Soot structure and flow characteristics in turbulent non-premixed methane flames stabilised on a bluff-body

The soot properties of methane in turbulent regimes are not well characterised but are highly desirable. Methane is the main constituent of natural gas that is broadly used in many industrial combustors. Investigation of turbulent methane flames under well-defined boundary conditions is therefore useful for interpreting soot formation in practical burners and can be used for further model development. This study presents a joint experimental and numerical study of a series of turbulent non-premixed bluff-body flames fuelled with pure methane for three values of the momentum flux ratio of fuel jet to co-flowing air. Soot volume fraction (SVF) and flowfield are measured simultaneously using planar laser-induced incandescence (P-LII) and 2D-polarised particle image velocimetry (P-PIV). Additionally, time-averaged temperature, mixture fraction, OH and C2H2 concentrations are estimated numerically using RANS models. The global flame structure for all three flames features a recirculation zone with a double-vortex structure, a jet-propagating zone, and a neck zone connecting the two regions. The soot distribution within the recirculation zone shows clear distinct features, which is attributed to the mean mixture fraction distribution in this zone. Increasing the momentum flux ratio shifts the location of the mean stoichiometric mixture fraction to the rich inner vortex core, leading to a distinct peak of the total integrated soot in the inner vortex of the recirculation zone that is not observed in other cases. Also, it is deduced that the soot inception starts earlier in the recirculation zone for the flame with the highest momentum flux ratio and in the jet zone for the other two flames. Much higher soot concentration and lower intermittency are found with ethylenebased flames stabilised on the same burner and with the same operating conditions. In addition, the study has generated a database of soot and flowfield results, which can be helpful for future model validations.