Construction methods for slope protection of open-cut tunnels

Abstract

This invention discloses a construction method for slope protection in open-cut tunnels, belonging to the field of tunnel engineering construction and slope protection technology. Addressing the problems of insufficient monitoring coverage leading to delayed slippage warnings, static support parameters causing anchorage failure, and poor coordination between grouting and drainage exacerbating slope instability in traditional methods, this invention proposes pre-embedding distributed fiber optic sensors on both sides of the tunnel axis at the excavation front. Real-time monitoring is used to activate the hydraulic servo anchor support system when the strain increment reaches 0.04-0.06% or the seepage rate reaches 4-6 mm / h. The initial prestress of the anchors is 45-55 kN and dynamically adjusted. After each stage of excavation, high-pressure grouting is used to seal the slope surface, with nano-silica incorporated into the grouting holes. Simultaneously, a gravel filter layer and a blind ditch with drainage pipes are laid. Adjustment is terminated when the displacement rate is <0.5 mm / d and the seepage rate is <2 mm / h for three consecutive days. This method is suitable for slope stability control in water-rich strata and weak interlayer areas, and can simultaneously achieve deformation monitoring, dynamic support, and seepage control.

Description

Technical Field

[0001] This invention belongs to the field of tunnel engineering construction and slope protection technology, specifically relating to a construction method for slope protection of open-cut tunnels. Background Technology

[0002] In open-cut tunnel engineering, slope stability control is a core issue of construction safety. Traditional slope protection methods often use point displacement monitoring combined with static anchor bolt support, but there are significant shortcomings in practical applications: First, conventional monitoring methods (such as inclinometers and earth pressure cells) have low deployment density and limited coverage, making it difficult to capture local deformation and seepage anomalies in weak interlayer areas, resulting in delayed early warning of slippage risks; Second, anchor bolt support often adopts a fixed prestressed design, which cannot be adjusted in real time according to the dynamically changing stress and seepage fields during excavation, easily leading to insufficient support stiffness or excessive restraint, causing secondary slope deformation; In addition, grouting reinforcement technology often creates reinforcement blind spots due to insufficient grout penetration distance and poor matching with seepage channels, while the gradation design and water flow capacity maintenance of drainage blind ditches are insufficient, easily leading to rapid decline in drainage efficiency due to filter layer siltation.

[0003] The root causes of the above problems are: 1) Insufficient spatial resolution of the monitoring system, making it difficult to accurately locate potential slip surfaces and seepage paths; 2) Lack of a dynamic coupling mechanism between support parameters and geological response, leading to a disconnect between mechanical control and deformation development; 3) Failure to coordinate the performance of grouting materials with the evolution of the seepage field, resulting in a lack of guarantee for the drainage structure's anti-clogging ability and long-term service performance. Especially in areas with water-rich sand layers or weak interlayers, seepage erosion accelerates the weakening of the soil structure, while traditional grouting materials struggle to effectively penetrate and consolidate under dynamic water conditions, further exacerbating the risk of slope instability. Existing technologies attempt to improve the situation by increasing the number of monitoring points or raising the grouting pressure, but the former significantly increases construction costs, and the latter may damage the original structure of the soil and rock, neither of which fundamentally solves the problem of slope stability control under the coupling of multiple physical fields. Summary of the Invention

[0004] To address the problems of insufficient monitoring coverage, static support parameters, and poor coordination between grouting and drainage in traditional slope protection methods, this invention provides a construction method for slope protection in open-cut tunnels. This method is based on real-time monitoring by distributed fiber optic sensors, dynamic support by hydraulic servo anchors, and coordinated control of high-pressure grouting and drainage blind ditches to control slope stability.

[0005] To achieve these objectives and other advantages of the present invention, the present invention provides a method for slope protection construction of open-cut tunnels, comprising the following steps:

[0006] 1) Before the slope excavation, pre-embed distributed optical fiber sensors on both sides of the slope along the tunnel axis. The deployment depth of the distributed optical fiber sensors is 0.6-0.7 times the design height of the slope. The horizontal spacing between adjacent sensors is 2.5-3.5m and the vertical spacing is 1.8-2.2m. The sensors are connected to the ground data acquisition terminal through lines, and the response delay is ≤3 seconds.

