 9.3.2 of Krüger et al.'s "The Lattice Boltzmann Method", Springer 2017 (http://www.lbmbook.com).
Sample code is available at [https://github.com/lbm-principles-practice/code/](https://github.com/lbm-principles-practice/code/blob/master/chapter9/shanchen.cpp).
%% Cell type:markdown id: tags:
## Parameters
%% Cell type:code id: tags:
``` python
N = 64
omega_a = 1.
g_aa = -4.7
rho0 = 1.
stencil = get_stencil("D2Q9")
stencil = LBStencil(Stencil.D2Q9)
weights = get_weights(stencil, c_s_sq=sp.Rational(1,3))
```
%% Cell type:markdown id: tags:
## Data structures
%% Cell type:code id: tags:
``` python
dim = len(stencil[0])
dh = ps.create_data_handling((N,)*dim, periodicity=True, default_target='cpu')
src = dh.add_array('src', values_per_cell=len(stencil))
dst = dh.add_array_like('dst', 'src')
ρ = dh.add_array('rho')
```
%% Cell type:markdown id: tags:
## Force & combined velocity
%% Cell type:markdown id: tags:
The force on the fluid is
$\vec{F}_A(\vec{x})=-\psi(\rho_A(\vec{x}))g_{AA}\sum\limits_{i=1}^{q}w_i\psi(\rho_A(\vec{x}+\vec{c}_i))\vec{c}_i$
with
$\psi(\rho)=\rho_0\left[1-\exp(-\rho/\rho_0)\right]$.
%% Cell type:code id: tags:
``` python
def psi(dens):
return rho0 * (1. - sp.exp(-dens / rho0));
```
%% Cell type:code id: tags:
``` python
zero_vec = sp.Matrix([0] * dh.dim)
force = sum((psi(ρ[d]) * w_d * sp.Matrix(d)
for d, w_d in zip(stencil, weights)), zero_vec) * psi(ρ.center) * -1 * g_aa
```
%% Cell type:markdown id: tags:
## Kernels
%% Cell type:code id: tags:
``` python
collision = create_lb_update_rule(stencil=stencil,
relaxation_rate=omega_a,
compressible=True,
force_model='guo',
force=force,
kernel_type='collide_only',
lbm_config = LBMConfig(stencil=stencil, relaxation_rate=omega_a, compressible=True,
force_model=ForceModel.GUO, force=force, kernel_type='collide_only')
collision = create_lb_update_rule(lbm_config=lbm_config,
optimization={'symbolic_field': src})
stream = create_stream_pull_with_output_kernel(collision.method, src, dst, {'density': ρ})
opts = {'cpu_openmp': False,
'target': dh.default_target}
stream_kernel = ps.create_kernel(stream, **opts).compile()
collision_kernel = ps.create_kernel(collision, **opts).compile()
```
%% Cell type:markdown id: tags:
## Initialization
%% Cell type:code id: tags:
``` python
method_without_force = create_lb_method(stencil=stencil, relaxation_rate=omega_a, compressible=True)
method_without_force = create_lb_method(LBMConfig(stencil=stencil, relaxation_rate=omega_a, compressible=True))
init_assignments = macroscopic_values_setter(method_without_force, velocity=(0, 0),
pdfs=src.center_vector, density=ρ.center)
init_kernel = ps.create_kernel(init_assignments, ghost_layers=0).compile()
```
%% Cell type:code id: tags:
``` python
def init():
for x in range(N):
for y in range(N):
if (x-N/2)**2 + (y-N/2)**2 <= 15**2:
dh.fill(ρ.name, 2.1, slice_obj=[x,y])
else:
dh.fill(ρ.name, 0.15, slice_obj=[x,y])
dh.run_kernel(init_kernel)
```
%% Cell type:markdown id: tags:
## Timeloop
%% Cell type:code id: tags:
``` python
sync_pdfs = dh.synchronization_function([src.name])
sync_ρs = dh.synchronization_function([ρ.name])
def time_loop(steps):
dh.all_to_gpu()
for i in range(steps):
sync_ρs()
dh.run_kernel(collision_kernel)
sync_pdfs()
dh.run_kernel(stream_kernel)
dh.swap(src.name, dst.name)
dh.all_to_cpu()
```
%% Cell type:code id: tags:
``` python
def plot_ρs():
plt.figure(dpi=200)
plt.title("$\\rho$")
plt.scalar_field(dh.gather_array(ρ.name), vmin=0, vmax=2.5)
plt.colorbar()
```
%% Cell type:markdown id: tags:
## Run the simulation
### Initial state
%% Cell type:code id: tags:
``` python
init()
plot_ρs()
```
%% Output
%% Cell type:markdown id: tags:
### Check the first time step against reference data
The reference data was obtained with the [sample code](https://github.com/lbm-principles-practice/code/blob/master/chapter9/shanchen.cpp) after making the following changes:
```c++
const int nsteps = 1000;
const int noutput = 1;
```
Remove the next cell if you changed the parameters at the beginning of this notebook.
%% Cell type:code id: tags:
``` python
init()
time_loop(1)
ref = np.array([0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.136756, 0.220324, 1.2382, 2.26247, 2.26183, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.1, 2.26183, 2.26247, 1.2382, 0.220324, 0.136756, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15, 0.15])
assert np.allclose(dh.gather_array(ρ.name)[N//2], ref)
```
%% Cell type:markdown id: tags:
### Run the simulation until converged
%% Cell type:code id: tags:
``` python
init()
time_loop(1000)
plot_ρs()
```
%% Output
......
This source diff could not be displayed because it is too large. You can view the blob instead.
This source diff could not be displayed because it is too large. You can view the blob instead.
This diff is collapsed.
This diff is collapsed.
......@@ -5,6 +5,7 @@ API Reference
:maxdepth: 1
scenarios.rst
enums.rst
kernelcreation.rst
methodcreation.rst
stencils.rst
......
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment