Computational Technology Resources - CCC

Z. Yang, T. Wang, Y. Wang, "Aerodynamic Prediction and Optimization of Vortex Generator for 600km/h Maglev Train Based on Deep Learning", in J. Pombo, (Editor), "Proceedings of the Seventh International Conference on Railway Technology: Research, Development and Maintenance ", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 15, Paper 5.9, 2026, doi:10.4203/ccc.15.5.9

Keywords: maglev trains, aerodynamic force, vortex generator, drag reduction, deep learning, U-net.

Abstract

As high-speed maglev trains target operational speeds of 600 km/h, the complex wake flow and flow separation at the tail car significantly contribute to aerodynamic drag and stability challenges. This study proposes an integrated framework combining Computational Fluid Dynamics (CFD) and deep learning for the rapid prediction and optimization of the tail flow field. Passive flow control is implemented via vortex generators (VGs) installed on the tail car. A high-fidelity dataset was constructed using CFD simulations based on the Shear Stress Transport turbulence model, covering speeds from 400 to 600 km/h and various VG deflection angles. Subsequently, a U-Net convolutional neural network was developed to map the relationship between geometric/boundary parameters and the resulting wake vorticity contours and aerodynamic coefficients. Results indicate that the U-Net model achieves high accuracy in predicting complex wake topologies with a significant reduction in computational cost compared to traditional solvers. Specifically, an optimal deflection angle of 10° was identified, yielding maximum drag reduction. The findings demonstrate that the proposed deep-learning-based approach effectively accelerates the design and evaluation of bionic-inspired passive flow control structures for ultra-high-speed maglev systems.

download the full-text of this paper (PDF, 8 pages, 532 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCC CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Conferences ISSN 2753-3239 CCC: 15 PROCEEDINGS OF THE SEVENTH INTERNATIONAL CONFERENCE ON RAILWAY TECHNOLOGY: RESEARCH, DEVELOPMENT AND MAINTENANCE Edited by: J. Pombo Paper 5.9 Aerodynamic Prediction and Optimization of Vortex Generator for 600km/h Maglev Train Based on Deep Learning Z. Yang¹, T. Wang² and Y. Wang¹ ¹School of Traffic and Transportation Engineering, Central South University, Hunan, China ²College of Mechanical and Vehicle Engineering, Hunan University, China doi:10.4203/ccc.15.5.9 Full Bibliographic Reference for this paper Z. Yang, T. Wang, Y. Wang, "Aerodynamic Prediction and Optimization of Vortex Generator for 600km/h Maglev Train Based on Deep Learning", in J. Pombo, (Editor), "Proceedings of the Seventh International Conference on Railway Technology: Research, Development and Maintenance ", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 15, Paper 5.9, 2026, doi:10.4203/ccc.15.5.9 Keywords: maglev trains, aerodynamic force, vortex generator, drag reduction, deep learning, U-net. Abstract As high-speed maglev trains target operational speeds of 600 km/h, the complex wake flow and flow separation at the tail car significantly contribute to aerodynamic drag and stability challenges. This study proposes an integrated framework combining Computational Fluid Dynamics (CFD) and deep learning for the rapid prediction and optimization of the tail flow field. Passive flow control is implemented via vortex generators (VGs) installed on the tail car. A high-fidelity dataset was constructed using CFD simulations based on the Shear Stress Transport turbulence model, covering speeds from 400 to 600 km/h and various VG deflection angles. Subsequently, a U-Net convolutional neural network was developed to map the relationship between geometric/boundary parameters and the resulting wake vorticity contours and aerodynamic coefficients. Results indicate that the U-Net model achieves high accuracy in predicting complex wake topologies with a significant reduction in computational cost compared to traditional solvers. Specifically, an optimal deflection angle of 10° was identified, yielding maximum drag reduction. The findings demonstrate that the proposed deep-learning-based approach effectively accelerates the design and evaluation of bionic-inspired passive flow control structures for ultra-high-speed maglev systems. download the full-text of this paper (PDF, 8 pages, 532 Kb) go to the previous paper go to the next paper return to the table of contents return to the volume description
Back to top	©Civil-Comp Limited 2023 - terms & conditions