Marine soft soil with pebble rcd multi-field coupling simulation method

Abstract

The present application relates to geotechnical engineering and engineering simulation technical field, provide marine soft soil sandwich pebble RCD multi-field coupling simulation method, including the following: step 1, form multi-source input parameter table;Step 2, build fluid domain, particle domain and three-domain collaborative model of soil domain on unified solution platform;Step 3, copy the reverse circulation loop and load the real working condition boundary;Step 4, realize three-domain strong coupling calculation with multi-scale and sub-cycle strategy;Step 5, online updating and recalibration of parameters and model;Step 6, give the hole wall stability and sedimentation risk criterion;Step 7, give the construction parameter window and recommended process curve;Step 8, output control index results and early warning rules;Step 9, the calibrated model and parameter library are archived for engineering reuse.The present application provides a set of multi-field coupling numerical simulation method and tool basis which can be calculated, calibrated and optimized, and has good engineering adaptability, popularization and economic benefits.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of geotechnical engineering and engineering simulation, and particularly relates to a multi-field coupling simulation method for marine soft soil with pebble RCD. BACKGROUND

[0002] The foundation engineering of urban rail transit, cross-sea channel and high-rise buildings in coastal areas is generally in marine soft soil. Such soil has the characteristics of high water content, large void ratio, long consolidation time, low shear strength, small permeability coefficient, strong structure and sensitivity, etc. It is often mixed with pebble (gravel) lens or interlayer formed by tidal current transport and wave action of shoreline, which is a spatially discontinuous and unevenly thick and connected stratum combination. Under the above stratum conditions, the construction of cast-in-place piles is prone to induce risks such as hole collapse, shrinkage, drill biting, slurry absorption (leakage), sediment overrunning and concrete pouring defects, which becomes a key problem for engineering quality and construction period control.

[0003] In complex strata, reverse circulation drilling (RCD) can effectively improve the drilling slurry discharge efficiency, improve the hole bottom cleanliness and reduce the hole collapse probability due to its internal tube negative pressure slurry absorption and annular supply loop characteristics, and has been widely used in marine soft soil with pebble layer. However, the construction mechanism of RCD involves multi-physical field and multi-scale processes such as non-Newtonian mud flow, discrete particle migration, hole wall-soil consolidation seepage response, drill tool-flow field-particle interaction, etc. The traditional empirical design and single-field calculation method cannot comprehensively and accurately predict and optimize the construction safety and efficiency of RCD.

[0004] The existing technology and its limitations are embodied as follows: (1) traditional experience or semi-experience design method, which depends on experience charts and engineering analogy to determine mud density, viscosity, pump capacity and rotating speed, etc. This kind of method is difficult to reflect the randomness and non-steady-state loop characteristics of the gravel, and is poor in adaptability to tides or time-varying boundaries. The parameter window is often conservative or distorted. (2) Single-phase or simplified fluid model, which often regards the in-hole medium as a homogeneous Newtonian fluid, and ignores the yield stress and thixotropy. The local three-dimensional flow field in the inlet area and the starting conditions of the sediment are not described, and the difference in hole cleaning effect under the same pump capacity cannot be explained. (3) Mud hydraulics and pipe network are calculated separately. The ground-drilling tool pipe network is processed by one-dimensional steady-state hydraulic analysis, which is decoupled from the three-dimensional in-hole flow field, and it is difficult to accurately reflect the key dynamics such as in-pipe negative pressure fluctuation, cavitation margin, and inlet pressure feedback. (4) The particle size effect is ignored, and the gravel and drilling sediment are equivalent to continuous medium or average particle size parameters, which cannot characterize the micro-mechanism of screen gap blockage, local accumulation, jamming or restarting. (5) The body side is only checked for static force limit: the hole wall stability is often estimated by the static water head difference and the simplified Mohr-Coulomb criterion, which is not coupled with the seepage consolidation and transient pore pressure response; it is difficult to evaluate the time evolution and risk crossing of the effective stress caused by short-time stop-pump, lifting / lowering of drilling tools, etc. (6) Lack of multi-objective parameter optimization: the current practice is mainly based on single indicators such as sediment thickness or pump pressure, and a systematic trade-off between "hole collapse risk-hole cleaning efficiency-energy consumption-progress" has not been established, which is not robust for different gravel content and particle size combinations. (7) Insufficient closed-loop with field data: even if numerical simulation is introduced, it is mostly offline analysis, lacking online calibration and recalculation mechanism with real-time data such as pump pressure, backflow density / viscosity, in-pipe negative pressure, torque, current, sediment amount, etc., leading to gradual accumulation of model deviation from reality.

[0005] In summary of the above status, the industry urgently needs a multi-field coupled simulation method that can couple "non-Newtonian mud flow-discrete particle transport-soil seepage consolidation-hole wall stability-pipe network transient hydraulics" in a unified framework. SUMMARY

[0006] To solve the problems in the background art, the marine soft soil gravel RCD multi-field coupled simulation method is provided, which provides a set of calculable, calibratable and optimal multi-field coupled numerical simulation method and tool basis for reverse circulation bored pile construction in marine soft soil gravel layer conditions, and has good engineering adaptability, popularization and economic benefits.

[0007] To achieve the above purpose, the following scheme is adopted:

[0008] The marine soft soil gravel RCD multi-field coupled simulation method comprises the following steps:

[0009] Step 1, collecting material parameters to form a multi-source input parameter table;

[0010] Step 2: Construct a multi-field coupling model and establish a three-domain collaborative model of fluid domain, particle domain and soil domain in a unified solution platform;

[0011] Step 3: Replicate the reverse circulation loop and load the actual working condition boundary to establish a mud closed loop consistent with the actual construction, and set the drilling tool working condition to be loaded in the form of time history.

