An open channel flow sediment incipient motion full flow field synchronous measurement system and method

Abstract

The present application relates to the technical field of fluid dynamics test, and discloses a kind of open channel flow silt starting full flow field synchronous measurement system and method, the method comprises: configuration silt and tracer particle, set up high-speed camera and laser calibration coordinate mapping.Increase flow to the critical state of particle starting, synchronously collect multi-view image, split image to generate mask, reconstruct three-dimensional concentration field and particle coordinate using MART / SART algorithm, calculate three-dimensional velocity field of water flow, track particle trajectory, analyze velocity slip, vortex structure and starting space-time correlation, determine critical shear stress.The present application can not only obtain full three-dimensional, three-component transient velocity field of flow field near bed surface under silt starting critical state in a non-invasive manner, but also can identify, segment and independently measure velocity field of water tracer particle phase and moving silt particle phase synchronously, realize separation and coupling analysis of gas-liquid-solid multiphase motion information.

Description

Technical Field

[0001] This invention relates to the field of fluid dynamics testing technology, and more specifically, to a system and method for synchronous measurement of the entire flow field of sediment initiation in open channel flow. Background Technology

[0002] Sediment initiation is a core fundamental issue in river dynamics, coastal engineering, and geological hazard prevention. Accurate determination of its critical conditions relies on a deep understanding of the microstructure of the near-bed flow field. Traditional studies on sediment initiation often depend on measuring time-averaged velocities using point-type velocimeters (such as ADV) or obtaining planar velocity fields based on two-dimensional particle image velocimetry. However, these methods have inherent limitations: Spatial dimensional limitations: Conventional 2D2C or 2D3C PIV can only measure the velocity components within the optical plane of a single laser sheet, and cannot obtain the velocity components perpendicular to the optical direction of the sheet or the complete three-dimensional vortex structure. However, sediment initiation is often induced by strong three-dimensional vortex structures (such as horseshoe vortices and hairpin vortices) near the bed surface, and their vortex axes are often oblique to the mainstream direction. Two-dimensional measurements will lose key dynamic information.

[0003] Limitations of phase resolution: Existing PIV techniques mostly assume the fluid is a single phase, or roughly distinguish tracer particles from coarse particles through post-processing. At the critical state of sediment initiation, especially with fine-grained sediment or under aeration conditions, the flow field exhibits complex multiphase characteristics with the coexistence of a water phase, a moving sand phase, and possibly even a bubble phase. Traditional methods struggle to accurately and synchronously measure the individual velocities of each phase (water flow, sediment particles) in the same spatiotemporal coordinate system, and cannot quantify the velocity slip and interaction forces between particles and water flow.

[0004] Insufficient analysis of microscopic mechanisms: Two-dimensional measurements are difficult to accurately characterize the spatial morphology and intensity evolution of the vortex structure that envelops or lifts particles, and cannot accurately track the trajectory, rotation, and initiation mode (sliding, rolling, leaping) of individual sediment particles in three-dimensional space, which limits the explanation of the initiation criteria from the perspective of mechanical mechanisms.

[0005] Therefore, it is particularly important to design a three-dimensional detection system or method for sediment initiation. Summary of the Invention

[0006] The purpose of this invention is to provide a synchronous measurement system and method for the entire flow field of sediment initiation in open channel flow, in order to solve the above-mentioned problems.

[0007] On the one hand, the present invention provides a method for synchronous measurement of the entire flow field of sediment initiation in open channel flow, comprising the following steps: S1. Configuration and Calibration: In the test section, mud and sand particles and tracer particles were laid in the water tank. High-speed cameras were set up at different angles and lasers were set up for illumination. The high-speed cameras were turned on to obtain several calibration images. The mapping relationship between the pixel coordinates of each high-speed camera and the three-dimensional world coordinates was established based on the calibration images. S2. Synchronous acquisition of three-dimensional flow field data under critical state: The flow rate of the test section water tank was gradually increased. When the first sediment particle moved, it was recorded as the critical state. All high-speed cameras and lasers were triggered simultaneously to acquire a continuous time series of multi-view image sequences. S3. Multiphase Flow Image Segmentation and 3D Reconstruction: The acquired multi-view image sequence identifies tracer particles and sediment particles and segments the tracer particle region and sediment particle region to generate a mask image; based on the MART iterative algorithm or SART iterative algorithm, the three-dimensional concentration field of tracer particles and the three-dimensional spatial coordinates of each sediment particle in the entire measurement body are reconstructed step by step over time. S4. Calculation and Mechanism Analysis of Multiphase Three-Dimensional Velocity Field: 4.1 Calculation of water flow phase velocity field: Based on the three-dimensional concentration field of the continuous time step, the transient three-dimensional three-component water velocity vector field U(x,y,z,t) is calculated by applying the three-dimensional multi-scale cross-correlation algorithm, where x, y and z represent position coordinates and t represents time; 4.2 Sediment phase motion analysis: Three-dimensional particle tracking is performed on the three-dimensional spatial coordinates, and the same particle at adjacent time points is matched to obtain the three-dimensional trajectory, instantaneous velocity vector Vp(t), and angular velocity of each moving sediment particle; 4.3 Multiphase Flow Coupling Analysis: Extract the instantaneous velocity Up(t) of the water flow phase at the location of each moving sediment particle at time t, and calculate the velocity slip vector between the sediment particles and the water flow; extract the three-dimensional vortex structure and analyze the spatiotemporal correlation between the vortex core position, intensity, evolution and particle initiation position and time; statistically analyze the relationship between the spatial distribution probability of particle initiation under critical conditions and local hydrodynamic parameters; based on the three-dimensional three-component water velocity vector field U(x,y,z,t), the bed shear stress tensor is determined by fitting the near-wall velocity profile or directly calculating the bed shear stress tensor, and the critical shear stress or frictional velocity for sediment initiation is determined based on the synchronously observed particle initiation.

[0008] Furthermore, step S1 specifically includes: 1.1 Lay mud and sand particles of the target particle size at the bottom of the test section water tank to form a flat bed surface; 1.2 Install four high-speed cameras around the measuring body, and adjust their positions and angles. Under Scheimpflug conditions, the measuring body is the area from the flat bed surface to a height several times the target particle size above it. 1.3 Adjust the laser output until a cubic light spot uniformly covers the measurement body; 1.4 Place the three-dimensional calibration plate within the measurement body, synchronously trigger all high-speed cameras, acquire several calibration images from different angles, establish the mapping relationship between the pixel coordinates of each high-speed camera and the three-dimensional world coordinates, and evaluate the calibration error. 1.5. Release tracer particles into the water.

[0009] Furthermore, step S2 specifically includes: 2.1 Gradually increase the flow rate of the test section's water tank; 2.2 When the first sediment particle moves on the leveled bed surface, it is recorded as the critical start-up state. At the same time, all high-speed cameras and lasers are triggered by the synchronous controller to synchronously acquire a continuous time series of multi-view image sequences at a fixed frame rate. 2.3. Increase the flow rate of the test section water tank and repeat the process of obtaining multi-view image sequences in step 2.2 to obtain three-dimensional flow field data under different intensities of start-up conditions; Furthermore, step S3 specifically includes: 3.1 Process the images from each high-speed camera at each moment in the acquired multi-view image sequence, identify and segment the tracer particle region and the sediment particle region, and generate tracer particle images and sediment particle images. 3.2 For the segmented tracer particle image and sediment particle image, respectively, the MART algorithm or SART algorithm is used to reconstruct the three-dimensional concentration field of tracer particles and the three-dimensional spatial coordinates of each sediment particle in the entire measurement body step by step.

[0010] Further, in step S4, the velocity slip vector ΔV = Vp - Up, where ΔV represents the velocity slip vector, Vp is the instantaneous velocity vector of the sediment particles, and Up is the instantaneous velocity vector of the water flow phase.

