peclet-dem (peclet.dem)

PyPI version Python License: MIT CI DOI

Performance-portable Discrete Element Method (DEM) particle simulation: an XPBD solver with SDF-based point-shell collision detection. Built on Kokkos + ArborX, so the same source runs on CUDA, HIP (AMD/LUMI), and OpenMP backends (selected at build time by the install prefix). Optional MPI for domain partitioning, with nanobind Python bindings (zero-copy, via scikit-build-core) for scripting and visualization.

The CUDA implementation was retired (2026-06): the Kokkos peclet.dem module was validated against it before the CUDA sources were removed. Restore point: git tag pre-cuda-retirement.

Features

  • Hybrid XPBD Solver: Two-pass velocity/position solver for stable high-density packing.
  • SDF Collision: Point-shell collision detection using Signed Distance Fields (supports analytic shapes like hollow cylinders).
  • Periodicity: Full periodic boundary conditions (Ghost Particles).
  • Python Bindings: Control simulation logic, data initialization, and export entirely from Python.
  • MPI Support: Optional Multi-GPU/Node support via domain decomposition.

Folder Structure

├── CMakeLists.txt              # Build configuration (find_package Kokkos + ArborX)
├── src                         # Kokkos sources (header-only, namespace peclet::dem)
│   ├── dem_bindings.cpp        # nanobind module entry point (the `peclet.dem` module)
│   ├── sim.hpp                 # Simulation facade + the demStep XPBD substep
│   ├── integration.hpp         # Time integration & prediction
│   ├── broadphase_arborx.hpp   # ArborX BVH broad-phase
│   ├── narrowphase.hpp         # Narrow-phase point-shell-vs-SDF collision
│   ├── solver_velocity.hpp     # Velocity solver kernels
│   ├── solver_position.hpp     # Position solver kernels (XPBD overlap removal)
│   ├── solver_friction.hpp     # Coulomb friction cluster
│   ├── output_sdf.hpp          # SDF/VTI grid generation (Eikonal)
│   ├── shapes_portable.hpp     # Analytic shapes (sphere / hollow cylinder / box)
│   ├── io.hpp                  # LAMMPS-dump + SDF-VTI export
│   └── mpi_halo.hpp            # Distributed particle halo (core), gated PECLET_DEM_MPI
├── tests                       # C++ unit tests: kokkos/ (kernels), arborx/, kokkos_mpi/
├── docs                        # Documentation
└── *.py                        # Python verification/example scripts (verify_*.py)

Prerequisites

  • Linux
  • CMake >= 3.24
  • Kokkos 5.x + ArborX (C++20) — provisioned by ../tools/bootstrap_deps.sh into ../extern/install/<backend> (nvidia-cuda / host-openmp / lumi-hip). A hard build dependency.
  • nanobind + scikit-build-core (found via the active Python interpreter; see pyproject.toml)
  • a backend compiler: nvcc (CUDA) on PATH, hipcc (ROCm), or just a host C++ compiler (OpenMP)
  • Python >= 3.10
  • MPI (optional, -DDEM_MPI=ON) — OpenMPI or MPICH

Build Instructions

python -m venv .venv && source .venv/bin/activate && pip install nanobind numpy
export PATH=/usr/local/cuda-13.2/bin:$PATH        # if building the CUDA backend

# Canonical: build + install the module via scikit-build-core
CMAKE_PREFIX_PATH="$PWD/../extern/install/nvidia-cuda" pip install .

# Or a dev cmake build (nanobind is found via the active interpreter, no cmakedir needed):
cmake -B build -S . -DCMAKE_PREFIX_PATH="$PWD/../extern/install/nvidia-cuda"
cmake --build build -j$(nproc)

Swap the prefix to ../extern/install/host-openmp for the OpenMP backend. -DDEM_MPI=ON links MPI and exposes the distributed step (init_mpi / enable_mpi_step / step_mpi), including dynamic load balancing — enable_mpi_step(..., rebalance_every=N) or an explicit rebalance() re-decomposes by particle count (weighted ORB) and migrates ownership so each rank keeps a near-equal share.

The compiled peclet.dem extension is placed in build/peclet/dem/; run scripts with build/ on PYTHONPATH (import peclet.dem).

Running Simulations

Example scripts are provided in the root directory:

# Add build artifact to python path if needed (or symlink it)
export PYTHONPATH=$PYTHONPATH:$(pwd)/build

# Run a verification script
python verify_packing_hollow_cylinders.py

Output & Visualization

The simulation supports two primary output formats:

1. LAMMPS + STL (Ovito)

For particle visualization (especially non-spherical shapes), we use the LAMMPS dump format combined with an STL mesh.

  1. Generate Output: The simulation writes dump.custom.* files.
  2. Generate Shape: Run python generate_particles.py to create particle_shape.stl.
  3. Visualize:
    • Open Ovito.
    • Load the dump.custom.* sequence.
    • Add a Particle Types modifier.
    • Set the shape visualization to Mesh/User-defined and load particle_shape.stl.
    • Ovito will automatically scale the mesh by the particle radius.

See docs/visualization.md for a detailed guide.