[0012] Step 4: Achieve strong coupling calculations among the fluid domain, particle domain, and soil domain using a multi-scale and sub-cycle strategy;

[0013] Step 5: Use on-site monitoring to update and recalibrate the parameters and model online;

[0014] Step 6: Based on the effective stress path and local flow field, provide the criteria for judging the stability of the borehole wall and the risk of sedimentation;

[0015] Step 7: Conduct multi-objective optimization with safety, efficiency and energy consumption as the goals, and provide construction parameter windows and recommended process curves;

[0016] Step 8: Output control indicator results and early warning rules to achieve linkage parameter adjustment;

[0017] Step 9: Archive the calibrated parameters and model into the engineering parameter library for engineering reuse.

[0018] Optionally, in step 1, the collected material parameters refer to the collection and supplementation of stratigraphic, geotechnical, hydrological, mud, and equipment parameters obtained from experiments, followed by consistency verification and outlier removal. The stratigraphic, geotechnical, hydrological, mud, and equipment parameters specifically include:

[0019] In terms of strata, the layering, thickness variation, and spatial distribution of gravel lenses in marine soft soil were clarified based on exploration boreholes, core sampling, and in-situ testing. Gravel was analyzed by sieving and imaging methods to statistically determine particle size distribution, flatness factor, volume fraction, and lens thickness-connectivity index.

[0020] In terms of geotechnical engineering, state parameters, strength parameters, seepage-consolidation parameters, historical parameters, and disturbance sensitivity parameters were obtained through indoor triaxial consolidated undrained creep and permeability tests.

[0021] In terms of hydrology, establish the time-varying boundary relationship between groundwater level and tides;

[0022] Regarding the mud, the target range and temperature dependence of mud density, yield stress, consistency, index and thixotropic recovery coefficient were obtained through formula pre-selection and rotary rheometer testing.

[0023] Equipment, side drawing aperture, annulus width, inner tube diameter, inlet opening rate, screen size, ground pipe network configuration and along or local loss coefficient, and record pump curve, rated power and net positive suction head NPSH parameters, where the available net positive suction head The necessary net positive suction head is calculated according to the system static pressure head, liquid level, along and local loss and medium saturated steam pressure Given by the pump characteristic curve, the control criterion is .

[0024] Optionally, in step 2, the fluid domain, the particle domain and the soil domain are weakly coupled or partitioned strongly coupled through sharing geometry or topological relationship and exchanging interface quantity, and an immersed boundary or ALE dynamic mesh is introduced at the hole wall for describing the feedback of the slight deformation of the hole wall to the near-wall flow field; wherein the fluid domain is for mud flow, and the mud rheology follows Herschel-Bulkley constitutive:

[0025] ,

[0026] Wherein, in numerical implementation, the tensor form is adopted:

[0027] ,

[0028] In the formula, The yield stress is K The consistency is The shear stress tensor is The equivalent shear rate is D The strain rate tensor is n; the flow index is The apparent viscosity is

[0029] The particle domain adopts the discrete element method for pebbles and drilling sludge, and allows the activation of the bonding-unbonding mechanism in the high shear zone;

[0030] The soil domain selects the modified Cambridge model MCC or the double structure Cambridge damage coupling model DSC+DC, adopts displacement-pore pressure or displacement-relative displacement consolidation seepage coupling, and sets the Brinkman-Darcy transition layer in the near-field of the hole wall for describing the weak seepage or shear transition.

[0031] Optionally, in step 3, the reverse circulation loop is copied as the reverse circulation boundary of the annulus supply-hole bottom inlet-inner tube negative pressure return, wherein the annulus end is provided with a constant pressure or constant flow boundary, and the along or local loss is taken into account; The hole bottom inlet area is subjected to geometric constraints of opening rate and screen size, and local loss and inlet contraction effect are used to reflect shear enhancement and vortex formation; The inner tube end is provided with a negative pressure sludge suction boundary, and is coupled with the pump curve and NPSH constraint; The backsludge condition at the ground end adopts liquid level or back pressure of the separation device;

[0032] The drilling tool working condition is loaded in time sequence form, including rotating speed, drilling pressure, footage rate and pump volume, and is allowed to give a short rhythm sequence of drilling-cleaning-measuring and pump stop or start transient state.

[0033] Optionally, in step 4, the multi-scale and sub-cycle strategy specifically includes:

[0034] Step 4.1, three-dimensional CFD is used to refine the hole bottom inlet and near-wall area, and axisymmetric or simplified area is used for the far field; three-dimensional CFD is coupled with one-dimensional transient pipe network equation through boundary flux and pressure feedback;

[0035] Step 4.2, two-way momentum exchange between CFD and DEM is realized based on voxelized grid, the effect of particles on fluid is realized through drag, lift or additional mass force; the effect of fluid on particles is realized through local velocity gradient and turbulent kinetic energy of fluid; DEM sub-step length is set to meet the stable condition of particle collision, and is synchronized with CFD step length according to the sub-cycle ratio of 10 to 100;

[0036] Step 4.3, FEM and CFD are coupled by exchanging hole pressure and wall shear stress through hole wall interface, and relaxation iteration method is used to suppress numerical oscillation in strong coupling;

[0037] Step 4.4, time discrete step length is determined based on CFL condition, and local adaptive grid and adaptive time step adjustment are triggered in high gradient area of fluid inlet to capture the migration of sediment core and resuspension boundary.

[0038] Optionally, in step 5, the parameters include rheological parameters and soil parameters, and the updating and calibration steps of the rheological parameters and soil parameters specifically include: the calibration of rheological parameters is based on the rheological test results obtained in step 1 to obtain the initial value of parameters; and the return flow density or viscosity measured on site is used to update and calibrate the Kalman or Bayesian filter; the calibration of soil parameters is based on the soil basic parameters in step 1 as the prior, and the MCC or DSC parameters are re-calibrated by short pile test or pore forming response inversion to obtain the posterior estimation;

[0039] The model updating and calibration steps specifically include: key parameters and boundary disturbances are updated online by using unscented Kalman filter, Bayesian update or particle filter, and when the residual error exceeds the threshold, parameter re-identification and simulation restart are triggered.