[0011] Compared with the prior art, the beneficial effects of the present invention are as follows: Achieving true three-dimensional full flow field measurement: breaking through the limitations of two-dimensional planar measurement, for the first time in sediment initiation experiment, the complete instantaneous three-dimensional three-component velocity field near the bed surface was obtained, which can accurately analyze the spatial morphology, intensity and transient evolution process of the key three-dimensional vortex structure that induces sediment initiation.

[0012] The simultaneous separation and measurement of gas-liquid-solid multiphase motion information was achieved: Through AI image segmentation and tomographic reconstruction technology, the Eulerian velocity field of water (tracer particles) and the Lagrangian trajectory and velocity of moving sediment particles were separated and measured independently at the same time and in the same three-dimensional spatial coordinate system, directly quantifying the interphase interaction and providing direct experimental evidence for establishing a multiphase flow initiation model based on mechanical mechanisms.

[0013] High measurement accuracy and spatiotemporal resolution: Employing a high-energy continuous laser and a high-frame-rate, global shutter camera, coupled with nanosecond-level synchronous control, it can clearly capture the microscopic flow details and particle motion at the moment of sediment initiation; three-dimensional tomographic reconstruction and cross-correlation algorithms ensure the high accuracy of the three-dimensional velocity field.

[0014] The system boasts strong versatility and an advanced methodology: It is not only applicable to sediment initiation studies under clear water conditions but can also be extended to more complex gas-liquid-solid multiphase flow scenarios involving aeration and fine particle suspension. Its integrated tools, such as AI segmentation, 3D PTV, and transient eddy analysis, represent the cutting edge of PIV technology in multiphase flow applications.

[0015] On the other hand, the present invention also proposes a synchronous measurement system for the entire flow field of sediment initiation in open channel flow, used to realize the above-mentioned synchronous measurement method for the entire flow field of sediment initiation in open channel flow, including: The experimental water tank subsystem includes a test section water tank, a flow stabilizing device, and a flow and water level control system; the bottom of the test section water tank is covered with the sediment to be tested to form a flat bed surface; The three-dimensional volume illumination subsystem uses a high-energy continuous laser with an output wavelength of 532nm. The laser beam is expanded and focused through a cylindrical lens group and a spherical lens to form a cubic illumination area with a thickness of 5-10mm, which is used to illuminate the measuring body in the water tank that covers the flat bed surface and the water flow above. The multi-view synchronous image acquisition subsystem includes at least four identical high-speed cameras arranged around the measuring body at different spatial azimuth angles and equipped with Scheimpflug adapters to adjust the imaging plane. All high-speed cameras are connected to a synchronization controller. The tracer and particle subsystem includes tracer particles with a density matched to water, used to mark water flow motion, the tracer particles having a particle size of 10-20 µm; and sediment particles as the object of observation. The intelligent data processing and measurement subsystem includes: The multiphase image segmentation module is configured to segment the original image based on a deep learning algorithm and distinguish different phase states, including background, tracer particles, mud and sand particles and bubbles. The three-dimensional tomography reconstruction module is configured to reconstruct the instantaneous distribution of tracer particles and sediment particles in three-dimensional space based on calibration images acquired by a high-speed camera and using the MART or SART algorithm. The multiphase three-dimensional velocity field calculation module is configured to use a three-dimensional multigrid cross-correlation algorithm to calculate the three-dimensional three-component velocity field of the continuous phase of the water body for the reconstructed tracer particle cloud; and to use a three-dimensional particle tracking velocimetry algorithm to obtain the three-dimensional trajectory, velocity and angular velocity of each sediment particle for the reconstructed individual sediment particles. The post-processing and diagnostic module is configured to provide three-dimensional eddy identification criteria, three-dimensional pulsating kinetic energy calculation, Reynolds stress tensor calculation, interphase velocity slip statistics, particle initiation probability analysis, and three-dimensional streamline and pulsating velocity field visualization tools. The calibration and evaluation module is configured to perform three-dimensional calibration and evaluate calibration errors.

[0016] Furthermore, the parameters of the high-speed camera are as follows: full-frame resolution of not less than 2560×1920 pixels, global shutter, full-frame frame rate of not less than 2000fps, minimum exposure time ≤100ns, focal length lens of 100mm / F2.8 for fine observation and 24mm / F1.4 for wide field of view, and 532nm narrowband filter.

[0017] It should be noted that the beneficial effects of the open channel flow sediment initiation synchronous measurement system and its method are exactly the same, and will not be repeated here. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A functional framework diagram of a synchronous measurement system for the entire flow field of sediment initiation in an open channel, provided in an embodiment of the present invention; Figure 2 This is a scene diagram of a water tank subsystem provided in an embodiment of the present invention; Figure 3 The lighting scene diagram is provided for the three-dimensional volume lighting subsystem in the embodiments of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] This invention provides a method for synchronous measurement of the entire flow field of sediment initiation in open channel flow, comprising the following steps: S1. Configuration and Calibration: In the test section, mud and sand particles and tracer particles were laid in the water tank. High-speed cameras were set up at different angles and lasers were set up for illumination. The high-speed cameras were turned on to obtain several calibration images. The mapping relationship between the pixel coordinates of each high-speed camera and the three-dimensional world coordinates was established based on the calibration images. S2. Synchronous acquisition of three-dimensional flow field data under critical state: The flow rate of the test section water tank was gradually increased. When the first sediment particle moved, it was recorded as the critical state. All high-speed cameras and lasers were triggered simultaneously to acquire a continuous time series of multi-view image sequences. S3. Multiphase Flow Image Segmentation and 3D Reconstruction: The acquired multi-view image sequence identifies tracer particles and sediment particles and segments the tracer particle region and sediment particle region to generate a mask image; based on the MART iterative algorithm or SART iterative algorithm, the three-dimensional concentration field of tracer particles and the three-dimensional spatial coordinates of each sediment particle in the entire measurement body are reconstructed step by step over time. S4. Calculation and Mechanism Analysis of Multiphase Three-Dimensional Velocity Field: 4.1 Calculation of water flow phase velocity field: Based on the three-dimensional concentration field of the continuous time step, the transient three-dimensional three-component water velocity vector field U(x,y,z,t) is calculated by applying the three-dimensional multi-scale cross-correlation algorithm, where x, y and z represent position coordinates and t represents time; 4.2 Sediment phase motion analysis: Three-dimensional particle tracking is performed on the three-dimensional spatial coordinates, and the same particle at adjacent time points is matched to obtain the three-dimensional trajectory, instantaneous velocity vector Vp(t), and angular velocity of each moving sediment particle; 4.3 Multiphase Flow Coupling Analysis: The instantaneous velocity Up(t) of the water flow phase at the location of each moving sediment particle at time t is extracted, and the velocity slip vector between the sediment particle and the water flow is calculated. A three-dimensional vortex structure is extracted, and the spatiotemporal correlation between the vortex core position, intensity, evolution, and particle initiation position and time is analyzed. The extraction of the three-dimensional vortex structure includes calculating the velocity gradient tensor based on the three-dimensional three-component water velocity vector field, determining the three-dimensional vortex structure region using the Q criterion, determining the vortex core position based on the local maxima of the Q value, determining the vortex intensity based on the volume integral of the Q value, and determining the vortex structure boundary based on the Q threshold connected regions. The relationship between the spatial distribution probability of particle initiation under critical conditions and local hydrodynamic parameters is statistically analyzed. Based on the three-dimensional three-component water velocity vector field U(x,y,z,t), the bed shear stress tensor is calculated by fitting the near-wall velocity profile or directly calculating the bed shear stress tensor. Based on the synchronously observed particle initiation, the critical shear stress or frictional velocity for sediment initiation is determined.

