Soft soil confined water layer precipitation water-soil coupling physical model test device and test method

Abstract

The application discloses a soft soil confined water layer dewatering water-soil coupling physical model test device and test method, which comprises four core systems of a high-rigidity visual model box system, a confined water pressure precision simulation servo system based on an intelligent algorithm, a dewatering well group and layered boundary water head control system and a same-site integrated water-soil coupling in-situ test system. The application realizes high-precision dynamic continuous adjustment of the confined water head through GA-Kalman PID intelligent closed-loop control, and breaks through the static limitation of traditional gravity type water head control. The absolute synchronous measurement of total stress and pore water pressure at the same space point is realized through a MEMS same-site integrated sensor, and the space misplacement error of a traditional separated sensor is eliminated. Meanwhile, a "point-line-surface-body" multi-dimensional panoramic monitoring system is constructed, which can accurately reproduce the whole process of water-soil coupling in the deep foundation pit dewatering in a coastal soft soil area, and provides a high-precision indoor physical simulation platform for the disaster prevention and control of deep foundation pit dewatering.

Description

Technical Field

[0001] This invention relates to the fields of environmental geological disaster prevention and control technology, geotechnical engineering, and underground engineering testing technology. Specifically, it relates to a physical model test device and test method for water-soil coupling in dewatering of soft soil confined aquifers, which is particularly suitable for indoor physical simulation of water-soil coupling in deep foundation pit dewatering projects in thick soft soil strata in coastal areas. Background Technology

[0002] As global urbanization deepens, underground engineering faces increasingly severe challenges from deep and complex environments. Coastal areas are characterized by thick layers of soft soil, with a typical "soft on top, hard below" geological structure and alternating layers of aquifers, forming a complex hydrogeological system. In such environments, when constructing deep foundation pits, if the weight of the weakly permeable layer at the bottom of the pit is insufficient to resist the pressure of the underlying confined water head, engineering accidents such as sudden heaving and piping are highly likely to occur.

[0003] To ensure the stability of the foundation pit, dewatering and pressure reduction measures are typically used in engineering to lower the confined water head. However, dewatering causes the pore water pressure in the soil to dissipate, increasing the effective stress in the soil and leading to compaction, consolidation, and three-dimensional deformation. This is a typical water-soil coupling process: the dynamic changes in groundwater seepage alter the pore pressure field, causing stress redistribution and deformation in the soil; while soil compression changes porosity and permeability, which in turn affects the seepage path and hydraulic gradient. This interaction often leads to delayed uneven settlement and ground fissures around the foundation pit, seriously threatening the safety of adjacent buildings and underground pipelines.

[0004] To deeply reveal the deformation mechanism of confined aquifers caused by precipitation, physical model testing, as an intuitive and highly integrated research method, can eliminate the interference of complex environments under laboratory conditions by using scaled-down physical models, and reproduce the water and soil response laws under specific hydraulic boundary conditions, thus possessing irreplaceable scientific value. However, existing precipitation physical model testing devices have significant technical limitations: Firstly, in terms of simulating confined aquifer pressure and hydraulic boundary, existing experiments mostly use traditional Marriott bottles, simple overflow pipe tanks, or manually adjusted valves to control the head height of the model boundary. This can only provide a relatively constant static boundary water pressure, and cannot achieve continuous, dynamic, and precise feedback adjustment of water pressure during precipitation. It is difficult to accurately simulate the real complex working conditions of dynamic rise and fall of the head of confined aquifers during precipitation, and it is even more impossible to reproduce nonlinear hydrological processes such as transient strong discharge precipitation and tidal fluctuation seepage.

[0005] Secondly, in terms of in-situ monitoring and mechanism measurement of soil-water coupling, existing experiments typically place pore water pressure sensors and micro earth pressure cells separately in adjacent but different spatial locations within the soil. The stress and seepage states at these two locations differ significantly, making it impossible to obtain absolutely synchronous evolution data of total stress and pore water pressure at the same spatial coordinate point. This results in a large spatial misalignment error in the calculation of effective stress. Furthermore, traditional large-volume sensors severely disturb the natural structure of the undisturbed soil, causing uncertain changes in the contact state between the sensor and the soil interface, making it difficult to accurately reflect the dynamic soil-water coupling response during the formation of the precipitation funnel.

[0006] Meanwhile, the existing experimental equipment has a single monitoring system, which makes it difficult to achieve panoramic and full life cycle monitoring of soil deformation from point to volume. It cannot fully reveal the transmission mechanism of stratum deformation caused by precipitation, which restricts the development of the water-soil coupling theory of precipitation in soft soil deep foundation pits and the progress of engineering disaster prevention and control technology. Summary of the Invention

[0007] The purpose of this invention is to overcome the aforementioned defects in the existing technology and provide a physical model test device and method for water-soil coupling during dewatering in soft soil confined aquifers. On the one hand, it solves the technical pain points of low accuracy in confined water pressure simulation, slow dynamic response, and inability to reproduce non-steady flow boundaries in existing physical model tests. On the other hand, it overcomes the technical bottleneck of the difficulty in accurately obtaining water-soil coupling parameters at the same location using separate sensors. At the same time, it constructs a multi-dimensional panoramic monitoring system to achieve high-precision and high-fidelity indoor physical simulation of the entire process of water-soil coupling during dewatering in soft soil confined aquifers. This provides a solid and accurate experimental foundation and scientific basis for ground settlement control caused by deep foundation pit dewatering, underground structure deformation analysis, and groundwater resource protection.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In the first aspect, the present invention provides a physical model test device for water-soil coupling in soft soil confined aquifers, which integrates four major functional systems: a high-rigidity visualization model box system, a servo system for accurate simulation of confined water pressure based on intelligent algorithms, a group of dewatering wells and a layered boundary water head control system, and an in-situ integrated water-soil coupling test system.

