A microfluidic chip and simulation method based on soil penetration erosion whole process simulation

Abstract

The application discloses a kind of based on soil penetration erosion whole process simulation microfluidic chip and simulation method, belong to geotechnical and environmental remediation technical field, it is in to solve the technical problem that existing technology cannot simulate the dynamic erosion process of soil truly.The microfluidic chip includes upper base layer, lower base layer, liquid injection pipe, liquid outflow pipe and the flow-through hole opened in both ends, the top surface of lower base layer is provided with liquid flow channel, trapezoidal groove, the long side of trapezoidal groove is parallel with liquid flow channel and is communicated, liquid injection pipe, liquid outflow pipe are communicated with both ends of liquid flow channel respectively by flow-through hole;Upper base layer is fixed with lower base layer by bonding.In simulation, soil is filled in trapezoidal groove and is shaped before simulation;When simulating, liquid is injected into the liquid injection pipe of microfluidic chip, the soil in the trapezoidal groove of penetration, erosion side is eroded by liquid flowing through the process of liquid flow channel, and the process of erosion, migration, collision, agglomeration of soil particles by liquid is observed and recorded in real time by high-speed microscopic imaging module.

Description

Technical Field

[0001] This invention belongs to the field of geotechnical and environmental remediation technology, and relates to soil infiltration and erosion simulation, and particularly to a microfluidic chip and simulation method based on the simulation of the entire process of soil infiltration and erosion. Background Technology

[0002] Soil erosion is a significant issue in agriculture and environmental management, involving complex processes such as the separation, migration, and deposition of soil particles. While extensive research has been conducted on the mechanisms of soil erosion during rainfall, the microscopic erosion processes remain largely unclear.

[0003] Patent application number 2024117713423 discloses an experimental device for simulating acid rain erosion of silt-stabilized soil, which includes an experimental unit. The experimental unit includes a sample space for placing solidified soil samples, a sample sensor set in the sample space for monitoring sample parameters, an acid tank for placing acid, and a spraying device for extracting acid from the acid tank and spraying the acid onto the outer surface of the solidified soil sample. The acid is a liquid prepared to a specific pH level. The spraying device sprays the acid onto the surface of the solidified soil sample, and the acid penetrates from the surface of the solidified soil sample into the interior of the solidified soil sample to simulate the erosion effect of acid rain on the solidified soil sample. The sample sensor is used to detect the erosion parameters of the solidified soil sample by the acid inside the solidified soil sample.

[0004] This experimental setup simulates acid rain by manually mixing acid to a specific pH level and spraying it onto the surface of a solidified soil sample to mimic the erosion effect of acid rain on the solidified silt. A sample sensor placed inside the solidified soil sample can collect sample parameter information to analyze the erosion parameters of the acid on the solidified soil sample.

[0005] In existing technologies, microfluidic chips can enable the observation of the movement of micron-sized particles on a small scale. Several studies have established soil pore networks based on microfluidic chips, using fluorescent labeling and transparent materials to investigate colloidal transport and biochemical reactions within soil pores. The high precision and controllability of microfluidic chips can help reveal the microscopic mechanisms of soil separation and transport processes at the microscale, overcoming the limitations of traditional experimental methods.

[0006] Patent application number 202510415525X discloses a centrifugal microfluidic chip, comprising: a chip body; an upstream cavity located on the chip body; a downstream cavity located on the chip body, with the distal end of the upstream cavity connected to the downstream cavity via a flow channel; and a repeatable switching valve disposed on the flow channel, containing a phase change material, which can repeatedly block or connect the upstream and downstream cavities by controlling heating and rotation speed. When it is necessary to block the flow channel, the phase change material is heated, controlling the rotation of the centrifugal microfluidic chip. At a first preset speed, the liquid phase change material moves from the proximal end to the distal end, sealing the upstream and downstream cavities. The microfluidic chip consists of a flow channel between the liquid and the phase change material (PCM). Heating is then stopped, allowing the liquid PCM to solidify. When it is necessary to connect the upstream and downstream cavities, the solid PCM is heated. Under negative pressure, the PCM changes from solid to liquid and flows towards the energy storage device through the third channel. Heating is stopped once the upstream and downstream cavities are connected, allowing the liquid PCM to solidify. Alternatively, the centrifugal microfluidic chip can be heated while simultaneously controlling its rotation speed to ensure the PCM flows at a certain velocity towards the energy storage device through the third channel. Heating is stopped once the upstream and downstream cavities are connected, allowing the liquid PCM to solidify. This microfluidic chip, composed of a three-channel system, a PCM, and a connected energy storage device, can conveniently achieve controlled batch release and precise quantitative release of upstream liquid. Its simple structure facilitates reduced manufacturing costs and simplifies the complexity of fluid controlled release.

[0007] The aforementioned centrifugal microfluidic chip is used for the separation of liquids such as reagents and samples, but its application in the study of soil erosion still faces technical bottlenecks: it is difficult to capture microscopic dynamic mechanisms such as particle-level fluid drag and pore water pressure changes, and the artificially controlled boundary conditions deviate significantly from the heterogeneity of the actual soil. In addition, it can only present static seepage scenarios and cannot dynamically reproduce the coupling effect of particle transport and pore evolution during erosion, and it lacks real-time multi-parameter synchronous monitoring capabilities. Summary of the Invention

[0008] The purpose of this invention is to solve the technical problem that existing technologies cannot realistically simulate the dynamic erosion process of soil, and to provide a microfluidic chip and simulation method based on the simulation of the entire process of soil infiltration and erosion.

