Radial Traverses Research Articles

Experimental and Computational Investigation of an Advanced Casing Treatment in a Single-Stage High-Speed Compressor

Abstract An advanced casing treatment (CT) design was tested in a single-stage, high-speed axial compressor. The geometry of the casing included bent-axial slots applied near the rotor leading edge. Compressor performance and stall margin (SM) measurements were acquired for both casing treatment and for the smooth wall case at four different corrected speeds. The measurements also included rotor-exit radial traverses of unsteady total pressure. The results indicated an increase in rotor pressure ratio at all operating conditions with the application of the casing treatment. The stall margin also increased at all rotor speeds. Efficiency increases were observed with the casing treatment at design conditions and some off-design conditions. Numerical simulations of the IGV, rotor, and stator were conducted with and without the casing treatment. These results served to validate the efficacy of the part-wheel unsteady Reynolds-averaged Navier–Stokes (URANS) simulations to predict the effects of the casing treatment. The simulations also allowed a detailed study of the flow physics related to the interactions between the rotor and casing.

Dew point measurement at the Stuttgart University ATF

Highly efficient conditioning of the air supplied to the test vehicle is a vital task of an altitude test facility (ATF). The time needed to achieve stable conditions as well as the required amounts of energy and other supplies to achieve those conditions are means to measure the effectiveness of the ATF operation. Optimizing the operation of single facility components only delivers local optima but not a total operational optimum. Hence, the optimized operation of highly interlinked groups of components delivering a defined functionality has to be achieved. A thorough understanding of the operational characteristics of the components is a prerequisite for such an optimization. The complex systems required for cooling the air supplied to test vehicles down to high-altitude temperature conditions is an example for such a complex optimization task. In a first step, the operational envelope of one of the Stuttgart ATF air coolers has been characterized experimentally. This included the analysis of the startup procedures. A new measurement setup to determine the dew point within the Stuttgart University Altitude Test Facility has been implemented for this purpose. Experiments were conducted to measure the dew point using a precise hygrometer. A radial traversing system was implemented to enable measurements at different positions downstream of the air cooler. Since the determination of the dew point takes place in a wide temperature range, a probe for sampling and conditioning of the air was developed. The measured data are of high quality. The achieved results form the basis of the operational optimization of the Stuttgart ATF air cooler system.

CFD Simulations of the New MEXICO Rotor Experiment under Yawed Flow

This paper describes the conduction of CFD simulations with DLR’s flow solver TAU for comparison with the experimental data that has been used within the third phase of the MexNext-III project. More precisely computations for three different inflow velocities under yawed flow condition have been conducted and compared to the experimental results. Important aspects of the simulations such as a sufficiently long simulation time or a sufficiently small time-step are discussed. At first local flow quantities such as the flow velocity along axial and radial traverses, pressure distributions at four radial stations and normal and tangential forces at these sections are compared. Subsequently the integral rotor quantities i.e. rotor thrust and rotor torque are revised. The integration of the CFD results is performed using the complete surface dataset as well as two subsets. The comparison proves that the agreement can substantially be improved if the subset resembles the experimental dataset. Furthermore the integration of the experimental dataset is reviewed and a weak spot in the current reconstruction procedure of the trailing edge pressure is identified. The proposed alternative provides evidence that further improvements are possible. The final comparison shows an acceptable agreement for the medium and high inflow speed case yielding 0% and 10% difference for integral thrust and 9% and 14% for integral torque. The results for the low inflow speed case are divers: while the agreement for the integral thrust is excellent, it is less satisfactory for the integral torque which amounts to 41%.

