Particula

A simple, fast, and powerful particle simulator for aerosol science.

Requires: Python 3.12+

Documentation | Examples | PyPI

Installation

pip install particula

or via conda:

conda install -c conda-forge particula

Quick Start

import particula as par

# Build an aerosol system
aerosol = (
    par.AerosolBuilder()
    .set_atmosphere(atmosphere)
    .set_particles(particles)
    .build()
)

# Run dynamics (chainable with | operator)
process = par.dynamics.Condensation(strategy) | par.dynamics.Coagulation(strategy)
aerosol = process.execute(aerosol, time_step=10, sub_steps=1000)

Migration / What's New

See the migration guide for current APIs, compatibility notes, and direct-kernel troubleshooting.

Code Structure

particula/
├── gas/           # Gas phase: species, vapor pressure, atmosphere
├── particles/     # Particle representations & distributions
├── dynamics/      # Time-dependent processes
│   ├── coagulation/
│   ├── condensation/
│   └── wall_loss/
├── activity/      # Activity coefficients, phase separation
├── equilibria/    # Gas-particle partitioning
└── util/          # Constants, validation, unit conversion

Documentation Guide

Examples by Topic

• Aerosol — Building and inspecting aerosol objects
• Dynamics — Coagulation, condensation, wall loss simulations
• Equilibria — Gas-particle partitioning calculations
• Gas Phase — Vapor pressure, species properties
• Particle Phase — Size distributions, optical properties
• Simulations — Full end-to-end scientific scenarios

Featured Examples

• Aerosol Tutorial — Learn how to build gas species, atmospheres, particle distributions, and combine them into an Aerosol object.

• CPU Latent-Heat Condensation Bookkeeping — CPU-only walkthrough showing diagnostic latent-heat bookkeeping from real condensation mass transfer with no temperature feedback.

• Direct-Warp GPU Nucleation — Explicit-transfer, fixed-capacity low-level example using the package- exported nucleation_step_gpu. It requires Warp and is separate from the CPU-only Nucleation runnable; it provides no CPU fallback.

• Organic Partitioning & Coagulation — Full simulation of secondary organic aerosol (SOA) formation from 10 organic vapors, followed by Brownian coagulation over 10 minutes.

• Cloud Chamber Cycles — Multi-cycle cloud droplet activation demonstrating κ-Köhler theory across 3 seed compositions (Ammonium Sulfate, Sucrose, Mixed), showing how hygroscopicity affects activation at different supersaturations.

Features

• Gas & Particle Phases — Full thermodynamic modeling with swappable strategies
• Dynamics — Coagulation, condensation, wall loss, dilution
• Flexible Representations — Discrete bins, continuous PDF, particle-resolved
• Builder Pattern — Clean, validated object construction with unit conversion
• Composable Processes — Chain runnables with | operator
• Condensation Utilities — Non-isothermal helpers via particula.dynamics.get_thermal_resistance_factor, particula.dynamics.get_mass_transfer_rate_latent_heat, and particula.dynamics.get_latent_heat_energy_released
• Condensation Strategies — CondensationIsothermal plus CondensationLatentHeat with latent-heat-corrected mass_transfer_rate()/rate() and step() energy tracking via last_latent_heat_energy, with optional dynamic_viscosity override
• Latent Heat Factories — Build constant, linear, and power-law latent heat strategies via particula.gas.LatentHeatFactory with unit-aware builders and gas-phase exports for upcoming non-isothermal workflows

Citation

If you use Particula in your research, please cite:

Particula [Computer software]. DOI: 10.5281/zenodo.6634653

License

MIT