Measurements from flame chemiluminescence tomography of forced laminar premixed propane flames

Abstract

The unsteady, visible, chemiluminescence fields of two non-axisymmetric, forced, laminar, premixed, propane–air flames are tomographically reconstructed using between 3 and 36 equally-spaced views. Algorithms for measuring flame surface area, flame curvature, flame thickness, and the normal component of the flame propagation velocity (surface speed) are demonstrated. The sensitivity of each measurement to the number of views used in the reconstruction is then assessed. For both flames studied, the difference between flame surface area fluctuations measured using 36 views and those measured using as few as 9 views was less than 1%. For the other three quantities, a measurement sample is acquired over the entire flame surface at one phase of the forcing cycle for each flame. The sensitivity to the number of views is compared by assessing the similarity of measurement distributions obtained using the maximum number of views to those obtained using fewer views. The surface speed measurement distribution is found to converge fastest as the number of views was increased, though results for mean curvature are similar. However, the flame thickness measurement distribution was found to have significantly slower convergence and more than twice the number of views are required to measure flame thickness compared to curvature or surface speed. The demonstrated measurement algorithms are generally applicable to the chemiluminescence fields of wrinkled, premixed flames. The results suggest that for laboratory-scale, weakly turbulent, premixed, jet flames, statistical measurements of flame curvature and surface speed may be accurately obtained using as few as six views, while greater than 20 views are likely to be required to obtain useful measurements of flame thickness.