YFYiwei Fan

Yiwei Fan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Yiwei Fan has a total of 22 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Free-surface breaking dynamics and morphology of waves generated by the semi-submerged wedge-shaped body

    by Minglei Ji, Haihua Deng, Zhiben Shen, Yiwei Fan, Yunfei Kuai, Yang Han

    The breaking waves at the bow and stern represent a fundamental free-surface phenomenon involving violent interface deformation, air entrainment, and strong two-phase coupling. The temporal evolution and spatial distribution of wave breaking are fundamental to understanding water-air mixing flows around ships. The present study adopts the square stern of high-speed ships as a reference and selects a wedge model as the research object in order to investigate the characteristics of interface breaking and the flow morphology. The free-surface waves around the wedge model were measured using a binocular vision system. A systematic analysis was conducted to investigate the impact of various parameters on the flow patterns in the bow and stern regions. The parameters investigated included draft depth (D), stern width (B), attack angle (α), and draft Froude number FrD=U/gD. The results show that the wave profiles have strong similarity when FrD is chosen as the dominant parameter. The associated parameters of FrD and geometric parameters are found to have a significant influence on the morphological characteristics of the waves.

  • Probabilistic physics-constrained reconstruction of three-dimensional flow fields

    by Shengyu Tang, Weihao Xia, Yiwei Fan, Changdong Yu, Shuaiyu Bao, Yiming Cao, Yang Han

    Accurate reconstruction of three-dimensional flow fields is essential for resolving fine-scale flow structures and understanding unsteady flow physics in ship and ocean engineering applications. However, while direct numerical simulation provides high-fidelity solutions of the Navier–Stokes equations, its prohibitive computational cost limits its applicability to large-scale and high-Reynolds-number flows. Reduced-order and data-driven reconstruction approaches offer improved efficiency, yet often fail to preserve inter-slice coherence, fine-scale vortical structures, and physical consistency when extrapolating beyond the training regime. To address these challenges from a physical-consistency perspective, this study investigates a physics-constrained framework for three-dimensional flow-field reconstruction. The governing physical constraints are embedded into the reconstruction process through a probabilistic formulation derived from the vorticity–velocity representation of the Navier–Stokes equations, enabling improved enforcement of incompressibility and dynamical consistency. A spatiotemporal reconstruction architecture is constructed to capture multi-scale spatial features and temporal evolution characteristics of unsteady flows, while a physics-informed attention mechanism is employed to enhance the recovery of fine-scale flow structures in regions with strong spatial gradients. The proposed framework is assessed using turbulent flow data from the Johns Hopkins Turbulence Database, transient boundary layer flows, and experimental tomographic particle image velocimetry measurements. Quantitative comparisons with high-fidelity reference data demonstrate that the physics-constrained reconstruction significantly improves the preservation of fine-scale flow structures and extrapolation performance compared with conventional deterministic approaches, indicating its effectiveness for physically consistent three-dimensional flow-field reconstruction at reduced computational cost.


  • Effects of bow geometry on bubble generation and transport around a research vessel

    by Peng Xu, Zejun Liang, Yiwei Fan, Linfeng Chen, Kunde Yang

    This study investigates how different bow geometries affect bubble generation and transport, a critical factor in sonar performance degradation due to bubble sweep-down. A coupled Eulerian–Lagrangian numerical framework with improved delayed detached eddy simulation method was developed. A staged numerical approach was adopted, combining the volume of fluid–Lagrangian multiphase model transition model for bubble generation under waves with a pre-seeded particle method for transport analysis in calm water. Results reveal that the intensity of bow–wave interaction is the decisive factor governing bubble generation. Model B, featuring a bow bulb, produced the highest bubble count due to intense impact at the bulb forefront and subsequent cavity collapse entrainment. In contrast, the axe-bow design of model A exhibited superior wave-piercing capability, resulting in minimal bubble generation. Regarding transport pathways, the wake flow and vortex structures dictated by the bow geometry are of paramount importance. For model B, the downward flow deflection induced by the bulb, combined with a discrete and low-strength bilge vortex, led to chaotic bubble sweep-down near the bow and widespread dispersion across the vessel's bottom. Conversely, model C, incorporating a flow guide channel and a robust, coherent bilge vortex structure, effectively confined the bubbles within the bilge flow field, thereby significantly suppressing their invasion into the critical flat bottom region. This research elucidates the underlying physics of how bow geometry affects bubble sweep-down, providing a theoretical foundation and a robust numerical tool for the optimal design of ship bows with low bubble sweep-down and enhanced acoustic stealth characteristics.

  • Robust Fluid Motion Estimator Based on Attentional Transformer

    by Changdong Yu, Yongpeng Chang, Xiao Liang, Yiwei Fan

    Estimating complex fluid motions from successive images, i.e., Particle Image Velocimetry (PIV), is of great research significance in physics and engineering applications. Although deep learning-based methods have made great progress for fluid motion estimation, it still remains a challenge in terms of robustness and generalization ability. To address this issue, we present a novel fluid flow estimator called RFMFlowNet in this paper, to robustly estimate complex fluid motion using attentional Transformer. Concretely, the feature encoder based on Transformer is customized to enhance features of two frames globally and stably. Here we also enlarge the receptive field of the encoder according to the characteristics of the flow image to extract more effective information. Then, efficient global matching is performed using 4D correlation volume. Furthermore, a refined flow field is iteratively predicted using a GRU-based optimizer. Extensive experiments, including on synthetic and real-world images, are performed to assess the proposed approach. Experimental results demonstrate that the RFMFlowNet achieves new state-of-the-art performance on the public dataset. Meanwhile, our model enjoys great robustness and generalization ability to challenging flow images, reasonably predicting the evolution trend of fluid flows.


