A method and system for predicting the pull-out bearing capacity of slurry mixed piles

Abstract

The present application provides a method and system for predicting the pull-out bearing capacity of a slurry mixed pile. The method is modeled by integrating geomechanics with monitoring data. First, the geometric parameters of the interface of multiple heterogeneous soil layers and mechanical tests are obtained to construct a database, and the gradient identification and collaborative verification of the pile deformation and load data are combined to improve reliability. The longitudinal segmented units are divided and the anti-slip parameters are adjusted to generate a spatial mapping table, and a debonding dynamics model with interlayer blocking effects is established to simulate the debonding extension process and output load evolution data. The maximum load in the critical state is determined based on data convergence analysis as the predicted value of the pull-out bearing capacity. The technical solution provided by the present application realizes the accurate prediction of the pull-out bearing capacity of pile foundations in complex soil layers through dynamic parameter mapping and debonding process simulation, overcoming the defect that the traditional model is insufficient in characterizing interlayer heterogeneity and dynamic debonding behavior.

Description

Technical Field

[0001] The present application relates to the technical field of pile foundation bearing capacity analysis and prediction, and in particular to a method and system for predicting the pull-out bearing capacity of a slurry mixed pile. Background Art

[0002] Accurately predicting the pullout bearing capacity of slurry-mixed piles in complex soil layers composed of alternating sand and clay is a key challenge in the design of anti-floating underground structures. In practical engineering, the pile-soil interface is prone to debonding and slippage due to factors such as varying soil composition and uneven penetration of the solidifying slurry. Traditional static models struggle to characterize the dynamic impact of interlayer heterogeneity on debonding propagation. Therefore, predictive methods that integrate multi-source data and dynamically simulate the debonding process are urgently needed to improve reliability.

[0003] The current mainstream approach uses a layered homogenization model combined with static equilibrium theory. This approach simplifies multiple soil layers into homogenized units, calculates the anti-slip forces of each layer based on average interface parameters obtained from laboratory tests, and then superimposes these to predict the overall pullout bearing capacity. This approach incorporates an interlayer shear force reduction factor to correct for heterogeneity and uses pile strain monitoring data to invert the load distribution.

[0004] Although existing layered homogenization models achieve computational efficiency by simplifying soil layer units and using static equilibrium theory, their homogenization assumptions ignore the blocking effect of soil transition zones and the spatial differences in debonding behavior, leading to the accumulation of errors in interlayer shear force transmission. Furthermore, the static equilibrium framework cannot characterize the nonlinear dynamic process of debonding from local triggering to multi-layer cascade expansion, making it difficult to capture the critical state mutation characteristics. Furthermore, the reduction coefficients based on empirical calibration have poor adaptability to new slurry ratios or complex soil layer combinations, ultimately resulting in high dispersion and insufficient reliability in the prediction results. Summary of the Invention

[0005] The present application provides a method and system for predicting the pull-out bearing capacity of a slurry mixed pile, which is used to solve the problem in the prior art that it is impossible to accurately characterize the dynamic effects of interlayer heterogeneity and the nonlinear process of debonding extension.

[0006] In a first aspect, the present application provides a method for predicting the pull-out bearing capacity of a slurry-mixed pile, comprising:

[0007] Obtain the geometric characteristic parameters of the interface structure formed by solidified slurry in multi-layer heterogeneous soil layers at different ratios;

[0008] Determine the interface mechanical parameters of the multi-layer heterogeneous soil layer through mechanical tests, and construct an interface mechanical database in combination with the geometric characteristic parameters;

[0009] Collecting pile deformation monitoring data and load monitoring data, generating deformation gradient features by identifying deformation mutation thresholds, and collaboratively verifying the load monitoring data and deformation gradient features;

[0010] Dividing the pile into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layer, adjusting the anti-slip parameters in the interface mechanics database based on the proportion of soil layer components in each segmented unit, and generating a spatial anti-slip parameter mapping table;

[0011] Establishing an interface debonding dynamics model, by inputting the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics, combining the interlayer blocking effect parameters to simulate the spatially differentiated expansion process of debonding behavior, and outputting evolution data related to load and depth;

[0012] Based on the convergence relationship between the evolution data and the deformation gradient characteristics, the maximum load threshold corresponding to the critical extension state of debonding is determined as the prediction value of the pull-out bearing capacity.

[0013] Optionally, the collecting of pile deformation monitoring data and load monitoring data, generating a deformation gradient feature by identifying a deformation mutation threshold, and collaboratively verifying the load monitoring data and the deformation gradient feature, includes:

[0014] Synchronously collect multi-dimensional deformation monitoring data of the pile surface and load monitoring data of the pile top at fixed time intervals;

[0015] Performing continuous spatial scanning on the deformation monitoring data, locating adjacent areas where deformation values ​​suddenly change as mutation areas, extracting deformation amplitudes and direction change angles in the mutation areas, and generating a set of deformation mutation thresholds;

[0016] According to the amplification extreme value points in the deformation mutation threshold set, the deformation gradient intervals are divided along the longitudinal direction of the pile body, and the rate of change of the deformation amplification with depth in each deformation gradient interval is calculated to generate the deformation gradient feature;

[0017] Performing event synchronization matching on the load monitoring data according to the acquisition timestamp and the generation timestamp of the deformation gradient feature, and screening out the synchronous change interval of the load increase and the deformation gradient increase;

[0018] The load data and deformation gradient characteristics within the synchronous change interval are verified for amplitude ratio. If the ratio of the load increase to the deformation gradient increase exceeds the preset fluctuation range, the abnormal data segment is eliminated and the verified deformation gradient characteristics are output.

[0019] Optionally, the pile body is divided into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layer, and the anti-slip parameters in the interface mechanics database are adjusted based on the proportion of soil layer components in each segmented unit to generate a spatial anti-slip parameter mapping table, including:

[0020] Extracting the longitudinal distribution profile of the soil layer around the pile based on the drilling exploration data, dividing the pile body into multiple concentric ring-shaped segmented units at preset depth intervals, calculating the volume ratio of sand and clay within each segmented unit, and converting the volume ratio into a soil layer component ratio coefficient;

[0021] Extracting foundation anti-slip parameters corresponding to sand and clay from the interface mechanics database, linearly superimposing the foundation anti-slip parameters corresponding to sand and clay according to the soil layer component ratio coefficient, and generating anti-slip composite parameters adapted to the current segmented unit;

[0022] Arranging the depth position of each segmented unit and the corresponding anti-slip composite parameter in a vertical order to form a one-dimensional anti-slip parameter mapping table with depth as the index;

[0023] Based on the one-dimensional anti-slip parameter mapping table, the anti-slip composite parameters of adjacent segmented units are nonlinearly interpolated according to the soil layer transition gradient to generate a continuous anti-slip parameter field;

[0024] The continuous anti-slip parameter field is spatially bound to the three-dimensional geometric model of the pile body, and a global anti-slip parameter distribution map of the pile-soil interface is output as a spatial anti-slip parameter mapping table.

[0025] Optionally, the interface debonding dynamics model is established by inputting the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics, combining the interlayer blocking effect parameters to simulate the spatially differentiated expansion process of the debonding behavior, and outputting the evolution data of the load-depth correlation, including:

[0026] Based on the spatial anti-slip parameter mapping table, a pile-soil interface mechanical network with segmented units as nodes is constructed;

[0027] Inputting the verified deformation gradient characteristics into the pile-soil interface mechanical network, calculating the difference between the deformation gradient and the anti-slip composite parameter at each node, and generating the node debonding trigger criterion;

[0028] Based on the interlayer blocking effect parameters, the resistance weight of the debonding extension between adjacent nodes is defined. If the current node meets the debonding trigger criterion, the probability of the debonding extending to the adjacent node is calculated according to the resistance weight.

[0029] Using the load monitoring data as a time-series driving signal, the debonding state of each node is iteratively updated. When the debonding extends from the surface nodes to the deep nodes, the load value of each deep node when the debonding is triggered is recorded;

[0030] The load values ​​and corresponding depths at the time of debonding triggering of all nodes are aggregated to generate the evolution data of the load as the debonding depth increases.

[0031] Optionally, aggregating the load values ​​and corresponding depths of all nodes when debonding is triggered to generate load evolution data as the debonding depth increases includes:

[0032] Traversing all nodes in the pile-soil interface mechanical network, sorting the nodes from shallow to deep according to their depth, and extracting the load value and depth position when debonding is first triggered at each node;

[0033] The depth intervals between adjacent nodes are checked for continuity. If the depth difference between adjacent nodes exceeds a preset threshold, a virtual node is inserted into the interval and the load value of the previous node is linearly distributed to the virtual node according to the depth difference.

[0034] Integrating the load values ​​and depths of the actual nodes and the virtual nodes in a triggering order to generate an initial load depth sequence;

[0035] Detecting the presence of multiple load values ​​at the same depth in the initial load depth sequence, retaining the maximum load value and eliminating redundant data points to form a final data point;

[0036] The final data points are arranged in order of increasing depth to generate continuous and monotonic load evolution data with increasing debonding depth.

[0037] Optionally, determining a maximum load threshold corresponding to a critical debonding extension state as a predicted value of pull-out bearing capacity based on a convergence relationship between the evolution data and the deformation gradient characteristics includes:

[0038] Extracting the amplification slope of the load changing with the debonding depth from the evolution data, and marking the depth point where the amplification slope reverses direction for the first time as the first critical point;

[0039] Extracting the deformation increase corresponding to the depth of the first critical point from the deformation gradient feature, and marking it as the second critical point if the deformation increase exceeds a set multiple of the historical average increase;

[0040] The depth difference between the first critical point and the second critical point is recorded. When the depth difference is less than a preset tolerance, the load value corresponding to the two is determined to be the maximum load threshold of the critical extension state of debonding. When the depth difference exceeds the preset tolerance, the load application time is extended until a stable point where the load increase returns to zero appears in the evolution data. The last load peak before the stable point is used as the maximum load threshold, and the maximum load threshold is output as the predicted value of the pull-out bearing capacity of the pile foundation.

