B. Rotavera, E. L. Petersen
Fuel, Vol. 115
Publication year: 2014


Ignition delay times were measured behind reflected shock waves under highly dilute conditions (99% Ar) for 10 ternary blends of methyl octanoate, n-nonane, and methylcyclohexane (MCH), using a Design of Experiments approach, namely an L9 array. Measurements were obtained over a wide range of temperature (1258 < T (K) < 1630), pressure (1.5 < P (atm) < 9.5), and equivalence ratio (ϕ = 0.5, 1.0, 2.0). The series of experiments focused on the effects of variation in relative fuel volume fractions to assess respective influence of hydrocarbon class (i.e. methyl ester, n-alkane, cycloalkane) on ignition delay times of blended fuel. The L9 array utilized for the design of the experimental test matrix led to the development of an empirical ignition delay time correlation which was then employed for a broad series of calculations examining the effects of fuel-ratio variation. Results of the calculations show that addition of methyl octanoate to blends of n-nonane and MCH lowers the power-law pressure dependence of ignition delay times of such blends (the latter two species share similar dependence on pressure). The correlation also shows MCH as the most dominant of the three fuels in the blends studied with respect to controlling ignition delay time, a finding supported by ignition delay time-sensitivity calculations. Calculation of blending effects, defined herein as deviation from neat-fuel behavior, using the correlation indicated several key features of the studied blends. First, for fixed MCH concentration in a ternary blend, correlation calculations revealed an insensitivity of ignition delay time to variation in the ratio of methyl octanoate to n-nonane for all conditions. Second, blending effects on ignition delay times depend strongly on blend composition and on experimental conditions (pressure, equivalence ratio, temperature), indicating fuel- and condition-specific complexities, although were generally more pronounced at lower temperature. Third, of the three experimental conditions varied, ignition delay times of fuel blends exhibited the most sensitivity to equivalence ratio. To support experimental results, the study also involved compilation of a detailed chemical kinetics mechanism (4815 reactions, 1082 species) by integrating pertinent reaction chemistry from two independently validated mechanisms into a base model for n-nonane oxidation from a previous study by the authors. Model-based calculations indicate that under stoichiometric conditions near 10 atm, blending effects are negligible above 1350 K yet become increasingly significant towards lower temperature. Lastly, to assess the importance of blending effects, ignition delay times of fuel blends measured experimentally were compared against times calculated using a linear combination method (i.e. superposition of constituent-fuel ignition delay times). In general, superposition of neat-fuel trends cannot be utilized to reproduce ignition trends of the fuel blends herein, highlighting the importance of and continued need for blended-fuel studies.