Gas Turbine Laboratory Reports

Asymptotic analysis of numerical wave propagation in finite difference equations

Sat, 01 Jan 1983 00:00:00 GMT

Asymptotic analysis of numerical wave propagation in finite difference equations Giles, M. (Michael); Thompkins, William T. An asymptotic technique is developed for analysing the propagation and dissipation of wave-like solutions to finite difference equations. It is shown that for each fixed complex frequency there are usually several wave solutions with different wavenumbers and the slowly varying amplitude of each satisfies an asymptotic amplitude equation which includes the effects of smoothly varying coefficients in the finite difference equation's. The local group velocity appears in this equation as the velocity of convection of the amplitude. Asymptotic boundary conditions coupling the amplitudes of the different wave solutions are also derived. A wave packet theory is developed which predicts the motion, and interaction at boundaries, of wavepackets, wave-like disturbances of finite length. Comparison with numerical experiments demonstrates the success and limitations of the theory. Finally an asymptotic global stability analysis is developed which gives results which agree with other stability analyses and which can be applied to a wider range of problems. March 1983; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1983; Includes bibliographical references (page 134)

A numerical analysis of 3-D inviscid stator/rotor interactions using non-reflecting boundary conditions

Wed, 01 Jan 1992 00:00:00 GMT

A numerical analysis of 3-D inviscid stator/rotor interactions using non-reflecting boundary conditions Saxer, André P. (André Pierre) This dissertation presents a method for the computation of three-dimensional inviscid, transonic steady and unsteady flows, primarily in axial flow turbines. The work is divided into two major contributions. The first is an algorithm for the solution of the 3-D Euler equations which incorporates a second-order accurate numerical smoothing for non-uniform grids and steady-state non-reflecting boundary conditions. Fourier analysis applied to the linearized Euler equations is used to develop novel quasi-3-D non-reflecting boundary conditions at the inflow/outflow and at the stator/rotor interface. The accuracy, effectiveness and robustness of the boundary condition formulation is demonstrated through several subsonic and transonic test cases and through comparison with the standard 1-D formulation. The second contribution consists in the study of three specific flow phenomena occurring in an axial flow turbine.; First, the steady-state effects of an inlet spanwise stagnation temperature gradient in a transonic stage are analyzed. The mechanism for the migration of the temperature as well as the extent of the non-uniformity are assessed. Then, the secondary flow produced by a combined thermal and vortical inlet distortion on a downstream moving rotor is studied. The extent of the radial mixing for steady and unsteady flow is assessed as a function of the strength of the inlet disturbance. The third case is an analysis of the steady, unsteady and time-averaged flow fields in a highly loaded industrial transonic turbine stage. In particular, the unsteady shock interaction due to the impact of the stator trailing edge shock wave off the downstream rotor is studied. From the last two cases it is concluded that in many aspects the time-averaged results are extremely close to the steady-state values, even with strong unsteady shock interaction.; For each case the mechanisms for the creation of the secondary flow and deviations from a steady, uniform inlet conditions flow field are presented and analyzed. March 1992; Includes bibliographical references (pages 230-239)

Parametric dependencies of aeroengine flutter for flutter clearance applications

Sat, 01 Jan 2000 00:00:00 GMT

Parametric dependencies of aeroengine flutter for flutter clearance applications Khalak, Asif, 1972- This thesis describes the effects of operational parameters upon aeroengine flutter stability. The study is composed of three parts: theoretical development of relevant parameters, exploration of a computational model, and analysis of fully scaled test data. Results from these studies are used to develop a rational flutter clearance methodology-a test procedure to ensure flutter-free operation. It is shown, under conditions relevant to aeroengines, that four nondimensional parameters are necessary and sufficient for flutter stability assessment of a given rotor geometry. We introduce a new parameter, termed the reduced damping, g/p*, which collapses the combined effects of mechanical damping and mass ratio (blade mass to fluid inertia). Furthermore, the introduction of the compressible reduced frequency, K*, makes it possible to uniquely separate the corrected performance map from the non-dimensional operating environment (including inlet temperature and pressure).; Simultaneous plots of the performance map of corrected mass flow and corrected speed, (mc,nc), with the (K*, g/p*) map provide a dimensionally complete and fully integrated view of flutter stability, as demonstrated in the context of a historic multimission engine. A parametric, .computational study was conducted using a 2D, linearized unsteady, compressible, potential flow model of a vibrating cascade. This study showed the independent effects of Mach number, inlet flow angle, and reduced frequency upon flutter stability in terms of critical reduced damping, which corroborates the 4D view of flutter stability. Test data from a full-scale transonic fan, spanning the full 4D parameter space, were also analyzed. A novel boundary fitting tool was developed for data processing, which can handle the generic case of sparse, multidimensional, binary data.; The results indicate that the inlet pressure does not alone determine the flight condition effects upon flutter, which necessitates the use of the complete 4D parameter set. Such a complete view of the flutter boundary is constructed, and sensitivities with respect to various parameters are estimated. A rational flutter clearance procedure is proposed. Trends in K* and g/p* allow one to rapidly determine the worst-cases for testing a given design. One may also use sensitivities to extend the results of sea level static (SLS) testing, if the worst case is relatively close to the SLS condition. August 2000; Includes bibliographical references (pages 223-228)

Measurements of rotor stalling in a matched and a mismatched multistage compressor

Sun, 01 Jan 1995 00:00:00 GMT

Measurements of rotor stalling in a matched and a mismatched multistage compressor Silkowski, Peter D. (Peter Daniel) This paper presents the results from a set of experiments on stall inception in multistage axial flow compressors. The experiments were tailored to investigate phenomena having a wide range of time and length scales. This range of scales was motivated by two previously observed paths to stall. Parametric changes such as tip clearance, inlet distortion and mismatch were carried out to demonstrate the importance of component coupling in the stall inception process. Evidence is presented for the importance of the local compressor characteristic in determining where and when the initiation of the stall inception process will occur. Although the stall inception process may begin as a localized event, its growth into rotating stall is governed by the environment established by the coupling of the various compression system components. Finally, the tip flow field, specifically the rotor tip leakage jet, is shown to be a key feature in the stall inception process. April 1995; Includes bibliographical references (page 21)

A computational model for rotating stall and inlet distortions in multistage compressors

Fri, 01 Jan 1999 00:00:00 GMT

A computational model for rotating stall and inlet distortions in multistage compressors Gong, Yifang, 1964- This thesis presents the conceptualization and development of a computational model for describing three-dimensional non-linear disturbances associated with instability and inlet distortion in multistage compressors. Specifically, the model is aimed at simulating the non-linear aspects of short wavelength stall inception, part span stall cells, and compressor response to three-dimensional inlet distortions. The computed results demonstrated the first-of-a-kind capability for simulating short wavelength stall inception in multistage compressors.; The adequacy of the model is demonstrated by its application to reproduce the following phenomena: (1) response of a compressor to a square-wave total pressure inlet distortion; (2) behavior of long wavelength small amplitude disturbances in compressors; (3) short wavelength stall inception in a multistage compressor and the occurrence of rotating stall inception on the negatively sloped portion of the compressor characteristic; (4) progressive stalling behavior in the first stage in a mismatched multistage compressor; (5) change of stall inception type (from modal to spike and vice versa) due to IGV stagger angle variation, and "unique rotor tip incidence" at these points where the compressor stalls through short wavelength disturbances. The model has been applied to determine the parametric dependence of instability inception behavior in terms of amplitude and spatial distribution of initial disturbance, and intra-blade-row gaps.; It is found that reducing the inter-blade row gaps suppresses the growth of short wavelength disturbances. It is also concluded from these parametric investigations that each local component group (rotor and its two adjacent stators) has its own instability point (i.e. conditions at which disturbances are sustained) for short wavelength disturbances, with the instability point for the compressor set by the most unstable component group. For completeness, the methodology has been extended to describe finite amplitude disturbances in high-speed compressors. Results are presented for the response of a transonic compressor subjected to inlet distortions. March 1999; Includes bibliographical references (pages 175-182)

Effects of inlet conditions on centrifugal diffuser performance

Wed, 01 Jan 1997 00:00:00 GMT

Effects of inlet conditions on centrifugal diffuser performance Deniz, Sabri This report examines the influence of inlet flow conditions, including Mach number, flow angle, blockage, and axial flow non-uniformity, on the performance and operating range of a straight channel centrifugal compressor diffuser. The research was carried out in a unique facility specifically developed to provide the diffuser with a controlled inlet flow. The tests were carried out for inlet Mach number up to values greater than unity and for a range of inlet flow angles up to the onset of rotating stall. It was found that expressing the overall diffuser pressure recovery coefficient, defined using availability or mass averaged inlet total pressure, as a function of momentum averaged diffuser inlet flow angle yields a relationship which is essentially independent of diffuser inlet flow distortion, blockage, or Mach number. Further, the operating range of the diffuser was limited by the onset of rotating stall at a momentum averaged diffuser inlet flow angle ([alpha]crit = 70.5 ±0.5), which was also independent of the inlet flow field axial distortion and Mach number. The straight channel diffuser was designed to be comparable to a previously tested discrete passage diffuser and the performance of the two was compared; the overall pressure recovery of the former was found to be roughly 10% higher than that of the latter. Both diffuser types, straight channel and discrete passage diffuser showed similar behavior regarding the insensitivity of the performance and operating range to inlet flow axial non-uniformities and Mach number. The report also presents information on recent developments in the area of centrifugal compressor diffusers together with a detailed review of the open literature. March 1997; Includes bibliographical references (pages 169-190)

Effects of non-axisymmetric tip clearance on axial compressor performance and stability

Wed, 01 Jan 1997 00:00:00 GMT

Effects of non-axisymmetric tip clearance on axial compressor performance and stability Graf, Martin Bowyer; Wong, Thomas S. (Thomas Sheung); Greitzer, E. M. (Edward M.), 1941-; Marble, Frank E.; Tan, Choon Sooi; Shin, Hyoun-Woo; Wisler, David C., 1941- The effects of circumferentially non-uniform tip clearance on axial compressor performance and stability have been investigated experimentally and analytically. A theoretical model for compressor behavior with non-axisymmetric tip clearance has been developed and used to design a series of first-of-a-kind experiments on a four-stage, low speed compressor. The experiments and computational results together show clearly the central physical features and controlling parameters of compressor response to non-axisymmetric tip clearance. It was found that the loss in stall margin was more severe than that estimated based on average clearance. The stall point was, in fact, closer to that obtained with uniform clearance at the maximum clearance level. The circumferential length scale of the tip clearance (and accompanying flow asymmetry) was an important factor in determining the stall margin reduction.; For the same average clearance, the loss in peak pressure rise was 50% higher for an asymmetry with fundamental wavelength equal to the compressor circumference than with wavelength equal to one-half the circumference. The clearance asymmetry had much less of an effect on peak efficiency; the measured maximum efficiency decrease obtained was less than 0.4 percent compared to the 8% decrease in peak pressure rise due to the asymmetric clearance. The efficiency penalty due to non-axisymmetric tip clearance was thus close to that obtained with a uniform clearance at the circumferentially-averaged level. The theoretical model accurately captured the decreases in both steady-state pressure rise and stable operating range which are associated with clearance asymmetry.; It also gave a good description of the observed trends of (i) increasing velocity asymmetry with decreasing compressor flow, and (ii) decreasing effect of clearance asymmetry with decreasing dominant wavelength of the clearance distribution. The time resolved data showed that the spatial structure of the pre-stall propagating disturbances in the compressor annulus was well represented and that the stability limiting process could be linked to the unsteady structure of these disturbance modes. The model was also utilized for parametric studies to define how compressor performance and stability is affected by the circumferential distribution of clearance, steady-state compressor pressure-rise characteristic, and system dynamic parameters. Sensitivity to clearance asymmetry was found to fall off strongly with the (asymmetry-related) reduced frequency and to increase with peak pressure rise and increasing curvature of the characteristic near the peak. September 1997; Statement of responsibility on title-page reads: M.B. Graf, T.S. Wong, E.M. Greitzer, F.E. Marble, C.S. Tan, H-W Shin, D.C. Wisler; Includes bibliographical references (pages 34-35)

Active stabilization of rotating stall and surge in a transonic single stage axial compressor

Wed, 01 Jan 1997 00:00:00 GMT

Active stabilization of rotating stall and surge in a transonic single stage axial compressor Weigl, Harald Jürgen Rotating stall and surge have been stabilized in a transonic single-stage axial compressor using active feedback control. The control strategy is to sense upstream wall static pressure patterns and feed back the measured signal to an annular array of twelve separately modulated air injectors upstream of the rotor. At tip relative Mach numbers of 1.0 and 1.5, the control achieves a 11% and 3.5% reduction in stalling mass flow respectively. With control, the steady amount of injected air is equal to 3.6% of the design compressor mass flow. At a tip Mach number of 1.0 the stall inception dynamics and effective active control schemes are similar to results for low-speed axial compressors. The range extension can be achieved by individually damping the first and second spatial harmonics of the pre-stall rotating stall perturbations using constant gain feedback. The pre-stall dynamics are different at a tip Mach number of 1.5 than at the lower speed.; Both one-dimensional (surge) and two-dimensional (rotating stall) perturbations must be stabilized to increase the compressor operating range. At design speed, the instability is initiated by approximately 10 rotor revolutions of rotating stall followed by classic surge cycles. In accord with the results from a compressible stall inception analysis which has been applied to the compressor with actuation and refined based on forced response measurements, the zeroth, first, and second pre-stall harmonics each include more than one lightly damped mode which can grow into the large amplitude instability. Forced response testing has identified several modes traveling up to 150% of rotor speed for the first three spatial harmonics. Simple constant gain control cannot damp all of these modes and thus cannot stabilize the compressor at this speed.; Robust dynamic control is therefore used to stabilize the multiple modes which comprise the first three harmonic perturbations in this transonic region of operation. Eigenvalue perturbations are incorporated into an H [infinity] control design to directly address uncertainty in the dominant eigenvalue stability. July 1997; Includes bibliographical references (pages 297-302)

The effect of upstream rotor vortical disturbances on the time-average performance of axial compressor stators

Wed, 01 Jan 1997 00:00:00 GMT

The effect of upstream rotor vortical disturbances on the time-average performance of axial compressor stators Valkov, Theodore V. (Theodore Valkov) Time-accurate Reynolds-averaged Navier-Stokes simulations have been carried out to investigate the impact of upstream rotor wakes and tip leakage vortices on the loss of a compressor stator. The objective is to (1) identify the unsteady flow mechanisms responsible for performance changes, (2) quantify these changes, and (3) extract useful design information. There are two generic mechanisms with significant impact on performance: reversible recovery of the energy in the disturbances (beneficial) and non-transitional boundary layer response (detrimental). For both wakes and tip leakage vortices, the impact of these mechanisms can be described in the same two-dimensional terms. In presence of unsteady flow, the efficiency of the design under consideration is 0.2 points higher than that obtained using a mixing-out steady flow approximation (0.5 points recovery benefit minus 0.3 points from boundary layer response). The effects of tip vortices and wakes are of comparable importance.; The impact of stator interaction with upstream wakes and vortices depends on the following parameters: axial spacing, loading, aid the frequency of wake fluctuations in the rotor frame. At reduced spacing, this impact becomes significant. For a spacing of 0.07 chords, stage efficiency is 0.6 points higher relative to the steady flow (1.2 points recovery benefit minus 0.6 points from boundary layer response). About 1/2 to 2/3 of the efficiency gain observed experimentally can be attributed to the interaction with upstream wakes and vortices. In the relative frame, wakes fluctuate in time. For fluctuation frequencies between 0.3-0.8 times the blade passing frequency, recovery does not occur, and there is a significant difference between the effects of fluctuating and steady wakes (with the same ensemble-averaged properties). The most important aspect of the tip vortex is the velocity defect, which is perceived by the stator in the same manner as a wake.; A model of recovery and boundary layer response in an embedded stage indicates that a mixing-out steady flow approximation underestimates stage efficiency by 0.3-0.5 points (for typical designs) and by 0.6-1.0 points (for closely-spaced blade rows). A region in design space exists where interaction has a beneficial and relatively constant impact on efficiency. Outside this region, interaction benefits rapidly disappear. For a typical blading diffusion factor of 0.45, the beneficial region is approximately delimited by the de Haller criterion. The detrimental aspects of boundary layer response may be mitigated by (a) selective removal of boundary layer fluid from the suction surface, or (b) tailoring of the blade loading to reduce loss in the front part of the blade. August 1997; Includes bibliographical references (pages 121-130)

A numerical investigation of a flutter in a transonic fan

Mon, 01 Jan 1996 00:00:00 GMT

A numerical investigation of a flutter in a transonic fan Isomura, Kousuke The mechanism of the bending mode flutter of a modern transonic fan has been studied using a quasi-3D viscous unsteady code. The type of flutter in the scope of this research is that for a highly loaded blade with a tip relative Mach number just above unity, commonly referred to as transonic stall flutter. This type of flutter is often encountered in modern wide chord fans without a part span shroud. The code written as a part of this research uses an upwinding scheme with Roe's 3rd-order flux differencing, and Johnson and King's turbulence model with later modification by Johnson and Coakley. An extensive series of code validation calculations were performed and the reliability of the code has been verified against data and other calculational procedures. The calculations of the flow in this fan revealed that the source of the flutter is an oscillation of the passage shock, rather than a stall. As blade loading increases, the passage shock moves forward. Just before the passage shock unstarts, the stability of the passage shock decreases, and the shock oscillates at a large amplitude between unstarted position and started position with small blade vibration. The shock foot of the oscillating passage shock on the blade pressure surface exerts the dominant blade exciting force. October 1996; Includes bibliographical references (pages 155-158)

Blade passage flow structure effects on axial compressor rotating stall inception

Mon, 01 Jan 1996 00:00:00 GMT

Blade passage flow structure effects on axial compressor rotating stall inception Hoying, Donald Andrew A new computational approach has been developed to study the inception of rotating stall in axial compressors. Using this approach the flow structures within the compressor blade passages have been examined in order to determine their influence on the process of rotating stall inception. Both two and three-dimensional numerical simulations were carried out. The two-dimensional computations showed a long wave-length (or modal) type of stall inception which was found to be well described by existing compressor stability models. The numerical results were used to directly confirm the various assumptions used in the formulation of the stability models. The three-dimensional computations of rotating stall displayed a short lengthscale type of stall inception with the same character as that seen in experiments. The central feature of the flow associated with the development of the short lengthscale stall cell was the tip clearance vortex moving forward of the blade row leading edge. Vortex kinematic arguments were used to provide a physical explanation of this motion. The resulting criteria for the inception of the short length-scale stall depends upon local flow phenomena related to the tip clearance flow. Thus, unlike the modal stall situation, the flow structure within the blade passages must be addressed when describing the stability of an axial compression system to short length-scale disturbances. December 1996; Includes bibliographical references (pages 125-130)

Rotor wake behavior in a transonic compressor stage and its effect on the loading and performance of the stator

Tue, 01 Jan 1980 00:00:00 GMT

Rotor wake behavior in a transonic compressor stage and its effect on the loading and performance of the stator Durali, Mohammad The structure and behavior of wakes from a transonic compressor rotor and their effects on the loading and performance of the downstream stator have been studied experimentally. The rotor is 2 feet in diameter with a tip Mach number of 1.23 and a measured pressure ratio of 1.66. Time and space resolved measurements have been completed of the rotor and stator outlet flows, as well as of the pressure distribution on the surface of the stator blades. The data is analyzed and the rotor-stator viscous interaction is studied both qualitatively and quantitatively. It is found that the wakes from this rotor have large flow angle and flow Mach number variations from the mean flow, significant pressure fluctuations and a large degree of variation from blade to blade and from hub to tip. There is a significant total pressure defect and practically no static pressure variation associated with the stator wakes. Wakes from the rotor exist nearly undiminished in the exit flow of the stator and decay in the annular duct behind the stator. The pressure at all the points along the chord over each of the stator blades' surfaces, fluctuates nearly in phase in response to the rotor wakes, that is the unsteady chordwise pressure distribution is determined by the changes in angle of incidence to the blade and not by the local velocity fluctuations within the passage. There are large unsteady forces on the stator blade, induced by the rotor wakes, as high as 25% of the steady forces. The stator force due to the rotor wakes lags the incidence of the wake on the leading edge by approximately 1800 for most radii. April 1980; Also issued as: Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1980; Includes bibliographical references (leaves 44-45)

