ALPHANSO: ALPHA Neutron SOurces

Open-source Python package for modeling ($\alpha$,n) neutron source terms.

Motivation

Legacy ($\alpha$,n) codes such as SOURCES-4C are written in FORTRAN 77 with nuclear data from the 1980s and have not been maintained since 2002. ALPHANSO provides a modern, open-source replacement built on up-to-date evaluated nuclear data libraries stored in the GNDS format. It covers all naturally occurring target nuclides and agrees well with experimental data.

Applications include reactor design, nuclear safeguards, radioactive waste management, nuclear astrophysics, and dark matter detection experiments.

Key Features

• Multiple Geometry Types: Beam, homogeneous, interface, and sandwich configurations
• Accurate Physics: Up-to-date nuclear data with customizable data sources
• Complete Output: Neutron yields and energy spectra, plus the alpha-induced prompt gamma source term (yields and discrete lines, validated against SaG4n/Geant4)
• Command-Line & Python API: Use via CLI or integrate into Python workflows

Installation

From PyPI

pip install alphanso

On first use, ALPHANSO automatically downloads ~1.1 GB of nuclear data files and caches them locally (e.g., ~/.local/share/alphanso/ on Linux). To pre-download the data:

alphanso download-data

From Source

git clone https://github.com/alphanso-org/alphanso.git
cd alphanso
pip install -e .

When installed from source, ALPHANSO uses the bundled alphanso/data/ directory directly (no download needed).

Custom Data Location

Set the ALPHANSO_DATA_DIR environment variable to override the data directory:

export ALPHANSO_DATA_DIR=/path/to/nuclear/data

Use alphanso data-info to check the current data paths and status.

Quick Start

Command-Line Interface

Run ALPHANSO with a YAML configuration file:

alphanso path/to/config.yaml

Or equivalently:

python3 -m alphanso path/to/config.yaml

Results are saved to alphanso_output/<config_name>/. You can specify a custom output directory as a second CLI argument: alphanso config.yaml my_output/.

Example YAML Configurations

Beam — monoenergetic alpha beam on a thick target:

name: "Be-9 Beam Calculation"
calc_type: "beam"
matdef:
  Be-9: 1.0
beam_energy: 5.0

Homogeneous — uniform mixture of alpha emitters and targets:

name: "Homogeneous Source"
calc_type: "homogeneous"
matdef:
  Pu-239: 0.3
  Pu-238: 0.2
  Be-9: 0.5

Interface — planar interface between source and target regions:

name: "Interface Geometry"
calc_type: "interface"
source_matdef:
  Pu-238: 1.0
source_density: 19.8
target_matdef:
  Be-9: 1.0

Sandwich — multi-layer geometry with intermediate layers:

name: "Multi-Layer Sandwich"
calc_type: "sandwich"
source_matdef:
  Pu-238: 1.0
source_density: 19.8
target_matdef:
  Be-9: 1.0
intermediate_layers:
  - matdef: { C-13: 1.0 }
    density: 2.26
    thickness: 1.0e-4
  - matdef: { Al-27: 1.0 }
    density: 2.70
    thickness: 1.0e-4

Python API

from alphanso.transport import Transport

config = {
    "calc_type": "beam",
    "matdef": {"Be-9": 1.0},
    "beam_energy": 5.0
}

results = Transport.calculate(config)
print(f"Neutron yield: {results['an_yield']:.3e} n/alpha")

Transport.calculate() accepts optional keys in the config dict:

• name (str): Label for identification in CLI output
• output_dir (str): Directory to save results.yaml.
• save_data_files (bool, default: True): Whether to save output files when output_dir is specified. Set to False to skip file saving.

Running Tests

pytest

Calculation Types

ALPHANSO supports four calculation geometries, each designed for specific physical scenarios.

1. Beam Calculations (calc_type: "beam")

Monoenergetic or polyenergetic alpha beam incident on a thick target.

Required Parameters:

• matdef (dict): Target material composition (ZAID or element names with mass fractions)
• beam_energy (float): Alpha beam energy in MeV (monoenergetic)
• beam_intensities (list, optional): List of [energy, intensity] pairs for polyenergetic beams (use instead of beam_energy)

Optional Parameters:

• num_alpha_groups (int, default: 15000): Number of alpha energy groups
• min_alpha_energy (float, default: 1e-11): Minimum alpha energy in MeV
• max_alpha_energy (float, default: 15): Maximum alpha energy in MeV
• neutron_energy_bins (list or NumPy array, optional): Neutron energy bin edges in MeV. Three-element lists are shorthand [start, stop, num_points] for linearly spaced edges; lists of other lengths and all NumPy arrays specify explicit edges. Default: 101 edges spanning 0–15 MeV (100 bins).
• an_xs_data_dir (str): Path to ($\alpha$,n) cross-section data directory
• stopping_power_data_dir (str): Path to stopping power data directory

Results:

• an_yield: Neutron production rate (n/$\alpha$)
• an_spectrum: Normalized neutron energy spectrum
• an_spectrum_absolute: Absolute neutron spectrum
• neutron_energy_bins: Energy bins for spectrum (MeV)

---

2. Homogeneous Calculations (calc_type: "homogeneous")

Uniform mixture of alpha-emitting isotopes and target materials.

