Cyclic ethers are intermediates formed from unimolecular reaction of carbon-centered hydroperoxy-substituted radicals (Q̇OOH), which are central to low-temperature chain-branching. Because cyclic ethers are isomer-specific proxies for Q̇OOH, detailed prescription of chemical reactions describing the subsequent consumption mechanisms is required for accurate combustion modeling. However, the most common approach in the development of chemical kinetics mechanisms is to use a simplified set of elementary steps that neglect the formation and subsequent reaction of cyclic ether radicals, which creates a source of mechanism truncation error. As a consequence, quantitative discrepancies between model predictions and experimental species profiles of cyclic ethers are ubiquitous and span a range of hydrocarbons such as n-butane, n-pentane, n-hexane, cyclohexane, hexene isomers, and hexanal. Moreover, uncertainties in species profile predictions of cyclic ethers translate directly to uncertainty in ignition predictions.
For the explicit purpose of determining the extent to which cyclic ether consumption mechanisms affect combustion modeling, the present work examines the chemical kinetics underpinning such discrepancies using, as a representative case, a subset of cyclic ethers produced from n-butane oxidation. Detailed sub-mechanisms are developed using Reaction Mechanism Generator (RMG) for ethyloxirane and 2,3-dimethyloxirane, which form from unimolecular decomposition of b-Q̇OOH radicals during n-butane combustion. The sub-mechanisms prescribe consumption reactions for both cyclic ethers, including ȮH-initiated H-abstraction, O2-addition, ROȮ isomerization, among other reactions, and were integrated with the NUIGMech1.1 mechanism to examine model predictions of species profiles and ignition delay times.
Inclusion of the sub-mechanisms led to closer consistency between model predictions and experimental species profiles and also affected ignition predictions. Sensitivity analysis shows that rates of ȮH-initiated H-abstraction are critical for determining temperature dependence of species profiles of cyclic ethers, which may serve as an indicator for the importance of branching fractions in the initiation step ȮH + n-butane → H2O + 1-butyl/2-butyl. Moreover, flux towards ketohydroperoxide formation from n-butane increased upon addition of the sub-mechanisms, as determined by ignition delay time simulations and rate-of-production analyses conducted on ȮH.
The results herein demonstrate that detailed sub-mechanisms and accurate H-abstraction rates from cyclic ethers are necessary for high-fidelity predictions of chemical kinetics for combustion modeling. Continued refinement of detailed reaction mechanisms is required in order to produce accurate models for combustion that serve as a starting point for mechanism reduction techniques applied either to detailed mechanisms or to sub-mechanisms for consequent integration. Such techniques are required to enable the modeling of reactive flows that incorporate computational fluid dynamics at practical conditions.