[0023] For ease of understanding, the following explains some key terms in this embodiment: Sediment initiation in open channel flow refers to the phenomenon where sediment particles at the bottom of the channel begin to move from a static state under the influence of water flow. This phenomenon is central to the study of hydraulics and sediment dynamics, and understanding its critical conditions and microscopic mechanisms is significant for fields such as water conservancy engineering and environmental management.

[0024] Simultaneous measurement across the entire flow field refers to the simultaneous measurement of the motion states of different phases (such as the water phase and the sediment particle phase) in a fluid within the same time and space. This measurement method aims to obtain physical quantities such as velocity and concentration at various points in the flow field, and simultaneously capture the motion trajectory and state of sediment particles, in order to achieve a comprehensive analysis of multiphase flow.

[0025] A high-speed camera is a device capable of acquiring images at high frame rates. It is used to capture rapidly changing physical phenomena, such as the instantaneous motion of tracer particles and sediment particles in flowing water, thereby providing temporally resolved image data for subsequent velocity field calculations and motion analysis.

[0026] In this method, a laser is used as the illumination source, illuminating the measurement area by emitting a laser beam. Its purpose is to make tracer particles and sediment particles visible in the field of view of the high-speed camera, facilitating image acquisition and subsequent processing.

[0027] Tracer particles are tiny particles with a density similar to water, which are released into water flows to mark the movement of the water. By tracking the movement of tracer particles, information about the velocity field of the water flow can be indirectly obtained.

[0028] Sediment particles are the objects observed and analyzed in this method, and their initiation, movement, and transport under the action of water flow are the core of the research.

[0029] The MART iterative algorithm, or SART iterative algorithm, is one of two algebraic reconstruction techniques used to reconstruct the distribution of objects in three-dimensional space from multi-view projection images. In this method, it is used to reconstruct the three-dimensional concentration field of tracer particles and the three-dimensional spatial coordinates of sediment particles.

[0030] A three-dimensional concentration field refers to the distribution density of tracer particles or sediment particles in three-dimensional space. By reconstructing the three-dimensional concentration field, the spatial distribution of tracer particles within the measurement area can be quantified, and then used to calculate the water flow velocity field.

[0031] Three-dimensional spatial coordinates refer to the position of sediment particles in a three-dimensional measurement space. By obtaining the time-step three-dimensional spatial coordinates of sediment particles, their motion trajectory can be tracked and analyzed.

[0032] The three-dimensional multi-scale cross-correlation algorithm is an image processing algorithm used to calculate the velocity field of fluids. It calculates the instantaneous velocity vector field of the fluid in three-dimensional space by analyzing the displacement of tracer particle distribution in continuous time-step images.

[0033] 3D particle tracking is a technique that determines the trajectory, velocity, and angular velocity of a particle in three-dimensional space by identifying and matching the positions of the same particle in consecutive image frames.

[0034] The velocity slip vector is the vector difference between the instantaneous velocity vector of a sediment particle and the instantaneous velocity vector of the water flow phase at the particle's location. This vector reflects the motion between the particle and the fluid and is a parameter for analyzing the forces acting on the particle and the interactions between phases.

[0035] A three-dimensional vortex structure refers to a three-dimensional region in a fluid that exhibits rotational motion characteristics. Sediment initiation is often associated with three-dimensional vortex structures near the bed surface. Analyzing their location, intensity, and evolution helps to reveal the microscopic mechanisms of sediment initiation. In this embodiment, the three-dimensional vortex structure is identified using the Q-criterion. The Q-criterion, based on the velocity gradient tensor of the three-dimensional three-component water velocity vector field, identifies spatially continuous regions where local rotational effects dominate over local strain effects as three-dimensional vortex structure regions.

[0036] This method begins with configuration and calibration. Specifically, sediment particles are laid at the bottom of the test section's water tank, and tracer particles are released into the water. Then, high-speed cameras are positioned at different angles around the measurement area, and lasers are configured for illumination. To establish the correspondence between image pixels and actual 3D space, the high-speed cameras are activated to acquire several calibration images. Based on these calibration images, a mapping relationship between the pixel coordinates of each high-speed camera and 3D world coordinates can be established. For example, a calibration board based on known geometry can be used for image acquisition, and camera parameters and distortion correction parameters can be calculated using computer vision algorithms.

[0037] After configuration and calibration, synchronous acquisition of three-dimensional flow field data under critical conditions was performed. This process involved gradually increasing the flow rate of the test section's flume to simulate different flow conditions. When the first sediment particle movement was observed at the bottom of the flume, this state was recorded as the critical state of sediment initiation. Under this critical state, all high-speed cameras and lasers were synchronously triggered to ensure that the instantaneous state of the flow field was captured at the same moment. This resulted in the acquisition of a continuous time-series multi-view image sequence containing motion information of tracer particles and sediment particles in the water flow.

[0038] Subsequently, multiphase flow image segmentation and 3D reconstruction are performed on the acquired multi-view image sequences. This step first identifies tracer particles and sediment particles in the image sequence and segments them from the background, thereby generating mask images of the tracer particle region and the sediment particle region. For example, image processing techniques based on threshold segmentation or edge detection can be used. Based on this, using the MART iterative algorithm or SART iterative algorithm, the 3D concentration field of tracer particles and the 3D spatial coordinates of each sediment particle within the entire measurement body are reconstructed step-by-step. These algorithms, through iterative optimization, invert the multi-view 2D image information into the distribution of physical quantities in 3D space.

[0039] Finally, multiphase three-dimensional velocity field calculations and mechanism analysis are performed. This step is further divided into water flow phase velocity field calculation, sediment phase motion analysis, and multiphase flow coupling analysis.

[0040] Specifically, for calculating the water flow velocity field, based on the three-dimensional concentration field at continuous time steps, a three-dimensional multi-scale cross-correlation algorithm is applied to calculate the transient three-dimensional three-component water velocity vector field U(x,y,z,t). Here, x, y, and z represent position coordinates, and t represents time. This algorithm calculates the water flow velocity by analyzing the displacement of tracer particles within continuous time steps.

[0041] For sediment facies motion analysis, three-dimensional particle tracking is performed on the three-dimensional spatial coordinates. By matching the same particle at adjacent time points, the three-dimensional trajectory, instantaneous velocity vector Vp(t), and angular velocity of each moving sediment particle can be obtained. For example, a prediction-matching based tracking algorithm can be used.