[0041] Optionally, the determining of the interface mechanical parameters of the multi-layer heterogeneous soil layers by mechanical testing and the construction of an interface mechanical database in combination with the geometric characteristic parameters include:

[0042] preparing a composite soil sample containing alternating layers of sand and clay, and embedding a solidifying slurry into the composite soil sample to form a simulated pile-soil interface;

[0043] Applying multi-level tensile loads to the simulated pile-soil interface, monitoring the load value and displacement during the interface separation process, and extracting the peak point of the load-displacement curve as the interface tensile strength;

[0044] After the interface is completely separated, the friction sliding resistance of the separation surface is measured, and the interface friction angle is calculated based on the relationship between the sliding displacement and the friction sliding resistance;

[0045] Extracting the surface undulation height and contact area ratio of the simulated pile-soil interface to generate a set of geometric feature parameters;

[0046] Under each set of ratios, the interface tensile strength and interface friction angle set are parameter-matched to establish interface mechanical parameters indexed by the ratio number;

[0047] Aggregate the geometric feature parameter sets and the interface mechanical parameters under all ratio numbers to generate an interface mechanical database containing the ratio, geometry and mechanical mapping relationship.

[0048] In a second aspect, the present application provides a prediction system for the pull-out bearing capacity of a slurry mixed pile, comprising:

[0049] An acquisition module is used to obtain the geometric characteristic parameters of the interface structure formed by the solidified slurry in multiple heterogeneous soil layers at different ratios;

[0050] The acquisition module further comprises determining the interface mechanical parameters of the multi-layer heterogeneous soil layer through mechanical testing, and constructing an interface mechanical database in combination with the geometric characteristic parameters;

[0051] A verification module collects pile deformation monitoring data and load monitoring data, generates deformation gradient features by identifying deformation mutation thresholds, and performs collaborative verification on the load monitoring data and deformation gradient features;

[0052] a generation module that divides the pile into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layer, adjusts the anti-slip parameters in the interface mechanics database based on the proportion of soil layer components in each segmented unit, and generates a spatial anti-slip parameter mapping table;

[0053] The evolution module establishes an interface debonding dynamics model, inputs the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics, combines the interlayer blocking effect parameters to simulate the spatially differentiated expansion process of the debonding behavior, and outputs the evolution data associated with the load and depth;

[0054] The prediction module determines, based on the convergence relationship between the evolution data and the deformation gradient characteristics, a maximum load threshold corresponding to the critical extension state of debonding as a predicted value of the pull-out bearing capacity.

[0055] In a third aspect, an embodiment of the present application provides a computing device comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for predicting the pull-out bearing capacity of a mud mixed pile as described in the first aspect above.

[0056] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, the method for predicting the pull-out bearing capacity of a mud-mixed pile as described in the first aspect is implemented.

[0057] In an embodiment of the present application, geometric characteristic parameters of the interface structure formed by solidified slurry in multi-layer heterogeneous soil layers at different proportions are obtained; the interface mechanical parameters of the multi-layer heterogeneous soil layers are measured through mechanical tests, and an interface mechanical database is constructed in combination with the geometric characteristic parameters; pile deformation monitoring data and load monitoring data are collected, deformation gradient characteristics are generated by identifying deformation mutation thresholds, and the load monitoring data and deformation gradient characteristics are collaboratively verified; the pile body is divided into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layers, and the anti-slip parameters in the interface mechanical database are adjusted based on the proportion of soil layer components in each segmented unit to generate a spatial anti-slip parameter mapping table; an interface debonding dynamics model is established, and the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics are input, combined with interlayer blocking effect parameters to simulate the spatially differentiated expansion process of debonding behavior, and output load-depth correlation evolution data; based on the convergence relationship between the evolution data and the deformation gradient characteristics, the maximum load threshold corresponding to the critical debonding expansion state is determined as the pullout bearing capacity prediction value.

[0058] The technical solution of this application has the following beneficial effects:

[0059] This application constructs a multidimensional database by acquiring the geometric characteristic parameters of the interface of multiple heterogeneous soil layers and mechanical test data, and combines the gradient characteristics of pile deformation and load monitoring data for collaborative verification to achieve high-reliability fusion of dynamic monitoring data; divides the longitudinal segmented units based on the spatial distribution of the soil layer and dynamically adapts the anti-slip parameters to generate a mapping table, breaking through the static limitations of the homogenized model; integrates the interlayer blocking effect and three-dimensional parameter mapping through the interface debonding dynamics model, accurately simulates the spatially differentiated expansion process of debonding behavior, combines the evolution data with the convergence analysis of deformation characteristics, effectively captures the critical state mutation point, and ultimately achieves high-precision prediction of pull-out bearing capacity in complex soil layers, significantly improving the comprehensive characterization capability of interlayer heterogeneity, dynamic debonding behavior and load evolution nonlinearity.

[0060] Furthermore, multi-dimensional deformation data of the pile body and pile top load data are collected synchronously at fixed time intervals. Continuous spatial scanning is used to locate the deformation mutation area and extract the amplitude and direction angle to generate a mutation threshold set. The deformation gradient interval is divided vertically based on the threshold extreme point, and the rate of change of deformation amplitude with depth in each interval is calculated to form a gradient feature. The load data and deformation gradient feature are matched according to the timestamp, and the intervals with synchronous changes in load and deformation amplitude are screened. Abnormal data segments are eliminated through amplitude ratio verification, and a high-confidence verified gradient feature is output. Through dynamic synchronous monitoring and multi-dimensional deformation mutation threshold identification, abnormal data caused by equipment errors or external interference are effectively eliminated, and the temporal and spatial evolution laws of load changes and deformation gradients are accurately correlated. The amplitude ratio verification mechanism is combined to strengthen data consistency, significantly improve the authenticity and reliability of deformation gradient features, provide high-precision input for subsequent debonding dynamics models, and enhance the anti-interference ability of pull-out bearing capacity prediction and the accuracy of critical state capture.

[0061] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0063] Figure 1 A flow chart showing a method for predicting the pull-out bearing capacity of a slurry mixed pile provided in the present application is shown;

[0064] Figure 2 A scenario diagram showing a method for predicting the pull-out bearing capacity of a slurry mixed pile provided in the present application is shown;

[0065] Figure 3 A schematic diagram of the structure of a prediction system for the pull-out bearing capacity of a slurry mixed pile provided by the present application is shown;

[0066] Figure 4 A schematic structural diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0067] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0068] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0069] The study found that there are significant challenges in predicting the pullout bearing capacity of slurry-mixed piles in complex soil layers with alternating layers of sand and clay. Traditional static models have difficulty accurately characterizing the slip behavior of the pile-soil interface because they ignore the dynamic effects of interlayer heterogeneity and the nonlinear process of debonding expansion. Although the existing layered homogenization model improves computational efficiency by simplifying soil layer homogeneity and superimposing static equilibrium, its homogenization assumption conceals the blocking effect of the transition zone and the spatial differences in debonding. The static framework is unable to simulate the cascading debonding expansion process. In addition, the reliance on empirical reduction factors leads to poor adaptability to new slurries or complex soil layers, ultimately resulting in high discreteness and insufficient reliability of the prediction results.

[0070] In response to the above problems, this application proposes a method for predicting the pull-out bearing capacity of mud-mixed piles. The core of this method is to break through the limitations of traditional static models through collaborative modeling of geometric and mechanical parameters, spatiotemporal monitoring data verification, and three-dimensional differentiated debonding simulation. First, the geometric parameters of the interface of multi-layer heterogeneous soil layers are obtained and an interface mechanics database is constructed through mechanical tests; the deformation and load data of the pile body are collected simultaneously, and the reliability of the data is improved through the identification of deformation mutation thresholds and the collaborative verification of gradient characteristics; the longitudinal segmented units are divided based on the spatial distribution of the soil layer, and the anti-slip parameters are dynamically adjusted to generate a three-dimensional mapping table; a debonding dynamics model with interlayer blocking effect is established to simulate the spatially differentiated expansion process of debonding behavior and output load depth evolution data; finally, the maximum load threshold of the critical state is accurately determined through the convergence analysis of the evolution data and deformation characteristics. This method effectively characterizes the blocking effect of interlayer heterogeneity on interface slip through the multi-source fusion of geometric mechanical parameter database and dynamic monitoring data, solving the spatial error accumulation problem of the homogenized model. Based on the three-dimensional differentiated simulation of the debonding dynamics model, it completely restores the nonlinear process of debonding from local triggering to multi-layer cascade expansion, breaking through the bottleneck of capturing mutation characteristics in the static equilibrium framework. At the same time, through the dynamic adaptation of soil layer segmentation parameters and data-driven verification mechanism, it reduces the dependence on empirical reduction factors, significantly improves the adaptability of new slurry ratios and complex soil layer combinations, and realizes the accurate prediction of the pull-out bearing capacity of pile foundations in complex soil layers.

[0071] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0072] Figure 1 A flowchart of a method for predicting the pull-out bearing capacity of a slurry mixed pile is provided for an embodiment of the present application. Figure 1 As shown, the method includes:

[0073] 101. Obtain the geometric characteristic parameters of the interface structure formed by solidified slurry in multi-layer heterogeneous soil layers at different ratios;

[0074] In the above scheme, geometric characteristic parameters refer to the morphological quantitative indicators of the interface structure between the solidified slurry and the soil layer in multi-layer heterogeneous soil layers, including interface penetration width, interface roughness, branch network density and interface inclination. These parameters are obtained through three-dimensional scanning and image analysis technology and are used to characterize the spatial heterogeneity of the interface structure under different slurry ratios.

[0075] In the embodiment of the present application, first, three-dimensional imaging of the multi-layer heterogeneous soil solidification slurry complex sampled on site was performed using CT scanning technology to obtain high-resolution images of the interface structure under different slurry ratios;

[0076] Secondly, a digital image edge detection algorithm is used to perform layered processing on the three-dimensional image to identify the contact boundary between the slurry and the soil layer, and to extract the interface penetration width and branch network density.

