Optimization method for hole filling process of steel sheet pile based on 3D printing and numerical simulation

Abstract

The application belongs to the technical field of foundation pit construction, and particularly relates to a steel sheet pile hole filling process optimization method based on 3D printing and numerical simulation. In view of the problems that the internal form of the hole is difficult to detect, the filling is blind and the effect is poor in the prior art, the application replicates the three-dimensional form of the hole by pouring a solidifiable material in the real hole, and obtains point cloud data through three-dimensional scanning after taking out; a numerical model is constructed based on the data and drives 3D printing to prepare a high-fidelity physical model; a filling test is carried out on the physical model and parameters such as compactness are detected, and a corresponding numerical model is established on a platform such as PFC, and the simulation model is quantitatively calibrated by using the test results to form a high-credibility digital twin model; finally, the model is used to efficiently carry out a multi-factor virtual proportioning test to quickly determine an optimal filling scheme.

Description

Technical Field

[0001] This application relates to the field of foundation pit construction technology, specifically to an optimization method for filling the extraction hole of steel sheet piles based on 3D printing and numerical simulation. Background Technology

[0002] Before excavation of the foundation pit, steel sheet piles are inserted into the soil around the pit to form a ground support structure. Steel sheet pile support is a common method for foundation pit support, and it is also widely used in underground utility tunnels and tunnel support.

[0003] In actual construction, sheet piles should be extracted and transferred promptly after the support is completed. During the extraction process, due to the strong adhesion of the hard soil layer to the sheet piles, extraction holes will be formed at the extraction location. These extraction holes need to be filled in a timely manner to avoid collapse accidents.

[0004] However, the extraction holes are distributed along the entire length of the sheet pile, with a narrow space and a length of up to twenty meters, making physical morphology detection extremely difficult. Therefore, most of the filling process for extraction holes is essentially a "blind" filling, lacking theoretical understanding of the internal morphology of the extraction holes.

[0005] Currently, the main methods for treating pull-out holes include the filling method and the bottom-up method. Common filling materials include sand and gravel, and cement grout. However, there are many considerations during the filling process, such as whether the soil bearing capacity is calculated correctly during grouting, whether the grout mix ratio is appropriate, and whether the sand and gravel particle gradation is appropriate.

[0006] Numerical simulation technology can provide new insights into the process of extracting holes. By establishing a soil-structure interaction model, it can simulate the deformation and stability evolution of holes under different filling materials, grouting pressures, and other parameters, aiding in the initial screening of process parameters. However, the reliability of numerical simulation results highly depends on the matching degree of the constitutive model and the rationality of the boundary conditions. In actual engineering, the spatial variability of soil parameters and complex factors such as construction disturbances are difficult to fully quantify and characterize, making their application in engineering only theoretically possible. The application of 3D printing technology in the field of civil engineering is gradually expanding, but the traditional combination of 3D printing and numerical simulation is mostly one-way verification and has not yet formed a systematic closed loop, naturally failing to solve the systemic problems in the treatment of extracting holes. Summary of the Invention

[0007] This application provides an optimization method for filling the extraction hole of steel sheet piles based on 3D printing and numerical simulation, which organically combines 3D printing technology with numerical simulation models to systematically solve the problems existing in hole treatment.

[0008] The technical solution of this application is as follows: An optimization method for filling the extraction cavity of steel sheet piles based on 3D printing and numerical simulation includes the following steps: Step 1: Remove the sheet pile to form a pull-out hole, fill the pull-out hole with curing material, and after the curing material has solidified, remove it and scan its surface to obtain the three-dimensional data of the solidified material. Step 2: Based on the three-dimensional data of the solidified material, establish a numerical model of the pull-out hole, input the numerical model into the 3D printing system, and print the solid model of the pull-out hole; Step 3: Fill the solid model according to the preset filling scheme of the hole to conduct the test, and check the test parameters of the filling after the test; Step 4: Establish a three-dimensional numerical model based on the numerical model and simulate the preset filling scheme. Calibrate the three-dimensional numerical model according to the test results of the physical model. After calibration, obtain the digital twin model of the hole. Step 5: Conduct multi-condition, multi-parameter, and multi-factor ratio simulation experiments based on the digital twin model to obtain optimized parameters for the hole removal process.

[0009] Furthermore, the optimization methods also include: Step 6: Using the density of the hole filling as the target parameter, perform variation source analysis on the target parameter, construct the factor sensitivity index and the relative deviation evaluation index between levels, and establish an optimized parameter combination based on the factor sensitivity index and the relative deviation evaluation index between levels.

