On Flow Structure in an Oscillating-Triangular-Jet Nozzle: Conditionally-Averaged Wall Pressure

A jet flow from a triangular-inlet orifice can be made to oscillate or “precess” by partially confining it with a short cylindrical chamber. However, for a narrow range of chamber lengths, the jet is deflected but does not oscillate. Earlier flow-visualisation experiments with this stationary-deflected triangular jet (SDTJ) have revealed a complex flow pattern inside the chamber. The aim of this study is to discover by conditionally averaging if there is a similar flow pattern in the fully oscillating jet flow. A backward-facing pressure probe placed in the reattachingflow region of the chamber surface is used as a detector of jet-flow oscillation. With the signal from this detector and simultaneous wall-pressure measurements, we have obtained conditionally-averaged pressure distributions on the wall of the oscillating-triangular-jet (OTJ) nozzle. The OTJ surfacepressure distributions have the same features as the timeaveraged SDTJ surface-pressure distribution, and so the inferred OTJ surface-flow topology has the same features as the SDTJ. However, for the OTJ, the chamber is twice as long, and the conditionally-averaged pressure distribution is much more asymmetrical about a mirror plane drawn through the axis of the nozzle and the detector probe. We conclude that the net radial force due to the asymmetric pressure distribution is large enough to produce significant deflection of the OTJ flow in directions normal to the chamber axis. Introduction A device consisting of a circular inlet orifice and a cylindrical chamber with an exit lip (Figure 1) can produce a naturallyoscillating jet flow. This device is known as the “fluidicprecessing-jet” (FPJ) nozzle. Nathan et al. [4] show that, for the FPJ to oscillate reliably, the inlet-expansion ratio (D/de1) must be larger than about 5.0, and the length ratio of the chamber must be in the range 2.6∼L/D < ∼2.8. The nozzle is usually made with an exit-lip height of 0.1D. Nathan’s early studies of the FPJ [4] identified several flow features inside the chamber. 1. The first of these features is asymmetric deflection of the jet flow from the inlet orifice, and reattachment of the deflected jet to the side wall of the chamber (Figure 1). reverse flow Induced bifurcation ring, Time−averaged negative− NB1 point,N Instantaneous reattachment Swirl precession Jet−flow Figure 1: The fluidic-precessing-jet (FPJ) nozzle: typical geometry and early flow-visualisation results. 2. There is a circumferential negative-bifurcation line between the path traced by the precessing reattachment point and the inlet plane of the nozzle. 3. There is a torus of strong tangential swirl between the negative-bifurcation line and the inlet plane of the nozzle. The swirling fluid spirals inward and is entrained into the jet flow from the inlet orifice. 4. Fluid entrained into the jet flow is ejected from the chamber. The partial vacuum thus produced induces a reverse flow through the exit plane of the nozzle. Mi et al. [3] found that an FPJ-like flow oscillation is obtained with a number of different inlet-orifice shapes. In particular, with an equilateral-triangular orifice (Figure 2), oscillation is obtained at much lower expansion ratios than with the FPJ. The oscillating-triangular-jet (OTJ) nozzle therefore has much lower pressure loss than the FPJ nozzle. In a parametric study of the OTJ nozzle, Lee et al. [2] have verified that continuous oscillation occurs at much smaller expansion ratios than for the FPJ. Lee et al. [1] also observe that, for a narrow range of nozzle geometries (D/de1=3.5, 1.0<L/D∼1.25), the jet is deflected towards the wall but does not oscillate. In the stationarydeflected flow, “time-averaged” observations capture the effects of jet deflection and only details of turbulence motion are removed. Figure 3(a) is a typical surface-flow image of this “stationary-deflected triangular jet” (SDTJ). The image shows that the SDTJ has the same four features as listed above, but it also reveals new detail. As shown in Figure 3(c), there is a sink focus (Fa, Fb) upstream and on each side of the jet-reattachment node (N). These foci counter rotate and are of unequal size. The reverse flow through the chamber-exit plane is attracted to the larger focus. The surface-flow pattern clearly has no symmetry, and this is a reminder that the instantaneous, time-dependent oscillating-jet flow should also be strongly asymmetric. The similarities between the OTJ and the SDTJ, and the propensity of the SDTJ to oscillate tempt us to believe that the OTJ flow pattern is much the same as the SDTJ flow pattern, but we need more evidence. Therefore, the aim of the research reported in this paper is to perform conditionally-sampled experiments which record the internal pressure distribution on the