Evolutionary behavior of sprays produced by pressure atomizers

Abstract

Drop siae and velocity distributions were obtained for a pressure atomizer operating at 415 KF' a and a flowrate of 2.Zltf min. The nozzle produced a hollow-cone spray pattern with large drops travelling at high velocities confined to a relatively narrow conical sheet. These f a s t moving drops entrained air into the spray which served to convect the smaller drops t o the center of the spray cone. Measurements of the drop size and the axial and radial velocity components were obtained at radial and axial stations within the spray. The drop angles of trajectory for each size clws were calculated at representative stations. At the periphery, these results revealed that the large drops travelled at an angle of -30' corresponding t o the spray cone angle whereas the smallest drops moved at a mean angle of approximately +So with excursions to +IS0. With lower atomization pressures, a recirculation region at the center of the spray cone within a few centimeters from the exit of the nozzle was measured. Simultaneous measurements of the drop size and velocity revealed the axial velocity relaxation behavior for the various size classes. These results emphasize the need in the predictions of sprays to not only include the characteristics of the spray formation but also the dynamics of the spray drops and their interaction with the ambient turbulent flow. INTRODUCTIONDetailed mewurements of fuel spray drop size and velocity distributions remain LS an important contribution toward the understanding and improvement of spray combustion. Although numerous experimental investigations have been conducted, few, if any, have provided the complete data needed to facilitate development of advanced gas turbine combustors and to provide a sufficient base for testing and guiding the computational efforts. The required data must include simultaneous drop siae and velocity measurements at points near the plane of injection and at numerous downstream planes. Measurements of the gaseous mean velocity and turbulence intensity in the presence of the polydispersed phase are also required. Even in the case of simple pressure swirl atomizers, it is well-known that the spray formation is complicated by the entrainment and inflow of surrounding air, relaxation behavior of the various size classes, drop-drop interactions and

Experiments on Spray Combustion in a Gas Turbine Model Combustor

“…Mean axial drop velocity profiles for the non-burning case show a typical doublehump shape which was also observed by others in hollow-cone sprays, for example, see Bachalo et al (1986) and McVey et al (1989). Negative mean drop velocity are seen to exist around the spray center line for axial locations larger than 25 mm, indicating the existence of a toroidal recirculation zone.…”

On the measurement of particle concentration near a stagnation point

“…In the terminology of Bachalo et al (1986), this is a "spatially-averaged" value of d32. At z/D>0.8, where (3) is similar within about 3 ~tm.…”

“…m-3) of size class "i", where the class includes diameters between d~ and d~ + ~ d. If the instrument measures n~ particles in this class over a sampling time, T~, then Saffman (1987) and Bachalo et al (1986) propose that, for one-dimensional flow normal to the fringes,…”

“…The number density is a derived, not a directly measured, quantity by optical instruments which detect single particles, such as the phaseDoppler anemometer. The purpose of this note is to compare the method used by Saffman (1987) and Bachalo et al (1986) to find the number density with an alternative approach, originally derived for the measurement of the number density of seeding particles in laser-Doppler anemometry (Capp 1983). Our conclusion will be that, particularly near mean stagnation points of the flow, only the method used by Capp gives reasonable answers.…”

Performance and operating envelope of imaging and scattering particle sizing instruments

“…However, data from single particle counters must be corrected before it can be compared with other types of instruments that are not sensitive to the particles' velocity (7). Also, measurements should be corrected before calculating number density or total mass.…”