NASA Glenn (pyglenn) Thermodynamic Import

This notebook shows how to bring ideal-gas thermodynamic data from the NASA Glenn (CEA) database into difflow using pyglenn, via difflow.pyglenn_import.

pyglenn stores NASA-9 polynomials for ~2030 species. The adapter samples pyglenn’s computed Cp(T) and least-squares fits difflow’s cubic Cp = a + bT + cT^2 + dT^3, and maps molecular weight and heat of formation. pyglenn carries no critical properties, so for a real-gas CubicThermo we pair its SpeciesData with CriticalProperties from elsewhere (last section).

Executable example. pyglenn is an optional dependency. So this notebook runs without it, the cell below falls back to a small mock of pyglenn’s documented API. Everything downstream — the difflow adapter, the Cp fit, IdealThermo/CubicThermo — is the real difflow code. With pyglenn installed, drop the calc= argument to use real data.

import jax
jax.config.update('jax_enable_x64', True)
import jax.numpy as jnp
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline

from difflow.pyglenn_import import import_species_data, list_available_species
from difflow.thermo import IdealThermo, CubicThermo
from difflow.eos import PengRobinson, CriticalProperties

WARNING:2026-07-14 07:07:36,873:jax._src.xla_bridge:905: Platform 'mps' is experimental and not all JAX functionality may be correctly supported!

1. Connect to pyglenn (or a mock)

The real API is pyglenn.ThermochemicalCalculator with get_available_species(name) and calculate_properties(species_id, T). The mock implements exactly that surface.

# Realistic ideal-gas Cp(T) cubics (J/mol/K) for the mock's species.
_MOCK = {
    'O2':  dict(id=1, phase='G', molecular_weight=31.998, heat_of_formation_298K=0.0,
                cp=(28.1, 5.0e-3, 1.0e-6, -3.0e-10)),
    'CO2': dict(id=2, phase='G', molecular_weight=44.009, heat_of_formation_298K=-393510.0,
                cp=(22.3, 5.98e-2, -3.5e-5, 7.5e-9)),
    'H2O': dict(id=3, phase='G', molecular_weight=18.015, heat_of_formation_298K=-241826.0,
                cp=(30.5, 9.6e-3, 1.2e-6, -1.1e-9)),
}

class MockGlennCalculator:
    """Minimal stand-in for pyglenn.ThermochemicalCalculator."""
    def connect(self):
        return True
    def close(self):
        pass
    def __enter__(self):
        return self
    def __exit__(self, *a):
        self.close()
    def get_available_species(self, query):
        q = query.lower()
        return [dict(name=n, **{k: v for k, v in d.items() if k != 'cp'})
                for n, d in _MOCK.items() if q in n.lower()]
    def calculate_properties(self, species_id, T):
        a, b, c, d = next(v['cp'] for v in _MOCK.values() if v['id'] == species_id)
        T = float(T)
        return {'cp': a + b*T + c*T**2 + d*T**3, 'h_relative': 0.0, 's': 0.0}

try:
    from pyglenn import ThermochemicalCalculator
    CALC = None          # adapter opens its own real calculator
    SOURCE = 'real pyglenn'
except ImportError:
    CALC = MockGlennCalculator()
    SOURCE = 'a mock of the pyglenn API (pip install pyglenn for real data)'
print('Data source:', SOURCE)

Data source: a mock of the pyglenn API (pip install pyglenn for real data)

2. Import ideal-gas SpeciesData

species = ['O2', 'CO2', 'H2O']
data = import_species_data(species, calc=CALC)
for name, sd in data.items():
    print(f'{name:4s}  MW={sd.MW:7.3f} g/mol   Hf={sd.Hf/1000:9.2f} kJ/mol   '
          f'Cp_coeffs={tuple(round(c, 6) for c in sd.Cp_coeffs)}')

O2    MW= 31.998 g/mol   Hf=     0.00 kJ/mol   Cp_coeffs=(28.1, 0.005, 1e-06, -0.0)
CO2   MW= 44.009 g/mol   Hf=  -393.51 kJ/mol   Cp_coeffs=(22.3, 0.0598, -3.5e-05, 0.0)
H2O   MW= 18.015 g/mol   Hf=  -241.83 kJ/mol   Cp_coeffs=(30.5, 0.0096, 1e-06, -0.0)

3. Build an IdealThermo and check the Cp fit

The cubic fit should track the source Cp closely across the window.

thermo = IdealThermo(data)
T = np.linspace(300, 1000, 60)
fig, ax = plt.subplots(figsize=(6, 4))
for name in species:
    ax.plot(T, [float(thermo.Cp(name, t)) for t in T], label=name)
ax.set_xlabel('Temperature (K)'); ax.set_ylabel('Ideal-gas $C_p$ (J/mol/K)')
ax.set_title('difflow $C_p$ fit from NASA Glenn data'); ax.legend()
plt.tight_layout(); plt.show()

4. Real-gas enthalpy: pair with critical properties

pyglenn has no Tc/Pc/omega, so we supply CriticalProperties (here inline; in practice from difflow.database or the Cantera import) and wrap everything in a CubicThermo — ideal-gas enthalpy from the pyglenn-sourced Cp, plus the Peng-Robinson departure.

crit = {
    'O2':  CriticalProperties(name='O2',  Tc=154.6, Pc=5.04e6,  omega=0.022, MW=31.998),
    'CO2': CriticalProperties(name='CO2', Tc=304.2, Pc=7.38e6,  omega=0.224, MW=44.009),
    'H2O': CriticalProperties(name='H2O', Tc=647.1, Pc=22.06e6, omega=0.345, MW=18.015),
}
cubic = CubicThermo(IdealThermo(data), PengRobinson(crit))
flows = {'O2': 1.0, 'CO2': 1.0, 'H2O': 1.0}
Tk, P = jnp.array(400.0), jnp.array(5e6)
H_ideal = cubic.stream_enthalpy(flows, Tk, P=None)
H_real = cubic.stream_enthalpy(flows, Tk, phase='vapor', P=P)
print(f'Ideal-gas enthalpy : {float(H_ideal):12.1f} W')
print(f'Real-gas (PR dep.) : {float(H_real):12.1f} W')
print(f'Departure          : {float(H_real - H_ideal):12.1f} W')

Ideal-gas enthalpy :      10501.9 W
Real-gas (PR dep.) :       6174.7 W
Departure          :      -4327.2 W

Summary

• difflow.pyglenn_import.import_species_data turns NASA Glenn ideal-gas data into difflow SpeciesData (cubic Cp fit + MW + Hf).

• pyglenn provides no critical properties; pair with CriticalProperties for an EOS / CubicThermo.

• Everything downstream is JAX-differentiable difflow code.