Gerald L. Morrison

Instability process in low Reynolds number supersonic jets

The objective of the present research was to characterize the instability of supersonic jets of Mach numbers 1.4, 2.1, and 2.5 in the Reynolds number range around 8000. In this Reynolds number range the jet instability has its maximum coherence so that its properties as well as the acoustic properties can be most clearly identified. Growth rate, wavelength, and wave orientation of dominant spectral components of the instability for the three Mach number jets were measured in order to characterize the instability. The saturation and subsequent decay on the instability are coincident with a drastic decrease in the coherence of the instability. Associated research shows that the phenomenon of rapid growth and decay is of fundamental importance in the noise generation process.

Reynolds Number Dependence in Supersonic Jet Noise

An experimental study of the noise production by high speed jets over a wide range of Reynolds numbers has been performed. Two jets of nominal Mach numbers 1.5 and 2.3 were run over a Reynolds number range from 5300 to 107,000. Microphone measurements of the radiated noise and hot-wire measurements of the flow fluctuations demonstrate that at low Reynolds numbers coherent flow instabilities produce a dominant portion of the noise. In the nominal Mach number 2.3 jet these instability waves convect downstream supersonically with respect to the ambient air. In the nominal Mach number 1.5 jet the instabilities convect downstream subsonically. In both cases however, sound pressure level amplitude contours show that the low Reynolds number jets radiate noise comparable to intermediate and high Reynolds number jets. These measurements constitute substantial evidence that a flow instability model of the dominant noise generators may be appropriate for conventional high Reynolds number supersonic jets. a0 C D d M m n r Re St u U Nomenclature speed of sound outside jet wavespeed in the downstream direction diameter of the jet effective diameter of the jet Mach number of the jet at the exit normalized mass velocity fluctuations = azimuthal mode number = radial distance from jet centerline = Reynolds number = p Ud/n. = Strouhal number =fd/ (/(/is frequency) = local velocity =mean centerline velocity of the jet at the nozzle exit

Five-hole pressure probe analysis technique

Abstract A refined calibration technique is presented for five-hole pressure probes operating in the non-nulling mode. The four 3D calibration surface equations required to reduce data obtained from the probe are curve-fit using a 3D curve-fitting program. The relatively simple equations are quick and easy to use for data reduction. The shape of the 3D surfaces are useful in determining if a probe should not be used due to any machining abnormality or damage a probe has sustained. The contours can also be used to determine the range of flow angles a particular probe can measure.

Comparison of orifice and slotted plate flowmeters

The performance of a standard β = 0.50 orifice flowmeter is compared to the same flowmeter with a slotted orifice plate replacing the standard orifice plate. The slotted orifice plate has the same total open area as the standard plate and consists of three concentric rings, each of which contains several radial slots. The flow upstream of both orifice plates in preconditioned using a concentric pipe device to produce a wide range of axial velocity profiles without swirl. The discharge coefficient of the standard plate varies from the base value by −1% to +6%, while that of the slotted orifice plate varies by only ±0.25%. When swirl is generated upstream of the orifice plates, the discharge coefficients vary by 5% and 2% for the standard and slotted orifice plates, respectively. These data indicate that the slotted orifice flowmeter is superior to the standard orifice flowmeter in maintaining its calibration over a wide range of inlet flow conditions. The slotted orifice plate can be a ‘drop in’ replacement for a standard orifice plate.

Response of a slotted orifice flow meter to an air/water mixture

Abstract The ability of a flow meter to respond predictably to the presence of liquid and gas is important to the natural gas industry and to users of steam. In both cases, the gas can become saturated and some liquids can condense in the line. The response of orifice flow meters to the presence of liquids is erratic and produces considerable uncertainty. Turbine flow meters can sustain severe damage when subjected to two phase flow. The slotted orifice flow meter has been developed to address the problem of upstream flow conditioning. This device has been shown to be insensitive to the upstream velocity profile. To further evaluate the flow meter for use by the natural gas industry, the effects of adding liquid to a gas flow upon the meter performance has been investigated by subjecting a slotted orifice flow meter with an equivalent β ratio of 0.50 to a two phase flow consisting of air and water.