[0007] 2) When the distributed fiber optic sensor detects that the slope strain increment reaches 0.04-0.06% (based on historical landslide data statistics) and the seepage rate reaches 4-6 mm / h (critical value of dynamic water pressure), the hydraulic servo anchor support system is activated. The installation angle of the hydraulic servo anchor is 15°-25°, the anchor length is 1.1-1.4 times the excavation depth, the anchor spacing is 1.8-2.2m×1.8-2.2m grid arrangement, the initial prestress is set to 45-55kN, and the step size is dynamically adjusted according to the real-time monitored strain rate of 0.01-0.03% / h, with a step size of 5-15kN.

[0008] 3) After each slope is excavated to the design elevation, a high-pressure grouting machine is used to seal the slope surface with grout. The grouting pressure is dynamically adjusted according to the seepage rate: when the seepage rate is 4-6 mm / h, the pressure is 0.8-1.0 MPa; when the seepage rate is >6 mm / h, the pressure is increased to 1.2-1.4 MPa. The hole spacing of the grouting holes is 1.3-1.7 m, and the grouting material is silicate cement slurry mixed with 2-8% nano silica.

[0009] 4) Lay longitudinal drainage blind ditches along the toe of the slope, lay a three-grade gravel filter layer at the bottom of the ditch, install HDPE perforated drainage pipes on top of the filter layer, and wrap the pipes with double layers of geotextile.

[0010] 5) Continuously monitor slope deformation data using distributed fiber optic sensors. When the displacement rate is <0.5mm / d and the seepage rate is <2mm / h for 3 consecutive days, terminate the hydraulic servo anchor prestress adjustment in stages: in the first stage, reduce the step size to 5kN and maintain it for 24 hours; in the second stage, completely stop the adjustment and continue monitoring for 48 hours.

[0011] Preferably, the construction method for slope protection of the open-cut tunnel is as follows:

[0012] 1) In areas with weak interlayers, the deployment depth of distributed fiber optic sensors is increased to 0.8-0.9 times the design height of the slope, the horizontal spacing is reduced to 2.0-3.0m, and the vertical spacing is reduced to 1.5-2.0m;

[0013] 2) When pre-embedding distributed fiber optic sensors, stainless steel fixing clips are installed every 0.5m along the sensor axis, and an epoxy resin bonding layer with a thickness of 10-15mm is filled between the clips and the slope soil.

[0014] 3) The ground data acquisition terminal performs moving average filtering on the monitoring data of the distributed fiber optic sensors. The filtering window time is 10-15 minutes, and outliers outside the ±3σ range are removed.

[0015] 4) When three adjacent distributed fiber optic sensors on the same monitoring profile simultaneously detect a strain increment ≥ 0.04% and a seepage rate ≥ 4 mm / h, the start-up conditions of the hydraulic servo anchor bolt support system are triggered.

[0016] Preferably, the construction method for slope protection of the open-cut tunnel is as follows:

[0017] 1) During the prestress adjustment process of the hydraulic servo anchor, when the distributed fiber optic sensor detects that the slope strain rate exceeds 0.03% / h, the adjustment step size is increased to 12-15kN; when the strain rate is less than 0.01% / h, the adjustment step size is decreased to 5-8kN.

[0018] 2) The anchoring section of the hydraulic servo anchor is equipped with a dual hydraulic cylinder cooperative control system. The main hydraulic cylinder applies axial tension, and the auxiliary hydraulic cylinder applies lateral pressure at an angle of 30°-45° to the rock stratum. The pressure ratio of the main and auxiliary hydraulic cylinders is dynamically controlled by the real-time calculated rock stratum strain gradient, and the range is 1:0.3-0.5.

[0019] 3) The surface of the anchor rod is machined with a continuous thread structure, the thread pitch is 15-20mm, the thread depth is 2-3mm, the thread area covers the entire length of the grouting section, and grout penetration holes with a diameter of 3-5mm are opened at the root of the thread, with a hole spacing of 50-80mm.

[0020] 4) When the anchor head pressure sensor detects that the contact stress fluctuation exceeds 20% of the initial value, the anchor spacing densification mechanism is automatically triggered, reducing the grid spacing to 1.5-1.8m × 1.5-1.8m.

[0021] Preferably, the construction method for slope protection of the open-cut tunnel is as follows:

[0022] 1) The ground data acquisition terminal inputs the strain increment, seepage rate and standard deviation of contact stress fluctuation at the anchor head monitored by the distributed optical fiber sensor into a Bayesian network model trained based on historical slope engineering data. The model input nodes are defined as: strain increment → seepage rate → contact stress fluctuation → instability probability. The mean square error of the output instability probability value is ≤0.005.

