Assessment of soot particle-size imaging with LII at Diesel engine conditions

Abstract

International audienceTwo-time-step laser-induced incandescence (LII) imaging was performed in Diesel engine-relevant combustion to investigate its applicability for spatially-resolved measurements of soot primary particle sizes. The method is based on evaluating gated LII signals acquired with two cameras consecutively after the laser pulse and using LII modeling to deduce the particle size from the ratio of local signals. Based on a theoretical analysis, optimized detection times and durations were chosen to minimize measurement uncertainties. Experiments were conducted in a high-temperature high-pressure constant-volume pre-combustion vessel under the Engine Combustion Network's (ECN) "Spray A" conditions at 61–68 bar with additional parametric variations of injection pressure, gas temperature, and composition. The LII measurements were supported by pyrometric imaging measurements of particle heat-up temperatures. The results were compared to particle-size and size-dispersion measurements from transmission electron microscopy (TEM) of soot thermophoretically sampled at multiple axial distances from the injector. The discrepancies between the two measurement techniques are discussed to analyze uncertainties and related error sources of the two diagnostics. It is found that in such environment where particles are small and pressure is high, LII signal decay times are such that LII with standard nanosecond laser and detector equipment suffers from a strong bias towards large particles

A Progress Review on Soot Experiments and Modeling in the Engine Combustion Network (ECN)

“…No assessment of primary particle size as a function of location within the spray flame can be made from these data because the entire cylinder volume was sampled; however, if we consider the samples extracted shortly after TDC as representative of the early formation stage and those extracted at the latest timing after TDC as representative of the oxidation stage, these results are consistent with the findings of Cenker et al [14] described above.…”

“…Soot particles formed during high-pressure spray combustion were collected on TEM grids by IFPEN and shipped to MU for TEM analysis [14]. During sampling, the probe was positioned slightly off the main jet flow axis such that only half of the jet impinged onto the probe, while the other half of the spray was minimally disturbed.…”

“…More information on this subject can be found in Manin et al [11]. With regard to primary particle size, TEM analysis published in Cenker et al [14] showed a count median diameter of 9.1 nm for soot sampled 60-mm from the injector orifice under the Spray A condition. While this was the only spatial location sampled for Spray A, TEM analysis was also performed on samples extracted at three axial locations, namely, 36 mm, 45 mm, and 60 mm from the 21% O 2 parametric variant of Spray A.…”

“…This is in spite of the TEM measurements presented above indicating that both d p and N p change as a function of axial position. To quantify the potential error introduced in these assumptions, we note that for the range of d p measured by Kook and Pickett [15], Aizawa et al [16], and Cenker et al [14] in combusting spray flames (7.3 nm-18.4 nm) the influence of changing d p on k e is approximately 4%. Further, for the apparently small primary particle diameters associated with these flames, k e changes by around 3% when varying N p from 5 to 150.…”

Sensitivity analysis for soot particle size imaging with laser-induced incandescence at high pressure

“…On the other hand, the camera gate should be long enough to collect sufficient signal to minimize the signal-to-noise ratio. In recent experiments [45] performed at the conditions mentioned in this present work, 15 ns was identified as the shortest gate that yields sufficiently strong signal at conditions comparable to Table 1, and therefore, this value is taken as the base gate width in this study. Also, with conventional ICCD cameras, significantly shorter gates are often strongly affected by the not well-defined rising and trailing edge.…”