Role of flow alignment and inlet blockage on vaned diffuser performance

Wed, 01 Jan 1997 00:00:00 GMT

Role of flow alignment and inlet blockage on vaned diffuser performance Phillips, Michael Stephen A computational investigation of the effects of inlet conditions on straight-channel diffuser performance is undertaken. The steady, three-dimensional, Navier-Stokes solver used for the investigation is found to adequately model the performance of a diffuser that has been previously examined experimentally. Results indicate that, contrary to the established view, vaned diffuser channel performance is weakly dependent on throat blockage. Rather, channel pressure rise is strongly affected by flow angle alignment with the diffuser centerline; misalignment of the flow can cause separation and reduced channel performance. This result challenges current design methods, and indicates that the designer is capable of sculpting the diffuser vanes to change the flow angle alignment, thus enabling control of both performance and range. In support of experimental results, overall diffuser performance is found to be largely independent of inlet axial distortion. Inlet nonuniformities are attenuated within the diffuser channel due to a spanwise work transfer which energizes regions of high flow angle misalignment, thus preventing the development of localized channel stall, and preserving good diffuser performance. This result indicates that axially twisted vanes, which are tailored for nonuniform inlet flow, may be unnecessary; simple untwisted vanes display no loss of performance when subjected to severe inlet distortion. September 1997; Includes bibliographical references (pages 81-84)

Dynamic control of rotating stall in axial flow compressors using aeromechanical feedback

Fri, 01 Jan 1993 00:00:00 GMT

Dynamic control of rotating stall in axial flow compressors using aeromechanical feedback Gysling, Daniel L. Dynamic control of rotating stall in an axial flow compressor has been implemented using aeromechanical feedback. The control strategy developed used an array of wall jets upstream of a single stage compressor which were regulated by locally reacting reed valves. These reed valves responded to pressure perturbations in the flow that were associated with small amplitude perturbations that precede rotating stall. The control strategy was designed such that the combined system of compressor plus the reed valve controller was stable in previously unstable operating conditions. A 10% decrease in the stalling flow coefficient was achieved using this dynamic feedback control strategy, and the stable flow range was extended with no noticeable change in the steady state performance of the compression system.; The experimental demonstration is the first use of aeromechanical feedback to extend the stable operating range of an axial flow compressor, as well as the first use of locally reacting feedback and dynamic compensation techniques to stabilize rotating stall in an axial flow compressor. The design of the experiment was based on a two-dimensional model of the rotating stall dynamics which incorporated the effect of aeromechanical feedback. The physical mechanism responsible for rotating stall in axial flow compressors was examined with focus on the role of dynamic feedback in stabilizing compression system instability. The effectiveness of the aeromechanical control strategy was predicted, and experimentally demonstrated, to be a function of a set of non-dimensional control parameters that determine the interaction of the control strategy and the rotating stall dynamics.; Predictions based on linear stability analyses and non-linear numerical simulations agreed qualitatively with the steady state and time resolved experimental data. During the experimental investigations, large amplitude, one-dimensional acoustic oscillations were observed in the compression system with aeromechanical feedback stabilization. Based on these observations, the role of the compression system parameters in the acoustic oscillations was examined analytically and a method was developed to reduce these oscillations. The mechanism responsible for the generation of self-excited acoustic oscillations, and the implications for dynamic control of compression system instabilities was also examined. August 1993; Includes bibliographical references (pages 250-252)

Active control of rotating stall in axial compressors

Wed, 01 Jan 1992 00:00:00 GMT

Active control of rotating stall in axial compressors Paduano, James D. (James Donald) An active control system has been implemented on a low-speed single-stage axial compressor. This control system stabilizes the perturbations which normally lead to rotating stall, thus extending the range of operation of the compressor to a flow coefficient 23% below the natural stall flow coefficient. Sensing of the perturbations which precede stall is accomplished using a circumferential set of hot wires mounted at an axial station ahead of the rotors. Actuation is accomplished by a set of 12 high-response inlet guide vanes, whose individual deflections can be controlled independently by a digital computer. The feedback scheme used in this work is motivated by a recently developed model for rotating stall. In this model, the perturbation axial velocity as a function of circumferential position at any axial station completely determines the state of the system.; Furthermore, circumferential sinusoidal waves of the perturbation are the fundamental eigenmodes, and these eigenmodes develop as rotating waves around the annulus. When the eigenvalues associated with these eigenmodes become unstable, the system diverges into rotating stall. Therefore, feedback stabilization of these rotating waves is used as the technique to eliminate rotating stall. The model for rotating stall is extended to include the effects of high-response inlet-guide-vane actuation. This model is then converted from a system of partial differential equations (PDEs) to a set of ordinary differential equations (ODEs). This conversion will be shown to yield a model to which standard control and identification techniques can be applied. The experimental investigation consisted of two main parts. The first part is systematic identification of the relevant compressor dynamics.; The procedures described yield an accurate model of the compressor input-output behavior over the frequency range of interest (DC to twice rotor rotation frequency), for flow coefficients which span the entire range of stable and unstable (stabilized) operation of the system. This identification verifies the basic behavior predicted by the model and provides quantitative information for control system design. The second part of the experimental investigation is active stabilization. The goal of active control is to extend the range of the compressor to the lowest flow coefficient possible. It is shown that, in the compressor studied, the circumferential sinusoidal modes go unstable in succession - first mode, followed by second, followed by third, as flow coefficient is reduced. Thus, additional range of the compressor is gained for each additional mode stabilized. March 1992; Includes bibliographical references (p. 231-233)

Active control of rotating stall in a three-stage axial compressor

Fri, 01 Jan 1993 00:00:00 GMT

Active control of rotating stall in a three-stage axial compressor Haynes, Joel M. Stall inception in a three-stage axial compressor has been suppressed over a range of previously unstable operating points through the feedback of velocity perturbations to the inlet flow field. Perturbations were generated using 12 individually actuated guide vanes at the compressor inlet. The operating range was extended by 7.8% to a slope of 0.9 on the pressure rise characteristic. Over this range data describing the compressor's pressure rise and torque were collected. Flow field measurements upstream of the compressor revealed the excitation of spatial harmonics in the annular flow field before stall inception. With or without feedback, a spatial mode was observed to grow into a stall cell without a discontinuity of amplitude or position. The decrease in the stall inception mass flow as a result of damping the critical spatial mode indicated the importance of spatial modes in the stall inception process in this compressor.; Destabilization was originally caused by the first mode. After stabilizing an under-damped mode, the flow range was extended until the next sequential mode became unstable. The independent behavior of the modes before stall inception is described by four versions of a small disturbance model of compressor dynamics adapted from that of Moore and Greitzer. The two more sophisticated versions are presented here for the first time. The models require a description of the compressor's geometry, knowledge of its pressure rise characteristic, and some versions require a lag parameter characterizing the pressure rise response lag. The response lag parameter which enabled the models to most accurately predict the open-loop compressor dynamics was consistent with published values. The open-loop modal dynamics were determined experimentally and could be accurately described by the dynamics represented in the simplest model.; The open-loop dynamics were measured for the first three modes over a range of stable and formerly unstable operating points. The more elaborate models gave accurate predictions of the closed-loop compressor performance, and the most accurate one predicted the flow range extension with less than 1.5% error. June 1993; Includes bibliographical references (pages 183-185)

Control of the unsteady flow in a stator blade row interacting with upstream moving wakes

Fri, 01 Jan 1993 00:00:00 GMT

Control of the unsteady flow in a stator blade row interacting with upstream moving wakes Valkov, Theodore V. (Theodore Valkov) A computational study of the unsteady flow in a 2-D stator blade row interacting with upstream rotor wakes has been carried out. A direct spectral-element Navier-Stokes solver has been used for the laminar flow regime (Re<10,000). Turbulent calculations (Re>106 are based on the Baldwin-Lomax turbulence model. The rotor wakes are represented by velocity distortions moving along the inlet boundary of the computational domain. After interception, the rotor wake migrates towards the pressure surface of the stator blades where it forms a pair of counter-rotating vortices. A moving series of such vortex pairs is the dominant form of unsteady flow over the pressure surface. The unsteady flow over the suction surface is characterized by a street of co-rotating vortices, produced in the leading edge region. These vortices consist of boundary layer fluid distorted and detached by the passing wakes.; Downstream of the leading edge, each of these vortices induces an associated, opposite-sign vortex. The blade loading fluctuations arising from wake interaction, are of two kinds. First, a strong pressure pulse occurs on the leading edge upon wake interception. This pulse is a potential flow effect associated with the excess tangential velocity in the wake. Second, a moving pattern of pressure fluctuations, associated with the vortices, is present over the blade surface. The pressure fluctuations are negative on the suction surface, and positive on the pressure surface. The unsteady flow features over the suction surface can be adequately represented by linearized perturbation calculations, where the disturbance flow associated with the wakes is linearized about a steady viscous flow. Three parameters influence the unsteady flow over the suction surface-stator blade loading, excess wake momentum in the stator frame, and wake reduced frequency.; The strength of the disturbance flow vortices is directly proportional to the wake momentum and decreases at higher reduced frequencies. An adverse pressure gradient results in stronger vortices and pressure fluctuations. On the pressure surface, the amount of unsteady flow depends on the excess wake momentum only. Strategies for controlling the unsteady flow are simulated using appropriate blade surface boundary conditions. Fluid removal from the suction surface prevents formation of vortices and reduces the associated loading disturbances. Fluid injection from the pressure surface reduces the pressure fluctuations there. January 1993; Includes bibliographical references (pages 63-66)

Feedback stabilization of compression systems

Fri, 01 Jan 1993 00:00:00 GMT

Feedback stabilization of compression systems Simon, Jonathan S. (Jonathan Seth) An experimental and analytical study was conducted on the use of feedback stabilization to extend the stable operating range of compression and pumping systems. Prevention of surge, in aircraft gas turbine engines, an instability characterized by violent system wide oscillations in pressure, and flow, primarily motivated this research. However, such instabilities arise as a result of the non-monotonic pressure/flow relations of the axial and/or centrifugal compressors used in gas turbine engines and the results of this research can therefore be applied to many other compression and pumping systems which also employ these types of devices. The frequency response of a centrifugal compressor was measured to validate a lumped parameter model of this component. The results showed that the model is useful for control design and analysis purposes.; Modeling the flow development process as a first order lag was shown to account for observed phenomena that were not predicted by a quasi-steady model. Theoretical limitations to the control of this class of systems that are imposed by bounded actuation and stability robustness requirements were characterized and quantified. The limitations are shown to depend strongly on the choice of actuator and sensor. The influence of actuator/sensor selection was further defined by systematically evaluating a diverse range of available options. The practical difficulty, as well as the number of viable options, were found to depend primarily on the slope of the compressor pressure rise characteristic and the system B-parameter, which represents the relative amount of fluid compliance to inertia.; An important result is that only options which can be considered dynamically close-coupled to the compressor are viable for systems with B-parameter and non-dimensional slope substantially greater than unity. One particular close-coupled scheme, a control valve at the compressor exit with compressor mass flow feedback, was examined in detail and experimentally demonstrated. Although the system of interest can also be stabilized with a passive flow restriction, it is shown that the use of feedback control reduces the steady state pressure loss across the valve required for stabilization. The benefits are greatest for systems with relatively steep slopes at high flow. Includes bibliographical references (pages 244-250)

Computational simulation of unsteady flow in transonic compressor rotor

Wed, 01 Jan 1986 00:00:00 GMT

Computational simulation of unsteady flow in transonic compressor rotor Owen, Philip Ray The unsteady flow about a section of a modern first stage transonic compressor rotor was simulated using a finite difference approximation to the two-dimensional, Reynolds averaged, unsteady, compressible, viscous Navier-Stokes equations. The computation was performed in both steady state and time-accurate modes, and the results compared. The time-accurate results were analyzed in some detail. Two frequency regimes were observed. High frequency unsteadiness due to vortex shedding was found at frequencies varying between 11 KHz and 19 KHz. A low frequency cycle was also observed at 365 Hz. The low frequency cycle produced significant variations in blade force and moment. It also modulated the strength and frequency of the vortex shedding. Arguments were advanced to explain the mechanics of the vortex street formation in terms of a single free shear layer instability. The variations in shedding strength and frequency were related to movement of the separation point. A wholly satisfactory normalization of the frequencies was not found. The low frequency cycle was analyzed as a quasi-steady sequence of events stemming from movement of a shock wave spanning the blade passage. The possibility was entertained that the cycle was due to purely numerical sources, but no likely mechanism was found. October 1986; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1986; Includes bibliographical references (pages 57-58)

Aerodynamics of aircraft engines : stride and stumbles

Wed, 01 Jan 1992 00:00:00 GMT

Aerodynamics of aircraft engines : stride and stumbles Cumpsty, N. A. Summary: Attempts to understand and predict the aerodynamic behaviour of compressors and turbines in aircraft gas turbines have been encouraged by the intense competitive pressure which exists. Many of the apparently most difficult problems have been overcome using suitable numerical analysis, for example the calculation of three-dimensional transonic flows has been particularly successful. It is seemingly paradoxical that the numerical methods are relatively very good for flows which are traditionally regarded as difficult, but do less well at predicting efficiency when the flow is well behaved in the conventional aerodynamical sense, such as fully attached flows. The numerical methods do not necessarily give insight into the flow that is seen as most helpful to the designer and it can be useful to complement them with simpler approaches to the problem which seek to capture the essential features of the flow. The aerodynamics of aircraft engine fans are used to illustrate these points. Although numerical methods have been very successful with aeroengine fan blading they have been less successful with the multistage compressor. The reasons for this, primarily the difficulty of prescribing the boundary conditions, are discussed in this paper. The analysis of flow in multistage compressors still stumbles along with empirical methods, much of it based on data published over twenty years ago. A second area where stumbling has occurred is the prediction of flutter of blading, particularly fan blading; as recently as 1990 there were major in-flight failures due to flutter of a fan in civil airline service. To this day there is no reliable method of predicting the operating boundaries of flutter, and testing the engine over the entire operating range of altitude and speed is the only reliable method of ensuring safe operation. September 1992; Includes bibliographical references (leaves 21-22)

A numerical analysis of 3-D inviscid stator/rotor interactions using non-reflecting boundary conditions

Wed, 01 Jan 1992 00:00:00 GMT

An analytical and numerical study of the second-order effects of unsteadiness on the performance of turbomachines

Wed, 01 Jan 1992 00:00:00 GMT

An analytical and numerical study of the second-order effects of unsteadiness on the performance of turbomachines Fritsch, Gerd A linear approach in two dimensions is used to investigate the second-order effects of unsteadiness on the efficiency of turbomachines. The three main themes are the identification of physical nature and location of unsteady loss mechanisms, the magnitude of the associated losses and their effect on the time-mean efficiency, and the assessment of the modeling accuracy of numerical simulations with respect to unsteady loss. A mathematically rigorous link is established between linear waves in a compressible, two-dimensional flow and the efficiency drop associated with their dissipation. The analysis is applied to the mixing loss at the interface in a steady simulation of rotor/stator interaction in a turbine and to the study of unsteady loss mechanisms. Two unsteady loss mechanisms are considered. Unsteady Circulation Loss, i.e.; the transfer of mean-flow energy to kinetic energy associated with vorticity shed at the trailing edge in response to an unsteady circulation, was first considered by Keller (1935) and later by Kemp and Sears (1955). Keller's original work is extended to compressible, homentropic flows. The use of simulations to obtain circulation amplitudes avoids the limitations of thin-airfoil theory and yields a loss measure realistic for modern turbomachines. For the Unsteady Viscous Loss mechanism, i.e. the dissipation induced by pressure waves in unsteady boundary layers, the high-reduced-frequency limit and a near-wall approximation are used to obtain the local velocity distribution in the laminar Stokes sublayer and the corresponding time-mean dissipation. The input to the model are the unsteady pressure gradients along a blade surface obtained from an unsteady simulation. A numerical study of the errors due to modeling approximation is included.; Both sources of loss are small but not negligible. It is found that numerical smoothing shifts the principal locus of unsteady dissipation from boundary layers to the freestream, reducing the magnitude of the loss models input and the predicted loss. June 1992; Based on the author's Sc. D. thesis, Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1992; Includes bibliographical references (pages 147-152)

Experimental investigation of flow distortion effects on the performance of radial discrete-passage diffusers

Tue, 01 Jan 1991 00:00:00 GMT

Experimental investigation of flow distortion effects on the performance of radial discrete-passage diffusers Filipenco, Victor Gregory A swirling-radial-flow generator has been developed for the study of fluid-dynamic phenomena in radial passage- and vaneless-diffusers. A unique feature of the swirling-flow generator is the capability of providing a wide range of diffuser inlet flow conditions. This is accomplished by means of a very-high-solidity rotating radial-outflow nozzle cascade in conjunction with annular cross-flow injection/suction slots in the flow-path walls immediately upstream and downstream of the rotor blading. The rotor generates a shockless and weak-wake axisymmetric transonic swirling flow which can be tailored to provide a desired level of diffuser inlet flow-field axial distortion by means of cross-flow injection and/or suction through the annular slots.; A complete test-facility was designed and constructed based on this concept and was utilized to study effects of inlet flow-field axial distortion on the pressure-recovery performance and stability of a modem high-performance gas-turbine-engine radial discrete-passage diffuser. It was shown that the diffuser pressure-recovery coefficient, if based on the inlet availability-averaged total pressure, correlates well with the diffuser inlet momentum-averaged flow angle independent of flow-field axial distortion and Mach number over the wide flow parameter range investigated. It was argued that the generally accepted high sensitivity of diffuser pressure recovery performance to inlet flow distortion and boundary-layer blockage is largely due to inappropriate quantification of the diffuser inlet flow-field parameters.; Time resolved pressure measurements in the vaneless space between the rotor and diffuser showed that the diffuser operating range is limited by the onset of rotating stall triggered by the loss of flow stability in the diffuser, independent of the rotor operating point (if overall compression system instability did not occur first). It was found that the loss of flow stability in the diffuser occurred at a critical value of the diffuser inlet momentum-averaged flow angle and corresponding overall-diffuser pressure recovery coefficient (based on the availability-averaged total pressure), independent of inlet flow-field distortion and Mach number, over the wide flow parameter range investigated.; A simple analytical consideration of an idealized diffuser consisting of a constant-area mixing duct followed by an isentropic-flow diffuser was used to show that the observed insensitivity of the diffuser pressure-recovery performance to inlet flow-field distortion can be attributed to rapid mixing of the flow. It was shown that measurements of the static pressure distribution along the centerline of an individual diffuser-passage support this hypothesis. September 1991; Based on the author's Ph. D. thesis, Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1991; Includes bibliographical references (pages 185-189)

UNSFLO : a numerical method for the calculation of unsteady flow in turbomachinery

Tue, 01 Jan 1991 00:00:00 GMT

UNSFLO : a numerical method for the calculation of unsteady flow in turbomachinery Giles, M. (Michael) Introduction: There are four principal sources of unsteadiness in a single stage of a turbomachine in which there is one row of stationary blades (stators) and one row of moving blades (rotors). As shown in Fig. 1.1, wake/rotor interaction causes unsteadiness because the stator wakes, which one can consider to be approximately steady in the stator frame of reference, are unsteady in the rotor frame of reference since the rotor is moving through the wakes and chopping them into pieces. This causes unsteady forces on the rotor blades and generates unsteady pressure waves. Although the stator wakes are generated by viscosity, the subsequent interaction with the rotor blades is primarily an inviscid process and so can be modelled by the inviscid equations of motion. This allows two different approaches in numerical modelling. The first is to perform a full unsteady Navier-Stokes calculation of the stator and rotor blades.; The second is to perform an unsteady inviscid calculation for just the rotor blade row, with the wakes being somehow specified as unsteady inflow boundary conditions. This latter approach is computationally much more efficient, but assumes that one is not concerned about the unsteady heat transfer and other viscous effects on the rotor blades. Potential stator/rotor interaction causes unsteadiness due to the fact that the pressure in the region between the stator and rotor blade rows can be decomposed approximately into a part that is steady and uniform, a part that is non-uniform but steady in the rotor frame (due to the lift on the rotor blades) and a part that is non-uniform but steady in the stator frame (due to the lift on the stator blades).; As the rotor blades move, the stator trailing edges experience an unsteady pressure due to the non-uniform part that is locked to the rotors, and the rotor leading edges experience an unsteady pressure due to the non-uniform part that is locked to the stators. This is a purely inviscid interaction which is why it is labelled a "potential" interaction. There are again two approaches to modelling this interaction. The first is an unsteady, inviscid calculation of the stator and rotor blade rows. The second is an unsteady, inviscid calculation of just one of the blade rows, either the stator or the rotor, with the unsteady pressure being specified as a boundary condition. The latter approach is more efficient, but unfortunately the situation in which the potential stator/rotor interaction becomes important is when the spacing between the stator and rotor rows is extremely small, and/or there are shock waves moving in the region between them.; Consequently, one does-not usually know what values to specify as unsteady boundary conditions. The first two sources of unsteadiness were both due to the relative motion of the stator and rotor rows. The remaining two sources are not. The viscous flow past a blunt turbine trailing edge results in vortex shedding, very similar to the Karman vortex street shed behind a cylinder. In fact real wakes lie somewhere between the two idealized limits of a Karman vortex street and a turbulent wake with steady mean velocity profile. It is believed that provided the integrated loss is identical the choice of model does not affect the subsequent interaction with the downstream rotor blade row. However, this is an assumption which needs to be investigated sometime in the future. The importance of vortex shedding lies in the calculation of the average pressure around the blunt trailing edge, which determines the base pressure loss, a significant component of the overall loss.; There is also experimental evidence to suggest that the vortex shedding can be greatly amplified under some conditions by the potential stator/rotor interaction. Finally, there can be unsteadiness due to the motion of the stator or rotor blades. The primary concern here is the avoidance of flutter. This is a condition in which a small oscillation of the blade produces an unsteady force and moment on the blade which due to its phase relationship to the motion does work on the blade and so increases the amplitude of the blade's unsteady motion. This can rapidly lead to very large amplitude blade vibrations, and ultimately blade failure. May 1991; Includes bibliographical references (pages 89-91)

