Mohammad Samimy

Exploring strategies for closed-loop cavity flow control

One of the current three main thrust areas of the Collaborative Center of Control Science (CCCS) at The Ohio State University is feedback control of aerodynamic flows. Synergistic capabilities of the flow control team include all of the required multidisciplinary areas of flow simulations, low-dimensional and reduced-order modeling, controller design, and experimental integration and implementation of the components along with actuators and sensors. The initial application chosen for study is closed-loop control of shallow subsonic cavity flows. We have made significant progress in the development of various components necessary for reduced-order model based control strategy, which will be presented and discussed in this paper. Stochastic estimation was used to show that surface pressure measurements along with the reduced-order model based on flow-field variables can be used for closed-loop control. Linear controllers such as H ∞ , Smith predictor, and PID were implemented experimentally with various degrees of success. The results showed limitations of linear controllers for cavity flow with inherent nonlinear dynamics. Detailed experimental work further explored the physics and showed the highly non-linear nature of the cavity flow and the effects of forcing on the flow structure.

Analysis of the spectral relationships of cavity tones in subsonic resonant cavity flows

The understanding of the self-sustained flow-acoustic coupling mechanism in flows over shallow rectangular cavities is of great interest owing to its various practical applications. The ability to understand and predict the resonant frequencies in such flows has recently been advanced through contributions from signal processing theory and by viewing the Rossiter tones as the product of an amplitude modulation process between a fundamental aeroacoustic loop frequency (fa) and a modulating lower frequency. The results obtained using this approach applied to detailed and high-quality spectral data of shallow cavity flow over the Mach number range of 0.20–0.65 are presented and discussed. The new approach, while not a predictive technique, is used to clearly identify all the tones (Rossiter modes, their harmonics, and harmonics of fa) observed in the pressure spectra and to show relationships between the tones. The asymptotic growth with Mach number of fa and the small-step changes of the modulating lower fr...

Control of Subsonic Cavity Flows by Neural Networks - Analytical Models and Experimental Validation

Flow control is attracting an increasing attention of researchers from a wide spectrum of specialties because of its interdisciplinary nature and the associated challenges. One of the main goals of The Collaborative Center of Control Science at The Ohio State University is to bring together researchers from different disciplines to advance the science and technology of flow control. This paper approaches the control of subsonic cavity flow, a study case we have selected, from a computational intelligence point of view, and offers a solution that displays an interconnected neural architecture. The structures of identification and control, together with the experimental implementation are discussed. The model and the controller have very simple structural configurations indicating that a significant saving on computation is possible. Experimental testing of a neural emulator and of a directlysynthesized neurocontroller indicates that the emulator can accurately reproduce a reference signal measured in the cavity floor under different operating conditions. Based on preliminary results, the neurocontroller appears to be marginally effective and produces spectral peak reductions analogous to those previously observed by the authors using linearcontrol techniques. The current research will continue to improve the capability of the neural emulator and of the neurocontroller.

Closed-Loop Active Flow Control - A Collaborative Approach

The Collaborative Center of Control Science (CCCS) at The Ohio State University was founded very recently with funding from the Air Force Research Laboratory to conduct multidisciplinary research in the area of feedback control, with applications such as cooperative control of unmanned air vehicles (UAVs), guidance and control of hypersonic vehicles, and closed-loop active flow control. The last topic is the subject of this paper. The goal of this effort is to develop tools and methodologies for the use of closedloop aerodynamic flow control to manipulate the flow over maneuvering air vehicles and ultimately to control the maneuvers of the vehicles themselves. It is well known in the scientific community that this is a challenging task and requires expertise in flow simulation, low dimensional modeling of the flow, controller design, and experimental integration and implementation of these components along with actuators and sensors. The CCCS flow control team possesses synergistic capabilities in all these areas, and all parties have been intimately involved in the project from the beginning, a radical departure from the traditional approach whereby an experiment is designed and constructed, data are collected, a model is developed, and a control law is designed, i.e. the system is assembled for validation in a sequential fashion. The first problem chosen for study, control of the noise created by a shallow cavity placed in a flow, has specific relevance to the needs of the Air Force. For example, significant pressure fluctuations in an aircraft weapon bay can lead to structural damage to the air vehicle, to the stores carried in the cavity, and especially to the electronics carried onboard the stores. The team has been working together for a relatively short period of time. Nevertheless, significant progress has been made in the development of various components of the closed-loop cavity flow control problem. The paper will present and discuss the progress made to date and future plans.

Further Development of Feedback Control of Cavity Flow Using Experimental Based Reduced Order Model

*† ‡ § ** †† ‡‡ In our recent work we presented preliminary results for subsonic cavity flow control using a reduced-order model based feedback control derived from experimental measurements. The model was developed using the Proper Orthogonal Decomposition of PIV images in conjunction with the Galerkin projection of the Navier-Stokes equations onto the resulting spatial eigenfunctions. A linear-quadratic optimal controller was designed to reduce cavity flow resonance by controlling the time coefficient and tested in the experiments. The stochastic estimation method was used for real-time estimation of the corresponding time coefficients from 4 dynamic surface pressure measurements. The results obtained showed that the controller was capable of reducing the cavity flow resonance at the design Mach 0.3 flow, as well as at other flows with slightly different Mach number. In the present work we present several improvements made to the method. The reduced order model was derived from a larger set of PIV measurements and we used 6 sensors for the stochastic estimation of the instantaneous time coefficients. The reduced order model so obtained shows a better convergence of the time coefficients. This combined with the 6sensor estimation improves the control performance while using a scaling factor closer to the theoretically expected value. The controller also performed better in off design flow conditions.