[0009] The high-rigidity visualization model box system includes a high-rigidity welded frame, a visualization panel, a double-layer drag-reducing structure, and a bottom pressurized water-bearing overhead layer. The high-rigidity welded frame is constructed from H-beams and channel steel, ensuring that the box deformation under full-load water and soil pressure is controlled within millimeters. The visualization panel uses high-transmittance explosion-proof tempered glass fixed to the front and rear sides of the frame, facilitating non-contact, high-resolution full-section measurement of the soil's internal displacement field using PIV / DIC technology. The double-layer drag-reducing structure is laid on... The left, right, and rear walls inside the model box are specifically designed as follows: industrial-grade high-molecular-weight silicone grease is evenly applied to the surface of the steel plate or glass, and a flexible polytetrafluoroethylene film is laid in two layers to significantly reduce the interference of frictional effects at the model boundaries on soil deformation; the bottom pressurized water support layer is a height-adjustable structure, filled with coarse gravel of a certain gradation, and interlaced with porous permeable blind pipes, and covered with a high-strength stainless steel porous support plate and a 300-mesh micron-level nylon geotextile filter screen to achieve uniform water distribution and stable replenishment of pressurized water.

[0010] The pressurized water pressure precision simulation servo system based on intelligent algorithms includes a servo constant pressure water circuit hardware module and a GA-Kalman PID dynamic intelligent control core. The servo constant pressure water circuit hardware module comprises a large-capacity stainless steel water storage tank, a high-precision variable frequency submersible pump, an acrylic glass buffer pressure-stabilizing cylinder, a miniature high-frequency electronic proportional regulating valve, and a high-frequency electromagnetic flowmeter, connected sequentially through pipelines to form a closed-loop control water circuit. The outlet of the acrylic glass buffer pressure-stabilizing cylinder is connected to the pressurized water support layer at the bottom of the model tank through pipelines. Its interior is equipped with a sealed air chamber, which can convert the pumped dynamic water pressure into a stable static water pressure, eliminating water flow pulsation interference. The GA-KalmanPID dynamic intelligent control core includes an industrial control computer, a PLC motherboard, and a miniature water pressure sensor. The miniature water pressure sensor is deployed at the boundary between the confined aquifer and the aquifer, acquiring water head pressure data in real time at a kilohertz frequency. The PLC motherboard is embedded with a Kalman filter PID control module optimized by a genetic algorithm (GA). The Kalman filter algorithm effectively filters out high-frequency noise interference caused by pump mechanical pulsation, valve action, and soil seepage instability, accurately predicting water pressure state parameters. The genetic algorithm optimizes the PID control parameters (proportional gain, integral time, and derivative time) online in real time, solving the integral saturation and system oscillation problems of traditional PID control in nonlinear seepage scenarios. Ultimately, it achieves high-precision, stepless, continuous, and steady-state control of the rise and fall of the confined aquifer head, with a water pressure control accuracy of ±0.5 mm head. It can also accurately reproduce the evolution of complex transient drop funnels caused by pumping from multiple wells, as well as the dynamic interference conditions of groundwater under tidal influences such as sinusoidal fluctuations.

[0011] The dewatering well group and stratified boundary head control system includes a miniature anti-sand flow group well pumping device and an independent stratified water level control tank. The miniature anti-sand flow group well pumping device includes a multi-channel high-precision digital peristaltic pump matrix and a stainless steel microporous filter pipe. The stainless steel microporous filter pipe is designed with strict geometric similarity ratio for open area and blind zone, supporting free switching between full-penetration and non-penetration modes. It is tightly wrapped with multiple layers of micron-level nylon filter screens, which can effectively prevent sand flow and piping during pumping. The multi-channel high-precision digital peristaltic pump matrix is ​​connected to the stainless steel microporous filter pipe, which can support precise closed-loop control of extremely small flow rates, simulating the working conditions of dewatering wells or recharge well groups inside and outside the foundation pit. The independent stratified water level control tank is made of methyl methacrylate board. The tank is physically separated into three or more independent sections with adjustable height, corresponding to the unconfined aquifer, weakly permeable layer and confined aquifer inside the model. Each section is connected to the inlet and outlet reserved on the side plate of the model box by multiple flexible silicone tubes, so as to realize independent control and replenishment of hydraulic boundary conditions (constant head, constant flow, etc.) of different aquifers, and completely cut off unexpected interlayer hydraulic interference.

[0012] The in-situ water and soil coupling test system includes an in-situ water and soil pressure sensor and a macro- and micro-scale full-field deformation monitoring matrix. The integrated soil and water pressure sensor is a single-crystal silicon piezoresistive micro-sensor based on MEMS (Micro-Electro-Mechanical Systems). The probe is made of titanium alloy with an outer diameter of only 12mm and a thickness of only 5mm, which can significantly reduce the disturbance to the soil. The sensing end face of the sensor is divided into two completely isolated sensing micro-regions by high-precision machining. One side is the total stress detection surface, covered with a flat titanium alloy stress-bearing diaphragm, which directly contacts the soil skeleton to sense the total normal stress. The other side is the pore water pressure detection surface, with a recessed front end and embedded with high-density bronze sintered permeable stone, allowing only free pore water to enter the internal independent silicon diaphragm cavity to sense pure pore water pressure. Polycrystalline silicon piezoresistors are etched on the back of the diaphragms of both sensing micro-regions and connected to independent Wheatstone bridges. The sensor also integrates a PT100 micro temperature compensation chip, which can eliminate the temperature drift caused by temperature fluctuations, and finally achieve absolute synchronous measurement of total stress and pore water pressure at the same spatial coordinate point on a macroscopic scale. The macro-micro full-field deformation monitoring matrix integrates a contact settlement monitoring component and a non-contact full-field displacement monitoring component. The contact settlement monitoring component includes a lightweight settlement magnetic ring pre-embedded in the soil layer and a high-precision linear variable differential transformer (LVDT) displacement meter. The LVDT measurement accuracy is better than 0.01 mm and is used to measure the absolute one-dimensional vertical settlement evolution of the surface and internal layers. The non-contact full-field displacement monitoring component includes multiple 4K high-definition industrial CCD cameras deployed on the outside of the model box's visualization panel to capture the movement trajectory of tracers on the inside of the model box's transparent panel. Combined with PIV technology, continuous tracking and calculation of the full-field displacement vector are performed. Finally, through sensor point data, LVDT line data, and PIV surface data, a multi-dimensional water and soil coupling observation matrix of "point-line-surface-volume" is constructed.