Newton solution of steady two-dimensional transonic flow

Tue, 01 Jan 1985 00:00:00 GMT

Newton solution of steady two-dimensional transonic flow Giles, M. (Michael) A new method is developed for the solution of the steady, two-dimensional Euler equations for transonic flows. The discrete steady-state equations are derived in conservative finite-volume form on an intrinsic streamline grid, and are solved using Newton's method. Direct solution of the linear system of Newton equations is shown to be more efficient than iterative solution. Test cases include duct, cascade, and isolated airfoil flows, and demonstrate the speed and robustness of the method. The accuracy of the solutions is verified by comparison against values obtained analytically, experimentally and by other numerical methods. October 1985; Also issued as: Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1985; Includes bibliographical references (pages 167-169)

Analysis and design of transonic cascades with splitter vanes

Tue, 01 Jan 1991 00:00:00 GMT

Analysis and design of transonic cascades with splitter vanes Youngren, H. H. (Harold Hayes) A new computational method, MISES, is developed for turbomachinery design and analysis applications. The method is based on the fully coupled viscous /inviscid method, ISES, and is applicable to blade-to-blade analysis of axial fan and compressor stator or rotors with optional splitter vanes. Quasi-three dimensional effects for stream surface radius, streamtube thickness and wheel rotation may be included. The flow is modeled with the steady Euler equations and the integral boundary layer equations. A robust Newton-Raphson method is used to solve the coupled non-linear system of equations, requiring only several minutes for solution on a typical workstation. Design options are implemented for either single surface or camber redesign. The method is exercised by comparison with transonic cascade tests to validate the quasi-three-dimensional formulation. The results show excellent correlation to measured pressure distributions and loss levels. The multiple blade capability is demonstrated by comparison to test data for a supersonic cascade with splitter vane. New splitter vane configurations for improving the performance of the supersonic cascade are explored, resulting in large increases in turning and reduced loss. Includes bibliographical references (pages 162-164)

Condensation of supersaturated organic vapors in a supersonic nozzle

Sun, 01 Jan 1967 00:00:00 GMT

Condensation of supersaturated organic vapors in a supersonic nozzle Dawson, Daniel Bogert Experiments were performed involving condensation of supersaturated benzene and chloroform vapors in a supersonic nozzle, with compressed air as the carrier gas. Experiments showed that the magnitude of the water vapor content of the carrier air made no observable difference in the condensation behavior of either fluid. It was demonstrated that addition of small amounts of these fluids to the carrier air tended to reduce the thickness of the boundary layer in the nozzle. Comparison of experimental results with theory show, without making any adjustments to physical properties of condensate droplets to account for size, that incidence of condensation for chloroform can be predicted by the revised theory of nucleation, whereas benzene incidence can be predicted by neither revised nor classical theory. These results, combined with prior data on other fluids, show that at present neither theory seems to be generally applicable. In support of previous conclusions, the problem may well be the assumption that bulk properties may be assigned to small (30 - 50 molecules) droplets of condensate. (Author). April 1967; Includes bibliographical references

Blade scale effects of tip leakage

Mon, 01 Jan 1990 00:00:00 GMT

Blade scale effects of tip leakage Martinez-Sanchez, Manuel; Gauthier, R. P. The effects of blade-tip leakage in a turbine are investigated by modeling the stage as an incomplete actuator disk. It is found that the spanwise flow redistribution due to the gap is such as to produce a uniform unloading of the blades, despite the very concentrated leakage. Partial lift retention at the blade tip is accounted for based on a leakage jet-free stream collision model which successfully predicts the roll-up of the leakage flow. The predicted efficiency loss due to the gap correlates well with experimental data. October 1990; Includes bibliographical references (leaves 43-44)

A systematic study of supersonic jet noise

Fri, 01 Jan 1971 00:00:00 GMT

A systematic study of supersonic jet noise Louis, Jean F. (Jean François); Letty, Richard M.; Patel, Jayantilal R. The purpose of this work is to study the acoustic fields associated with two different nozzle configurations; a rectangular and a circular. Both nozzles are designed with the same exit Mach number and have an identical momentum and energy flux. By presenting a comparison of the two nozzles, it proposed to establish and identify the dominant noise generating mechanisms. A basic difference in shape changes the relative importance of different noise mechanisms. The other main aim of this study is to establish scaling laws of supersonic jet noise. A shock tube is a very versatile apparatus for such an analysis. By first changing the driver, driven pressure and molecular weights, a wide range of stagnation pressures and temperatures could be achieved. The case with which these conditions are simulated is, however, traded off with the short test time, of the order of milliseconds. A short test time allows the use of a heat sink nozzle and eliminates the use of an anechoic chamber.; So far tests have been made in the range of 1000-5000°R, for different levels of expansion and an exit Mach number of 2.7. In comparing the two nozzles, it is found that the rectangular nozzle is indeed quieter than the circular nozzle. The rectangular nozzle is more effective under overexpanded conditions, and a factor of 1.6 in acoustic efficiency at low temperature (1200°R) and about 3 at high temperature is related to a rapid deceleration of the jet through a system of strong shocks. The low acoustic efficiency of the overexpanded rectangular jet is related to a rapid deceleration of the jet through a system of strong shocks. At high temperature, this effect is not observed because an important density ratio exists across the shear layer which becomes very unstable due to the Taylor instability.; For both the circular and rectangular nozzle, the effect of temperature showed an increase in the directivity angle at high temperature which is correlated to an increase in eddy convective velocity, rather than refraction due to density gradients, which seems to play a secondary role. The low temperature overexpanded jet showed a difference of about 2.6 db in the OPWL between the two nozzles. However, at this condition, for the rectangular nozzle, a difference of 8 db between the maximum and minimum noise direction is observed. Hence, a suitable orientation of the nozzle could cause a considerable reduction in the noise level. The rectangular nozzle seems to be very effective under overexpanded conditions. The scaling laws, which are in the preliminary stages, were developed for the change in the OPWL as a function of stagnation pressure. For the circular nozzle, additional scaling was done for temperature and acoustic efficiency.; These scaling laws need to be verified for additional temperatures. Also, further work should be initiated in the potential use of the rectangular nozzle as a noise suppressor and as a model for better comprehension of noise generating mechanisms. December 1971; Includes bibliographical references (leaves 39-40)

Separation of turbulent, incompressible flow from a curved, backward-facing step

Sat, 01 Jan 1966 00:00:00 GMT

Separation of turbulent, incompressible flow from a curved, backward-facing step Nice, George Roland; Tseng, W.-Y. (Wu-Yang); Moses, Hal Lynwood An experimental investigation of turbulent, incompressible flow separation over curved and sharp, backward-facing steps is presented with results for various step heights. Mean velocities in the separating boundary layer as well as the downstream shear layer were recorded. The static pressure in the separated region was determined with a spherical probe. With the curved step, the boundary layer separated at approximately 28 degrees: the reattachment lengths were somewhat less and the base pressures slightly higher than those with the sharp step. October 1966; Includes bibliographical references

Unsteady radial transport in a transonic compressor stage

Sun, 01 Jan 1989 00:00:00 GMT

Unsteady radial transport in a transonic compressor stage Kotidis, Petros Anestis A technique, based on the operation of a dual-hot wire-aspirating probe with frequency response of at least 18 kHz and uncertainty less than 0.005 to full scale in mass fraction units, has been developed to measure time resolved concentration in unsteady, compressible flows. The goal of the experimental part of this research effort was to obtain time resolved measurements of spanwise transport in a transonic compressor. This was achieved by injecting a circumferentially oriented, thin sheet of tracer gas in front of the compressor and detecting the foreign fluid molecules at the rotor exit The experiments were conducted at the MIT Blowdown Facility using the Air Force High Through Flow Compressor Stage as the test article. During a preliminary data reduction procedure, it was discovered that the signals from the probe's hot wires lag in time with respect to the signal from the companion total pressure probe.; A correction incorporated in the data reduction schemes to account for this, eliminated most of the negative entropy regions observed in previous experiments with this probe. Several conclusions have been drawn from the experimental observations. First, up to 5% of the compressor mass flow moved along the blade span. Second, the migrating fluid was found primarily in the blade wakes at the measurement location. Third, this fluid moved towards both hub and tip in the blade wakes. Fourth, the radially convected fluid had high entropy, much higher than that of the average flowfield. Fifth, the "inviscid core" fluid moves preferentially towards the suction side of the blade passage and away from the pressure side. A simple model was developed to explain the spanwise fluid transport Gertz's 2-D wake vortex street model was extended into a quasi 3-D form.; The 2-D model was fitted to the data at four spanwise locations and the spanwise variation of the parameters of the vortex street (such as vortex strength and core size) were determined. The model fit showed the shedding frequencies to be the same [17 (+/-) 0.4 kHz] at all four spanwise locations, suggesting that the vortex shedding is coherent along the span. The spanwise pressure gradient created by the variation of vortex strength led to substantial spanwise transport in the vortex cores. The model predicted the transport to the hub, but underestimated the transport to the tip by a factor of five. The measured spanwise transport can explain the previously observed discrepancy between predicted (viscous+normal shock losses) and measured spanwise distributions of adiabatic efficiency in the tip regions of transonic compressors (assuming the radial outflow to be primarily in regions of separated flow on the blades). September 1989; Includes bibliographical references (pages 212-219)

Induced strain actuation of composite plates

Sun, 01 Jan 1989 00:00:00 GMT

Induced strain actuation of composite plates Lazarus, Kenneth B.; Crawley, Edward F. Two models of induced strain plate actuator/substrate systems are developed and verified experimentally. Equations relating the actuation strains produced by the strain actuators to the induced strains found in the system are derived for both models. In addition, the plate strain energy relations are also developed. Exact and approximate solutions are formulated for isotropic and anisotropic plate systems. Exact solutions are found for actuator/substrate systems with free-free, free-free boundary conditions, and a general procedure for solving the strain energy equations with a Rayleigh-Ritz approximate solution is formulated for systems with arbitrary boundary conditions and external loads. Specific solutions are detailed for cantilever plate systems, including a discussion of the assumed modes selected. A model for predicting the actuation strains produced by a specific class of induced strain actuators, piezoceramic actuators, is also developed.; The non-linear properties of piezoceramics are discussed and the important effects of such non-linearities are accounted for by developing a strain dependent model for the actuation strains created in piezoceramics. The models developed were verified through experimentation via two sets of plate test articles. The first set of simple test articles were used to verify the accuracy of the basic induced strain actuation models, the strain dependence of piezoceramic actuation strains, and a semi-empirical solution procedure. The second, more representative, set of large cantilever plate test articles verified the ability of the models to predict the strains induced in systems with extensive stiffness couplings and complicated boundary conditions, and the Ritz model.; Agreement between the solutions predicted by the induced strain actuator models and the experimentally measured deformations was excellent, verifying the effectiveness of using induced strain actuation for shape control of structures such as aeroelastic lifting surfaces and components of intelligent structures. March 1989; Originally written by Mr. Lazarus as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1989; Includes bibliographical references (pages 126-128)

Surge dynamics and unsteady flow phenomena in centrifugal compressors

Fri, 01 Jan 1988 00:00:00 GMT

Surge dynamics and unsteady flow phenomena in centrifugal compressors Fink, David Allan Detailed time resolved measurements of centrifugal compressor surge has been obtained on an automotive turbocharger for two very different compression systems one with a large downstream volume and one with a much smaller downstream volume. These measurements show impeller stall at the inducer tips to be a key phenomena in initiating surge. The inducer tip stall, which is dominant over other types of stall in the compressor, is observed to be non-rotating and asymmetric due to the presence of an asymmetric downstream volute. The most severe stalling of the impeller occurs at a circumferential position nearest the volute tongue position and is due to a circumferential flow distortion set up by the . volute. The vaneless diffuser is seen to be destabilizing but does not initiate surge by abrupt stalling. Rotating stall was found to be unimportant in surge initiation. New evidence is presented concerning the dynamic behavior of the compressor characteristics in surge operation.; Instantaneous compressor characteristics in surge when operating in a large volume(large B-parameter) system are found to be flatter than the time averaged ones for a small volume(small B-parameter) stabilized system. A physical mechanism accounting for the difference between the two measured characteristics is the slow development time and differing circumferential extent of the inducer stall present. The flatness of the large B characteristic contrasts with the characteristic of an axial. compressor operating in surge and leads to slow growth of the surge massflow instability. A dynamic model has been developed which includes effects of speed variations, compressibility, and time lags. The inclusion of speed variations changes the time domain behavior of the compression system substantially from the results obtained with constant speed.; A precursor period of mild surge whose length depends on the amount of throttling is shown to be present before deep surge and is due to the speed variations. Both speed variations and time lags in compressor behavior are shown to introduce a stabilizing effect on compressor behavior in mild surge. The results of this model agree qualitatively as well as quantitatively with the measured experimental system dynamic behavior. The dynamic behavior observed has some properties of importance for the new field of active control of surge instabilities in centrifugal compressors. The flatness of the instantaneous compressor characteristic, the existence of a mild surge precursor period, and slow growth of the surge instability are favorable conditions for a relatively simple active control strategy to stabilize the compression system and eliminate surge. Also the dynamic model developed may be useful for exploring alternative active control strategies. June 1988; Includes bibliographical references

A coupled Euler/Navier-Stokes algorithm for 2-D unsteady transonic shock/boundary-layer interaction

Sun, 01 Jan 1989 00:00:00 GMT

A coupled Euler/Navier-Stokes algorithm for 2-D unsteady transonic shock/boundary-layer interaction Allmaras, Steven R. (Steven Richard) This thesis presents a coupled Euler/Navier-Stokes algorithm for solving 2-D unsteady transonic flows. The flowfield is described by a Defect formulation, where separate Euler and Navier-Stokes algorithms are used on overlapping grids and are coupled through wall transpiration fluxes. The work is separated into three major contributions. The first contribution is a new algorithm for the solution of the 2-D unsteady Euler equations. The algorithm incorporates flux-splitting to capture shocks crisply and with minimal oscillations. To reduce numerical errors, grid independent second order accuracy is achieved for both steady and unsteady flows. This is done by a formulation in which both solution averages and gradients are stored for each cell. The scheme allows no decoupled modes; hence, no explicitly added artificial is necessary. The second contribution is a Thin-Shear-Layer Navier-Stokes algorithm for viscous regions.; The algorithm uses two-point differencing across the boundary layer, which is second order accurate for both inviscid and viscous terms on nonsmooth grids. Fluxsplitting is used for the streamwise discretization to capture shocks. The spatial discretization of this scheme also admits no decoupled modes and does not require added artificial dissipation. A semi-implicit time integration is employed, which allows a time step determined by the streamwise grid spacing only. The algorithm uses a dynamic coordinate rescaling to evolve the viscous grid to the changing boundary layer thickness. The final contribution of the work is a explicit relaxation procedure for coupling the Euler and Navier-Stokes algorithms together. The coupling is through boundary conditions-specified outer edge values for the viscous solution and wall transpiration fluxes for the Euler solution. Computational results are presented for a series of duct geometries.; The test cases are used to demonstrate the accuracy of the Euler algorithm, the Navier-Stokes algorithm, and the fully coupled algorithm. Results are compared with analytic theory, experimental results, and other computational methods. March 1989; Originally issued as: Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1989; Includes bibliographical references (pages 238-243)

Effects of compressor endwall suction and blowing on stability enhancement

Fri, 01 Jan 1988 00:00:00 GMT

Effects of compressor endwall suction and blowing on stability enhancement Lee, Norman Kai Wing An experimental investigation was carried out to examine the effects on stall margin of flow injection into, and removal out of, the endwall region of an axial compressor blade row. A main goal was to identify the mechanism by which wall treatment suppresses stall in turbomachines. To simulate the relative motion between blade and treatment, the injection and removal took place through a slotted hub rotating beneath a cantilevered stator row. Overall performance data and detailed (time-averaged) flow-field measurements were obtained. Both injection and removal increased the stalling pressure rise, but neither was as effective as the complete treatment. This implies that removal of high blockage flow is not the only reason behind the observed stall margin improvement in a casing or hub treatment, and that injection also contributes to stall suppression. The results also indicated that, for a given variation of injection, the increase in stall pressure rise is linked to the streamwise momentum of the injected flow. January 1988; Includes bibliographical references

UNSFLO : a numerical method for unsteady inviscid flow in turbomachinery

Fri, 01 Jan 1988 00:00:00 GMT

UNSFLO : a numerical method for unsteady inviscid flow in turbomachinery Giles, M. (Michael) Introduction: There are four principal sources of unsteadiness in a single stage of a turbomachine in which there is one row of stationary blades (stators) and one row of moving blades (rotors). As shown in Fig. 1.1, wake/rotor interaction causes unsteadiness because the stator wakes, which one can assume to be approximately steady in the stator frame of reference, are unsteady in the rotor frame of reference since the rotor is moving through the wakes and chopping them into pieces. This causes unsteady forces on the rotor blades and generates unsteady pressure waves. Although the stator wakes are generated by viscosity, the subsequent interaction with the rotor blades is primarily an inviscid process and so can be modelled by the inviscid equations of motion. This allows two different approaches in numerical modelling. The first is to perform a full unsteady Navier-Stokes calculation of the stator and rotor blades.; The second is to perform an unsteady inviscid calculation for just the rotor blade row, with the wakes being somehow specified as unsteady inflow boundary conditions. This latter approach is computationally much more efficient, but assumes that one is not concerned about the unsteady heat transfer and other viscous effects on the rotor blades. Potential stator/rotor interaction causes unsteadiness due to the fact that the pressure in the region between the stator and rotor blade rows can be decomposed approximately into a part that is steady and uniform, a part that is non-uniform but steady in the rotor frame (due to the lift on the rotor blades) and a part that is non-uniform but steady in the stator frame (due to the lift on the stator blades).; As the rotor blades move, the stator trailing edges experience an unsteady pressure due to the non-uniform part that is locked to the rotors, and the rotor leading edges experience an unsteady pressure due to the non-uniform part that is locked to the stators. This is a purely inviscid interaction which is why it is labelled a "potential" interaction. There are again two approaches to modelling this interaction. The first is an unsteady, inviscid calculation of the stator and rotor blade rows. The second is an unsteady, inviscid calculation of just one of the blade rows, either the stator or the rotor, with the unsteady pressure being specified as a boundary condition. The latter approach is more efficient, but unfortunately the situation in which the potential stator/rotor interaction becomes important is when the spacing between the stator and rotor rows is extremely small, and/or there are shock waves moving in the region between them.; Consequently, one does not usually know what values to specify as unsteady boundary conditions. The first two sources of unsteadiness were both due to the relative motion of the stator and rotor rows. The remaining two sources are not. The viscous flow past a blunt turbine trailing edge results in vortex shedding, very similar to the Karman vortex street shed behind a cylinder. In fact real wakes lie somewhere between the two idealized limits of a Karman vortex street and a turbulent wake with steady mean velocity profile. It is believed that provided the integrated loss is identical the choice of model does not affect the subsequent interaction with the downstream rotor blade row. However, this is an assumption which needs to be investigated sometime in the future. The importance of vortex shedding lies in the calculation of the average pressure around the blunt trailing edge, which determines the base pressure loss, a significant component of the overall loss.; There is also experimental evidence to suggest that the vortex shedding can be greatly amplified under some conditions by the potential stator/rotor interaction. Finally, there can be unsteadiness due to the motion of the stator or rotor blades. The primary concern here is the avoidance of flutter. This is a condition in which a small oscillation of the blade produces an unsteady force and moment on the blade which due to its phase relationship to the motion does work on the blade and so increases the amplitude of the blade's unsteady motion. This can rapidly lead to very large amplitude blade vibrations, and ultimately blade failure. October 1988; Includes bibliographical references

Two-dimensional transonic aerodynamic design and analysis using the Euler equations

