This thesis reports on the structure of the flow inside a nozzle which produces a naturally oscillating jet flow. The nozzle consists of a short cylindrical chamber with a concentric triangular-inlet orifice at one end and a circular exit lip at the other end. This triangular-jet nozzle was developed from the “fluidic-precessing-jet” (FPJ) nozzle, which has a similar arrangement of components, but has a circular rather than a triangular inlet. For reliably oscillating flow, the FPJ nozzle should have an inlet-to-chamber expansion ratio of at least 5.0, a chamber lengthto- diameter ratio between 2.6 and 2.8, and an exit-lip height of about 0.1 chamber diameters. The triangular-jet nozzle produces a continuously and aperiodically oscillating jet flow which is different from the FPJ flow. The oscillation occurs at smaller inlet-to-chamber expansion ratios (2.1 ≲ D /de₁ ≲ 3.5) and over a wider range of chamber lengths (2.0 ≲ L /D ≲ 2.5). The initial spreading angle of the jet flow is smaller, but is still much larger than that of non-oscillating, axisymmetric turbulent-jet flows. In addition, the external “oscillating-triangular-jet” (OTJ) flow has preferred azimuthal directions which are aligned with the three corners of the orifice. The kinetic-energy-loss coefficient of the OTJ nozzle is much smaller than that of the FPJ nozzle because oscillation occurs at much smaller inlet-to-chamber expansion ratios. For a narrow range of length-to-diameter ratios (1.00 ≲ L/D ≲ 1.25), the triangular-inlet nozzle can also produce a non-oscillating or “stationary deflected triangular jet” (SDTJ) which reattaches asymmetrically to the inside surface of the cylindrical chamber. The SDTJ has a weak tendency to oscillate, which suggests that flow patterns required for self-excited oscillation are already present in the SDTJ flow. Surface-flow visualisation and surface-pressure measurements in the SDTJ nozzle have provided the location of critical points and bifurcation lines on the chamber wall, and from this the topology of the SDTJ flow is deduced. Some details of the flow such as a jet-reattachment node near the chamber exit and a strong swirl adjacent to the inlet orifice are known from previous studies of the FPJ flow, but there are many newly observed features. The most easily identified of these are two sink-focus separation points, one on each side of the reattachment node but closer to the inlet plane. The foci counter rotate and are of unequal size. Reverse flow through the exit plane of the chamber is attracted to the larger focus. The vortex core rising from each focus is entrained by the reattaching-jet (SDTJ) flow and is drawn out of the chamber. A backward-facing pressure probe placed in the OTJ “reattaching-flow” region of chamber wall can be used as a reliable detector of jet-flow oscillation. Cross-correlating the signal from this detector probe with simultaneous static-pressure measurements elsewhere on the chamber wall gives a conditionally-averaged pressure on the wall of the OTJ chamber. The OTJ wall-pressure distribution has the same features as the SDTJ surface-pressure distribution, but it has greater asymmetry about a mirror plane drawn through the chamber axis and the detector probe. An array of three backward-facing pressure probes has been used as an “event detector” for conditionally-sampled (PIV) measurements of non-axial velocity components in cross-sections of the OTJ nozzle. The event-detection scheme responds only to a preselected (counter-clockwise) direction of motion of the oscillating-jet flow. The streamline patterns constructed from the conditionally-sampled measurements confirm the presence of the jet-reattachment node, the swirl and the sink foci identified from the SDTJ surface-flow visualisation. The shear-layer interaction between the jet from the triangular orifice and the swirl (adjacent to the inlet plane) produces strong longitudinal vortices in the ensemble-averaged flow. The jet flow distributes these vortices through the length of the chamber. Vortex cores representing the vortices are reconstructed by tracking streamline foci from one PIV cross-section plane to another. The tracking process includes the connection and termination of vortex cores in a manner which is consistent with the Helmholtz vortex law. In this flow field, the vortex core produced by the swirl and the vortex core rising from the larger sink-focus vortex on the chamber wall are connected to form a loop. The extent to which this vortex loop is contained within the chamber determines whether or not the flow is oscillating. If only a small fraction (e.g. 8%) of the vortex circulation passes through the exit plane of the nozzle, the loop is trapped inside the chamber and the deflected jet oscillates. If the length of the chamber is halved, about 35% of vortex circulation escapes from the nozzle and the oscillation stops. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1353005 / Thesis (Ph.D.) - University of Adelaide, School of Mechanical Engineering, 2009