[0013] Secondly, the present invention also provides a test method for a physical model of water-soil coupling during precipitation in soft soil confined aquifers, implemented based on the above-mentioned test apparatus, specifically including the following steps: S1. Derivation of Similarity Criteria and Preparation of Model Materials: First, based on Buckingham's π theorem and Biot's fluid-structure interaction control equation, the geometric similarity ratio of the experiment was determined. To ensure the rationality of the seepage and consolidation time within the laboratory timescale, the unit weight similarity ratio was maintained at 1:1. This led to the derivation of key parameters such as the elastic modulus similarity ratio, permeability coefficient similarity ratio, and time similarity ratio. Second, geological materials were prepared on a scaled-down basis for the silty clay and confined silt layers in the target simulation area. The overlying weakly permeable soft soil layer was prepared by mixing kaolin, bentonite, talc powder, and undisturbed soft soil in a specific ratio, adjusting its liquid limit, plastic limit, and permeability coefficient to match the target formation parameters. The underlying confined aquifer was prepared by mixing standard quartz sand according to graded mix to meet the high permeability design requirements.

[0014] S2. Deployment of In-situ Sensors and Precise Model Filling: The lower sand layer is filled in the model box using underwater sand dropping or layered compaction. In the designed precipitation-affected area, an integrated soil and water pressure sensor is precisely implanted using a micro-support. At the interface between the sand and clay layers, a rubber film or bentonite waterproof blanket with specific openings is laid for physical water-stopping and sealing, strictly controlling the location and scale of interlayer overflow. Subsequently, the upper soft soil layer is slowly filled using mud hydrostatic consolidation or layered preloading. Finally, an equivalent overburden load or flexible water bag is applied to the top to complete the normal consolidation initialization of the geostress field.

[0015] S3. High-precision confined water pressure servo initialization and calibration: Connect all sensors and perform zero-point calibration, turn on the GA-Kalman PID servo water pressure control system, and steadily raise the water pressure at the boundary of the confined water layer to the predetermined initial head height. At the same time, maintain the groundwater level of the upper soft soil layer saturated through the stratified water level control tank. Let the entire system stand for 48~72 hours until all LVDT, pore pressure and earth pressure data no longer drift, and complete the steady-state reconstruction of the initial water and soil stress field of the model.

[0016] S4. Dynamic Stage Simulation of Precipitation-Soil Coupling: This process is divided into sub-stages such as stepped precipitation, continuous precipitation, and pumping cessation and recovery. Through programmable control of multi-channel peristaltic pumps, water is pumped in stages according to the set flow curve or target drawdown curve to simulate the equivalent drawdown of different excavation stages of the foundation pit. During the forced drainage period, the GA-KalmanPID servo system accurately controls the replenishment water volume and water pressure of the boundary according to the set external boundary conditions. A sinusoidal fluctuation curve can be input to simulate the dynamic disturbance of groundwater caused by tides in nearshore foundation pits.

[0017] S5. Multidimensional Data Acquisition and Analysis of Soil-Water Coupling Mechanism: The high-speed data acquisition instrument synchronously and continuously records pore water pressure, total earth pressure, stratified settlement, well flow rate, and head change data at a set frequency. Using the same-point data from the integrated sensor, the effective stress evolution three-dimensional trajectory of any measuring point within the fallout funnel is calculated in real time. Combined with the full-field deformation cloud map extracted by PIV, the nonlinear synergistic relationship between the water release hysteresis consolidation deformation of the weakly permeable layer and the instantaneous compression deformation of the confined aquifer is quantitatively analyzed, revealing the physical transmission mechanism of the "soil arching effect" and differential settlement caused by water pressure imbalance.

[0018] Compared with the prior art, the present invention has the following advantages: 1. This invention innovatively adopts a PID closed-loop electro-hydraulic servo control system based on genetic algorithm and Kalman filter intelligent optimization, which completely overcomes the shortcomings of traditional gravity water tanks such as Marriott bottles in simulating nonlinear dynamic water level changes and severe response lag. The system significantly reduces external disturbances caused by fluid pulsation and non-Darcy flow in porous media. It can not only accurately simulate the staged steady-state precipitation of confined aquifers, but also perfectly reproduce complex transient engineering conditions such as transient forced drainage, tidal fluctuation seepage, and confined water recharge. The water pressure control accuracy can reach ±0.5mm head, effectively improving the control accuracy and dynamic response of confined water.

[0019] 2. This invention introduces a dual-sensor integrated single-crystal silicon piezoresistive micro-sensor based on MEMS technology into the precipitation physics model, fundamentally solving the spatial geometric misalignment error and severe soil structure disturbance caused by the separate arrangement of pore pressure gauges and earth pressure gauges in traditional tests. This enables the dissipation curve of pore water pressure and the transfer curve of total stress in the soil to achieve perfect data closure at the same microscopic geometric point, and the calculation accuracy of effective stress increment achieves an exponential leap. This provides the most direct and accurate measured calibration data for verifying advanced soil mechanics constitutive models, and truly realizes the non-destructive measurement of water-soil coupling mechanical parameters at absolutely "same location".