[0040] Optionally, in step 6, the borehole wall stability criterion is based on the strength reduction or Mohr-Coulomb safety factor Fs, which specifically includes: calculating the effective stress path time history response of the soil borehole wall by coupling the borehole water head variation and seepage consolidation process; identifying the safety factor F s across the early warning threshold; when the safety factor F s approaches or is lower than the preset early warning threshold, outputting a borehole wall risk index R c , and locking the problem deep section;

[0041] The sediment risk criterion specifically includes: calculating the shear stress between the wall surface and the bed surface in the three-dimensional local flow field at the borehole bottom inlet; comparing the shear stress with the critical incipient shear stress of the particles to determine the deposition area and resuspension area of the sediment; updating the sediment thickness and its planar morphology based on the bed surface evolution equation and local mass conservation; at the same time, calculating and comparing the available net positive suction head NPSH avail and the required net positive suction head NPSH req ; when NPSH avail <NPSH req , determining the risk of cavitation in the circulating loop; if the margin is insufficient, adjusting the pump flow Q or the constant pressure boundary H a in linkage, and triggering a short-time pulsating cleaning operation.

[0042] Optionally, step 7 specifically includes: selecting mud density, yield stress, pump capacity, rotational speed, drilling pressure, footage rate, and inlet opening rate as decision variables; using NSGA-II or equivalent multi-objective evolutionary algorithm to search for the Pareto frontier; using a surrogate model to accelerate evaluation during optimization, outputting a contour map of the constructible parameter window and a recommended process curve, and generating a robustness comparison for different gravel content or particle size combinations.

[0043] Optionally, in step 8, the control index results include the parameter window and the control index, the sensitivity matrix under the boundary scenario, and the recommended curve; the early warning rule identifies the signs of inlet blockage, stuck drill pipe, and hole collapse by setting joint features.

[0044] Optionally, step 9 specifically includes: archiving the calibrated material parameters, equipment characteristics, optimization results, and sensitivity maps to the engineering parameter library to form a retrievable entry of stratum type-gravel statistics-equipment configuration-optimal parameter window, exporting a standardized calculation report and storing it as a readable engineering template; subsequent engineering is quickly initialized and calibrated online based on the readable engineering template by replacing a small number of parameters.

[0045] The present application has the beneficial effects that the scheme provides a multi-field coupling numerical simulation method and tool basis which can be calculated, calibrated and optimized for the construction of reverse circulation bored piles under the condition of marine soft soil with pebble layers, specifically:

[0046] (1) Coupling the non-Newtonian mud flow field, discrete particle migration field and marine soft soil seepage-stress field in the same numerical platform, the simultaneous solution of the whole process of reverse circulation drilling is realized, the effective stress path change of the hole wall under complex working conditions can be simulated and predicted in real time, and the advanced early warning of hole collapse risk is realized.

[0047] (2) The reverse circulation mechanism and boundary can be truly expressed, the reverse circulation boundary of annulus supply-hole bottom inlet-negative pressure up-return of inner pipe and transient pipe hydraulic equation are adopted to depict the dynamic relationship among loop pressure drop, negative pressure and up-return flow, which restores the loop fluid dynamics characteristics and reproduces the dynamic influence of field operation, providing simulation basis for subsequent accurate analysis of loop performance, hole wall stability and risk warning.

[0048] (3) The particle size effect of pebble and drilling slag is depicted, that is, the interaction of particles, fluid, hole wall and drilling tool is explicitly described in the numerical model, and the screen gap blockage, local accumulation and restart conditions are quantified, thereby solving the problem of ignoring the particle size effect in the prior art.

[0049] (4) A hole wall stability time course determination method is constructed, forming a time course safety criterion suitable for non-steady state construction operation; a quantitative criterion for slurry formation and hole cleaning capacity is established, which can guide hole cleaning quality control.

[0050] (5) A multi-objective parameter optimization mechanism is formed, and the output of the construction parameter window and the recommended process curve can be directly connected to the field operation, solving the problem that the current method does not establish a systematic trade-off among hole collapse risk, hole cleaning efficiency, energy consumption and progress.

[0051] (6) Online calibration and recalculation with field data are realized: the monitoring data such as pump pressure, backflow density / viscosity, inner pipe negative pressure, torque and slurry amount are introduced into the model to construct the online parameter determination and recalculation process, supporting digital twin adaptive adjustment and risk warning.

[0052] (7) A software form of engineering reusability is provided, that is, the corresponding simulation system / device and its computer program and storage medium are given, which is convenient for rapid deployment and reuse in different projects. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 is the overall flowchart of the method of the present application;

[0054] Figure 2A schematic diagram of three-domain coupling of fluid domain, particle domain and soil domain in the embodiment of the present application;

[0055] Figure 3 A schematic diagram of local flow field and particle starting criterion in the bottom inlet area in the embodiment of the present application;

[0056] Figure 4 A schematic diagram of the calculation domain and boundary conditions of the hole wall in the embodiment of the present application;

[0057] Figure 5 A schematic diagram of the construction parameter window and recommended process curve output of multi-objective optimization in the embodiment of the present application;

[0058] Figure 6 A schematic diagram of online calibration and parameter adjustment logic in the embodiment of the present application;

[0059] Figure 7 A schematic diagram of numerical simulation of anti-circulation bored pile construction in the embodiment of the present application. DETAILED DESCRIPTION

[0060] In order to make the present application clearer and more understandable, the following describes the present application in detail with reference to the accompanying drawings and examples. It should be understood that the given examples are only one of the implementation manners, and do not represent all the examples.

