-
A Standing-Wave Model for the Low-Frequency Dynamics of a Turbulent Separation Bubble
Authors:
Lukas M. Fuchs,
Ben Steinfurth,
Jakob G. R. von Saldern,
Julien Weiss,
Kilian Oberleithner
Abstract:
This study investigates the low-frequency dynamics of a turbulent separation bubble (TSB) forming over a backward-facing ramp, with a focus on very-large-scale coherent structures associated with the so-called 'breathing motion'. Using time-resolved particle image velocimetry (PIV) in both streamwise and spanwise planes, we examine the role of sidewall confinement, an aspect largely overlooked in…
▽ More
This study investigates the low-frequency dynamics of a turbulent separation bubble (TSB) forming over a backward-facing ramp, with a focus on very-large-scale coherent structures associated with the so-called 'breathing motion'. Using time-resolved particle image velocimetry (PIV) in both streamwise and spanwise planes, we examine the role of sidewall confinement, an aspect largely overlooked in previous research. Spectral proper orthogonal decomposition (SPOD) of the streamwise velocity field reveals a dominant low-rank mode at very low Strouhal numbers ($St =$ 0.01-0.05), consistent with prior observations of TSB breathing. Strikingly, the spanwise-oriented PIV data uncover a previously unreported standing wave pattern, characterized by discrete spanwise wavenumbers and nodal/antinodal structures, suggesting the presence of spanwise resonance. To explain these observations, we construct a resolvent-based model that imposes free-slip conditions at the sidewall locations by superposing left- and right-traveling three-dimensional modes. The model accurately reproduces the spanwise structure and frequency content of the measured SPOD modes, demonstrating that sidewall reflections lead to the formation of standing wave-like patterns. The analysis further indicates that the observed dynamics are caused by the lift-up mechanism driving the formation of large-scale streaks that span the domain width. Our findings highlight the critical influence of spanwise boundary conditions on the selection and structure of low-frequency modes in TSBs. This has direct implications for both experimental and numerical studies, particularly those relying on spanwise-periodic boundary conditions, and offers a low-order framework for predicting sidewall-induced modal dynamics in separated flows.
△ Less
Submitted 30 May, 2025;
originally announced May 2025.
-
Amplifying vortex shedding for energy harvesting with active flow control
Authors:
Varun Varma Jaganath,
Ben Steinfurth
Abstract:
Energy harvesting from vortex-induced vibrations is a promising technology that relies on the vibrations of bluff bodies due to vortex shedding. Increasing the vibration amplitude at a given free stream kinetic energy is therefore equivalent to enhancing the efficiency of the harvesting device. In this study, we assess the potential of alternate slot blowing to amplify force fluctuations. Pressuri…
▽ More
Energy harvesting from vortex-induced vibrations is a promising technology that relies on the vibrations of bluff bodies due to vortex shedding. Increasing the vibration amplitude at a given free stream kinetic energy is therefore equivalent to enhancing the efficiency of the harvesting device. In this study, we assess the potential of alternate slot blowing to amplify force fluctuations. Pressurized air is ejected alternatingly from the top and bottom parts of the cylinder. Through experimentation in a low-speed wind tunnel ($Re=8,000$), we show that the magnitude of lift fluctuations can be enhanced by up to a factor of three compared to the unforced flow when the actuation is aligned with the natural vortex shedding frequency. Velocity field measurements indicate that this is caused by strong streamline bending whereas, at a higher forcing frequency, vortex shedding is suppressed. The results presented in this article suggest that a significant increase in the dynamic load acting on a cylinder can be achieved with carefully chosen active flow control parameters, thereby promoting future energy harvesting applications.
△ Less
Submitted 25 December, 2024;
originally announced December 2024.
-
Optimizing pulsed blowing parameters for active separation control in a one-sided diffuser using reinforcement learning
Authors:
Alexandra Müller,
Tobias Schesny,
Ben Steinfurth,
Julien Weiss
Abstract:
Reinforcement learning is employed to optimize the periodic forcing signal of a pulsed blowing system that controls flow separation in a fully-turbulent $Re_θ= 1000$ diffuser flow. Based on the state of the wind tunnel experiment that is determined with wall shear-stress measurements, Proximal Policy Optimization is used to iteratively adjust the forcing signal. Out of the reward functions investi…
▽ More
Reinforcement learning is employed to optimize the periodic forcing signal of a pulsed blowing system that controls flow separation in a fully-turbulent $Re_θ= 1000$ diffuser flow. Based on the state of the wind tunnel experiment that is determined with wall shear-stress measurements, Proximal Policy Optimization is used to iteratively adjust the forcing signal. Out of the reward functions investigated in this study, the incremental reduction of flow reversal per action is shown to be the most sample efficient. Less than 100 episodes are required to find the parameter combination that ensures the highest control authority for a fixed mass flow consumption. Fully consistent with recent studies, the algorithm suggests that the mass flow is used most efficiently when the actuation signal is characterized by a low duty cycle where the pulse duration is small compared to the pulsation period. The results presented in this paper promote the application of reinforcement learning for optimization tasks based on turbulent, experimental data.
△ Less
Submitted 10 December, 2024;
originally announced December 2024.
-
Automatic extraction of wall streamlines from oil-flow visualizations using a convolutional neural network
Authors:
Jonas Schulte-Sasse,
Ben Steinfurth,
Julien Weiss
Abstract:
Oil-flow visualizations represent a simple means to reveal time-averaged wall streamline patterns. Yet, the evaluation of such images can be a time-consuming process and is subjective to human perception. In this study, we present a fast and robust method to obtain quantitative insight based on qualitative oil-flow visualizations. Using a convolutional neural network, the local flow direction is p…
▽ More
Oil-flow visualizations represent a simple means to reveal time-averaged wall streamline patterns. Yet, the evaluation of such images can be a time-consuming process and is subjective to human perception. In this study, we present a fast and robust method to obtain quantitative insight based on qualitative oil-flow visualizations. Using a convolutional neural network, the local flow direction is predicted based on the oil-flow texture. This was achieved with supervised training based on an extensive dataset involving approximately one million image patches that cover variations of the flow direction, the wall shear-stress magnitude and the oil-flow mixture. For a test dataset that is distinct from the training data, the mean prediction error of the flow direction is as low as three degrees. A reliable performance is also noted when the model is applied to oil-flow visualizations from the literature, demonstrating the generalizability required for an application in diverse flow configurations.
△ Less
Submitted 10 December, 2024;
originally announced December 2024.