Uncertainty estimates in DGV systems due to pixel location and velocity gradients

The accuracy of Doppler global velocimeter systems is dependent upon many parameters. The ability to obtain accurate values of the light intensity at specified locations in space can easily be the largest contributor to uncertainty in the resultant velocity measurement. The effects of light intensity spatial gradients and the accuracy of spatial locations of the individual measurements are analysed to provide estimates of the magnitude of these uncertainties and to determine how these uncertainties can be reduced.

Beta ratio, swirl and Reynolds number dependence of wall pressure in orifice flowmeters

Abstract Experimental work has been performed in an effort to gain a better understanding of the flow field inside orifice flowmeters and the pressure field generated on the walls of the pipe and orifice plate. As a part of a larger study, extensive wall pressure measurements have been made on the pipe wall from four pipe diameters upstream of the orifice plate to six pipe diameters downstream, as well as on both the upstream and downstream faces of the orifice plate. These measurements were performed for Reynolds numbers of 54 700; 91 100 and 122 800; for beta ratios of 0.50 and 0.75 with air as the working fluid. An adjustable swirl plate was installed, which was used to impart varying amounts of swirl into the flow upstream of the orifice plate. For each swirl case, Pitot and static pressure probes were used to characterize the upstream flow field while the pipe wall and orifice plate surface pressures were measured.

Sound spectra of gas dispersion in an agitated tank

Abstract Sound spectra of gas dispersion phenomena were obtained using a hydrophone and a spectrum analyzer at three locations (the sparge ring, the impeller and a position in the bulk of the tank) in an agitated tank as a function of gassing rate and impeller rotational speed. For the sparge ring, the spectra made it possible to distinguish between gas sparging controlled by the sparge ring itself and the natural volume pulsation frequency of the bubbles and gas sparging controlled by the impeller flow. Sound pressure level spectra, obtained for an impeller under various flooding and non-flooding conditions, showed harmonies in the sound pressure levels only for non-flooding conditions. The number and amplitude of the harmonics diminished as the impeller approached flooding. The sound pressure level spectra for a position in the bulk of the tank resonated with the impeller rotational speed in some cases in which the impeller was not flooded. In other cases in which the impeller was flooded, the sound pressure level spectra of the bulk of the tank resembled the impeller spectra at frequencies below 500 Hz.

Installation effects upon orifice flowmeters

Abstract An experimental study has been undertaken to quantify the effect of the inlet velocity distribution upon the coefficient of discharge, C d . A two inch (50.8 mm) diameter orifice run was operated at a Reynolds number of 91 000 with a beta ratio, β, of 0.75. The upstream pipe section was replaced with a one inch pipe mounted concentrically inside the two inch pipe. The mass flowrate was held constant by an array of sonic nozzles upstream of the concentric pipes and was split between the two. By varying the ratio of the flow split, various inlet velocity profiles were generated. The results show that the change in coefficient of discharge is related to first-, second- and third-order moments of momentum: ∫ 0 R ∫ 0 2π i=1,2,3, Analysis of data presented by Morrow et al. (Flow Measurement and Instrumentation 2 (1) (1991) 14–20) shows the same relationship. This paper proposes the use of this correlation to develop criteria for correcting the discharge coefficient given the variation of the inlet velocity profile from ‘fully developed’ flow. The velocity profile can be measured at the upstream flange tap location with the orifice plate removed, and that profile can be used to generate the moment of momentum to be used to correct the coefficient of discharge.

Leakage Optimization of Labyrinth Seals Using a Navier-Stokes Code

It has been demonstrated that design optimization of labyrinth seals using the present numerical model is quite beneficial. The results shown include important, but previously unknown effects on the leakage rate, especially that of step height. Further, complete details are given of a very effective seal designed using this technique. Measurements using turbine flow meters revealed that the optimized configuration gives a 60 and 79 percent leakage reduction over the baseline design at laboratory and higher (extrapolated) seal pressure drops, respectively.