[0010] Furthermore, in step three, the test parameters of the filler are obtained based on non-destructive testing and / or destructive dissection methods, respectively. The test parameters include overall density, defect location, and contact state with the pull-out hole wall.

[0011] Furthermore, in step four, multiple sets of cross-sectional size samples are selected in the solid model and the three-dimensional numerical model, and multiple sets of parallel samples are selected for each cross-section. Consolidated undrained shear tests under multi-level normal stress are carried out to obtain the shear strength curves and shear strength parameters of the solid model and the three-dimensional numerical model. The shear strength parameters include cohesion and internal friction angle.

[0012] Furthermore, in step four, a two-way verification is performed on the solid model and the 3D numerical model based on the Coulomb-Moore theoretical curve to improve the credibility of the digital twin model. The two-way verification specifically includes the following sub-steps: Step 1: Perform cohesion deviation rate verification and internal friction angle deviation rate verification. ; ; In the formula, Indicates the cohesion deviation rate. Indicates the deviation rate of the internal friction angle. Represent the cohesion and internal friction angle of the three-dimensional numerical model under the experiment, respectively. These represent the cohesion and internal friction angle of the solid model under test, respectively; Step 2: For the shear stress-normal stress curves of the 3D numerical model and the solid model, calculate the shear stress deviation at the corresponding normal stress level. : ; In the formula, This represents the shear stress in a three-dimensional numerical model. Represents the shear stress of the solid model. Indicates the normal stress level number. Indicates the number of samples; Step 3: Constructing an overall curve similarity evaluation index : ; In the formula, m represents the number of normal stress levels, and n represents the number of samples.

[0013] The ratio of the area under the curves of the solid model and the 3D numerical model is: ; In the formula, This represents the area of ​​the curve in the three-dimensional numerical model. The area of ​​the curve in the solid model; Step 4: Set the verification threshold. When all verification indicators meet the corresponding verification thresholds, the current 3D numerical model is determined to be a digital twin model.

[0014] Furthermore, in step four, when performing bidirectional verification of the solid model and the three-dimensional numerical model, the multiple sets of cross sections are uniformly selected along the axial direction of the pull-out hole, and the number of multiple sets of parallel samples taken for each cross section is not less than 3 sets.

[0015] Furthermore, in step five, the multi-factor proportioning test includes test factors such as material proportions, construction techniques, and environmental conditions.

[0016] Furthermore, in step five, when the filling material for the pulled-out hole is cement grout, the optimized parameters for the pulled-out hole treatment process are grout ratio, filling speed, filling process, and grouting pressure. When the filling material for the extraction hole is sand and gravel, the optimized parameters for the extraction hole treatment process are particle size distribution, filling speed, filling process, and filling pressure.

[0017] Furthermore, in step one, the curing material is bisphenol A type epoxy resin.

[0018] Furthermore, in step six, after obtaining the filling density data under each test condition through multi-factor ratio experiment simulation, the average filling density of each test factor at different levels is calculated, and the level with the largest average value is selected as the preferred level of the test factor; among them, the test factor with the most significant influence is determined by the one with the largest corresponding factor sensitivity index F value.

[0019] Due to the adoption of the above technical solution, the beneficial effects of this application are as follows: 1. This application does not simply use 3D printed entities to verify numerical simulation results. Instead, it systematically obtains the shear strength parameters of the physical model and the 3D numerical model under different normal stresses by conducting consolidated undrained shear tests on multiple axial sections. Several quantitative verifications are then performed based on Coulomb-Moore theory, including cohesion deviation rate, internal friction angle deviation rate, shear stress deviation under progressive normal stress, and the ratio of the overall shear strength curve area. This closed-loop verification system ensures that the 3D numerical model is highly consistent with the actual hole geometry and material behavior at the mechanical response level, thereby generating a digital twin model that can be used for engineering decision-making. Compared to traditional simulation methods that rely solely on empirical parameters or single-point verification, the 3D numerical model used in this application for multi-condition simulation has higher reliability.

[0020] 2. This application utilizes a calibrated digital twin model to systematically conduct multi-dimensional coupled mix design simulations in a virtual environment, encompassing material proportions, construction techniques, and environmental conditions. The highly reliable three-dimensional numerical model eliminates the need for repeated physical model creation or on-site trial and error, enabling rapid analysis of stability and density evolution under various working conditions. Traditional "blind filling" methods cannot predict grout diffusion paths or sand and gravel settling voids. For pull-out holes exceeding twenty meters in depth, with confined spaces and irregular shapes, this method uses high-fidelity simulation to reveal the formation mechanism of filling defects in advance, significantly reducing rework risks and improving filling efficiency.

