A study of temperature measurement using coherent anti-Stokes Raman spectroscopy.

The aim of this work is to increase the applicability of Coherent Anti-Stokes Raman Spectroscopy (CARS) to temperature measurement in practical devices. Particular emphasis is placed on combustion thermometry and high pressure steam systems are also considered. A study is made of the temperature measurement accuracy attainable in the range 290 to 1050 K, using broadband CARS. Accuracies of 1 - 2% are attained, and laser cross coherence effects are found to be important. The determination of temperature probability density functions is of great importance to combustion science. Their measurement using single shot CARS requires the analysis of very large numbers of spectra. A study is made of fast methods of data analysis and the temperature measurement precision attainable using them. A very rapid data analysis method suitable for use in fluctuating temperature, pressure and concentration environments is developed. The temperature precision attainable using CARS is limited by CARS signal noise. For systems with high temperature fluctuations, detector counting statistics are found to make a dominant contribution to this. The spread in measured temperature probability density function width due to signal noise is characterised for the CARS system used, as a function of CARS signal strength for the temperature range 290 to 1050 K. A fast CARS signal analysis method is applied to map temperatures and temperature fluctuations in the flame zone of a turbulent oil spray furnace. The temperature measurements are compared with Discrete Droplet and Continuous Droplet oil spray model predictions (Stopford, 1984) with good agreement, particularly in the former case. In the post flame region, where turbulent fluctuations are less severe, averaged measurements of H[2]O concentrations were made