The factors governing stability and combustion intensity in "mechanically" atomised oil flames.

Problems of combustion instability and the need for wider range in firing rates have drawn attention to the need for more fundamental knowledge of the performance and the characteristics of pressure-jet atomized oil flames. A secondary problem is the complication and inconvenience of measurements on full scale equipment. In the present study various operational conditions for a typical modern burner with a pressure-jet atomizer and a swirled air supply were investigated. The variables chosen were fuel rate and spray angle. Data obtained consisted of gas composition, velocity and temperature measurements. The development of a laboratory unit was considered desirable to provide a means of facilitating measurements on flames of this type and a rig was constructed which consisted of a suitable combustion box and 3/10ths scale burner. It was shown that the matching of the spray angle to the air register aerodynamics affected the flame characteristics very markedly, Regions of high unburnt fuel occurred due to fuel penetration and poor mixing outside and downstream of the main part of the flame. The effect of widening the spray angle which usually occurs with wide range types of pressure-jet atomizers at the lower fuel rates was shown to be particularly unsatisfactory, and to effect the black smoke limit. The general course of mixing and the progress of combustion through the flame was deduced from calculation from the gas composition results. The distribution and effects of recirculation were shown from the velocity measurements. These were related to data for simple jets. In particular, it was deduced that internal recirculation in the core was largely responsible for flame stabilization. Theoretically derived scaling criteria were investigated by comparing full scale and model results. These showed that reasonable comparisons were possible between the two, with flames scaled either on the basis of similarity in velocity and relative momentum between spray and air, or on similarity in residence times in the flame. The former was thought probably to be better, Cold flow measurements were made, and showed differences in the size of the central recirculation core between hot and cold but not in the peak velocity to peak velocity diameter early in the flame. Combustion oscillation measurements on a marine boiler produced data which suggested strongly that these were of the "non-acoustic" type. None of the theories so far suggested appear to - ii - account with much accuracy for the observed data. In the course of experimental work various special Instruments for flow measurement were developed. It was concluded that the choice of spray distribution to suit a particular aerodynamic pattern was vital to the optimum performance of the burner. In all the cases investigated this was found to be to some extent imperfect. It is considered that decided improvements could be made to the air distribution in burners of this type. The following possibilities are suggested solutions, the use of a higher degree of swirl and a confining quarl, upstream air injection or the use of multiple oil nozzles and by avoiding devices where the spray angle varies widely over the operating range