[0020] 3. This invention deeply integrates high-precision water and soil pressure sensing at the microscopic isotope, high-precision LVDT stratified settlement monitoring, and high-resolution PIV / DIC displacement vector tracking technology with transparent observation windows. This cross-scale measurement matrix can reproduce the entire life cycle physical process of water level drawdown funnel evolution caused by precipitation, fine particle transport within the soil, shear zone formation and development, and uneven settlement of the surface / strata in a panoramic and comprehensive manner, thereby constructing a multi-field panoramic visualization fusion test system of "point-line-surface-volume".

[0021] 4. The low-friction double-layer drag reduction system, the fine interlayer hydraulic isolation sealing structure, the layered multi-zone water supply control network, and the well group simulation module with microporous anti-flowing sand design designed in this invention perfectly replicate the complex binary geological structure characteristics of some coastal areas, which are "upper high water sensitivity, high compressibility, weak permeability soft soil + lower high pressure, strong permeability, easy-flowing sand silt". This provides a high-precision indoor physical simulation platform for solving the prevention and control of major engineering disasters caused by dewatering of deep foundation pits. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the overall structure of the experimental device of the present invention.

[0023] Figure 2 This is a control logic block diagram of the pressurized water pressure precision simulation servo system in this invention.

[0024] Figure 3This is a schematic cross-sectional view of the integrated water and soil pressure sensor in this invention.

[0025] Figure 4 This is a schematic diagram of the cross-sectional structure of the independent layered water level control tank in this invention. Detailed Implementation

[0026] The present invention will be further described below.

[0027] This embodiment provides a physical model test device for soil-water coupling during dewatering in soft soil confined aquifers, the overall structure of which is as follows: Figure 1 As shown, it includes four core functional systems, and the specific implementation structure is as follows: 1. High-rigidity visual model box system: The main frame of the model box is welded from 10mm thick Q235BH steel and channel steel. The overall internal dimensions are 2000mm long × 1000mm wide × 1500mm high. After finite element verification, the maximum deformation of the box under full load water and soil pressure is less than 0.8mm, which meets the test stiffness requirements.

[0028] The front and rear side panels of the model box are made of 25mm thick high-transmittance explosion-proof tempered glass with a transmittance of ≥92%, which can clearly observe the internal deformation of the soil. At the same time, it can be combined with PIV / DIC technology to realize non-contact full-section measurement of the two-dimensional displacement field inside the soil.

[0029] The left, right and rear walls inside the model box are treated with double drag reduction: first, a 1mm thick industrial-grade high-molecular silicone grease is evenly applied to the surface of the steel plate, and then a 0.2mm thick flexible polytetrafluoroethylene film is laid in two layers. The boundary friction coefficient can be reduced to below 0.1, which greatly reduces the interference of boundary effects on soil consolidation deformation.

[0030] The bottom of the model box is equipped with a 200mm high pressurized water-bearing air-supported layer, which is filled with 5-10mm graded coarse gravel and staggered with DN20 perforated water-permeable blind pipes with an opening rate of ≥15%. The top of the air-supported layer is covered with a 5mm thick stainless steel perforated support plate (opening rate ≥30%). 300-mesh and 500-mesh double-layer nylon geotextile filters are laid on top of the support plate to achieve uniform water distribution of the pressurized water and prevent soil particles from entering the air-supported layer and clogging the pipes.

[0031] 2. A precise servo system for simulating pressurized water pressure based on intelligent algorithms: The structure and control logic of this system are as follows... Figure 2 As shown, it consists of two parts: a servo constant pressure water circuit hardware module and a GA-Kalman PID dynamic intelligent control core.

[0032] The servo constant pressure water circuit hardware module consists of a 500L large-capacity 304 stainless steel water tank, equipped with a high-precision variable frequency submersible pump with a head of 20m and a flow rate of 5m³ / h. The outlet is sequentially connected to a DN25 acrylic buffer and pressure-stabilizing cylinder, a miniature high-frequency electronic proportional regulating valve with a response speed ≤5ms, and a high-frequency electromagnetic flowmeter with an accuracy of 0.5%. Finally, the water is piped to the pressurized water support layer at the bottom of the model tank, forming a closed-loop control water circuit. The acrylic buffer and pressure-stabilizing cylinder has an inner diameter of 200mm and a height of 500mm, with a 10% reserved sealed air chamber inside to utilize the buffering effect of the hydraulic accumulator, completely smoothing out the minute pulsations of dynamic water pressure and converting them into a stable static water pressure output.

[0033] GA-Kalman PID Dynamic Intelligent Control Core: Equipped with an industrial control computer and a PLC motherboard with a DSP chip, it deploys three high-precision diffused silicon miniature water pressure sensors in the confined water layer, with a sampling frequency of 1000Hz and a measurement accuracy of ±0.2%FS, to collect the actual water head data of the confined water layer in real time.

[0034] Its core control feedback logic is as follows: The set objective function for the confined head is defined as follows: (The curve can be any complex curve that varies with time). A high-precision diffused silicon miniature water pressure sensor, arranged in the pressure-bearing overhead layer, collects the actual pore pressure at a frequency of 1000Hz and converts it into the actual physical water head. The system calculates the current water pressure error at any time. .

[0035] However, due to the significant damping and time lag (i.e., flow resistance effect) in the seepage of fluids in porous soil, and the unavoidable mechanical pulsation characteristics of water pumps during pressurization, directly using conventional industrial PID (proportional-integral-derivative) controllers can easily lead to severe integral saturation, pressure overshoot oscillations, or even system instability and collapse. Therefore, the digital control motherboard of this invention creatively embeds an advanced PID intelligent control algorithm combining genetic algorithms and discrete Kalman filtering. Specifically, since the raw sensor signals contain a large amount of high-frequency white noise (caused by pump impeller rotation and microscopic turbulence), the Kalman filter equation is first used in the time update and measurement update cycle to extract the true and smooth water pressure state parameters: State prediction equation: Error covariance prediction: State update equation: Kalman gain: in, This represents the actual stress state. Represents state prediction, This represents the prediction of the error covariance.

