| OCR Text |
Show - 2 - parameters, some of which are hard to evaluate, should be given. Generally, the estimation based on reliable data seems practically impossible. An experimental study on sound emlSSlOn from turbulent premixed flames was carried out by Smith and Kilham[4] . They found that the sound noise level was proportional to the product of the burner diameter, flow velocity, and burning velocity. The spectrum of the sound was confined in a low frequency region, and the peak frequency was inferred to be expressed as the Strouhal number, in terms of the burner diameter, flow and burning velocities , and the frequency maxim urn. They only presented the averag d normalized combustion-noise spectrum based on an empirical survey of spectra. The sound emlSSlOn during transient flame propagation was studied by Thomas and Williams[5] . They examined the sound emission froIn flammable gas babbles and showed that the sound pressure depended linearly on flame radius and on the square of burning velocity. In their discussion, the flammable gas babbles, which they assumed to be nl0nopole sources, were considered to correspond to the flame front disturbances in usual turbulent burner flames. They inferred that their results are valid at turbulent flames. In turbulent burner flames, flame front disturbances, which might be monopole sources, continue to be generated and consumed. To verify the nOIse emlSSIOn mechanisms in turbulent flames presumed in the monopole source theory, the correlation between the fluctuations of the sound pressure and light emission has been examined[6-8]. Hurle et al. adopted a bandpass width of 50 Hz to 1 kHz[6]. The upper limit was determined because the distortion of the time-derivative of a light emission intensity signal caused by electrical noise became serious at frequency beyond 1 kHz( at most 4 kHz). They pointed out that for pressure waves emitted by individual combustion elements, the period necessary for the pressure wave to propagate from the source to the point of measurement became comparable to the corresponding quarter-periods, and the individual pressure waves no longer had the same relative phase differences at a point in the far -field. Katsuki et al. limited til observing frequency to levels below l.9 kHz in order to prevent the resultant phase difference from spoiling one-to-on orrespondenc s among the signals of sound pressure, ion current, and OH light emission[8] . How ver, no further explanation |
