Using reynolds¶

Requirements¶

Python 3.6 or later
OpenFoam version 4.x
macOS 10.10 or later
Ubuntu 14.x or later

Pre-requisites¶

You need to install OpenFoam version 4.x on macOS or Ubuntu.

Please note that reynolds depends on an OpenFoam sparsbundle on macOS.

The following are recommended guides for installing openfoam:

Installation¶

Clone the repository from Github:

git clone git@github.com:dmsurti/reynolds.git

Install the requirements:
```
pip install -r requirements.txt
```
Run setup.py:
```
python setup.py install
```

Starting OpenFoam¶

To use OpenFoam in your python environment, you can use the FoamRunner class which will source the openfoam environment variables. See the code listing below:

from reynolds.foam.start import FoamRunner

foam_runner = FoamRunner()
foam_runner.start()

This loads your environment with various openFoam utilities such as blockMesh and various solvers so they can be executed with a python process using POpen.

Generating blockMeshDict¶

To generate a blockMeshDict for the cavity tutorial, you could do:

from reynolds.dict.parser import ReynoldsFoamDict

block_mesh_dict = ReynoldsFoamDict('blockMeshDict.foam')
self.assertIsNotNone(block_mesh_dict)

# add vertices
vertices = []
vertices.append([0, 0, 0])
vertices.append([1, 0, 0])
vertices.append([1, 1, 0])
vertices.append([0, 1, 0])
vertices.append([0, 0, 0.1])
vertices.append([1, 0, 0.1])
vertices.append([1, 1, 0.1])
vertices.append([0, 1, 0.1])
block_mesh_dict['vertices'] = vertices

# add blocks
blocks = []
blocks.append('hex')
blocks.append([0, 1, 2, 3, 4, 5, 6, 7])
blocks.append([20, 20, 1])
blocks.append('simpleGrading')
blocks.append([1, 1, 1])
block_mesh_dict['blocks'] = blocks
self.assertEqual(block_mesh_dict['blocks'],
                  ['hex', [0, 1, 2, 3, 4, 5, 6, 7], '(20 20 1)',
                  'simpleGrading', '(1 1 1)'])

# add edges
self.assertEqual(block_mesh_dict['edges'], [])
edges = []
edges.append('arc')
edges.append(1)
edges.append(5)
edges.append([1.1, 0.0, 0.5])
block_mesh_dict['edges'] = edges

boundary = []
# add moving wall
boundary.append('movingWall')
moving_wall = {}
moving_wall['faces'] = [[3, 7, 6, 2]]
moving_wall['type'] = 'wall'
boundary.append(moving_wall)
# add fixed walls
boundary.append('fixedWalls')
fixed_walls = {}
fixed_walls['faces'] = [[0, 4, 7, 3], [2, 6, 5, 1], [1, 5, 4, 0]]
fixed_walls['type'] = 'wall'
boundary.append(fixed_walls)
# add front and back
boundary.append('frontAndBack')
front_and_back = {}
front_and_back['faces'] = [[0, 3, 2, 1], [4, 5, 6, 7]]
front_and_back['type'] = 'empty'
boundary.append(front_and_back)
block_mesh_dict['boundary'] = boundary

# add mergePatchPairs
mergePatchPairs = []
mergePatchPairs.append(['inlet1', 'outlet1'])
mergePatchPairs.append(['inlet2', 'outlet2'])
block_mesh_dict['mergePatchPairs'] = mergePatchPairs

print(block_mesh_dict)

The above generates an in memory blockMeshDict. To write this to a file on disk, you can do:

# case_dir is the absolute path to your case directory on disk
file_path = os.path.join(case_dir, 'system', 'blockMeshDict')

with open(file_path, 'w') as f:
    f.write(str(block_mesh_dict))

Running a solver¶

You can run any openfoam solver available in the openfoam environment which has been sourced, see Installation instructions. For example, to run the icoFoam solver used in the cavity tutorial, you can do:

from reynolds.foam.cmd_runner import FoamCmdRunner

# case_dir is the absolute path to your case directory on disk
solver_runner = FoamCmdRunner(cmd_name='icoFoam', case_dir=cavity_case_dir)
for info in solver_runner.run():
    pass # client can stream this info live
if solver_runner.run_status: # All is well
    print("Success")
else:
    print("Failure")

On exactly the same lines, you can run any other OpenFoam command such as blockMesh using the FoamCmdRunner.

Running with Blender using Docker¶

You can use Blender with an add-on that invokes this reynolds API to start openfoam, generate a blockMeshDict and run a solver.

The simplest way to run Blender with this addon is to use this Docker file, which can be installed on Ubuntu, and runs the Blender GUI with this add-on.

You can refer to the docker image repository homepage for instructional videos.