[0021] 3. Step six of this application uses filling density as the core target parameter and employs analysis of variance to calculate the sensitivity index (F-value) and relative deviation between levels for each experimental factor. This not only identifies the key factors that have the most significant impact on density (those with the largest F-value) but also determines the optimal level for each factor (corresponding to the highest average density), thus establishing a complete optimization chain from data-driven decision output. For example, when it is found that the "filling speed" is much more sensitive to density than "ambient temperature," the construction party can prioritize controlling the grouting rate rather than investing resources in temperature control measures, achieving precise resource allocation and lean process optimization. Attached Figure Description

[0022] The accompanying drawings, which are provided to further illustrate this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application.

[0023] Figure 1 Flowchart of the optimized method for filling the extraction hole of sheet pile based on 3D printing and numerical simulation provided for this application; Figure 2 This is a schematic diagram of the curing material filling process in an embodiment of this application; Figure 3 This is a schematic diagram of the cured material after molding according to an embodiment of this application; Figure 4 This is a schematic diagram of a three-dimensional numerical model of the hole in an embodiment of this application; Figure 5 This is a diagram showing the nonlinear shear strength of an embodiment of this application. Detailed Implementation

[0024] Based on the background technology described, as shown in the appendix Figure 1 As shown, this application provides an optimization method for filling the extraction hole of steel sheet piles based on 3D printing and numerical simulation, including the following steps: Step 1: Remove the sheet pile to form a pull-out hole, fill the pull-out hole with curing material, and after the curing material has solidified, remove it and scan its surface to obtain the three-dimensional data of the solidified material.

[0025] As attached Figure 2 As shown, after the bottom sheet piles are removed on site, a pre-set pipe is inserted into the extraction hole, and a filling and solidification material is poured in from bottom to top, so that it is constrained and formed in situ inside the extraction hole, thereby replicating the hole space structure to provide a sample for the subsequent construction of a high-precision physical model.

[0026] In practice, sheet pile extraction typically employs a vibration method: first, a vibratory hammer is used to vibrate the sheet pile interlocks to loosen them and reduce the resistance of the surrounding soil. Then, the sheet pile is gradually extracted while continuously vibrating. If a sheet pile is difficult to extract, a diesel hammer can be used to drive it down 100-300mm, followed by alternating vibration and extraction with the vibratory hammer. When the sheet pile is pulled slightly above the foundation slab, extraction is paused, and the vibratory hammer is used to vibrate continuously for several minutes to fill as much of the surrounding soil as possible into the cavity. It should be noted that filling the cavity does not affect the accuracy of the subsequent solid model. The cavity morphology includes the irregular three-dimensional structure formed during sheet pile extraction due to vibration disturbance, soil stress release, and local backfilling. This morphology accurately reflects the actual boundary conditions required for subsequent grouting processes.

[0027] For bottom-up pouring, the curing material can be resin-based, rubber-based, or cement grout. Resin-based materials are preferred due to their lower viscosity, higher fluidity, smaller curing shrinkage, and superior mechanical properties compared to rubber-based materials and cement grout. In specific implementation, bisphenol A epoxy resin, unsaturated polyester resin, and phenolic resin can be selected as resin-based materials. Among these, bisphenol A epoxy resin has a more mature process and is more operable. The bisphenol A epoxy resin pouring process is as follows: The curing agent is slowly added to the resin while stirring for 3-5 minutes to ensure uniform mixing. The stirring speed should not be too fast to avoid generating too many air bubbles. The mixed resin is then poured into a vacuum degassing device and degassed under a vacuum of -0.1 MPa for 5-10 minutes until all air bubbles disappear. Pouring is then carried out at a slow, uniform rate of 0.5-1.0 L / min to ensure the resin fully penetrates every corner of the pores. The pouring pressure is controlled at 0.1-0.3 MPa; the pressure should not be too high to avoid damaging the pore structure. After pouring, pre-cur at room temperature for 2-4 hours to allow the resin to initially gel. Then, raise the temperature to 60-80℃ for post-curing for 4-8 hours to ensure complete curing. After curing, allow it to cool naturally to room temperature to avoid rapid cooling and the generation of internal stress. Once the resin is fully cured, carefully remove the pouring mold. Perform surface treatments such as grinding and polishing on the poured body to obtain a smooth surface. After the resin has cured, use a vibration method to pull out the cured bisphenol A epoxy resin.

