Study of surrogates for conventional and synthetic aviation jet fuels

Deepti Singh, Purdue University

Abstract

An experimental apparatus to measure laminar flame speeds and Markstein lengths of a fuel/air mixture using a spherical flame configuration, was developed to study both gaseous and liquid fuels at high temperature. Extensive validation of the apparatus was performed through laminar flame speed measurements of H2/CO/Air (Syngas/air) flames at room temperature and pressure, at which reliable data are available in literature. For Syngas, range of available measurements at several conditions was increased. The apparatus was tested for liquid fuel measurements with n-Decane and then measurements were provided for other pure and complex fuels. Compositions of Syngas with H2 content varying from 5% to 75%, with the remaining CO, were studied. Excellent agreement was found between present measurement and data available in literature. Further, measurements of laminar flame speed were obtained for two fuel compositions (5%H2 -95%CO and 5%H2-95%CO) with preheat temperatures of 400 K and 500 K, for a wide range of equivalence ratios. Measurements were compared with simulations using the GRI Mech 3.0, H2/CO Davis mechanism and the San Diego mechanism. The H2/CO Davis and San Diego mechanisms best represent experimental data over most conditions. However, at preheat temperatures of 500K, large discrepancies were seen between the measured and simulated flame speeds for the high Hydrogen content fuel. Additionally, the effect of water addition to these two fuel-air mixtures was studied at 400 K. The two mixtures showed different responses to water addition. Sensitivity analysis and species profiles were used to explain these differences. For liquid fuels, two methods of fuel injection were explored, one based on pre-vaporizing the fuel outside the chamber and the other based on direct fuel injection into the chamber. Some differences were observed with measurements obtained using the two methods for complex fuels but not pure single component fuels. Non-linear extrapolation techniques were applied in the extraction of laminar flame speed from stretch flame speed data, due to high stretch rates and Lewis numbers of the mixtures. Comparison with linear extrapolation methodology, showed an over prediction to the extent of 3-5% using the latter. n-Decane and Methylcyclohexane, that are popular fuels in several surrogate mixtures for complex gasoline and jet fuels, were studied at 400 K and 1 atm. Comparisons were made with simulations performed using the CHEMKIN PREMIX code, with several recently developed chemical kinetic schemes. Significant differences were seen in the simulation results using the various schemes. Laminar flame speeds of Jet-A were measured in the spherical flame apparatus at 400 K and 1 atm in order to examine the validity of surrogate mixtures to simulate flame speeds. Several surrogate mixture models proposed in literature for Jet-A were simulated using the existing kinetic schemes, some of which are validated using n-Decane and MCH experimental data. These models contain two to six pure components. It was seen that the three component mixture containing Benzene, Hexane, n-Decane was able to best predict measurements of Jet-A. Additionally, the six component surrogate did not provide any advantage in simulating the combustion properties. Similarly, laminar flame speed measurements of synthetic jet fuel S-8 at similar conditions were compared with the only available surrogate model in literature, containing n-Decane and iso-Octane. Markstein lengths of the fuels were also measured at the conditions stated above. The fuels showed a decreasing trend with increasing equivalence ratio. Similarity in the flame speed measurements and Markstein lengths of n-Decane, Jet-A and S-8, indicates that pure n-Decane can be used as a single component surrogate to simulate these properties of the complex fuels. To extend the range of measurements for n-Decane fuels, in order to provide additional information on transport properties, N2 was replaced with Helium. Flame speed measurements were made up to an equivalence ratio of 1.6. This resulted in an increase in the flame speed as well as Markstein lengths at a particular equivalence ratio. Flame speed measurements made in this apparatus were consistently lower than measurements obtained using counter-flow flame configurations, for the same experimental conditions. Differences between present measurements and past measurements made using a counter-flow flame configuration could be attributed to radiation heat losses that need further investigation and quantification. The differences were primarily seen in the heavier liquid fuels but not in the Syngas/air mixtures.

Degree

M.S.E.

Advisors

Qiao, Purdue University.

Subject Area

Aerospace engineering|Energy

COinS