[0023] 2) When the instability probability output by the Bayesian network is ≥85% or the real-time strain rate is <0.05% / h, the following operations are triggered synchronously:

[0024] a. The adjustment step of the hydraulic servo anchor bolt is increased to 12-15kN (each adjustment ≤5kN, interval ≥10 minutes).

[0025] b. The grouting pressure of the high-pressure grouting machine is increased to 1.2-1.4 MPa;

[0026] c. The opening ratio of HDPE perforated drainage pipes is increased to 6-8%, and the hole diameter is adjusted to 8-10mm;

[0027] 3) The Bayesian network model is trained based on historical slope engineering data. The input parameters include strain increment, seepage rate, and standard deviation of contact stress fluctuation. The output is the instability probability value.

[0028] 4) When the instability probability is <60% and the displacement rate is <0.5mm / d for 3 consecutive days, restore the anchor bolt step length to 8-12kN, reduce the grouting pressure to 0.8-1.0MPa, and adjust the drainage pipe opening rate back to 3-5%.

[0029] Preferably, when the Bayesian network model has an instability probability of ≥95% or communication is interrupted, it automatically switches to the preset static support mode (anchor bolt step length is fixed at 10kN, grouting pressure is 1.0MPa, and drainage pipe opening rate is 5%).

[0030] Preferably, the construction method for slope protection of the open-cut tunnel is as follows:

[0031] 1) In the preparation process of the silicate cement slurry, nano-silica is incorporated in two stages:

[0032] a. In the first stage, 1-2% by mass of nano-silica and water-reducing agent are added together to the mixing water and ultrasonically dispersed;

[0033] b. In the second stage, add the remaining 1-2% nano-silica, mix it with silicate cement, and continue stirring.

[0034] 2) A medium-coarse sand transition layer is added to the bottom of the gravel filter layer of the drainage blind ditch, and geotextile is laid between the sand layer and the gravel layer.

[0035] 3) Within 24 hours after grouting, backwash the drainage blind ditch with a water pressure of ≤0.4MPa and a flushing duration of 10-20 minutes until the drainage pipe has a water flow capacity of ≥5L / (m·s).

[0036] 4) When the hydraulic servo anchor bolt adjustment step length is ≥12kN, 0.2-0.5% of early strength agent is added to the grout simultaneously to shorten the initial setting time of the grout to 45-60 minutes.

[0037] Preferably, the construction method for slope protection of the open-cut tunnel is as follows:

[0038] 1) During the slope excavation process, the temperature data of the rock strata is monitored in real time by distributed optical fiber sensors, and the data is fused and analyzed with strain and seepage data to establish a temperature-strain-seepage coupling model.

[0039] 2) When the temperature rises abnormally (ΔT≥3℃ / 24h, baseline value is the average value of monitoring over the 30 days prior to construction ±1℃) and is accompanied by a strain rate >0.02% / h, the following emergency measures shall be triggered:

[0040] a. Increase the hydraulic servo anchor bolt adjustment step size to 15kN, and refine the anchor bolt grid spacing to 1.5m×1.5m;

[0041] b. Add 0.5-1.0% of an expanding agent to the grout so that the volume expansion rate of the grout after solidification is ≥5%;

[0042] c. The opening rate of HDPE perforated pipes for drainage blind drains is increased to 8-10%, and the hole diameter is increased to 10-12mm;

[0043] 3) The temperature-strain-seepage coupling model is constructed based on finite element analysis. The input parameters include the thermal conductivity of the rock mass, porosity, and radius of curvature of the seepage path. The output is the potential thermally induced slip risk level.

[0044] 4) After the emergency measures are implemented, if the temperature recovers to the baseline value ±1℃ and the strain rate is <0.01% / h within 48 hours, the anchor bolt step length, grouting pressure and drainage pipe opening ratio shall be gradually restored to the initial set values.

[0045] Preferably, the construction method for slope protection of the open-cut tunnel is as follows:

[0046] 1) After construction, based on long-term monitoring data from distributed fiber optic sensors, a graded early warning mechanism for slope stability is established: when the displacement rate is stable at 0.3-0.5 mm / d for 7 consecutive days, it is judged as a Level 1 warning, and a weekly manual inspection is initiated to check whether the prestress loss of the anchor bolts exceeds 10%; when the daily displacement rate is >0.6 mm / d or the seepage rate is >6 mm / h, it is judged as a Level 2 warning, and local grouting and anchor bolt prestress compensation are immediately implemented; when the displacement rate is >0.8 mm / d and the seepage rate is >8 mm / h for 3 consecutive days, it is judged as a Level 3 warning, and the anchor bolt grid is fully densified to 1.5m×1.5m, and temporary drainage wells are added.