This thesis reports on the structure of the flow inside a nozzle which produces a naturally oscillating jet flow. The nozzle consists of a short cylindrical chamber with a concentric triangular-inlet orifice at one end and a circular exit lip at the other end. This triangular-jet nozzle was developed from the “fluidic-precessing-jet” (FPJ) nozzle, which has a similar arrangement of components, but has a circular rather than a triangular inlet. For reliably oscillating flow, the FPJ nozzle should have an inlet-to-chamber expansion ratio of at least 5.0, a chamber lengthto- diameter ratio between 2.6 and 2.8, and an exit-lip height of about 0.1 chamber diameters. The triangular-jet nozzle produces a continuously and aperiodically oscillating jet flow which is different from the FPJ flow. The oscillation occurs at smaller inlet-to-chamber expansion ratios (2.1 ≲ D /de₁ ≲ 3.5) and over a wider range of chamber lengths (2.0 ≲ L /D ≲ 2.5). The initial spreading angle of the jet flow is smaller, but is still much larger than that of non-oscillating, axisymmetric turbulent-jet flows. In addition, the external “oscillating-triangular-jet” (OTJ) flow has preferred azimuthal directions which are aligned with the three corners of the orifice. The kinetic-energy-loss coefficient of the OTJ nozzle is much smaller than that of the FPJ nozzle because oscillation occurs at much smaller inlet-to-chamber expansion ratios. For a narrow range of length-to-diameter ratios (1.00 ≲ L/D ≲ 1.25), the triangular-inlet nozzle can also produce a non-oscillating or “stationary deflected triangular jet” (SDTJ) which reattaches asymmetrically to the inside surface of the cylindrical chamber. The SDTJ has a weak tendency to oscillate, which suggests that flow patterns required for self-excited oscillation are already present in the SDTJ flow. Surface-flow visualisation and surface-pressure measurements in the SDTJ nozzle have provided the location of critical points and bifurcation lines on the chamber wall, and from this the topology of the SDTJ flow is deduced. Some details of the flow such as a jet-reattachment node near the chamber exit and a strong swirl adjacent to the inlet orifice are known from previous studies of the FPJ flow, but there are many newly observed features. The most easily identified of these are two sink-focus separation points, one on each side of the reattachment node but closer to the inlet plane. The foci counter rotate and are of unequal size. Reverse flow through the exit plane of the chamber is attracted to the larger focus. The vortex core rising from each focus is entrained by the reattaching-jet (SDTJ) flow and is drawn out of the chamber. A backward-facing pressure probe placed in the OTJ “reattaching-flow” region of chamber wall can be used as a reliable detector of jet-flow oscillation. Cross-correlating the signal from this detector probe with simultaneous static-pressure measurements elsewhere on the chamber wall gives a conditionally-averaged pressure on the wall of the OTJ chamber. The OTJ wall-pressure distribution has the same features as the SDTJ surface-pressure distribution, but it has greater asymmetry about a mirror plane drawn through the chamber axis and the detector probe. An array of three backward-facing pressure probes has been used as an “event detector” for conditionally-sampled (PIV) measurements of non-axial velocity components in cross-sections of the OTJ nozzle. The event-detection scheme responds only to a preselected (counter-clockwise) direction of motion of the oscillating-jet flow. The streamline patterns constructed from the conditionally-sampled measurements confirm the presence of the jet-reattachment node, the swirl and the sink foci identified from the SDTJ surface-flow visualisation. The shear-layer interaction between the jet from the triangular orifice and the swirl (adjacent to the inlet plane) produces strong longitudinal vortices in the ensemble-averaged flow. The jet flow distributes these vortices through the length of the chamber. Vortex cores representing the vortices are reconstructed by tracking streamline foci from one PIV cross-section plane to another. The tracking process includes the connection and termination of vortex cores in a manner which is consistent with the Helmholtz vortex law. In this flow field, the vortex core produced by the swirl and the vortex core rising from the larger sink-focus vortex on the chamber wall are connected to form a loop. The extent to which this vortex loop is contained within the chamber determines whether or not the flow is oscillating. If only a small fraction (e.g. 8%) of the vortex circulation passes through the exit plane of the nozzle, the loop is trapped inside the chamber and the deflected jet oscillates. If the length of the chamber is halved, about 35% of vortex circulation escapes from the nozzle and the oscillation stops. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1353005 / Thesis (Ph.D.) - University of Adelaide, School of Mechanical Engineering, 2009