Wed, 01 Jan 1986 00:00:00 GMT

Two-dimensional transonic aerodynamic design and analysis using the Euler equations Drela, Mark A method is developed for the solution of the steady two-dimensional Euler equations with viscous corrections for transonic design and analysis problems. The steady finite volume integral equations are formulated on an intrinsic streamline grid, and are solved using a global Newton method. Conservative differencing together with artificial bulk viscosity in supersonic regions permit correct shock capturing. The design capability of the method stems from the streamline-based grid and Newton solution method, which allow both direct and inverse boundary conditions and constraints to be readily applied to the governing equations. For all boundary condition types, the effects of boundary layers and wakes on the inviscid flow are modeled by the displacement thickness concept. The boundary layer and wake parameters are described by compressible integral boundary layer equations which are coupled to the inviscid flow and are included in the global Newton solution scheme. This coupling procedure gives stable convergence for flows with limited separation regions. A transition criterion based on the Orr-Sommerfeld equation is developed and applied to transitional separation bubbles. Accurate drag predictions are obtained for subsonic and shocked transonic airfoils. Design examples involving airfoils and cascades are presented. February 1986; Originally presented as the author's thesis (doctoral-Massachusetts Institute of Technology)--under the title: Two-dimensional transonic aerodynamic design and analysis using the Euler equations; Includes bibliographical references (pages 139-143)

Experimental investigation of Stator hub treatment in an axial flow compressor

Thu, 01 Jan 1981 00:00:00 GMT

Experimental investigation of Stator hub treatment in an axial flow compressor Prell, Mark E. An experiment was carried out to examine the effects, on stator stall margin and performance, of a slotted hub treatment rotating beneath the stator of an axial flow compressor. The compressor was run with this hub treatment and the results compared to those taken with a smooth rotating hub. It was determined that, for the configuration tested, the hub treatment was ineffective in the improvement of stall margin but resulted in a measurably higher static pressure rise across the stator and a significant decrease in flow deviation and blockage in the stator midspan region. Although it is the hub section of the stator that sets the stall limit in this configuration, measurements of the stator exit flow field indicated that the type of stall that is occurring is a blade stall rather than a pure wall stall. Absence of a wall stall is thus seen as a key possibility for the lack of stall margin improvement. July 1981; Includes bibliographical references (page 29)

Measurements of aerodynamic damping on the MIT transonic rotor

Thu, 01 Jan 1981 00:00:00 GMT

Measurements of aerodynamic damping on the MIT transonic rotor Crawley, Edward F. A method has been developed and demonstrated for the direct measurement of aerodynamic forcing and aerodynamic damping of a transonic compressor. The method is based on the inverse solution of the structural dynamic equations of motion of the blade disk system in order to determine the forces acting on the system. The disturbing and damping forces acting on a given blade are determined if the equations of motion are expressed in individual blade coordinates. If the structural dynamic equations are transformed to multi blade coordinates, the damping can be measured for blade-disk modes, and related to a reduced frequency and inter blade phase angle. To measure the aerodynamic damping in this way, the free response to a known excitation is studied. This method of force determination was demonstrated using a specially instrumented version of the MIT Transonic Compressor run in the MIT Blowdown Compressor Test Facility.; Unique on-rotor instrumentation included piezoelectric displacement transducers to monitor the displacement of each blade, three accelerometers to measure in plane motion of the disk and a leading edge mounted total pressure transducer. Resonance tests performed prior to installation of the rotor in the tunnel indicate that the blade disk structural interaction is dominated by the rigid body inertial coupling of the disk. An analytical model was developed for this inertial coupling. The model was verified by extensive testing of the tuned and severely mistuned rotor. No regions of aeroelastic instability were found while testing the rotor in the Blowdown Facility, but three forms of forced vibration were encountered. When operated in rotating stall, the blades were strongly excited at the fundamental frequency of stall cell excitation and those higher harmonics in proximity to blade resonances.; At the fundamental frequency, the blade bending loading decreased as the blade entered the stall cell and increased as smooth flow was reestablished over the blade. In runs near the operating point, the rotor was aerodynamically excited by a controlled two-per-revolution fixed upstream disturbance. The disturbance was sharply terminated midway through the test and the ring down of the rotor monitored. Analysis of the data in terms of multiblade modes led to a direct measurement of aerodynamic damping for several interblade phase angles. During all runs, the third circumferential harmonic of the blade displacement was strongly excited by wakes shed from three evenly spaced upstream struts. The addition of a two per revolution fixed upstream disturbance caused a marked decrease in the third harmonic response, suggesting a nonlinear mechanism either in the upstream wake production or in the aerodynamic response of the rotor.; It may therefore be possible to alleviate some forced vibrations by the deliberate introduction of upstream disturbances. February 1981; Includes bibliographical references (page 80)

Sound generation from a transonic compressor stage

Thu, 01 Jan 1981 00:00:00 GMT

Sound generation from a transonic compressor stage Schaffner, Joan Elsa; Ingard, K. Uno Some aspects of the sound pressure wave characteristics produced by a transonic compressor stage are studied by the use of a simplified mathematical model involving flow parallel to a moving corrugated board. The critical parameters of the system are correlated with the various unknowns in the mathematical model resulting in a suitable representation of the physical system. The analysis is first performed for the rotor alone, calculating the sound pressure wave upstream and downstream from the rotor, spinning at a tip speed Mach number of 1.2. The model is then modified to account for the stator blade row, 3/4" axially displaced from the rotor. A study of the reflective characteristics of the stator blade row is performed with the result that the reflective characteristic due to straightening of the flow is much more significant than that due to the physical hardware. The complete description for the sound pressure wave in the three regions: upstream of the rotor, downstream of the stator, and between the rotor and the stator, is determined. The results of this analysis are presented in the form of graphs showing: 1) the reflection coefficients, RF and RS , as functions of tip speed, 2) the pressure wave magnitude as a function of rotor tip speed for specific values of flow speed, 3) the axial dependence of the pressure wave magnitude in the region between the rotor and stator, 4) the ratios PO/P1 , PO/P2 , and P1/P2 as functions of Mt . Experimental data obtained prior to this analysis at the M.I.T. Blowdown Facility is presented and shown to correlate quite favorably with the theoretical determination of the pressure ratios presented. July 1981; Includes bibliographical references (pages 64-65)

Flow visualization study of the inlet vortex phenomenon

Thu, 01 Jan 1981 00:00:00 GMT

Flow visualization study of the inlet vortex phenomenon De Siervi, Francesca The inlet vortex phenomenon was experimentally investigated in a water tunnel, using hydrogen bubble flow visualization techniques. Several inlet ambient flow combinations were studied, including an inlet and ground plane configuration in a known shear as well as an irrotational flow, and a double (twin) inlet configuration in an irrotational flow. This latter situation created a boundary layer free analogue of the ground plane and enabled investigation of inlet vortex formation in flow essentially free of ambient vorticity. The three dimensional inlet flow field and the vortex formation mechanisms were determined by marking material lines and observing their path and deformation as they are convected from a far upstream location into the inlet. Two basic mechanisms of inlet vortex generation were found. For flows possessing a vertical component of ambient vorticity, the amplification of this vorticity as the vortex lines are stretched and drawn into the inlet results in the formation of an inlet vortex. However, an inlet vortex does not require the presence of ambient vorticity to form. It can also be created in an irrotational flow, with an inlet in crosswind. In this situation, it is accompanied by a variation in circulation along the axial length of the inlet. The ratio of inlet velocity to upstream velocity is an important parameter in determining the generation of an inlet vortex for both mechanisms. July 1981; Includes bibliographical references (page 99)

Noise and performance of propellers for light aircraft : final report

Tue, 01 Jan 1980 00:00:00 GMT

Noise and performance of propellers for light aircraft : final report Succi, G. P. (George Peter); Larrabee, E. Eugene; Dunbeck, Peter Brian; Munro, David Herbert; Zimmer, Jeffrey Alan; Ingard, K. Uno; Kerrebrock, Jack L. Introduction and Summary: The project "Noise and Performance of Propellers for Light Aircraft," Contract #NASl-15154 between NASA Langley and MIT, has now been completed, and the main results obtained are summarized in this report and its appendices. The primary practical objective of the study was to explore the possibility of reducing the noise from a general aviation type propeller without altering significantly its aerodynamic performance or the engine characteristics. After an extensive study of this question, involving aerodynamic and acoustic theory, design, construction and wind tunnel testing of model propellers, design and manufacturing of full scale propellers and, finally, flight tests, we are pleased to report that for one of the propellers tested an overall reduction of 4.8 dBA as measured in a flight test was achieved.; The theory deals with aerodynamics and acoustics of lightly loaded propellers with subsonic tip speeds and includes studies of the effects of sweeping the blades, altering the radial load distribution, and changing the number of blades. These studies lead to new insight into the general problem of sound generation from moving bodies. Of particular value are the algorithms, which are well suited for computer coding. The wind tunnel tests involved three propellers, 1/4 scale, including a replica of a fixed pitch propeller used on a 150 HP single engine airplane. The other two propellers were designed to have the peak radial load distribution shifted inboard. The acoustic wind tunnel which was used in these tests enabled measurement not only of the radiated sound field but also the thrust and torque of the propeller. In addition, the load distribution was determined indirectly from wake surveys.; Sound pressure signatures were obtained at different locations and speeds (up to a tip Mach number of 0.75) and compared with theoretical predictions in which only the shape and motion of the propeller were needed as input parameters; no empirical adjustments were made. Agreement to within a few percent was obtained throughout except in the presence of a transonic "buzz" instability which was encountered within a narrow speed range. On the basis of the theoretical analysis and its verification in the model tests, a two-bladed fixed pitch propeller was designed for a 150 HP single engine airplane. Flight tests with this propeller indicated about the same performance as the production propeller for that airplane, but the maximum sound level during a full power flyover at 1000 feet was found to be 4.8 dBA lower. A second propeller, with three blades and fixed pitch, was designed for the Ohio State University 180 HP single engine airplane.; Flight tests of this propeller have not yet been made at this time. July 1980; Project Manager: G. P. Succi ; Contributors: E.E. Larrabee, P.D. [i.e. P. B.] Dunbeck, D.H. Munro, J.A. Zimmer; Principal Investigators: K.U. Ingard, J.L. Kerrebrock; Includes bibliographical references (pages 22-23); Final report. February 24, 1978 to July 31, 1980

Time accurate internal flow solutions of the thin shear layer equations

Thu, 01 Jan 1981 00:00:00 GMT

Time accurate internal flow solutions of the thin shear layer equations Bush, Robert Hull An implicit factored algorithm for the solution of the thin shear layer approximation of the Navier-Stokes equations is described and explicit boundary conditions are developed for internal flow problems. This scheme is compared to theoretical predictions and experimental data, as well as to other more thoroughly tested numerical schemes. The examples presented demonstrate the ability of the factored algorithm to accurately predict internal flow fields and provide insight into the difficulties associated with the numerical simulation of internal flow fields. February 1981; Includes bibliographical references (page 58)

Design of the MIT Breakdown Turbine Facility

Tue, 01 Jan 1985 00:00:00 GMT

Design of the MIT Breakdown Turbine Facility Epstein, Alan H. (Alan Henry); Guenette, Gerald R. (Gerald Roger); Norton, Robert J. G. Summary: This report details the design, construction, and preliminary testing of a short duration (0.4 sec) test facility capable of testing 0.5 meter diameter, film cooled, high work aircraft turbine stages under conditions which rigorously simulate actual engine conditions. The simulation capability of the facility extends up to 40 atm inlet pressure at 2500 0K (40000F) turbine inlet temperatures. The facility is intended primarily for the exploration of unsteady, three-dimensional fluid mechanics and heat transfer in modern turbine stages. April 1985; Includes bibliographical references (page 151)

Unconventional blade design for compressors

Wed, 01 Jan 1969 00:00:00 GMT

Unconventional blade design for compressors Gopalakrishnan, S. One of the factors causing the low efficiency of axial flow compressors operating at decreased flow rates is the flow separation at blade extremities. To postpone this tendency to separate to lower flow rates, compressor blades were designed with increased chord at the blade extremities, which are immersed in the hub and casing boundary layers. The experimental results have shown that the stage static pressure rise is not significantly altered when the modification is used, but the rotor and stage efficiencies are significantly higher at low flow rates. A lifting vortex, similar to that encountered on delta wings at moderate angles of attack, has been observed near the hub region of the modified rotor. The observed differences in the turning through the rotor have been explained satisfactorily using a model incorporating the effects of the lifting vortex. The model also demonstrates that the total drag of the modified rotor can be expected to be smaller than that of the conventional rotor at low flow rates. In the Appendix, some theories based on the lifting line approach are described, which predict the behavior of a stationary cascade facing an inlet flow with spanwise variations in the angle of attack. The comparison of the theoretical results with the very limited amount of experimental data available is considered satisfactory. May 1969; Also issued as: Thesis (Sc. D.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1969; Includes bibliographical references (leaves 64-66)

Some observations of the vibrations of slender rotating shafts

Fri, 01 Jan 1982 00:00:00 GMT

Some observations of the vibrations of slender rotating shafts Dugundji, John; Buellesbach, Monica M.; Wright, Mary A. (Mary Anne Jackson), 1941- The linear theory of a slender, initially bowed, rotating shaft is reviewed for both free and forced vibrations, and found to compare well with a simple experiment on such a shaft. The shaft behavior passing through the critical speed is described in detail, and the maximum bowed-out static deflection of the shaft was found dependent on the external damping and the initial bowing. The amplitude of the oscillatory deflections of the shaft due to gravity loads increased somewhat near the critical speed, but these increases were small compared to the large static deflection of the shaft. During rapid passage through the critical speed, low frequency whirling modes were excited transiently. At higher rotation speeds, the second critical speed was observed, and also the first mode was excited subharmonically and appeared as a backward whirl mode relative to the rotating shaft. March 1982; Includes bibliographical references (page 29)

Quantitative investigation of inlet vortex flow field

Sun, 01 Jan 1984 00:00:00 GMT

Quantitative investigation of inlet vortex flow field Shin, Hyoun-Woo; Shippee, Cheryl (Cheryl Lynn) A quantitative investigation of the flow field of an inlet in cross-wind has been carried out for various operating conditions, including flow regimes representative of aircraft engine inlets at take-off. The measurements were made to show the influence of two non-dimensional parameters, height of the inlet above the ground to the inlet diameter ratio (H/D), and velocity ratio (Ui/Ue), on the vortex strength (i.e., circulation) and the position of the vortex inside the inlet. In general, as H/D is decreased and/or Ui/Ue is increased, the vortex strengthens. In addition, at an operating condition where a strong inlet vortex was present, the circulation of both the inlet and trailing vortices was determined. These were found to be essentially equal in magnitude but opposite in sign, verifying a previous hypothesis concerning the vortical structure of the flow. Qualitative flow visualization has been done to examine two aspects of the inlet/trailing vortex flow field. One is the nature of the transition between regimes in which an inlet vortex is present and those in which no inlet vortex exists. The other deals with the extent of the separated region on the inlet body. The results of this study were used to support a conceptual model of the formation of an inlet/trailing vortex system for an inlet in a cross-wind. March 1984; Includes bibliographical references (leaf 126)

Channel flow modeling of impingement cooling of a rotating turbine blade

Sun, 01 Jan 1984 00:00:00 GMT

Channel flow modeling of impingement cooling of a rotating turbine blade Koo, JaHye Jenny Local heat transfer distributions in impingement cooling have been measured by Kreatsoulas [1] and Preiser [2] for a range of conditions which model those in actual turbine blades, including the effects of rotation. These data were reported as local Nusselt numbers, but referred to coolant supply conditions. By means of a channel flow modeling of the flow in the supply and impingement passages, the same data are here presented in terms of local Nusselt number distributions such as are used in design. The results in this form are compared to the nonrotating impingement results of Chupp [3] and to the rotating but non-impingement results of Morris [4]. Rotation reduces the mean Nusselt numbers from these found by Chupp by about 30 percent, and introduces important radial variations which are sensitive to rotation and to leading edge stagger angle. December 1984; Includes bibliographical references (leaf 29)

An investigation of rotating stall in a single stage axial compressor

Sat, 01 Jan 1955 00:00:00 GMT

An investigation of rotating stall in a single stage axial compressor Montgomery, Stephen Ross; Braun, Joseph J. The rotating stall characteristics of a single stage axial flow compressor were investigated. The number of stall cells and their propagation velocities were found with and without stator blades. The measured velocities were compared with those predicted by Stenning's theory, assuming the downstream pressure fluctuations to be negligible, and correlation within 25% was obtained over a wide range of stall patterns. It was found that the pressure fluctuations caused by rotating stall were less downstream of the rotor than upstream; the minimum reduction across the rotor was 40% with stator blades, and 75% without stator blades. It was also found that, for the compressor tested, the stator blades decreased, the number of stall cells and tended to induce rotating stall at larger mass flow rates. May 1955; Thesis written jointly by both authors: Thesis (M.S.)--Massachusetts Institute of Technology. Dept. of Mechanical Engineering, 1955; Includes bibliographical references

FORTRAN program for calculating three-dimensional, inviscid, rotational flows with shock waves in axial compressor blade rows : I--user's manual

Thu, 01 Jan 1981 00:00:00 GMT

FORTRAN program for calculating three-dimensional, inviscid, rotational flows with shock waves in axial compressor blade rows : I--user's manual Thompkins, William T. Summary: A FORTRAN-IV computer program has been developed for the calculation of the inviscid transonic/supersonic flow field in a fully three-dimensional blade passage of an axial compressor rotor or stator. Rotors may have dampers (part-span shrouds). MacCormack's explicit time-marching method is used to solve the unsteady Euler equations on a finite difference mesh. This technique captures shocks and smears them over several grid points. Input quantities are blade row geometry, operating conditions and thermodynamic quantities. Output quantities are three velocity components, density and internal energy at each mesh point. Other flow quantities are calculated from these variables. A short graphics package is included with the code, and may be used to display the finite difference grid, blade geometry and static pressure contour plots on blade-to-blade calculation surfaces or blade suction and pressure surfaces. Flows in four transonic compressor rotors have been analyzed and compared with exit flow field measurements and intra-blade static density measurements obtained with a gas fluorescence technique. These comparisons have generally shown that the computed flow fields accurately model the experimentally determined passage shock positions and overall aerodynamic performance. The computer code was developed and generally run on a large minicomputer system, a Digital Equipment Corporation PDP-ll/70, with run times of two to three days. The code has also been run on several main-frame computers (IBM 3033, IBM 360/678, UNIVAC 1110, CDC 7600 and a CRAY-1). Typical run times on an IBM 3033 have been found to be 5-10 hours. September 1981; Includes bibliographical references (pages 81-82)

A study of outlet guide vanes

Tue, 01 Jan 1963 00:00:00 GMT

A study of outlet guide vanes Qvale, Einar Björn A criterion for design of unstalled compressor blades has been proposed, and a set of outlet guide vanes has been designed according to this criterion and has been tested. A typical compressor stator and outlet guide vane configuration has been tested and compared with the new design. The tests show that the new design gives considerably higher pressure coefficients and lower pressure losses. June 1963; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1963; Includes bibliographic references

Secondary flow and losses in a compressor cascade

Wed, 01 Jan 1958 00:00:00 GMT

Secondary flow and losses in a compressor cascade Soderberg, Olof E. Three ways in which secondary flow can be generated in a straight compressor cascade have been investigated. 1. Wakeflow. The inlet flow is characterized by a constant inlet angle and a varying stagnation pressure over the span. 2. Skewed flow. The inlet flow is characterized by a constant stagnation pressure and a varying inlet angle over the span. 3. Skewed wakeflow. The inlet flow is characterized by a variation of both stagnation pressure and inlet angle over the span. For a certain combination of inlet angle and stagnation pressure distribution (presented in a formula) in the skewed wakeflow case no secondary flow is generated behind the cascade. The kinetic energy of the secondary flow was found to be very small in all three cases. The secondary flow itself did not create any losses but the blades stalled in the skewed flow layer and in the skewed wakeflow layer causing great losses. The stream pressure and the tangential blade force were lower in the disturbed flow region. A small perturbation theory (for a non-viscous, incompressible fluid) had been developed to describe analytically the secondary flow. Formulae readily adaptable for numerical calculations of flow deviation angles, kinetic energy of the secondary flow, and tangential blade force are presented. Good and satisfactory correlation of theory and experiment was found. Applied to compressor design the results imply: The secondary flow is very small and may occur as an overturning or an underturning of the flow at the casings depending on the actual design. The kinetic energy of the secondary flow may be neglected when considering the losses. The ends of the blades stall in the skewed boundary layer at the casings causing losses. This could be reduced by twisting the blade ends to account for the increased incidence angle. The stream pressure varies in spanwise direction through the boundary layer on the casings. August 1958; Also issued as: Thesis (Sc. D.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1958; Includes bibliographical references

Three-dimensional laminar boundary layer in curved channels with acceleration

Sun, 01 Jan 1956 00:00:00 GMT

Three-dimensional laminar boundary layer in curved channels with acceleration Senoo, Y. A theory is developed for two families of three-dimensional laminar boundary layers; namely, for the boundary layer on the parallel plane end walls of a curved channel with logarithmic spiral side walls, and for the boundary layer on the plane end wall of a concentric circular-arc channel having a particular family of accelerated or decelerated main flows. The second case shows the influence of acceleration and deceleration of a curved main flow. Numerical calculations show that acceleration makes the boundary layer thin and deceleration makes it thick, but the variation of thickness due to pressure gradient is very small compared with that in the two-dimensional case. The first case can be compared to the flow in a cascade. In this case, the variation of the width of the channel is directly related to the variation of the main flow velocity. According to the calculation, the boundary layer is thicker in an accelerated flow through a converging logarithmic spiral channel than in the decelerated flow through the same channel in the opposite direction. It is suspected that converging side walls make the end-wall boundary layer thick and that the effect of convergence is dominant over the effect of accelerated main flow. Experimental data on the end wall of a turbine nozzle cascade were compared with theoretical predictions, with fair agreement across the nozzle and along the center line of the nozzle. November 1956; Includes bibliographical references

