Collection of sphere packing and meshing algorithms.

Installation

pip3 install bola

The c++ code requires the math library Eigen3 to be installed and a dependency of the python gmsh package is libglu. So, you may need to run (debian/ubuntu-based):

sudo apt update
sudo apt -y install libeigen3-dev libglu1

Alternatively, you can follow the steps of the CI.

Examples

Particle size according to bola.psd.GradingCurve (sieve lines):

gc = bola.psd.GradingCurves.fuller(d_min=4., d_max=16)
box = (32.0, 32.0, 32.0)
radii = psd.sample_grading_curve(gc, V=0.5 * np.prod(box))

Initial packing using bola.packing.rsa (random sequential addition)

spheres = bola.packing.rsa(radii, box)

Maximize particle distance using bola.packing.edmd (event-driven molecular-dynamics)

sim = bola.packing.edmd(box, spheres, growth_rate=0.1)
while sim.t() < 10.0:
    sim.process(100 * len(radii))
    sim.synchronize(True)
    print(packing.stats_string(sim))
distant_spheres = sim.spheres()
distant_spheres[:, 3] = radii # new centers, old radii

Mesh via gmsh using bola.mesh

bola.mesh.create(
    box, new_spheres, bola.mesh.GmshOptions(
        mesh_size_matrix=2.0, mesh_size_aggregates=2.0, out="mesh.xdmf")
)