[0077] Then, using 3D surface reconstruction technology, a quantitative model of interface roughness was generated based on point cloud registration and surface fitting, and the distribution characteristics of the interface inclination angle were calculated through principal component analysis.

[0078] Finally, the penetration width, roughness, branch density and inclination data under different ratios are integrated into a structured database to form a geometric feature parameter set, which provides spatial morphological input for subsequent mechanical parameter matching and anti-slip analysis.

[0079] In practical applications, first, a high-resolution industrial CT scanner such as the X5000 series is used to perform three-dimensional tomography on the mud-mixed pile-soil complex samples drilled on site to obtain the original image data of the penetration interface structure of the solidified slurry in the alternating layers of sand and clay under different slurry ratios, and the resolution is controlled within 10μm to capture the microscopic morphology; then, an image processing module based on the Canny edge detection algorithm is used to perform layered slice analysis on the three-dimensional scanning data, and the contact boundary between the slurry and the soil layer is identified through multi-threshold adaptive grayscale segmentation technology to extract the lateral penetration Width, branch network density and interface contour geometric characteristics; on this basis, the layered slice data are reconstructed into a three-dimensional spatial model using point cloud registration technology, and the interface roughness is quantified by combining the non-uniform rational B-spline surface fitting method. The statistical mean and variance of the angle between the slurry diffusion direction and the pile axis are calculated through principal component analysis to obtain the interface inclination distribution; finally, the penetration width, roughness, branch density and inclination data under different ratios are classified and stored according to the soil layer type, and a structured geometric feature parameter database is constructed to provide spatial heterogeneity input for subsequent mechanical parameter matching and anti-slip analysis.

[0080] The above-mentioned 101 overall solution, through high-precision three-dimensional imaging and quantitative analysis technology, systematically obtains the geometric characteristic parameters of the interface structure under different slurry ratios, overcoming the blind spots of traditional manual measurement in characterizing microscopic morphologies such as seepage branches and roughness; through the construction of a structured parameter set, it provides spatially heterogeneous input data for subsequent dynamic anti-slip analysis, supports the precise modeling of the blocking effect between layers and the debonding extension path, and fundamentally improves the reliability and adaptability of the pull-out bearing capacity prediction in complex soil layers.

[0081] 102. Determine the interface mechanical parameters of the multi-layer heterogeneous soil layer through mechanical testing, and construct an interface mechanical database based on the geometric characteristic parameters;

[0082] Optionally, step 102 may specifically include the following steps:

[0083] 1021. Prepare a composite soil sample containing alternating layers of sand and clay, and embed a solidifying slurry into the composite soil sample to form a simulated pile-soil interface;

[0084] 1022. Applying multi-level tensile loads to the simulated pile-soil interface, monitoring the load value and displacement during the interface separation process, and extracting the peak point of the load-displacement curve as the interface tensile strength;

[0085] 1023. After the interface is completely separated, measure the frictional sliding resistance of the separation surface, and calculate the interface friction angle based on the relationship between the sliding displacement and the frictional sliding resistance;

[0086] 1024. Extract the surface fluctuation height and contact area ratio of the simulated pile-soil interface to generate a set of geometric feature parameters;

[0087] 1025. Perform parameter matching on the interface tensile strength and interface friction angle set under each set of ratios to establish interface mechanical parameters indexed by the ratio number;

[0088] 1026. Aggregate the geometric feature parameter sets and the interface mechanical parameters under all the ratio numbers to generate an interface mechanical database including the ratio, geometry, and mechanical mapping relationship.

[0089] In the above scheme, the simulated pile-soil interface refers to an artificially fabricated pile-like contact surface, recreating the actual stratigraphic structure of alternating sand and clay layers, including the cross-linked permeability of the solidified slurry and soil. This can be used to replicate the mechanical and geometric properties of the actual pile-soil interface. The interfacial tensile strength refers to the maximum bearing capacity of the pile-soil interface when it separates under tensile load and can be used to characterize the interface's ultimate resistance to debonding failure. The interfacial friction angle, the angular parameter corresponding to the ratio of sliding resistance to normal stress during the sliding phase of the pile-soil interface, can be used to quantify the residual friction characteristics after debonding. The geometric characteristic parameter set refers to a set of quantitative indicators describing the surface morphology of the pile-soil interface. It includes surface relief height distribution data acquired through 3D laser scanning, including contact area ratio and roughness fractal dimension, and can be used to reveal the spatial influence of geometric morphology on the mechanical behavior of the interface. The interface mechanics database is a linked data system integrating interface mechanical and geometric parameters for different mixes. It contains mechanical and geometric parameters indexed by mix number and includes a multidimensional relational data structure, providing high-fidelity input for dynamic prediction of pullout bearing capacity.

[0090] In the embodiment of the present application, first, step 1021 is used to determine the thickness ratio of the sand and clay layers based on the drilling data of the engineering site, and a pneumatic layered compactor is used to fill the soil sample layer by layer. The thickness error of each layer after compaction is controlled within a certain range to ensure that the transition morphology between layers is consistent with the actual stratum. A vertical hole is drilled in the center of the soil sample, and a curing slurry of a preset ratio is injected. The slurry is allowed to stand and cure for 72 hours to form a simulated pile body with a diameter proportional to the prototype of the pile body, ensuring that the slurry penetration range and the cross-linking morphology of the soil body are close to the actual working conditions. A CT scan is performed on the cured pile-soil interface to confirm that there are no voids or cracks, and the shear wave velocity test is used to verify that the deviation between the density of the soil sample layers and the field data is less than a certain preset value.

[0091] Next, in step 1022, the composite soil sample was secured to the upper and lower clamps of the tensile testing machine. The loading rate was set to 0.1 mm / s, with a range extending to 1.2 times the estimated ultimate load. The force sensor and displacement meter were simultaneously activated. Loading was performed in a 10 kN gradient, with each load level maintained for 30 seconds to eliminate creep interference. The load-displacement curve was recorded. A filtering algorithm was used to smooth noise, and a local extremum search algorithm was used to identify the peak point of the curve. The load value corresponding to the peak point was used as the interfacial tensile strength.

[0092] Then, in step 1023, a normal pressure is applied to the separated interface to simulate the soil's own weight stress. The pressure is maintained constant by the servo hydraulic system. Pulling is performed along the sliding direction of the interface at a rate of 0.5 mm / s, and the sliding resistance versus displacement curve is recorded. A moving average algorithm is used to smooth the data and extract the mean resistance value during the stable phase. The interfacial friction angle is calculated according to Coulomb's friction law. The specific calculation formula is as follows: , where the normal pressure is , the sliding resistance is , the interface friction angle is .

[0093] Next, in step 1024, a line laser scanner is used to scan the separated interface from multiple angles to generate 3D point cloud data. The multi-view data is then registered using the ICP algorithm to reconstruct the complete interface topography. A fractal dimension algorithm is used to calculate the surface height distribution, extracting the maximum peak-to-valley difference, average roughness, and fractal dimension. The 3D point cloud data is projected onto a 2D plane, and a threshold segmentation algorithm is used to distinguish contact and non-contact areas. The actual percentage of contact pixels is counted, and the spatial distribution uniformity index of the contact area is calculated to generate a set of geometric feature parameters.

[0094] Next, in step 1025, parameters such as interfacial tensile strength, interfacial friction angle, undulation height, and contact area ratio for a single mix are encoded in JSON format as key-value pairs, with the mix number serving as a unique identifier. For example, for a mix with a sand content of 40%, a clay content of 60%, and a slurry mix number of P7-3, the mix number is "S40-C60-P7-3." The key-value pair stored in JSON format is {"tensile strength": 38,"friction angle": 14,"undulation height": 0.8}. The parameter set is stored in a relational database, and a primary-key and foreign-key association table is established to support joint queries based on multiple conditions, such as sand content ratio and slurry type.

[0095] Finally, in step 1026, all mix parameter sets are deduplicated and verified, outliers are removed, and an interface mechanics database is generated. A three-dimensional data cube is constructed within the interface mechanics database. The dimensions include mix number, soil layer ratio, and slurry type, and the metrics include mechanical and geometric parameters. This data cube is exposed to the debonding dynamics model through an API interface.

[0096] In practical applications, taking the alternating layers of 45% sand and 55% clay as an example, the interface mechanics database was constructed using the method of the present invention. First, according to the layer thickness ratio of the geological survey report, a pneumatic layered compactor was used to prepare a composite soil sample, and the sand layer and the clay layer were laid alternately. The single layer thickness was 15 cm for sand and 18 cm for clay, with a total height of 1.2 m. A 10 cm diameter solidified slurry of cement and bentonite was pre-buried in the center of the soil sample in a ratio of 7:3 and a water-cement ratio of 0.6. The cement was cured for 72 hours to form a simulated pile-soil interface, and the shear wave velocity test was used to verify that the density deviation was less than the preset value. The soil sample was installed in a 200kN tensile testing machine, loaded in stages at a rate of 0.1 mm / s, and the load-displacement curve was collected simultaneously. After filtering and noise reduction, the peak load was identified as 53.7 kN, the tensile strength was calculated to be 53.7 kN, and the critical strain energy was calculated. A normal pressure of 120 kPa was applied to the separation interface, and the sliding traction was performed at a rate of 0.5 mm / s. The average sliding resistance was measured to be 34.2 kPa, and the interfacial friction angle was calculated to be 15.8. A line laser scanner was used to acquire 3D point cloud data of the interface. After ICP registration and denoising, the average surface relief height was calculated to be 1.2 mm. These parameters were bound to the mix number "S45-C55-P7-3" and stored in JSON format as {"tensile strength": 53.7,"friction angle": 15.8,"relief height": 1.2}. This mix was aggregated with data from 12 other mixes to construct a 3D relational table containing fields such as soil layer ratio, slurry type, mechanical parameters, and geometric parameters. This table was then exposed to the debonding dynamics model via an API.