An experimental study of the wake of a model wind turbine using phase-averaging

The wake of a model wind turbine with a rotor diameter of 0.9 m has been investigated at the design condition (tip speed ratio = 6) in a closed return wind tunnel with a cross-section of 1.8 by 2.7 m. The three bladed rotor was operated in a low turbulence intensity uniform flow. Velocity data were obtained in the wake using a four-wire hot-wire probe at 18 radial traverses from x/D=0.22 to 5. All three components of the velocity vector were resolved instantaneously for a wide range of flow angles and velocity magnitudes. In addition to time averaged data, phase-averages with respect to rotor position have been used to extract periodic phase-averaged motions from the flow. In this way the downstream development of the tip vortices was extracted. The phase-averaging method allows the local phase-averaged and turbulent parts of the stresses associated with the vortical motion to be studied, as well as their production terms. The results show that the periodic phase-averaged structures initially contains most of the turbulent kinetic energy and dominate the edge of the wake. The helical vortex structures interact and pair-up, merge and finally breaks up within x/D=3. From this point the phase-averaged motion is lost and the radial transport of momentum across the wake is seen to increase significantly. By x/D=5 the large scale diffusion has removed the sharp edge of the wake that was characteristic in the initial region, and developed the approximately Gaussian velocity defect which is characteristic of bluff body wakes.

Droplet dynamics and size characterization of high-velocity airblast atomization

Airblast atomizers are especially useful and commonplace in liquid fuel combustion applications. However, the spray formation processes, the droplet dynamics and the final drop size distributions are still not sufficiently understood due to the coupled gas-liquid interactions and turbulence generation. Therefore, empirical and semi-empirical approaches are typically used to estimate the global spray parameters. To develop a physical understanding of the spray evolution, a plain-jet airblast atomizer was investigated in an atmospheric spray rig using the phase-Doppler technique. The simultaneous drop size and axial and radial velocity components were measured on radial traverses across the spray at various axial distances from the nozzle for a range of atomizing pressures. The droplet turbulent and mean kinetic energies were found to be proportional to the atomizing pressure. Hence, the scatter of the radial motion of the droplets increased with the atomizing pressure. A droplet stability analysis was performed to locate the regions characterized by ongoing secondary atomization. The volume-to-surface diameter, D32, of the fully developed spray was compared with estimates provided by five published formulae. The role of liquid viscosity, hence the Ohnesorge number, was found to be negligible in the investigated regime. Three commonly used size distribution functions were fitted to the measured data to analyze their dependence on the atomizing pressure. The Gamma distribution function was found to give the best approximation to the atomization process.

The Effect of Airfoil Clocking on Efficiency and Noise of Low Pressure Turbines

The effect of airfoil clocking (stator-stator interaction) on efficiency and noise of low pressure turbines (LPT) was investigated experimentally in a multistage turbine high-speed rig. The rig consisted of three stages of a state-of-the-art LPT. The stages were characterized by a very high wall-slope angle, reverse cut-off design, very high lift, and very high aspect ratio airfoils. The rig had identical blade count for the second and third stators. The circumferential position of the second stator was individually adjusted with respect to the third stator. Eight different circumferential clocking locations over one pitch were back-to-back tested. The rig was heavily instrumented with miniature five hole probes, hot wires, hot films, total pressure and temperature rakes, pressure tappings on the airfoil surface, two array of Kulites in a rotatory module, etc. Every clocking location was tested with the same instrumentation and at the same operating conditions with the intention of determining the impact of the clocking on the overall efficiency and noise. Due to the large amount of data, the results of this test will be reported in several papers. The present paper contains the impact on the overall efficiency, radial traverses, static pressure fields on the airfoils and averaged sound pressure levels in the duct. The comparison of the results suggests that the efficiency is weakly affected by clocking; however the effect on noise is noticeable for some acoustic tones at certain operating conditions.

Investigating the Cause of Computational Fluid Dynamics Deficiencies in Accurately Predicting the Efficiency and Performance of High Pressure Turbines: A Combined Experimental and Numerical Study