[0009] To achieve the above objectives, the present invention specifically adopts the following technical solution:

[0010] A microfluidic chip based on the simulation of the entire process of soil infiltration and erosion includes an upper layer, a lower layer, a liquid injection pipe, and a liquid outflow pipe. Both the inlet and outlet ends of the upper layer and the lower layer are provided with flow holes.

[0011] The top surface of the lower base layer has a liquid flow channel and a trapezoidal groove, with the long side of the trapezoidal groove parallel to and connected to the liquid flow channel; the upper base layer is attached to the lower base layer, and the liquid injection pipe and liquid outflow pipe are connected to the two ends of the liquid flow channel through the flow holes at both ends of the lower base layer.

[0012] Furthermore, there is an angle α between the bottom surface of the liquid flow channel and the lower surface of the lower base layer; there is an angle α between the bottom surface of the trapezoidal groove and the lower surface of the lower base layer; the bottom surface of the liquid flow channel and the bottom surface of the trapezoidal groove gradually decrease in height along the liquid flow direction, and the angle α is less than 5°.

[0013] Furthermore, the bottom surface of the liquid flow channel is curved, and the radius of curvature of the bottom surface of the liquid flow channel gradually increases along the liquid flow direction;

[0014] The bottom surface of the trapezoidal groove is curved, and the radius of curvature of the bottom surface of the trapezoidal groove gradually increases along the liquid flow direction.

[0015] Furthermore, a flow guide protrusion is provided on the bottom surface of the liquid flow channel at the position corresponding to the flow channel of the trapezoidal groove.

[0016] Furthermore, along the direction close to the short side of the trapezoidal groove, the height of the guide protrusion above the bottom surface gradually decreases.

[0017] A method for simulating the entire process of soil infiltration and erosion is proposed, which uses the aforementioned microfluidic chip for simulating the entire process of soil infiltration and erosion to simulate the process of soil infiltration and erosion.

[0018] Before the simulation, the trapezoidal trough was filled with soil and shaped.

[0019] During the simulation, liquid is injected into the liquid injection tube of the microfluidic chip and flows out through the liquid outlet tube after passing through the liquid channel. As the liquid flows through the liquid channel, it penetrates and erodes the soil in the trapezoidal groove on one side. The process of erosion, migration, collision and aggregation of soil particles by the liquid is observed and recorded in real time by a high-speed microscopic imaging module.

[0020] Furthermore, the specific steps for simulating soil infiltration and erosion are as follows:

[0021] Step S1: Adhere the upper substrate to the lower substrate, install the liquid injection pipe and liquid outflow pipe and connect them to the liquid flow channel to obtain the microfluidic chip;

[0022] Step S2: The test soil is air-dried and sieved, water is added and mixed well to obtain a soil-water mixture solution; the soil-water mixture solution is filled into the trapezoidal groove of the microfluidic chip and dried and shaped.

[0023] Step S3: Liquid is injected into the liquid injection pipe by a precision pressure-driven pump. After passing through the liquid flow channel, the liquid flows out from the liquid outlet pipe. During the process of the liquid flowing through the liquid flow channel, it permeates and erodes the soil in the trapezoidal trench on one side.

[0024] Step S4: By controlling the pressure of the precision pressure drive pump, the flow rate of the injected liquid is adjusted to simulate the stripping and migration process of soil under different rainfall intensities; the high-speed microscopic imaging module is used to observe and record the process of erosion, migration, collision and aggregation of soil particles by the liquid in real time.

[0025] Step S5: Automatically identify and quantify the erosion rate and pattern of the images acquired by the high-speed microscopic imaging module.

[0026] Furthermore, in step S5, an image multi-particle motion tracking and recognition method is used for recognition. The specific steps of the image multi-particle motion tracking and recognition method are as follows:

[0027] Step S5-1, Particle identification and localization;

[0028] A multi-threaded parallel processing closed loop is constructed using the Rolling-ball background subtraction algorithm and the particle recognition function of the TrackPy library. The multi-threaded parallel processing closed loop is used to process the images acquired by the imaging module to obtain the coordinates (x, y) and particle size d of each frame.

[0029] Step S5-2, cross-frame particle correlation and tracking;

[0030] Using the built-in tracking algorithm of the TrackPy library, and combining Kalman filtering and nearest neighbor search tree strategies, the identification points of the same particle in different frames are associated to form a continuous trajectory with particle ID and temporal coordinates.

[0031] Step S5-3, Dynamic behavior determination;

[0032] Using a collision threshold as the trigger condition, candidate collision pairs are identified based on a nearest neighbor search tree, and the distance between the two particles in subsequent frames is tracked. If the distance between the two particles is consistently below the threshold and their relative velocity decreases monotonically, they are considered to be effectively adhered and counted. Otherwise, they are classified as instantaneous collisions. Ultimately, the erosion dynamics parameters of the entire process of "separation-transportation-reaggregation" of soil particles are quantitatively analyzed, and the erosion rate and mode are quantified.

[0033] Furthermore, in step S5-1, when processing the image acquired by the imaging module using a multi-threaded parallel processing closed loop, the specific steps are as follows:

[0034] Step S5-1-1 involves using a preprocessing pipeline consisting of color image grayscale conversion, color inversion, adaptive histogram equalization, and Gaussian filtering to perform reflection filling, enhancement, and masking processing on the image acquired by the high-speed microscopic imaging module.

[0035] Step S5-1-2: Based on the preset particle size sequence, the particle locator is called in parallel to fuse multi-scale candidate particle sets;

[0036] Step S5-1-3: Deduplication is achieved through pixel grid clustering. While ensuring full particle size coverage, false targets are eliminated, and the coordinates (x, y) and particle size d of each frame are output.