[0097] The 102-point overall solution, through the preparation of composite soil samples and the integration of multidimensional parameters, constructs a high-precision interface mechanics database, systematically addressing the challenge of predicting pullout resistance in complex soil layers. First, a simulated pile-soil interface is prepared based on alternating sand-clay layers to replicate the actual stratum structure. Multi-stage tensile testing accurately extracts the interfacial tensile strength, and combined with dynamic monitoring of sliding resistance and friction angle calculation, quantifies the residual friction characteristics after debonding. 3D laser scanning and image analysis techniques capture geometric features such as surface relief height and contact area ratio, revealing the spatial influence of topography on mechanical behavior. Furthermore, using mix number as an index, mechanical parameters are dynamically matched with geometric features to form a structured multidimensional database, overcoming the limitations of traditional homogenized models that simplify interlayer heterogeneity and transition zone blocking effects. Ultimately, the coupled mechanical and geometric data supports spatial mapping of anti-slip parameters and debonding dynamics simulations, significantly improving the adaptability of complex mixes and soil layer combinations, reducing reliance on empirical coefficients, and providing a data foundation for pullout bearing capacity prediction that combines theoretical rigor with engineering practicality.

[0098] 103. Collecting pile deformation monitoring data and load monitoring data, generating deformation gradient features by identifying a sudden deformation change threshold, and collaboratively verifying the load monitoring data and the deformation gradient features;

[0099] Optionally, step 103 may specifically include the following steps:

[0100] 1031. Synchronously collect multi-dimensional deformation monitoring data of the pile body surface and load monitoring data of the pile top at fixed time intervals;

[0101] 1032. Perform continuous spatial scanning on the deformation monitoring data, locate adjacent areas where deformation values ​​suddenly change as mutation areas, extract deformation amplitudes and direction change angles in the mutation areas, and generate a set of deformation mutation thresholds;

[0102] 1033. Divide the pile longitudinally into deformation gradient intervals based on the amplification extreme value points in the deformation mutation threshold set, calculate the rate of change of deformation amplification with depth in each deformation gradient interval, and generate deformation gradient features;

[0103] 1034. Perform event synchronization matching on the load monitoring data according to the acquisition timestamp and the generation timestamp of the deformation gradient feature, and filter out the synchronous change interval of the load increase and the deformation gradient increase;

[0104] 1035. Perform amplitude ratio verification on the load data and the deformation gradient characteristics within the synchronous change interval. If the ratio of the load increase to the deformation gradient increase exceeds a preset fluctuation range, remove the abnormal data segment and output the verified deformation gradient characteristics.

[0105] In the above scheme, deformation monitoring data refers to the multi-dimensional displacement and strain signals collected by sensors on the pile surface. These signals include dynamic changes in axial tension, radial compression, and bending deformation, and can be used to characterize the structural response of the pile under pullout loads. Load monitoring data refers to the instantaneous pullout load values ​​recorded by sensors at the top of the pile. These data include load magnitude, direction of action, and temporal evolution trends, and are used to quantify the mechanical input of external loads to the pile-soil system. The deformation mutation threshold set refers to the set of characteristic parameters of the local mutation region of the pile deformation extracted through spatial scanning. These parameters include the deformation amplitude mutation amount, mutation direction angle, and spatial coordinate information, and are used to locate the starting point and extension path of debonding and slippage behavior. The deformation gradient characteristic refers to the distribution of the rate of change of deformation increase with depth within the deformation gradient interval divided along the longitudinal direction of the pile. These characteristics include gradient slope, spatial continuity, and strength attenuation characteristics, and are used to describe the three-dimensional extension strength and interlayer blocking effect of debonding behavior.

[0106] In this embodiment, synchronous data acquisition is first achieved through a distributed sensor network in step 1031. An array of fiber Bragg grating sensors is evenly distributed across the pile surface, symmetrically spaced at a fixed axial spacing. Multi-dimensional deformation data, such as axial strain, radial displacement, and bending curvature, is collected at a fixed sampling frequency. Simultaneously, a high-precision hydraulic sensor is installed at the top of the pile to record instantaneous load values ​​at the same sampling frequency. Data synchronization is achieved through a clock protocol, ensuring that the timestamp alignment error of the deformation and load data is less than 1ms.

[0107] Next, dynamic mutation detection based on a spatial sliding window is performed, employing a sliding window algorithm to continuously scan the pile deformation data in space. The standard deviation of the deformation values ​​is calculated within each window. If the rate of change in the standard deviation of adjacent windows exceeds a certain threshold, the region is marked as a mutation area. The deformation amplitude and directional change angle of the mutation area are further extracted. The deformation amplitude, directional change angle, and spatial coordinates of all mutation areas constitute the set of deformation mutation thresholds.

[0108] Then, in step 1033, based on the extreme deformation increase points in the mutation threshold set, the deformation gradient intervals are divided along the longitudinal direction of the pile body, for example, every 0.5m is an interval. Within each interval, the least squares method is used to perform a linear fit between the deformation increase and the depth to calculate the deformation gradient slope. The specific calculation formula is as follows: , where k is the deformation gradient slope, is the deformation amplification, The depth change is then calculated, and its confidence is finally generated, which includes the depth range, gradient slope and confidence of each interval.

[0109] Next, in step 1034, the time series of the load monitoring data is aligned with the timestamps generated by the deformation gradient features. A dynamic time warping algorithm is used to compensate for sensor response delays, such as the mechanical hysteresis between hydraulic and fiber optic sensors. After time alignment, the Pearson correlation coefficient between the load increase and the deformation gradient increase is calculated using a sliding window. Periods with a Pearson correlation coefficient greater than 0.8 are selected as synchronous change intervals, and low-correlation noise segments are eliminated.

[0110] Finally, in step 1035, the ratio of the load increase to the deformation gradient increase is calculated frame by frame within the synchronous variation interval. The ratio is preset to a reasonable fluctuation range of 0.8 to 1.2, which can be adjusted based on the slurry ratio. If the ratio in consecutive fixed frames exceeds this range, it is determined to be an abnormal segment. The abnormal segment data is smoothed and repaired using a filter. If the repaired segment still does not meet the threshold, it is directly removed. Finally, the verified deformation gradient characteristic curve and associated load data are output.

[0111] In practical application, a microstrain gauge array was first placed symmetrically in two groups, 0.1 m apart axially, on the surface of a simulated pile with a diameter of 50 mm and a height of 400 mm in the laboratory. Axial strain and radial displacement data were collected at a sampling frequency of 10 Hz. A microload sensor with a 50 kN range was installed on the pile top to simultaneously record the load data. Deformation and load time stamps were synchronized using hardware trigger signals. During the construction loading phase, when the load was gradually increased to 8 kN, a sudden change in the standard deviation of deformation at a depth of 0.15 m was detected, reaching a rate of 65%. The deformation increase of 0.15 mm and a 42° angle pointing toward the clay layer in this region were extracted and included in the set of sudden change thresholds. Gradient intervals were divided based on the extreme value of the sudden change. The least squares method was used to fit the relationship between the deformation increase and depth in the range of 0.1 to 0.3 m. The calculated gradient slope was 0.15 mm / m, indicating that debonding behavior accelerated upward from this region. After aligning the load data using a dynamic time warping algorithm, the Pearson correlation coefficient between the deformation gradient increase and the load increase during the load increase from 6 kN to 9 kN period was 0.85, identifying it as a valid synchronization interval. During verification, it was found that the ratio of the load increase to the deformation gradient increase within a 0.2 s period reached 1.52 for three consecutive frames, exceeding the preset threshold of 0.8 to 1.2. This was confirmed to be interference from the loading device motor vibration. If the threshold was still exceeded after filtering and smoothing, the abnormal segment was removed. The final output of the verified deformation gradient characteristic curve showed that when the load reached 10.5 kN, the gradient slope in the 0.3-0.4 m interval suddenly increased to 0.22 mm / m. Combined with the convergence analysis of the evolution data, the maximum load threshold corresponding to the critical debonding state was determined to be 11 kN, which deviated only 1.8% from the laboratory destructive test result of 11.2 kN, verifying the high accuracy and reliability of this scheme in predicting pullout bearing capacity in complex soil layers.

[0112] The above-mentioned 103 overall solution, by synchronously collecting multi-dimensional deformation and load monitoring data of the pile body, combined with spatial mutation threshold identification and gradient feature extraction, dynamically correlates load changes and deformation evolution laws, and significantly improves data reliability. Based on the precise positioning of the deformation mutation area and the division of gradient intervals, it effectively characterizes the spatial blocking effect of interlayer heterogeneity on debonding behavior; through the time stamp synchronization matching and amplitude ratio verification mechanism, abnormal interference such as sensor drift and external impact is eliminated to ensure the consistency of the dynamic response of load deformation. The high-confidence deformation gradient characteristics and evolution data finally output provide accurate input for the debonding dynamics model, completely restore the nonlinear process of debonding from local triggering to multi-layer cascade expansion, break through the limitations of traditional static models in capturing critical state mutation characteristics, and achieve high-precision prediction of pull-out bearing capacity in complex soil layers, with both engineering applicability and anti-interference ability.

[0113] 104. Divide the pile into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layer, adjust the anti-slip parameters in the interface mechanics database based on the proportion of soil layer components in each segmented unit, and generate a spatial anti-slip parameter mapping table;

[0114] Optionally, step 104 may specifically include the following steps:

[0115] 1041. Extract the longitudinal distribution profile of the soil layer around the pile based on the drilling exploration data, divide the pile body into a plurality of concentric ring-shaped segmented units at preset depth intervals, calculate the volume ratio of sand and clay in each segmented unit, and convert the volume ratio into a soil layer component ratio coefficient;

[0116] 1042. Extracting foundation anti-slip parameters corresponding to sand and clay from the interface mechanics database, linearly superimposing the foundation anti-slip parameters corresponding to sand and clay according to the soil layer component ratio coefficient, and generating anti-slip composite parameters adapted to the current segmented unit;

[0117] 1043. Arrange the depth position of each segmented unit and the corresponding anti-slip composite parameter in a vertical order to form a one-dimensional anti-slip parameter mapping table with depth as the index;

[0118] 1044. Based on the one-dimensional anti-slip parameter mapping table, nonlinearly interpolate the anti-slip composite parameters of adjacent segmented units according to the soil layer transition gradient to generate a continuous anti-slip parameter field;

[0119] 1045. Spatially bind the continuous anti-slip parameter field to the three-dimensional geometric model of the pile body, and output a global anti-slip parameter distribution map of the pile-soil interface as a spatial anti-slip parameter mapping table.