[0042] For multiphase flow coupling analysis, the instantaneous velocity Up(t) of the water phase at the location of each moving sediment particle at time t is extracted, and the velocity slip vector between the sediment particles and the water flow is calculated. Furthermore, based on the three-dimensional three-component water velocity vector field U(x,y,z,t), the Q criterion is used to extract the three-dimensional vortex structure, and the spatiotemporal correlation between the position, intensity, evolution of the vortex core and the particle initiation position and time is analyzed. Specifically, when extracting the three-dimensional vortex structure using the Q criterion, spatial difference is first performed on the velocity components of adjacent grid points in the three-dimensional three-component water velocity vector field U(x,y,z,t) to obtain the velocity gradient tensor at each grid point. U; the velocity gradient tensor U is decomposed into a strain rate tensor S and a rotation rate tensor Ω; the Q value at each grid point is calculated according to Q = 1 / 2 (||Ω||² - ||S||²), where ||Ω|| represents the magnitude of the rotation rate tensor and ||S|| represents the magnitude of the strain rate tensor. When determining the three-dimensional vortex structure region, the 70th percentile of all positive Q values ​​in the same time step is taken as the Q threshold Q. TWhen a positive Q value does not exist, the three-dimensional vortex structure region is not output at this time step. A positive Q value greater than Q... T Furthermore, a three-dimensional vortex structure region is defined by a set of voxels connected to twenty-six neighborhoods in three-dimensional space; the outer surface of this connected volume is defined as the vortex structure boundary. When determining the vortex core location and vortex intensity, for each three-dimensional vortex structure region, the coordinates of the grid point with the largest Q value within the region are determined as the vortex core coordinates; when multiple grid points with the same maximum Q value exist within the same region, the grid point with the smallest distance from the geometric center of the three-dimensional vortex structure region is selected as the vortex core coordinates. The sum of the products of the Q values ​​of each grid point within the three-dimensional vortex structure region and the corresponding voxel volume is taken as the vortex intensity, and the maximum Q value within the region is taken as the peak intensity. When tracking the temporal evolution of vortex structures, for two adjacent time steps, the distance between the vortex core coordinates of the previous time step and the candidate vortex core coordinates of the next time step is first calculated. If this distance is no greater than five times the velocity field grid spacing, and the voxel overlap ratio of the two three-dimensional vortex structure regions is no less than 20%, the candidate vortex structure of the next time step is determined to be a continuation of the vortex structure of the previous time step. If multiple candidate vortex structures exist, the candidate vortex structure with the largest voxel overlap ratio is selected first. If the voxel overlap ratios are the same, the candidate vortex structure with the smallest vortex intensity difference is selected. This yields the vortex core coordinate sequence, vortex structure boundary sequence, and vortex intensity sequence. When performing spatiotemporal correlation analysis between vortex structures and sediment particle initiation, the three-dimensional spatial coordinates corresponding to the initiation time of sediment particles are compared with the three-dimensional vortex structure region at the same time step. When the three-dimensional spatial coordinates of sediment particles are located within the boundary of the vortex structure, or the distance between the three-dimensional spatial coordinates of sediment particles and the coordinates of the vortex core is not greater than twice the target particle size, and the time difference between the particle initiation time and the vortex structure passing through the particle's neighborhood is not greater than one image sampling time interval, it is determined that the sediment particle initiation event and the three-dimensional vortex structure have a spatiotemporal correlation.

[0043] This method achieves simultaneous three-dimensional measurement of the entire flow field, including both the water flow phase and the sediment particle phase, during sediment initiation in open channel flow, overcoming the limitations of traditional methods in terms of spatial dimension, phase resolution, and microscopic mechanism analysis. As a result, it can acquire the three-dimensional velocity field of the water flow, the three-dimensional motion trajectory and velocity of sediment particles, quantify the velocity slip between particles and the water flow, and analyze the spatiotemporal correlation between the three-dimensional vortex structure and particle initiation. This provides data support for determining the critical conditions for sediment initiation and elucidating the microscopic mechanisms.

[0044] In the synchronous measurement method of the entire flow field of sediment initiation in open channel flow, configuration and calibration are fundamental to ensuring measurement accuracy and reliability. However, if the configuration process lacks refined and standardized operations, especially if the definition of the measurement volume, camera layout, illumination uniformity, and calibration accuracy are not adequately considered, it may lead to poor quality of subsequently acquired data, thereby affecting the accurate judgment of the critical state of sediment initiation and the reliability of flow field analysis.

[0045] In this regard, this application further proposes a specific implementation method for step S1, including: First, sediment particles of the target size are laid at the bottom of the test section's flume to form a smooth bed surface. This step aims to provide a uniform and repeatable initial boundary condition for the sediment initiation experiment. By precisely controlling the particle size distribution of the sediment and ensuring a smooth bed surface, local flow field disturbances caused by bed irregularities can be effectively avoided, thus ensuring the accuracy and comparability of sediment initiation studies. This is typically achieved through fine sieving of the sediment and careful laying and leveling underwater using tools such as scrapers.

[0046] Secondly, four high-speed cameras were installed around the measurement body, and their positions and angles were adjusted to meet the Scheimpflug condition. The measurement body was precisely defined as the region extending from the flat bed surface to a height several times the target particle size above it. Installing four high-speed cameras provided multi-view observation capabilities, which is fundamental to achieving high-precision 3D reconstruction, ensuring sufficient field-of-view overlap and viewing angle diversity. The Scheimpflug condition refers to the condition where, when the object plane, lens plane, and image plane are not parallel, tilting the lens to make these three planes intersect by a line, thus ensuring that the image on the entire object plane is clearly focused. In 3D measurement, especially at tilted angles, this is crucial for maintaining the clarity of the depth direction of the measurement body, effectively avoiding image blurring caused by depth-of-field limitations, thereby improving the accuracy of particle identification and 3D reconstruction. Precisely limiting the measurement body to the region extending from the flat bed surface to a height several times the target particle size above it allows measurement resources to be concentrated in the core area where sediment initiation occurs, thereby improving the measurement resolution and accuracy of this critical area.

[0047] Next, the laser output is adjusted until a cubic light spot uniformly covers the measurement object. The laser is used to illuminate tracer particles and sediment in the water flow. Forming a cubic light spot that uniformly covers the measurement object means that the illumination area precisely matches the measurement object and that the light intensity distribution is uniform. Uniform illumination is a crucial prerequisite for ensuring image quality, accurate particle identification, and subsequent image segmentation and 3D reconstruction. Non-uniform light spots lead to large differences in image brightness, increasing the difficulty and error of image processing. This is typically achieved by precisely adjusting the laser power, the position of beam-shaping optical elements (such as cylindrical lens groups and spherical lenses), and the focal length.

[0048] Subsequently, the 3D calibration board is placed within the measurement body, and all high-speed cameras are simultaneously triggered to acquire several calibration images from different angles. Based on these images, a mapping relationship between the pixel coordinates of each high-speed camera and 3D world coordinates is established, and the calibration error is evaluated. The 3D calibration board typically has markers with known precise 3D coordinates; placing it within the measurement body ensures that the calibration covers the actual measurement area. Simultaneous triggering of all high-speed cameras ensures that all cameras capture the calibration board at the same time, while acquiring multiple images from different angles provides sufficient information for accurate calibration. Calibration, converting pixel coordinates on a 2D image into world coordinates in 3D space, is a necessary step for 3D reconstruction. Evaluating the calibration error quantifies the accuracy of the measurement system, providing a reliable basis for subsequent data analysis.

[0049] Finally, tracer particles are released into the water. These particles are used to mark the water flow. Releasing them into the water, ensuring their uniform distribution, and allowing them to move with the flow forms the basis for subsequent calculations of the water flow phase velocity field. The particle size, density, and optical properties of the tracer particles must be matched to the water flow characteristics and measurement methods to ensure they accurately follow the flow and are easily captured by the camera.

[0050] Through the detailed configuration and calibration steps described above, this application ensures a high degree of standardization of the experimental environment and the accuracy of the measurement system. Specifically, laying a smooth sediment bed with the target particle size provides uniform and repeatable initial conditions for sediment initiation; four high-speed cameras, combined with Scheimpflug conditions and a precisely defined measurement volume, greatly improve the clarity of image acquisition and the accuracy of 3D reconstruction, effectively avoiding problems such as depth-of-field blurring and unclear measurement areas; uniform cubic spot illumination ensures clear visibility of tracer particles and sediment particles throughout the measurement volume, providing high-quality raw data for image segmentation; and precise calibration and error assessment based on a 3D calibration plate lay the foundation for accurate conversion from pixel coordinates to 3D world coordinates, thereby ensuring the reliability of subsequent 3D reconstruction and velocity field calculations. The deployment of tracer particles provides necessary markers for calculating the velocity field of the water flow phase. These measures work together to significantly improve the accuracy and reliability of the synchronous measurement method for the entire flow field of open channel sediment initiation, providing high-quality experimental data for in-depth analysis of the sediment initiation mechanism.