In single phase, steady, turbulent axisymmetric jets, the time-averaged velocity field can be characterised by the decay in centreline velocity and increased spread with increasing distance from the jet orifice. In a two-phase or ‘particle-laden’ jet, the particles will modulate the jet turbulence and exchange momentum with the gas phase. Consequently, these effects reduce both the centreline velocity decay and spreading rates with respect to the single-phase jet. Empirical exponential scaling factors were found by previous authors to describe the reduced centreline decay and spreading rates well for low Stokes numbers. In this thesis, power-law scaling factors are found to scale well a wide range of centreline velocity decay and spreading rate data published over the past 40 years, for a wide range of Stokes numbers. The power-law scaling is composed of three different regimes. For low Stokes numbers St₀ ≲20, it is found that the gas phase centreline velocity, u₀/uc collapses if plotted as a function of x/D(1 + Ø₀)⁻¹, and the velocity profile half widths r₁/ ₂ collapse if plotted as a function of x/D(1+Ø₀)⁻¹. Here, u₀ is the exit velocity, Ø₀ is the exit mass loading, x is the axial coordinate and D is the pipe diameter. For intermediate Stokes numbers, u₀/uc collapses if plotted as a function of x/D(1 + Ø₀)⁻¹ and r₁/ ₂ collapses if plotted as a function of x/D(1 + Ø₀)⁻¹/². For high Stokes numbers St₀ ≳ 200, u₀/uc collapses if plotted as a function of x/D(1 + Ø₀)⁻¹/² and the half width is approximately independent of Ø₀. In addition to the velocity of the gas phase, other aspects of particle- laden jets are found to be amenable to scaling by power-law functions. It is found that reported solid phase mass flux data scales similarly to gas phase measurements. Limited solid phase concentration and entrainment measurements reported in the literature are also found to scale by power-law functions. Whereas that limited data was obtained from the literature, measurements of the distribution of particles in particle-laden jets were conducted to further assess the validity of the scaling regimes to the solid phase. A planar light scattering technique is conducted to measure the distribution of particles in an axisymmetric jet and their subsequent scaling (or lack thereof) are reported for a variation in Ø₀, Stokes number and gas phase jet exit density. For Stokes numbers based on the pipe friction velocity St* ₀ ∼ 1, half widths of particle distributions were found to scale with x/D(1+Ø₀)⁻¹/² . The apparent centreline concentration was found to be independent of Ø₀ at this same St* ₀ . For Stokes numbers based on the pipe friction velocity St*₀ < 1, half widths are independent of Ø₀. The effect of the other parameters, i.e. Stokes number and density ratio, on centreline distributions and half widths are also investigated. Measurements of particle distributions, delivered via an annular channel, in a triangular oscillating jet (OJ) flow are also reported for a variation in momentum ratio, the ratio of OJ momentum to channel momentum and mass loading. The results of the variation in momentum ratio on particle distributions are compared with an existing precessing jet (PJ) study. It is the aim of this study to determine the experimental conditions for which the OJ nozzle is superior to the PJ nozzle. The use of an OJ nozzle is preferable at an industrial scale by virtue of its lower driving pressure compared with a PJ nozzle. It is found that particle distributions in a PJ flow spread at a greater rate with increasing momentum ratio compared with the spread of particles in an OJ flow. However, at momentum ratios approximately less than unity, the absolute spread from an OJ is greater. This also corresponds to nozzle driving pressure less than approximately 10kPA. For an increase in mass loading, the spread of particle distribution in the OJ decreases and recirculation increases. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1337352 / Thesis (M.Eng.Sc.) -- University of Adelaide, School of Mechanical Engineering, 2008

In single phase, steady, turbulent axisymmetric jets, the time-averaged velocity field can be characterised by the decay in centreline velocity and increased spread with increasing distance from the jet orifice. In a two-phase or ‘particle-laden’ jet, the particles will modulate the jet turbulence and exchange momentum with the gas phase. Consequently, these effects reduce both the centreline velocity decay and spreading rates with respect to the single-phase jet. Empirical exponential scaling factors were found by previous authors to describe the reduced centreline decay and spreading rates well for low Stokes numbers. In this thesis, power-law scaling factors are found to scale well a wide range of centreline velocity decay and spreading rate data published over the past 40 years, for a wide range of Stokes numbers. The power-law scaling is composed of three different regimes. For low Stokes numbers St₀ ≲20, it is found that the gas phase centreline velocity, u₀/uc collapses if plotted as a function of x/D(1 + Ø₀)⁻¹, and the