Despite the tremendous progress over the past three decades in the area of turbomachinery computational fluid dynamics, there are still substantial differences between the experimental and the numerical results pertaining to the individual flow quantities. These differences are integrally noticeable in terms of major discrepancies in aerodynamic losses, efficiency, and performance of the turbomachines. As a consequence, engine manufacturers are compelled to frequently calibrate their simulation package by performing a series of experiments before issuing efficiency and performance guaranty. This paper aims at identifying the quantities, whose simulation inaccuracies are preeminently responsible for the aforementioned differences. This task requires (a) a meticulous experimental investigation of all individual thermofluid quantities and their interactions, resulting in an integral behavior of the turbomachine in terms of efficiency and performance; (b) a detailed numerical investigation using appropriate grid densities based on simulation sensitivity; and (c) steady and transient simulations to ensure their impact on the final numerical results. To perform the above experimental and numerical tasks, a two-stage, high-pressure axial turbine rotor has been designed and inserted into the TPFL turbine research facility for generating benchmark data to compare with the numerical results. Detailed interstage radial and circumferential traversing presents a complete flow picture of the second stage. Performance measurements were carried out for design and off-design rotational speed. For comparison with numerical simulations, the turbine was numerically modeled using a commercial code. An extensive mesh sensitivity study was performed to achieve a grid-independent accuracy for both steady and transient analysis.

Identification of Spinning Mode in the Unsteady Flow Field of a Low Pressure Turbine

This paper presents an experimental investigation of the vane-blade unsteady interaction in an unshrouded low pressure (LP) turbine research rig with uneven blade/vane count (72 blades and 96 vanes). The rig was designed in cooperation with MTU Aero Engines and considerable efforts were put on the adjustment of all relevant model parameters. In particular blade count ratio, airfoil aspect ratio, reduced mass flow, reduced speed, and Mach and Reynolds numbers were chosen to reproduce the full scale LP turbine at take off condition. Measurements by means of a fast-response pressure probe were performed adopting a phase-locked acquisition technique in order to provide the time resolved flow field downstream of the turbine rotor. The probe has been fully traversed both in circumferential and radial traverses. The rotor exit is characterized by strong perturbations due to the tip leakage vortex and the rotor blade wake. Circumferential nonuniformities due to the upstream vane wake and to the downstream exit guide vane potential effects are also identified. Furthermore, in the present configuration with an uneven blade/vane count the nonuniformities due to the stator and rotor row are misaligned along the whole turbine circumference and create a spinning mode that rotates in direction opposite to the rotor at a high frequency. The aeroacoustic theory is employed to explain such further unsteady pattern. The variations of the exit flow angle within a cycle of such pattern are not negligible and almost comparable to the ones within the blade passing period.

A High Temperature High Bandwidth Fast Response Total Pressure Probe for Measurements in a Multistage Axial Compressor

Over the last decades, fast response aerodynamic probes have been recognized as a robust measurement technique to provide time-resolved flow field data in turbomachinery environments. Still, most of the existing probe designs are restricted to low temperature applications (<120 °C) either because of sensor temperature range limitations or packaging issues. Measurements in turbomachines also require a small probe size often with a very high bandwidth which are conflicting constraints difficult to satisfy simultaneously. This contribution therefore presents the development of a novel miniature (∅ 2.5 mm ) high temperature single sensor total pressure probe, designed for operation up to 250 °C with a very high bandwidth of 250 kHz. The probe main element is a 1.7 mm diameter commercial piezoresistive transducer placed in a Pitot type arrangement with a flush mounted sensor to provide the highest bandwidth. The details of the probe design are presented as well as the probe calibrations in pressure and in temperature. The effects of using a thermal compensation module or a sense resistor to monitor the temperature drift are described in the context of measurement uncertainty. The probes were characterized in terms of aerodynamic characteristics versus flow angle and Mach number. Shock tube tests have shown a dynamic response of the probe with sensor resonance frequencies well over 300 kHz, with a flat frequency range up to 250 kHz. Two probe prototypes were manufactured and first used in the 3½-stage high speed axial compressor CREATE of the LMFA at École Centrale de Lyon in France. The probes were traversed at each interblade row plane up to temperatures of 180 °C and absolute pressure of 3 bars. The probe was able to resolve the high blade passing frequencies (∼16 kHz) and several harmonics including rotor-stator interaction frequencies up to 200 kHz. Besides the average total pressure distributions from the radial traverses, phase-locked averages and random unsteadiness are presented. The probe spatial and temporal resolutions are discussed in the context of those results.