[0036] High-purity water head error signal after Kalman filtering This data is then input into the PID control loop. At this point, because soil permeability changes non-linearly with compression, the traditional fixed PID gain parameter is no longer applicable. The system uses a background genetic algorithm (GA) to minimize the integral of the pressure error as the fitness function, and continuously optimizes online in real-time to find the optimal PID control parameter (proportional gain). Integral Time Differential time The optimized PID controller outputs a high-frequency PWM (Pulse Width Modulation) signal to precisely drive the valve core opening of the high-frequency electronic proportional control valve, thus finely regulating the flow rate into the plexiglass buffer pressure-stabilizing cylinder. The plexiglass cylinder utilizes its extremely small, sealed air cavity to act as a flexible damping buffer, similar to a hydraulic accumulator, to completely smooth out the minute pulsations of dynamic water pressure, ultimately transforming them into extremely stable hydrostatic pressure applied to the bottom pressure-bearing elevated layer of the model box.

[0037] This system can achieve two core control modes: one is steady-state servo control, which locks the control error of the pressurized head within ±0.5mm when pumping at a constant drawdown depth. Even if there is water loss at the seepage boundary, the system can complete automatic water replenishment balance within 100ms; the other is transient unsteady flow simulation, which can directly input any time-head curve such as sine wave or engineering measured drawdown sequence to accurately simulate complex working conditions such as tidal fluctuations and transient forced drainage.

[0038] 3. Dewatering Well Group and Layered Boundary Head Control System: Micro Sand-prevention Group Pumping Device: Equipped with an 8-channel high-precision digital peristaltic pump matrix, with a flow control range of 0.001~100mL / min and a control accuracy of ±0.5%, it can independently control the pumping / recharge flow of a single well; the pumping well pipe adopts 316L stainless steel microporous pipe with an outer diameter of 16mm, designed with an opening rate of 12% according to a geometric similarity ratio of 1:50, and set with an adjustable blind zone, supporting full penetration / non-penetration mode switching. The filter pipe is wrapped with a 300-mesh + 500-mesh double-layer nylon filter screen to effectively prevent sand flow and piping during the pumping process.

[0039] Independent stratified water level control tank: such as Figure 4As shown, the tank is made of 20mm thick methyl methacrylate board, with overall dimensions of 1500mm long × 300mm wide × 1200mm high. The interior of the tank is divided into three independent, height-adjustable sections by physical partitions, corresponding to the shallow groundwater layer, the weakly permeable layer, and the confined aquifer inside the model, respectively. Each section is equipped with an overflow port and a scale. Each section is connected to the reserved inlet and outlet ports at the corresponding elevations on the side plate of the model box by φ8 flexible silicone tubes. The boundary head of each aquifer can be set independently to achieve accurate simulation of constant head / constant flow boundary, while completely severing the unexpected hydraulic connection between layers.

[0040] 4. In-situ Coupling Testing System for Soil and Water: This system includes an integrated soil and water pressure sensor and a macro- and micro-scale full-field deformation monitoring matrix. The structure of the integrated soil and water pressure sensor is as follows: Figure 3 As shown.

[0041] The integrated soil and water pressure sensor is manufactured using MEMS (Micro-Electro-Mechanical Systems) photolithography technology. The probe body is made of titanium alloy, with an outer diameter of 12mm and a thickness of 5mm, significantly smaller than traditional sensors, greatly reducing soil disturbance. The sensor's circular sensing end face is machined into two completely isolated semi-circular micro-regions: one side is the total stress detection surface, covered with a 0.1mm thick flat titanium alloy stress-bearing diaphragm, directly contacting the soil skeleton to sense the total normal stress σ at the measurement point; the other side is the pore water pressure detection surface, with a recessed front end inlaid with high-density bronze sintered permeable stone. The permeable stone has a pore diameter of 5μm, completely blocking the soil skeleton and allowing only free pore water to enter the internal independent silicon diaphragm cavity to sense the pore water pressure u at the measurement point. Four sets of polycrystalline silicon varistors are ion-implanted on the back of the diaphragms in both micro-regions and connected to form independent Wheatstone bridges. The sensor also integrates a PT100 micro temperature compensation chip to eliminate temperature drift effects within a ±10℃ range. Two independent electrical signals are transmitted synchronously to a 24-bit high-speed data acquisition instrument through a multi-core Teflon shielded cable without time difference, enabling absolute synchronous measurement of total stress and pore water pressure at the same spatial point, and the effective stress σ'=σ-u can be calculated in real time at the millisecond level.

[0042] Macro- and micro-scale full-field deformation monitoring matrix: The contact settlement monitoring component includes 12 sets of high-precision LVDT displacement gauges with a measurement range of 0~200mm and an accuracy better than 0.01mm. Combined with lightweight settlement magnetic rings pre-embedded at various elevations in the soil layer, it achieves real-time continuous monitoring of surface and stratified settlement. The non-contact full-field displacement monitoring component includes two 4K high-speed industrial CCD cameras with a frame rate of 60fps, deployed on the outside of the tempered glass of the model box. Combined with fluorescent tracer particles pre-embedded inside the soil, it achieves continuous tracking of the soil displacement vector across the entire field using PIV technology, with a spatial resolution of up to 0.1mm / pixel. Finally, through point-like soil and water pressure sensors, line-like LVDT settlement monitoring, and area-like PIV full-field displacement monitoring, a multi-dimensional soil and water coupled observation matrix of "point-line-area-volume" is constructed.