[0047] 2) The following parameters need to be recorded during manual inspection: the fluctuation range of contact stress at the anchor head, the width of cracks in the grouting body, and the water-carrying capacity of the drainage blind channel. The subsequent construction parameters should be dynamically adjusted based on the inspection results.

[0048] 3) The grouting material is quick-setting silicate cement grout with an initial setting time of ≤1 hour, and 4-6% nano silica is added. The grouting hole positions are preferentially selected from the nodes in the area with the largest strain increment.

[0049] 4) Warning cancellation conditions: After maintenance, the displacement rate is <0.3mm / d and the seepage rate is <2mm / h for 7 consecutive days, and the prestress loss of the anchor bolt is ≤5%.

[0050] Preferably, the construction method for slope protection of the open-cut tunnel is as follows:

[0051] 1) Dynamically adjust anchor bolt support parameters according to different geological conditions:

[0052] In water-rich sandy areas (seepage rate of about 7.2 mm / h): the length of the anchor bolts is increased to 1.3-1.5 times the excavation depth, the initial prestress is increased to 55-65 kN, and the grouting pressure is set to 1.2-1.5 MPa;

[0053] In weak interlayer areas: the anchor mesh spacing is increased to 1.5m×1.5m, the grouting hole depth is increased to 0.6-0.7 times the design height of the slope, and 6-8% nano silica is added to improve the impermeability of the grout;

[0054] In areas with well-developed rock fractures: the anchor bolt installation angle is adjusted to be orthogonal to the fracture direction (error ≤ ±5°), and the spacing of the grout penetration holes is reduced to 30-50mm;

[0055] 2) During the rainy season or periods of rising groundwater levels, implement the following preventative maintenance measures:

[0056] Increase the opening rate of the HDPE perforated pipe for drainage blind drain to 8-10%, enlarge the hole diameter to 10-12mm, and perform backflushing once a week (water pressure 0.6-0.8MPa, lasting 20-30 minutes).

[0057] The frequency of anchor bolt prestress monitoring has been increased from once a day to three times a day. When a single monitoring shows a prestress loss of ≥5%, a compensation step of 5-8kN is immediately applied.

[0058] A temporary diversion channel (200-300mm deep, slope ≥5%) is added inside the intercepting ditch at the top of the slope to prevent surface runoff from seeping into the slope.

[0059] 3) The matching relationship between the amount of nano-silica and the grouting pressure is as follows:

[0060] When the grouting pressure is 0.8-1.0 MPa, 4-5% nano-silica is added; when the grouting pressure is 1.2-1.5 MPa, 6-8% nano-silica is added.

[0061] The present invention has at least the following beneficial effects:

[0062] 1. The distributed fiber optic sensor of this invention, combined with moving average filtering, achieves a monitoring data integrity rate exceeding 98%, reduces the early warning response time to within 5 minutes, and controls the missed detection rate to within 3%. Compared to traditional point sensors, the landslide early warning accuracy is improved by 40%.

[0063] 2. The hydraulic servo anchor bolt of the present invention is based on the step-by-step adjustment of strain rate (5-15kN) and the coordinated control of dual hydraulic cylinders (pressure ratio 1:0.3-0.5), which enables the support stiffness to match the stratum deformation in real time, reduces the anchor bolt breakage rate from 12% to below 5%, and reduces the interface slippage to within 0.5mm.

[0064] 3. The phased incorporation of nano-silica in this invention enables the slurry to achieve a compressive strength of 20.8 MPa (ASTM C109) after 28 days; the three-graded gravel filter layer and dynamic porosity adjustment (3-10%) ensure that the drainage efficiency is stable at over 5 L / (m·s), reducing the risk of seepage erosion by 60%.

[0065] 4. The Bayesian network of this invention integrates multiple parameters (strain, seepage, temperature) to achieve linkage control, reducing the false trigger rate from 15% to 5%; the graded early warning mechanism (displacement 0.3-0.8mm / d) significantly enhances the adaptability of the project to weak interlayers, water-rich layers and seasonal changes.

[0066] 5. The dynamic parameter adjustment of this invention reduces ineffective grouting by 30% and saves 1.2 tons of cement per 100 linear meters; the intelligent termination conditions (displacement + seepage + contact stress) reduce the need for concealed maintenance, reduce the total cost by 25%, and improve construction efficiency by 20%.