Implications of garnet resorption for the Lu-Hf garnet geochronometer: an example from the contact aureole of the Makhavinekh Lake Pluton, Labrador

In the contact aureole of the Makhavinekh Lake Pluton (MLP), Labrador, garnet resorption caused redistribution of Lu and loss of Hf, creating spuriously young Lu-Hf garnet ages. Garnet grew during granulite facies regional metamorphism at 1860-1850 Ma. At 1322 Ma, garnet rims were replaced by coronas of cordierite and orthopyroxene during contact metamorphism. Garnet-ilmenite Lu-Hf geochronology using bulk-garnet separates yields apparent ages that young from 1876 ± 21 Ma at 4025 m from the contact to 1396 ± 8 Ma at 450 m from the contact. Toward the contact, garnet crystals are progressively more resorbed. Concentrations of Lu measured by LA-ICP-MS along radial traverses on central sections through relict garnet decrease gently away from the cores but rise steeply within 50-200 lm of the edges of the relicts. Enrichments of Lu in rims of relict garnet demonstrates strong partitioning of Lu into garnet during resorption and modest intracrystalline diffusion. Hafnium distributions could not be measured, but considering the strong incompatibility of Hf with garnet, it is likely that nearly all Hf in resorbed portions of the garnet was lost from the crystals. Lu-Hf ages in the aureole are thus controlled predominantly by this retention of Lu and loss of Hf during garnet resorption. This deduction was tested with a simple numerical model in which the partial retention of Lu and loss of Hf is tracked as a population of garnet is resorbed. Assuming a spherical geometry for garnet porphyroblasts, Rayleigh fractionation is used to approximate initial Lu zoning profiles ranging from flat to steeply decreasing toward garnet rims. The model simulates: (i) Lu-Hf decay for a specified period before resorption; (ii) instantaneous resorption with retention of Lu and loss of Hf from the resorbed portion of the crystal and (iii) Lu-Hf decay during a specified period after resorption. Several parameters influence the modelled age, but garnet resorption and Lu retention are the primary factors. When all other parameters are held constant, larger amounts of resorption and higher degrees of Lu retention produce younger apparent ages (false ages). Similarly, flatter initial Lu profiles yield younger apparent ages as a consequence of the larger proportion of Lu and Hf that resides in the outer portions of the porphyroblast. The difference between the apparent and actual ages is greater if the duration of the pre-resorption decay period is large relative to the post-resorption decay period. Larger crystals in a Gaussian crystal-size distribution (CSD) generally dominate the Lu-Hf budget and produce an older apparent age relative to the age of the mean crystal size. Compared to a symmetrical Gaussian CSD, positively skewed CSDs result in reduced resorption of large crystals and produce an older apparent age. Application of the model to the MLP aureole, positing growth at 1850 Ma and resorption at 1320 Ma, yields model ages that young from 1850 to 1374 Ma toward the contact, in good agreement with the apparent ages determined from geochronology.

Unsteady Pressure Measurements With a Fast Response Cooled Probe in High Temperature Gas Turbine Environments

This paper presents the first experimental engine and test rig results obtained from a fast response cooled total pressure probe. The first objective of the probe design was to favor continuous immersion of the probe into the engine to obtain a time series of pressure with a high bandwidth and, therefore, statistically representative average fluctuations at the blade passing frequency. The probe is water cooled by a high pressure cooling system and uses a conventional piezoresistive pressure sensor, which yields, therefore, both time-averaged and time-resolved pressures. The initial design target was to gain the capability of performing measurements at the temperature conditions typically found at high pressure turbine exit (800–1100°C) with a bandwidth of at least 40 kHz and in the long term at combustor exit (2000 K or higher). The probe was first traversed at the turbine exit of a Rolls-Royce Viper turbojet engine at exhaust temperatures around 750°C and absolute pressure of 2.1 bars. The probe was able to resolve the high blade passing frequency (≈23 kHz) and several harmonics of up to 100 kHz. Besides the average total pressure distributions rom the radial traverses, phase-locked averages and random unsteadiness are presented. The probe was also used in a virtual three-hole mode yielding unsteady yaw angle, static pressure, and Mach number. The same probe was used for measurements in a Rolls-Royce intermediate pressure burner rig. Traverses were performed inside the flame tube of a kerosene burner at temperatures above 1600°C. The probe successfully measured the total pressure distribution in the flame tube and typical frequencies of combustion instabilities were identified during rumble conditions. The cooling performance of the probe is compared with estimations at the design stage and found to be in good agreement. The frequency response of the probe is compared with cold shock-tube results and a significant increase in the natural frequency of the line-cavity system formed by the conduction cooled screen in front of the miniature pressure sensor were observed.