[0061] Example 1

[0062] In combination Figures 1-6 , the embodiment of the present application provides a marine soft soil with pebble RCD multi-field coupling simulation method, which aims to provide a set of multi-field coupling numerical simulation method and tool basis that can be calculated, calibrated and optimized for the construction of anti-circulation bored piles under the condition of marine soft soil with pebble layers. The method comprises the following steps:

[0063] Step 1, collecting material parameters: collecting and supplementing test obtained stratum, soil, hydrology, mud and equipment parameters, and performing consistency checking and outlier elimination to form a multi-source input parameter table. The stratum, soil, hydrology, mud and equipment parameters specifically include:

[0064] In terms of stratum, based on exploration holes, coring and in-situ testing, the stratification, thickness fluctuation and spatial distribution of marine soft soil lens with pebbles are determined; the particle size distribution D10 / D50 / D90, flatness coefficient, volume fraction V g and lens thickness-connectivity index of pebble are comprehensively counted, and four key characteristic indexes of particle size distribution, shape, content and spatial structure of pebble are counted.

[0065] In terms of soil, state parameters (γ sat, w, e0), strength parameters (c, φ), seepage-consolidation parameter k, historical parameter OCR and disturbance sensitivity parameter.

[0066] In hydrology, the time-varying boundary relationship between groundwater level and tide is established, and the high water level during construction and the rainstorm infiltration scenario are determined.

[0067] In mud, the target interval and temperature dependence of mud density , yield stress , consistency K, index n and thixotropic recovery coefficient are obtained through formula preselection and rotary rheometer test.

[0068] In equipment, side drawing aperture , annulus width, inner tube diameter, inlet opening rate , screen gap size s, ground pipe network configuration and along or local loss coefficient, and record pump curve, rated power and net positive suction head NPSH parameters, where, available net positive suction head The necessary net positive suction head is calculated according to the system static pressure head, liquid level, along and local loss, and medium saturated steam pressure. The necessary net positive suction head is calculated according to the system static pressure head, liquid level, along and local loss, and medium saturated steam pressure. .

[0069] After consistency checking and abnormal value elimination of the above data, the tabular data of "input parameter-source-uncertainty" are formed, which are used as the basis for numerical model initialization and subsequent calibration.

[0070] Step 2, build a multi-field coupling model, build a three-domain collaborative model of fluid domain, particle domain and soil domain in a unified solving platform, and realize weak coupling or partition strong coupling through sharing geometric / topological relationship and exchanging interface quantities (volume fraction, momentum source term, pore pressure and effective stress), as shown in Figure 2 The immersed boundary or arbitrary Lagrangian-Eulerian method ALE dynamic mesh is introduced at the hole wall to describe the feedback of the slight deformation of the hole wall to the near-wall flow field.

[0071] Among them, the fluid domain is for mud flow, and the mud rheology follows the Herschel-Bulkley (Herschel-Bulkley) constitutive equation (with thixotropic term) to solve the local three-dimensional flow field in the annulus and hole bottom, and a temperature correction interface for viscosity is reserved:

[0072] ,

[0073] Among them, in the numerical implementation, the tensor form is adopted:

[0074] ,

[0075] In the formula, is the yield stress, Kfor consistency, for shear stress tensor, for equivalent shear rate, D for strain rate tensor; n is flow index; for apparent viscosity;

[0076] The particle domain discretizes the gravel and drilling slurry explicitly, adopts the contact model of normal / tangential spring-damper, Coulomb friction and rolling damping, and allows the bond-debond mechanism to be activated in the high shear zone.

[0077] The soil domain selects the modified Cambridge model (MCC) or the dual-structure Cambridge damage coupling model (DSC+DC) combined constitutive model, adopts the displacement-pore pressure u-p or displacement-relative displacement u-w consolidation seepage coupling, and sets the Brinkman-Darcy transition layer near the pore wall for describing the weak seepage / shear transition.

[0078] This step realizes the simultaneous solution of the whole process of the reverse circulation drilling by coupling the non-Newtonian mud flow field, the discrete particle migration field and the marine soft soil seepage-stress field in the same numerical platform.

[0079] Step 3, copy the reverse circulation loop and load the real working condition boundary, establish a mud closed loop consistent with the actual construction, and set the drilling tool working condition in the form of time history, so that the reverse circulation drilling simulation is consistent with the actual construction scene, which not only restores the hydrodynamic characteristics of the loop, but also reproduces the dynamic influence of the field operation, providing a simulation basis for subsequent accurate analysis of loop performance, hole wall stability and risk warning, such as Figure 4 .

[0080] The reverse circulation drilling pile (RCD) loop is usually composed of a mud pool, an annular downlink, a hole bottom inlet area, an inner tube uplink and a ground separation device, therefore, the embodiment constructs the reverse circulation boundary of annular supply-hole bottom inlet-inner tube negative pressure uplink based on the RCD loop, wherein the annular end is provided with a constant pressure or constant flow boundary , and the along-the-way or local loss is taken into account; the hole bottom inlet area is applied with the geometric constraints of opening rate α and screen joint size s, and the local loss ζ and inlet contraction effect are adopted to reflect shear enhancement and vortex formation; the inner tube end is applied with a negative pressure slurry suction boundary , and the pump curve and NPSH constraint are coupled to avoid cavitation and gas resistance; the ground end returns to the slurry condition with a liquid level or a separation device back pressure.

[0081] The drilling tool working condition is loaded in the form of time history, including rotating speed N(t), drilling pressure WOB(t), footage rate ROP(t) and pump volume Q(t), and allows to give a short rhythm sequence of drilling-cleaning-measuring and pump stop or start transient, which can reproduce the influence of field operation on the loop dynamics and hole wall effective stress path.