[0028] After removing the cured material, a 3D laser scanner is used to perform high-density point cloud sampling on its surface to obtain complete 3D surface point cloud data. Specifically, this includes: attaching reflective markers to the surface of the cured material; powering on the 3D laser scanner and connecting it to the computer, setting necessary scanning parameters such as scanning accuracy and scanning spacing; scanning the reflective markers and the surface of the cured material; and saving the 3D surface point cloud scan data of the cavity filling.

[0029] Numerical modeling was performed based on the 3D surface point cloud scanning data of the cavity filling body: The scanned point cloud data was imported into the post-processing software Geomagic Studio, where discontinuous points and other noise were removed, and the point cloud was encapsulated into a triangular mesh surface; the irregular model boundary was trimmed to obtain a clear and regular study area; using the "plane best fit" function, a dominant plane of the model was fitted to the standard coordinate system plane of the software, and the origin of the coordinate system was set at the center of the fitted plane; starting from the center of the coordinate system, the trimming area was gradually expanded from the inside out according to the set initial size and fixed step size; after each trimming, the fitted plane of the retained part needed to be recalibrated to ensure the consistency of the data benchmark, and then the model file at this size was saved; based on the above progressive sampling method, a complete sample file of the removed cavity filling body surface was finally obtained.

[0030] Step Two: As attached Figure 3As shown, a numerical model of the pull-out hole is established based on the three-dimensional data of the cured material. This numerical model is then input into a 3D printing system to print a solid model of the pull-out hole. By performing Boolean operations in software, the numerical model of the pull-out hole can be replicated based on the three-dimensional data of the cured material. It should be explained that the technical solution of this application simulates the actual pull-out hole through the geometric shape of the three-dimensional solid model, thereby simulating the actual boundary conditions of the filling material. Subsequently, the actual physical parameters of the filler under this filling scheme can be obtained through non-destructive and destructive testing methods.

[0031] Step 3: Fill the solid model according to the preset filling scheme of the hole to conduct the test, and check the test parameters of the filling after the test; It should be noted that the test parameters of the filler are obtained based on non-destructive testing (NDT) and / or destructive dissection methods. These parameters include overall density, defect location, and contact state with the pull-out hole wall. NDT methods utilize ultrasonic CT to assess the density, distribution of internal voids and cracks, contact with the hole wall, and overall integrity of the solid model. Destructive dissection methods precisely measure the contact surface thickness, material penetration range, and verify NDT results. They allow for direct observation of the internal structure, offer high precision, and allow for sampling for material performance testing. The destructive testing process is as follows: A cutting machine is used to cut along the hole axis or perpendicular to the axis, with water cooling during the cutting process; the cut surface is directly observed and recorded, primarily focusing on the filler's overall integrity, defect distribution, and interface contact state; after sampling the filler section, a uniaxial compressive strength test or a direct shear test is performed to obtain the mechanical parameters of the cured filler. It should be explained that destructive dissection methods are more reliable than NDT methods and can be used to correct NDT results.

[0032] Step 4: Establish a three-dimensional numerical model based on the numerical model and simulate the preset filling scheme. Calibrate the three-dimensional numerical model according to the test results of the physical model. After calibration, obtain a digital twin model of the hole.

[0033] As attached Figure 4 As shown, in practical implementation, the numerical model from step two is imported into the PFC software to generate a three-dimensional numerical model in PCF. It should be explained that in step four, the preset filling scheme simulated based on the three-dimensional numerical model refers to the filling scheme in the actual construction process, with the aim of reproducing the filling process in the simulation model.

[0034] In step four, multiple sets of cross-sectional dimension samples are selected in both the solid model and the three-dimensional numerical model. For each cross-section, multiple sets of parallel samples are selected. Consolidated undrained shear tests under multi-level normal stress are then conducted to obtain the shear strength curves and shear strength parameters for both the solid model and the three-dimensional numerical model. The shear strength parameters include cohesion and the angle of internal friction. It should be noted that in step four, when performing bidirectional verification of the solid model and the three-dimensional numerical model, the multiple sets of cross-sections are uniformly selected along the axial direction of the pull-out hole, and the number of parallel samples taken for each cross-section is no less than three sets. Parallel samples refer to repeated samples obtained from different locations on the same model or through repeated tests.