[0043] Based on the test apparatus described in Example 1, a water-soil coupling physical model test method is provided for dewatering projects in deep foundation pits in typical soft soil strata in Shanghai. The specific steps are as follows: S1. Derivation of Similarity Criterion and Preparation of Model Materials: Based on Buckingham's π theorem and Biot's fluid-structure interaction control equations, the geometric similarity ratio is determined. Effective stress similarity ratio: Pore ​​water pressure similarity ratio: Elastic modulus similarity ratio: Time (consolidation) similarity ratio: Permeability coefficient similarity ratio: .

[0044] Model material preparation was carried out for the fourth layer of silty clay and the seventh layer of confined silt in Shanghai: The seventh confined aquifer: It is prepared by mixing standard quartz sand (0.25~0.5mm and 0.51mm) at a mass ratio of 3:7. The measured permeability coefficient k = 5 × 10⁻⁶. -3 cm / s, which meets the similarity ratio requirement with the prototype stratigraphic parameters; The fourth layer, a weakly permeable soft soil layer, consisted of undisturbed Shanghai silty clay, kaolin, bentonite, and talc powder mixed in a mass ratio of 5:3:1:1. Deionized water was added and stirred to form a slurry. After standing for 24 hours, the measured liquid limit (WL) was 45.2%, the plastic limit (WP) was 23.8%, and the permeability coefficient (k) was 1.0 × 10⁻⁶. -7 cm / s, matching the prototype stratigraphic parameters.

[0045] S2. Placement of Co-location Sensors and Precise Model Filling: The lower 200mm thick confined aquifer sand layer was filled inside the model box using the underwater sand-falling method, with the relative density D controlled during the filling process. r=0.65; In the area affected by the designed dewatering well, a co-located integrated soil and water pressure sensor is precisely implanted at measuring points at different elevations and radial distances using a customized micro acrylic bracket. The sensor cable is led out along the side wall of the model box to avoid disturbing the soil.

[0046] A 1mm thick rubber film is laid on top of the sand layer, with openings reserved only in the pre-designed overflow area, and the remaining areas are sealed with bentonite waterproof blankets to strictly control the interlayer overflow parameters.

[0047] Subsequently, the upper 800mm thick soft soil layer was filled using the mud slurry hydrostatic consolidation method, in four layers. Each layer was left to stand for 72 hours after filling to complete self-weight consolidation. Finally, a flexible water bag was laid on top of the soil layer, and an equivalent overburden load of 20kPa was applied. After standing for 14 days, the normal consolidation initialization of the geostress field was completed.

[0048] S3. High-precision confined water pressure servo initialization and calibration: Connect all sensors and data acquisition instruments to complete the zero-point calibration and sensitivity calibration of the entire system; turn on the GA-Kalman PID servo water pressure control system to steadily raise the boundary water pressure of the confined aquifer to 80kPa (equivalent to a prototype 5m water head burial depth) at a rate of 0.5kPa / min, while controlling the water level of the upper unconfined layer to 100mm below the soil surface through the stratified water level control tank to keep the soil saturated.

[0049] The entire system was left to stand still for 72 hours, and the LVDT, pore water pressure and soil pressure data were monitored in real time until the data drift was less than 0.1% of the sensor's full scale for 12 consecutive hours, thus completing the steady-state reconstruction of the initial soil and water stress field of the model.

[0050] S4. Simulation of dynamic stages of precipitation-soil coupling: The experiment is divided into three sub-stages: The stepped precipitation stage: the pumping wells in the control center are pumped in stages through a multi-channel peristaltic pump matrix to achieve equivalent drawdown depths of 150mm, 300mm and 400mm in sequence. Each drawdown depth is maintained for 24 hours, and the next stage is started after the water head stabilizes. Continuous precipitation phase: Maintain the target drawdown depth of 400mm and pump water continuously for 72 hours to simulate the continuous precipitation conditions during the foundation pit excavation. Stop pumping recovery phase: shut down the water pumps, stop pumping, and monitor the lag process of the confined water head recovery and the formation deformation rebound, lasting 72 hours.

[0051] Throughout the precipitation process, the GA-KalmanPID servo system maintains a constant confined water head at the far boundary of the model box to simulate infinite recharge boundary conditions; at the same time, it can input a sinusoidal head curve to simulate the dynamic impact of Huangpu River tides on the groundwater in the nearshore foundation pit.

[0052] S5. Multidimensional data acquisition and analysis of soil-water coupling mechanism: Throughout the experiment, the high-speed data acquisition instrument synchronously and continuously acquired data from all the co-located integrated soil-water pressure sensors, LVDT displacement gauges, flow meters, and water pressure sensors at a frequency of 100Hz, and calculated the three-dimensional trajectory of effective stress evolution at each measuring point in real time; the industrial CCD camera acquired soil tracer particle images at a frequency of 10s / frame, and after the experiment, the PIV software was used to process the soil to obtain the full-field displacement cloud map and shear zone distribution characteristics.