[0120] In the above scheme, the soil layer component proportion coefficient refers to the weight parameter that quantifies the volume ratio of sand and clay in the heterogeneous soil layer, including the complementary relationship between the sand coefficient and the clay coefficient, which can be used to dynamically adjust the anti-slip parameters to adapt to the spatial heterogeneity of the soil layer. The anti-slip composite parameter refers to the comprehensive mechanical parameter that characterizes the equivalent anti-slip capacity of the heterogeneous soil layer, including the linear weighted superposition result of the anti-slip parameters of the sand and clay foundations, including the proportional fusion value of the friction angle and cohesion, which can be used to describe the interfacial anti-slip strength of the mixed soil layer. The one-dimensional anti-slip parameter mapping table refers to a set of longitudinal anti-slip parameters indexed by depth, which can be used to preliminarily reflect the longitudinal variation law of the anti-slip capacity of the pile-soil interface. The continuous anti-slip parameter field refers to the spatial distribution field that describes the gradual characteristics of the anti-slip capacity of the soil layer around the pile, which can be used to eliminate the segmented mutation error of the discrete model. The spatial anti-slip parameter mapping table refers to a three-dimensional dynamic description model of the global anti-slip capacity of the pile-soil interface. It contains the binding relationship between the continuous anti-slip parameter field and the pile geometry model, including the distribution heat map of the parameters in three-dimensional space, which can be used to drive high-precision simulation of the debonding dynamics model.

[0121] In the embodiment of the present application, first, through step 1041, the longitudinal distribution profile of the soil layer around the pile is extracted based on the borehole exploration data, and the alternating distribution pattern of the sand and clay layers is identified by geological interpretation software. The pile body is divided into concentric annular segmented units at preset depth intervals, and the geometry of each unit is defined by the pile diameter and the segment depth. The volume ratio of sand and clay in each segmented unit is calculated by a three-dimensional volume integration algorithm. For example, the volume of sand in the unit accounts for 70% and clay accounts for 30%, and is normalized to the soil layer component ratio coefficient, with a sand coefficient of 0.7 and a clay coefficient of 0.3. This process relies on spatial interpolation technology to fill the gaps in the borehole data and ensure the accuracy of the stratification.

[0122] Next, in step 1042, the basic anti-slip parameters of sand and clay are extracted from the interface mechanics database, and the two parameters are linearly weighted superimposed based on the component ratio coefficient to generate the anti-slip composite parameters adapted to the current segmented unit. The formula for the linear weighted superposition process is as follows: and , where the sand coefficient and clay coefficient are their proportions in the complex soil.

[0123] Next, in step 1043, the composite anti-slip parameters generated in step 1042 are arranged in longitudinal order with their corresponding segmented unit depth positions to form a one-dimensional anti-slip parameter mapping table. This one-dimensional anti-slip parameter mapping table, indexed by depth and structured as a two-dimensional array, reflects the discrete distribution characteristics of the pile's longitudinal anti-slip capacity. This process enables parameter storage and rapid retrieval using a database management tool.

[0124] Again, through step 1044, based on the discrete segmentation characteristics of the one-dimensional mapping table, the cubic spline interpolation method is used to perform nonlinear interpolation on the anti-slip composite parameters of adjacent segmented units. Combined with the soil layer transition gradient, such as the slope of the transition zone from sand to clay, a continuous anti-slip parameter field is generated to eliminate parameter mutations between discrete segments.

[0125] Finally, in step 1045, the continuous anti-slip parameter field is spatially bound to the three-dimensional geometric model of the pile using a finite element modeling tool. Mesh mapping techniques are then used to map the parameter field to the mesh nodes at the pile-soil interface. For example, the pile surface is divided into a finite element mesh, with each node associated with a corresponding anti-slip parameter value. A visualization engine is then used to render and generate a global anti-slip parameter distribution heat map. This mapping table can be directly imported into the debonding dynamics model to drive a three-dimensional differentiated simulation of debonding behavior.

[0126] In a practical application, a scaled-down model pile in a laboratory simulation was required to penetrate an alternating layer of sand and clay approximately 1.2 m thick (reduced to a 1:10 scale). First, a soil profile around the model pile was extracted using CT scan data, and image analysis software was used to identify the alternating distribution of sand and clay. The pile was then divided into 24 concentric annular segments at 0.05 m depth intervals. A two-dimensional area integration algorithm was used to calculate the soil composition within each segment. For example, within the segment with a depth of 0.2-0.25 m, sand accounted for 78.6% of the area, while clay accounted for 21.4%. Bilinear interpolation was used to fill in local data gaps, ensuring a stratification accuracy error of less than 3%. Laboratory-grade parameters were extracted from the interface mechanics database: a friction angle of 32° and cohesion of 5 kPa for sand; and an angle of 18° and cohesion of 20 kPa for clay. By linearly superposing the component coefficients, the composite friction angle of 28.7° and composite cohesion of 8.63 kPa were calculated for the segment with a depth of 0.2-0.25 m, generating the anti-slip parameters for this segment. After element-by-element calculations, a one-dimensional mapping table was constructed, and a quadratic spline interpolation algorithm combined with transition gradient correction weights was used. For the depth range of 0.25 to 0.3 m, for example, adjacent element parameters at 0.25 m corresponded to 8.63 kPa and at 0.3 m corresponded to 11.2 kPa. The interpolation function calculated 9.84 kPa at a depth of 0.275 m. Finally, the continuous field was imported into a simplified finite element model, and the pile surface was divided into 1,200 quadrilateral mesh nodes. After binding the node parameters, a microscopic imaging thermogram was generated, showing the high anti-slip strength of the sandy soil zone in the middle of the pile and the gradient strength variation of the clay zone at the bottom. The simulation results showed that the critical debonding load was 12.5 kN, which deviated only 2.3% from the laboratory failure test result of 12.8 kN, validating the high-precision prediction capability of the proposed scheme in complex soil layers.

[0127] The above-mentioned 104 overall solution achieves high-precision prediction of the pull-out bearing capacity of mud-mixed piles by dynamically integrating geological data and mechanical parameters. Based on the borehole exploration data, the longitudinal segmented units are divided and the proportional coefficients of the soil layer components are calculated. The anti-slip composite parameters are generated by combining the linear superposition of sand-clay foundation parameters, and a one-dimensional anti-slip parameter mapping table is constructed. Nonlinear interpolation technology is further used to eliminate the mutation errors between discrete segments, generate a continuous anti-slip parameter field, and accurately characterize the gradient change characteristics of the soil layer transition zone. Finally, through three-dimensional spatial binding, the global anti-slip parameter distribution map is output to provide high-resolution input for the debonding dynamics model. This method breaks through the static assumptions of the traditional homogenization model, effectively solves the problems of shear force accumulation errors and nonlinear characterization of debonding extension caused by interlayer heterogeneity, significantly improves the adaptability of complex soil layer combinations and new slurry ratios, and provides reliable theoretical support and engineering decision-making basis for the anti-floating design of underground structures.

[0128] 105. Establish an interface debonding dynamics model by inputting the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics, combining the interlayer blocking effect parameters to simulate the spatially differentiated expansion process of the debonding behavior, and outputting the evolution data of the correlation between load and depth;

[0129] Optionally, step 105 may specifically include the following steps:

[0130] 1051. Based on the spatial anti-slip parameter mapping table, construct a pile-soil interface mechanical network with segmented units as nodes;

[0131] 1052. Input the verified deformation gradient characteristics into the pile-soil interface mechanical network, calculate the difference between the deformation gradient and the anti-slip composite parameter at each node, and generate the node debonding trigger criterion;

[0132] 1053. Define a resistance weight for debonding extension between adjacent nodes based on the interlayer blocking effect parameter. If the current node meets the debonding trigger criterion, calculate the probability of debonding extension to the adjacent node based on the resistance weight.

[0133] 1054. Using the load monitoring data as a time-series driving signal, iteratively update the debonding state of each node. When the debonding extends from the surface nodes to the deep nodes, record the load value of each deep node when the debonding is triggered.

[0134] 1055. Aggregate the load values ​​and corresponding depths of all nodes when debonding is triggered, and generate the evolution data of the load as the debonding depth increases.

[0135] Among them, step 1055 may specifically include the following processes: traversing all nodes in the pile-soil interface mechanical network, sorting the nodes from shallow to deep according to their depth, and extracting the load value and depth position of each node when debonding is first triggered; performing continuity check on the depth intervals between adjacent nodes, and if the depth difference between adjacent nodes exceeds a preset threshold, inserting a virtual node in the interval, and linearly distributing the load value of the preceding node to the virtual node according to the depth difference; integrating the load values ​​and depths of the actual node and the virtual node in the triggering order to generate an initial load depth sequence; detecting the presence of multiple load values ​​at the same depth in the initial load depth sequence, retaining the maximum load value and eliminating redundant data points to form a final data point; arranging the final data points in ascending order of depth to generate continuous and monotonic load evolution data as the debonding depth increases.

