There is a growing need for the clean generation of electricity in the world, and increased efficiency is one way to achieve cleaner generation. Increased efficiency may be achieved through an improved understanding of the heat flux of participating media in combustion environments. Real-time in-situ optical measurements of gas temperature and concentrations in combustion environments is needed. Optical methods do not disturb the flow characteristics and are not subject to the temperature limitation of current methods. Simpler, less-costly optical measurements than current methods would increase the ability to apply them in more circumstances. This work explores the ability to simultaneously measure gas temperature and H2O concentration via integrated spectral intensity ratios in regions where H2O is the dominant participating gas. This work considered combustion flows with and without fuel and soot particles, and is an extension of work previously performed by Ellis et al. [1]. Five different combustion regimes were used to investigate the robustness of the infrared intensity integral method first presented by Ellis et al. [1]. These included Post-Flame Natural Gas (PFNG), Post-Flame Medium Wood (PFMW), Post-Flame Fine Wood (PFFW), In-Flame Natural Gas (IFNG), and In-Flame Fine Wood (IFFW). Optical spectra were collected as a function of path length for each regime. Methods for processing the spectra to obtain gas temperature, gas concentration, broadband temperature, and broadband emissivity were developed. A one-dimensional spectral intensity model that allowed for specular reflection, and investigated differences between measured and modeled spectral intensities was created. It was concluded that excellent agreement (within 2.5%) was achieved between optical and suction pyrometer gas temperatures as long as 1) the optical probe and cold target used were well-aligned 2) the path length was greater than 0.3 m and 3) the intensity from broadband emitters within the path was smaller than the gas intensity. Shorter path lengths between 0.15 – 0.3 m produced reasonable temperatures with 7 % error while path lengths of 0.05 m or less were as much as 15% in error or the signal would not effectively process. Water vapor concentration was less accurate being at best within 20% (relative) of expected values. The accurate determination of concentration requires first an accurate temperature concentration as well low broadband participation. Some optical concentrations were in error as much as 85%. The 1-D model was compared to the measurement and it was found that the model peaks were sharper and shifted 0.167 cm-1 compared to the measured data. The reason for the shift can be attributed to the uncertainty of the reference laser frequency used in the FTIR. No conclusion was found for the cause of the sharper peaks in the model. The integrated area of bands used to find temperature and concentration matched well between the model and measured spectrum being typically within 3%.



College and Department

Ira A. Fulton College of Engineering and Technology; Mechanical Engineering



Date Submitted


Document Type





radiation, spectral, water vapor, temperature, concentration, emission, intensity