[0053] Through the high-precision data obtained at the same location in this embodiment, researchers were able to reveal the water-soil coupling dynamic mechanism at depths far exceeding those of conventional experiments: At the instant precipitation begins, due to the extremely high Darcy permeability coefficient of the ⑦th silt layer, the confined water pressure within it rapidly dissipates towards the precipitation well (manifested as a steep drop in the pore pressure signal of the integrated sensor); however, due to the extremely low permeability coefficient of the overlying ④th layer of silty soft soil (approximately...), the water pressure dissipates rapidly towards the precipitation well. The water cannot drain quickly, and the dissipation of pore water pressure within the soil exhibits a significant "time lag" lasting from tens of minutes to several hours. The physical mechanism analysis reveals that, at the instant of a sharp drop in bottom water pressure, due to the initial cohesion of the soft soil's structure, the total soil pressure at that point did not immediately undergo a large-scale downward transfer or unloading. As pumping time progresses, the drop funnel continues to expand, causing minute volumetric shrinkage in the bottom sand layer. Under gravity, the upper soft soil loses some hydraulic support and begins to undergo extremely significant downward drag compaction (at this point, the LVDT records rapid displacement, and the PIV speckle plot shows a strong downward strain vector arrow). It is during this stage that the "effective stress versus time evolution curve" is plotted by real-time subtraction of data from the same point. The curve ("curve") exhibits a highly nonlinear dynamic step response. Combined with the full-field shear strain cloud map extracted by PIV, it can be clearly and quantitatively confirmed that the "soil arching effect" mechanism is induced within the soil due to the extreme imbalance of water pressure gradient. That is, the shear stress inside the soft soil transfers part of the weight of the overlying strata to the surrounding rigid undisturbed zone, resulting in a peculiar pattern of "smaller at the top and larger at the bottom (i.e., deeper settlement is greater than shallow settlement)" in the surface settlement directly above the center of the precipitation funnel. If traditional non-isolated point separation sensors are used, the misaligned signals collected will completely mask this nonlinear step characteristic, and may even lead to the derivation of incorrect consolidation equation parameters.

[0054] In summary, based on the collected multi-dimensional data, this study quantitatively analyzes the nonlinear synergistic relationship between the delayed consolidation deformation of the soft soil weak permeability layer and the instantaneous compression deformation of the confined aquifer, reveals the formation mechanism of the "soil arching effect" during the evolution of the precipitation funnel, and the transmission law of differential settlement from deep to the surface, providing experimental basis and theoretical support for the prevention and control of ground settlement in deep foundation pits in soft soil areas of Shanghai.

[0055] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A physical model test device for water-soil coupling during dewatering in soft soil confined aquifers, characterized in that, The device includes a model box body, a water pressure control unit, a dewatering unit, and a testing unit. It is characterized in that the device integrates a high-rigidity visual model box system, a pressurized water pressure precision simulation servo system, a dewatering well group and a layered boundary water head control system, and a co-located integrated water and soil coupling in-situ testing system. The high-rigidity visual model box system includes a high-rigidity welded frame, a visual panel, a double-layer drag-reducing structure, and a bottom pressurized water-bearing overhead layer. The high-rigidity welded frame is formed by welding H-beams and channel steel. The visual panel is made of high-transmittance explosion-proof tempered glass and fixed to the front and rear sides of the frame. The double-layer drag-reducing structure is laid on the inner side wall of the model box. The bottom pressurized water-bearing overhead layer is filled with graded coarse gravel and interlaced with porous permeable blind pipes. The top of the overhead layer is covered with a stainless steel porous support plate and a micron-level nylon geotextile filter. The pressurized water pressure precision simulation servo system includes a servo constant pressure water circuit hardware module and a GA-Kalman PID dynamic intelligent control core. The servo constant pressure water circuit hardware module is connected to the bottom pressurized water overhead layer through a pipeline to form a closed-loop control water circuit. The GA-Kalman PID dynamic intelligent control core is electrically connected to the servo constant pressure water circuit hardware module to achieve high-precision dynamic closed-loop control of the pressurized water head. The dewatering well group and stratified boundary head control system includes a miniature anti-sand flow group well pumping device and an independent stratified water level control tank. The miniature anti-sand flow group well pumping device is installed inside the model box to simulate the dewatering / recharge conditions of the foundation pit group wells. The independent stratified water level control tank is set outside the model box and is connected to the stratified inlet and outlet reserved on the side plate of the model box through flexible pipelines to realize independent control of the hydraulic boundaries of different aquifers. The in-situ water and soil coupling testing system includes an in-situ integrated water and soil pressure sensor and a macro-micro full-field deformation monitoring matrix. The in-situ integrated water and soil pressure sensor is embedded in the soil of the model and is used for the synchronous acquisition of total stress and pore water pressure at the same spatial point. The macro-micro full-field deformation monitoring matrix integrates a contact settlement monitoring component and a non-contact full-field displacement monitoring component for multi-dimensional monitoring of soil deformation.

2. The experimental apparatus according to claim 1, characterized in that, The visualization panel of the high-rigidity visualization model box system is used in conjunction with Particle Image Velocimetry (PIV) and Digital Image Correlation (DIC) technologies to achieve non-contact full-section measurement of the displacement field inside the soil. The double-layer drag reduction structure includes an industrial-grade polymer silicone grease lubricating layer and a double-layer flexible polytetrafluoroethylene (PTFE) film layer. The bottom pressurized water-bearing overhead layer is a height-adjustable structure, and its top is covered with a nylon geotextile filter.

3. The experimental apparatus according to claim 1, characterized in that, The servo constant pressure water circuit hardware module includes a large-capacity stainless steel water storage tank, a high-precision variable frequency submersible pump, an acrylic glass buffer pressure stabilizing cylinder, a miniature high-frequency electronic proportional regulating valve, and a high-frequency electromagnetic flow meter, which are connected in sequence through pipelines. The acrylic glass buffer pressure stabilizing cylinder has a sealed air chamber inside, which is used to convert dynamic water pressure into stable static water pressure. Its outlet is connected to the bottom pressurized water overhead layer through pipelines.

4. The experimental apparatus according to claim 1, characterized in that, The GA-KalmanPID dynamic intelligent control core includes an industrial control computer, a PLC motherboard, and a miniature water pressure sensor. The miniature water pressure sensor is deployed at the boundary between the confined water layer and the aquifer to collect water head pressure data in real time. The PLC motherboard has an embedded Kalman filter PID control module optimized by a genetic algorithm (GA). The control module filters out high-frequency noise from the collected signal using the Kalman filter algorithm, optimizes the PID control parameters online in real time using the genetic algorithm, and outputs control signals to the miniature high-frequency electronic proportional regulating valve and the variable frequency submersible pump to achieve stepless continuous steady-state control of the confined water head and transient unsteady flow simulation.

