Laminar burning velocity of NH3 + N2 + O2 and NH3 + Ar + O2 flames at ultra-lean and rich conditions

•Ultra-lean and rich NH3SL was measured with different xO2 and diluent types.•11 mechanisms were used but none could reproduce all the present experimental data.•The mixed mechanism behavior indicates the necessity of broad-range SL measurement.•Sensitivity and uncertainty analyses give suggestions on future mechanism updating. Ammonia (NH3) has received increasing attention as a carbon-free alternative fuel. The laminar burning velocities (SL) of NH3-containing flames have been extensively investigated, resulting in the development, updating, or optimization of many mechanisms. In this study, we aimed to test the validity of the available mechanisms by measuring the SL of ammonia flames at ultra-lean (ϕ = 0.3–0.6) and rich (ϕ = 1.5–1.8) conditions, which were not covered by literature experiments and thus outside the validation targets of the literature mechanisms. The measurements were conducted using the heat flux method at 1 atm and 298 K, with careful evaluation of uncertainties in laminar burning velocities, oxygen ratios, and equivalence ratios. In addition to NH3 + O2 + N2, we also tested NH3 + O2 + Ar flames to provide data at broader conditions. Simulations were carried out using eleven kinetic mechanisms that have been validated against the NH3 laminar burning velocities when first published, and none of them reproduces well the present experimental data with mixed behavior observed, indicating the necessity of SL measurements with different oxygen ratios and diluent types, as well as that there is still space for mechanism updating. A-factor reaction sensitivity analyses were also conducted using the eleven mechanisms, providing uncertainty ranges of rate constants based on their variations in the different mechanisms. Our analyses suggest that future mechanism updating should focus on especially on the temperature dependences of the sensitive reactions with large rate constant uncertainties. In addition, mechanism updating targeting fuel-rich side data will be more challenging than the lean side due to a larger number of large-uncertainty reactions.