Computational analysis of incompressible turbulent flow in an idealised swirl combustor

Isothermal turbulent swirling flow in a water test rig, representing an idealised swirl combustor, has been investigated experimentally and numerically. The Reynolds number based on combustor inlet diameter and mean axial velocity was 4600. Two cases were investigated at two different swirl intensities. Time-averaged velocities and RMS turbulence intensities were measured by Laser Doppler Anemometer (LDA), along radial traverses at different axial stations. In the three-dimensional, transient computations, Large Eddy Simulations (LES) and URANS Reynolds Stress Models (RSM) have basically been employed as modelling strategies for turbulence. To model subgrid-scale (SGS) turbulence for LES, the models owing to Smagorinsky and Voke were used. In one of the cases, Detached Eddy Simulations (DES) were also applied. The predictions have been compared with the measurements. It has been observed that LES provides the best overall accuracy, where no significant differences between the Smagorinsky and Voke models could be discerned.

Experimental and numerical investigation of isothermal flow in an idealized swirl combustor

PurposeThe main purpose of the paper is the validation of different modelling strategies for turbulent swirling flow of an incompressible fluid in an idealized swirl combustor.Design/methodology/approachExperiments have been performed and computations carried out for a water test rig, for a Reynolds number of 4,600 based on combustor inlet mean axial velocity and diameter. Two cases have been investigated, one low swirl and the other high swirl intensity. Measurements of time‐averaged velocity components and corresponding rms turbulence intensities were measured using laser Doppler anemometer, along radial traverses at different axial locations. In the three‐dimensional, unsteady computations, large eddy simulation (LES) and URANS (Unsteady Reynolds Averaged Navier‐Stokes Equations or Reynolds Averaged Numerical Simulations) RSMs (Reynolds‐stress models) are basically employed as modelling strategies for turbulence. To model subgrid‐scale turbulence for LES, the models due to Smagorinsky and Voke are used. No‐model LES and coarse‐grid direct numerical simulation computations are also performed for one of the cases.FindingsThe predictions are compared with the measurements and reveal that LES provided the best overall accuracy for all of the cases, whereas no significant difference between the Smagorinsky and Voke models are observed for the time‐averaged velocity components.Originality/valueThis paper provides additional valuable information on the performance of various modelling strategies for turbulent swirling flows.

Measurements of mean and fluctuating temperature in an underexpanded jet using electrostrictive laser-induced gratings

Instantaneous temperature measurements were obtained in an underexpanded jet using electrostrictive laser-induced gratings. Evaluation of the technique under static, low-pressure conditions provided a baseline uncertainty or precision for single-shot temperature measurements of 4.4% of the local mean temperature, which represents the minimum detectable temperature fluctuation. The underexpanded jet was operated at a nozzle pressure ratio of 2.39 and a fully expanded jet Mach number of 1.19. Data were acquired along the centerline and over two radial traverses through the shear layer. Mean temperature data agree well with expectations, describing the shock-cell structure and the compressible shear layer. The growth in shear-layer width with downstream distance can be identified in the mean and fluctuating temperature measurements. Temperature fluctuations are near the baseline detection limit in the jet core and surrounding ambient air, and reach a maximum in the shear layer. The temperature fluctuation measurements compare well with previous computational and experimental work, confirming the application of the technique to a turbulent, supersonic flow.