[0037] Furthermore, in step S5-1, the Rolling-ball background subtraction algorithm is expressed as:

[0038] ;

[0039] in, This represents the pixel value of the current frame image. Represents the pixel values ​​of the background model. This represents the threshold for foreground detection;

[0040] In step S5-2, a Kalman filter is used for filtering; the state equation of the Kalman filter is:

[0041] ;

[0042] The observation equation is:

[0043] ;

[0044] The prediction equation is:

[0045] ;

[0046] ;

[0047] The gain matrix being updated is:

[0048] ;

[0049] The status during the update is now updated as follows:

[0050] ;

[0051] The covariance update in the update is as follows:

[0052] ;

[0053] in, This represents the state transition matrix, which describes the changes in the state from time k to time k+1. This represents the state vector at time k; This represents the noise input matrix, describing how the noise affects the state. Indicates system noise; The observation matrix represents the relationship between the state and the observations. Indicates observation noise; This represents the optimal state estimate at time k after all observations are known. Let represent the error covariance matrix of the state estimate at time k; Represents the process noise covariance matrix; Indicates transpose; Let represent the error covariance matrix of the state estimation from time k to time k+1; This represents the error or noise level at time k+1; This represents the optimal state estimate after all observations are known from time k to time k+1; This represents the actual observed value obtained at k+1; Represents the identity matrix;

[0054] In step S5-2, the nearest neighbor search tree strategy specifically includes:

[0055] Step S5-2-1: Calculate the sample spacing in the tree nodes. :

[0056] ;

[0057] in, Indicates the total number of samples; , This represents the coordinates of soil particles within a series of frames. , Indicates particles Two coordinate points in consecutive frames;

[0058] Step S5-2-2, divide the sample into left and right subtrees, represented as:

[0059] ;

[0060] in, Represents a dataset, Indicates XX, Indicates XX; This represents the coordinates of the sample in the j-th dimension; This represents the median of the j-th dimension, used to construct a balanced tree structure;

[0061] Step S5-2-3, during the search process, if the samples in the tree node With query point If the distance to the nearest neighbor is less than the current best distance, then update the nearest neighbor:

[0062] ;

[0063] like If so, the recording of "candidate collision pairs" is triggered, and the subsequent relative velocity determination will proceed.

[0064] in, This represents the current optimal distance. Point The nearest neighbor, Indicates the collision detection threshold;

[0065] Step S5-2-4, if the query point With a certain subtree center The distance is greater than the radius of the sphere The sum of the distances to the current best result is used for pruning:

[0066] ;

[0067] in, Indicates query point With a certain subtree center The distance.

[0068] The beneficial effects of this invention are as follows:

[0069] 1. In this invention, a trapezoidal groove is set inside the microfluidic chip, and the test soil is filled into the trapezoidal groove. The upper and lower narrow structure of the trapezoidal groove can realistically simulate the slope erosion gradient. The liquid flows in from the flow hole and through the liquid channel and then flows through the trapezoidal groove. The liquid penetrates the test soil longitudinally. The constriction design at the bottom of the trapezoidal groove can also suppress the instantaneous "plugging" collapse of the entire test soil, making the erosion process smoother and more controllable. The trapezoidal groove with the inverted trapezoidal single cavity design can be flexibly filled with different types / sizes of particles and corresponding solutions as needed. Through precise control, it can quickly simulate various erosion conditions from gentle slope flow to steep slope rapid flow, which significantly improves the efficiency of simulating the entire dynamic erosion process of the test soil, reduces hardware costs, and realizes independent or coupled scanning tests of multiple parameters (such as soil quality, slope, and hydrodynamics) under the same cavity geometry and boundary conditions. It provides an unprecedented high-throughput and highly comparable experimental platform for studying the soil erosion mechanism under complex multi-factor interaction.

[0070] 2. In this invention, the bottom surfaces of the liquid flow channel and the trapezoidal groove are set as curved surfaces, and the flow guiding protrusions are set. The purpose is to guide the liquid and make it deflect to one side of the trapezoidal groove, so as to better penetrate and erode the experimental soil in the trapezoidal groove and improve the efficiency of simulating the whole process of dynamic soil erosion.

[0071] 3. In this invention, the precision pressure-driven pump can achieve precise control of ultra-low flow rates at the level of 0.1 μL / min. Combined with the temperature and pressure compensation unit, the impact of environmental fluctuations is reduced in real time, which significantly improves the controllability of the test process, the reliability of the data, and the overall operating efficiency.

[0072] 4. In this invention, an image multi-particle motion tracking and recognition method is used to realize the quantitative analysis of erosion dynamic parameters of the entire process of soil particle "separation-transportation-re-aggregation" from the original image sequence. This enables high-throughput, high-precision real-time quantitative analysis of complex soil erosion simulation processes, improving the efficiency and accuracy of simulating the entire process of dynamic soil erosion. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of the structure of the present invention;

[0074] Figure 2 This is a schematic diagram of the structure of the present invention;

[0075] Figure 3 This is a schematic diagram of the lower base layer structure in this invention;

[0076] Figure 4 yes Figure 3 A schematic diagram of the cross-section;

[0077] Figure 5 yes Figure 3 A schematic diagram of the cross-section;

[0078] Figure 6 This is a schematic diagram of the soil infiltration and erosion simulation system of the present invention;

[0079] Figure 7 These are before-and-after comparison images of particle migration during the experiment of this invention;

[0080] Figure 8 This is a grayscale image of particle migration during the experiment of this invention;

[0081] Figure 9 These are particle migration images identified during the experiments of this invention;

[0082] Figure 10 This is a distribution diagram of the erosion particle size of loess particles under different flow rates and NaCl concentrations during the experiment of this invention;

[0083] Figure 11 This is a grayscale image showing the erosion movement of soil particles in Songhu, Xinjiang, under different flow rates and NaCl concentrations during the experiments of this invention.