[0051] In response, this application further proposes a step for synchronous acquisition of three-dimensional flow field data under critical conditions, including: gradually increasing the flow rate of the test section water tank; when the first sediment particle moves on the leveled bed surface, it is recorded as a critical start-up state, and at the same time, all high-speed cameras and lasers are triggered by the synchronous controller to synchronously acquire a continuous time series of multi-view image sequences at a fixed frame rate; the flow rate of the test section water tank is increased, and the process of acquiring multi-view image sequences is repeated to obtain three-dimensional flow field data under different intensity start-up states.

[0052] Specifically, gradually increasing the flow rate in the test section of the flume aims to simulate the sediment initiation process by controlling the hydrodynamic conditions. Gradually increasing the flow rate means that the water velocity and shear stress will gradually increase until they reach a critical condition sufficient to drive the sediment particles to move. This gradual increase helps to accurately capture the critical point at which sediment particles move from rest to their first movement, avoiding missing the crucial initiation moment due to sudden changes in flow rate. This can be achieved by adjusting the pump speed, adjusting the opening of the flume inlet gate, or changing the flume slope, ensuring a stable and controllable increase in flow rate.

[0053] When the first sediment particle moves on the leveled bed surface, it is recorded as a critical start-up state. Simultaneously, a synchronous controller triggers all high-speed cameras and lasers to synchronously acquire a continuous time-series multi-view image sequence at a fixed frame rate. This is crucial for identifying and recording the critical start-up state of sediment. The determination of the "first sediment particle movement" can be achieved through observation assisted by a high-resolution camera or real-time monitoring based on image processing algorithms. Once the minute displacement of the first particle is observed or detected, it is immediately marked as a critical start-up state. At this point, the synchronous controller plays a central role, ensuring that all high-speed cameras and lasers are precisely triggered at the same moment to begin image acquisition at a preset fixed frame rate. The fixed frame rate ensures the uniformity of the time-series data, which is essential for subsequent velocity field calculations and trajectory tracking. Acquiring a continuous time-series multi-view image sequence means not only capturing the start-up instant but also recording the flow field and particle motion over a period before and after start-up to analyze changes in the water flow structure and the initial motion characteristics of the particles.

[0054] By increasing the flow rate in the test flume and repeatedly acquiring multi-view image sequences, three-dimensional flow field data under different initiation intensities were obtained. The aim was to acquire comprehensive data on sediment initiation under varying flow intensities to study the universality and variability of the initiation mechanism. After initially capturing the critical initiation state, further increasing the flow rate in the test flume simulated stronger flow conditions, leading to the initiation of more sediment particles and even large-scale sediment transport. Each increase in flow rate was followed by repeating the monitoring and synchronous acquisition process, resulting in multi-view image sequences corresponding to different initiation intensities (e.g., mild initiation, moderate initiation, widespread initiation, etc.). This systematic data acquisition allows for the construction of a database encompassing various initiation states, providing rich data support for subsequent analysis of the relationship between different hydrodynamic parameters and sediment initiation probabilities and modes.

[0055] In the synchronous measurement of the entire flow field during sediment initiation in open channel flow, accurately extracting information on tracer particles and sediment particles from multi-view image sequences and precisely reconstructing them into three-dimensional spatial data is crucial for subsequent flow field and particle motion analysis. However, in complex two-phase flow environments, tracer particles and sediment particles may have different optical properties, sizes, and distribution densities, and the images may contain noise, occlusion, and other issues. This poses a challenge to efficiently and accurately segmenting and reconstructing the three-dimensional images acquired by all high-speed cameras at every moment, especially when it is necessary to distinguish different phase states and obtain their respective precise three-dimensional information.

[0056] To address this, this application further proposes optimized steps for multiphase flow image segmentation and 3D reconstruction. Specifically, these steps include: processing the images from each high-speed camera at each time step in the acquired multi-view image sequence to identify and segment the tracer particle region and the sediment particle region, generating tracer particle images and sediment particle images; and reconstructing the 3D concentration field of tracer particles and the 3D spatial coordinates of each sediment particle in the entire measurement body step by step using the MART algorithm or SART algorithm for the segmented tracer particle images and sediment particle images, respectively.

[0057] This process involves processing images from each high-speed camera at each moment in the acquired multi-view image sequence to identify and segment tracer particle regions and sediment particle regions, generating tracer particle images and sediment particle images. The aim is to accurately separate tracer particles and sediment particles from the original multi-view image sequence, providing clean and accurate 2D projection data for subsequent 3D reconstruction. Specifically, for each image acquired by each high-speed camera at each moment, image processing techniques can be used for analysis. For example, deep learning-based image segmentation algorithms, such as U-Net and Mask R-CNN, can be used to classify pixels in the image using pre-trained models, thereby identifying and segmenting tracer particle regions, sediment particle regions, and background regions. Alternatively, traditional image processing methods, such as thresholding, edge detection, and connected component analysis, can be combined to differentiate tracer particles and sediment particles based on their size, shape, brightness, and other characteristics. Through these processes, images containing only tracer particles (tracer particle images) and images containing only sediment particles (sediment particle images) can be generated, effectively removing background interference and providing dedicated data input for subsequent processing in different phases.

[0058] Based on this, for the segmented tracer particle images and sediment particle images, the MART or SART algorithm is used to reconstruct the three-dimensional concentration field of tracer particles and the three-dimensional spatial coordinates of each sediment particle within the entire measurement volume step by step. This step aims to transform the segmented two-dimensional image information into physical quantities in three-dimensional space. Specifically, for tracer particle images, since tracer particles are usually numerous and densely distributed, the main objective is to reconstruct their three-dimensional concentration field within the measurement volume. This can be achieved by dividing the measurement volume into a series of voxels and then using algebraic reconstruction algorithms such as MART (Multiplicative Algebraic Reconstruction Technique) or SART (Simultaneous Algebraic Reconstruction Technique) to iteratively invert the concentration value of tracer particles within each voxel based on the projection intensity of tracer particles from multiple viewpoints, thereby obtaining the transient three-dimensional concentration field. For sediment particle images, since sediment particles are discrete individuals, the goal is to obtain the precise three-dimensional spatial coordinates of each sediment particle. This can be achieved by identifying the two-dimensional projection of each sediment particle from different perspectives, and then using multi-view geometry principles combined with MART or SART algorithms for iterative optimization to accurately determine the three-dimensional center coordinates of each sediment particle. Time-step reconstruction ensures continuous capture of the flow field and particle motion dynamics, providing fundamental data for subsequent transient analysis.

[0059] The above technical solution refines and optimizes the image segmentation and 3D reconstruction process of multiphase flow. First, meticulous processing of images acquired by each high-speed camera at each time step accurately identifies and segments the tracer particle region and sediment particle region, effectively avoiding confusion between different phases and providing high-quality input data for subsequent processing. Second, 3D reconstruction is performed using either the MART or SART algorithm, tailored to the different characteristics of tracer particles and sediment particles. For tracer particles, reconstructing their 3D concentration field accurately reflects the transient distribution and structure of the water flow phase; for sediment particles, reconstructing their 3D spatial coordinates precisely obtains the instantaneous position of each particle. This separate processing and reconstruction method ensures accurate acquisition of key information for both the water flow phase and the sediment phase, significantly improving the accuracy and reliability of the 3D flow field and sediment particle motion data, providing a solid data foundation for subsequent multiphase flow coupling analysis and sediment initiation mechanism research.

[0060] In some of the embodiments described above in this application, although the instantaneous velocity field of the water flow phase and the instantaneous velocity vector of the sediment particles can be obtained separately, there is a lack of a direct and quantitative index to characterize the relative motion between the two when analyzing the interaction between the water flow and the sediment particles. This limits the in-depth understanding and accurate modeling of the sediment initiation mechanism.