velocity profile half widths r₁/ ₂ collapse if plotted as a function of x/D(1+Ø₀)⁻¹. Here, u₀ is the exit velocity, Ø₀ is the exit mass loading, x is the axial coordinate and D is the pipe diameter. For intermediate Stokes numbers, u₀/uc collapses if plotted as a function of x/D(1 + Ø₀)⁻¹ and r₁/ ₂ collapses if plotted as a function of x/D(1 + Ø₀)⁻¹/². For high Stokes numbers St₀ ≳ 200, u₀/uc collapses if plotted as a function of x/D(1 + Ø₀)⁻¹/² and the half width is approximately independent of Ø₀. In addition to the velocity of the gas phase, other aspects of particle- laden jets are found to be amenable to scaling by power-law functions. It is found that reported solid phase mass flux data scales similarly to gas phase measurements. Limited solid phase concentration and entrainment measurements reported in the literature are also found to scale by power-law functions. Whereas that limited data was obtained from the literature, measurements of the distribution of particles in particle-laden jets were conducted to further assess the validity of the scaling regimes to the solid phase. A planar light scattering technique is conducted to measure the distribution of particles in an axisymmetric jet and their subsequent scaling (or lack thereof) are reported for a variation in Ø₀, Stokes number and gas phase jet exit density. For Stokes numbers based on the pipe friction velocity St* ₀ ∼ 1, half widths of particle distributions were found to scale with x/D(1+Ø₀)⁻¹/² . The apparent centreline concentration was found to be independent of Ø₀ at this same St* ₀ . For Stokes numbers based on the pipe friction velocity St*₀ < 1, half widths are independent of Ø₀. The effect of the other parameters, i.e. Stokes number and density ratio, on centreline distributions and half widths are also investigated. Measurements of particle distributions, delivered via an annular channel, in a triangular oscillating jet (OJ) flow are also reported for a variation in momentum ratio, the ratio of OJ momentum to channel momentum and mass loading. The results of the variation in momentum ratio on particle distributions are compared with an existing precessing jet (PJ) study. It is the aim of this study to determine the experimental conditions for which the OJ nozzle is superior to the PJ nozzle. The use of an OJ nozzle is preferable at an industrial scale by virtue of its lower driving pressure compared with a PJ nozzle. It is found that particle distributions in a PJ flow spread at a greater rate with increasing momentum ratio compared with the spread of particles in an OJ flow. However, at momentum ratios approximately less than unity, the absolute spread from an OJ is greater. This also corresponds to nozzle driving pressure less than approximately 10kPA. For an increase in mass loading, the spread of particle distribution in the OJ decreases and recirculation increases. / http://proxy.library.adelaide.edu.au/login?url= http://library.adelaide.edu.au/cgi-bin/Pwebrecon.cgi?BBID=1337352 / Thesis (M.Eng.Sc.) -- University of Adelaide, School of Mechanical Engineering, 2008

The Stability And Control CONfiguration (SACCON) model represents an emerging trend in airplane design where the classical tube, wing and empennage are replaced by a single tailless configuration. The challenge is to assure that these designs are stable and controllable. Nonlinear aerodynamic behavior is observed on the SACCON at higher incidence angles due to leading edge vortex structures. Active Flow Control (AFC) used in preliminary design represents a promising solution to the longitudinal stability problems and this was demonstrated experimentally on a semi span model. AFC can be used to trim the SACCON in pitch and it alters forces and moments comparable to common control surface deflections. A combination of AFC and control surface deflection may increase the overall efficiency and opens up a variety of maneuvering possibilities. This implies that AFC should be treated concomitantly with other design parameters and should be considered in the preliminary design process already and not as an add-on tool. Integral force and moment data was supplemented by observations using Pressure Sensitive Paint (PSP) and flow visualization. Two arrays of individually controlled sweeping jets, one located along the leading edge and the other along the flap hinge provided the AFC input needed to alter the flow. The array positioned over the flap-hinge of the model was most effective in stabilizing the wing by decreasing the pitching moment at lower and intermediate angles of incidence. This effect was achieved by reducing the spanwise flow on the swept back portion of the wing through jet-entrainment that also affected the leading edge vortex. Leading edge actuation showed some beneficial effects by inhibiting the formation of the leading edge vortex near the wing tip. A preliminary study using suction was carried out. The tests were carried out at Mach numbers smaller than 0.2 and Reynolds numbers based on the root chord of the model that approached 10⁶.