[0084] The attached figures are labeled as follows:

[0085] 1-Upper base layer, 2-Lower base layer, 3-Flow hole, 4-Liquid flow channel, 5-Trapezoidal groove, 6-Liquid injection pipe, 7-Liquid outflow pipe, 8-Bottom surface, 9-Guiding protrusion. Detailed Implementation

[0086] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 some embodiments of the present invention, but not all embodiments.

[0087] Therefore, all other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0088] Example 1

[0089] This embodiment provides a microfluidic chip for simulating the entire process of soil infiltration and erosion. Figures 1-3 As shown, it includes a microfluidic chip body containing horizontally arranged microchannels. The microfluidic chip body includes an upper layer 1 and a lower layer 2. The top surface of the lower layer 2 has a downwardly recessed liquid channel 4 and a trapezoidal groove 5 located on one side of the liquid channel 4 and communicating with the liquid channel 4. The long side of the trapezoidal groove 5 is parallel to and communicates with the liquid channel 4, and the short side of the trapezoidal groove 5 protrudes to the side opposite to the liquid channel 4, forming an inverted trapezoidal structure. Flow holes 3 are provided at both ends (inlet and outlet) of the upper layer 1 and the lower layer 2. The flow holes 3 on the upper layer 1 and the lower layer 2, located on the same side, are sequentially connected to the liquid channel 4. Liquid injection pipes 6 and liquid outflow pipes 7 are respectively provided at the flow holes 3 at both ends of the upper layer 1. The liquid injection pipes 6, flow holes 3, liquid channels 4, and liquid outflow pipes 7 are sequentially connected.

[0090] The upper base layer 1 and the lower base layer 2 are bonded together (and can be glued and fixed after bonding) to form a microfluidic chip body containing microchannels inside.

[0091] In this embodiment, the inner diameter of both the liquid outflow pipe 7 and the liquid inflow pipe 6 is 1 mm. The lower base layer 2 is made of PDMS material, the length of the liquid flow channel 4 is 42 mm, and the sum of the widths of the liquid flow channel 4 and the trapezoidal groove 5 is 3.5 mm. The upper base layer 1 is made of silica glass.

[0092] By setting a trapezoidal groove 5 inside the microfluidic chip and filling the test soil into the trapezoidal groove 5, the upper-wide and lower-narrow structure of the trapezoidal groove 5 can realistically simulate the slope erosion gradient. The liquid flows in from the flow hole 3 and through the liquid channel 4 and then through the trapezoidal groove 5. The liquid penetrates the test soil longitudinally. The constriction design at the bottom of the trapezoidal groove 5 can also suppress the instantaneous "plugging" collapse of the entire test soil, making the erosion process smoother and more controllable. The trapezoidal groove 5 with this inverted trapezoidal single-cavity design can be flexibly filled with different types / sizes of particles and corresponding solutions as needed. Through precise control, it can quickly simulate various erosion conditions from gentle slope flow to steep slope rapid flow, which significantly improves the efficiency of simulating the entire dynamic erosion process of the test soil, reduces hardware costs, and realizes independent or coupled scanning tests of multiple parameters (such as soil quality, slope, and hydrodynamics) under the same cavity geometry and boundary conditions. It provides an unprecedented high-throughput and highly comparable experimental platform for studying the soil erosion mechanism under complex multi-factor interaction.

[0093] Example 2

[0094] To change / adjust the flow velocity of the liquid in the channel, based on Example 1, the bottom surfaces of both the liquid channel 4 and the trapezoidal groove 5 are set as inclined surfaces, and along the liquid flow direction, the bottom surface of the upstream section is higher than that of the downstream section. For example... Figure 4 As shown, specifically:

[0095] An angle α exists between the bottom surface 8 of the liquid flow channel 4 and the lower surface of the lower base layer 2; an angle α also exists between the bottom surface 8 of the trapezoidal groove 5 and the lower surface of the lower base layer 2. The angles α at both locations are equal and less than 5°. By setting the bottom surfaces 8 of the liquid flow channel 4 and 5 to gradually decrease in height along the liquid flow direction, and making the bottom surface 8 of the trapezoidal groove 5 flush with the bottom surface 8 of the corresponding area of ​​the liquid flow channel 4, it is beneficial for the liquid to gradually accelerate downstream, improving the penetration and erosion effect of the liquid on the experimental soil within one side of the trapezoidal groove 5.

[0096] Preferably, a guide protrusion 9 can be provided on the bottom surface 8 of the liquid flow channel 4 at the position corresponding to the flow channel of the trapezoidal groove 5, and the height of the guide protrusion 9 above the bottom surface 8 gradually decreases along the direction close to the short side of the trapezoidal groove 5 (that is, the height of the guide protrusion 9 becomes smaller and smaller along the direction close to the short side of the trapezoidal groove 5). With this structure, the liquid can be guided and deflected to one side of the trapezoidal groove 5, thereby better penetrating and eroding the experimental soil in the trapezoidal groove.

