Augmentation of piezoelectric thin-film flapping velocimetry turbulence strength detection via machine learning DOI Creative Commons

Ted Sian Lee,

Ean Hin Ooi, Wei Sea Chang

et al.

Physics of Fluids, Journal Year: 2025, Volume and Issue: 37(1)

Published: Jan. 1, 2025

Qualitatively evaluating the fundamental mechanical characteristics of square-fractal-grid (SFG)-generated turbulent flow using piezoelectric thin-film flapping velocimetry (PTFV) is rather time-consuming. More importantly, its sensitivity in detecting high-frequency, fine-scale fluctuations constrained by high-speed camera specifications. To reduce dependency on imaging future PTFV implementations, regression models are trained with supervised machine learning to determine correlation between piezoelectric-generated voltage V and corresponding local equivalent velocity fluctuation. Using tip deflection δ data as predictors responses, respectively, Trilayered Neural Network (TNN) emerges best-performing model compared linear regression, trees, support vector machines, Gaussian process ensembles trees. TNN from (i) lower quarter, (ii) bottom left corner, (iii) central opening SFG-grid provide accurate predictions insert-induced centerline streamwise cross-sectional lateral turbulence intensity root mean square-δ, average errors not exceeding 5%. The output predicted response, which considers small-scale across entire surface, better expresses integral length scale (38% smaller) forcing (270% greater), particularly at corner SFG where eddies significant. Furthermore, effectively captures occasional extensive excitation forces large-scale eddies, resulting a more balanced force distribution. In short, this study paves path for comprehensive expedited dynamics characterization detection via PTFV, potential deployment high Reynolds number flows generated various grid configurations.

Language: Английский

Enhancing property prediction and process optimization in building materials through machine learning: A review DOI Creative Commons
Konstantinos I. Stergiou, Charis Ntakolia,

Paris Varytis

et al.

Computational Materials Science, Journal Year: 2023, Volume and Issue: 220, P. 112031 - 112031

Published: Jan. 25, 2023

Analysis and design, as the most critical components in material science, require a highly rigorous approach to assure long-term success. Due recent increase amount of available experimental data, large databases now contain depth knowledge on important properties materials. The use this information, combined with Machine Learning (ML) solutions, can enhance materials’ manufacturing process efficiency. Indeed, ML predict properties, minimize time cost laboratory testing, well optimize processes. This paper aims give an up-to-date review literature how models are used buildings’ (thermal, mechanical, optical) production lines for: a) Phase Change Materials (PCMs), b) Thermoelectric generators (TEGs), c) Customizable 3D-components, d) Advanced cement/concrete-based materials, e) Aerogels, f) Insulation made from waste g) Multifunctional component materials (MCs), h) Solar active building envelopes (SAE), i) Omniphobic coatings. showed that ML-driven approaches for prediction buildings optimization have grown rapidly, providing information insights be utilized industry maximize efficiency while reducing CO2 emissions, resulting more productive environmentally friendly era.

Language: Английский

Citations

92
A Review of Physics-Informed Machine Learning in Fluid Mechanics DOI Creative Commons
Pushan Sharma, Wai Tong Chung, Bassem Akoush

et al.

Energies, Journal Year: 2023, Volume and Issue: 16(5), P. 2343 - 2343

Published: Feb. 28, 2023

Physics-informed machine-learning (PIML) enables the integration of domain knowledge with machine learning (ML) algorithms, which results in higher data efficiency and more stable predictions. This provides opportunities for augmenting—and even replacing—high-fidelity numerical simulations complex turbulent flows, are often expensive due to requirement high temporal spatial resolution. In this review, we (i) provide an introduction historical perspective ML methods, particular neural networks (NN), (ii) examine existing PIML applications fluid mechanics problems, especially Reynolds number (iii) demonstrate utility techniques through a case study, (iv) discuss challenges developing mechanics.

Language: Английский

Citations

89
Multi-cavitation states diagnosis of the vortex pump using a combined DT-CWT-VMD and BO-LW-KNN based on motor current signals DOI

Weitao Zeng,

Peijian Zhou, Yanzhao Wu

et al.

IEEE Sensors Journal, Journal Year: 2024, Volume and Issue: 24(19), P. 30690 - 30705

Published: Aug. 26, 2024

Language: Английский

Citations

38
Complex nonlinear dynamics and vibration suppression of conceptual airfoil models: A state-of-the-art overview DOI
Qi Liu, Yong Xu, Jürgen Kurths

et al.

