Real-Time Adaptive Feedback Control of a Supersonic Dual-Stream Jet

Melissa Yeung, Yiyang Sun · 2025

Read Real-Time Adaptive Feedback Control of a Supersonic Dual-Stream Jet by Melissa Yeung, Yiyang Sun (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-03236.txt.

Introduction

A supersonic jet can sing a loud, steady tone. Melissa Yeung and Yiyang Sun work on controlling that tone in this 2025 work. Their jet has two streams, a fast core and a slower bypass, which mix into a shear layer.

Shed vortices in that layer cause the high-frequency tone. The authors apply adaptive control in real time to tame the flow. Feedback that adjusts on the fly suits flows that change fast. Quieter, steadier jets matter for aircraft engines and for test rigs on the ground. Real-time feedback suits jets whose tone changes with thrust and flight conditions.

Catalog record

CREATORMelissa Yeung, Yiyang Sun
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE63 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-03236

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Real-Time Adaptive Feedback Control of a Supersonic Dual-Stream Jet Melissa Yeung, Yiyang Sun Abstract Adaptive control is applied to a supersonic dual-stream jet flow comprised of Mach 1.6 core and Mach 1.0 bypass streams that mix to form a supersonic shear layer. The vortices shed are the source of a high-frequency tone that persists throughout the flow. The intricate flow dynamics motivates the need for an elaborate and efficient actuation system to suppress the tone and weaken the propagating shock train. The present work utilizes online dynamic mode decomposition, which estimates the system dynamics as a locally linear evolution. Snapshot matrices are constructed using sensor measurements, facilitating economical and real-time computations, which are continuously updated and used in a feedback control model. Adaptive control is found to efficiently target the resonant tone with little disturbance to the mean features. The framework is not sensitive to sensor placements, enabling actuator design under physically realizable spatial locations in practical implementation. To reflect physical limitations, constraints are imposed on the controller model. It is found that the restricted controller yields greater vortex suppression due to repeated transitory stabilization of the shear layer instability. Statistical analysis reveals intermittent low-pressure events are responsible for the characteristic frequency, which are largely suppressed by adaptive feedback control. Nomenclature A = area \bm{A} = discrete-time state transition matrix \mathcal{A} = continuous-time state transition matrix \bm{B} = discrete-time control input matrix \mathcal{B} = continuous-time control input matrix C_{\mu} = coefficient of momentum C_{T} = coefficient of thrust c = speed of sound D = computational domain length D_{h} = hydraulic diameter of rectangular jet exit f = frequency F_{T} = thrust force \bm{K} = LQR feedback control gain matrix M = Mach number m = total number of snapshots n = state dimension P = pressure P_{xx} = autospectral density p = number of control inputs q = dynamic pressure R = specific gas constant Re = Reynolds number St = Strouhal number T = temperature t = time u = streamwise velocity \bm{u} = control input W = spanwise splitter plate width w = online DMD window size \bm{X} = snapshot matrix \bm{x} = state variable \bm{Y} = shifted snapshot matrix \delta_{\text{SP}} = splitter plate thickness \delta\tau = event temporal extent \gamma = specific heat ratio \Lambda = event amplitude \mu = dynamic viscosity \Omega = online DMD window weighting factor \phi = signal reconstruction \psi = angle of actuation \rho = density \tau = event time \xi = event definition 1 Introduction Modern supersonic engines designed for next-generation aircraft have become increasingly complex over the last few decades to meet demanding flight requirements while maintaining propulsive efficiency. A variable-cycle engine proposed by Simmons [ 28 ] is shown schematically in figure 1 ( a ). The arrangement features a rectangular cross-section nozzle to facilitate seamless integration into an airframe. Additionally, a unique independently modulated cooled bypass stream is designed to continuously adapt to the engine inlet demands while acting as a heat sink to dissipate aircraft heat loads. In the present work, we analyze and control the flow at the engine exit, where the core and fan streams are assumed to be fully mixed before entering the nozzle, and are simply referred to as the core stream. The nozzle geometry, shown in figure 1 ( b ), contains a single-sided expansion ramp (SERN), a splitter plate separating the two canonical flows, and an aft-deck plate to mimic the structure of an aircraft. Figure 1: ( a ) Illustration of a generic three-stream engine [ 28 ] . ( b ) Nozzle configuration. The engine configuration shown in figure 1 ( b ) has been the topic of study over the past decade with joint experimental [ 21 , 1 , 11 , 18 ] and computational efforts [ 29 , 10 , 41 , 40 , 33 ] . A signature high-frequency tone of approximately 34 kHz is detected in the far-field acoustics, which originates from the instability generated by the mixing of the core and bypass streams. Various flow control techniques have been applied with the goal of suppressing the inherent instabilities that dominate this flow, ranging from passive [ 11 , 10 ] to active control methods [ 18 ] . Most notably, Gist [ 11 ] and Doshi [ 10 ] experimentally and computationally introduced a spanwise wavenumber along the trailing edge of the splitter plate, as this region was determined to be most receptive to external forcing [ 10 , 40 , 33 ] . This induced streamwise vorticity that broke down the coherent structures responsible for the high-frequency tone. The success of this passive control technique prompted an investigation into active flow control techniques [ 18 , 40 ] , with the ultimate goal of achieving the same control outcomes through an adaptive sensor-actuator control system, to enable the control at a wider range of operating conditions. Kelly [ 18 ] experimentally introduced an array of steady-blowing micro-jet actuators along the aft-deck that imitated the troughs of the sinusoidal wave used in Gist’s wavy splitter plate. Kelly was able to influence the location of the shock structures, and in some instances split the shock structures in two segments. Yeung et al. [ 40 ] computationally introduced the same micro-jets at various locations on the surfaces of the splitter plate trailing edge and found the strength of the shocks could be influenced by control, consequently attenuating the resultant shock-induced flow separation. While the aforementioned studies detail a comprehensive parametric campaign of steady-blowing actuators and provide significant insights into the actuator location, angle, and mass flow effects, such an approach may not be energy efficient from a practical standpoint and face the risk of failure during off-design flight conditions. To address this gap, the present work investigates a data-driven approach, where a modal decomposition method is used to build a closed-loop feedback control system. The reduced-order model of the system is continuously (i.e., online) updated using measurements obtained from local probes in the flow. This approach allows the flow to be driven towards a desirable state with minimal energy input compared to open-loop approaches [ 8 ] . Various data-driven modal decomposition techniques [ 31 ] have been historically applied to a multitude of fluid flow problems [ 32 , 16 ] . These techniques remain an attractive post-processing tool due to their ability to decompose turbulent flow fields into coherent modes that highlight physically important flow features. The present work investigates the use of online dynamic mode decomposition (DMD) [ 42 ] in which the DMD algorithm is efficiently computed in real-time and subjected to rank–1 updates, allowing a time-varying state-space model to be generated. This model more accurately reflects the dynamics of the system as it evolves under the influence of external forcing, as opposed to maintaining a time-invariant dynamical model, such as one based on a time-averaged flow state, for feedback control. In these instances, the time-invariant model no longer accurately represents the system once subjected to forcing. Online DMD has been used to suppress a laminar separation bubble using measurements from unsteady surface pressure transducers [ 8 ] , and holds promise for adaptive control of supersonic flows. In the

Read the complete text