[0035] The solid model and the 3D numerical model are verified bidirectionally based on the Coulomb-Moore theory curve, which includes the following steps: Step 1: Perform cohesion deviation rate verification and internal friction angle deviation rate verification. ; ; In the formula, Indicates the cohesion deviation rate. Indicates the deviation rate of the internal friction angle. Represent the cohesion and internal friction angle of the three-dimensional numerical model under the experiment, respectively. These represent the cohesion and internal friction angle of the solid model under test, respectively; Step 2: For the shear stress-normal stress curves of the 3D numerical model and the solid model, calculate the shear stress deviation at the corresponding normal stress level. : ; In the formula, This represents the shear stress in a three-dimensional numerical model. Represents the shear stress of the solid model. Indicates the normal stress level number. Indicates the number of samples; Step 3: Constructing an overall curve similarity evaluation index : ; In the formula, m represents the number of normal stress levels, and n represents the number of samples.

[0036] The ratio of the area under the curves of the solid model and the 3D numerical model is: ; In the formula, This represents the area of ​​the curve in the three-dimensional numerical model. The area of ​​the curve in the solid model; Step 4: Set the verification threshold. When all verification indicators meet the corresponding verification thresholds, the current 3D numerical model is determined to be a digital twin model.

[0037] In practical implementation, the verification threshold is set to <5% and <3%; <6%; ≥94% and For a twin digital model to be ∈[95%, 105%], it must meet all of the above verification thresholds. The following explains the basis for setting the verification thresholds: <5%: Small changes in cohesion have a significant impact on the stability of shallow soil. A deviation rate of 5% is an acceptable critical value for engineering projects, ensuring that the strength prediction does not deviate from the actual working conditions.

[0038] <3%: The internal friction angle is more sensitive to the shear strength of soil, so a stricter 3% threshold is set to control the overall shape error of the shear strength curve.

[0039] <6%: To address the local deviation of shear stress under various levels of normal stress, the 6% threshold can prevent local distortion of the stress-strain curve and ensure that the three-dimensional numerical model can accurately reproduce the nonlinear response of the solid model.

[0040] ≥94%: The 94% threshold requires the overall deviation to be controlled within 6%, which meets the conventional requirements for model fit in statistics and ensures the consistency of the curve trend.

[0041] ∈[95%, 105%]: The upper limit of 105% prevents the three-dimensional numerical model from becoming over-hardened, while the lower limit of 95% avoids softening, ensuring that the model is equivalent to the solid model in terms of macroscopic mechanical behavior.

[0042] Step 5: Conduct multi-condition, multi-parameter, and multi-factor ratio simulation experiments based on the digital twin model to obtain optimized parameters for the hole removal process.

[0043] In step five, the multi-factor proportioning test includes test factors such as material proportions, construction technology, and environmental conditions. When the filling material for the pulled-out hole is cement grout, the optimized parameters for the pulled-out hole treatment process are grout proportion, filling speed, filling process, and grouting pressure; when the filling material for the pulled-out hole is sand and gravel, the optimized parameters for the pulled-out hole treatment process are particle size distribution, filling speed, filling process, and filling pressure.

[0044] This embodiment simulated and tested various grouting processes, including bottom-up cement mortar method, direct cement mortar filling method, bottom-up sand and gravel method, and direct sand and gravel filling method. The experimental schemes for each process are shown in Tables 1 to 4. It should be noted that the grouting and sand filling speed is determined based on published relevant results, and the filling materials include ordinary silicate pure cement slurry and uniformly clean and dry sand and gravel.

[0045] Table 1 Test Scheme for Cement Mortar Bottom-Up Method Table 2 Test Scheme for Direct Filling Method of Cement Mortar Table 3 Test Scheme for the Bottom-Up Method of Sand and Gravel Table 4 Test Scheme for Direct Impact Method of Sand and Gravel Step 6: Using the density of the hole filling as the target parameter, perform variation source analysis on the target parameter, construct the factor sensitivity index and the relative deviation evaluation index between levels, and establish an optimized parameter combination based on the factor sensitivity index and the relative deviation evaluation index between levels.

[0046] In step six, after obtaining the filling density data under each test condition through multi-factor ratio experiment simulation, the average filling density of each test factor at different levels is calculated, and the level with the largest average value is selected as the preferred level of the test factor; among them, the test factor with the most significant influence is determined by the one with the largest corresponding factor sensitivity index F value.

