High frequency spectra of three single-stage, high-speed compressor rotors were analyzed for behavior indicative of rotating stall. The compressor rotors were straight, backward swept, and forward swept. Power integrated over time showed promise as a prestall warning pararmeter. For the straight and swept back rotors, stall warning times varied from several seconds to 0.5 seconds. The forward swept rotor however, was difficult to characterize since surge played an important role in the stalling characteristic and stall appeared to originate at the hub of the rotor, away from pressure transducers located on the casing.

This paper presents a study of the dynamics for a single-stage, high-speed compressor core, specifically, the NASA Lewis rotor stage 35. Due to the overall blading design for this advanced core compressor, each stage has considerable tip loading and higher speed than most compressor designs, thus the compressor operates closer to stall. Due to the operation of this compressor close to the stall line, it is important to quickly predict the onset of stall. The onset of rotating stall is explained as bifurcations in the dynamics of axial compressors. Data taken from the compressor during a rotating stall event is analyzed. Through the use of a dimension analysis technique, the attractor dimension is determined during the bifurcations leading to rotating stall. The intent of this study is to examine the behavior of precursive stall events so as to predict the entrance into rotating stall. This information may provide a better means to identify, avoid, or control the catastrophic event of rotating stall formation in high-speed compressor cores.

This report describes the experiemental design and validation of sensing and actuation hardware to be incorporated into a NASA Lewis Research Center high-speed compressor test rig. The purpose of the control-augmented rig will be to investigate the dynamics of rotating stall in high-speed compressors, and to demonstrate stabilization of the perturbations which lead to rotating stall and surge. The overall experimental design is first described. Then the design of jet injection actuation is presented, including mechanical/fluid mechanical design rules, bandwidth limitations imposed by the electromagnetic valve, and by the fluid mechanics, and experimental validation of the actuation system. Specialized probes for 3D high-response flow measurements are then discussed, along with experimental validation of their performance. Finally, procedures for modeling and measurement of aerodynamic oscillation modes will be described.

A numerical technique capable of simulating blade-scale compression system flow instabilities over time-scales spanning tens of rotor revolutions is presented. Simulations of stall inception, growth to fully developed rotating stall, and evidence for hysteresis, secondary instabilities, and other nonlinear phenomena are presented. Signal processing techniques for flow asymmetry characterization are discussed in the context of obtaining low-order representations of the flow disturbances with the ultimate goal of active stall suppression.

This paper presents a novel eigenstructure assignment approach for sythesizing robust fault detection and isolation (FDI) systems with known inputs. After formulating the FDI problem in eigenstructure assigment, we proceed to develop a parametric characterization of all allowable eigenspaces for disturbance decoupling to achieve robust fault detection. In addition to the structured uncertainties, the robustness of the diagnostic observer to unstructured modeling errors is discussed. A numerical algorithm is further proposed to suppress the effects due to the unstructured uncertainties. The overall robustness of the diagnostic strategy is verfied through simulation studies on jet engine systems.

In this paper, we discuss a computational approach to sensor/estimator design for feedback control of fluid dynamic systems. This approach is based on combining minmax compensator design with piecewise constant approximations of 'optimal feedback gains'. A driven cavity flow control problem is presented to illustrate the idea and to demonstrate the feasibility of this approach.

We demonstrate a control strategy for increasing the mean passage time between 'bursts' of a perturbed heteroclinic attractor. Perturbed heteroclinic attractors were proposed by as a model for the dynamics of coherent structures in the turbulent wall layer. It is observed that most of the turbulent drag is produced by 'bursts' in the dynamics of coherent structures, hence increasing the mean time between bursts corresponds directly to drag reduction. The control is actuated by perturbations of the wall topography. We show how one may derive the form of the control input in dynamical systems phase space from the time-dependent boundary perturbation. This is done through a conformal change of coordinates. Assuming that the magnitude of the control input is small, we formulate a robust strategy for the increase of the passage time close to the hyperbolic fixed points of the attractor. We demonstrate the strategy on a 1D PDE, the Kuramoto-Sivashinsky equation. Initial simulations show an increase of 7% in the mean passage time. Our results are global nonlinear control results for a system exhibiting nontrivial dynamics. We further present initial results on estimation and control using partial system observations. In the application at hand, we will have a limited amount of information derived from wall-mounted sensors, thus requiring estimation models. We show results of control using and extended Kalman filter. The control algorithm used yields excellent results including a complete suppression of bursting when large control forces are introduced.