[0061] In this regard, this application further proposes that in step S4, the relative motion between sediment particles and water flow is quantified by calculating the velocity slip vector ΔV, wherein the velocity slip vector ΔV is equal to the instantaneous velocity vector Vp of the sediment particles minus the instantaneous velocity vector Up of the water flow phase.

[0062] The velocity slip vector ΔV is a physical quantity describing the relative velocity between sediment particles and the surrounding water flow. It is calculated by subtracting the instantaneous velocity vector Up of the water flow at the particle's location from the particle's instantaneous velocity vector Vp. The instantaneous velocity vector Vp is obtained by three-dimensionally tracking the sediment particles, matching the same particle at adjacent time points, and calculating the ratio of its displacement to the time interval. The instantaneous velocity vector Up of the water flow is extracted from the transient three-dimensional three-component water velocity vector field U(x,y,z,t) calculated using a three-dimensional multi-scale cross-correlation algorithm, taking the velocity value corresponding to the sediment particle's current three-dimensional spatial coordinates (x,y,z) and time t. Since the positions of sediment particles do not usually coincide perfectly with the grid points of the velocity field, when extracting Up, the eight adjacent velocity field grid points surrounding the current three-dimensional spatial coordinates of the sediment particle are selected. Trilinear interpolation is then performed based on the relative distances between the sediment particle coordinates and the eight adjacent velocity field grid points to obtain the instantaneous velocity Up(t) of the water flow at the particle's location. When the particle is near the boundary of the measurement volume and cannot form a complete eight-point surrounding grid, the nearest effective velocity field grid point is selected for distance-weighted interpolation. In this way, the instantaneous motion state of each sediment particle relative to its surrounding water flow can be accurately quantified.

[0063] In sediment initiation studies, traditional methods are limited by spatial dimension, phase resolution, and the ability to analyze microscopic mechanisms, making it difficult to obtain a complete three-dimensional flow field and simultaneously measure multiphase motion. The core innovation of this application lies in combining a multi-view synchronous image acquisition subsystem with a three-dimensional volume illumination subsystem in a collaborative manner, and introducing an intelligent data processing and measurement subsystem. This enables synchronous and accurate measurement of the water flow phase and sediment particle phase in three-dimensional space, overcoming the limitations of traditional methods in terms of spatial dimension, phase resolution, and microscopic mechanism analysis, and achieving the effect of comprehensively revealing the microscopic dynamic mechanism of sediment initiation.

[0064] See Figure 1As shown in the embodiments of this application, a synchronous measurement system for the entire flow field of sediment initiation in open channel flow is also disclosed. This system is used to realize the aforementioned synchronous measurement method for the entire flow field of sediment initiation in open channel flow. It includes: an experimental flume subsystem, comprising a test flume, a flow stabilization device, and a flow and water level control system; the bottom of the test flume is covered with the sediment to be measured, forming a flat bed surface; a three-dimensional volume illumination subsystem, employing a high-energy continuous laser with an output wavelength of 532nm, which expands and focuses the laser beam through a cylindrical lens group and a spherical lens to form a cubic illumination area with a thickness of 5-10mm, used to illuminate the measurement body in the flume covering the flat bed surface and the water flow above; a multi-view synchronous image acquisition subsystem, comprising at least four identical high-speed cameras arranged around the measurement body at different spatial azimuth angles, equipped with Scheimpflug adapters to adjust the imaging plane, all high-speed cameras connected to a synchronous controller; a tracer and particle subsystem, comprising tracer particles matched to the water density for marking water flow motion, the tracer particles having a particle size of 10-20µm; and the object of observation. The intelligent data processing and measurement subsystem for sediment particles includes: a multiphase image segmentation module configured to segment the original image based on a deep learning algorithm, distinguishing different phase states, including background, tracer particles, sediment particles, and bubbles; a three-dimensional tomography reconstruction module configured to reconstruct the instantaneous distribution of tracer particles and sediment particles in three-dimensional space within the measurement body based on calibration images acquired by a high-speed camera, using the MART or SART algorithm; a multiphase three-dimensional velocity field calculation module configured to calculate the three-dimensional three-component velocity field of the continuous phase of the water body using a three-dimensional multi-grid cross-correlation algorithm for the reconstructed tracer particle cloud; and to obtain the three-dimensional trajectory, velocity, and angular velocity of each sediment particle using a three-dimensional particle tracking velocimetry algorithm for the reconstructed individual sediment particles; a post-processing and diagnostic module configured to provide three-dimensional eddy recognition criteria, three-dimensional pulsating kinetic energy calculation, Reynolds stress tensor calculation, interphase velocity slip statistics, particle initiation probability analysis, and three-dimensional streamline and pulsating velocity field visualization tools; and a calibration and evaluation module configured to perform three-dimensional calibration and evaluate calibration errors.

[0065] This system, through the collaborative operation of a multi-view high-speed camera and a cubic illumination area, ensures complete coverage of the measured object in three-dimensional space, effectively overcoming the spatial dimensional limitations of traditional two-dimensional measurements. It can acquire velocity components perpendicular to the mainstream direction and the complete three-dimensional vortex structure. A deep learning-based multiphase image segmentation module accurately distinguishes phases such as water flow, sediment particles, and bubbles, solving the limitation of phase resolution and enabling synchronous measurement of the velocity of each phase. A three-dimensional tomography reconstruction module, combined with MART or SART algorithms, reconstructs the three-dimensional concentration field of tracer particles and the three-dimensional spatial coordinates of sediment particles. A multiphase three-dimensional velocity field calculation module further quantifies the water flow velocity field and particle trajectory, enabling accurate calculation of the velocity slip vectors of particles and water flow. The post-processing and diagnostic module identifies three-dimensional vortex structures based on the Q-criterion, determines the vortex structure boundary based on the Q-threshold connected region, determines the vortex core coordinates based on the local maxima of the Q value, and performs correlation analysis based on the vortex intensity sequence and the spatiotemporal characteristics of particle initiation, compensating for the deficiency in microscopic mechanism analysis. Therefore, this system provides reliable technical support for determining the critical shear stress for sediment initiation and elucidating the microscopic dynamic mechanism, significantly improving the accuracy and depth of sediment initiation research.

[0066] In some embodiments described above in this application, the multi-view synchronous image acquisition subsystem is configured to use a high-speed camera for image acquisition. However, in practical applications, if the parameters of the high-speed camera are not sufficiently optimized—for example, if the resolution is insufficient, motion blur exists, or the signal-to-noise ratio is low—it will be difficult to accurately capture the rapid, transient, and subtle physical processes of sediment particle initiation in open channel flow. This will directly affect the accuracy of subsequent image segmentation, 3D reconstruction, and velocity field calculation, thereby limiting the in-depth analysis of sediment initiation mechanisms and the accurate determination of critical conditions.

[0067] In this regard, this application further specifies the parameters of the aforementioned high-speed camera, namely, a full-frame resolution of not less than 2560×1920 pixels, a global shutter, a full-frame frame rate of not less than 2000fps, a minimum exposure time of ≤100ns, a focal length lens of 100mm / F2.8 for fine observation and 24mm / F1.4 for wide field of view observation, and is equipped with a 532nm narrowband filter.