Testing palaeo-environmental proxies in Jurassic belemnites: Mg/Ca, Sr/Ca, Na/Ca, δ 18O and δ 13C

We test the hypothesis that the noise in belemnite records through time results from changing climate. We do so by comparing palaeo-proxies (Mg/Ca, Sr/Ca, Na/Ca, δ 18O, δ 13C) in two species of Toarcian belemnites, Acrocoelites ( Odontobelus) vulgaris (14 specimens) and Acrocoelites ( Acrocoelites) subtenuis (10 specimens), deposited in a period of time that was short relative to the timescale of climate change. The specimens are interpreted to have been regurgitated from the stomach of a marine predator and so were deposited together in a geological instant of Early Jurassic time. In A. ( O.) vulgaris, isotopic and elemental palaeo-proxies group tightly, suggesting that this belemnite occupied a restricted environmental niche. In A. ( A.) subtenuis, elemental and isotopic compositions spread more widely, suggesting that this species occupied a greater range of habitats. Possible sexual dimorphism in A. ( O.) vulgaris is reflected in bimodal Mg/Ca and differing intra-rostral distribution of Mg in putative dimorphs; concentrations decline by 30% from apical line to rim in specimens with high Mg/Ca, but are largely invariant radially in specimens with lower Mg/Ca. Sexual dimorphism in A. ( A.) subtenuis yields a bimodality in size, Mg/Ca, and intra-rostral distribution of Mg; larger specimens have a higher Mg/Ca, and Mg concentrations that increase along radial traverses from apical line to rim, whilst the smaller specimens show a mid-rostral minimum in Mg concentration. Radial profiles of concentrations for Sr and Na vary little in all forms. Toarcian belemnites of these species thus show strong biofractionation effects on Mg/Ca, and little or no observable biofractionation of Sr/Ca or Na/Ca. It follows that much of the noise in belemnite records results from biofractionation, and from a cosmopolitan life-style of some belemnite species, rather than climate-change. For Toarcian belemnites, we suggest that Sr/Ca is the best indicator of palaeo-temperature.

Radial Resistivity/IP Surveys Using a Downhole Current Electrode

The sensitivity and resolution of surface resistivity and Induced Polarization (IP) surveys can be severely restricted by the conductivity and thickness of overburden material. Electrode arrays using a current electrode buried below the overburden can increase the sensitivity and resolution of surveys conducted in such environments. A current electrode is located down a drillhole, and resistivity and chargeability measurements are made on the surface along survey lines radial to the drillhole. Reconnaissance resistivity/IP surveys using the downhole radial electrode array were conducted in the Cambro-Ordovician Mt Windsor Volcanics near Charters Towers in Queensland, an area covered by electrically conductive unconsolidated sediments of the Late Tertiary Campaspe Beds. The surveys demonstrated the effectiveness of the array in detecting anomalies from buried base metal targets. Data were acquired from a number of surveys conducted around a series of adjacent drillholes to provide continuous coverage across the prospective area. The data from individual radial traverses show the sounding response of the ground as well as the ground response along the profile. A method was devised for removing the sounding response, which is local to each drillhole, from each radial traverse so that the data from each traverse could be merged to form a continuous data set. A number of issues have been identified for future research which include the optimum depth of the buried current electrode, removal of the sounding response, and interpretation of the survey data.

Application of magnetic and spectrometry methods in the detection of human activity in soils: a case study at the archaeological site of Kitros (Northern Greece)

AbstractIn the present study the archaeological site of (Louloudies–Kitros) (northern Greece) is investigated by the use of magnetic and spectrometry geophysical methods. Soil samples were collected from several trenches within the archaeological site, as well as from radial traverses around kilns, which were used for glass production. Moreover, samples were collected from a long traverse, starting from the middle of the archaeological site and extending outside the site, in order to identify the limits of the settlement.Measurements were successful in identifying the different occupation levels and the boundaries of the archaeological site. High values of magnetic susceptibility and iron oxide content correlated well with the periods of intense human activity. Around the kilns, the mean values of the magnetic susceptibility indicated the effect of the fire mechanism in the surrounding area and gave some first evidence concerning the use of the kilns. Comparison of the magnetic and chemical properties of the anthropogenic soils suggests that their enhancement may be used as an index for locating areas and features of archaeological interest.Additionally, archaeomagnetic data have been obtained from a pottery and a glass kiln of the site. The magnetic directions (D = 355°, I = 62°, a95 = 3.6 and D = 356°, I = 51°, a95 = 2.3) were correlated to the reference curves compiled for Bulgaria and showed a satisfactory consistency for the declination, but the inclination values of the recorded magnetic field were divergent by about 10°. Copyright © 2001 John Wiley & Sons, Ltd.

