Development of fluorescence-based techniques for quantitative measurements of combustion species

The work presented in this thesis covers how laser-induced fluorescence (LIF) and photofragmentation laser-induced fluorescence (PF-LIF) can be used to determine quantitative species concentrations in different combustion environments. To attain species concentrations with LIF it is of vital importance to investigate the influence of collisional quenching on the fluorescence signal strength, which can be done by measuring the fluorescence lifetime. A method for simultaneous measurements of fluorescence lifetimes of two species present along a line is described and discussed. The experimental setup is based on picosecond laser pulses from a dual optical parametric generator/amplifier (OPG/OPA) system tuned to excite two different species, whose fluorescence signals are detected with a streak camera. The concept is demonstrated for fluorescence lifetime measurements of CO and OH in laminar methane/air flames on a Bunsen-type burner. The measured one-dimensional lifetime profiles generally agree well with lifetimes calculated from quenching cross sections found in the literature and quencher concentrations predicted by the GRI 3.0 chemical mechanism. For OH there is a systematic deviation of ~30% between calculated and experimental lifetimes in the product zone. It is found that this discrepancy is mainly due to the adiabatic assumption regarding the flame and uncertainty in H2O quenching cross section. The second technique, i.e. PF-LIF, is used to study H2O2 and HO2, which both are molecules lacking accessible bound excited states. Here, a pump laser pulse of 266-nm wavelength dissociates the molecules into OH fragments, which after a short time delay (nanosecond time scale), are probed with LIF using a second laser pulse tuned to an OH absorption line. PF-LIF is for the first time used for two-dimensional imaging of HO2 in laminar flames and H2O2 in an homogenous charged compression (HCCI) engine. In methane/air flames on a Bunsen-type burner, relative species concentrations of HO2, H2O2 and CH3O2 are achieved via comparison of experimental signal profiles with simulated concentrations predicted by the Konnov detailed C/H/N/O reaction mechanism. An interfering OH signal contribution is observed in the product zone and found to originate from hot CO2. In the HCCI experiments, quantitative H2O2 concentrations at different piston positions are attained via an on-line calibration procedure. In terms of mass fraction levels, the crank-angle resolved experimental data agree well with profiles resulting from simulations using the software Digital Analysis of Reaction System (DARS), while shapes and profiles deviate slightly, which mainly is due to signal interference from HO2