[0136] In the above scheme, the interlayer retardation effect parameter defines the weighted parameter that defines the resistance to debonding propagation due to compositional differences between adjacent soil layers. It includes dynamic characteristic quantities such as the resistance coefficient and energy attenuation factor, and is used to quantify the degree of resistance and difficulty of debonding across different soil layers. The pile-soil interface mechanical network is a networked mechanical model with segmented elements as nodes. Nodes store the anti-slip parameters and deformation gradient characteristic data for the corresponding segments. Edges define the soil layer transition relationship and retardation effect weights between adjacent nodes, dynamically simulating the spatial propagation path of debonding. The debonding trigger criterion is the difference threshold between the deformation gradient and the anti-slip composite parameter at a node. If the difference exceeds the critical value, debonding is triggered, indicating the mechanical imbalance of local interfacial slip. Virtual nodes are virtual data points inserted when the depth gap between adjacent actual nodes exceeds a preset threshold. The load values ​​of the preceding nodes are linearly interpolated to fill the discrete gaps in the debonding depth evolution data, ensuring the continuity and monotonicity of the load-depth curve.

[0137] In an embodiment of the present application, first, through step 1051, based on the spatial anti-slip parameter mapping table, the longitudinally continuous segmented units of the pile body are abstracted as nodes of the pile-soil interface mechanical network. Each node corresponds to a specific depth interval of the pile body, and its attributes include the anti-slip parameters of the soil layer in the segment and the soil layer type. Adjacent nodes are connected by edges, and the attributes of the edges are assigned interlayer blocking effect parameters according to the actual soil layer transition type. For example, in the sand-clay alternating layer, if the sand segment of node A is adjacent to the clay segment of node B, the blocking weight of edge AB is set to 0.8, which characterizes the resistance strength of the sand to the clay extension debonding.

[0138] Then, in step 1052, the verified deformation gradient characteristic is input into the mechanical network, and the difference between the deformation gradient and the anti-slip composite parameter is calculated node by node. If the difference at a node exceeds a preset threshold, the node is determined to meet the debonding trigger condition. For example, if the anti-slip composite parameter of node C is 0.15, the deformation gradient is 0.25% / m, and the difference is 0.1, which exceeds the threshold of 0.05, node C is marked as a debonding trigger point.

[0139] Next, the debonding extension rule is defined based on the interlayer blocking effect parameters. If debonding is triggered at the current node, the adjacent nodes are traversed and the extension probability is calculated based on the resistance weights in the edge attributes. For example, when node C extends to node D, if the blocking weight of edge CD is 0.6, the extension probability is 40%. If the difference in the anti-slip composite parameter at node D does not reach the threshold, the triggering condition is recalculated by superimposing the current load increment. This probabilistic screening determines the range of nodes to which debonding may extend in the next iteration.

[0140] Next, in step 1054, the collected load monitoring data is used as a time-series driving signal to iteratively update the state of each node in the network. For example, at the 10th second, the load increases to 150kN, triggering debonding at node C. At the 15th second, the load increases to 155kN, triggering debonding at node D due to the cumulative extension probability meeting the conditions. During each iteration, the depth of the triggering node and the corresponding load value are recorded. For example, when node C is triggered, the depth is 3m and the load is 150kN, forming a spatiotemporal trajectory of the debonding extension.

[0141] Finally, in step 1055, the load-depth data for all triggered nodes is aggregated. Virtual nodes are inserted for adjacent nodes with excessively large depth intervals, and load values ​​are assigned using linear interpolation. If low load values ​​are triggered multiple times at the same depth, these values ​​are considered redundant and removed, generating a continuous, monotonic load-depth evolution curve. For example, if the actual trigger point, node A, has a depth of 1m and a load of 100kN, and node B has a depth of 3m and a load of 140kN, a virtual node is inserted at a depth of 2m to generate an evolution sequence from 1m to 2m to 3m, fully reflecting the progressive relationship between load and debonding depth.

[0142] In practical application, a model pile with a diameter of 30 mm and a height of 1.5 m was first divided into 50 longitudinal segments based on CT scan data to construct a mechanical network for the pile-soil interface. Node attributes were dynamically assigned based on the laboratory mix soil layer: a friction coefficient of 0.25 and a cohesion of 1.2 kPa for the sand segment, and a friction coefficient of 0.4 and a cohesion of 1.8 kPa for the clay segment. Edges adjacent to nodes in the sand-clay transition zone were assigned a retardation weight of 0.7, and edges in the clay-sand transition zone were assigned a retardation weight of 0.6. For example, a node at a depth of 0.5 m was a sand segment, and its adjacent node at a depth of 0.53 m was a clay segment. The edge attribute retardation weight was set to 0.7. Data was collected using distributed optical fiber and a micro-load meter at the pile top. Deformation monitoring revealed a sudden change at a depth of 0.8 m, with a deformation increase of 0.035% / m. This change was verified and then input into the mechanical network. Calculations revealed that the anti-slip composite parameter at the 0.8 m node was 0.336 kPa·m, and the deformation gradient difference of 0.014, exceeding the threshold of 0.01, triggered debonding. As the curve traversed adjacent nodes, it expanded to the 0.77 m and 0.83 m nodes. The probability of expansion was 25% for the 0.77 m node in the sand-clay transition zone, and 35% for the 0.83 m node in the pure clay segment. In the third iteration, when the load increased to 18 kN, debonding was triggered at the 0.83 m node, recording a load of 18.2 kN. When the load was gradually increased to 21 kN, debonding extended to a depth of 1.1 m, with two virtual nodes inserted at 0.03 m intervals. The resulting evolution curve shows that the load increases from 18 kN at 0.8 m to 21 kN at 1.1 m. The slope change indicates that the resistance increase in the clay segment is 22% higher than that in the sand segment, deviating by 2.3% from the laboratory failure test result of 21.4 kN.

[0143] The overall solution described above, 105, achieves refined prediction of pullout bearing capacity in complex soil layers by constructing a pile-soil interface mechanical network and a dynamic debonding expansion model. First, a segmented unit node network is established based on a spatial anti-slip parameter mapping table. This integrates soil layer heterogeneity parameters and deformation gradient characteristics to accurately quantify the debonding triggering conditions at each node. Interlayer retardation effect parameters are used to define the weights of adjacent node expansion resistance. Combined with load time-series drive signals, the nonlinear process of debonding from local triggering to multi-layer cascade expansion is iteratively simulated, dynamically recording load-depth correlation data. When further aggregating debonding trigger point information, virtual node interpolation and redundant data elimination mechanisms are introduced to ensure the continuity and monotonicity of the evolution curve. This method breaks through the static assumptions of traditional homogenized models, fully characterizing the retardation effect in the soil transition zone and the differences in the spatial debonding expansion paths. This significantly improves the accuracy of critical load determination and provides a highly reliable dynamic evolution basis for pile foundation pullout design under complex geological conditions.

[0144] 106. Based on the convergence relationship between the evolution data and the deformation gradient characteristics, determine the maximum load threshold corresponding to the critical debonding extension state as the predicted value of the pull-out bearing capacity.

[0145] Optionally, step 106 may specifically include the following steps:

[0146] 1061. Extract the amplification slope of the load changing with the debonding depth in the evolution data, and mark the depth point where the amplification slope reverses direction for the first time as the first critical point;

[0147] 1062. Extract the deformation increase corresponding to the depth of the first critical point from the deformation gradient feature. If the deformation increase exceeds a set multiple of the historical average increase, mark it as the second critical point.

[0148] 1063. Record the depth difference between the first critical point and the second critical point. When the depth difference is less than a preset tolerance, determine that the load value corresponding to the two is the maximum load threshold of the critical extension state of debonding. When the depth difference exceeds the preset tolerance, extend the load application time until a stable point where the load increase returns to zero appears in the evolution data. Use the last load peak before the stable point as the maximum load threshold, and output the maximum load threshold as the predicted value of the pull-out bearing capacity of the pile foundation.

[0149] In the above scheme, the maximum load threshold refers to the load characteristic value reflecting the bearing capacity limit of the pile foundation under the critical extension state of debonding. It includes the peak characteristics of the load and depth evolution during the debonding extension process and the dynamic equilibrium state indicator, which can serve as the core basis for predicting pullout bearing capacity. The first critical point refers to the depth location characteristic at which the load increase slope first reverses direction in the load evolution data. It includes the spatial signature of the load transfer path mutation and the slope direction reversal signal, which is used to identify the initial triggering stage of debonding extension. The second critical point refers to the monitoring location characteristic of the deformation gradient characteristic corresponding to the depth of the first critical point where the deformation increase is significantly abnormal. It includes the intensity and direction of the deformation mutation caused by local debonding and is used to verify the spatiotemporal correlation between the load mutation and the deformation response. The stable point refers to the depth location characteristic at which the load increase returns to zero and remains stable in the load evolution data. It includes the state indicator of the debonding extension reaching dynamic equilibrium and the load attenuation convergence signal, which is used to determine the termination stage of debonding and revise the maximum load threshold.

[0150] In an embodiment of the present application, first, a continuous curve of load variation with depth is extracted from the evolution data output by the debonding dynamics model through step 1061, and the load increase slope between adjacent depth points is calculated using a sliding window difference algorithm. Specifically, the window length is set to 3 depth sampling points, the first two points in the window are used to calculate the initial slope, and the last two points are used to calculate the subsequent slope. When the first reversal of the slope direction in the window is detected, the depth of the center point of the window is marked as the first critical point. For example, if the initial slope in a window is +8kN / m and the subsequent slope is -3kN / m, the depth of 8.6m corresponding to the window is determined to be the first critical point, indicating that the load transfer path here has undergone a sudden change due to debonding triggering.

[0151] Secondly, based on the depth position of the first critical point, the deformation increase value corresponding to that depth is extracted from the deformation gradient feature. The historical average increase of the depth point is calculated through a sliding time window mean filter. If the current deformation increase exceeds the set multiple of the historical mean, it is determined to be the second critical point. For example, if the historical mean is 0.15mm / m, the current deformation increase is 0.4mm / m, and the set multiple is 2, the second critical point mark is triggered. This step verifies the correlation between load mutation and deformation mutation through spatiotemporal matching to avoid misjudgment due to local noise.