2. VTI (ParaView)

For visualizing fields (like the Signed Distance Field or occupancy grids), the simulation exports VTI files (.vti).

  1. Generate Output: Use Simulation.export_sdf("filename.vti", resolution=...).
  2. Visualize:
    • Open ParaView.
    • Load the .vti file.
    • Use "Volume" representation or "Slice" filter to inspect the field.

Status

The single-GPU engine is complete and validated: it reaches stable high-density (random close) packing, and energy is conserved to ~0.3% (see docs/packing_investigation.md). Friction is stabilized for spheres; body-body tangential friction is a known follow-up (currently weaker than ideal). Active work is at-scale multi-GPU/MPI tuning.

[!NOTE] The distributed (MPI) step is validated against the single-rank result (tests/kokkos_mpi, np=1,2,4 on OpenMP + CUDA) and supports dynamic load rebalancing; remaining MPI work is at-scale multi-GPU tuning.

Project links

Data verified by PyPI on Sep 6, 2026
Data provided by the project maintainers, verified at the time the release was uploaded to PyPI.

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Latest release

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for eajfpeters from gravatar.com eajfpeters

Credits

Author: Frank Peters

GitHub Statistics

Data verified by PyPI on Sep 6, 2026
The GitHub source repository was provided by the project maintainers and verified by PyPI at the time of upload. Stars, forks, and open issues/PRs are derived from that repository and have not been independently verified.

License expression

MIT
View SPDX License List

Requires

Python >=3.10

Report project as malware

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

peclet_dem-0.5.1.tar.gz (286.8 kB view details)

Uploaded Source

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

peclet_dem-0.5.1-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (1.1 MB view details)

Uploaded CPython 3.13manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

peclet_dem-0.5.1-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (1.1 MB view details)

Uploaded CPython 3.12manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

peclet_dem-0.5.1-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (1.1 MB view details)

Uploaded CPython 3.11manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

peclet_dem-0.5.1-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (1.1 MB view details)

Uploaded CPython 3.10manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

File details

Details for the file peclet_dem-0.5.1.tar.gz.

File metadata

  • Download URL: peclet_dem-0.5.1.tar.gz
  • Upload date:
  • Size: 286.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for peclet_dem-0.5.1.tar.gz
Algorithm Hash digest
SHA256 8aa3b38dbc6222e50b640e038915d63836c34bdae865e67b33be2fdf2e0494fa
MD5 b222e19e01f27cb7a41263b2c5612276
BLAKE2b-256 16c14e7293f9374a505130eb90056d5434bad39fdf74dafcc0c1c7bfc72992f4

See more details on using hashes here.

Provenance

The following attestation bundles were made for peclet_dem-0.5.1.tar.gz:

Publisher: release.yml on computational-chemical-engineering/peclet-dem

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file peclet_dem-0.5.1-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for peclet_dem-0.5.1-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 721a6ba9024f36aa6c5787f44aa112177c27b5f1ef2f6df21d99ef038d0cd224
MD5 b7179b0c613b7e2c0c630eadb06816df
BLAKE2b-256 5b22556ec0578511781c6a59014ecd8f945ec26558e0341f81d63ff13eb4f83c

See more details on using hashes here.

Provenance

The following attestation bundles were made for peclet_dem-0.5.1-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: release.yml on computational-chemical-engineering/peclet-dem

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file peclet_dem-0.5.1-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for peclet_dem-0.5.1-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 b106f09f6ab08ef7bef6c99ca2b180fac977052d162583a98f90b607067a8a08
MD5 19fe0dbfb25817c154016858e1595f11
BLAKE2b-256 8e20c4f486c20aa9d106c414fb829017fa6c789d77d7a99cb507e01cc1b46713

See more details on using hashes here.

Provenance

The following attestation bundles were made for peclet_dem-0.5.1-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: release.yml on computational-chemical-engineering/peclet-dem

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file peclet_dem-0.5.1-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for peclet_dem-0.5.1-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 0ccc7f040d43443a15cb1802b21c78fab6adff3256447a1b930120c56b62300c
MD5 256a269f9a5fb5a2172df5371829117f
BLAKE2b-256 232f771bd1934591b2d5e6122b565a69e6bbb4b22fac5dd61e1d823bc52e28d1

See more details on using hashes here.

Provenance

The following attestation bundles were made for peclet_dem-0.5.1-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: release.yml on computational-chemical-engineering/peclet-dem

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file peclet_dem-0.5.1-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for peclet_dem-0.5.1-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 9bd4995303dad39fd89c4304856f92974dfa7c1aa1d3d358f496beabf1a7fe7c
MD5 c6669b887cd8f97aac852a55bb3b731f
BLAKE2b-256 5b8088cdfa349d848c1a4a5af868683f7032d0bb1a8b05959d09682717342375

See more details on using hashes here.

Provenance

The following attestation bundles were made for peclet_dem-0.5.1-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: release.yml on computational-chemical-engineering/peclet-dem

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Release history Release notifications | RSS feed

This release

0.5.1 This release

5 files

0.5.0

5 files

0.4.0

5 files

0.3.2

5 files

0.3.1

5 files

0.3.0

5 files

0.2.1

5 files

0.2.0

5 files

0.1.0

5 files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page