Cascade performance with accelerated or decelerated axial velocity

Thu, 01 Jan 1959 00:00:00 GMT

Cascade performance with accelerated or decelerated axial velocity Kubota, Shigeo A theoretical method to estimate the effect of axial velocity change through a cascade was investigated. The change of axial velocity was reproduced by distributing sinks and sources within the blade passages, and the conclusions are set forth in some simple formulae. Some graphs for the numerical evaluation of the performance of NACA 65 series cascades were prepared, and several examples were compared with experimental data. September 1959; Includes bibliographical references (page 27)

Boundary layer separation ; preliminary report

Mon, 01 Jan 1962 00:00:00 GMT

Boundary layer separation ; preliminary report Moses, Hal Lynwood The phenomenon of incompressible boundary layer separation and the existing methods of predicting it are discussed. The failure of these theories in many cases clearly indicates a need for further investigation. A program is proposed that, it is hoped, will improve the present situation. The study is directed primarily at turbulent flow, but the laminar case is treated as well. A theoretical method is presented which involves the ability of the boundary layer to transfer momentum to the fluid near the wall by shear stress. The apparatus, which has already been built for the experimental investigation, is described. Due to its flexibility, the apparatus should prove valuable in comparing and improving methods of predicting separation. May 1962; Includes bibliographical references; Preliminary report

A study of the end wall boundary layer in an axial compressor blade row

Sat, 01 Jan 1955 00:00:00 GMT

A study of the end wall boundary layer in an axial compressor blade row Moore, Raymond W. (Raymond William); Richardson, David L. Turbulent end wall boundary layers in the hub region of axial compressor blade rows have been studied in the Gas Turbine Laboratory. Both analytical and experimental work are presented. An attempt was made to predict the growth of the boundary layer momentum thickness as it passes through a cascade of blades using the momentum integral equation and several assumptions This calculation was compared with measurements taken along an assumed free streamline in a stationary cascade of blades supplied with an artificially skewed inlet boundary layer. From this comparison, the relative influence of individual terms in the momentum integral equation is deduced. The effects of inlet skewing and cross flow in the boundary layer in the region between blades is described. Two methods of representing main and cross flow velocity profiles are compared with experimental data. October 1955; Errata sheet inserted in front; At head of title: Shear flow in bends; Includes bibliographical references (page 12)

Boundary layer on an airfoil in a cascade

Wed, 01 Jan 1958 00:00:00 GMT

Boundary layer on an airfoil in a cascade Peterson, Carl R. A study of the turbulent boundary layer on an airfoil in a cascade is presented. The major portion of the observations are on the suction surface of the airfoil. The effect of added diffusion across the cascade due to a three-dimensional flow is included. Empirical equations describing the boundary layer in terms of momentum and displacement thickness growth as a function of free-stream diffusion are developed. Shape factor is also related to free-stream diffusion. A reasonable check with an existing separation criterion is observed. An appendix contains a discussion of chord-wise integrating of various boundary layer parameters. December 1958; This report was originally the authors master's thesis: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1958); Includes bibliographical references

A cascade tunnel for investigation of rotating stall

Fri, 01 Jan 1954 00:00:00 GMT

A cascade tunnel for investigation of rotating stall Kriebel, Anthony R.; Stenning, Alan H. (Alan Hugh) A test apparatus which will reproduce the rotating stall phenomenon found in axial compressors has been constructed and installed in the main wind tunnel circuit of the Gas Turbine Laboratory. The test section consists of a two dimensional, circular, radial outflow cascade between two flat plates. The cascade was designed to reproduce the pressure distribution, and hence the stall characteristics, of a typical rectilinear cascade. The air entry angle to the cascade is controlled by a variable angle nozzle ring so that the cascade may be stalled and unstalled in operation, Provision has been made for removal of wall boundary layers by suction, and for changing the setting of the cascade stagger angle. By a simple modification, the blades may be supported elastically to permit torsional vibrations. Independent variation of the Reynolds number and Mach number of the air stream is possible. Visual observations of the flow using Schlieren and Interferometer equipment can be made through windows in the walls of the test section. In preliminary tests, rotating stall has been observed and the research program will commence in August of 1954. JUN 1954 --hand-stamped on title-page; 'Stall flutter phenomena in axial flow machines' terminal report -- title-page; Includes bibliographical references; Terminal report; June 1954

Temperature wake in an adverse pressure gradient

Sun, 01 Jan 1961 00:00:00 GMT

Temperature wake in an adverse pressure gradient Sturek, Walter B. (Walter Beynon) A study is made of the flow pattern resulting from the presence of a temperature wake in a uniform velocity field subjected to various adverse pressure gradients. Dimensionless temperature profiles and velocity profiles are presented. A velocity defect appears in the region of high temperature, and the magnitude of the adverse pressure gradient over the range investigated has no effect on the spreading of the temperature wake. A simple theoretical approach is presented which correlates well with the experimental results. The implications of the experimental results and the theoretical approach are discussed with reference to an axial compressor with a temperature wake in the fluid stream at the inlet. June 1961; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1961; Includes bibliographical references

Comparative study of tip clearance effects in compressors and turbines

Sun, 01 Jan 1961 00:00:00 GMT

Comparative study of tip clearance effects in compressors and turbines Yokoyama, Eiji The influence of tip clearance on the lift distribution along a blade span was studied on linear compressor and turbine cascades. The wall boundary layer was removed by means of the image technique. The lift distribution in the span direction was found to be almost similar for both compressor and turbine cascades having the similar velocity diagrams. The lift acting on the blades increases toward the blade tip. The phenomena were explained theoretically by considering the velocity induced by tip vortices on the blade surface. May 1961; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1961; Includes bibliographical references

Stall propagation in axial compressors

Sat, 01 Jan 1955 00:00:00 GMT

Stall propagation in axial compressors Stenning, Alan H. (Alan Hugh) A theory of stall propagation in a cascade of airfoils of high solidity has been developed which includes the effects of finite blade chord and of the boundary layer response to changes in angle of attack. The theory is valid for small perturbations in velocity about a mean flow- condition with finite pressure rise across the cascade, provided that the pressure fluctuations behind the cascade are much smaller than those ahead of the cascade. The solution for the velocity of stall propagation indicates that the velocity increases with the wave length of the stall cell, tending towards a limiting value for very large stall cells. The wave length of the stall cells at the beginning of rotating stall is dependent on the magnitude of the boundary layer time delay. The theory predicts propagation velocities within 25% over a wide range of wave lengths for a stationary circular cascade and a rotating blade row which have been tested. The experiments have confirmed that an increase in wave length is accompanied by an increase in propagation velocity if other parameters are unchanged. April 1955; Includes bibliographical references

Stall propagation in a cascade of airfoils

Sun, 01 Jan 1956 00:00:00 GMT

Stall propagation in a cascade of airfoils Kriebel, Anthony R. An experimental investigation of stall propagation in a stationary circular cascade by means of high speed schlieren and interferometer photography is described. This investigation suggests an analytical approach to the problem which is valid only for an isolated blade row in an infinite flow field but which is not restricted to small unsteady perturbations or an assumed simplified cascade geometry. Conditions necessary for the existence of the assumed type of stall cells are described and equations are derived for the velocity of stall cell propagation. The propagation velocities predicted for the theoretical potential flow-model correlate with all the experimental values measured in an isolated rotor within 15%. Analysis of the flow model predicts a tendency for the assumed type of stall cell to split with increasing incidence of the mean flow through the blade row which appears to correlate with the experimental observation of a trend for increasing numbers of cells in the rotor. August 1956; Also issued as: Thesis (Sc.D.) Massachusetts Institute of Technology. Dept. of Mechanical Engineering, 1956; Includes bibliographical references

The effect of tip clearance on an axial flow fan

Sat, 01 Jan 1955 00:00:00 GMT

The effect of tip clearance on an axial flow fan Ryan, Francis J. (Francis James); Ōhashi, Hiroshi The effect of tip clearance on the performance of a single stage axial flow (6000 cfm) fan was investigated for tip clearances ranging from 0.003 in. to 0.252 in. It was found that, at the rated flow, the stagnation pressure rise (mass flow weighted) across the fan rotor was a maximum for the smallest tip clearance and a minimum for the largest tip clearance. However, it was found also that at the rated flow, the stagnation pressure rise was larger at a tip clearance of 0.022 in. than at a tip clearance of 0.015 in. At a flow of 4000 cfm (two thirds of rated flow) the effect of tip clearance was found to be negligible over the entire range of clearances tested. At flows in excess of 6500 cfm, the stagnation pressure rise was highest for a tip clearance of 0.022 in. Plotting a performance map of stagnation pressure rise across the rotor versus volume flow also revealed that as the peak pressures became higher, these peaks shifted to lower values of flow. This would indicate a shifting of the surge line vertically and horizontally on the performance map as the tip clearance is varied. The flow pattern in the original fan, as manufactured, was investigated also. This revealed that, at the rated flow, the suction sides of the stator blades were stalled over approximately the after one-half of their surface while the pressure sides were completely unstalled. This study also showed that, at all flows tested, the flow pattern at the rotor tip was somewhat turbulent, while the pattern at the rotor root was stable at high flows but became very unstable at low flows. At flows below 000 cfm reversal of flow started near the rotor tip. May 1955; AT head of title: Three-dimensional flow in turbo-machine research; Includes bibliographical references (page 18)

A light scattering investigation of droplet growth in nozzle condensation

Wed, 01 Jan 1969 00:00:00 GMT

A light scattering investigation of droplet growth in nozzle condensation Roberts, Richard, Sc. D. Massachusetts Institute of Technology /, by Richard Roberts An experimental and theoretical study has been made of the condensation of water vapor (with air carrier) in a supersonic nozzle in order to investigate the possible existence of condensate droplets which are substantially larger than predicted by the standard application of classical condensation theory. Droplet size was measured using light scattering techniques, which when combined with the total mass concentration of condensate, provided limits on the maximum and average droplet size. It was found that approximately one part in 10 of the droplet concentration reached a size a factor of 10 greater than predicted by the classical theory . . .The maximum droplet size, furthermore, was not seen to decrease proportionately as the nucleation zone was approached, indicating that the larger droplets are formed during the early stages of condensation. Inconclusive evidence suggests that this occurs following the completion of nucleation but before the vapor supply is exhausted. A calculation procedure which allowed the separation of the nuclei into a distribution of sizes, arising from a varying stability criterion and radius dependent growth rate, resulted in the establishment of a qualitatively correct distribution shape but no theoretical substantiation of an aging or coarsening mechanism. A separate application of Brownian coagulation theory to surface-averaged condensation theory resulted in the prediction that the average droplet size increased by a factor of between 2.5 and 4. No conclusion could be drawn concerning the actual existence of this size increase due to the level of uncertainty in the determination of average droplet size. February 1969; Also issued as: Thesis (Sc. D.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1969; Includes bibliographical references (leaves 82-87)

Boundary layer effects on airfoil lift

Wed, 01 Jan 1958 00:00:00 GMT

Boundary layer effects on airfoil lift Schneider, Kurt H. (Kurt Herbert) A study of the discrepancy in lift between potential flow calculations and experiments was made. In potential flow the fluid is assumed to follow the contour of the airfoil, forming a stagnation point at the geometric center of the trailing edge. In the flow of a viscous fluid around an airfoil a boundary layer forms along the surface of the airfoil changing its effective shape. The pressure at the trailing edge is approximately equal to the static pressure at upstream infinity. Detail stagnation and static pressure and flow direction surveys were made near the trailing edge on a 50 inch chord NACA 0015 airfoil with trailing edge thickness increased to three per cent chord. By the use of thin airfoil theory, modified to include terms. of second order, good agreement with published experiments in lift and pressure distribution is obtained. Using Teledeltos paper, an electrical analog of the flow around the NACA 0015 airfoil in the tunnel was made. A method was developed for determining velocity directly from the analog. September 1958; Includes bibliographical references (leaves 48-49)

The design of axial inducers for turbo-pumps

Wed, 01 Jan 1958 00:00:00 GMT

The design of axial inducers for turbo-pumps Stenning, Alan H. (Alan Hugh) In many turbopump applications it is desirable to run the pump at the highest possible speed to minimise the size and weight of the unit and facilitate matching with a drive turbine. Frequently, a limit on rotational speed is imposed by pump cavitation with its associated deterioration in performance and structural damage. For conventional single sided centrifugal pumps cavitation occurs when the suction specific speed ... exceeds 8-10,000 (1) so that for such pumps the maximum rotational speed without cavitation is determined by the flow and the suction head available. To permit operation at higher speeds an inducer or boost pump may be mounted in front of the main pump (f ig. 1). A typical axial inducer is simply a very lightly loaded axial pump which raises the pressure of the fluid sufficiently to avoid cavitation in the main pump. It has been found possible to operate inducers successfully at suction specific speeds up to 30,000, so that considerable reductions in total pump weight can be achieved when they are employed. Of course, since the inducer is simply a lightly loaded pump which is capable of handling a cavitating fluid, the functions of inducer and main pump may be combined by designing the main pump so that the inlet is lightly loaded. However, it is not always convenient to do this and in many applications conventional centrifugal pumps are still used, preceded by a separate inducer. The objective of this report is to put the design of such inducers on a rational basis by developing a method of calculating blade shapes for the optimum pressure distribution. February 1958; Includes bibliographical references (page 15)

Experiments on turbulent boundary layers along a circular cylinder with and without separation

Sun, 01 Jan 1967 00:00:00 GMT

Experiments on turbulent boundary layers along a circular cylinder with and without separation Fernholz, Hans; Gibson, Paul. Massachusetts Institute of Technology. Gas Turbine Laboratory Summary: Experiments were conducted in a turbulent boundary-layer near separation along a circular cylinder with the flow in the axial direction. The pressure gradient along the axis of the cylinder could be varied such that it was possible to maintain three boundary-layer configurations close to separation or with regions of reversed flow: 1. A turbulent boundary-layer with skin friction zero. 2. A turbulent boundary layer with a separated region and reattachment further downstream with skin friction zero. 3. A turbulent boundary layer with a region of small but constant skin friction and normal separation. Pressure and skin friction along the cylinder wall,. as well as mean velocity profiles in the boundary-layer, were measured. August 1967; Includes bibliographical references

The three-dimensional turbulent boundary layer in a free vortex diffuser

Wed, 01 Jan 1958 00:00:00 GMT

The three-dimensional turbulent boundary layer in a free vortex diffuser Gardow, Ernest Bernhard Experimental velocity profiles and static pressure measurements were obtained from the flow in a free vortex diffuser. The purpose of the study was to obtain additional three dimensional boundary layer data and to compare it with existing theories. Free vortex diffuser behavior under varying inlet angles was also studied. The boundary layer profiles correlated well with an existing model for three dimensional boundary layers generated by transverse pressure gradients. From this model boundary layer cross flow can be approximately predicted using a shearless flow analysis for the outer part of the boundary layer and an empirical correlation for the inner portion. At high inlet swirl angles, a separation ring followed by reattachment was found in the diffuser. January 1958; Addendum sheet inserted; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1958; Includes bibliographical references (page 27)

A two-dimensional design method for highly-loaded blades in turbomachines

A two-dimensional design method for highly-loaded blades in turbomachines Dang, Thong Quoc A practical design method for highly-loaded blades in an isolated cascade is presented in this thesis. The flow is assumed to be incompressible and inviscid. The upstream inlet flow condition is taken to be uniform. The present goals of this research are to provide a practical numerical code for the design problem, and a non-linear theory which can be easily expanded to three-dimensions. The theory is based in part on the Clebsh formulation. The blade profile is determined iteratively through the blade boundary conditions using a "smoothing" technique. A practical numerical code is presented for the design problem using "partial smoothing". The program gives very fast convergence solutions with satisfactory accuracy for practical solidity range.; A practical design method for highly-loaded blades in an isolated cascade is presented in this thesis. The flow is assumed to be incompressible and inviscid. The upstream inlet flow condition is taken to be uniform. The present goals of this research are to provide a practical numerical code for the design problem, and a non-linear theory which can be easily expanded to three-dimensions. The theory is based in part on the Clebsh formulation. The blade profile is determined iteratively through the blade boundary conditions using a "smoothing" technique. A practical numerical code is presented for the design problem using "partial smoothing". The program gives very fast convergence solutions with satisfactory accuracy for practical solidity range. Includes bibliographical references; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1983.

Rotor wake transport in turbomachine stators

Fri, 01 Jan 1971 00:00:00 GMT

Rotor wake transport in turbomachine stators Kumar, Ajay; Kerrebrock, Jack L. The mechanism of rotor wake interaction with stators has been examined experimentally by using helium, injected into the rotor wakes, as a tracer for the wake fluid. Time averaged helium Drofiles downstream of the stator, measured with a thermal conductivity cell, indicate the time averaged distribution of rotor wake fluid at the stator exit. The results are in qualitative agreement with the wake transport theory of Kerrebrock and Mikolajczak, but indicate the need for accounting for differential radial drifts of the wake fluid which encounters the motion and pressure sides of the stator blades. They also indicate that the wake transport theory is valid only when the stators flow is not separated. February 1971; Includes bibliographical references (leaf 12)

A new transformation and integration scheme for the compressible boundary layer equations, and solution behavior at separation

A new transformation and integration scheme for the compressible boundary layer equations, and solution behavior at separation Drela, Mark A new coordinate and variable transformation for the two-dimensional boundary layer equations is presented. The normal coordinate is stretched with a scaling length determined by the local solution. The boundary layer thickness is then essentially constant in computational space for the most types of flows, including separation bubbles and rapidly growing turbulent boundary layers. Similarity solutions can be obtained for all wedge flows. Two finite difference schemes are presented: the Shifted Box Scheme and the Double-Shifted Box Scheme. Both schemes are more resistant to streamwise profile oscillations than the standard Keller's Box Scheme. All governing equations, including the turbulence model, are solved simultaneously as a fully coupled system. This is faster and more robust than conventional weak-coupling iteration schemes. The solution scheme implementation presented makes no restriction on one boundary condition. Any point or integral quantity such as edge velocity, wall shear, displacement thickness, or some functional relationship between two or more of such quantities can be prescribed. The behavior of the boundary layer solution near separation is investigated. It is demonstrated that non-unique solutions always exist whenever an adverse pressure gradient is specified. This bifurcation of the solution is responsible for inability of calculations with prescribed pressure or edge velocity to be carried past separation. Includes bibliographical references

Optimal mistuning for enhanced aeroelastic stability of transonic fans

Sat, 01 Jan 1983 00:00:00 GMT

Optimal mistuning for enhanced aeroelastic stability of transonic fans Hall, Kenneth C.; Crawley, Edward F. An inverse design procedure was developed for the design of a mistuned rotor. The design requirements are that the stability margin of the eigenvalues of the aeroelastic system be greater than or equal to some minimum stability margin, and that the mass added to each blade be positive. The objective was to achieve these requirements with a minimal amount of mistuning. Hence, the problem was posed as a constrained optimization problem. The constrained minimization problem was solved by the technique of mathematical programming via augmented Lagrangians. The unconstrained minimization phase of this technique was solved by the variable metric method of Broyden, Fletcher, and Shanno. The bladed disk was modelled as being composed of a rigid disk mounted on a rigid shaft. Each of the blades were modelled with a single tosional degree of freedom. Adamcyzk and Goldstein's linearized aerodynamic model for the unsteady moment coefficients in a supersonic cascade was applied at the typical section. The resulting non-self-adjoint eigenvalue problem is of the form Aq = XBq. The eigenvalues and eigenvectors of this eigenvalue problem were found by a fourth-order Runge-Kutta line integration of the derivatives of the eigenvalues and eigenvectors. It was shown that mass mistuning does not introduce damping into the system, and that a necessary but not sufficient condition for stability is that the blade be self damped. The results of the optimization showed that an optimally mistuned rotor can achieve a given stability margin for a much lower level of mistuning than alternate mistuning. However, it was shown that optimal mistuning is sensitive to errors in mistuning. Small errors in the implementation of optimal mistuning can severely reduce the gains in stability achieved by optimal mistuning. Alternate mistuning, on the other hand, is relatively insensitive to errors in mistune. Carried out in the Gas Turbine and Plasma Dynamics Laboratory, MIT, supported by the NASA Lewis Research Center under grant NSG-3079; November 1983; N84-16180 --Microfiche header; Bibliography: p. 94-96