Influence of liquid flow conditions on spray characteristics of internal-mixing twin-fluid atomizers

Abstract For an internal-mixing twin-fluid atomizer with coaxial liquid feed the dependence of the spray characteristics on the liquid flow conditions, i.e. laminar, turbulent or cavitating, has been investigated. Results obtained with Phase Doppler Analyzer, high speed photography and pressure drop measurements were correlated with the nozzle geometry and the flow rates of liquid and gas. The investigated nozzle allowed variation of five geometric parameters, the most important ones being the liquid inlet port diameter, the nozzle outlet port diameter and the length of the outlet port. The volume flow rate of the liquid V L ranged from 5 to 100 L/h, and of the compressed air V A from 1 to 20 mn3/h. For visualizing the internal flow, an optically accessible planar nozzle of comparable geometry has been used. It was observed that flow conditions of the liquid jet leaving the inlet port have a strong influence on the distributions of the D32 droplet diameter along radial traverses across the spray. For short outlet ports (lo/do ≲ 2), the maximal D32 for laminar liquid jet is always found on the spray axis, whereas a turbulent liquid jet leads to nearly even profiles. For long outlet ports (lo/do ≳ 4) even profiles are found in either case.

Three-Dimensional Flowfield Downstream of an Axial-Flow Turbine Rotor

A systematic and comprehensive investigation was performed to provide detailed data on the threedimensional viscous flow phenomena downstream of a modern turbine rotor and to understand the flow physics such as the origin, nature, development of wakes, secondary flow, and leakage flow. The experiment was carried out in the Axial Flow Turbine Research Facility (AFTRF) at The Pennsylvania State University, with velocity measurements taken with a 3-D LDV System. Two radial traverses at 1% and 10% of chord downstream of the rotor were performed. Sufficient spatial resolution was maintained to resolve blade wake, secondary flow, and tip leakage flow. The wake deficit is found to be substantial, especially at 1% of chord downstream of the rotor. At this location, negative axial velocity occurs near the tip, suggesting flow separation in the tip clearance region. Cross-correlations are mainly associated with the velocity gradient of the wake deficit. The radial velocities, both in the wake and in the endwall region, are found to be substantial. Two counter-rotating secondary flows are identified in the blade passage, with one occupying the half span close to the casing and the other occupying the half span close to the hub. The tip leakage flow is restricted to 10% immersion from the blade tip. There are strong vorticity distributions associated with these secondary flows and tip leakage flow. The passage averaged data are in good agreement with design values.

Experimental Investigation of the Flow Field in a Multistage Axial Flow Compressor

The nature of the flow field in a three stage axial flow compressor, including a detailed survey at the exit of an embedded stator as well as the overall performance of the compressor is presented and interpreted in this paper. The measurements include area traverse of a miniature five hole probe (1.07 mm dia) downstream of stator 2, radial traverses of a miniature five hole probe at the inlet, downstream of stator 3 and at the exit of the compressor at various circumferential locations, area traverse of a low response thermocouple probe downstream of stator 2, radial traverses of a single sensor hot‐wire probe at the inlet, and casing static pressure measurements at various circumferential and axial locations across the compressor at the peak efficiency operating point. Mean velocity, pressure and total temperature contours as well as secondary flow contours at the exit of the stator 2 are reported and interpreted. Secondary flow contours show the migration of fluid particles toward the core of the low pressure regions located near the suction side casing endwall corner.