Francis M. Haas

Development of Reduced Kinetic Models for Petroleum-Derived and Alternative Jet Fuels

The surrogate fuel concept to replicate the detailed gas phase combustion behaviors of conventional and alternative jet aviation fuels in numerical combustion models is extended and tested in specific examples of synthetic jet fuels derived from coal and natural gas, and also to the pressure and equivalence ratio dependences of the combustion responses of conventional Jet–A fuel. The formulation of surrogate fuels for Syntroleum S-8, Shell SPK and Sasol IPK, is described. Assuming these compositions, a detailed chemical kinetic model construction previously elaborated upon is extended and tested against reference data sets of shock tube ignition delay and laminar burning velocity. Calculations with the detailed kinetic model, containing 3147 species correctly represent the experimentally measured reactivity of the target fuels for shock tube ignition delay. The model also captures trends in the ignition delay for a reference Jet-A as a function of pressure and equivalence ratio. The earlier reported detailed model is expanded to encompass a range of n-alkane carbon numbers up to C16 and iso-cetane. The expanded model is validated against available shock tube ignition delay in detailed form and against laminar burning velocity datasets using a series of numerically reduced models of decreasing dimension for n-hexadecane, iso-cetane, and their mixtures. Though the detailed model reproduces the general kinetic behavior for the ignition delays of each jet fuel, the predicted values are generally longer than experimental results. A series of reduced models of the order of 100 species in size, are produced for simulation of flame environments. Calculations for laminar premixed flames for each jet fuel are similar with burning velocities for IPK flames marginally lower than those for the conventional Jet-A which in turn are marginally lower than those for S-8. The requirement for severely reduced, but high fidelity chemical kinetic numerical schemes that retain predictive capacities for the combustion behaviors of real liquid transportation fuels is addressed through the introduction of a strategy to produce “compact” models of the order of 35 species. The strategy utilizes calculations of the detailed model construct as a fundamental and scientific standard, to which engineering approximations achieved through adjusting reaction rates and omitting or diverting the fate of select reaction pathways at high carbon numbers are applied. The strategy is tested for the exemplar real fuel test case of the S-8 ignition delay and laminar

“Virtual” Smoke Point Determination of Alternative Aviation Kerosenes by Threshold Sooting Index TSI) Methods

One benefit attributable to blending of alternative jet fuels into conventional petroderived kerosenes is reduced sooting tendency relative to the conventional unblended kerosenes. This benefit is largely due to the lower aromatics content of the alternative fuels/blends, and it is desirable with respect to sooting tendency limits embedded in aviation turbine fuel specifications (e.g., ASTM D1655 and D7566). However, the relatively high smoke points of many alternative jet fuels are not directly measurable by the prevailing ASTM D1322 smoke point (SP) method and its international variants. This frustrates characterization of these alternative fuels and prediction of their blending properties. The present work addresses an extrapolative “virtual” smoke point (VSP) technique for determination of smoke points compatible with the ASTM D1322 standard. Importantly, this compatibility removes the significant ambiguity historically associated with measurements of non-standard smoke points, herein categorically designated SP*. The VSP approach invokes the linear-by-mole blending functional basis of the Threshold Sooting Index (TSI), which has been empirically demonstrated in the literature for both defined molecular species and complex hydrocarbon fluids. If the average molecular weights (MWs) of the blending component fuels are known, then the VSPs of low-sooting tendency fuels can be forecast using D1322 SP measurements. This is demonstrated here for iso-octane and ndodecane as illustrative pure-component test fuels, as well as the full boiling range POSF 7720, a camelina sativa-derived hydrotreated jet fuel (HRJ). In part, the approach is facilitated by the ability to easily determine the average molecular weight of a fuel using a method recently developed by this laboratory.

Rate Coefficient Determinations for H + NO2 → OH + NO from High Pressure Flow Reactor Measurements

Rate coefficients for the reaction H + NO2 → OH + NO (R1) have been determined over the nominal temperature and pressure ranges of 737-882 K and 10-20 atm, respectively, from measurements in two different flow reactor facilities: one laminar and one turbulent. Considering the existing database of experimental k1 measurements, the present conditions add measurements of k1 at previously unconsidered temperatures between ∼820-880 K, as well as at pressures that exceed existing measurements by over an order of magnitude. Experimental measurements of NOx-perturbed H2 oxidation have been interpreted by a quasi-steady state NOx plateau (QSSP) method. At the QSSP conditions considered here, overall reactivity is sensitive only to the rates of R1 and H + O2 + M → HO2 + M (R2.M). Consequently, the ratio of k1 to k2.M may be extracted as a simple algebraic function of measured NO2, O2, and total gas concentrations with only minimal complication (within measurement uncertainty) due to treatment of overall gas composition M that differs slightly from pure bath gas B. Absolute values of k1 have been determined with reference to the relatively well-known, pressure-dependent rate coefficients of R2.B for B = Ar and N2. Rate coefficients for the title reaction determined from present experimental interpretation of both laminar and turbulent flow reactor results appear to be in very good agreement around a representative value of 1.05 × 10(14) cm(3) mol(-1) s(-1) (1.74 × 10(-10) cm(3) molecule(-1) s(-1)). Further, the results of this study agree both with existing low pressure flash photolysis k1 determinations of Ko and Fontijn (J. Phys. Chem. 95 3984) near 760 K as well as a present fit to the theoretical expression of Su et al. (J. Phys. Chem. A 106 8261). These results indicate that, over the temperature range considered in this study and up to at least 20 atm, net chemistry due to stabilization of the H-NO2 reaction intermediate to form isomers of HNO2 may proceed at negligible rates compared to R1.