Chirped-Probe-Pulse Femtosecond Coherent Anti-Stokes Raman Scattering For Gas-Phase Temperature Measurements In High-Pressure Kerosene/Air Flames
S. Legros, B. Barviau, F. Grisch
Normandie Univ., UNIROUEN, INSA Rouen, CNRS, CORIA, 76000 Rouen, France
Air traffic is increasing and its related emission is a major concern for the stability of the planet's climate. Innovative combustion systems need to be developed to continue to enable mobility while respecting the environment. The qualification of new propulsive systems requires the analysis of the associated combustion and one of the quantities of major interest is temperature. The present work exposes the development of the chirped-probe-pulse femtosecond coherent anti-Stokes Raman scattering (CPP-fs-CARS) thermometry at 1 kHz and its application in a semi-industrial test bench. Temperature measurements in a kerosene/air flame at a pressure of 0.75 MPa were carried out. CARS is a third order nonlinear optical diagnostic renown for high accuracy temperature determination. CPP-fs-CARS rely on the analyzis of the frequency-spread dephasing rate after the initial excitation of the Raman coherence on N2. This pump-probe method needs three input pulses. The two first excitation pulses are Fourier transform limited and present a temporal width of 100 fs at 800 nm and 675 nm. The frequency difference matches 𝑁2 molecular vibrational energy gap. Then the evolution of the coherence generated in the medium is probed with a delayed picosecond which encompass a sufficient part of the coherence evolution. In the CPP configuration, this longer probe pulse results from a 100 fs, 675 nm temporally stretched through a propagation in 30 cm glass rod. Following the energy conservation principle, the interaction of the probe with the excited medium results in the generation of a CARS signal, presenting a spectral shape sensitive to the temperature. Nevertheless, the spectral shape is also dependant on the incident pulses spectro-temporal features implying that spectral phases of the three input pulses need to be precisely evaluated. A polynomial fourth order function is then used to describe each pulse phase shape. Moreover, the molecular bath encompassing the nitrogen molecule contribute in a certain extend to the CARS spectrum. To deal with such a number of adjustable parameters, a genetic algorithm is used to tune those parameters. Finally, temperature is deduced from the single-shot CARS spectra. First results will report the application of the diagnostic in atmospheric pressure well known environments and the associated data processing. Such environments lead to found a temperature accuracy determined in a near-adiabatic laminar flame better than 1.5% at 2250 K. Those results demonstrate the applicability and usefulness of CPP-fs CARS and are pursued with the investigation of the high-pressure two-phase flames. The whole CPP-fs-CARS setup was then moved close to the representative aircraft combustor engine facility and adapted for pressurized environment. Measurements where performed at several locations in the plane orthogonal to the flame propagation and at several distances from the injection system. The data processing of the measurements in such harsh conditions was then adapted to disregard spectra resulting from the presence of liquid droplets. Finally, temperature evolution was extracted from single shot measurements and probability density function (PDF) are reported at several location of interest enabling to retrieve not only the mean temperature, but also important information on flame behaviour at different stages of its evolution within the combustion chamber length.