[0082] Step 4, in order to balance accuracy and efficiency, a multi-scale and sub-cycle strategy is used to realize the strong coupling calculation of the fluid domain, the particle domain and the soil domain, so as to depict the particle scale effect of gravel and drill cuttings, that is, the interaction between particles and fluid, hole wall and drilling tool is explicitly described in the numerical model, and the conditions of screen gap blockage, local accumulation and restart are quantified, so as to solve the problem that the particle scale effect is ignored in the prior art. The steps specifically include:

[0083] Step 4.1, three-dimensional CFD is used to refine the hole bottom inlet and the near-wall area, and axisymmetric or simplified area is used for the far field; three-dimensional CFD is coupled with one-dimensional transient pipe network equation through boundary flux and pressure feedback to ensure the consistency of loop pressure drop.

[0084] Step 4.2, based on the voxelized grid, the two-way momentum exchange between CFD and DEM is realized, the effect of particles on fluid is realized through drag, lift or additional mass force; the effect of fluid on particles is realized through local velocity gradient and turbulent kinetic energy of fluid; the DEM sub-step length is set to meet the stable condition of particle collision, and is synchronized with the CFD step length according to the sub-cycle ratio of 10 to 100, which can be expressed as:

[0085] ,

[0086] In the formula, is the fluid-particle interaction force, is the drag coefficient, is the fluid velocity vector, is the particle velocity vector; is the particle density; is the fluid density, is the particle volume, is the gravitational acceleration.

[0087] Step 4.3, the coupling between FEM and CFD is realized by exchanging hole pressure and wall shear stress through hole wall interface, and the relaxation iteration method is used to suppress numerical oscillation in strong coupling.

[0088] Step 4.4, the time discrete step length is determined based on the CFL Courant-Friedrichs-Lewy condition, and local adaptive grid and adaptive time step adjustment are triggered in the high gradient area of fluid inlet to capture the migration of sediment core and resuspension boundary.

[0089] This step establishes a quantitative criterion for sediment formation and hole cleaning capacity: a particle start / resuspension judgment and sediment thickness prediction model is given in the local three-dimensional flow field at the hole bottom inlet, which is used to guide the hole cleaning quality control.

[0090] Step 5 involves online updating and recalibrating of parameters and the model using on-site monitoring. This step incorporates on-site monitoring data such as pump pressure, reflux density / viscosity, inner pipe negative pressure, torque, and sediment volume into the model to construct an online parameter calibration and recalculation process. This supports digital twin adaptive adjustment and risk warning, such as... Figure 6 .

[0091] Specifically, the parameters include rheological parameters and soil parameters. The update and calibration steps for the rheological and soil parameters specifically include: updating the rheological parameters (mud density)... Yield stress The calibration of the consistency K, exponent n, and thixotropic recovery coefficient (γ) is based on the rheological test results obtained in step 1, and initial values ​​of the parameters are fitted; then, Kalman or Bayesian filtering is applied to update them using the reflow density or viscosity measured online in the field; the calibration of the soil parameters is based on the basic soil parameters (γ) obtained in step 1. sat (e, w, e0, c, φ, k, etc.) are priors, obtained through short pile tests or borehole response inversion. Recalibrate after the posterior estimates of the parameters of the corrected Cambridge model MCC or the perturbation state conceptual model DSC are completed.

[0092] The specific steps for updating and calibrating the model include: during field operation, collecting signals such as pump pressure, flow rate, inner pipe negative pressure, torque / current, return sediment, and liquid level; constructing a state vector x and an observation vector y; and using unscented Kalman filtering (or Bayesian update / particle filtering) to update key parameters and boundary disturbances online. When the residual statistics exceed the threshold, parameter recalibration and simulation restart are triggered to ensure that the consistency between the model and the field is maintained over time.

[0093] Step 6: Based on the effective stress path and local flow field, provide criteria for borehole wall stability and sediment risk. This step establishes a time-history method for determining borehole wall stability, forming a time-history safety criterion applicable to unsteady-state construction operations; and establishes quantitative criteria for sediment formation and borehole cleaning capability to guide borehole cleaning quality control.

[0094] Specifically, the stability criterion for borehole walls is based on strength reduction or the Mohr-Coulomb safety factor. With this as the core, its formula is:

[0095] ,

[0096] In the formula, For effective cohesion, For effective confining pressure, This represents the maximum shear stress at the borehole wall. It is the tangent of the effective internal friction angle.

[0097] Furthermore, the borehole wall stability criterion specifically includes: calculating the effective stress path time history response of the soil borehole wall by coupling the changes in water head within the borehole with the seepage consolidation process; and based on the effective stress path time history response, identifying the safety factor caused by at least one unsteady-state operation, such as pump shutdown, pump startup, raising or lowering of the drill string, or fluid level fluctuations. Behavior that crosses the warning threshold; when the safety factor When the value approaches or falls below the preset warning threshold, the output hole wall risk index is activated. And pinpoint the deeper aspects of the problem;

[0098] The specific criteria for determining sediment risk include: calculating the shear stress between the wall and the bed surface within the three-dimensional local flow field at the borehole inlet; and applying the shear stress... Critical starting stress of particles By comparing the sedimentation and resuspension zones, the deposition and resuspension zones of the sludge can be determined, such as... Figure 3 Update sediment thickness based on bed evolution equation and local mass conservation. Its planar morphology; simultaneously, the available net positive suction head (NPSH) is calculated and compared. With required net positive suction head (NPSH) ;when At that time, determine the risk of cavitation in the circulation loop; if the margin is insufficient, adjust the pump flow rate Q or the constant pressure boundary accordingly. This triggers a short-term pulsed cleaning operation.

[0099] Step 7: Conduct multi-objective optimization with safety, efficiency, and energy consumption as the goals, and provide construction parameter windows and recommended process curves, such as... Figure 5 Finding an acceptable balance between safety, efficiency, and energy consumption, without sacrificing safety for efficiency or blindly controlling energy consumption to the point of delay, and with the output parameter windows and process curves directly connected to on-site operations, solves the shortcomings of current practices that do not establish a systematic trade-off between "hole collapse risk, hole cleaning efficiency, energy consumption, and schedule".

