Systematic effects of Transport-Parameter uncertainty on the laminar burning velocity of NH3/H2 mixtures

Transport parameters, the Lennard-Jones well depth ε/k, collision diameter σ and dipole moment μ, are an essential part of every combustion mechanism file, yet unlike the rate constants they are largely inherited from mid-century empirical correlations and have never been independently validated. Taking NH 3 /H 2 as the core system, this study quantifies on the Cantera platform how that uncertainty propagates to the laminar burning velocity (S_L) and to the extinction strain rate. An audit of mainstream mechanisms reveals two systematic biases: the ε/k of H, H 2 and OH is underestimated by 73–88 % relative to first-principles values, and the dipole moments of polar species such as OH, NH 2 and HNO are missing. The same entries recur, numerically identical, across mechanisms of independent kinetic origin, showing the bias to be inherited rather than incidental. Its consequences are not constant but evolve with the effective Lewis number, being amplified in hydrogen-rich flames (Le < 1) and negligible for pure ammonia. Isolating transport strictly within one mechanism, with kinetics, thermodynamics and numerics held identical, correction of the H/H 2 /OH parameters shifts S_L by up to + 6.8 % on the rich side; the shift is systematic and direction-consistent, but it lies within the experimental scatter and can be reproduced by an admissible change of a single rate constant, so it is reported as a quantified uncertainty rather than as improved accuracy. The extinction strain rate responds on the same order, rising from 26.4 to 4719 s⁻1 as the hydrogen fraction increases from 0 to 0.7. Extended along an n-alkane series computed within a single mechanism, the influence of the fuel’s own transport parameters decays monotonically with molecular weight, because the fuel is pyrolyzed before reaching the reaction zone; two independent diagnostics, the sign change of the pyrolysis lead distance and the 1 % sensitivity threshold, place the onset of this shielding in the C 4 –C 6 interval. A diffusion Damköhler number defined a priori orders the alkane series correctly out of sample, but fails a diluent-substitution test in which the sensitivity tracks flame temperature instead, bounding its scope to comparison within a fuel family.