Abstract
The influence of beam steering and signal trapping on the accuracy of soot volume fractions measured using planar laser-induced incandescence (LII) has been investigated in turbulent non-premixed sooting flames at atmospheric pressure. In turbulent non-premixed C2H4/air flames, the influence of local de-focusing/focusing of the laser sheet from beam steering can result in the underestimate of the averaged LII signal by 30 %, even when operating within the so-called plateau regime of laser fluence. Beam steering was also found to be significant in both the upstream region of C2H4/air flames and non-reacting C2H4 flows, because the fuel has a relatively high refractive index compared with ambient air. The extent of beam steering at different heights of reacting and isothermal flows as well as its dependence on exit Reynolds number (Re) has been measured. The measurements reveal that even at low turbulence levels (2000 < Re < 3000), beam steering effects can be significant. Also found is that the LII signal at a 450 nm wavelength can be attenuated by a few per cent at high soot loading regions in turbulent flames due to signal trapping. Finally, the feasibility of directly evaluating the signal attenuation via planar LII results was assessed by comparing the virtual soot attenuation calculated based on the planar LII result with that measured using light-of-sight extinction.