Effect of m-xylene on soot formation in high pressure diffusion flames.

摘要

A set of benchmark experimental data has been acquired to study the effect of m-xylene on soot formation in an ethylene flame diluted with nitrogen. The data will be used for comparisons with simulations performed using well-established numerical codes. This investigation is part of a bigger program, aimed at developing a complete chemical mechanism to describe soot formation from a JP-8 surrogate mixture consisting of n-dodecane and m-xylene. The soot volume fractions and qualitative fluorescence profiles of polycyclic aromatic hydrocarbons (PAH) in three sets of flames are measured using LII and broadband PLIF with excitation in the UV, at pressures between 1 and 5 atm. Quantitative soot volume fractions are obtained by calibrating the LII signal with radial laser extinction measurements at 3 atm.;Three sets of flames are established with different concentration levels of m-xylene. The first flame is established using ethylene as the fuel, diluted with a large quantity of nitrogen intended to lower soot loading. The lower soot concentrations resulted in reduced optical densities for the flames enabling easier optical access for laser diagnostics. The second set of measurements is made on a similar flame (Flame 2), but with 2.5% of the carbon flow rate in the fuel stream being provided by m-xylene. In the third and final flame reported (Flame 3), the concentration level of m-xylene was doubled such that 5% of the carbon flowing to the flame being contributed by the m-xylene.;The flame heights are unchanged with increasing pressure, which is in agreement with laminar diffusion flame theory and previous studies. However, the flame heights are seen to increase with increasing levels of m-xylene and the distances from the fuel tube to the locations of maximum soot volume fraction are seen to vary inversely with pressure.;Flame 1 had a peak soot volume fraction of approximately 1.9 ppm at 5 atm. This doubled to 4 ppm when 2.5% of the fuel carbon atoms were substituted with m-xylene, and 4.4 ppm when the contribution was 5%.;The addition of m-xylene resulted in earlier detection of soot concentrations, possibly indicating an earlier onset of soot. This is believed to be due to the introduction of an aromatic compound ( m-xylene) which provides a faster route to the initial soot particle. The peak soot volume fractions in the flame varied with the pressure following a relationship of the form fnmax∝Pn , where n varied between 3 and 1.88, decreasing with increasing values of pressure. These n-values are higher than previously reported values in range of 1.0 and 1.2.;The PLIF measurements were made in the 320--380 nm and 420--480 nm wavelength bands and were not corrected for quenching effects, which become important at higher pressures. The LIF signals in the shorter wavelength band are representative of the small PAH (1--2 rings) and the LIF signal in the longer wavelength band (more than 2 rings), is representative of the large PAH. Comparisons are drawn between the concentrations of the small and large PAH, assuming that the quenching rates are similar in the two wavelength regions.;The relative locations of the peak PAH and soot concentrations are consistent with previously reported measurements on diffusion flames. In both radial and axial directions, the small PAH concentrations peak before the large PAH, which in turn peak ahead of the soot volume fractions.;The addition of m-xylene, which fluoresces strongly in the shorter wavelength band. results in a bright region of fluorescence near the tip of the fuel tube, highlighting the dissipation of m-xylene in the flame. The presence of m-xylene in the flame increases the concentration of both small and large PAH in the flame. Doubling the concentration of m-xylene in the fuel stream is seen to double to concentration of small PAH in the flame, but results in a comparatively smaller increase in the large PAH concentrations.