[0047] The multiple indicators obtained from the multi-factor proportioning test are converted into a comprehensive score to construct an evaluation index: taking the bottom-up multi-factor proportioning test scheme of cement slurry as an example: that is, three factors and three levels.

[0048] Slurry mix ratio A: Level 1 (0.8:1), Level 2 (1:1), Level 3 (1.2:1); Pipe diameter B: Horizontal 1 (25mm), Horizontal 2 (32mm), Horizontal 3 (38mm); Injection pressure C: Level 1 (0MPa), Level 2 (0.5MPa), Level 3 (1MPa); Based on the data obtained from the multi-factor proportioning test for the filling density of the sheet pile pull-out hole, the filling density of Experiments 1 to 9 can be obtained. That is, to obtain the average value of the filling density. .

[0049] Calculate the relative deviation between levels of each factor , , ; Slurry ratio A: Experiment (1.2.3) ; Experiment (4.5.6) ; Experiments (7.8.9) ; Pipe diameter B: Experiment (1.4.7) ; Experiment (2.5.8) ; Experiment (3.6.9) ; Injection pressure C: Experiment (1.6.8) ; Experiment (2.4.9) ; Experiment (3.5.7) ; Calculate the sum of squared deviations from the mean, S: Total sum of squared deviations from the mean: ; Sum of squares of deviations from the mean of slurry mix proportion A: ; Sum of squares of deviations of pipe diameter B from the mean: ; Sum of squares of the deviations from the mean of the charging pressure C: ; Sum of squared errors from the mean ; Calculate the degrees of freedom W, mean square M, and factor sensitivity index F: Total degrees of freedom ; Degrees of freedom of each factor ; Error degrees of freedom ; Mean Square ; Mean Square ; F-value = .

[0050] Example 1: A foundation pit project for an underground utility tunnel in a certain city has a depth of 18m and uses U-shaped Larssen sheet piles for support. The sheet piles are 22m long and spaced 0.8m apart. After the foundation pit support was completed, the sheet piles were removed using a vibration method. During the removal process, a narrow, elongated cavity approximately 20m long with a cross-sectional dimension matching the sheet piles was formed in the stratum. The stratum in the area is mainly silty clay with local sand interlayers. There are municipal pipelines and existing buildings within a 5m radius. Therefore, the removed cavity needs to be efficiently and densely filled to prevent ground subsidence and damage to the pipelines.

[0051] Step 1: Preparation of porous solid samples Sheet pile extraction: A vibratory hammer is used to vibrate the sheet pile interlocks to loosen the surrounding soil before continuous vibration extraction. For two sheet piles that are difficult to extract, a diesel hammer is first used to drive them down 200mm, followed by alternating vibration with the vibratory hammer. When the sheet pile is pulled to 500mm above the foundation slab, extraction is paused, and the vibratory hammer is used to vibrate continuously for 3 minutes to initially fill part of the holes with soil around the pile.

[0052] Filler Injection: Bisphenol A type epoxy resin was selected as the curable filler. The hardener was slowly added to the resin at a resin:hardener ratio of 10:1, while stirring for 4 minutes. Then, degassing was performed for 8 minutes under a vacuum of -0.1 MPa. The filler was then injected from bottom to top through a pre-designed pipeline at a rate of 0.8 L / min, with the injection pressure controlled at 0.2 MPa. After injection, pre-curing was carried out at room temperature for 3 hours, followed by curing at 70°C for 6 hours, and then allowed to cool naturally to room temperature.

[0053] Sample removal: The cured epoxy resin filler was pulled out as a whole using a vibration method to obtain a complete solid sample with holes.

[0054] Step 2: Construction of Digital Hole Model Point cloud data acquisition: Reflective markers are pasted on the surface of the cavity filling body. The 3D laser scanner is connected to the computer, and the scanning accuracy is set to 0.1mm and the scanning interval is 0.5mm. The reflective markers and the surface of the filling body are scanned, and the 3D surface point cloud data is saved.

[0055] Model processing: The point cloud data was imported into Geomagic Studio software, discontinuous noise points were removed, sharp points were smoothed and reflective points were filled, and the data was encapsulated into a triangular mesh surface. After trimming irregular boundaries, the bottom surface was fitted to the XY plane, and a coordinate system was established with the center of the fitted plane as the origin. A complete digital hole model was obtained through progressive sampling.

[0056] Step 3: 3D Printing Solid Model and Solid Model Solid model printing: The digital hole model is imported into the 3D printing control system, and printing materials that match the mechanical properties of the soil are selected to prepare a 1:5 scale solid physical model (hole length 4m, cross-sectional dimensions scaled proportionally).