[0067] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0068] Figure 1 This is a flowchart of a construction method for slope protection of open-cut tunnels according to the present invention. Detailed Implementation

[0069] The present invention will be further described in detail below with reference to embodiments, so that those skilled in the art can implement it based on the description.

[0070] like Figure 1 As shown, a method for slope protection construction of an open-cut tunnel according to the present invention includes the following steps:

[0071] 1) Before the slope is excavated, distributed optical fiber sensors are pre-embedded on both sides of the slope along the tunnel axis. The depth of the distributed optical fiber sensors is 0.6-0.7 times the design height of the slope. The horizontal spacing between adjacent sensors is 2.5-3.5m and the vertical spacing is 1.8-2.2m. The sensors are connected to the ground data acquisition terminal.

[0072] 2) When the distributed fiber optic sensor detects that the slope strain increment reaches 0.04-0.06% and the seepage rate reaches 4-6 mm / h, the hydraulic servo anchor support system is activated. The installation angle of the hydraulic servo anchor is 15°-25°, the anchor length is 1.1-1.4 times the excavation depth, the anchor spacing is a grid arrangement of 1.8-2.2m×1.8-2.2m, the initial prestress is set to 45-55kN, and the step size is dynamically adjusted according to the real-time monitored strain rate of 0.01-0.03% / h, with a step size of 5-15kN.

[0073] 3) After each slope is excavated to the design elevation, a high-pressure grouting machine is used to seal the slope surface with grout. The grouting pressure is dynamically adjusted according to the seepage rate: when the seepage rate is 4-6 mm / h, the pressure is 0.8-1.0 MPa; when the seepage rate is >6 mm / h, the pressure is increased to 1.2-1.4 MPa. The hole spacing of the grouting holes is 1.3-1.7 m, the row spacing is 0.866 times the hole spacing, the drilling direction is at an angle of 5°-10° to the slope normal, and the grouting material is silicate cement slurry mixed with 2-8% nano silica.

[0074] 4) Lay longitudinal drainage blind ditches along the toe of the slope, with cross-sectional dimensions of 280mm×280mm to 320mm×320mm. Lay a three-grade gravel filter layer at the bottom of the ditch (lower layer 8-12mm, middle layer 12-18mm, upper layer 18-22mm). Install HDPE perforated drainage pipes (diameter 140-160mm, opening rate 3-8%) on top of the filter layer, and wrap the pipes with double layers of geotextile.

[0075] (5) Continuously monitor slope deformation data using distributed fiber optic sensors. When the displacement rate is <0.5 mm / d and the seepage rate is <2 mm / h for 3 consecutive days, terminate the hydraulic servo anchor prestress adjustment in stages: the first stage reduces the step size to 5 kN and maintains it for 24 hours; the second stage completely stops the adjustment and continues to monitor for 48 hours.

[0076] Specifically, distributed fiber optic sensors can be pre-embedded along both sides of the tunnel axis slope, with a deployment depth of 0.6-0.7 times the designed slope height. For example, when the slope height is 10 meters, the burial depth can be 6-7 meters. The horizontal spacing can be 2.5-3.5 meters, and the vertical spacing can be 1.8-2.2 meters. The sensors can be MOI SM125 type, connected to a ground data acquisition terminal, such as the Campbell CR6. The sensors can be installed 0.5-1.0 meters from the slope edge, and the data acquisition terminal can be deployed in a temporary monitoring station at the top of the slope. Monitoring data includes strain increment and seepage rate. The strain increment threshold can be set to 0.04-0.06%, and the seepage rate threshold can be set to 4-6 mm / hour. Inclined tube monitoring methods have limited coverage. This implementation scheme achieves continuous monitoring of the entire slope using distributed fiber optic sensors. The parameter setting method is based on historical landslide data analysis; for example, in a certain area, the strain increment before 10 landslides all exceeded 0.04%.

[0077] The installation angle of the hydraulic servo anchor bolt can be selected from 15-25 degrees, and the anchor bolt length can be selected as 1.1-1.4 times the excavation depth. For example, when the excavation depth is 8 meters, the anchor bolt length can be 8.8-10.4 meters. The anchor bolt spacing can be arranged in a grid of 1.8-2.2 meters × 1.8-2.2 meters, and the initial prestress can be set to 45-55 kN. The dynamic adjustment step size can be selected as 8-12 kN, for example, when the strain rate is 0.02% / hour, the step size is adjusted to 10 kN. The hydraulic system can use an Enerpac RC-106 pump station, and the anchor bolt body can be made of HRB400 grade threaded steel. The anchor bolt can be driven in at an angle perpendicular to the slope, with the anchoring section penetrating into the moderately weathered rock layer. The hydraulic control unit can be installed on the side of the construction platform. Compared with static anchor bolt support, this implementation scheme adjusts the prestress through real-time data. In functional testing, the adjustment delay needs to be controlled within 3 minutes, for example, by simulating input strain data to verify the response time.