[0097] Example 3

[0098] To change / adjust the flow rate of the liquid in the flow channel, a different implementation method than Embodiment 2 is provided. That is, as... Figure 5As shown, based on Example 1, the bottom surface 8 of the liquid flow channel 4 is set as a curved surface, and the radius of curvature of the bottom surface 8 of the liquid flow channel 4 gradually increases along the liquid flow direction; similarly, the bottom surface 8 of the trapezoidal groove 5 is also set as a curved surface, and the radius of curvature of the bottom surface 8 of the trapezoidal groove 5 gradually increases along the liquid flow direction. By setting the bottom surface 8 of the liquid flow channel 4 and the bottom surface of the trapezoidal groove 5 as curved surfaces, and increasing the radius of curvature in the downstream direction, it is beneficial for the liquid to gradually accelerate downstream, thereby improving the penetration and erosion effect of the liquid on the test soil in the trapezoidal groove 5 on one side.

[0099] Preferably, a flow-guiding protrusion 9 can be provided on the bottom surface 8 of the liquid flow channel 4 at the position corresponding to the flow channel of the trapezoidal groove 5, and the height of the flow-guiding protrusion 9 above the bottom surface 8 gradually decreases along the short side direction near the trapezoidal groove 5 (that is, the protrusion height of the flow-guiding protrusion 9 becomes smaller and smaller along the short side direction near the trapezoidal groove 5). With this structure, the liquid can be guided, causing the liquid to deflect to one side of the trapezoidal groove, thereby enabling better penetration and erosion of the experimental soil in the trapezoidal groove.

[0100] Example 4

[0101] This embodiment provides a soil infiltration and erosion simulation system based on a microfluidic chip, such as... Figure 6 As shown, it includes:

[0102] The microfluidic chip belonging to any of Examples 1-3 is used to fill soil and inject liquid to permeate and erode the soil;

[0103] A precision pressure-driven pump is used to pump liquid into a microfluidic chip and regulate the flow rate and pressure of the liquid.

[0104] The high-speed microscopic imaging module includes an atomic force microscope and a light source, used to capture images of soil particle motion within the microchannels of a microfluidic chip in real time.

[0105] The data monitoring and acquisition system, including sensors and analysis modules, is used to quantify erosion rates and patterns.

[0106] The specific method for using this simulation system is as follows:

[0107] The trapezoidal groove 5 of the microfluidic chip is placed vertically (with the long side of the trapezoidal groove 5 on top and the short side on the bottom), and the prepared montmorillonite suspension is slowly injected into the trapezoidal groove 5, so that the montmorillonite particles accumulate in the trapezoidal groove 5 under the action of gravity, and are dried and shaped at 50°C after settling.

[0108] The data monitoring and acquisition system is electrically connected to the precision pressure-driven pump. The sampling tube, the precision pressure-driven pump, and the microfluidic chip are connected through pipes. The other end of the microfluidic chip is connected to the waste liquid collection tube.

[0109] The pure water in the sampling tube is injected into the liquid channel 4 through the liquid injection tube 6 and flow hole 3 of the microfluidic chip by a precision pressure driven pump, and the fluid flow rate is controlled by adjusting the numerical control of the precision pressure driven pump so that the fluid flows through the liquid channel 4 at a flow rate of 2uL / min.

[0110] After the soil was fully saturated, the precision pressure-driven pump was repeatedly adjusted to maintain the liquid flow rate in liquid channel 4 at 10 μL / min, 15 μL / min, and 20 μL / min for a period of time, simulating the infiltration and erosion of the soil at different flow rates. An atomic force microscope combined with a high-speed microscopic imaging module was used to capture, observe, and record images of soil particle erosion, migration, collision, and aggregation at the observation section for subsequent analysis and processing (e.g., ...). Figure 7 , Figure 8 (as shown)

[0111] The data monitoring and acquisition system analyzes and processes the acquired images, and uses an image multi-particle motion tracking and recognition method to accurately identify effective particles (such as...). Figure 9 As shown in the figure, the erosion rate and pattern are quantified.

[0112] Experimental results show that the simulation system in this embodiment can clearly capture the particle initiation, migration, and deposition processes, and output particle size distribution maps (such as...). Figure 10 (As shown).

[0113] Example 5

[0114] This embodiment provides a method for simulating the entire process of soil infiltration and erosion, which uses the aforementioned microfluidic chip or simulation system including a microfluidic chip to simulate the process of soil infiltration and erosion.

[0115] Before the simulation, soil was filled into trapezoidal groove 5 and shaped.

[0116] During the simulation, liquid is injected into the liquid injection tube 6 of the microfluidic chip and flows out through the liquid flow channel 4 and then out through the liquid outflow tube 7. As the liquid flows through the liquid flow channel 4, it penetrates and erodes the soil in the trapezoidal groove 5 on one side. The process of erosion, migration, collision and aggregation of soil particles by the liquid is observed and recorded in real time by the high-speed microscopic imaging module.

[0117] The specific steps are as follows:

[0118] Step S1: Adhere the upper base layer 1 and the lower base layer 2, install the liquid injection pipe 6 and the liquid outflow pipe 7 and connect them to the liquid flow channel 4 to obtain the microfluidic chip;

[0119] Step S2: The test soil is air-dried and sieved, water is added and mixed well to obtain a soil-water mixture solution; the soil-water mixture solution is filled into the trapezoidal groove 5 of the microfluidic chip and dried and shaped.

[0120] In step S3, liquid is injected into the liquid injection pipe 6 by a precision pressure driven pump. After passing through the liquid flow channel 4, the liquid flows out from the liquid outlet pipe 7. During the process of the liquid flowing through the liquid flow channel 4, it permeates and erodes the soil in the trapezoidal trench 5 on one side.