[0100] The specific steps include: minimizing the risk of borehole collapse. or maximum ), minimum sediment ( The objectives are "lowest energy consumption (energy consumption per unit advance E) and highest efficiency (time-based advance rate η)," ensuring that the results align with actual engineering needs rather than pursuing a single indicator. Adjustable parameters are determined, and mud density is selected. Yield stress Pump flow rate Q, rotational speed N, drilling pressure WOB, drilling rate ROP, inlet opening ratio The decision variables are determined by NSGA-II or equivalent multi-objective evolutionary algorithms, which are used to search for the Pareto front and select the Pareto optimal solution among the three objectives of safety, efficiency, and energy consumption. A surrogate model (Gaussian process / radial basis function) is employed to accelerate the evaluation process, outputting contour maps of constructable parameter windows and recommended process curves (e.g., based on the time of day). Rise moderately (with Q, down-regulation of N and ROP), and generated robustness comparisons for different combinations of pebble content / size.

[0101] Step 8: Output control index results and early warning rules to achieve linkage parameter adjustment. The method in this embodiment can transform the model's calculation results into directly usable engineering tools, and at the same time, through real-time risk identification, provide early warning of potential construction hazards and specific solutions.

[0102] To transform the complex optimization results and parameter relationships of the model into outcomes that engineers can quickly understand, the model's control index results are output in two categories: reports and graphs. The first category includes parameter windows and control indices (such as...). Combinatorial domain of Q, N, and ROP , The first is the NPSH margin ≥ 1.2), and the second is the sensitivity matrix and recommendation curve under the boundary scenario.

[0103] The warning rules identify signs of inlet blockage, stuck drill, and borehole collapse risk by setting combined characteristics. Specifically, when the combined characteristics of "continuous decrease in inner tube negative pressure + increased pump flow fluctuation + increased torque" are detected, it is judged as an early sign of inlet blockage / stuck drill, and a warning is issued, along with "short-term pulsating pressurization + backwashing + reducing ROP / increasing..." (Slight) "Linkage suggestion"; when "liquid level drops + pore pressure recovers slowly + If both the "downward warning value" and the "downward warning value" are present, it indicates a risk of borehole collapse and recommends "increasing the mud head, temporarily halting the advance, and implementing short-cycle borehole cleaning." All warnings include actionable parameter adjustment ranges and estimated quantified impact values.

[0104] Step 9, the calibrated parameters are archived to the engineering parameter library for engineering reuse. Specifically, after completing the simulation and calibration of one project, the identified material parameters, equipment characteristics, optimization results and sensitivity atlas are archived to the engineering parameter library to form a retrievable entry of "stratum type- pebble statistics-equipment configuration-optimal parameter window"; a standardized calculation report (including input source, grid and numerical strategy, convergence and residual statistics, key atlas and control suggestions) is simultaneously exported and stored as an engineering template on a readable storage medium. Subsequent similar projects only need to replace a small amount of survey and equipment parameters, so that the model can be quickly initialized and calibrated online after the first shift data arrives, significantly shortening the modeling-optimization-production cycle and maintaining a closed-loop operation with the field monitoring.

[0105] Therefore, the method of the embodiment follows the layer-by-layer advancing logic of input-modeling-control edge-strong coupling-calibration-evaluation-optimization-output-reuse, and provides a set of computable, calibratable and optimizable multi-field coupling numerical simulation method and tool basis. The traditional, experience-dependent and passive response construction mode is changed into a new construction mode of data-driven, model-guided and active pre-control, and finally the safe, efficient and economic pile foundation construction goal is achieved.

[0106] The application will be further described below in combination with a typical engineering example, but the protection scope of the application is not limited thereto.

[0107] Regarding the engineering background: the bored pile construction project of a certain public and railway dual-purpose bridge engineering in Wenzhou, large-diameter reverse circulation bored piles (Φ1200 mm) are used, and the single pile hole forming depth is about 36 m. The site geological conditions are complex, and from top to bottom, there are miscellaneous fill layer, thick layer of marine silt soft soil, silty clay interbedded pebble lens and bearing stratum silty clay. Among them, the particle size distribution range of the interbedded pebble layer is wide (D50 is about 80-120 mm), the spatial distribution is uneven, the content is 15%-25%, and the construction risk is high. There have been many quality problems such as hole collapse, drill pipe sticking and sediment overrunning in history. The mud density and pump capacity window set by the traditional experience method is difficult to adapt to the dynamic construction environment under the joint action of "soft soil-granular interlayer".

[0108] In order to overcome the above technical problems, the "marine soft soil interbedded pebble RCD multi-field coupling simulation method" proposed in the application is applied, and the specific implementation process is as follows:

[0109] Based on the application, a three-dimensional calculation model is established, which integrates the CFD-DEM-FEM three-domain coupling strategy, and comprehensively depicts the mud flow, particle migration and hole wall stress response process. The marine soft soil adopts the disturbed state concept and Duncan-Chang combined model, the gravel interlayer is explicitly modeled as a discrete particle system, the drilling fluid is expressed by the Herschel-Bulkley constitutive model, and the ALE dynamic grid technology is used to handle the deformation and seepage boundary interaction. The input parameters of the model are provided with initial values by indoor rheological test and CU triaxial test, and the on-site measured pump pressure, backflow density, negative pressure, sediment thickness and other data are used for online calibration. Combined with the NSGA-II algorithm, multi-objective optimization is carried out, taking the hole wall safety factor, sediment thickness, unit footage energy consumption and construction efficiency as the objective function, and outputting the safe and constructible interval and recommended process curve of multiple parameter combinations.