Flutter and forced response of mistuned rotors using standing wave analysis

Sat, 01 Jan 1983 00:00:00 GMT

Flutter and forced response of mistuned rotors using standing wave analysis Bundas, David J.; Dugundji, John The torsion flutter and forced response of tuned and mistuned cascades is examined using a standing wave approach, as opposed to the traditional traveling wave approach used in cascades aerodynamic models. The motion of the blades and the corresponding cascade aerodynamic loads are expressed in terms of standing wave modes and arbitrary transient motion, by fitting the sinusoidal force coefficients in terms of ratios of polynomials in the Laplace transform variable, sometimes referred to as Pade approximants. Whitehead's two dimensional, incompressible aerodynamic model is expressed in this transient form and is used to solve the flutter and forced response problems. Results obtained with the transient, standing wave analysis for the flutter and forced response are similar to those obtained by traveling wave analyses, but they yield the transient decay rate associated with vibrations of the blades, as opposed to the structural damping required for flutter obtained by the traveling method. The standing wave analysis presented here may prove to be more versatile for dealing with certain applications such as mistuned rotors, "localized" blade flutter, low engine order excitation, transient impulses on the rotor, and coupling in with forced response and dynamic shaft problems. First published as the author's thesis (M.S.) in the M.I.T. Dept. of Aeronautics and Astronautics, 1983; Massachusetts Inst. of Tech. --Microfiche header; March 1983; N84-24586 --Microfiche header; NASA Grant no. NAG3-214; Includes bibliographical references

Lifting-line theory for subsonic axial compressor rotors

Sat, 01 Jan 1972 00:00:00 GMT

Lifting-line theory for subsonic axial compressor rotors McCune, James E. (James Elliot); Dharwadkar, Shashikant Prabhakar The three-dimensional compressible vortex theory of an axial compressor rotor or ducted fan is extended by relating blade loading to blade geometry in the lifting-line approximation. The resulting integral equation, which is valid up to high subsonic Mach numbers, is solved for both design and off-design problems. It is shown that three-dimensional effects must be taken into account, for rotors with non-uniform spanwise loading, in order to obtain accurate predictions of flow angles and other performance parameters. July 1972; Includes bibliographical references (leaf 25)

The time dependent magnetohydrodynamic generator

Tue, 01 Jan 1974 00:00:00 GMT

The time dependent magnetohydrodynamic generator Oliver, David A. (David Anthony), 1939- Introduction: Low magnetic Reynolds number magnetohydrodynamic generators as would be utilized in power plants are conceived of as steady state D.C. devices. Unsteady phenomena may exist in such generators however. Fluctuations induced fluid mechanically or through combustion may develop into amplifying magnetoacoustic instabilities. In addition to possible instabilities appearing in the bulk of the flow, there exist other important unsteady situations in which the generator must operate. These include the transient response of the generator to changes in load conditions, start up and shut down, coupled unsteady interactions of the generator with the power grid, and the response of the generator to faults such as electrode wall break down or the sudden imposition of a short along the insulator wall. Many of these unsteady situations involve large changes in the amplitudes of the fluid mechanical and electrical variables and therefore require a large amplitude unsteady theory.; In the present work we present a description of unsteady quasi one dimensional magnetohydrodynamic generator flows and propose a highly accurate explicit time dependent method of predicting the time response of such flows. This method of calculation is capable of treating MHD flows under subsonic, supersonic, and transonic flow conditions, arbitrary nonuniformities in electric fields and currents, strong interaction parameters, and with normal shocks present in the duct. In Part II a formulation of the appropriate magnetohydrodynamic fluid equations for quasi-one-dimensional flow is given. In Part III a description of the Lorentz forces and Lorentz power in the flow is given. In Part IV a finite difference operator for the unsteady nonlinear MHD equations is proposed and its stability and accuracy characteristics are described.; In Parts VI and VII, illustrations of this analysis are presented for two unsteady generator situations of contemporary interest: (1) the growth and evolution of large amplitude magnetoacoustic fluctuations under conditions of strong interaction, and (2) the behavior of the generator to changes in loading. March 1974; Includes bibliographical references (page 24)

The display of multi-dimensional flows in turbomachinery

Sat, 01 Jan 1972 00:00:00 GMT

The display of multi-dimensional flows in turbomachinery Probst, Harris Carey; Oliver, David A. (David Anthony), 1939- Introduction: As part of a general development of computational techniques for multidimensional flows through turbomachine passages and over turbomachine blading at the M.I.T. Gas Turbine Laboratory, display techniques and programs have been developed for the visualization and interpretation of the computed flows. These display programs have been developed for the Adage AGT-30 system implemented at the Electronic Systems Laboratory at M.I.T. In the present report two major programs are described and illustrated. These are programs for the display of the turbomachine blading and flow passage geometry and for the display of contours of any scalar fluid variable in the flow passage. The manner in which the system is used is as follows. The turbomachine blading and flow passage geometrical data exist on an Adage tape. The output of the fluid dynamic calculation which is presently done on the OS/370 system is connected on OS/370 to an Adage compatible tape. This tape which represents the computed flow is then viewed on the Adage system. August 1972

Noise generation by shock-turbulence interaction

Thu, 01 Jan 1970 00:00:00 GMT

Noise generation by shock-turbulence interaction Kiang, David Tien Sik; Tam, Christopher Kwong Wah; Kerrebrock, Jack L. The noise produced by convection of turbulence through an oblique shock wave has been measured and compared to theoretical predictions by Ribner and Kerrebrock. There is excellent agreement with the theoretical prediction that, for a fixed turbulent input, the downstream noise pressure (divided by the mean pressure), should first increase very rapidly, and then decrease as the normal Mach number of the shock is increased from unity to values of the order of 1.5. This behavior implies that a part of the noise from supersonic jets should behave similarly, with a sharp increase, then a decrease as the nozzle pressure ratio is raised from unity. October 1970; Includes bibliographical references (leaf 10)

Attenuation of sound in lined ducts

Tue, 01 Jan 1974 00:00:00 GMT

Attenuation of sound in lined ducts Cho, Young-chung; Ingard, K. Uno Extensive computations have been carried out of the attenuation characteristics of resonator and porous type duct liners in rectangular and circular ducts. First the frequency dependence of the attenuation constant and the phase velocity of the fundamental duct mode are obtained for a large number of duct and liner parameters. Then, assuming that the fundamental mode is dominant in the lined duct element, the octave band transmission losses have been computed. The effect of the shape of the input spectrum is discussed and shown explicitly for three different spectra, namely, a "flat" spectrum and spectra with slopes of + 6 dB per octave and - 6 dB per octave, respectively. Finally the effect of the length of the duct liner on the octave band transmission loss has been computed. It is found that the octave band transmission loss does not increase linearly with the length of the duct liner, particularly in regions where the attenuation varies strongly with frequency. September 1974; Includes bibliographical references (page 264)

The prediction of inter-electrode breakdown in magnetohydrodynamic generators

Tue, 01 Jan 1974 00:00:00 GMT

The prediction of inter-electrode breakdown in magnetohydrodynamic generators Oliver, David A. (David Anthony), 1939- Introduction: In this paper, a method of calculation of inter-electrode electrical breakdown phenomena in magnetohydrodynamic generators is presented. Interelectrode breakdown is observed to occur in experimental MHD generators with channel pressures near atmospheric whenever the Hall voltage between adjacent electrode segments exceeds a critical value which varies between 30 and 100 volts. When breakdown occurs, the local Hall voltage across the segments decreases suddenly and a large axial current is believed to flow over the insulator segment. If the generator volume to surface area is large enough, the breakdown current can drive into the electrode wall under the action of the Lorentz force and cause physical destruction of the channel wall.; Since the breakdown development is an inherently unsteady electrical shorting occurring in general within a turbulent, compressible fluid mechanical boundary layer, the description of the flow will be framed in terms of the turbulent fluid equations with Lorentz forces present. The time scale for development of the breakdown arc for MHD generators operating under power generation conditions can be estimated to be of the order of 10-3 sec. The turbulent fluctuations at the Reynolds numbers of interest have time scales of the order of 10-6 sec. Thus, the fluid will be described in terms of mean velocity, pressure, temperature, etc., over the turbulent time scale, but these mean variables will be considered to be functions of time over the electrical breakdown time scale. The breakdown arc can in general be expected to be a three dimensional phenomenon. Experimental evidence bearing on this point in MHD generator channels is at this time still contradictory.; Examination of the electrode walls of combustion driven generators reveals definite spots along the magnetic field direction of the electrode at which arc damage occurs. In closed cycle generators, however, careful image converter diagnostic of the discharge structure in the magnetic field direction indicates that electrothermal arc streamers which are significantly nonuniform in the axial direction across the electrode are quite uniform in the magnetic field direction along the electrode and essentially layered or two dimensional structures. In the present study we shall treat the two dimensional breakdown arc within the context of a two dimensional boundary layer theory. Three dimensional models can follow the initial results of a two dimensional model if the two dimensional model proves inadequate in describing experimental observations. In Part II a formulation of the unsteady fluid equations for the electrode wall boundary layer is presented.; In Part III expressions for the turbulent transport terms are presented and discussed. In Part IV the electrical problem is formulated and the procedure for calculation of the instantaneous electric field and current distribution corresponding to the instantaneous distribution of velocity and electrical conductivity in the boundary layer region is presented. In Part V a computational procedure for the solution of the unsteady boundary layer equations with Lorentz forces is described. March 1974; Includes bibliographical references (page 24)

The M.I.T. Blowdown Compressor Facility

Sat, 01 Jan 1972 00:00:00 GMT

The M.I.T. Blowdown Compressor Facility Kerrebrock, Jack L. A Blowdown Compressor Test Facility has been developed which allows aerodynamic testing of full-scale compressor stages at low cost. The rotor is brought to speed in vacuum, a diaphragm is opened, and the test gas allowed to flow for a time of the order of one tenth second, during which the rotor is driven by its own inertia. Both "steady state" performance evaluation and detailed time resolution of the flow on the blade-passing time scale have been demonstrated in a preliminary way for a two-foot diameter transonic rotor with tangential Mach number of 1.2 and nominal pressure ratio of 1.6. May, 1972; Includes bibliographical references (leaf 70)

Small disturbances in compressor annuli with swirl, throughflow and entropy variation

Thu, 01 Jan 1970 00:00:00 GMT

Small disturbances in compressor annuli with swirl, throughflow and entropy variation Kerrebrock, Jack L. Introduction: Control of the radiation of sound from the compressors and turbines of jet engines depends to a great extent on understanding of the propagation of acoustical modes in the ducting, the general design approach being to choose the numbers of blades in interacting rows so that no propagating mode will be strongly excited. The conditions for non-propagation or "cutoff" are therefore critical to this procedure. Another application of modal analysis is found in the linear three dimensional flow theory of turbomachinery. Here the complete isentropic flow field of the compressor rotor is represented as a superposition of normal modes. To date, most such modal treatments have either neglected the effect of average flow velocity in the turbomachine duct, or considered the acoustical disturbances to propagate in a gas at rest in a coordinate system moving with the average flow velocity.; This approach is correct if the resulting (moving) coordinate system is inertial, but in general is not correct for rotating coordinate systems. In the context of compressor analyses, it is valid for uniform axial flow, as applied by McCune, for example, but incorrect for swirl, as applied by Morfey.Indeed, as we shall show, the classical technique of dividing small disturbances into the three classes of vorticity, entropy, and sound fluctuations, which do not interact to first order, is not valid in rotating flows. Thus, a generalization of the concepts of sound and turbulence is needea. Such a generalization will not be achieved in the present work, but it is hoped that a few steps will be made in this direction. The general equations for pressure disturbances in an inhomogeneous swirling gas have been given by Blokhintsev, who also obtained the general equation for an isentropic gas.; Apparently the only other analyses of pressure wave behavior in rotating fluids are those of Salant and Sozou. The former considered the effects of a solid body rotation on the symmetric normal modes, i.e., modes with no tangential nonuniformity. The latter treated the same type of disturbance in a Rankine vortex. The main purpose of the present analysis is to provide a consistent modal acoustic treatment for compressor annuli with large swirl and throughflow, and with radial variations of entropy. The mean flow will be assumed uniform in the axial and tangential directions, so that the results are applicable only sufficiently far upstream and downstream of blading that the first order variations in these directions have died out. As might be expected, the analysis is nevertheless somewhat complex.; While a general treatment will be given, for arbitrary radial variations of entropy and tangential and axial velocity, analytical solutions for the radial eigenfunctions are available only for some special cases. These do include three important cases, namely, 1) isentropic flow with solid body rotation and constant axial velocity, 2) isentropic flow with free vortex rotation and constant axial velocity, and 3) flow with negligible mean velocity but with radial entropy variation. The first of these represents the conditions behind inlet guide vanes, though not with complete consistency, as will be noted below. The second represents quite accurately the conditions behind high-work fan rotors, except for the effects of entropy variation. The last case gives some insight into the effects of such variations. October 1970; Includes bibliographical references (leaf 25)

The development of a perturbed, incompressible, turbulent boundary layer

Wed, 01 Jan 1969 00:00:00 GMT

The development of a perturbed, incompressible, turbulent boundary layer Handa, Hisayuki The response of a fully-developed equilibrium turbulent boundary layer to a small disturbance was observed experimentally under low-Mach number conditions: a turbulent boundary layer in an axisymmetric channel under zero pressure gradient was perturbed by a single fence like two dimensional protuberance, and the subsequent development of the velocity profile, the turbulent-shear-stress profile, and the wall shear stress was recorded by a constant-temperature hot-wire anemometer and a Preston tube. The height of the five roughness elements used ranged from 0.011 to 0.100 inches (2-15% of the original boundary layer thickness). The perturbation effects are observable only in the vicinity of their origin and each parameter undergoes an individual recovery process. The wall shear stress exhibits a unique style in its development. A perturbed turbulent-shear-stress profile shows a maximum as in the case of a turbulent boundary layer in an adverse pressure gradient. The analysis of the data has revealed a self-preserving feature of the developing velocity profile in form for small enough disturbances: the wall region of the boundary layer is in a local equilibrium after 30 roughness-heights and the unaffected outer layer retains its original characteristics. January 1969; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1969; Includes bibliographical references (leaves 27-29, 2nd group)

Condensation of supersaturated organic vapors in a supersonic nozzle

Sun, 01 Jan 1967 00:00:00 GMT

Wall-shear-stress and laminarisation in accelerated turbulent compressible boundary-layers

Mon, 01 Jan 1968 00:00:00 GMT

Wall-shear-stress and laminarisation in accelerated turbulent compressible boundary-layers Nash-Weber, James Ludlow Turbulent-laminar transition in compressible, steeply-accelerated, adiabatic, turbulent boundary layers on a smooth wall was investigated experimentally in the ranges of Mach and Reynold's numbers typical of nozzles used in propulsive devices. Correlation of the present and previously published data suggests that the transition of such a shear layer may be predicted by consideration of its trajectory on a plane having an acceleration parameter K and a Reynold's number R [sigma] 2 as coordinates. An ab initio design method has been developed, based on these findings, which will ensure laminar flow before and at the throat of a sufficiently small nozzle operating at sufficiently small total pressure. A new type of surface-pilot was developed and calibrated and used to measure wall shear stresses in both transitional and non transitional flows. Decrease of wall shear-stress in lamina rising flows was found. General-purpose computer programs for data-reduction, surface-pilot calibration and interpretation and boundary layer development predictions were developed. April 1968; Includes bibliographical references

Effect of swirl on turbulent jets in ducted streams

Mon, 01 Jan 1968 00:00:00 GMT

Effect of swirl on turbulent jets in ducted streams Morton, Harmon Lindsay An experimental and theoretical program of research has been carried out to determine the effect of swirl on the mixing of a turbulent jet in a ducted stream. The term swirl is used to describe a flow pattern within the jet where mean streamlines are spirals. It has been found experimentally that the mixing rate of a turbulent jet with a surrounding stream can be increased by a factor of three due to the presence of swirl. Three dimensionless parameters, one of which is the jet nozzle to test section diameter ratio, have been used to describe the experimental results. The well-known integral technique has been used to predict the mean flow field of the turbulent jet in a ducted stream. Velocity profiles and turbulent eddy viscosity have been evaluated from turbulent free jet data. The increased mixing due to swirl has been explained in part by the adverse pressure gradient along the jet centerline associated with a decaying swirl and also by an increase in the magnitude of the Reynolds shear stresses within the jet. The calculation procedure divides the flow into two regions - one before the jet attaches to the wall and one after. Two possible correlations giving first order effects of swirl strength on local eddy viscosity are found to give good results for both the ducted jet and the free jet (a jet exhausting into a still fluid). Comparison of data with theory for the axial position of jet attachment is good for all cases studied. December 1968; Includes bibliographical references (leaves 46-47)

On the theory of shear flow

On the theory of shear flow Hawthorne, W. R. (William R.) In this report are collected together some introductory notes on the theory of the three-dimensional, steady flow of an inviscid fluid. The equations of motion are discussed and transformed and expressions for the streamwise or secondary vorticity derived both for incompressible and compressible flow. A special form of the Clebsch Transformation is used to show that three linearizing approximations are possible, depending on the order of magnitude of the stagnation pressure gradient and the disturbance to the upstream flow. The equations for each approximation are developed in turn and are applied to some elementary examples of the flow through actuator discs, over slender bodies and thin airfoils--isolated and in cascade, the flow in curved channels, and about thick struts, airfoils and through cascades of large deflection. October 1966; Includes bibliographical references

Wave drag in transonic axial compressors

Sun, 01 Jan 1967 00:00:00 GMT

Wave drag in transonic axial compressors Okurounmu, Olufemi Profiles of fluid properties, including flow angle, static and total pressures have been obtained behind a free wheeling 8" diameter transonic rotor. The latter consists of 40 blades with double circular arc profiles, a hub to tip diameter ratio of 0.80 and a mean solidity of 0.94. The variation of stagger angle along the blade span is such that the local relative velocity is aligned with the blade chord at a design point corresponding to an inlet axial Mach number of 0.6 and a rotor speed of 35,000 RPM. The profiles show strong evidence of the existence of standing acoustic waves in the flow passage at transonic relative Mach numbers. A strong drag rise is observed at subsonic relative Mach numbers (MT) close to unity, the slope of the drag curve being negative at MT = 1, but turning positive again shortly after MT = 1. For MT < 1, a remarkably good correlation is observed between the spanwise mean of the measured drag and previous 2-D pressure drag measurements on similar profiles. At low supersonic relative Mach numbers, the drag due to shocks in the blade passages appears to overshadow the wave drag, and as yet, there is no definitive way to isolate the contribution of the wave drag experimentally. Computed values of the wave drag at these speeds, based on McCune's analysis for a non-lifting blade row, are "compared" with the measured drag. A linearized theory is presented for obtaining the induced drag of an axial compressor blade row subjected to- any- arbitrary blade loading distribution. The blades are replaced at their leading edges by bound vortex lines of varying strength along the span, and the induced drag obtained from the induced velocity field of the resulting trailing helical vortex sheets. Use of this lifting line approach restricts the useful range of the theory to relative Mach numbers less than unity, since such a quasi 2-D theory would not be applicable to transonic flows which are believed to be inherently three dimensional. November 1967; Also issued as: Thesis (Sc. D.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1968; Includes bibliographical references

Boundary layer separation in internal flow

Fri, 01 Jan 1965 00:00:00 GMT

Boundary layer separation in internal flow Moses, Hal Lynwood; Chappell, John R.; Goldberger, Tomas An investigation of incompressible turbulent boundary layer separation in internal flow is presented with experimental results for a variable geometry, two-dimensional diffuser and two conical diffusers. A simple analytical model is adopted, which consists of wall boundary layers and a one-dimensional, inviscid core. Several approximate boundary layer methods and the possibility of extending them into the separated region are examined. With a limited amount of separated flow, the calculated pressure agrees reasonably well with the experimental results and gives a fair indication of maximum diffuser performance. The limitations of the model to the more general problem, as well as the problem of singularities and downstream stability, are discussed. September 1965; Includes bibliographical references

Upstream history and apparent stress in turbulent boundary layers

Sat, 01 Jan 1966 00:00:00 GMT

Upstream history and apparent stress in turbulent boundary layers Goldberg, Perry A comprehensive experimental and analytical study specifically designed to investigate upstream history and apparent stresses in incompressible, two-dimensional, turbulent boundary layers has been conducted. Hot-wire measurements of turbulent shear stress and longitudinal turbulence intensity, as well as velocity profiles and wall shear stress measurements, were made for six different pressure distributions. It was found that the turbulent shear stress is dependent upon the upstream history of the flow and not a unique function of the local velocity profile. A simple equation for the dissipation integral . . . with a constant K was found to represent the data well. This expression was used with the mean-flow energy integral equation to obtain a practical method for predicting turbulent boundary layer behavior which accounts for upstream history. The predictions made with this method for the six pressure distributions of this study and for others extracted from the literature agreed well with the experimental data. May 1966; Also issued as: Massachusetts Institute of Technology. Dept. of Mechanical Engineering. Thesis. 1966. Ph.D; Includes bibliographical references (leaves 43-45, second group)

