An optical crosscorrelation technique for turbulent density measurements in flames

Abstract

A crossed-beam schlieren (cbs) system has been constructed and used to study the structure of an open, premixed propane-air flame that was pilot-stabilized on a two-dimensional burner. The cbs is composed of two perpendicular light beams that intersect at a point of interest in the flame. Each beam is continuously refracted along its path, thus providing an integrated measure of the disturbances. However, the integrated measures from both beams can be crosscorrelated to provide information representative only of the intersection locale. The system is theoretically capable of measuring the cbs crosscorrelation, Qs ; the components of s the three-dimensional "energy" spectrum; propagation velocity and, if the flow is isotropic, root mean square density fluctuation, and the two point fluctuating density covariance The envelope of test conditions was as follows: approach flow mean velocity (U) of 20-30 fps; approach flow relative turbulent intensities (u'/U) of 3, .6, or 10%; and equivalence ratios ([phi]) between 0.6 and 1.2. Flames with low speed (20 fps), low turbulence (3%) feeds were found to be composed of continuous fronts that oscillate symmetrically about the burner centerline at frequencies on the order of 20 hz. The influence of the oscillations could be reduced by running at higher turbulence levels and nearer stoichiometric ratios. Near test conditions of U=30 fps, u/U=6%, and [phi]=0.83, one effect of turbulence was to wrinkle but not totally disrupt the flame front. However, in the upper portions of the flame brush, discrete eddy combustion was identified on both Schlieren photographs and single beam oscilloscope traces. The combustion intensity as measured by Q , and spatial extent of reaction were measured for various test flames and demonstrated established trends. Propagation velocities of the flame front and combusting eddies were always found equal to the approach velocity. Q is identically the root mean square fluctuating density in isotropic fields. Under the isotropic assumption, maximum density fluctuations in the tests flames were estimated to be 0.25 amagat which corresponded to temperature fluctuations of 500[degrees]R. However, fluctuating density gradients across the flame were shown to be greater than gradients in the flow direction, demonstrating that isotropy assumptions in flames may be crude estimates.