[0078] The grouting pressure can be selected from 0.7 to 1.3 MPa, with the specific value adjusted according to the seepage rate: 0.8 to 1.0 MPa for a seepage rate of 4-6 mm / h; and 1.2 to 1.4 MPa for a seepage rate exceeding 6 mm / h. The hole spacing of the grouting holes can be 1.3 to 1.7 meters. The grouting material can be silicate cement slurry with a water-cement ratio of 0.4-0.5, such as Conch PO 42.5 cement, and mixed with 2-8% nano-silica, such as Evonik AEROSIL 200.

[0079] The cross-sectional dimensions of the drainage ditch can be selected from 280 mm × 280 mm to 320 mm × 320 mm. A gravel filter layer with a thickness of 90-110 mm and a gravel particle size of 8-22 mm can be laid at the bottom of the ditch. HDPE perforated drainage pipes with a diameter of 140-160 mm, such as Liansu PE100 grade pipes, can be installed on top of the filter layer. The pipes can be wrapped with double-layer geotextile, with a specification of 200 g / m². The drainage pipes can be laid along the toe of the slope, with a slope of 0.5%. Compared to traditional grouting processes, this implementation scheme dynamically adjusts the grouting pressure through the seepage rate. After grouting, the grout diffusion radius reaches more than 1.5 meters, and the filling rate of cracks exceeds 85%.

[0080] Termination of adjustment requires that the displacement rate be less than 0.5 mm / day for three consecutive days, and the seepage rate be less than 2 mm / hour. For example, if the continuous monitoring data for a slope are 0.4 mm / day, 0.3 mm / day, and 0.4 mm / day, and the seepage rate is 1.8 mm / hour, 1.5 mm / hour, and 1.7 mm / hour, it is considered to be in a stable state. The phased termination operation includes the first phase of reducing the anchor bolt step length to 5 kN and maintaining it for 24 hours, and the second phase of completely stopping adjustment and continuing monitoring for 48 hours. The fluctuation range of contact stress can be calculated using the slip standard deviation; for example, a standard deviation exceeding 0.8 kN within 6 hours is considered abnormal.

[0081] Compared to a single displacement rate threshold, this implementation plan combines seepage rate and contact stress data. In experimental verification, the displacement rate of a certain project stabilized below 0.3 mm / day within 48 hours after the adjustment was terminated.

[0082] Preferably, in the event of a fiber breakage in the distributed fiber optic sensor or a failure of the data acquisition terminal, a backup monitoring module is activated: wireless tilt sensors (accuracy ±0.1°) and piezometers (range 0-100kPa) are deployed at 20m intervals along the slope, and the data is transmitted to the emergency control terminal via LoRa to trigger the preset static anchor support plan (prestress is 80% of the design value).

[0083] Taking the slope engineering project at section K12+480 of a certain tunnel as an example, the slope is designed to be 12m high, the stratum is silty clay with sand, and the excavation depth is 9m. The specific steps are as follows:

[0084] (1) Deployment of distributed fiber optic sensors

[0085] A casing drilling rig was used to form a hole (100mm diameter), and a PVC retaining sleeve (2mm wall thickness) was inserted. Distributed fiber optic sensors (MOI sm125) were pre-embedded on both sides of the tunnel axis slope at a depth of 7.8m (0.65 times the slope height), with a horizontal spacing of 3.0m and a vertical spacing of 2.0m. After the sensors were installed, they were calibrated on-site using a standard strain beam (accuracy ±0.005%), and a constant temperature water bath (temperature fluctuation ±0.5℃) was used to compensate for temperature drift.

[0086] (2) Dynamic support of hydraulic servo anchor bolts

[0087] The anchor bolt is installed at a 20° angle, with a length of 10.8m (1.2 times the excavation depth), a grid spacing of 2.0m × 2.0m, and an initial prestress of 50kN. When a strain increment of 0.05% and a seepage rate of 5.2mm / h are detected, the anchor bolt step length is adjusted to 10kN. The hydraulic system (Enerpac RC-106) is equipped with a backup accumulator, which can maintain the prestress for ≥30 minutes in the event of a power outage.