[0121] Step S4: By controlling the pressure of the precision pressure drive pump, the flow rate of the injected liquid is adjusted to simulate the stripping and migration process of soil under different rainfall intensities; the high-speed microscopic imaging module is used to observe and record the process of erosion, migration, collision and aggregation of soil particles by the liquid in real time.

[0122] Step S5: Automatically identify and quantify the erosion rate and pattern of the images acquired by the high-speed microscopic imaging module.

[0123] In step S5, the image multi-particle motion tracking and recognition method is used for recognition. The specific steps of the image multi-particle motion tracking and recognition method are as follows:

[0124] Step S5-1, Particle identification and localization;

[0125] A multi-threaded parallel processing closed loop is constructed using the Rolling-ball background subtraction algorithm and the particle recognition function of the TrackPy library. The multi-threaded parallel processing closed loop is used to process the images acquired by the imaging module to obtain the coordinates (x, y) and particle size d of each frame.

[0126] When processing images acquired by the imaging module using a multi-threaded parallel processing closed loop, the specific steps are as follows:

[0127] Step S5-1-1 involves using a preprocessing pipeline consisting of color image grayscale conversion, color inversion, adaptive histogram equalization, and Gaussian filtering to perform reflection filling, enhancement, and masking processing on the image acquired by the high-speed microscopic imaging module.

[0128] Step S5-1-2: Based on the preset particle size sequence, the particle locator is called in parallel to fuse multi-scale candidate particle sets;

[0129] Step S5-1-3: Deduplication is achieved through pixel grid clustering. While ensuring full particle size coverage, false targets are eliminated, and the coordinates (x, y) and particle size d of each frame are output.

[0130] The Rolling-ball background subtraction algorithm is expressed as follows:

[0131] ;

[0132] in, This represents the pixel value of the current frame image. Represents the pixel values ​​of the background model. Indicates the threshold for foreground detection;

[0133] The remaining steps in step S5-1 can be performed using existing technologies, and those skilled in the art do not need to make any creative effort.

[0134] Step S5-2, cross-frame particle correlation and tracking;

[0135] Using the built-in tracking algorithm of the TrackPy library, and combining Kalman filtering and nearest neighbor search tree strategies, the identification points of the same particle in different frames are associated to form a continuous trajectory with particle ID and temporal coordinates.

[0136] In step S5-2, a Kalman filter is used for filtering; the state equation of the Kalman filter is:

[0137] ;

[0138] The observation equation is:

[0139] ;

[0140] The prediction equation is:

[0141] ;

[0142] ;

[0143] The gain matrix being updated is:

[0144] ;

[0145] The status during the update is now updated as follows:

[0146] ;

[0147] The covariance update in the update is as follows:

[0148] ;

[0149] in, This represents the state transition matrix, which describes the changes in the state from time k to time k+1. This represents the state vector at time k; This represents the noise input matrix, describing how the noise affects the state. Indicates system noise; The observation matrix represents the relationship between the state and the observations. Indicates observation noise; This represents the optimal state estimate at time k after all observations are known. Let represent the error covariance matrix of the state estimate at time k; Represents the process noise covariance matrix; Indicates transpose; Let represent the error covariance matrix of the state estimation from time k to time k+1; This represents the error or noise level at time k+1; This represents the optimal state estimate after all observations are known from time k to time k+1; This represents the actual observed value obtained at k+1; Represents the identity matrix;

[0150] In step S5-2, the nearest neighbor search tree strategy specifically includes:

[0151] Step S5-2-1: Calculate the sample spacing in the tree nodes. :

[0152] ;

[0153] in, Indicates the total number of samples; , This represents the coordinates of soil particles within a series of frames. , Indicates particles Two coordinate points in consecutive frames;

[0154] Step S5-2-2, divide the sample into left and right subtrees, represented as:

[0155] ;

[0156] in, Represents a dataset, Indicates XX, Indicates XX; This represents the coordinates of the sample in the j-th dimension; This represents the median of the j-th dimension, used to construct a balanced tree structure;

[0157] Step S5-2-3, during the search process, if the samples in the tree node With query point If the distance to the nearest neighbor is less than the current best distance, then update the nearest neighbor:

[0158] ;

[0159] like If so, the recording of "candidate collision pairs" is triggered, and the subsequent relative velocity determination will proceed.

[0160] in, This represents the current optimal distance. Point The nearest neighbor, Indicates the collision detection threshold;

[0161] Step S5-2-4, if the query point With a certain subtree center The distance is greater than the radius of the sphere The sum of the distances to the current best result is used for pruning:

[0162] ;

[0163] in, Indicates query point With a certain subtree center The distance.

[0164] The remaining steps in step S5-2 can be performed using existing technologies, and those skilled in the art do not need to make any creative effort.

[0165] Step S5-3, Dynamic behavior determination;

[0166] Using a collision threshold as the trigger condition, candidate collision pairs are identified based on a nearest neighbor search tree, and the distance between the two particles in subsequent frames is tracked. If the distance between the two particles is consistently below the threshold and their relative velocity decreases monotonically, they are considered to be effectively adhered and counted. Otherwise, they are classified as instantaneous collisions. Ultimately, the erosion dynamics parameters of the entire process of "separation-transportation-reaggregation" of soil particles are quantitatively analyzed, and the erosion rate and mode are quantified.

[0167] After determining whether soil particles collide instantaneously based on the change in distance between the two particles, existing technologies can be used to quantify and analyze erosion dynamic parameters, quantify erosion rate and mode, without requiring creative effort from those skilled in the art.