Separation of turbulent, incompressible flow from a curved, backward-facing step

Sat, 01 Jan 1966 00:00:00 GMT

Thin airfoils in rotational flow

Fri, 01 Jan 1965 00:00:00 GMT

Thin airfoils in rotational flow Kotansky, Donald Richard The problem of rotational or shear flow about thin airfoils has been investigated theoretically and experimentally. The theoretical approach is based on the concept of a lifting line in a bounded shear flow whose primary flow velocity profile may be expressed in terms of elementary linear, hyperbolic, and / or circular functions. The solution of the linearized equations of motion is reduced to the solution of a characteristic value problem whose form is dependent on the geometry of the primary flow. The characteristic value problem is solved for four different velocity profiles including that of a monotonic-matched linear profile ( a layer of constant vorticity fluid bounded by layers of uniform flow ) which serves as a model for the experimental shear flows. The experimental work includes the measurement of local lift coefficients and spanwise lift distributions on thin symmetrical airfoils in monotonic shear flows for three values of the ratio of airfoil chord to shear layer thickness. The results of the lifting line theory show good agreement with the experimental data within the range of applicability, i.e. within the linear region of the CL, a relationship and for flow geometries where the distortion ( spanwise convection ) of the surfaces of constant stagnation pressure is negligible. The assumption that the local lift coefficient is a function only of the local angle of attack and the two-dimensional characteristic of the airfoil section ( a fundamental assumption of lifting line theory ) is investigated experimentally through a consideration of local pressure coefficient distributions. An approximate correction to the lifting line theory is suggested for flows in which the distortion of surfaces of constant stagnation pressure cannot be neglected. September 1965; Also issued as: Massachusetts Institute of Technology. Dept. of Mechanical Engineering. Thesis. 1966. Sc.D; Includes bibliographical references (leaves 88-90)

Vortex decay in a conical diffuser

Wed, 01 Jan 1964 00:00:00 GMT

Vortex decay in a conical diffuser So, Kwan-Lok The behavior of vortex flow in a conical diffuser has been studied in this report. The flow considered here is a steady, incompressible, axially symmetric one. The present experimental investigation has resulted in the establishment of five flow regimes. These regimes represent three basic types of rotating flow and two transitional phenomena. In Regime 1 the flow is a laminar, one-celled vortex. In Regime 3 the flow is also a one-celled vortex, but turbulent, In Regime 5 the flow is a turbulent, two-celled vortex. Regime 2 represents the transition of a one-celled vortex flow from laminar to turbulent. This transition is characterized by the formation of a bubble along the axis. Regime 4 shows another transitional phenomenon. This transition is the breakdown of a two-celled vortex into a one-celled vortex, but the character of this breakdown is still not fully explored. These regimes can be considered as the steps that the flow in the diffuser must go through as the flow rate or swirl is increased. The number associated with these regimes indicates the order of change from one regime to another. In Regimes 1,3, and 5, similarity in the velocity profiles does exist, but not in Regimes 2 and 4. For simplicity Regime 3 was selected as the basis for a theoretical model. The initial requirement that the total head in the outer region of the flow be conserved proved unsuccessful. When this condition was dropped in favor of the moment of axial momentum, a theoretical development at least qualitatively in agreement with the observed vortex decay is achieved. It is felt that the neglect of turbulent shear stresses in the analysis is largely responsible for the discrepancy that exists between theory and experiment. September 1964; Errata sheet inserted; Includes bibliographical references; Preliminary report

Wake induced blade forces in turbo-machines

Mon, 01 Jan 1962 00:00:00 GMT

Wake induced blade forces in turbo-machines Lefcort, Malcolm D. (Malcolm David) The wake interaction problem in a turbo-machine has been simulated in a water table. The pressure distribution over a compressor blade during the passage of a wake shed by an upstream wake generator has been determined experimentally using specially developed piezo-electric pressure transducers. The results indicate that the theory developed by R. X. Meyer for wakes that are very thin with respect to the blade chord is generally valid. The theory has been extended to cover the case of wakes of finite thickness. The peak unsteady force amounted to at least 23% of the steady lift force assuming a steady lift coefficient of 1. November 1962; Includes bibliographical references (leaves 45-46)

Centrifugal effect on the boundary layer on a blade of an axial turbo-machine

Mon, 01 Jan 1962 00:00:00 GMT

Centrifugal effect on the boundary layer on a blade of an axial turbo-machine Olsson, Erik K. A. A study of the effect of centrifugal force acting on the turbulent boundary layer of a blade in a turbo-machine is presented. The study includes experimental work conducted in an apparatus featuring an annular cascade of compressor blades, designed to simulate the flow in a compressor. This is a case of three-dimensional turbulent boundary layer flow end the objects of the study were to find the magnitude of the cross-flow and gain a better understanding of the skewed boundary layers. It is found that the cross-flow which develops in the boundary layer is small with velocities of the order of 1/10 of the through-flow velocities. The study further resulted in the development of a model of skewed, turbulent velocity-profiles, involving two "universal" analytical functions and four parameters. This model has been used with success to describe the measured velocity-profiles. The use of the model and the general result that the cross-flow is small has made it possible to find solutions to the boundary layer integral equations written in a "streamline" coordinate system. Application of the calculation method for the experimental data has been made successfully. A loss parameter is defined and an extension to include the crossflow losses is suggested. April 1962; Issued also as: Thesis (Sc. D.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1962; Includes bibliographical references

An investigation of cavitating inducers for turbopumps

Thu, 01 Jan 1959 00:00:00 GMT

An investigation of cavitating inducers for turbopumps Mullan, Philip Joseph The experimental performance of two axial inducers is presented. One of the :designs is analytical and includes radial equilibrium considerations. The other is a simple helix, with a constant pitch angle. The performance data consists of two parts; conventional non-cavitating pump performance, and performance with cavitation. High speed pictures of the cavitation formations were taken. May 1959; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1959; Includes bibliographical references (leaf [18])

Skewed turbulent boundary layers

Thu, 01 Jan 1959 00:00:00 GMT

Skewed turbulent boundary layers Nelson, Warren G. (Warren George) The problem of predicting the growth of incompressible, skewed, turbulent boundary layers on smooth surfaces is considered. It is supposed that sufficient initial information and a complete knowledge of the external pressure distribution is given. A computation scheme is devised which makes use of the two momentum integral relations and an auxiliary equation. The formulation of the latter is based on the empirical observation that if a velocity profile is plotted in a holograph plane the outer portion is nearly always linear. A theoretical means of describing velocity distributions is formulated. These theoretical profiles contain two "universal"- functions and five parameters that are functions of the surface coordinates only. The "universal" functions are derived analytically and. compared with experiment. Finally, a comparison of theoretical and experimental velocity profiles suggests that the proposed computation scheme is only applicable to situations where gradients of flow quantities in the cross-flow direction are not too large. Therefore, the scheme is probably better suited to the external flow over wings, etc., than to flows in turbomachinery. August 1959; Includes bibliographical references (pages [48]-[49])

The influence of tip clearance on stall limits of a rectilinear cascade of compressor blades

Thu, 01 Jan 1959 00:00:00 GMT

The influence of tip clearance on stall limits of a rectilinear cascade of compressor blades Khabbaz, Ghassan R. (Ghassan Raji) An experimental study of the influence of tip clearance on the stall limits of compressor blades was conducted on a two dimensional rectilinear cascade. By using the mirror and image technique the end wall boundary layer in the gap was dispensed with. The clearance was found to relieve the pressure gradient and to retard stalling. The loading on the blade near the slot was found to be higher than that at a distance further along. August 1959; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1959; Includes bibliographical references (leaves 20-21)

Asymmetric inlet flow in axial turbomachines

Thu, 01 Jan 1959 00:00:00 GMT

Asymmetric inlet flow in axial turbomachines Seidel, Barry S. (Barry Stanley) A modified actuator disc analysis is made which, through an improved prediction of the blade forces, attempts to give closer correspondence with experiment than the previous theory. The fluid is assumed inviscid and incompressible. Perturbations to the two-dimensional flow through an isolated blade row are considered. The steady flow equations of motion and continuity are linearized. According to experiments conducted on an isolated compressor rotor, the present theory offers an improvement, compared to previous theory, in the prediction of distortion attenuation, effects of flow rate, and effects of varying chord/spacing ratio. May 1959; Includes bibliographical references (leaves 53-55)

A secondary flow approach to the inlet vortex flow field

A secondary flow approach to the inlet vortex flow field Viguier, Henri Charles A theoretical study is presented of the fluid mechanics of the inlet vortex (or ground vortex). The vorticity field associated with this phenomenon is investigated using a secondary flow approach. In this approach the flow is assumed to be composed of an irrotational primary flow and a weak shear flow, with the vortex filaments associated with the latter being regarded as convected (and deformed) by the former. The potential flow field induced by the inlet-ground plane combination is computed using the three-dimensional panel method code developed by Hess, Mack and Stockman. Using this analysis, material lines (which coincide with vortex lines) can be tracked between a far upstream location, where the vorticity can be taken as known, and the engine face location. The deformation of the material lines thus shows directly the generation and amplification of the streamwise component of vorticity which is responsible for the velocity distortion at the compressor face. Two representative flow configurations are considered, one with headwind only and one with the flow at forty-five degrees to the inlet axis of symmetry. The results so far yield only qualitative information; however they do appear to provide some insight into one mechanism of inlet vortex formation. Includes bibliographical references

Prediction of three-dimensional compressible turbulent boundary layers on transonic compressor blades

Prediction of three-dimensional compressible turbulent boundary layers on transonic compressor blades Usab, William J., Jr. (William James) A small crossflow approximation to the full three dimensional compressible turbulent boundary layer equations for turbomachine blade rows is developed by taking advantage of the nature of blade geometry and inviscid flow field when an intrinsic coordinate system is used. The resulting system of equations is solved by Keller's box scheme, providing the capability of numerically calculating compressible turbulent boundary layers on transonic compressor blades to a good approximation. The scheme is checked with two known solutions of incompressible flow over unloaded zero thickness blades. It is then applied to the first stage of a NASA Low-aspect-ratio rotor blade for which the inviscid flow field is available. The results give insight to the three-dimensional boundary layer character of transonic compressor blades, caused by an imbalance of centrifugal and Coriolis forces within the boundary layer. Includes bibliographical references

Experiments on turbulent boundary layers along a circular cylinder with and without separation

Sun, 01 Jan 1967 00:00:00 GMT

Detailed time resolved measurements and analysis of unsteady flow in a transonic compressor

Detailed time resolved measurements and analysis of unsteady flow in a transonic compressor Ng, Wing Detailed time and space resolved measurements for a transonic compressor stage have been completed in the MIT Blowdown Compressor Facility. The stage studied was a new first stage for a NASA two-stage machine which incorporated low-aspect-ratio blading. This rotor has an inlet hub/tip ratio of 0.375, aspect ratio of 1.56 and an inlet relative Mach number of 1.38. The purposes of the test were to compare the results obtained by the Blowdown technique with those found by steady state testing at NASA Lewis Research Center, and to provide new time resolved data on the blade-to-blade flow in the rotor and stator. Data were obtained by surveys with a five diaphragm high frequency response probe and by tip casing transducers. The stage was tested at 100% design speed. Time resolved estimates of efficiency were obtained by direct measurement of stagnation pressure together with calculation of stagnation temperature by the Euler equation using measured tangential flow Mach number. Test results showed that the rotor achieved an adiabatic efficiency of 0.895 at a total pressure ratio of 1.677. The stage achieved an adiabatic efficiency of 0.862 at a total pressure ratio of 1.658. Corrected mass flow at design was measured to be 33.3 Kg/s with respect to air. Time averaged flow quantities in general agree very well with results from steady state tests at NASA Lewis Research Center. A significant difference was observed in the variation of efficiency with radius, with a low efficiency region near mid-span not observed in the steady state testing. Moreover, the measured rotor efficiency in the "core flow" between the blade wakes for the supersonic region is lower than can be explained by normal shock losses. Large streamwise vorticity is observed at the blade trailing edge in the inner half of the annulus, which may be associated with shock termination at the sonic radius. Includes bibliographical references

Three-dimensional vorticity-induced flow effects in highly-loaded axial compressors

Tue, 01 Jan 1980 00:00:00 GMT

Three-dimensional vorticity-induced flow effects in highly-loaded axial compressors Tan, Choon Sooi A new analytical method is proposed for the study of flow through highly-loaded turbomachine stages. The technique is used in the present study in order to: (i) analyze the three-dimensional induced effects of the viscous blade wakes in an isolated rotor; and (ii) to study the effects of the passage of distorted flow through an axial compressor rotor or stator. In Part I (GT&PDL Report Number 141), it is found, in contrast with the more familiar situation behind aircraft wings, that the induced effects of the vorticity in the (viscous) wakes are important in practical axial turbomachinery; for example, the flow angles through highly-loaded rotors are modified to a significant extent by such wake effects. The induced disturbances grow in strength within a certain distance downstream of the blade row before beginning to decay inversely with such axial distance.; In agreement with earlier predictions, pressure disturbances and vorticity disturbances cannot be decoupled in swirling flow. Similarly, in part II (GT&PDL Report Number 151), it is found that major differences arise on comparing two-dimensional with three-dimensional analyses, both for rectilinear and for annular configurations. Further, only the last of these three-dimensional analyses can adequately describe the true flow phenomena in highly-loaded turbomachines. This is because such a description properly includes both centrifugal effects together with two important distinct types of vorticity: the trailing vorticity and the vorticity associated with any stagnation pressure gradients present. Such an analysis predicts, a strongly persisting downstream pressure field which in many cases increases before again beginning to decay inversely with the axial distance downstream, both for free-vortex stators and rotors.; By contrast, three-dimensional wheel flow analysis predicts indefinitely persisting downstream disturbances. Further, a purely two-dimensional theory indicates for a stator, the downstream static pressure to be uniform, while even a three-dimensional rectilinear cascade theory would predict only an exponentially decaying pressure field. The amplitude of the above persistent downstream disturbances decreases for free-vortex downstream flow as the number of significant circumferential harmonics of the inlet distortion increases. These analytical results agree well with available experimental data recently obtained in annular cascades. January, 1980; This is part 1. Part 2 issued as his Asymmetric inlet flows through axial compressors, GT & PDL report no. 151; Originally presented as the author's thesis, (Ph. D.)--in the M.I.T. Dept. of Aeronautics and Astronautics, 1978; Includes bibliographical references (pages 184-188)

Asymmetric inlet flows through axial compressors

Tue, 01 Jan 1980 00:00:00 GMT

Asymmetric inlet flows through axial compressors Tan, Choon Sooi A new analytical method is proposed for the study of flow through highly-loaded turbomachine stages. The technique is used in the present study in order to: (i) analyze the three-dimensional induced effects of the viscous blade wakes in an isolated rotor; and (ii) to study the effects of the passage of distorted flow through an axial compressor rotor or stator. In Part I (GT&PDL Report Number 141), it is found, in contrast with the more familiar situation behind aircraft wings, that the induced effects of the vorticity in the (viscous) wakes are important in practical axial turbomachinery; for example, the flow angles through highly-loaded rotors are modified to a significant extent by such wake effects. The induced disturbances grow in strength within a certain distance downstream of the blade row before beginning to decay inversely with such axial distance.; In agreement with earlier predictions, pressure disturbances and vorticity disturbances cannot be decoupled in swirling flow. Similarly, in part II (GT&PDL Report Number 151), it is found that major differences arise on comparing two-dimensional with three-dimensional analyses, both for rectilinear and for annular configurations. Further, only the last of these threedimensional analyses can adequately describe the true flow phenomena in highly-loaded turbomachines. This is because such a description properly includes both centrifugal effects together with two important distinct types of vorticity: the trailing vorticity and the vorticity associated with any stagnation pressure gradients present. Such an analysis predicts, a strongly persisting downstream pressure field which in many cases increases before again beginning to decay inversely with the axial distance downstream, both for free-vortex stators and rotors.; By contrast, three-dimensional wheel flow analysis predicts indefinitely persisting downstream disturbances. Further, a purely two-dimensional theory indicates for a stator, the downstream static pressure to be uniform, while even a three-dimensional rectilinear cascade theory would predict only an exponentially decaying pressure field. The amplitude of the above persistent downstream disturbances decreases for free-vortex downstream flow as the number of significant circumferential harmonics of the inlet distortion increases. These analytical results agree well with available experimental data recently obtained in annular cascades. January, 1980; This is Part 2. Part 1 issued as his Three-dimensional vorti-city-induced flow effects in highly-loaded axial compressors , GT & PDL report no. 131; Originally presented as part of the author's thesis (Ph. D.)--in the M.I.T. Dept. of Aeronautics and Astronautics, 1978; Includes bibliographical references (pages 130-134)

A computational study of the flow in a transonic axial compressor using an inviscid, three-dimensional finite difference

Mon, 01 Jan 1979 00:00:00 GMT

A computational study of the flow in a transonic axial compressor using an inviscid, three-dimensional finite difference Haymann-Haber, Guido A computational study of the flow in a Transonic Axial Compressor has been performed. This compressor has a tip Mach number of 1.2 and an inlet hub to tip ratio of 0.5. The numerical procedure used is a fully three-dimensional, inviscid, finite difference algorithm. MacCormack's two-step, explicit second order accurate scheme was used. A total of 30,600 mesh points were used. The results were compared to space and time resolved exit flow measurements, and to quantitative density visualization pictures. Among the most significant features resolved by the computation, was an unusual shock structure, which had earlier been observed in the experiments. The general agreement of the computation with The experiment is good, except in regions dominated by viscous flow. Many of the effects of viscosity can be anticipated from the inviscid flow field. May 1979; Originally presented as the author's thesis, M.S., in the M.I.T. Dept. of Aeronautics and Astronautics, 1979; Includes bibliographical references (leaf 82)

Three-dimensional incompressible and compressible Beltrami flow through a highly-loaded isolated rotor

Mon, 01 Jan 1979 00:00:00 GMT

Three-dimensional incompressible and compressible Beltrami flow through a highly-loaded isolated rotor Tan, Choon Sooi An analysis is proposed for the three-dimensional incompressible and compressible Beltrami flow through a heavily loaded isolated rotor. It is found that, in agreement with previous analyses on disturbances in swirling flows, the Beltrami vorticity-induced disturbances are not purely convected by the mean flow. The disturbances, which induce a persisting static pressure field, can be shown to grow linearly in strength close to the blade-row before beginning to decay inversely with axial distance downstream. This is again in agreement with previous analyses on swirling flow in which the disturbances had their origin in the viscous blade wakes. It is further shown that this analysis only agrees with the earlier analysis, in which vorticity-induced disturbances are assumed to be purely convected by the mean flow, in the limit of large numbers of blades and in a downstream region very close to the blade row. The two analyses differ considerably further downstream; that is, this analysis can predict the downstream evolution of the three-dimensional disturbances while the earlier cannot. October 1979; Includes bibliographical references (pages 78-79)

Turbulent mixing in swirling flow

Sun, 01 Jan 1978 00:00:00 GMT

Turbulent mixing in swirling flow Cheng, Wai Kong The effect of swirl on the mixing of two gas streams of the same density in coaxial annular geometry has been examined. The flow is studied via a fluorescent tracer, 2-3 biacetyl. Through direct excitation, collisional excitation, and collisional de-excitation of the tracer, the turbulent transport, molecularly mixed and unmixed regions in the flow are visualized. A three dimensional structure flattened in the azimuthal direction is observed. It is suggested that this structure is the result of the growth of the unstable azimuthal modes as limited by a turbulent eddy viscosity. September 1978; Originally presented as the author's thesis, Sc. D. in the M.I.T. Dept. of Aeronautics and Astronautics, 1979; Includes bibliographical references (leaf 49)

Flutter analysis of a tuned rotor with rigid and flexible disks

Mon, 01 Jan 1979 00:00:00 GMT

Flutter analysis of a tuned rotor with rigid and flexible disks Dugundji, John The flutter behavior of a simple tuned rotor with a rigid and a flexible disk is reviewed. In Part A, the rotor assembly is assumed to consist of a rigid disk with N uniform flexible blades attached around the circumference, so that the blades are coupled only by aerodynamic forces. Both traveling wave and standing wave flutter analyses are conducted, and are shown to be equivalent. The relations between traveling and standing wave air forces are described in detail. The standing wave analysis is shown to be more versatile for some applications than the simpler traveling wave analysis. Applications are made to pure bending flutter and pure torsion flutter of the rotor assembly. Comments are made on combined bending-torsion flutter. In Part B, the rotor disk is assumed flexible and shrouds may be present. The blades are here coupled structurally as well as aerodynamically. The corresponding vibration and flutter behavior is examined. July 1979; Includes bibliographical references (page 58)