[0068] Specifically, high-resolution image acquisition capabilities, i.e., a full-frame resolution of at least 2560×1920 pixels, are crucial for accurately capturing the subtle features of sediment and tracer particles. In multiphase flow image segmentation and 3D reconstruction, higher pixel density provides richer spatial information, thereby improving the accuracy of particle identification and the precision of 3D coordinate reconstruction. For example, within the measurement volume, even tiny tracer particles or newly initiated sediment particles can occupy sufficient pixels in the image, avoiding detail loss or misjudgment due to insufficient resolution. Global shutter technology ensures that all pixels on the image sensor begin and end exposure simultaneously. This is particularly important for high-speed moving sediment particles and tracer particles in water flow, as it effectively avoids image distortion caused by the rolling shutter effect. In rapidly changing flow fields, global shutter ensures that the positional information of all moving objects is synchronized within a single frame, thus providing accurate instantaneous position data for subsequent 3D reconstruction and velocity field calculation. A full-frame frame rate of at least 2000 fps enables the system to capture the rapid dynamic changes during the initiation of sediment particles and the transient pulsations of water flow. At the critical state of sediment initiation, particle movement often occurs instantaneously, and the water flow structure may also evolve within a very short time. A frame rate of at least 2000 frames per second provides sufficient temporal resolution to ensure continuous capture of every minute stage of particle movement from rest to motion, as well as the rapid evolution of the water flow velocity field, thus providing ample data points for accurate analysis of the initiation mechanism and calculation of instantaneous velocity. Extremely short exposure times, such as no more than 100 nanoseconds, are crucial for "freezing" high-speed motion. In high-speed flow fields, the instantaneous velocities of tracer particles and sediment particles can be extremely high. If the exposure time is too long, moving objects will create motion blur in the image, which will severely affect the accurate identification of particle positions and the accuracy of 3D reconstruction. By limiting the exposure time to within 100 nanoseconds, motion blur can be minimized, ensuring clear and sharp particle images even at the fastest flow rates. Two lens configurations with different focal lengths and apertures are offered: 100mm / F2.8 for fine observation and 24mm / F1.4 for wide-field observation, designed to meet different observation needs. The 100mm / F2.8 lens offers high magnification and a shallow depth of field, suitable for fine observation of specific areas (e.g., near the bed surface) to capture the microscopic details of sediment particle initiation and the fine structure of the near-wall flow field. The 24mm / F1.4 lens, on the other hand, provides a wider field of view and greater depth of field, suitable for observing the macroscopic flow field structure and particle distribution across the entire measurement volume, ensuring comprehensive coverage of the measurement area without sacrificing detail. This configuration provides flexibility, allowing researchers to choose the most appropriate observation scale according to their experimental objectives.A 532nm narrowband filter is positioned in front of the high-speed camera lens. Its main function is to selectively allow light with the same wavelength as the laser (output wavelength of 532nm) used in the 3D volume illumination subsystem to pass through, while effectively blocking ambient light of other wavelengths. This significantly improves the signal-to-noise ratio of the image, allowing the scattered light from tracer particles and sediment particles under laser illumination to be clearly captured without interference from background light or other stray light, thus ensuring image quality and improving the accuracy of subsequent image segmentation and 3D reconstruction.

[0069] Through the aforementioned technical solutions, the multi-view synchronous image acquisition subsystem can obtain high-quality, high temporal resolution, and high spatial resolution image data. High resolution and a global shutter ensure image clarity and distortion-free operation, laying the foundation for accurate identification and 3D reconstruction of tracer particles and sediment particles. High frame rate and extremely short exposure time guarantee complete capture of the transient process of sediment particle initiation, effectively avoiding motion blur and enabling accurate tracking of particle instantaneous velocity and trajectory. The configuration of lenses with different focal lengths allows the system to perform detailed observations of the microscopic mechanisms of sediment initiation while also comprehensively grasping the macroscopic flow field within the entire measurement body. A 532nm narrowband filter further improves the signal-to-noise ratio of the image, ensuring data reliability under complex lighting conditions. The synergistic effect of these optimized camera parameters greatly improves the accuracy and reliability of synchronous measurement of the entire flow field of sediment initiation in open channel flow. This allows researchers to analyze more deeply the critical conditions for sediment initiation, the interaction between particles and water flow, and the influence of vortex structure on initiation, thereby obtaining more accurate critical shear stress or frictional velocity for sediment initiation.

[0070] The following will explain the solution of this application with reference to specific embodiments: Example 1: This embodiment provides a three-dimensional measurement system for studying the initiation of 0.2mm quartz sand.

[0071] Water tank subsystem: A glass water tank measuring 20m long, 1m wide, and 0.5m high, equipped with a variable frequency water pump and a tailgate. (Scene illustration shown) Figure 2 As shown.

[0072] The 3D volumetric lighting subsystem employs a 30W continuous 532nm laser, which, through optical components, forms a cubic lighting area approximately 8mm thick and 150mm x 80mm in cross-section, covering a range of 0-30mm above the bed surface. The lighting scene is as follows: Figure 3 As shown.

[0073] Image acquisition subsystem: Employs four high-speed cameras with a resolution of 2560×1920 pixels and a full-frame frame rate of 2000fps. The cameras are arranged in a circular pattern at approximately 30-degree intervals, mounted on a finely adjustable professional pan-tilt head, and all are equipped with 100mm macro lenses, 532nm narrowband filters, and Scheimpflug adapters. A single 8-channel synchronous controller (1ns time accuracy) provides unified control and triggering.

[0074] Tracer particles: 20µm fluorescent polymer microspheres (density 1.05g / cm³), with a 550nm long-pass filter added in front of the camera to enhance the signal-to-noise ratio.

[0075] Data Processing Subsystem: A workstation equipped with an Intel i9-14900HX processor, 64GB of RAM, and an NVIDIA RTX 4080 graphics card. It runs commercial software with integrated AI segmentation and tomographic PIV functions (such as LaVision's DaVis 10+TomoPIV module, and custom-developed particle segmentation and 3D-PTV coupled analysis scripts).

[0076] The measurement method process is as follows: The sieved 0.2mm quartz sand was laid flat on the test section of the water tank, with a thickness of 20mm.

[0077] Four cameras and a laser bulk optical system were set up and calibrated for three-dimensional stereo calibration, with the reprojection error controlled within 0.1 pixels.

[0078] Slowly increase the flow rate to near the critical point. When individual particles on the bed surface begin to vibrate as observed through the preparatory experimental observation window or auxiliary monitoring, manually trigger the system to collect 2 seconds of data (4000 sets of three-dimensional snapshots).

[0079] Data processing: The software first performs AI background segmentation and particle recognition on the four images; then reconstructs the coordinates of the tracer particles and sediment particles respectively; performs 3D cross-correlation calculation on the tracer particles to obtain the transient 3D3C velocity field sequence of 800×400×50 grid points; and performs 3D-PTV tracking on the sediment particles.

[0080] Analysis: The velocity gradient tensor is calculated based on a transient 3D3C velocity field sequence with an 800×400×50 grid, and the Q-criterion is used to identify the three-dimensional vortex structure. Specifically, the 70th percentile of the positive Q value at each time step is taken as the Q threshold, denoted as Q0. T , with a positive Q value greater than Q TFurthermore, a three-dimensional vortex structure region was determined based on a set of voxels connected to twenty-six neighborhoods. The grid point with the largest Q value within the region was used as the vortex core coordinate, and the sum of the products of the Q value and the voxel volume within the region was used as the vortex intensity. For the three-dimensional vortex structure within a continuous time step, matching was performed according to the rule that the distance between the vortex core coordinates is no more than five times the velocity field grid spacing and the voxel overlap ratio is no less than 20%, resulting in a vortex structure evolution sequence. Visualization showed that the high-shear regions corresponding to the two high-Q regions on the sides of a hairpin-shaped three-dimensional vortex structure identified by the Q criterion coincided with the positions of several initially rolling sediment particles identified by 3D-PTV. Quantitative analysis showed that at the moment of particle initiation, the instantaneous shear stress of the local water flow at the particle's location was 2-3 times the average value, and the particle had a significant lateral velocity. The particle's initiation position was located within the boundary of the three-dimensional vortex structure, and the time difference between the initiation moment and the vortex structure passing through the particle's neighborhood was no more than one image sampling time interval. Therefore, it was determined that the sediment particle initiation event and the three-dimensional vortex structure were spatiotemporally correlated.