Required Parameters:

• matdef (dict): Material composition including both alpha emitters and targets

Optional Parameters:

• num_alpha_groups (int, default: 15000): Number of alpha energy groups
• min_alpha_energy (float, default: 1e-11): Minimum alpha energy in MeV
• max_alpha_energy (float, default: 15): Maximum alpha energy in MeV
• neutron_energy_bins (list or NumPy array, optional): Neutron energy bin edges in MeV. Three-element lists are shorthand [start, stop, num_points] for linearly spaced edges; lists of other lengths and all NumPy arrays specify explicit edges. Default: 101 edges spanning 0–15 MeV (100 bins).
• an_xs_data_dir (str): Path to cross-section data
• stopping_power_data_dir (str): Path to stopping power data
• decay_data_dir (str): Path to decay/branching data

Results:

• an_yield: Alpha-n neutron yield (n/s/g)
• sf_yield: Spontaneous fission neutron yield (n/s/g)
• combined_yield: Combined (alpha-n + SF) yield (n/s/g) - default output
• an_spectrum: Normalized alpha-n neutron spectrum
• sf_spectrum: Normalized spontaneous fission neutron spectrum
• combined_spectrum: Combined normalized neutron spectrum
• neutron_energy_bins: Energy bins (MeV)

---

3. Interface Calculations (calc_type: "interface")

Planar interface between an alpha-emitting source region and a target region.

Required Parameters:

• source_matdef (dict): Alpha source material composition
• source_density (float): Source density in g/cm^3
• target_matdef (dict): Target material composition

Optional Parameters:

• num_alpha_groups (int, default: 15000): Number of alpha energy groups
• min_alpha_energy (float, default: 1e-11): Minimum alpha energy in MeV
• max_alpha_energy (float, default: 15): Maximum alpha energy in MeV
• neutron_energy_bins (list or NumPy array, optional): Neutron energy bin edges in MeV. Three-element lists are shorthand [start, stop, num_points] for linearly spaced edges; lists of other lengths and all NumPy arrays specify explicit edges. Default: 101 edges spanning 0–15 MeV (100 bins).
• an_xs_data_dir (str): Cross-section data path
• stopping_power_data_dir (str): Stopping power data path
• decay_data_dir (str): Decay data path

Results:

• an_yield: Neutron yield per cm^2 of interface (n/s/cm^2)
• an_spectrum: Normalized spectrum
• an_spectrum_absolute: Absolute spectrum (n/s/cm^2/MeV)
• neutron_energy_bins: Energy bins (MeV)

---

4. Sandwich Calculations (calc_type: "sandwich")

Multi-layer sandwich geometry with volumetric formulation. Alpha source (Region A) -> Intermediate layers (Region B1, B2, ..., Bn) -> Target (Region C).

Required Parameters:

• source_matdef (dict): Alpha source material composition (Region A)
• source_density (float): Source density in g/cm^3
• target_matdef (dict): Target material composition (Region C)
• intermediate_layers (list of dicts): One or more intermediate layers, each containing:
  • matdef (dict): Layer material composition
  • density (float): Layer density in g/cm^3
  • thickness (float): Layer thickness in cm

Optional Parameters:

• n_angular_bins (int, default: 40): Angular bins for integration
• num_alpha_groups (int, default: 15000): Number of alpha energy groups
• min_alpha_energy (float, default: 1e-11): Minimum alpha energy in MeV
• max_alpha_energy (float, default: 15): Maximum alpha energy in MeV
• neutron_energy_bins (list or NumPy array, optional): Neutron energy bin edges in MeV. Three-element lists are shorthand [start, stop, num_points] for linearly spaced edges; lists of other lengths and all NumPy arrays specify explicit edges. Default: 101 edges spanning 0–15 MeV (100 bins).
• an_xs_data_dir (str): Cross-section data path
• stopping_power_data_dir (str): Stopping power data path
• decay_data_dir (str): Decay data path

Results:

• an_yield: Total neutron yield (n/s/cm^2)
• yield_target: Yield in target region (n/s/cm^2)
• yield_layers: Per-layer yield breakdown (list of n/s/cm^2)
• yield_ab_b: Yield at AB interface in first layer material
• yield_bc_b: Yield at BC interface in last layer material
• yield_bc_c: Yield at BC interface in target material
• an_spectrum: Normalized total spectrum
• an_spectrum_absolute: Absolute total spectrum
• spectrum_layers: Per-layer spectrum contributions
• neutron_energy_bins: Energy bins (MeV)

Material Definition Format

Materials are defined using dictionaries with isotope/element identifiers as keys and mass fractions as values.

