canterax 0.3.1

A differentiable, GPU-resident chemical kinetics library using JAX

Canterax

Differentiable, JAX-based chemical kinetics with a Cantera-like ideal-gas Solution API. https://canterax.readthedocs.io/en/latest/

Features

• Ideal-gas ThermoPhase-style Solution wrapper
• Cantera YAML mechanism loading
• NASA-7 thermodynamics and Arrhenius kinetics
• Cantera-like state setters including TP, HP, UV, SP, SV, TD, and DP
• Basis-aware aliases matching Cantera (basis = "mass" or "molar")
• Mixture viscosity and thermal_conductivity
• TP and HP equilibrium modes
• Differentiable reactor integration with JAX and Diffrax

Installation

From pypi:

pip install canterax

For local development:

git clone https://github.com/BenjiG03/canterax.git
cd canterax
pip install -e .

Quickstart

from canterax import Solution

gas = Solution("gri30.yaml")
gas.TPX = 1500.0, 101325.0, "CH4:1, O2:2, N2:7.52"

print(gas.cp_mass)
print(gas.cv_mass)
print(gas.enthalpy_mole)
print(gas.viscosity)
print(gas.thermal_conductivity)

Basis-aware aliases

gas = Solution("gri30.yaml")
gas.TPX = 1200.0, 101325.0, "H2:2, O2:1, N2:3.76"

gas.basis = "mass"
print(gas.h, gas.cp, gas.v)

gas.basis = "molar"
print(gas.h, gas.cp, gas.v)

State setters

gas = Solution("gri30.yaml")
gas.TPX = 1200.0, 2 * 101325.0, "CH4:1, O2:2, N2:7.52"

h, p, y = gas.HPY
gas.HPY = h, p, y

u, v, x = gas.UVX
gas.UVX = u, v, x

s, p = gas.SP
gas.SP = s, p

Equilibrium

gas = Solution("gri30.yaml")
gas.TPX = 2000.0, 101325.0, "CH4:1, O2:2, N2:7.52"
gas.equilibrate("TP")

gas = Solution("gri30.yaml")
gas.TPX = 1500.0, 101325.0, "CH4:1, O2:2, N2:7.52"
gas.equilibrate("HP")

Species-level thermodynamic arrays

The Solution object exposes the standard and partial molar properties commonly used from Cantera:

• standard_cp_R
• standard_enthalpies_RT
• standard_entropies_R
• standard_int_energies_RT
• standard_gibbs_RT
• partial_molar_cp
• partial_molar_enthalpies
• partial_molar_entropies
• partial_molar_int_energies
• chemical_potentials

Supported Solution surface

Implemented ideal-gas thermodynamic/state API:

• State setters/getters: TP, TPX, TPY, HP, HPX, HPY, UV, UVX, UVY, SP, SPX, SPY, SV, SVX, SVY, TD, TDX, TDY, DP, DPX, DPY
• Scalar properties: cp_mole, cp_mass, cv_mole, cv_mass, enthalpy_mole, enthalpy_mass, int_energy_mole, int_energy_mass, entropy_mole, entropy_mass, gibbs_mole, gibbs_mass, density_mole, density_mass, volume_mole, volume_mass, mean_molecular_weight, viscosity, thermal_conductivity
• Basis-aware aliases: h, u, s, g, cp, cv, v, density
• Metadata/helpers: thermo_model, phase_of_matter, reference_pressure, min_temp, max_temp, state_size, species_name, species_index, element_name, element_index, n_atoms, mass_fraction_dict, mole_fraction_dict

Not Currently Implemented:

• non-ideal activity-based thermo
• pure-fluid / saturation properties
• plasma/electron properties
• diffusion-coefficient transport APIs

Validation

The test suite validates the Canterax Solution object directly against Cantera:

• scalar thermodynamic properties
• species-level standard and partial molar arrays
• basis-aware aliases
• all implemented state setters
• TP and HP equilibrium
• mixture viscosity and thermal_conductivity

Run the suite with:

python -m pytest canterax/tests -q

License

MIT License.