[0081] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this invention, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0082] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for synchronous measurement of the entire flow field of sediment initiation in open channel flow, characterized in that, Includes the following steps: S1. Configuration and Calibration: In the test section, mud and sand particles and tracer particles were laid in the water tank. High-speed cameras were set up at different angles and lasers were set up for illumination. The high-speed cameras were turned on to obtain several calibration images. The mapping relationship between the pixel coordinates of each high-speed camera and the three-dimensional world coordinates was established based on the calibration images. S2. Synchronous acquisition of three-dimensional flow field data under critical state: The flow rate of the test section water tank was gradually increased. When the first sediment particle moved, it was recorded as the critical state. All high-speed cameras and lasers were triggered simultaneously to acquire a continuous time series of multi-view image sequences. S3. Multiphase Flow Image Segmentation and 3D Reconstruction: The acquired multi-view image sequence identifies tracer particles and sediment particles and segments the tracer particle region and sediment particle region to generate a mask image; based on the MART iterative algorithm or SART iterative algorithm, the three-dimensional concentration field of tracer particles and the three-dimensional spatial coordinates of each sediment particle in the entire measurement body are reconstructed step by step over time. S4. Calculation and Mechanism Analysis of Multiphase Three-Dimensional Velocity Field: 4.1 Calculation of water flow phase velocity field: Based on the three-dimensional concentration field of the continuous time step, the transient three-dimensional three-component water velocity vector field U(x,y,z,t) is calculated by applying the three-dimensional multi-scale cross-correlation algorithm, where x, y and z represent position coordinates and t represents time; 4.2 Sediment phase motion analysis: Three-dimensional particle tracking is performed on the three-dimensional spatial coordinates, and the same particle at adjacent time points is matched to obtain the three-dimensional trajectory, instantaneous velocity vector Vp(t), and angular velocity of each moving sediment particle; 4.3 Multiphase Flow Coupling Analysis: Extract the instantaneous velocity Up(t) of the water flow phase at the location of each moving sediment particle at time t, and calculate the velocity slip vector between the sediment particles and the water flow; extract the three-dimensional vortex structure and analyze the spatiotemporal correlation between the vortex core position, intensity, evolution and particle initiation position and time; statistically analyze the relationship between the spatial distribution probability of particle initiation under critical conditions and local hydrodynamic parameters; based on the three-dimensional three-component water velocity vector field U(x,y,z,t), the bed shear stress tensor is determined by fitting the near-wall velocity profile or directly calculating the bed shear stress tensor, and the critical shear stress or frictional velocity for sediment initiation is determined based on the synchronously observed particle initiation.

2. The method for synchronous measurement of the entire flow field of sediment initiation in open channel flow according to claim 1, characterized in that, Step S1 specifically includes: 1.1 Lay mud and sand particles of the target particle size at the bottom of the test section water tank to form a flat bed surface; 1.2 Install four high-speed cameras around the measuring body, and adjust their positions and angles. Under Scheimpflug conditions, the measuring body is the area from the flat bed surface to a height several times the target particle size above it. 1.3 Adjust the laser output until a cubic light spot uniformly covers the measurement body; 1.4 Place the three-dimensional calibration plate within the measurement body, synchronously trigger all high-speed cameras, acquire several calibration images from different angles, establish the mapping relationship between the pixel coordinates of each high-speed camera and the three-dimensional world coordinates, and evaluate the calibration error. 1.

5. Release tracer particles into the water.

3. The method for synchronous measurement of the entire flow field of sediment initiation in open channel flow according to claim 2, characterized in that, Step S2 specifically includes: 2.1 Gradually increase the flow rate of the test section's water tank; 2.2 When the first sediment particle moves on the leveled bed surface, it is recorded as the critical start-up state. At the same time, all high-speed cameras and lasers are triggered by the synchronous controller to synchronously acquire a continuous time series of multi-view image sequences at a fixed frame rate. 2.

3. Increase the flow rate of the test section water tank and repeat the process of obtaining multi-view image sequences in step 2.2 to obtain three-dimensional flow field data under different starting conditions.

4. The method for synchronous measurement of the entire flow field of sediment initiation in open channel flow according to claim 3, characterized in that, Step S3 specifically includes: 3.1 Process the images from each high-speed camera at each moment in the acquired multi-view image sequence, identify and segment the tracer particle region and the sediment particle region, and generate tracer particle images and sediment particle images. 3.2 For the segmented tracer particle image and sediment particle image, respectively, the MART algorithm or SART algorithm is used to reconstruct the three-dimensional concentration field of tracer particles and the three-dimensional spatial coordinates of each sediment particle in the entire measurement body step by step.

5. The method for synchronous measurement of the entire flow field of sediment initiation in open channel flow according to claim 4, characterized in that, In step S4, the velocity slip vector ΔV = Vp - Up, where ΔV represents the velocity slip vector, Vp is the instantaneous velocity vector of the sediment particles, and Up is the instantaneous velocity vector of the water flow phase.

6. A synchronous measurement system for the entire flow field of sediment initiation in open channel flow, used to implement the synchronous measurement method for the entire flow field of sediment initiation in open channel flow as described in any one of claims 1-5, characterized in that, include: The experimental water tank subsystem includes a test section water tank, a flow stabilizing device, and a flow and water level control system; the bottom of the test section water tank is covered with the sediment to be tested to form a flat bed surface; The three-dimensional volume illumination subsystem uses a high-energy continuous laser with an output wavelength of 532nm. The laser beam is expanded and focused through a cylindrical lens group and a spherical lens to form a cubic illumination area with a thickness of 5-10mm, which is used to illuminate the measuring body in the water tank that covers the flat bed surface and the water flow above. The multi-view synchronous image acquisition subsystem includes at least four identical high-speed cameras arranged around the measuring body at different spatial azimuth angles and equipped with Scheimpflug adapters to adjust the imaging plane. All high-speed cameras are connected to a synchronization controller. A tracer and particle subsystem, comprising tracer particles with a density matched to water for marking water flow motion, wherein the tracer particles have a particle size of 10-20 µm; And the sediment particles that are the objects of observation; The intelligent data processing and measurement subsystem includes: The multiphase image segmentation module is configured to segment the original image based on a deep learning algorithm and distinguish different phase states, including background, tracer particles, mud and sand particles and bubbles. The three-dimensional tomography reconstruction module is configured to reconstruct the instantaneous distribution of tracer particles and sediment particles in three-dimensional space based on calibration images acquired by a high-speed camera and using the MART or SART algorithm. The multiphase three-dimensional velocity field calculation module is configured to use a three-dimensional multigrid cross-correlation algorithm to calculate the three-dimensional three-component velocity field of the continuous phase of the water body for the reconstructed tracer particle cloud; and to use a three-dimensional particle tracking velocimetry algorithm to obtain the three-dimensional trajectory, velocity and angular velocity of each sediment particle for the reconstructed individual sediment particles. The post-processing and diagnostic module is configured to provide three-dimensional eddy identification criteria, three-dimensional pulsating kinetic energy calculation, Reynolds stress tensor calculation, interphase velocity slip statistics, particle initiation probability analysis, and three-dimensional streamline and pulsating velocity field visualization tools. The calibration and evaluation module is configured to perform three-dimensional calibration and evaluate calibration errors.

7. The open channel flow sediment initiation full-field synchronous measurement system according to claim 6, characterized in that, The parameters of the high-speed camera are as follows: full-frame resolution of not less than 2560×1920 pixels, global shutter, full-frame frame rate of not less than 2000fps, minimum exposure time ≤100ns, focal length lens of 100mm / F2.8 for fine observation and 24mm / F1.4 for wide field of view, 532nm narrowband filter.

Images