This paper is concerned with active control of discrete frequency noise generated by subsonic blade rows. Airfoil surface oscillating actuators are used to generate control propagating pressure waves. These control waves interact with the blade row interaction generated propagating acoustic waves, thereby, in principle canceling the acoustic waves and thus the far-field discrete frequency tones. A mathematical model is developed based on the LinSub flat plate unsteady aerodynamic analysis to investigate the feasibility of this technique and to evaluate actuator requirements. Then a series of experiments are described directed at both verifying the basic math model assumptions, and investigating the fundamentals of this active discrete frequency noise control technique.

Recent advances in room-temperature tunable diode lasers, fiberoptic beam transport, and sensitive detection strategies now permit in-stream sensing of numerous parameters relevant to aeropropulsion monitoring and control. These include density measurements of important flow constituents such as O₂, H₂O, CO₂, and NO_x. Based on path-averaged absorption measurements, the basic density measurements can be expanded to include other gasdynamic properties such as temperature and velocity. Simultaneous, multiparameter measurements allow determination of high order system parameters such as mass flux and thrust continuously and in real-time. This paper describes several sensor development efforts, exhaust mass flux, and emissions monitoring. Example measurements from laboratory configurations are presented along with performance projections for test-stand and flight systems. Integration issues with full-scale hardware and control opportunities are also discussed.

A high-frequency response dual hot-wire aspirating probe which can measure the time- resolved stagnation temperature and pressure in an unsteady high-speed gas flow is described. The probe consists of two coplanar constant temperature hot wires at different overheat ratios, operated in a channel with a choked exit. Thus, the constant Mach number by the wires is influenced only by free-stream total temperature and pressure. The probe was used in three separate experiments. The experiments include measurements of the flowfield at the exit of transonic fans during steady-state operation, measurements of turbulence intensity in the wake of a transonic turbine cascade, and measurements of the exit flowfield in a transonic core- compressor during rotating stall. The paper describes the performance of the probe in all three experiments. Brief descriptions of the experiments are given and exemplary data are presented.

A strategy for optimizing the performance of a combustor with respect ot volumetric heat release (increase) and pressure fluctuations (decrease) has been developed. This strategy utilized actuation and sensing techniques that simultaneoulsy control and measure volumetric heat release and pressure fluctuations. Combustor performance is explicitly defined in terms of a cost function that is a weighted combination of the mean volumetric heat release and the rms pressure fluctuation level. The control strategy performs an online minimization of the cost function by continuously seeking the optimal combination of static actuator settings and subsequently maintaining cost at a minimum when the combustor is subject to unknown inlet condition changes. The adaptivity of the strategy has been experimentally tested with unknown inlet condition changes such as flow disturbances, changes in equivalence ratio, and changes in inlet condition changes, and to be effective at finding the new optimal actuator settings that reminimize the cost function for large and small inlet condition changes.

This paper presents preliminary results on the use of low flow, high momentum, pulsed air injectors to control the onset of stall in a low-speed, axial flow compressor. By measuring the unsteady pressures in front of the rotor, the controller determines the magnitude and phase of a stall cell and controls the injection of air in front of the rotor face. Initial experimental results have verified that controller slightly extends the stall point of the compressor and virtually eliminates the hysteresis loop normally associated with stall. An explanation of this effect is proposed based on the quasi-steady effects of air injection on the compressor characteristic curve.

