Portal de Periódicos da CAPES

Quantitative planar laser-induced fluorescence of naphthalenes as fuel tracers

2005; Elsevier BV; Volume: 30; Issue: 1 Linguagem: Inglês

10.1016/j.proci.2004.08.263

1873-2704

Sebastian A. Kaiser, Marshall B. Long,

Laser-induced spectroscopy and plasma

In planar laser-induced fluorescence (PLIF), naphthalenes are commonly used as tracers for diesel-like fuels and fuel surrogates. Previously, such a technique was used to visualize the air–fuel mixing in a mesoscale (∼10 cm3 device volume) burner prototype running on dodecane as a single-component JP-8 surrogate. For quantification of this technique, the influence of temperature, oxygen quenching, and the presence of the fuel surrogate was investigated through spectrally resolved LIF of naphthalenes in a heated jet as well as through broadband imaging. For excitation at 266 nm, naphthalene and 1-methylnaphthalene show very similar fluorescence behavior. Fluorescence is not influenced by the presence of dodecane but is strongly quenched by oxygen. Due to this strong oxygen quenching, fluorescence intensity was found to be linear with equivalence ratio at all temperatures. Both increased temperature and increased oxygen concentrations lead to a spectral broadening and red-shift. This effect is minor for oxygen addition and saturates at about 10% oxygen, but is significant for temperature increase and continuous throughout the temperature range investigated. This red-shift was exploited to simultaneously image temperature and equivalence ratio. A dichroic beamsplitter was used to obtain a long-wavelength and a short-wavelength image whose ratio represents temperature. Calibrations were performed on the heated jet, and subsequently this two-color PLIF method was applied to the mesoscale burner. After averaging the images from 20 processed single-shot measurements, the equivalence ratio and temperature images obtained had signal/noise ratios of 25 and 11, respectively. Temperature correction of the equivalence ratio resulted in excessive noise. Likely inaccuracies in temperature were identified in very lean regions where the signal is low and sensitive to the background corrections employed.