[0088] (3) High-pressure grouting and construction of drainage blind ditches

[0089] The grouting pressure was 1.0 MPa, the hole spacing was 1.5 m, and the grout containing 3% nano silica (Evonik AEROSIL200) (water-cement ratio 0.45, curing conditions 20℃±2℃, humidity ≥95%) was tested according to ASTM C109 standard. The 28-day compressive strength was 20.8 MPa (compared to 16.5 MPa for the control group without nanomaterials). At a grouting pressure of 1.0 MPa and a seepage rate of 4-6 mm / h, the diffusion radius of the grout containing 3% nano silica was 1.5-3.5 m (measured by dye tracer method), while the diffusion radius of the control group without nanomaterials was 0.9-1.2 m. The drainage blind ditch had a cross-section of 300 mm × 300 mm, a gravel filter layer thickness of 100 mm, and HDPE drainage pipes (Liansu PE100) with a diameter of 150 mm, an opening rate of 6%, and a water flow capacity of 3 L / (m·s).

[0090] (4) Termination of support adjustment

[0091] The displacement rate was 0.4 mm / d, 0.3 mm / d, and 0.4 mm / d for three consecutive days, and the seepage rate dropped to 1.8 mm / h. In the first stage, the anchor bolt step length was reduced to 5 kN and maintained for 24 hours. In the second stage, the adjustment was completely stopped, and the displacement rate was stabilized at <0.5 mm / d within 48 hours.

[0092] This implementation method achieves full-slope monitoring through distributed fiber optic sensors, reducing the early warning response time to within 5 minutes and the missed detection rate to below 3%. Dynamic adjustment of hydraulic servo anchors matches the support stiffness with ground deformation, reducing the anchor breakage rate from 12% in traditional methods to below 5%. Linked control of grouting pressure and seepage rate reduces reinforcement blind zones by 80%, and the drainage capacity of blind channels remains stable at over 5 liters / (m·s). Multi-parameter termination conditions reduce the false alarm rate from 20% in traditional single-indicator methods to below 8%. This effectively solves the technical problems of insufficient monitoring coverage leading to delayed slippage early warnings in traditional slope protection methods; the inability of static anchor support to dynamically match ground deformation; and poor coordination between grouting and drainage systems resulting in reinforcement blind zones and low drainage efficiency.

[0093] Furthermore, in another embodiment, the present invention provides a method for slope protection construction of open-cut tunnels:

[0094] (1) In the weak interlayer area, the deployment depth of the distributed optical fiber sensor is increased to 0.8-0.9 times the design height of the slope, the horizontal spacing is reduced to 2.0-3.0m, and the vertical spacing is reduced to 1.5-2.0m;

[0095] (2) When pre-embedding the distributed optical fiber sensor, stainless steel fixing buckles (SUS304, size 50mm×20mm×2mm) are installed every 0.5m along the sensor axis. An epoxy resin bonding layer (Huntsman Araldite AW 106 type) with a thickness of 10-15mm is filled between the buckle and the slope rock and soil.

[0096] (3) The ground data acquisition terminal performs moving average filtering on the monitoring data of the distributed optical fiber sensor, with a filtering window time of 10-15 minutes, and removes outliers outside the ±3σ range;

[0097] (4) When three adjacent distributed fiber optic sensors on the same monitoring profile simultaneously detect a strain increment ≥ 0.04% and a seepage rate ≥ 4 mm / h, the start-up conditions of the hydraulic servo anchor bolt support system are triggered.

[0098] Specifically, in areas with weak interlayers, the deployment depth of distributed fiber optic sensors can be selected as 0.8-0.9 times the designed slope height. For example, when the slope height is 12 meters, the burial depth can be 9.6-10.8 meters. The horizontal spacing can be 2.0-3.0 meters, and the vertical spacing can be 1.5-2.0 meters. MOI SM125 sensors can be used, and Campbell CR6 data acquisition terminals can be used. Sensors can be buried along potential slip surfaces, avoiding fractured rock zones, and the data acquisition terminal can be installed in a monitoring station at the top of the slope. The deployment path can be determined based on a geological exploration report; for example, when the thickness of a weak interlayer is 1.2 meters, the sensor burial depth increases to 10 meters. Compared to inclinometer deployment, this implementation scheme improves monitoring accuracy through denser deployment and depth adjustment. The parameter setting method is based on the thickness of the weak interlayer; for example, when the interlayer thickness exceeds 1 meter, the deployment depth increases to 0.85 times the slope height.