5. The experimental apparatus according to claim 1, characterized in that, The micro-dust-prevention well pumping device includes a multi-channel high-precision digital peristaltic pump matrix and a stainless steel microporous filter pipe. The stainless steel microporous filter pipe is designed with an opening ratio and blind zone according to a geometric similarity ratio, supporting free switching between full-penetration and non-penetration modes. It is surrounded by multiple layers of micron-level nylon filter screen. The multi-channel high-precision digital peristaltic pump matrix is ​​connected to the stainless steel microporous filter pipe for precise closed-loop control of pumping / recharge flow.

6. The experimental apparatus according to claim 1, characterized in that, The independent stratified water level control tank is made of methyl methacrylate board. The tank is physically separated into at least three independent sections with adjustable height. Each independent section corresponds to the unconfined aquifer, the weakly permeable aquifer, and the confined aquifer inside the model, respectively, and is used to achieve independent control and replenishment of the hydraulic boundary conditions of different aquifers.

7. The experimental apparatus according to claim 1, characterized in that, The integrated soil and water pressure sensor is a single-crystal silicon piezoresistive micro-sensor based on MEMS (Micro-Electro-Mechanical Systems). The sensor's sensing end face is physically divided into two isolated sensing micro-regions, one of which is the total stress detection surface and the other is the pore water pressure detection surface. The total stress detection surface is equipped with a flat titanium alloy stress-bearing diaphragm for direct contact with the soil skeleton to sense the total soil pressure. The pore water pressure detection surface is inlaid with high-density bronze sintered permeable stone, and the permeable stone is internally connected to an independent silicon diaphragm cavity for sensing pore water pressure. Polycrystalline silicon piezoresistors are etched on the back of the diaphragms in both sensing micro-regions and are respectively connected to independent Wheatstone bridges. The sensor also integrates a PT100 temperature compensation chip to eliminate temperature drift interference.

8. The test apparatus according to claim 1, characterized in that, The contact settlement monitoring component of the macro-micro full-field deformation monitoring matrix includes a lightweight settlement magnetic ring and a high-precision linear variable differential transformer (LVDT) embedded in the soil layer. The LVDT is used to measure the vertical settlement of the surface and the soil layer. The non-contact full-field displacement monitoring component includes a 4K high-definition industrial CCD camera deployed on the outside of the visualization panel of the model box. The CCD camera is used to capture the movement trajectory of tracers inside the soil and, together with PIV technology, realizes the continuous tracking and calculation of the soil displacement vector in the whole field.

9. A physical model test method for dewatering and soil-water coupling in soft soil confined aquifers, characterized in that, The method, implemented using the physical model test device for dewatering and soil-water coupling in soft soil confined aquifers as described in any one of claims 1 to 8, includes the following steps: S1. Derivation of similarity criteria and preparation of model materials: Based on Buckingham's π theorem and Biot's fluid-structure interaction control equation, the geometric similarity ratio and unit weight similarity ratio of the experiment are determined, and the key parameters of similarity ratio of thermal modulus, similarity ratio of permeability coefficient and similarity ratio of time are derived; corresponding model soil materials are prepared for the weakly permeable soft soil layer and the confined aquifer of the target soft soil layer respectively. S2. Deployment of In-situ Sensors and Precise Model Filling: The lower confined aquifer is filled in the model box using underwater sandfall or layered compaction. In the design precipitation influence zone, an in-situ integrated soil and water pressure sensor is precisely implanted using a miniature support. A water-stop sealing structure is laid at the interface between the sand and clay layers to control the interlayer flow parameters. Subsequently, the upper weakly permeable soft soil layer is filled using mud hydrostatic consolidation or layered preloading. An equivalent overburden load is applied to the top of the soil layer to complete the normal consolidation initialization of the geostress field. S3. High-precision pressurized water pressure servo initialization and calibration: Connect all test components and complete zero-point calibration; turn on the pressurized water pressure precision simulation servo system, and steadily raise the water pressure at the boundary of the pressurized water layer to the preset initial head height, while maintaining the groundwater level of the upper soft soil layer saturated through the stratified water level control tank; let the system stand still until the monitoring data has no drift, and complete the steady-state reconstruction of the initial water and soil stress field of the model; S4. Dynamic stage simulation of precipitation-soil coupling: Through programmable control of multi-channel peristaltic pump matrix, staged pumping test is performed according to preset flow curve or target drawdown curve, which is divided into stepped precipitation, continuous precipitation and pumping stop recovery sub-stages; during precipitation, the pressurized water pressure precision simulation servo system precisely controls the boundary replenishment water volume and water pressure according to the set external boundary conditions. S5. Multidimensional Data Acquisition and Analysis of Soil-Water Coupling Mechanism: Data on soil-water pressure, stratified settlement, well flow rate, and head changes are synchronously and continuously acquired using a high-speed data acquisition instrument. Based on synchronous data from the same location using an integrated sensor, the three-dimensional trajectory of effective stress evolution at the measuring point is calculated in real time. Combined with the full-field deformation cloud map extracted using PIV technology, the nonlinear synergistic relationship between the delayed consolidation deformation of the weakly permeable layer and the instantaneous compression deformation of the confined aquifer is quantitatively analyzed, revealing the physical transmission mechanism of differential settlement of strata caused by precipitation.

10. The physical model test method for dewatering and soil-water coupling in soft soil confined aquifers according to claim 9, characterized in that, In step S1, the geometric similarity ratio is selected as 1:30 to 1:50, while the unit similarity ratio is kept at 1:1; the weakly permeable soft soil layer is prepared by mixing kaolin, bentonite, talc powder and undisturbed soft soil in a certain proportion, and its liquid limit, plastic limit and permeability coefficient are adjusted to match the target formation parameters; the confined aquifer is prepared by mixing standard quartz sand according to grade to meet the permeability design requirements.

Images