[0110] The simulation results show that: in the gravel interlayer lens section, if the traditional parameters (mud density <1.18 g / cm 3 , pump capacity <1.0 m 3 / min) are used, the inlet shear stress decreases rapidly, the particle accumulation intensifies, resulting in a sediment thickness of more than 70 mm, and the hole cleaning failure probability is greater than 65%. Under the recommended parameters (ρ =1.19-1.21 g / cm 3 , Q=1.2-1.5 m 3 / min, N=12-18 rpm, ROP=0.35-0.55 m / min), the sediment thickness is controlled within 30-35 mm, the hole wall safety factor is not less than 1.65 throughout the process, and the pump pressure fluctuation is maintained within ±5%, and the construction process is stable and smooth.

[0111] In the actual construction process, a digital twin control platform is simultaneously deployed to realize real-time linkage with the simulation model. On-site online density meter, torque sensor, negative pressure sensor and sediment detection system are configured, and the model completes rolling prediction and parameter recalculation every 2 minutes. In the case of a sudden heavy rainfall causing the groundwater level to rise by 0.5 m, the model identifies a 15% reduction in effective stress of the hole wall, and the system automatically issues a "increase mud density by 0.02 g / cm 3 , reduce footage speed" control suggestion, successfully avoiding the risk of hole collapse.

[0112] Regarding the implementation effect: through the application of the application, after a complete cycle verification, the method successfully constructs 12 bored piles in the plot, the pile body integrity rate reaches 100%, the sediment thickness is stably controlled within 25-35 mm, the average single pile construction time is shortened by about 15% compared with the traditional method, the unit footage energy consumption is reduced by about 12%, and the comprehensive construction cost of each pile is saved by 920 yuan. The project saves more than 100,000 yuan in material and energy consumption costs, and the construction risk control level and intelligent level are greatly improved.

[0113] In summary, the case fully verifies the comprehensive value of the method of the application in the optimization of construction process parameters, safety risk prediction, digital control and other aspects of cast-in-place pile construction in complex marine soft soil and pebble stratum, and has good engineering adaptability, popularization and economic benefits. The method can be further applied to pile foundation and deep foundation pit engineering construction under similar coastal soft soil, lens interlayer or structural uneven stratum conditions, and provides key support for construction safety and high-quality piling.

[0114] The specific embodiments of the application are described in detail above with reference to the drawings, but the application is not limited to the described embodiments. For those skilled in the art, various changes, modifications, replacements and variations of the embodiments without departing from the principles and spirits of the application still fall within the protection scope of the application.

Claims

1. A multi-field coupled RCD simulation method for marine soft soil with pebbles, characterized in that, Includes the following steps: Step 1: Collect material parameters to form a multi-source input parameter table; Step 2: Construct a multi-field coupling model and establish a three-domain collaborative model of fluid domain, particle domain and soil domain in a unified solution platform; Step 3: Replicate the reverse circulation loop and load the actual working condition boundary to establish a mud closed loop consistent with the actual construction, and set the drilling tool working condition to be loaded in the form of time history. Step 4: Achieve strong coupling calculations among the fluid domain, particle domain, and soil domain using a multi-scale and sub-cycle strategy. The multi-scale and sub-cycle strategy specifically includes: Step 4.1: Use 3D CFD computational fluid dynamics to refine the inlet and near-wall regions at the bottom of the orifice, and use axisymmetric or simplified regions for the far field; 3D CFD is coupled with one-dimensional transient pipe network equations through boundary flux and pressure feedback. Step 4.2: Based on the voxelized mesh, realize the bidirectional momentum exchange between CFD and DEM discrete element method. The effect of particles on fluid is reflected by drag, lift or additional mass force; the effect of fluid on particles is reflected by local velocity gradient and turbulent kinetic energy; set the DEM substep size to meet the particle collision stability condition, and synchronize it with the CFD step size according to the subcycle ratio of 10 to 100. Step 4.3: The coupling of FEM and CFD is achieved by exchanging pore pressure and wall shear stress at the pore wall interface, and the numerical oscillation during strong coupling is suppressed by using a relaxation iteration method. Step 4.4: Determine the time step size based on the CFL Courant-Friedrich-Levi condition, and trigger local adaptive mesh and adaptive time step adjustment in the high gradient region of the fluid inlet to capture the migration of the sediment core and resuspension boundary; Step 5: Use on-site monitoring to update and recalibrate the parameters and model online; Step 6: Based on the effective stress path and local flow field, criterion for borehole wall stability and sediment risk is given; the borehole wall stability criterion is based on strength reduction or the Mohr-Coulomb safety factor F. s The core of this approach includes: calculating the effective stress path time history response of the soil borehole wall by coupling the changes in water head and the seepage consolidation process within the borehole; and based on the effective stress path time history response, identifying the safety factor F caused by at least one unsteady-state operation, such as pump shutdown, pump startup, raising or lowering of the drill string, or fluctuations in water level. s Behavior that crosses the warning threshold; when the safety factor F s When the value approaches or falls below the preset warning threshold, the output hole wall risk index R is increased. c And pinpoint the deeper aspects of the problem; The specific criteria for determining sediment risk include: calculating the shear stress on the wall and bed surface within the three-dimensional local flow field at the borehole inlet; comparing the shear stress with the critical initiation stress of the particles to determine the sediment deposition and resuspension zones; updating the sediment thickness and planar morphology based on the bed evolution equation and local mass conservation; and simultaneously calculating and comparing the available net positive suction head (NPSH). avail With required net positive suction head (NPSH) req When NPSH avail < NPSH req At that time, determine the risk of cavitation in the circulation loop; if the margin is insufficient, adjust the pump flow rate Q or the constant pressure boundary H accordingly. a And trigger a short-term pulsed cleaning operation; Step 7: Conduct multi-objective optimization with safety, efficiency and energy consumption as the goals, and provide construction parameter windows and recommended process curves; Step 8: Output control indicator results and early warning rules to achieve linkage parameter adjustment; Step 9: Archive the calibrated parameters and model into the engineering parameter library for engineering reuse.