[0057] Filling process simulation: Cement grout was used as the filling material, and the bottom-up filling simulation was carried out according to the preset multi-factor ratio test plan, with the grouting speed controlled at 0.3L / min.

[0058] Fill effect check: Non-destructive testing: An ultrasonic coupling agent is applied to the outer wall of the model. A 50kHz ultrasonic probe is used to collect data in a fan-shaped rotation. A cross-sectional CT image is generated by a back-projection algorithm to detect the density of the filler and internal defects.

[0059] Destructive testing: The model was cut along the hole axis to observe the integrity of the filler and its contact with the hole wall. Samples were taken from the cross section to conduct a uniaxial compressive strength test, and the average compressive strength was measured to be 32 MPa.

[0060] Step 4: Numerical simulation model establishment and calibration Model construction: The digital cavity model is imported into PFC software, and a silty clay stratum model is constructed based on the on-site geotechnical investigation report. The discrete element method is used to simulate the cement grout filling process, and the grouting parameters and working conditions of the solid model are completely replicated.

[0061] Model calibration: Based on the Coulomb-Mohr strength theory, three sections were uniformly selected along the hole axis in both the solid model and the numerical simulation model. Three sets of parallel samples were taken from each section, and consolidated undrained shear tests were conducted under 10 levels of normal stress. (See attached diagram.) Figure 5 As shown, the nonlinear shear strength curves and key parameters of the two sets of samples are obtained and listed in Table 5.

[0062] Table 5. Nonlinear shear strength curves and key parameters of the two groups of samples. Technical parameter deviation verification: Solid model obtained: cohesion 18.6 kPa, internal friction angle 28.3°; The three-dimensional numerical model yielded the following results: cohesion 17.9 kPa, internal friction angle 27.6°. Deviation rate calculation: Cohesion deviation rate =3.8%, internal friction angle deviation rate =2.5%; Curve deviation verification: The point deviation rate of each shear stress meets the requirement of ≤6%, with a maximum deviation rate of 4.5%; Overall curve similarity index: ≥95.2%, meeting the preset accuracy standard; The area of ​​the curve in the solid model is 18623 kPa², and the area of ​​the curve in the numerical simulation is 18057 kPa². Area ratio 96.9%; In this verification, all indicators met the preset thresholds, indicating that the numerical simulation model and the physical model are highly matched, making it a highly reliable digital twin model.

[0063] Step 5: Numerical simulation and parameter optimization under multiple operating conditions Multi-factor mix design: Taking the bottom-up method of cement slurry as an example, the slurry mix ratio (A), pipe diameter (B), and filling pressure (C) are selected as influencing factors, with 3 levels for each, and a multi-factor mix design scheme is designed.

[0064] Numerical simulation: As shown in Table 6, nine sets of working conditions were simulated using a digital twin model to obtain the filling density data for each set of working conditions.

[0065] Table 6. Simulation results of filling density under 9 working conditions using the digital twin model. Data processing: Calculate the average filling density =91%.

[0066] Calculate the relative deviation between the levels of each factor. : =88.3%, = =94.3%, = =90.3%.

[0067] Calculate the sum of squares of deviations from the mean, degrees of freedom, mean square, and F-value, where the slurry mix ratio F-value = 12.6, the pipe diameter F-value = 4.8, and the filling pressure F-value = 6.3.

[0068] Step 6: Determining the optimal process parameters Based on value analysis The slurry mix ratio A2 (1:1), pipe diameter B2 (32mm), and filling pressure C2 (0.5MPa) correspond to... The value is the largest, and the F value of the slurry ratio is the largest, which means the influence is the most significant. Finally, the optimal combination of process parameters was determined to be: slurry ratio 1:1, pipe diameter 32mm, and filling pressure 0.5MPa.

[0069] On-site application results: Holes were filled on-site using optimal process parameters. After filling, ultrasonic CT testing showed a filling density of 96%, with no obvious voids or cracks. Monitoring data showed that ground settlement in the surrounding area was controlled within 3mm, and the building showed no deformation, meeting engineering safety requirements. Furthermore, compared to traditional empirical methods, material consumption was reduced by 15%, and the construction period was shortened by 20%, significantly improving the project's economy and efficiency.

[0070] For any parts not mentioned in this application, existing technologies may be used or referenced.