[0168] By employing an image-based multi-particle motion tracking and recognition method, we have achieved quantitative analysis of erosion dynamic parameters for the entire process of soil particle "separation-transportation-re-aggregation" from the original image sequence. This enables high-throughput, high-precision, real-time quantitative analysis of complex soil erosion simulation processes, improving the efficiency and accuracy of simulating the entire dynamic soil erosion process.

[0169] Test case

[0170] The microfluidic chip described in Example 1 and the simulation method described in Example 5 were used to simulate and analyze the infiltration and erosion process of local soil in the Songhu iron mine in Xinjiang in a NaCl solution system. The specific method is as follows:

[0171] The collected local soil samples from the Songhu Iron Mine in Xinjiang were air-dried and sieved (≤2mm) to remove large particles and organic residues. A water-soil mixture solution was then filled into trapezoidal trough 5 using a precision pressure-driven pump (existing technology) and dried and shaped.

[0172] Connections such as Figure 6 The simulation system shown.

[0173] A 1 mol / L NaCl solution is injected into the liquid channel 4 through the flow hole 3 using a precision pressure-driven pump. The flow rate of the fluid is controlled by adjusting the numerical control of the precision pressure-driven pump to an injection rate of 2 μL / min.

[0174] After the soil was fully saturated, the precision pressure drive pump was repeatedly adjusted to increase the flow rate, making the flow rate of the liquid in liquid channel 4 successively 10 uL / min, 15 uL / min, and 20 uL / min. An atomic force microscope combined with a high-speed microscopic imaging module was used to capture, observe, and record images of the soil particle erosion, migration, collision, and aggregation processes at the observation section, obtaining... Figure 7 The comparison images of particles before and after migration are shown, and Figure 8 The grayscale image of particle migration shown is provided for subsequent analysis and processing.

[0175] This paper utilizes an image-based multi-particle motion tracking and recognition method. It combines background subtraction (Rolling-ball algorithm) and particle tracking (TrackPy library) with a data monitoring and acquisition system to automatically identify captured images and accurately pinpoint valid particles. Figure 9 As shown, the erosion rate and pattern are quantified.

[0176] Experimental results show that the simulation system in this embodiment can clearly capture the particle initiation, migration, and deposition processes, yielding results such as... Figure 10 The diagram shows the erosion particle size distribution of loess particles under different flow rates and NaCl concentrations, and as shown in the figure. Figure 11 The image shows the grayscale diagram of the erosion movement of soil particles in Songhu, Xinjiang, under different flow velocities and NaCl concentrations.

Claims

1. A microfluidic chip based on the simulation of the entire process of soil infiltration and erosion, comprising an upper layer (1), a lower layer (2), a liquid injection pipe (6), and a liquid outflow pipe (7), wherein flow holes (3) are provided at both the inlet and outlet ends of the upper layer (1) and the lower layer (2), characterized in that: The top surface of the lower base layer (2) is provided with a liquid flow channel (4) and a trapezoidal groove (5). The long side of the trapezoidal groove (5) is parallel to and connected to the liquid flow channel (4). The upper base layer (1) is attached to the lower base layer (2). The liquid injection pipe (6) and the liquid outflow pipe (7) are connected to the two ends of the liquid flow channel (4) through the flow holes (3) at both ends of the lower base layer (2). The trapezoidal groove (5) is filled with soil and shaped. An angle α exists between the bottom surface (8) of the liquid flow channel (4) and the lower surface of the lower base layer (2); an angle α exists between the bottom surface (8) of the trapezoidal groove (5) and the lower surface of the lower base layer (2); the bottom surface (8) of the liquid flow channel (4) and the bottom surface (8) of the trapezoidal groove (5) gradually decrease in height along the liquid flow direction, and the angle α is less than 5°; or, the bottom surface (8) of the liquid flow channel (4) is a curved surface, and the radius of curvature of the bottom surface (8) of the liquid flow channel (4) gradually increases along the liquid flow direction; the bottom surface (8) of the trapezoidal groove (5) is a curved surface, and the radius of curvature of the bottom surface (8) of the trapezoidal groove (5) gradually increases along the liquid flow direction.

2. The microfluidic chip based on the simulation of the entire process of soil infiltration and erosion as described in claim 1, characterized in that: A flow guide protrusion (9) is provided on the bottom surface (8) of the liquid flow channel (4) at the position corresponding to the flow channel of the trapezoidal groove (5).

3. A microfluidic chip based on the simulation of the entire process of soil infiltration and erosion as described in claim 2, characterized in that: Along the short side of the trapezoidal groove (5), the height of the guide protrusion (9) above the bottom surface (8) gradually decreases.

4. A method for simulating the entire process of soil infiltration and erosion, characterized in that, The process of soil infiltration and erosion is simulated using the microfluidic chip based on the full process of soil infiltration and erosion as described in any one of claims 1-3; Before the simulation, soil was filled into the trapezoidal groove (5) and shaped. During the simulation, liquid is injected into the liquid injection tube (6) of the microfluidic chip and flows out through the liquid flow channel (4) and then out through the liquid outflow tube (7). During the process of the liquid flowing through the liquid flow channel (4), it penetrates and erodes the soil in the trapezoidal groove (5) on one side. The process of erosion, migration, collision and aggregation of soil particles by the liquid is observed and recorded in real time through the high-speed microscopic imaging module.