2. The RCD multi-field coupled simulation method for marine soft soil with pebbles according to claim 1, characterized in that: In step 1, the collected material parameters refer to the collection and supplementation of stratigraphic, geotechnical, hydrological, mud, and equipment parameters obtained from experiments, followed by consistency verification and outlier removal. The stratigraphic, geotechnical, hydrological, mud, and equipment parameters specifically include: In terms of strata, the layering, thickness variation, and spatial distribution of gravel lenses in marine soft soil were clarified based on exploration boreholes, core sampling, and in-situ testing. Gravel was analyzed by sieving and imaging methods to statistically determine particle size distribution, flatness factor, volume fraction, and lens thickness-connectivity index. In terms of geotechnical engineering, state parameters, strength parameters, seepage-consolidation parameters, historical parameters, and disturbance sensitivity parameters were obtained through indoor triaxial consolidated undrained creep and permeability tests. In terms of hydrology, establish the time-varying boundary relationship between groundwater level and tides; Regarding the mud, the target range and temperature dependence of mud density, yield stress, consistency, index and thixotropic recovery coefficient were obtained through formula pre-selection and rotary rheometer testing. Regarding equipment, plot the orifice diameter, annular width, inner pipe diameter, inlet opening ratio, screen slot size, ground pipe network configuration, and friction loss coefficients (both along and locally). Record the pump curve, rated power, and net positive suction head (NPSH) parameters, including the available net positive suction head (NPSH). The required net positive suction head (NPSH) is calculated based on the system static head, liquid level, friction loss and local loss, and saturated vapor pressure of the medium. The control criterion is given by the pump characteristic curve. .

3. The RCD multi-field coupled simulation method for marine soft soil with pebbles according to claim 1, characterized in that: In step 2, the fluid domain, particle domain, and soil domain are weakly coupled or strongly coupled in different regions through shared geometric or topological relationships and exchange interface quantities. Simultaneously, an immersion boundary or an arbitrary Lagrange-Euler method ALE dynamic mesh is introduced at the borehole wall to describe the feedback of minute deformations of the borehole wall to the near-wall flow field. The fluid domain is defined in relation to mud rheology, which follows the Herschel–Bulkley constitutive model. ， In the numerical implementation, tensor form is used: ， In the formula, For yield stress, K For consistency, For shear stress tensor, For equivalent shear rate, D Let n be the strain rate tensor; n is the flow index. Apparent viscosity; The particle domain employs the discrete element method for pebbles and drill cuttings, and allows activation of the bonding-debonding mechanism in the high shear zone; The soil domain uses the modified Cambridge model MCC or the dual-structure Cambridge damage coupling model DSC+DC, employing displacement-pore pressure or displacement-relative displacement consolidation-seepage coupling. A Brinkman-Darcy transition layer is set in the near field of the pore wall to describe weak permeability or shear transition.

4. The RCD multi-field coupled simulation method for marine soft soil with pebbles according to claim 1, characterized in that: Step 3, replicating the reverse circulation loop, involves constructing a reverse circulation boundary of annular replenishment-orifice bottom inlet-inner tube negative pressure return. The annular end is set with a constant pressure or constant flow boundary, and friction loss or local loss is taken into account. Geometric constraints of opening ratio and screen size are applied to the orifice bottom inlet region, using local loss and inlet contraction effects to reflect shear enhancement and vortex formation. A negative pressure slag suction boundary is applied to the inner tube end, coupled with pump curves and NPSH constraints. The surface end slurry return condition uses liquid level or separation device back pressure. The drilling tool operating conditions are loaded in time-history form, including rotational speed, drilling pressure, drilling speed and pump flow rate, and allow the provision of short rhythmic sequences of drilling-cleaning-measuring and transient states of pump stop or start.

5. The RCD multi-field coupled simulation method for marine soft soil with pebbles according to claim 2, characterized in that: In step 5, the parameters include rheological parameters and soil parameters. The specific steps for updating and calibrating the rheological and soil parameters include: calibrating the rheological parameters based on the rheological test results obtained in step 1, fitting the initial parameter values; and updating them using Kalman or Bayesian filtering based on the backflow density or viscosity measured online in the field; calibrating the soil parameters based on the soil foundation parameters in step 1 as priors, and completing the recalibration by obtaining the posterior estimates of the modified Cambridge Model MCC or the Disturbance State Conceptual Model DSC parameters through short pile tests or borehole response inversion. The specific steps for updating and calibrating the model include: using unscented Kalman filtering, Bayesian updating, or particle filtering to update key parameters and boundary perturbations online; triggering parameter recalibration and simulation restart when the residual exceeds a threshold.

6. The RCD multi-field coupled simulation method for marine soft soil with pebbles according to claim 1, characterized in that, Step 7 specifically includes: selecting mud density, yield stress, pump rate, rotation speed, drilling pressure, footage rate, and inlet opening ratio as decision variables; using NSGA-II or equivalent multi-objective evolutionary algorithm to search for the Pareto front; using a surrogate model to accelerate evaluation during the optimization process, outputting a contour map of the workable parameter window and a recommended process curve, and generating a robustness comparison for different gravel contents or particle size combinations.

7. The RCD multi-field coupled simulation method for marine soft soil with pebbles according to claim 1, characterized in that, In step 8, the control index results include parameter windows and control indices, sensitivity matrices and recommended curves under boundary scenarios; the early warning rules identify signs of entry blockage, stuck drill, and hole collapse risks by setting joint features.

8. The RCD multi-field coupled simulation method for marine soft soil with pebbles according to claim 1, characterized in that, Step 9 specifically includes: archiving the calibrated material parameters, equipment characteristics, optimization results, and sensitivity spectra into the engineering parameter library, forming searchable entries in the formation type-pebble statistics-equipment configuration-optimal parameter window, exporting a standardized calculation report, and storing it as a readable engineering template; subsequent engineering processes are based on the readable engineering template to quickly initialize and calibrate online by replacing parameters.

Images