During a ten-stage compressor rig test conducted at Wright-Patterson AFB, several instances of compressor surge were observed. While surge is known to occur in high-speed multi-stage compressors, very little transient data pertaining to such events exists in the open literature, exclusive of engine data. In an attempt to make more data of this type available to researchers, surge data from the ten-stage compressor test is presented and analyzed in this thesis. Graphical presentation and data analysis techniques are employed in an effort to characterize the surge behavior of this compressor. Furthermore, the predictions of a computer-based transient compressor model are compared to the data for study. In the course of reviewing the data included in this thesis, certain abnormalities were noted in the overall behavior of this compressor. During testing, several researchers found that the speed boundary between surge and rotating stall occurred between 80% and 81 % corrected design rotor speeds. 1hls boundary did not change when the compressor discharge volume was increased or decreased. This seemed to contradict accepted theory, which predicts a shift in the surge/rotating stall boundary with discharge volume changes. An investigation into the possible causes of this phenomenon was conducted as part of this thesis. Several theories were explored, including the possibility of excess volume communicating with the compressor during instability. Although the excess volume theory could not be proven, it remains the most likely cause of the usual surge/rotating stall boundary behavior. / Master of Science

An experimental investigation on an uncooled choked plasma torch using hydrogen/argon mixtures as a propellant was conducted. This low-power plasma torch was designed to be used as an ignitor and flameholder in supersonic combustion. The anode and cathode were made of two-percent thoriated tungsten, and no cooling system was required. Sonic flow through the nozzle was obtained by using a small throat diameter (0.813 mm). The plasma torch can operate stably over a wide range hydrogen/argon mixtures at power levels of 500 to 2000 W. Voltage-current characteristics of the arc are presented for discharge currents ranging from 5 to 40 A and for various flow rates and mixture fractions. The electrical input power is found to be a linear function of the hydrogen flow for a constant argon flow and for a current of 20 A. Measurements with a calorimeter reveal that the thermal efficiency, defined as the rate of increase of total enthalpy of the gas flowing through the torch divided by the electrical input power, is about 88 percent. / Master of Science

High-response pressure measurements of a high-speed, 10- stage, axial-flow compressor operating in rotating stall are analyzed. Procedures used to digitize analog voltages and calibrate pressure transducers are presented. From total and static pressures measured at the exit of the test compressor, stall cell Mach number distributions are calculated and used to study the effects of discharge throttle levels and variable vane changes on the 10th-stage rotating stall cells. Results indicate that significant transition zones exist between the reverse flow and peak Mach number of the stall cell cycle. Since the axial Mach numbers of the stall cell cycle are constantly changing, the amount of leading and trailing edge transition zones and fully unstalled flow zones are not easily defined. A method is devised to approximate the different flow zone ranges and correlate them to in-stall pressure characteristic behavior of the 10th stage of the test compressor. Changes in the time-averaged pressure characteristics are found to correlate with changes in the rotating stall flow zones. A lower pressure coefficient appears to correspond to an increase in the ratio of trailing to leading edge transition zone size and the average transition zone size. Results also suggest that recovery hysteresis in the test compressor is characterized by reverse flow in the rotating stall cell. / Master of Science

Dynamic compression system response is a major concern in the operability of aircraft gas turbine engines. Computer models have been developed to predict compressor response to changing operating conditions. These models require a knowledge of the steady state operating characteristics as inputs, which limits the ability to use them as predicting tools. The full range of dynamic axial flow compressor operation spans forward and reversed flow conditions. A model for predicting the wide flow range characteristics of axial flow compressor stages has been developed and a parametric study of the effect of changing design variables on steady state performance has been conducted. This model was applied to a 3-stage, low speed compressor with very favorable results and to a 10-stage, high speed compressor with mixed results. Conclusions were made regarding the inception of stalling and the effects associated with operating a stage in a multistage environment. It was also concluded that there are operating points of an isolated compressor stage that are not attainable when that stage is operated in a multi-stage environment. A stage located in a multi-stage environment can also operate at points which cannot be reached when the stage is operated in isolation. / Master of Science

No description available.

A low-speed single-stage axial-flow compressor rig was configured for tests with variable-length upstream casing struts. Two axial strut lengths were tested for their effect on the inlet flow field and stalling characteristics of the compressor under clean and distorted inlet conditions. Data were obtained from one stationary and six blade-mounted high-response pressure transducers, and from a circumferentially- traversed directional probe. The distorted inlet flow field was changed significantly by the longer struts, which effectively sectored the inlet. Deep stall was not observed for this configuration. Rather, the stalling characteristic was changed to what is termed progressive rotating stall, in which six stable stall cells and significant increases in static pressure rise were noted. A new technique for determining the number of rotating stall cells from the frequency content of stationary and rotating reference frame signals is presented. Blade pressure response as a function of aerodynamic loading is detailed. / M.S.

no abstract provided by author / Master of Science

A six-stage axial-flow compressor with a 550 feet per second tip speed and a flat operating characteristic (constant stagnation-pressure ratio at constant speed over the operating range of the compressor) was designed and tested. The design theory and test results are presented in this thesis. It was designed for a constant power input per pound of flow regardless of mass flow. The design specific weight flow was 21.1 pounds per second per square foot of frontal area with an atmospheric discharge at an overall stagnation-pressure ratio of 3.25 and an inlet hub-tip radius ratio of 0.7. In order to reach design conditions the blade setting angles were reset and the machining notches at the root of the first three rotor blades were filled. In an attempt to increase the flat operating range of the compressor, the blade setting angles of the first two stages were increased and those of the last two stages were decreased. Also, the solidity of the first rotor was decreased. / Master of Science

A preliminary theoretical and experimental investigation of the effect of velocity diagram parameters on inlet total-pressure distortions through a single-stage subsonic axial-flow compressor for incompressible flow has been conducted. The wake of 1/4-inch diameter rod, measured both upstream and downstream of a rotor, has been compared for various velocity diagrams. The measured downstream wake was also compared with the downstream wake estimated by using a derived equation subject to the assumption that (a) the undistorted and distorted flows enter the rotor with the same absolute direction, (b) the static pressures of the undistorted and distorted flows are equal at the inlet and also at the exit of the rotor, (c) the increase of blade angle of attack in the distorted flow is not sufficient to cause blade stall, and (d) the distorted flow is turned in the rotor passage to the same exit direction, relative to the rotor, as the undistorted flow.> / Master of Science

No description available.