[0071] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An optimization method for filling the extraction cavity of steel sheet piles based on 3D printing and numerical simulation, characterized in that, Includes the following steps: Step 1: Remove the sheet pile to form a pull-out hole, fill the pull-out hole with curing material, and after the curing material has solidified, remove it and scan its surface to obtain the three-dimensional data of the solidified material. Step 2: Based on the three-dimensional data of the solidified material, establish a numerical model of the pull-out hole, input the numerical model into the 3D printing system, and print the solid model of the pull-out hole; Step 3: Fill the solid model according to the preset filling scheme of the hole to conduct the test, and check the test parameters of the filling after the test; Step 4: Establish a three-dimensional numerical model based on the numerical model and simulate the preset filling scheme. Calibrate the three-dimensional numerical model according to the test results of the physical model. After calibration, obtain the digital twin model of the hole. Step 5: Conduct multi-condition, multi-parameter, and multi-factor ratio simulation experiments based on the digital twin model to obtain optimized parameters for the hole removal process.

2. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 1, characterized in that, Optimization methods also include: Step 6: Using the density of the hole filling as the target parameter, perform variation source analysis on the target parameter, construct the factor sensitivity index and the relative deviation evaluation index between levels, and establish an optimized parameter combination based on the factor sensitivity index and the relative deviation evaluation index between levels.

3. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 2, characterized in that, In step three, the test parameters of the filler are obtained based on non-destructive testing and / or destructive dissection methods, respectively. The test parameters include overall density, defect location, and contact state with the pull-out hole wall.

4. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 3, characterized in that, In step four, multiple sets of cross-sectional size samples are selected in the solid model and the three-dimensional numerical model. For each cross-section, multiple sets of parallel samples are selected. Consolidated undrained shear tests under multi-level normal stress are carried out to obtain the shear strength curves and shear strength parameters of the solid model and the three-dimensional numerical model. The shear strength parameters include cohesion and internal friction angle.

5. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 4, characterized in that, In step four, the solid model and the 3D numerical model are bidirectionally verified based on the Coulomb-Moore theoretical curve to improve the credibility of the digital twin model. The bidirectional verification specifically includes the following sub-steps: Step 1: Perform cohesion deviation rate verification and internal friction angle deviation rate verification. ； ； In the formula, Indicates the cohesion deviation rate. Indicates the deviation rate of the internal friction angle. Represent the cohesion and internal friction angle of the three-dimensional numerical model under the experiment, respectively. These represent the cohesion and internal friction angle of the solid model under test, respectively; Step 2: For the shear stress-normal stress curves of the 3D numerical model and the solid model, calculate the shear stress deviation at the corresponding normal stress level. : ； In the formula, This represents the shear stress in a three-dimensional numerical model. Represents the shear stress of the solid model. Indicates the normal stress level number. Indicates the number of samples; Step 3: Constructing an overall curve similarity evaluation index : ； In the formula, m represents the number of normal stress levels, and n represents the number of samples; The ratio of the area under the curves of the solid model and the 3D numerical model is: ； In the formula, This represents the area of ​​the curve in the three-dimensional numerical model. The area of ​​the curve in the solid model; Step 4: Set the verification threshold. When all verification indicators meet the corresponding verification thresholds, the current 3D numerical model is determined to be a digital twin model.

6. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 5, characterized in that, In step four, when performing bidirectional verification of the solid model and the three-dimensional numerical model, the multiple sets of cross sections are uniformly selected along the axial direction of the pull-out hole, and the number of multiple sets of parallel samples taken for each cross section is not less than 3 sets.

7. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 6, characterized in that, In step five, the multi-factor proportioning test includes test factors such as material proportions, construction technology, and environmental conditions.

8. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 7, characterized in that, In step five, when the filling material for the pull-out hole is cement grout, the optimized parameters for the pull-out hole treatment process are grout mix ratio, filling speed, filling process, and grouting pressure. When the filling material for the extraction hole is sand and gravel, the optimized parameters for the extraction hole treatment process are particle size distribution, filling speed, filling process, and filling pressure.

9. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 8, characterized in that, In step one, the curing material is bisphenol A type epoxy resin.

10. The method for optimizing the filling process of steel sheet pile pull-out holes based on 3D printing and numerical simulation according to claim 9, characterized in that, In step six, after obtaining the filling density data under each test condition through multi-factor ratio experiment simulation, the average filling density of each test factor at different levels is calculated, and the level with the largest average value is selected as the preferred level of the test factor; among them, the test factor with the most significant influence is determined by the one with the largest corresponding factor sensitivity index F value.

Images