[0152] Finally, the depth difference between the first critical point and the second critical point is calculated through step 1063. If the difference is less than the preset tolerance, the load value corresponding to the two is determined to be the maximum threshold; if it exceeds the tolerance, the load loading time is extended and the evolution data is continuously monitored until a stable point where the load increase returns to zero is detected. At this time, the last load peak before the stable point is extracted as the corrected maximum threshold, and the maximum load threshold is output as the predicted value of the pull-out bearing capacity of the pile foundation. For example, if the depth of the first critical point is 8.6m and the second critical point is 9.3m, and the depth difference is 0.7m, then continue loading until the load increase approaches zero, and select the peak value of 1250kN before the stable point as the final predicted value. This process uses a dynamic iterative correction mechanism to ensure that the threshold judgment takes into account the spatiotemporal consistency of load deformation and the final equilibrium characteristics of debonding extension.

[0153] In practical applications, the load-depth curve is extracted from the evolution data output by the debonding dynamics model at a sampling interval of 0.02 m. The slope is calculated using a sliding window difference algorithm with a window length of 5 points. For example, in the depth range of 0.84-0.90 m, the loads corresponding to the first two points at 0.84 m and 0.86 m are 10.5 kN and 11.2 kN, respectively, and the calculated initial slope is +35 kN / m. The loads corresponding to the second two points at 0.88 m and 0.90 m are 11.0 kN and 10.7 kN, respectively, and the calculated subsequent slope is -15 kN / m. The slope direction changes from positive to negative. The center point of the window at 0.86 m is determined to be the first critical point, indicating that the debonding of the clay layer triggers a sudden change in the load transfer path. The deformation gradient feature then extracted deformation increase data at a depth of 0.86m. Using a sliding time window mean filter, the historical average deformation increase at this point was calculated to be 0.018mm / m, with a window length of 50 groups as the monitoring period. The deformation increase during the current monitoring period reached 0.052mm / m, exceeding the threshold of 0.036mm / m by two times, triggering the second critical point marker. This validated the spatiotemporal correlation between load mutations and deformation mutations and eliminated misjudgments caused by equipment vibration and noise. The depth difference between the first and second critical points was then calculated to be 0.03m, less than the preset tolerance of 0.05m. The load value of 11.5kN corresponding to both was directly determined to be the maximum threshold, which was then output as the predicted pullout bearing capacity of the pile foundation. In another set of tests, the first critical point was 0.92m, corresponding to a deformation increase of 0.048mm / m. The second critical point was marked at 0.99m. The depth difference of 0.07m exceeded the tolerance. The system automatically extended the load loading time by 1.5 minutes, and monitored that the load increase returned to zero at a depth of 1.08m. The peak value of 12.8kN before the stability point was extracted as the corrected maximum threshold, and output it as the predicted value of the pull-out bearing capacity of the pile foundation.

[0154] The overall approach described above, 106, achieves precise determination of the critical extension state of debonding through the coordinated analysis of load evolution data and deformation gradient characteristics. First, the first critical point is identified based on the reversal of the load increase slope direction, locating the initial mutation location triggering debonding. Second, the second critical point is verified by comparing the deformation increase with the historical mean threshold, ensuring the spatiotemporal correlation of the load-deformation mutation. Finally, a depth difference tolerance determination and a stable point iterative correction mechanism are combined to dynamically adapt to the nonlinear process of debonding extension. This method breaks through the limitations of traditional models that rely on static superposition and empirical coefficients. Through multi-source data fusion and dynamic parameter correction, it effectively characterizes the blocking effect of interlayer heterogeneity on debonding behavior, avoiding the error accumulation caused by the homogenization assumption. Furthermore, it fully restores the dynamic path of debonding from local triggering to multi-stage extension, accurately capturing the critical state mutation characteristics, significantly improving the adaptability to complex soil layer combinations and new slurry ratios. Ultimately, it outputs a pullout bearing capacity prediction value that is both theoretically reasonable and engineering-verifiable, providing highly reliable support for the anti-floating design of underground structures.

[0155] The following is a complete example for steps 101 to 106. Figure 2 As shown, a laboratory-prepared composite soil sample with alternating sand and clay layers was constructed using a precision layered compaction device with an accuracy of ±0.1 mm. Cylindrical specimens with a diameter of 200 mm and a height of 400 mm were constructed. Sand and clay layers were laid alternately at 30 mm thicknesses. Four cement-bentonite slurries with varying ratios were injected via a high-pressure grouting system to form simulated solidified piles with a diameter of 30 mm. A laser profilometer was used to scan the solidified interface in three dimensions, and geometric parameters such as interface relief height, effective contact area ratio, and fractal dimension were extracted using image processing algorithms. Simultaneous microfocus CT scans revealed that at a water-to-solid ratio of 1.0, the slurry penetrated the sand layer to a depth of 22 mm, while the clay layer, due to porosity limitations, only penetrated 2.8 mm, reaching a peak contact area ratio of 82% at this point.

[0156] During the interfacial mechanical property testing phase, composite specimens were subjected to multi-mode loading using an electronic universal testing machine. During the tensile test, an axial load was applied at a rate of 0.1 mm / s. A laser displacement sensor was used at a sampling rate of 100 Hz to capture a bimodal load-displacement curve. The first peak of the curve, 0.5-1.8 MPa, corresponds to the bond failure at the clay-slurry interface, while the second peak, 1.2-2.6 MPa, reflects the frictional energy dissipation at the sand-slurry interface. Based on the Mohr-Coulomb criterion, the sliding resistance curves during the post-separation phase calculated the friction angle of the sand-slurry interface to be 29°-34°, significantly higher than the friction angle range of the clay interface (18°-22%). The established mechanical database contains 80 sets of structured data, each associated with five geometric parameters: interface relief height, contact area ratio, fractal dimension, penetration depth, and penetration uniformity; three mechanical parameters: tensile strength, friction angle, and residual strength; and two material ratio parameters: water-to-solid ratio and bentonite content. Data is stored in JSON format, enabling rapid retrieval of multi-dimensional parameters.

[0157] The deformation monitoring system consists of a 32-channel strain gauge networked with LVDT displacement sensors, with strain gauge arrays arranged at 50mm intervals on the surface of the simulated pile. When an axial load of 12kN was applied, wavelet transform detected a sudden strain change in the 120-150mm depth region. The axial strain gradient reached 0.08% / mm, 3.2 times the baseline value, and the radial shrinkage rate increased by 1.8 times. A dynamic time warping algorithm was used to align the load and deformation time series data, eliminating three groups of abnormal data segments caused by sensor temperature drift. Ultimately, a calibration data set with a correlation coefficient greater than 0.91 was obtained. Based on the layered soil model reconstructed from CT scans, the anti-slip parameters of each depth unit were calculated using a volume weighted method.

[0158] When establishing the discretized pile-soil interface model, the 400mm pile length was divided into 40 nodes with a spacing of 10mm, and each node was assigned an initial anti-slip strength. When the local shear stress was greater than or equal to the initial anti-slip strength, the debonding state was triggered. Numerical simulations showed that the initial debonding occurred at a shallow node with a depth of 20mm, and then expanded downward at a rate of 0.5mm / kN. However, it encountered resistance in the clay-dominated layer, that is, at a depth of 90-120mm, and an additional 4kN load was required to break through the interface resistance. By inserting virtual nodes to increase the spatial resolution to 2mm, the load depth evolution curve obtained showed a three-stage characteristic: a linear elastic section, a nonlinear transition section, and a critical instability section. When the load reached 22.3 kN, an inflection point in the evolution curve and a sudden increase in strain of 0.12% / mm appeared simultaneously at a depth of 85 mm. The spatial deviation between the two critical points was only 0.6 mm, which was less than the preset tolerance. Based on this, the pull-out bearing capacity threshold was determined to be 22.3 kN, which was within 3.5% of the actual measured value of 23.1 kN in the physical test.

[0159] To validate the reliability of the method, five repeated tests were conducted on specimens with the same mix ratio, resulting in a standard deviation of 0.7 kN in the measured bearing capacity. Furthermore, the correlation coefficient between the interface fractal dimension and the debonding trigger load reached 0.89. Sensitivity analysis showed that when the anti-slip parameter error was ±10%, the predicted bearing capacity fluctuated by ±8%. Therefore, the quality of the input parameters needs to be controlled through CT scans with an error of less than ±2%.

[0160] Figure 3 The present invention provides a schematic diagram of a system for predicting the pull-out bearing capacity of a slurry mixed pile. Figure 3 As shown, the system includes:

[0161] An acquisition module 31 is used to obtain geometric characteristic parameters of the interface structure formed by the solidified slurry in the multi-layer heterogeneous soil layer at different ratios;

[0162] The acquisition module 31 further includes determining the interface mechanical parameters of the multi-layer heterogeneous soil layer through mechanical tests, and constructing an interface mechanical database in combination with the geometric characteristic parameters;

[0163] Verification module 32 collects pile deformation monitoring data and load monitoring data, generates deformation gradient features by identifying deformation mutation thresholds, and performs collaborative verification on the load monitoring data and deformation gradient features;

[0164] A generation module 33 divides the pile into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layer, adjusts the anti-slip parameters in the interface mechanics database based on the proportion of soil layer components in each segmented unit, and generates a spatial anti-slip parameter mapping table;

[0165] Evolution module 34 establishes an interface debonding dynamics model by inputting the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics, combining the interlayer blocking effect parameters to simulate the spatially differentiated expansion process of the debonding behavior, and outputting evolution data related to load and depth;

[0166] The prediction module 35 determines the maximum load threshold corresponding to the critical extension state of debonding as the predicted value of the pull-out bearing capacity based on the convergence relationship between the evolution data and the deformation gradient characteristics.

[0167] Figure 3 The prediction system of the pull-out bearing capacity of the mud mixed pile can be executed Figure 1 The implementation principle and technical effects of the method for predicting the pull-out bearing capacity of a slurry-mixed pile described in the illustrated embodiment will not be elaborated on here. The specific manner in which each module and unit performs operations in the system for predicting the pull-out bearing capacity of a slurry-mixed pile in the above embodiment has been described in detail in the embodiments of the method and will not be elaborated on here.