Chaos An Interdisciplinary Journal of Nonlinear Science, Journal Year: 2022, Volume and Issue: 32(6)

Published: June 1, 2022

During the past few decades, several significant progresses have been made in exploring complex nonlinear dynamics and vibration suppression of conceptual aeroelastic airfoil models. Additionally, some new challenges arisen. To best author's knowledge, most studies are concerned with deterministic case; however, effects stochasticity encountered practical flight environments on dynamical behaviors systems neglected. Crucially, coupling interaction structure nonlinearities uncertainty fluctuations can lead to difficulties models, including accurate modeling, response solving, suppression. At same time, existing depend mainly a mathematical model established by physical mechanisms. Unfortunately, it is challenging even impossible obtain an wing engineering practice. The emergence data science machine learning provides opportunities for understanding from data-driven point view, such as prediction, control recorded data. Nevertheless, relevant problems not addressed well up now. This survey contributes conducting comprehensive overview recent developments toward suppression, especially stochastic dynamics, early warning, problems, two-dimensional models different structural nonlinearities. results summarized discussed. Besides, potential development directions that worth further exploration also highlighted.

Language: Английский

Citations

45
Deep-learning-based aerodynamic shape optimization of rotor airfoils to suppress dynamic stall DOI
Jiaqi Liu, Rongqian Chen, Jinhua Lou

et al.

Aerospace Science and Technology, Journal Year: 2023, Volume and Issue: 133, P. 108089 - 108089

Published: Jan. 2, 2023

Language: Английский

Citations

29
Ultralow Energy Consumption Angstrom-Fluidic Memristor DOI

Deli Shi,

Wenhui Wang,

Yizheng Liang

et al.

Nano Letters, Journal Year: 2023, Volume and Issue: 23(24), P. 11662 - 11668

Published: Dec. 8, 2023

The emergence of nanofluidic memristors has made a giant leap to mimic the neuromorphic functions biological neurons. Here, we report signaling using Angstrom-scale funnel-shaped channels with poly-l-lysine (PLL) assembled at nano-openings. We found frequency-dependent current–voltage characteristics under sweeping voltage, which represents diode in low frequencies, but it showed pinched current hysteresis as frequency increases. is strongly dependent on pH values weakly salt concentration. attributed entropy barrier PLL molecules entering and exiting Angstrom channels, resulting reversible voltage-gated open-close state transitions. successfully emulated synaptic adaptation Hebbian learning voltage spikes obtained minimum energy consumption 2–23 fJ each spike per channel. Our findings pave new way neuronal by consumption.

Language: Английский

Citations

29
Enhanced surrogate modelling of heat conduction problems using physics-informed neural network framework DOI

Seyedalborz Manavi,

Thomas Becker, Ehsan Fattahi

et al.

International Communications in Heat and Mass Transfer, Journal Year: 2023, Volume and Issue: 142, P. 106662 - 106662

Published: Feb. 12, 2023

Language: Английский

Citations

24
On the spatial prediction of the turbulent flow behind an array of cylinders via echo state networks DOI Creative Commons
Mohammad Sharifi Ghazijahani, Christian Cierpka

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 144, P. 110079 - 110079

Published: Jan. 23, 2025

Language: Английский

Citations

1
Reduced-order modeling of fluid flows with transformers DOI
AmirPouya Hemmasian, Amir Barati Farimani

Physics of Fluids, Journal Year: 2023, Volume and Issue: 35(5)

Published: May 1, 2023

Reduced-order modeling (ROM) of fluid flows has been an active area research for several decades. The huge computational cost direct numerical simulations motivated researchers to develop more efficient alternative methods, such as ROMs and other surrogate models. Similar many application areas, computer vision language modeling, machine learning data-driven methods have played important role in the development novel models dynamics. transformer is one state-of-the-art deep architectures that made breakthroughs areas artificial intelligence recent years, including but not limited natural processing, image video processing. In this work, we investigate capability architecture dynamics a ROM framework. We use convolutional autoencoder dimensionality reduction mechanism train model learn system's encoded state space. shows competitive results even turbulent datasets.

Language: Английский

Citations

19
Physics-infused deep neural network for solution of non-associative Drucker–Prager elastoplastic constitutive model DOI

Arunabha M. Roy,

Suman Guha, Veera Sundararaghavan

et al.

Journal of the Mechanics and Physics of Solids, Journal Year: 2024, Volume and Issue: 185, P. 105570 - 105570

Published: Feb. 12, 2024

Language: Английский

Citations

8