Stalled flow performance of a single stage transonic compressor

Sun, 01 Jan 1978 00:00:00 GMT

Stalled flow performance of a single stage transonic compressor Bartlett, Frederick Girard; Greitzer, E. M. (Edward M.), 1941- The stalled flow performance of a single stage transonic compressor is examined, using data from the MIT Gas Turbine Laboratory's Blowdown Compressor Facility. Measurements of the blowdown corrected weightflow are included, as well as stage exit static to inlet total pressure rise, and rotating stall cell measurements. A comparison of the flow blockage, as represented by stall cell circumferential extent, with the blowdown corrected weightflow is made. The 100% design speed stalled flow performance characteristic is presented. Radial traverse.data is also presented including flow angles, static and total pressures, and Mach number components. Originally presented as the first author's thesis, M.S. in the M.I.T. Dept. of Mechanical Engineering, 1978; Includes bibliographical references (page 48)

The interaction of sound with turbulent flow

Sat, 01 Jan 1977 00:00:00 GMT

The interaction of sound with turbulent flow Succi, G. P. (George Peter) Introduction: In the now classical theory of sound from turbulent flow by Lighthill boundaries were not considered. Further, it was assumed that the turbulent flow field was not altered by sound emission. For closed systems, such as cavities and waveguides, discrete modes may be excited. It is possible that the coupling between the flow field and acoustic modes is strong enough to alter the primary flow field. Under such conditions acoustically induced flow instabilities such as whistles occur. The objective of this thesis is to study the turbulent excitation of duct modes and the conditions for possible instabilities. Chapter One reviews the theory of turbulent excitation of sound in free space. The induced acoustic field is calculated in three ways. The sound field is determined by time-domain Green's function technique for both Lighthill's quadrupole source model and Ribner's equivalent distribution of monopoles.; The omission of turbulent shear interactions in an isotropic monopole model of the turbulence is demonstrated. The sound field is also determined by a frequency-domain Green's function technique for a monopole distribution. Time and frequency domain calculations yield identical results for the monopole source distribution. However, the frequency domain technique allows the analysis to be extended to Chapter Two examines the excitation of axial pipe modes by turbulent flow theoretically and experimentally. The experiment is performed by drawing air through a cylindrical pipe. The observations demonstrate that mode excitation diminishes as the flow speed increases. This is attributed to end losses which increase with flow speed. A Green's function, based on the measured pressure reflection coefficients, is used to predict the variation in spectra with flow. Chapter Three demonstrates the excitation of transverse modes in pipes experimentally.; Air is drawn through a small orifice into a rectangular duct. Pronounced asymmetric peaks are observed at the first few cutoff frequencies. The asymmetric nature of the peaks and relative spectral intensity are again explained by Green's function techniques. For high frequencies, where a large number of modes can propagate, the spectra resembles the free-space jet spectra. In this frequency range, the duct radiation impedance asymptotically approaches the free space impedance, hence the similar response to similar source distributions. Chapter Four examines feedback instabilities, cases where the emitted sound field alters the jet flow itself. The chapter concentrates on screech tones of circular orifices having length to diameter ratios between one half and two. The frequency dependence of the screech on Mach number and length is explained by a kinematic analysis of the feedback loop.; It is further demonstrated that the frequency of the feedback instability must be approximately equal to that of an acoustic made for screech to occur. Similar observations are presented with regard to air jets impinging on plates. In conclusion two mechanisms exist whereby the acoustic energy from a turbulent jet can be concentrated at select frequencies. The first is the selective response of a medium with boundaries to a random source. The spectral line shape for such cases is accounted for by Green's function techniques. The second mechanism is the feedback instability which requires coupling of the jet flow to the acoustic field. Here the modification of the jet flow must be considered to determine the excitation frequencies. It is recommended that future work be done on the screech instability. The influence of the acoustic cavity mode on the feedback instability should be examined in greater detail.; In particular, the convection speed of jet column disturbances with and without adjacent resonators should be determined experimentally. Furthermore, the mechanism which limits the amplitude of the screech should be determined. June 1977; Originally presented as the author's thesis, Ph. D. in the M.I.T. Dept. of Physics, 1977; Includes bibliographical references (pages 217-224)

Heat transfer measurements in turbines

Sat, 01 Jan 1972 00:00:00 GMT

Heat transfer measurements in turbines Louis, Jean F. (Jean François); Demirjian, Ara Manuel; Topping, Richard Francis A blowdown turbine facility and an experimental program for the study of heat transfer in cooled turbines are described. The blowdown facility introduces the concept of using short duration experiments to test turbines by scaling the metal surface temperatures down to room temperature, and consequently scaling pressures and mass flow rates so that Reynolds, Mach and Prandtl numbers are kept unchanged. Hence, the Nusselt number is also unchanged. The short operating time assures that the surfaces remain nearly isothermal and that fast instrumentation, pressure transducers and thin film gauges can be used to record average and unsteady pressures and heat transfer. For example, the instrumentation is able to resolve heat transfer and pressure fluctuation induced by rotating blades passing over the stationary shrouds. Both the operation and demonstration runs of the blowdown facility are described. Finally, a two dimensional steady flow film cooling experiment yields heat transfer coefficients under conditions which model the average tangential flow over the stationary shrouds of the turbine under test in the blowdown facility. The results indicate that the curvature of the shroud leads to an increase of heat transfer by 15% and the separation of the coolant flow at the injection edge reduces the cooling effectiveness just downstream of the injection slot. Includes bibliographical references (leaf 17)

Spherical pressure probe for retrieving freestream pressure and directional data

Sat, 01 Jan 1977 00:00:00 GMT

Spherical pressure probe for retrieving freestream pressure and directional data Figueiredo, William Arthur A spherical pressure probe designed to time resolve the total and static pressure and freestream radial and rotational flow angles in turbomachine flow fields was found to produce accurate results for flow angles over a range of 200 in both the radial and rotational directions. Reynolds number based on the probe sphere diameter had negligible effect over the range 14,000 to 92,000. There is a weak Mach number dependence, but the probe is usable up to a Mach number of 0.90. Calibration curves are plotted for M = 0.27, 0.7, and 0.9. These results were found by calibration in steady flow. From the probe dimensions and transducer response it is judged that the probe should have frequency response better than 30 kHz. August 1977; Includes bibliographical references (page 52)

Fluorescent visualization for turbomachine research

Sun, 01 Jan 1978 00:00:00 GMT

Fluorescent visualization for turbomachine research Epstein, Alan Harry The quantitative gas fluorescence flow visualization technique using 2,3 butanedione as a tracer, has been refined and improved. Upgrading of the imaging system is responsible for the principal improvement. The technique has been applied to air flows in order to demonstrate its suitability to conventional compressor -testing. The possibility of using butanedione to measure static gas temperature has been explored by modeling. It has been found to be feasible only when a time lag after excitation of 1 ms is acceptable. March 1978; Includes bibliographical references (page 21)

Calculation of a self-consistent, low frequency electrostatic field in the drift-kinetic approximation

Sat, 01 Jan 1977 00:00:00 GMT

Calculation of a self-consistent, low frequency electrostatic field in the drift-kinetic approximation Beasley, Cloyd Orris, 1933-; Meier, H. K. Massachusetts Institute of Technology; Rij, W. I. van. Oak Ridge National Laboratory; McCune, James E. (James Elliot) We derive an asymptotic series in [omega]p -2 , the inverse-square plasma frequency, for the self-consistent, low frequency electrostatic field in tori. The derivation is consistent with the drift-kinetic ordering and may be used in either instability or equilibrium calculations. We find that in a time-dependent formalism, the electric field is completely determined to first order in a drift-kinetic expansion. July 1977; Includes bibliographical references (pages 22-23)

Luminescent visualization of molecular and turbulent transport in a plane shear layer

Sat, 01 Jan 1977 00:00:00 GMT

Luminescent visualization of molecular and turbulent transport in a plane shear layer Bates, Stephen Cuyler The purpose of this work is to contribute to the fundamental understanding of fluid turbulence by visualizing its detailed flow structures. Examination of these coherent structures gives information about the turbulent flow that cannot be deduced from its statistics. This information should reduce the role of empiricism in the analysis of turbulence. The experimental method chosen is to visualize a turbulent plane free shear layer using stop-action photography of a phosphorescing trace gas. Choice of 1) direct photo-excitation, 2) collisional excitation, or 3) collisional de-excitation of the phosphorescencing gas with a planar light beam, permits identification of the emission with a cross-sectional map of the material from one stream that is 1) throughout the flow, 2) molecularly mixed with material from the other free stream (alone), or 3) molecularly unmixed. The plane shear layer visualized has been specified experimentally. Extant requirements for self-preservation are insufficient in general, and make the claim of self-preservation for the experimental flow only probable and not definite. A large data set using all three variations of the visualization technique show structures that imply a large amount of new information about turbulent mixing and turbulent processes. The data shows the structures to be simply connected, with slow variation out of the mean flow plane. Specifically, there is a simply connected region of mixed fluid that always separates material entering the layer from the free streams. Collisional excitation and quenching data strongly imply a turbulent mixing process of random bursting from the free stream, followed by internal viscous decay. The complementary process of turbulent entrainment is recorded in the quenching photos as nibbling of the free stream by the layer, together with a randomly occurring large local amplification of this nibbling, previously thought to be engulfment by the boundary. June 1977; Originally presented as the author's thesis, (Sc. D.)--in the M.I.T. Dept. of Aeronautics and Astronautics, 1977; Includes bibliographical references (pages 123-124)

Uniform-inlet three-dimensional transonic Beltrami flow through a ducted fan

Sat, 01 Jan 1977 00:00:00 GMT

Uniform-inlet three-dimensional transonic Beltrami flow through a ducted fan Cheng, Wai Kong Introduction: This report is a continuation of a line of analytical treatment of three dimensional flows in compressor or ducted fans. In the earlier theories [1,2,9,10], the blade row is modelled as a set of spinning lifting lines and the induced velocities are treated as acoustic perturbations. While these studies have been useful in advancing our knowledge of the three dimensional nature of the flow, it has been difficult until now to correlate the theory with experiments. The reason, of course, is that the strong overall swirl induced by such blade rows is by no means a small perturbation in practical applications. It is realized that the high pressure stage of a compressor usually has a large number of blades (~40 to 100). Therefore although the collective effect of all the bales can be a large disturbance, each blade contributes only a small disturbance. Thus a linearized theory is still possible.; To extend the previous acoustic theory to a rotor with large turning we may still represent the blades as superpositions of source and lifting lines, but we have to calculate the exit flow by linearizing about a non-zero (and large) swirl velocity profile. This is the approach taken by the theory proposed by McCune and Hawthorne [3]. They calculated the velocities induced by the trailing vorticity of a nonuniformly loaded rectilinear cascade for incompressible flow. This work was later generalized to the compressible case by Morton [4]. Cheng [5] treated incompressible flow in an annular geometry. In that analysis the blades are represented as lifting lines of nearly constant circulation, and the exit flow is therefore, to lowest order, of "free vortex" type. Linearizing about the free vortex flow, the velocity induced by the trailing vortex sheets due to nonuniform blade bonding are calculated in [5] to order 61.; It is the purpose of this report to treat the general compressible case, including the results of Ref. [5] as the incompressible limit. The result of this analysis also serves as the Green's function for constructing a lifting surface theory for transonic rotors with practical loading. Before going into the details of the theory, let us examine some simple pictures of its findings. The nonuniformly loaded blades shed off the excess circulation as wakes. The induced velocity of the wakes is found to cause a "downwash" at the blade which has the effect of partially nullifying the nonuniformity in loading. This can easily be understood by a consideration of the wake system of a blade. Fig. 1 shows a blade with the tip region more heavily loaded than the inboard stations. 'We can see that the wake induces a tangential velocity component which lowers the angle of attack at the high work (tip) region and increases that of the low work (hub) region.; Another major development is the mode matching of the upstream and downstream flow. The presence of the strong swirl makes the acoustic mode shapes downstream drastically different from those upstream. For example, we can have upstream hyperbolic modes in a transonic rotor while all the downstream modes are elliptic because the relative velocity is subsonic there. Simple mode-wise matching is no longer possible. In the present work, a method of mixed-mode matching is developed. A result is that a pure tone downstream (upstream) can excite a whole spectrum of tones upstream (downstream). In particular, any source downstream can excite the acoustic radiations upstream of a transonic rotor. Errata sheet inserted; Includes bibliographical references (page 50)

Comparison of three dimensional quasi-linear large swirl theory with measured outflow from a high-work compressor rotor

Wed, 01 Jan 1975 00:00:00 GMT

Comparison of three dimensional quasi-linear large swirl theory with measured outflow from a high-work compressor rotor Chen, Lee-Tzong; McCune, James E. (James Elliot) A three-dimensional perturbation theory for incompressible strongly swirling flow in an annulus is applied to predict the outflow from a high-work compressor rotor (1)(2). A comparison of the analytical result with the experimental result is presented. The theory treats inviscid, incompressible flow through a highly loaded blade row in a long annular duct with uniform inlet. Trailing vorticity is shed from each blade which is represented by a lifting line of bound vorticity. The flow field between successive sheets of vorticity is assumed to be irrotational. The theoretical results are compared to data obtained in the M.I.T. Blowdown Compressor Test Facility (3). The mean circulation distribution is estimated from the mean pitchwise Mach number obtained from the experiment. The agreement of the predicted mean axial and radial velocity with the experimental results represents one confirmation of the theory. The pitchwise-varying velocity are evaluated by the theory at an axial distance of .02 tip radius which corresponds to the probe position if the lifting line is placed slightly ahead, but almost along the trailing edge of the blade. The theory predicts well the pitchwise variation of the velocity due to the effect of shed vorticity which results from the spanwise variation of circulation. The effect is pronounced near the blade tips. Corrections of the theory due to compressibility are omitted here, but will be available shortly. September 1975; Includes bibliographical references (page 37)

Collapse of the inter-electrode breakdown arc in the magnetic field direction

Wed, 01 Jan 1975 00:00:00 GMT

Collapse of the inter-electrode breakdown arc in the magnetic field direction Oliver, David A. (David Anthony), 1939- A two-dimensional inter-electrode breakdown arc which is uniform in the magnetic field direction is shown to be unstable. The growth time for the instability is of the order of the arc development time. It is unlikely that a fully steady two-dimensional arc is ever established. The relationship of the instability to recent experiments revealing an inherently three-dimensional breakdown is discussed. July 1975; Includes bibliographical references (leaf 9)

An experimental and computational study of the flow in a transonic compressor rotor

Thu, 01 Jan 1976 00:00:00 GMT

An experimental and computational study of the flow in a transonic compressor rotor Thompkins, William T. A comprehensive investigation of the flow field produced by an isolated transonic compressor rotor has been completed. This rotor has an overall diameter of two feet, an inlet hub/tip ratio of 0.5, a tip Mach number of 1.2 and a total pressure ratio of 1.65. The time resolved three dimensional exit flow produced by this rotor was experimentally measured with sufficient spatial and temporal resolution to determine velocity components and pressures inside individual blade wakes and in the surrounding flow. A numerical calculation of the steady inviscid three dimensional through-flow was computed using MacCormack's second order accurate time-marching scheme. Comparisons between the numerical solution, the exit flow measurements, and measurements of the intra-blade static density field obtained by gas fluorescence showed that the inviscid computation accurately models transonic compressor aerodynamics and rotor blade pressure distributions in the upstream portion of the passages, the viscous effects influencing mainly the downstream portions. It is felt that such a computation procedure has great potential as a compressor design and development tool especially when coupled with a suitable boundary layer analysis. May 1976; Includes bibliographical references (pages 69-71)

Accurate, efficient difference operators for the turbulent field equations

Wed, 01 Jan 1975 00:00:00 GMT

Accurate, efficient difference operators for the turbulent field equations Oliver, David A. (David Anthony), 1939- Introduction: A potentially powerful method for predicting and describing turbulent flow fields is that of utilizing a finite number of moments of the instantaneous Navier-Stokes equations. This sequence of moment equations is closed by theoretical modelling of the unknown correlations at the closure moment level. The use of an eddy viscosity which is "closed" in terms of the mean velocity gradient represents a familiar first order closure. A wide variety of turbulent flow fields are currently under investigation including boundary layers, shear layers, vortices, and wakes. Surveys of the general methods utilizing various closure models may be found in Reynolds 11], Mellor and Herring [2], and Donaldson [3]. The field equations for the turbulent flow which result from these moment equations are generally lengthy, simultaneous, non-linear partial differential equations of the initial value class.; Very little attention has been given to the mathematical structure of these non-linear systems and to the development of accurate numerical methods for their solution. In many flows of interest, diffusion processes are always present in sufficient strength to give the equations a strong parabolic character. Most workers have therefore adopted fully implicit difference operators for the spatial derivatives and have then utilized iterative methods for solving the non-linear implicit equations. These fully implicit operators are apparently selected because of the unconditional stability which such operators provide for linear systems. In many highly non-equilibrium turbulent flows, the convective, production, and turbulent decay processes are stronger than the diffusive processes. In such situations the iterative techniques become slow in convergence or may not converge at all.; In addition, fully implicit difference operators are only first order accurate and require a fine mesh spacing to properly resolve the structure of the flow. In the present work we offer a new class of second order accurate non-linear difference operators which are consistent, unconditionally stable (in the extended sense for the non-linear system discussed below), and which do not require iterative techniques for the solution of non-linear implicit equations. Hence these operators are highly efficient. We begin in Part II by discussing simple non-linear diffusion and convections equations and their possible difference operators. In Part III we examine a more complex non-linear diffusion equation which is representative of that which arises in mixing length closure models. It is here that the essential feature of the difference operator proposed in this work is presented. In Part IV we turn attention directly to the turbulent field equation models currently under investigation.; Here we summarize these equations in a single archetypal form and we present the archetypal form of the difference operator for their solution. July 1975; Includes bibliographical references (page 18)

Exit flow from a transonic compressor rotor

Wed, 01 Jan 1975 00:00:00 GMT

Exit flow from a transonic compressor rotor Thompkins, William T.; Kerrebrock, Jack L. Summary: The three dimensional unsteady flow field behind a transonic compressor rotor with a design pressure ratio of 1.6 at a tip Mach number of 1.2 has been resolved on the blade-passing time scale, using the M.I.T. Blowdown Compressor Facility. Quantities determined were total and static pressures, tangential flow angle and radial flow angle. The spatial and temporal resolution achieved was sufficient to determine velocity components inside individual blade wakes and in the surrounding flow. From these measurements the flow structure is described at stations immediately behind the rotor and one chord downstream. Some dominant features of the flow just behind the rotor are large radial velocity components, large static pressure fluctuations near the blade wakes, and definite unsteadiness (in rotor coordinates) of the wakes. The wake behavior one chord downstream is described in terms of the effect of the strong mean swirl on the behavior of shear disturbances. In the outer portion of the annulus, where the mean flow approximates a solid body rotation, a strong, persistent oscillatory flow is found with 16 periods in the circumference as roughly predicted by theory. In the inner portion of the annulus the disturbances attenuate axially. September 1975; Includes bibliographical references (page 6-8)

Wake behavior downstream of a transonic compressor rotor

Tue, 01 Jan 1974 00:00:00 GMT

Wake behavior downstream of a transonic compressor rotor Stephens, Harry Elias Measurements of the static pressure resolved in radial and axial directions downstream of a transonic compressor rotor have been harmonically analyzed. The investigation is part of a program addressing the behavior of disturbances in strongly swirling flows. There is evidence that the wakes are radially outward and carry static pressure disturbances with them. The harmonic analysis of the static pressure measured at the tip radius indicates a shift toward higher frequency of the peak in the power spectrum as the flow proceeds downstream. It is suggested that this increasing frequency associated with the power peak results from the wakes being convected downstream in a fluid which has increasing swirl velocity as it travels away from the rotor. August 1974; Also issued as: Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1974; Includes bibliographical references (page 64)

Attenuation of sound in lined circular ducts

Wed, 01 Jan 1975 00:00:00 GMT

Attenuation of sound in lined circular ducts Cho, Young-chung; Ingard, K. Uno In the previous report, we have used approximate expressions for the wall impedance for the discussion of sound attenuation in lined circular ducts. For instance, Eq. (2.5) has been used for the wall impedance of a circular duct lined with a resonator with a resistive screen, and Eq. (2.8) for a circular duct lined with a porous material. If the ratio of the duct radius to the liner thickness (D/L) is large and the sound frequency is large, the impedance given in Eq. (2.5) or Eq. (2.8) is a good approximation for a lined circular duct. However, when either one of these conditions is not fulfilled, the radial spread of the wave in the liner imposes some effects on the sound attenuation, whereas no wave spreading takes place in the liner of a rectangular duct. In this addendum we derive expressions for the wall impedance of lined circular ducts, accounting for the cylindrical spreading of the waves within the liner. The assumption of a locally reacting surface is still made. On the basis of the impedance thus obtained, the attenuation characteristics of a circular lined duct are computed for a wide range of parameters. April 1975; This is an addendum to Gas Turbine Laboratory Report No. 119. --Preface