5. The method for simulating the entire process of soil infiltration and erosion as described in claim 4, characterized in that, The specific steps for simulating soil infiltration and erosion are as follows: Step S1: The upper base layer (1) and the lower base layer (2) are attached together, and the liquid injection pipe (6) and the liquid outflow pipe (7) are installed and connected to the liquid flow channel (4) to obtain the microfluidic chip; Step S2: The test soil is air-dried and sieved, water is added and mixed to obtain a soil-water mixture solution; the soil-water mixture solution is filled into the trapezoidal groove (5) of the microfluidic chip and dried and shaped. In step S3, liquid is injected into the liquid injection pipe (6) by a precision pressure driven pump. After passing through the liquid flow channel (4), the liquid flows out from the liquid outlet pipe (7). During the process of the liquid flowing through the liquid flow channel (4), it permeates and erodes the soil in the trapezoidal groove (5) on one side. Step S4: By controlling the pressure of the precision pressure drive pump, the flow rate of the injected liquid is adjusted to simulate the stripping and migration process of soil under different rainfall intensities; the high-speed microscopic imaging module is used to observe and record the process of erosion, migration, collision and aggregation of soil particles by the liquid in real time. Step S5: Automatically identify and quantify the erosion rate and pattern of the images acquired by the high-speed microscopic imaging module.

6. The method for simulating the entire process of soil infiltration and erosion as described in claim 5, characterized in that, In step S5, the image multi-particle motion tracking and recognition method is used for recognition. The specific steps of the image multi-particle motion tracking and recognition method are as follows: Step S5-1, Particle identification and localization; A multi-threaded parallel processing closed loop is constructed by using the Rolling-ball background subtraction algorithm and the particle recognition function of the TrackPy library. The multi-threaded parallel processing closed loop is used to process the images acquired by the imaging module to obtain the coordinates (x, y) and particle size d of each frame. Step S5-2, cross-frame particle correlation and tracking; Using the built-in tracking algorithm of the TrackPy library, and combining Kalman filtering and nearest neighbor search tree strategies, the identification points of the same particle in different frames are associated to form a continuous trajectory with particle ID and temporal coordinates. Step S5-3, Dynamic behavior determination; Using a collision threshold as the trigger condition, candidate collision pairs are identified based on a nearest neighbor search tree, and the distance between the two particles in subsequent frames is tracked. If the distance between the two particles is consistently below the threshold and their relative velocity decreases monotonically, they are determined to be effective adhesions and counted cumulatively; otherwise, they are classified as instantaneous collisions. Ultimately, the erosion dynamics parameters of the entire process of "separation-transportation-reaggregation" of soil particles are quantitatively analyzed, and the erosion rate and mode are quantified.

7. The method for simulating the entire process of soil infiltration and erosion as described in claim 6, characterized in that, In step S5-1, when processing the image acquired by the imaging module using a multi-threaded parallel processing closed loop, the specific steps are as follows: Step S5-1-1 involves using a preprocessing pipeline consisting of color image grayscale conversion, color inversion, adaptive histogram equalization, and Gaussian filtering to perform reflection filling, enhancement, and masking processing on the image acquired by the high-speed microscopic imaging module. Step S5-1-2: Based on the preset particle size sequence, the particle locator is called in parallel to fuse multi-scale candidate particle sets; Step S5-1-3: Deduplication is achieved through pixel grid clustering. While ensuring full particle size coverage, false targets are eliminated, and the coordinates (x, y) and particle size d of each frame are output.

8. The method for simulating the entire process of soil infiltration and erosion as described in claim 6, characterized in that, In step S5-1, the Rolling-ball background subtraction algorithm is expressed as follows: ； in, This represents the pixel value of the current frame image. Represents the pixel values ​​of the background model. Indicates the threshold for foreground detection; In step S5-2, a Kalman filter is used for filtering; the state equation of the Kalman filter is: ； The observation equation is: ； The prediction equation is: ； ； The gain matrix being updated is: ； The status during the update is now updated as follows: ； The covariance update in the update is as follows: ； in, This represents the state transition matrix, which describes the changes in the state from time k to time k+1. This represents the state vector at time k; This represents the noise input matrix, describing how the noise affects the state. Indicates system noise; The observation matrix represents the relationship between the state and the observations. Indicates observation noise; This represents the optimal state estimate at time k after all observations are known. Let represent the error covariance matrix of the state estimate at time k; Represents the process noise covariance matrix; Indicates transpose; Let represent the error covariance matrix of the state estimation from time k to time k+1; This represents the error or noise level at time k+1; This represents the optimal state estimate after all observations are known from time k to time k+1; This represents the actual observed value obtained at k+1; Represents the identity matrix; In step S5-2, the nearest neighbor search tree strategy specifically includes: Step S5-2-1: Calculate the sample spacing in the tree nodes. : ； in, Indicates the total number of samples; , This represents the coordinates of soil particles within a series of frames. , Indicates particles Two coordinate points in consecutive frames; Step S5-2-2, divide the sample into left and right subtrees, represented as: ； in, Represents a dataset, This represents the subset of samples contained in the left subtree. This represents the subset of samples contained in the right subtree; This represents the coordinates of the sample in the j-th dimension; This represents the median of the j-th dimension, used to construct a balanced tree structure; Step S5-2-3, during the search process, if the samples in the tree node With query point If the distance to the nearest neighbor is less than the current best distance, then update the nearest neighbor: ； like If so, the recording of "candidate collision pairs" is triggered, and the subsequent relative velocity determination will proceed. in, This represents the current optimal distance. Point The nearest neighbor, Indicates the collision detection threshold; Step S5-2-4, if the query point With a certain subtree center The distance is greater than the radius of the sphere The sum of the distances to the current best result is used for pruning: ； in, Indicates query point With a certain subtree center The distance.

Images