[0168] In one possible design, Figure 3 The prediction system of the pull-out bearing capacity of a slurry mixed pile in the embodiment shown can be implemented as a computing device, such as Figure 4 As shown, the computing device may include a storage component 41 and a processing component 42;

[0169] The storage component 41 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 42 .

[0170] The processing component 42 is used for the above Figure 1 The embodiment provides a method for predicting the pull-out bearing capacity of a slurry mixed pile.

[0171] The processing component 42 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0172] The storage component 41 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0173] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0174] The input / output interface provides an interface between the processing component and the peripheral interface module, which can be an output device, an input device, etc.

[0175] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0176] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0177] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The embodiment shown is a method for predicting the pull-out bearing capacity of a slurry mixed pile.

[0178] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0179] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0180] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0181] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for predicting the pull-out bearing capacity of a slurry mixed pile, characterized in that: include: Obtain the geometric characteristic parameters of the interface structure formed by solidified slurry in multi-layer heterogeneous soil layers at different ratios; Determine the interface mechanical parameters of the multi-layer heterogeneous soil layer through mechanical tests, and construct an interface mechanical database in combination with the geometric characteristic parameters; Collecting pile deformation monitoring data and load monitoring data, generating deformation gradient features by identifying deformation mutation thresholds, and collaboratively verifying the load monitoring data and deformation gradient features; Dividing the pile into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layer, adjusting the anti-slip parameters in the interface mechanics database based on the proportion of soil layer components in each segmented unit, and generating a spatial anti-slip parameter mapping table; Establishing an interface debonding dynamics model, by inputting the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics, combining the interlayer blocking effect parameters to simulate the spatially differentiated expansion process of debonding behavior, and outputting evolution data related to load and depth; Based on the convergence relationship between the evolution data and the deformation gradient characteristics, the maximum load threshold corresponding to the critical extension state of debonding is determined as the prediction value of the pull-out bearing capacity.

2. The method according to claim 1, characterized in that Collecting pile deformation monitoring data and load monitoring data, generating deformation gradient features by identifying deformation mutation thresholds, and collaboratively verifying the load monitoring data and deformation gradient features, including: Synchronously collect multi-dimensional deformation monitoring data of the pile surface and load monitoring data of the pile top at fixed time intervals; Performing continuous spatial scanning on the deformation monitoring data, locating adjacent areas where deformation values ​​suddenly change as mutation areas, extracting deformation amplitudes and direction change angles in the mutation areas, and generating a set of deformation mutation thresholds; According to the amplification extreme value points in the deformation mutation threshold set, the deformation gradient intervals are divided along the longitudinal direction of the pile body, and the rate of change of the deformation amplification with depth in each deformation gradient interval is calculated to generate the deformation gradient feature; Performing event synchronization matching on the load monitoring data according to the acquisition timestamp and the generation timestamp of the deformation gradient feature, and screening out the synchronous change interval of the load increase and the deformation gradient increase; The load data and deformation gradient characteristics within the synchronous change interval are verified for amplitude ratio. If the ratio of the load increase to the deformation gradient increase exceeds the preset fluctuation range, the abnormal data segment is eliminated and the verified deformation gradient characteristics are output.

3. The method according to claim 1, characterized in that The pile body is divided into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layer. The anti-slip parameters in the interface mechanics database are adjusted based on the proportion of soil layer components in each segmented unit to generate a spatial anti-slip parameter mapping table, including: Extracting the longitudinal distribution profile of the soil layer around the pile based on the drilling exploration data, dividing the pile body into multiple concentric ring-shaped segmented units at preset depth intervals, calculating the volume ratio of sand and clay within each segmented unit, and converting the volume ratio into a soil layer component ratio coefficient; Extracting foundation anti-slip parameters corresponding to sand and clay from the interface mechanics database, linearly superimposing the foundation anti-slip parameters corresponding to sand and clay according to the soil layer component ratio coefficient, and generating anti-slip composite parameters adapted to the current segmented unit; Arranging the depth position of each segmented unit and the corresponding anti-slip composite parameter in a vertical order to form a one-dimensional anti-slip parameter mapping table with depth as the index; Based on the one-dimensional anti-slip parameter mapping table, the anti-slip composite parameters of adjacent segmented units are nonlinearly interpolated according to the soil layer transition gradient to generate a continuous anti-slip parameter field; The continuous anti-slip parameter field is spatially bound to the three-dimensional geometric model of the pile body, and a global anti-slip parameter distribution map of the pile-soil interface is output as a spatial anti-slip parameter mapping table.

4. The method according to claim 1, wherein Establish an interface debonding dynamics model. By inputting the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics, combined with the interlayer blocking effect parameters, simulate the spatially differentiated expansion process of the debonding behavior and output the evolution data of the load-depth correlation, including: Based on the spatial anti-slip parameter mapping table, a pile-soil interface mechanical network with segmented units as nodes is constructed; Inputting the verified deformation gradient characteristics into the pile-soil interface mechanical network, calculating the difference between the deformation gradient and the anti-slip composite parameter at each node, and generating the node debonding trigger criterion; Based on the interlayer blocking effect parameters, the resistance weight of the debonding extension between adjacent nodes is defined. If the current node meets the debonding trigger criterion, the probability of the debonding extending to the adjacent node is calculated according to the resistance weight. Using the load monitoring data as a time-series driving signal, the debonding state of each node is iteratively updated. When the debonding extends from the surface nodes to the deep nodes, the load value of each deep node when the debonding is triggered is recorded; The load values ​​and corresponding depths at the time of debonding triggering of all nodes are aggregated to generate the evolution data of the load as the debonding depth increases.

5. The method according to claim 4, characterized in that Aggregate the load values ​​and corresponding depths at all nodes when debonding is triggered, and generate load evolution data as debonding depth increases, including: Traversing all nodes in the pile-soil interface mechanical network, sorting the nodes from shallow to deep according to their depth, and extracting the load value and depth position when debonding is first triggered at each node; The depth intervals between adjacent nodes are checked for continuity. If the depth difference between adjacent nodes exceeds a preset threshold, a virtual node is inserted into the interval and the load value of the previous node is linearly distributed to the virtual node according to the depth difference. Integrating the load values ​​and depths of the actual nodes and the virtual nodes in a triggering order to generate an initial load depth sequence; Detecting the presence of multiple load values ​​at the same depth in the initial load depth sequence, retaining the maximum load value and eliminating redundant data points to form a final data point; The final data points are arranged in order of increasing depth to generate continuous and monotonic load evolution data with increasing debonding depth.

6. The method according to claim 1, characterized in that Based on the convergence relationship between the evolution data and the deformation gradient characteristics, the maximum load threshold corresponding to the critical extension state of debonding is determined as the predicted value of the pull-out bearing capacity, including: Extracting the amplification slope of the load changing with the debonding depth from the evolution data, and marking the depth point where the amplification slope reverses direction for the first time as the first critical point; Extracting the deformation increase corresponding to the depth of the first critical point from the deformation gradient feature, and marking it as the second critical point if the deformation increase exceeds a set multiple of the historical average increase; The depth difference between the first critical point and the second critical point is recorded. When the depth difference is less than a preset tolerance, the load value corresponding to the two is determined to be the maximum load threshold of the critical extension state of debonding. When the depth difference exceeds the preset tolerance, the load application time is extended until a stable point where the load increase returns to zero appears in the evolution data. The last load peak before the stable point is used as the maximum load threshold, and the maximum load threshold is output as the predicted value of the pull-out bearing capacity of the pile foundation.

7. The method according to claim 1, characterized in that The interface mechanical parameters of the multi-layer heterogeneous soil layers are measured by mechanical tests, and an interface mechanical database is constructed in combination with the geometric characteristic parameters, including: preparing a composite soil sample containing alternating layers of sand and clay, and embedding a solidifying slurry into the composite soil sample to form a simulated pile-soil interface; Applying multi-level tensile loads to the simulated pile-soil interface, monitoring the load value and displacement during the interface separation process, and extracting the peak point of the load-displacement curve as the interface tensile strength; After the interface is completely separated, the friction sliding resistance of the separation surface is measured, and the interface friction angle is calculated based on the relationship between the sliding displacement and the friction sliding resistance; Extracting the surface undulation height and contact area ratio of the simulated pile-soil interface to generate a set of geometric feature parameters; Under each set of ratios, the interface tensile strength and interface friction angle set are parameter-matched to establish interface mechanical parameters indexed by the ratio number; Aggregate the geometric feature parameter sets and the interface mechanical parameters under all ratio numbers to generate an interface mechanical database containing the ratio, geometry and mechanical mapping relationship.

8. A prediction system for the pull-out bearing capacity of a slurry-mixed pile, characterized in that: include: An acquisition module is used to obtain the geometric characteristic parameters of the interface structure formed by the solidified slurry in multiple heterogeneous soil layers at different ratios; The acquisition module further comprises determining the interface mechanical parameters of the multi-layer heterogeneous soil layer through mechanical testing, and constructing an interface mechanical database in combination with the geometric characteristic parameters; A verification module collects pile deformation monitoring data and load monitoring data, generates deformation gradient features by identifying deformation mutation thresholds, and performs collaborative verification on the load monitoring data and deformation gradient features; a generation module that divides the pile into longitudinally continuous segmented units according to the spatial distribution characteristics of the soil layer, adjusts the anti-slip parameters in the interface mechanics database based on the proportion of soil layer components in each segmented unit, and generates a spatial anti-slip parameter mapping table; The evolution module establishes an interface debonding dynamics model, inputs the spatial anti-slip parameter mapping table and the verified deformation gradient characteristics, combines the interlayer blocking effect parameters to simulate the spatially differentiated expansion process of the debonding behavior, and outputs the evolution data associated with the load and depth; The prediction module determines, based on the convergence relationship between the evolution data and the deformation gradient characteristics, a maximum load threshold corresponding to the critical extension state of debonding as a predicted value of the pull-out bearing capacity.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for predicting the pull-out bearing capacity of a mud mixed pile as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the method for predicting the pull-out bearing capacity of a slurry mixed pile according to any one of claims 1 to 7 is implemented.

Images