An Experimental Study of Subsonic Cavity Flow - Physical Understanding and Control

We present the results of an experimental investigation that uses two different techniques for controlling a shallow cavity flow in the Mach number range 0.25-0.5. The first method is basically an open-loop design that relies on zero-net-mass forcing at an optimal frequency for suppressing the cavity flow resonance. The second method is a parallel-proportional with time delay controller, a linear control design that relies on real-time feedback from the flow to counteract the resonance. With properly tuned parameters, both methods are successful in reducing the cavity resonance for flows in the Mach number range explored. However the parallel-proportional controller exhibits a superior robustness with respect to departure of the Mach number from the design conditions, a signature of feedback control designs that are naturally more capable to handle changes of the open-loop plant. An additional benefit of the feedback control method is the lower power requirement to achieve comparable suppression of the resonance. An interpretation is presented of the physical mechanisms by which the optimal forcing frequency and the parallel-proportional with time delay controller reduce the cavity flow resonance. The results support the idea that the optimal forcing frequency control induces in the system a rapid switching between modes competing for the available energy that can be extracted from the mean flow. In the case of parallelproportional control mode switching is also observed which involves a larger range of frequencies and spreads more the extracted energy thus producing a flow with a quieter, more broadband spectral signature.

Modeling of subsonic cavity flows by neural networks

Influencing the behavior of a flow field is a core issue as its improvement can yield significant increase of the efficiency and performance of fluidic systems. On the other hand, the tools of classical control systems theory are not directly applicable to processes displaying spatial continuity as in fluid flows. The cavity flow is a good example of this and a recent research focus in aerospace science is its modeling and control. The objective is to develop a finite dimensional representative model for the system with appropriately defined inputs and outputs. Towards the goal of reconstructing the pressure fluctuations measured at the cavity floor, this paper demonstrates that given some history of inputs and outputs, a neural network based feedforward model can be developed such that the response of the neural network matches the measured response. The advantages of using such a model are the representational simplicity of the model, structural flexibility to enable controller design and the ability to store information in an interconnected structure.

Neural network-based modelling of subsonic cavity flows

A fundamental problem in the applications involved with aerodynamic flows is the difficulty in finding a suitable dynamical model containing the most significant information pertaining to the physical system. Especially in the design of feedback control systems, a representative model is a necessary tool constraining the applicable forms of control laws. This article addresses the modelling problem by the use of feedforward neural networks (NNs). Shallow cavity flows at different Mach numbers are considered, and a single NN admitting the Mach number as one of the external inputs is demonstrated to be capable of predicting the floor pressures. Simulations and real time experiments have been presented to support the learning and generalization claims introduced by NN-based models.

Influence of Stochastic Estimation on the Control of Subsonic Cavity Flow - A Preliminary Study

This work aims at understanding how the different elements involved in the feedback loop influence the overall control performance of a subsonic cavity flow based on reduced- order modeling. To this aim we compare preliminary and limited sets of experimental results obtained by modifying some relevant characteristics of the loop. Our results support the findings in the literature that use of quadratic stochastic estimation is preferable to the linear one for real-time update of the model parameters. They also seem to indicate the merit of using more than one time sample of the pressure for performing the real-time update of the model through stochastic estimation. The effect of using two different sets of pressure signals for the stochastic estimation also corroborates previous findings indicating the need for optimizing the number and the placement of the sensors used in the feedback control loop. Finally we observed that the characteristics of the actuator can alter significantly the overall control effect by introducing in the feedback loop additional, undesirable frequency components that are not modeled and hence controlled. A compensator for the actuator is currently being designed that will alleviate this problem thus enabling a clearer understanding of the overall control technique.

Reduced-Order Model-Based Feedback Control of Subsonic Cavity Flows — An Experimental Approach

The results of an ongoing research activity in the development and implementation of reduced-order model-based feedback control of subsonic cavity flows are presented and discussed. Particle image velocimetry data and the proper orthogonal decomposition technique are used to extract the most energetic flow features or POD eigenmodes. The Galerkin projection of the Navier-Stokes equations onto these modes is used to derive a set of ordinary nonlinear differential equations, which govern the time evolution of the modes, for the controller design. Stochastic estimation is used to correlate surface pressure data with flow field data and dynamic surface pressure measurements are used for real-time state estimation of the flow model. Three sets of PIV snapshots of a Mach 0.3 cavity flow were used to derive three reduced-order models for controller design: (1) snapshots from the baseline (no control) flow, (2) snapshots from an open-loop forced flow, and (3) combined snapshots from the cases 1 and 2. Linear-quadratic optimal controllers based on all three models were designed and tested experimentally. Real-time implementation shows a remarkable attenuation of the resonant tone and a redistribution of the energy into various modes with much lower energy levels.