A machine learning model is developed to establish wake patterns behind oscillating foils for energy harvesting. The role of the wake structure is particularly important for array deployments of oscillating foils since the unsteady wake highly influences the performance of downstream foils. This work explores 46 oscillating foil kinematics, with the goal of parameterizing the wake based on the input kinematic variables and grouping vortex wakes through image analysis of vorticity fields. A combination of a convolutional neural network with long short-term memory units is developed to classify the wakes into three classes. To fully verify the physical wake differences among foil kinematics, a convolutional autoencoder combined with [Formula: see text]-means++ clustering is used to reveal four wake patterns via an unsupervised method. Future work can use these patterns to predict the performance of foils placed in the wake and build optimal foil arrangements for tidal energy harvesting.
- Award ID(s):
- 1921594
- NSF-PAR ID:
- 10349017
- Date Published:
- Journal Name:
- AIAA AVIATION 2022 Forum
- Page Range / eLocation ID:
- 2022-4140
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
-
Numerical studies are presented on the propulsive performance and vortex dynamics of multiple hydrofoils pitching in an in-line configuration. The study is motivated by the quest to understand the hydrodynamics of multiple fin–fin interactions in fish swimming. Using the flow conditions (Strouhal and Reynolds numbers) obtained from a solitary pitching foil of zero net thrust, the effect of phase differences between neighboring foils on the hydrodynamic performance is examined both in position-fixed two- and three-foil systems at Reynolds number Re = 500. It is found that the threefoil system achieves a thrust enhancement up to 118% and an efficiency enhancement up to 115% compared to the two-foil system. Correspondingly, the leading-edge vortex (LEV) and the trailing-edge vortex (TEV) of the hindmost foil combine to form a ‘2P’ wake structure behind the three-foil system with the optimal phase differences instead of a ‘2S’ wake, a coherent wake pattern observed behind the optimal two-foil system. The finding suggests that a position-fixed three-foil system can generate a ‘2P’ wake to achieve the maximum thrust production and propulsive efficiency simultaneously by deliberately choosing the undulatory phase for each foil. When increasing Reynolds number to 1000, though the maximum thrust and propulsive efficiency are not achieved simultaneously, the most efficient case still produces more thrust than most of the other cases. Besides, the study on the effects of three-dimensionality shows that when the foils have a larger aspect ratio, the three-foil system has a better hydrodynamic performance, and it follows a similar trend as the two-dimensional (2D) foil system. This work aids in the future design of high-performance underwater vehicles with multiple controlled propulsion elements.more » « less
-
more » « less
In this study, we numerically investigate the effects of the tail-beat phase differences between the trailing fish and its neighboring fish on the hydrodynamic performance and wake dynamics in a two-dimensional high-density school. Foils undulating with a wavy-like motion are employed to mimic swimming fish. The phase difference varies from 0° to 360°. A sharp-interface immersed boundary method is used to simulate flows over the fish-like bodies and provide quantitative analysis of the hydrodynamic performance and wakes of the school. It is found that the highest net thrust and swimming efficiency can be reached at the same time in the fish school with a phase difference of 180°. In particular, when the phase difference is 90°, the trailing fish achieves the highest efficiency, 58% enhancement compared with a single fish, while it has the highest thrust production, increased by 108% over a single fish, at a phase difference of 0°. The performance and flow visualization results suggest that the phase of the trailing fish in the dense school can be controlled to improve thrust and propulsive efficiency, and these improvements occur through the hydrodynamic interactions with the vortices shed by the neighboring fish and the channel formed by the side fish. In addition, the investigation of the phase difference effects on the wake dynamics of schools performed in this work represents the first study in which the wake patterns for systems consisting of multiple undulating bodies are categorized. In particular, a reversed Bénard–von Kármán vortex wake is generated by the trailing fish in the school with a phase difference of 90°, while a Bénard–von Kármán vortex wake is produced when the phase difference is 0°. Results have revealed that the wake patterns are critical to predicting the hydrodynamic performance of a fish school and are highly dependent on the phase difference.
-
null (Ed.)Machine learning techniques have received attention in fluid dynamics in terms of predicting, clustering and classifying complex flow physics. One application has been the classification or clustering of various wake structures that emanate from blu˙ bodies such as cylinders or flapping foils, creating a rich diversity of vortex formations specific to flow conditions, geometry, and/or kinematics of the body. When utilizing oscillating foils to harvest energy from tidal or river flows, it is critical to understand the intricate and nonlinear relationship between flapping kinematics and the downstream vortex wake structure for optimal siting and operation of arrays. This paper develops a classification model to obtain groups of kinematics that contain similar wake patterns within the energy harvesting regime. Data is obtained through simulations of 27 unique oscillating foil kinematics for a total of 13,650 samples of the wake vorticity field. Within these samples three groups are visually labeled based on the relative angle of attack. A machine learning approach combining a convolutional neural network (CNN) with long short-term memory (LSTM) units is utilized to automatically classify the wakes into the three groups. The average accuracy on five test data subsets is 80% when the three visually labeled groups are used for classification. After analyzing the test subset with lowest accuracy, an update on the group division boundaries is proposed. With this update, the algorithm achieves an average accuracy of 90%, demonstrating that the three groups are able to discern distinct wake structures within a range of energy harvesting kinematics.more » « less
-
Experiments and computations are presented for a foil pitching about its leading edge near a planar, solid boundary. The foil is examined when it is constrained in space and when it is unconstrained or freely swimming in the cross-stream direction. It was found that the foil has stable equilibrium altitudes: the time-averaged lift is zero at certain altitudes and acts to return the foil to these equilibria. These stable equilibrium altitudes exist for both constrained and freely swimming foils and are independent of the initial conditions of the foil. In all cases, the equilibrium altitudes move farther from the ground when the Strouhal number is increased or the reduced frequency is decreased. Potential flow simulations predict the equilibrium altitudes to within 3 %–11 %, indicating that the equilibrium altitudes are primarily due to inviscid mechanisms. In fact, it is determined that stable equilibrium altitudes arise from an interplay among three time-averaged forces: a negative jet deflection circulatory force, a positive quasistatic circulatory force and a negative added mass force. At equilibrium, the foil exhibits a deflected wake and experiences a thrust enhancement of 4 %–17 % with no penalty in efficiency as compared to a pitching foil far from the ground. These newfound lateral stability characteristics suggest that unsteady ground effect may play a role in the control strategies of near-boundary fish and fish-inspired robots.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.2514/6.2022-4140