Supported Formats:

• Element-mass names: "Be-9", "C-13", "Al-27", "Pu-238"
• Natural element symbols: "C", "O" (auto-expands to natural isotopes using abundance data)
• ZAID integers: 4009 (Be-9), 6013 (C-13), 94238 (Pu-238)
• Natural element ZAIDs: 6000 (natural C), 8000 (natural O) — AAA=000 convention

Note: ZAID keys in Python dicts must be integers, not strings. Using string keys like "92235" will cause entries to be silently dropped.

Example:

matdef:
  Pu-238: 0.90
  Pu-239: 0.05
  Pu-240: 0.05

Or in Python:

matdef = {94238: 0.90, 94239: 0.05, 94240: 0.05}

Mass fractions should sum to 1.0 for single-phase materials, but can exceed 1.0 for compounds.

Output Files

When using the CLI, results are saved to alphanso_output/<config_name>/ as results.yaml.

Custom Data Sources

ALPHANSO allows you to use custom nuclear data:

an_xs_data_dir: "/path/to/cross_sections"
stopping_power_data_dir: "/path/to/stopping_data"
decay_data_dir: "/path/to/decay_data"

Paths can be absolute or relative to the project root. If not specified, ALPHANSO uses built-in default data.

To integrate new data formats, extend the parsers in alphanso/parsers.py.

Nuclear Data Sources

ALPHANSO's bundled nuclear data is derived from the following evaluated libraries and databases, whose authors and maintainers we gratefully acknowledge:

• ($\alpha$,n) cross sections and secondary-neutron distributions — the JENDL-5 $\alpha$ sub-library and JENDL/AN-2005 (Japan Atomic Energy Agency), the TENDL TALYS-based evaluated library, and ENDF/B-VIII.0, stored in GNDS format.
• Alpha-induced gamma production — the JENDLTENDL01 charged-particle library (JENDL/AN-2005 for the light target nuclides combined with TENDL-2017), distributed with the SaG4n code by E. Mendoza et al., together with residual-nucleus level schemes from the Geant4 neutron data library (G4NDL).
• Stopping powers — NIST ASTAR and SRIM.
• Decay and nuclear-structure data — ENDF/B-VIII.0 and RIPL-3 (IAEA).

The alpha-induced gamma source-term calculation uses the ($\alpha$,xn) data library and methodology of:

E. Mendoza, D. Cano-Ott, V. Pesudo, and R. Santorelli, "Neutron production induced by $\alpha$-decay with Geant4," Nuclear Instruments and Methods in Physics Research Section A 960, 163659 (2020). doi:10.1016/j.nima.2020.163659

Underlying evaluated libraries: TENDL (A. J. Koning et al., Nucl. Data Sheets 155, 1, 2019); JENDL-5 (O. Iwamoto et al., J. Nucl. Sci. Technol. 60, 1, 2023); Geant4 / G4NDL (S. Agostinelli et al., Nucl. Instrum. Methods A 506, 250, 2003).

Citation

If you use ALPHANSO in your research, please cite:

@article{rawal2026alphanso,
  title = {ALPHANSO: Open-source modeling of ($\alpha$, n) neutron source terms},
  author = {Divit Rawal and Anthony J. Nelson and William Zywiec and Daniel Siefman},
  journal = {Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment},
  year = {2026},
  pages = {171713},
  doi = {10.1016/j.nima.2026.171713},
  url = {https://www.sciencedirect.com/science/article/pii/S0168900226004390?via%3Dihub}
}

Contributing

Contributions are welcome. To contribute:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/my-feature)
3. Commit your changes and push to your fork
4. Open a pull request against main

Please open an issue first for bug reports or feature requests.

License

This project is licensed under the BSD-3-Clause License. See the LICENSE file for details.

Authors

• Divit Rawal - divit.rawal@berkeley.edu
• Anthony J. Nelson - nelson254@llnl.gov
• William Zywiec - zywiec1@llnl.gov
• Daniel Siefman - daniel.siefman@berkeley.edu

Support

For questions, issues, or feature requests, please open an issue on GitHub or contact the authors.