The paper applies Selective Modal Analysis (SMA) to a simple linearized model of the dynamics of a multistage axial flow compressor-throttle system. The technique allows quantification of the degree of participation of each system variable in system modes. This feature allows determination of the participation of each stage in an overall compression system mode. The main application is to determine the stage(s) most involved in a mode of instability, with the associated implications for control design.

Results are presented from an experiment designed to actively suppress a rotating stall condition in a low speed centrifugal compressor. The control system uses a circumferential array of twelve air injection jets located in the endwall of the compressor inlet. The system is able to suppress the eruption of a one-cell rotating stall condition in the compressor during stall initiation, but the eruption of high order stall conditions limits stable operating range extension. The controller is also capable of destabilizing the compressor and accelerating the eruption of the first spatial mode.

A plant model is required to design and develop a control system where the device to be controlled is a gas turbine engine. The desirable plant model is a wide-range, high-fidelity model which incorporates the fluid dynamic behavior of the engine internal flows. Such a model can predict the behavior of the engine in stable and unstable regimes; is easily altered to include design changes; and is suitable for the investigation of active controls for preventing instabilities and optimizing performace, especially in aircraft engines in connection with 'intelligent' engine concepts. Models of this kind are presently available and under continuing development. Both compression system and engine models may be constructed as one- or multi-dimensional flow simulations. When such models are linked with control models, a powerful analysis and development tool is produced. The usual control strategies for the stable, normal operating range may be investigated, as well as means to prevent and recover from unstable operation. The mathematical theory of detailed dynamic models is described and examples of specific simulations are given. The application of conventional and advanced control strategies to the simulation is shown by discussion and examples. The advantages of such model-controller combinations are illustrated.

A 0-450 Hertz bandwidth, voice coil actuated, proportional sleeve valve is designed to modulate air mass flow by controlling the throat area of a choked flow. The valve was designed to deliver a mass flow of 0.072 kg/s with a maximum valve throat area of 41 mm², a 689 kPA pressure difference across the valve, and 20 degree(s)C, air supply. The valve was developed with inexpensive, off-the-shelf components for use in ground-based forced response testing of compression systems. The design and operation of the valve are discussed and experimental test data of a prototype valve and air injector are compared to a mathematical model. Implementation of a set of eight of these valves in the compression system of a jet engine is discussed.

The problem of stabilization of plane Poiseuille flow is considered using techniques employed in control system analysis. The linearized Navier-Stokes equations are converted into control theoretic models using a Galerkin method. It is shown that control theoretic models provide a rich tool for designing control systems to stabilize flow as well as for understanding the physics of controlled transitional flows. Transfer function models provide information of optimal sensor locations and sensor types. State variable models provide information on how each mode, both stable and unstable, is affected by feedback control. State variable models can also provide insight into transitional fluid flow phenomena such as 'bypass' transition. In addition, control theory concepts such as observability and controllability are used to explain possible pitfalls in flow control. Numerical simulations are performed to validate the controllers designed using the control theoretic models.

Aeroengines operate in regimes for which both rotating stall and surge impose low flow operability limits. Thus, active control strategies designed to enhance operability of aeroengines must address both rotating stall and surge as well as their interaction. In this paper, a nonlinear control strategy is designed based on an analytical model to achieve simultaneous active control of rotating stall and surge in an axial flow compression system with relevant dynamics representative of modern aeroengines. The controller is experimentally validated on a 3-stage low-speed axial flow compression system. This rig is dynamically scaled to replicate the interaction between rotating stall and surge typical of modern aeroengines, and several experimental results are presented for this rig. For actuation, the control stategy utilizes a single plenum bleed valve with bandwidth on the order of the rotor frequency. For sensing, measurements of the circumferential asymmetry and annulus-averaged unsteadiness of the flow through the compressor are used. Experimental validation of simultaneous control of rotating stall and surge with minimal sensing and actuation requirements is viewed as an important step towards applying active control to enhance operability of compression systems in modern aeroengines.