  • FlowSRNet: A multi-scale integration network for super-resolution reconstruction of fluid flows

    by Xiaojun Bi, Ankang Liu, Yiwei Fan, Changdong Yu, Zefeng Zhang

    A wide range of research problems in physics and engineering involve the acquisition of high-resolution data. Recently, deep learning has proved to be a prospective technique for super-resolution (SR) reconstruction of fluid flows. General deep learning methods develop temporal multi-branch networks to improve SR accuracy while ignoring computational efficiency. Further, the generalization ability of the deep learning model in different fluid flow scenarios is still an unstudied issue. In this article, we propose an efficient multi-scale integration network called FlowSRNet to reconstruct the high-resolution flow fields. Specifically, we elaborately design a lightweight multi-scale aggregation block (LMAB) to capture multi-scale features of fluid data, which contains a parallel cascading architecture and feature aggregation module. The residual backbone of FlowSRNet is built by cascading the LMABs (cascaded blocks number N = 8) in a serial manner. Also, we present a small architecture LiteFlowSRNet (cascaded blocks number N = 2) for comparison. In addition, a corresponding SR dataset is constructed to train and test the proposed model, which contains different kinds of fluid flows. Finally, extensive experiments are performed on different fluid data to evaluate the performance of the proposed model. The results demonstrate that our approach achieves state-of-the-art SR performance on various fluid flow fields. Notably, our method enjoys merit of lightweight, which facilitates the development of the complicated calculation in computational fluid dynamics.


  • Time-Resolved Particle Image Velocimetry Algorithm Based on Deep Learning

    by Chunyu Guo, Yiwei Fan, Changdong Yu, Yang Han, Xiaojun Bi

    Time-resolved particle image velocimetry (TR-PIV) is an advanced fluid mechanics experiment technology, which can simultaneously measure the velocity field from multi-frame images to analyze the evolution of fluid over time. Deep learning technology has made great progress in the field of particle image velocimetry (PIV). However, as far as we know, no deep learning method has been adopted to calculate the velocity field of TR-PIV images (i.e., TR-PIV). In this article, we propose a novel cascaded convolutional neural network (CNN) called CascLiteFlowNet-R-en for TR-PIV estimation task. Furthermore, to train and optimize the model, we generated a challenge TR-PIV dataset of multi-frame velocity fields. The velocity field here changes according to the frames to simulate the real fluid scene. Finally, the proposed model has been verified on synthetic and experimental particle images. The results show that our proposed method achieves excellent performance, with competitive calculation accuracy and high calculation efficiency.


  • A Cascaded Convolutional Neural Network for Two-Phase Flow PIV of an Object Entering Water

    by Changdong Yu, Haozhe Luo, Yiwei Fan, Xiaojun Bi, Mingjie He

    In the two-phase flow particle image velocimetry (PIV) experiment of an object entering water, the mask of the noncomputed area and the calculation of velocity field in particle images are two key stages. Due to the complexity of the edge of the object in the particle image, the mask calibration is usually performed manually and then the PIV estimation is carried out. We propose a cascaded convolutional neural network (CNN) in this article to implement end-to-end two-phase flow fluid motion estimation. In the first stage, the image segmentation network U-Net is used to mask the noncomputational area of the image and extract the liquid phase. In the second stage, we adopt the improved deep optical flow network, which is known as recurrent allpairs field transforms (RAFT) to calculate the velocity field. What is more, the corresponding datasets are generated for training model parameters. Finally, our approach is tested on synthetic and experimental images. The experimental results indicate that our approach not only reaches accurate segmentation of the calculated liquid phase region but also achieves a high-precision velocity field calculation. Meanwhile, this cascade CNN model has high efficiency toward real-time estimation.


  • LightPIVNet: An Effective Convolutional Neural Network for Particle Image Velocimetry

    by Changdong Yu, Xiaojun Bi, Yiwei Fan, Yang Han, Yunfei Kuai

    Particle image velocimetry (PIV) plays a significant role in experimental fluid mechanics, which aims to extract the velocity fields from successive particle image pairs. Deep learning (DL) techniques have been proposed to solve such a fluid motion estimation problem. However, the existing DL methods put emphasis on accuracy while ignoring the problem of model redundancy and low computational efficiency. In this article, we propose a novel lightweight convolutional neural network called LightPIVNet for PIV estimation, which is targeted to improve the advanced optical flow model Recurrent All-Pairs Field Transforms (RAFT). Furthermore, considering the real fluid scene, we generated a new dataset PIV-Dataset-II to train our network, which increases the amount and variety of flow fields. Our approach has been verified and analyzed on synthetic and experimental particle images. The experimental results indicate that our method achieves state-of-the-art performance among all the DL methods and is much superior to those classical traditional methods, such as WIDIM and HS optical flow. Meanwhile, our model LightPIVNet reduces parameters by 40.2% and improves inference time by 14.3% compared to the current best DL model PIV-LiteFlowNet-en.