[0099] When pre-embedding sensors, stainless steel fixing clips can be installed every 0.5 meters along the axial direction. The clip size can be 50 mm × 20 mm × 2 mm, and the material can be SUS304 stainless steel. An epoxy resin bonding layer with a thickness of 10-15 mm can be filled between the clip and the soil / rock mass, such as Huntsman Araldite AW 106 epoxy resin. During installation, the clips can be fixed to the inner wall of the borehole using expansion bolts. After the epoxy resin is filled, it should be allowed to cure for 24 hours. The ground data acquisition terminal can perform moving average filtering on the sensor data. The filtering window time can be selected as 10-15 minutes, for example, 12 minutes. Outlier removal can use the ±3σ criterion. For example, if the original strain data for a certain period is: 105, 108, 250, 99, 104, 260, 98, 106, 255, 103, after filtering, outliers of 250 με, 260 με, and 255 με are removed, and the output average is 109 με. Compared to sensor fixing methods, this solution enhances coupling through clips and epoxy resin. During functional testing, sensor drift error can be controlled within ±0.005%.

[0100] The condition for triggering the hydraulic servo anchor bolt is that three adjacent sensors on the same cross-section simultaneously detect a strain increment exceeding 0.04% and a seepage rate reaching 4 mm / h. For example, on a cross-section, sensors S07, S08, and S09 are 2.5 meters apart horizontally. When they detect strain increments of 0.05%, 0.06%, and 0.04% respectively, and seepage rates of 4.5 mm / h, 4.8 mm / h, and 5.2 mm / h respectively, the triggering condition is met. Data verification requires at least three consecutive sampling cycles exceeding the threshold; for example, sampling once per minute, with data exceeding the threshold for three consecutive minutes. Compared to single-sensor triggering, this implementation scheme reduces the false triggering rate through multi-sensor joint judgment. In one experiment, the false triggering rate of the traditional method was around 15%, while the false triggering rate of this scheme was reduced to below 1.5%.

[0101] Taking the slope engineering at section K15+200 of a certain tunnel as an example, there is a weak interlayer (1.2m thick) in this area. The specific implementation steps are as follows:

[0102] (1) Sensor deployment in weak interlayer areas

[0103] A casing drill (model Atlas Copco Diamec 262) was used to drill a hole with a diameter of 110 mm. A PVC casing (wall thickness 2.5 mm, externally coated with lubricant) was inserted, and distributed fiber optic sensors (model MOI sm125) were pre-embedded at a depth of 0.85 times the slope height (12.75 m when the slope height is 15 m), with a horizontal spacing of 2.5 m and a vertical spacing of 1.8 m. Stainless steel clips were installed every 0.5 m along the sensor axis and fixed to the hole wall with expansion bolts (M10×80 mm). Epoxy resin (12 mm thickness) was then injected and allowed to cure for 24 hours.

[0104] (2) Data filtering and trigger condition verification

[0105] The ground data acquisition terminal (Campbell CR6) performs a moving average filter on the raw data (window time 12 minutes). For example, if the strain data for a certain period is: 105, 108, 250, 99, 104, 260, 98, 106, 255, 103, after filtering, outliers of 250με, 260με, and 255με are removed, and the output average is 109με. When sensors S07, S08, and S09 (horizontal spacing 2.5m) simultaneously detect strain increments of 0.05%, 0.06%, and 0.04%, and seepage rates of 4.5mm / h, 4.8mm / h, and 5.2mm / h, respectively, the anchor bolt is triggered.

[0106] (3) Emergency measures and reliability verification

[0107] The hydraulic servo anchor system is equipped with a backup solar power module (output power 200W) to ensure that the data acquisition and hydraulic control unit can continue to operate for ≥8 hours after a power outage; monitoring 30 days after construction shows that the sensor data integrity rate is ≥97% and the false trigger rate is <1%.

[0108] This implementation plan increases the sensor deployment depth in weak interlayer areas to 0.8-0.9 times the slope height, improving the data integrity rate from 85% to 98%. Stainless steel clips and epoxy resin fixing reduce sensor drift error to ±0.005%, improving data stability. Multi-sensor combined triggering conditions reduce the false trigger rate from 15% to below 1.5%, improving the accuracy of support initiation. Therefore, it effectively solves the problems of data distortion caused by insufficient sensor deployment depth and spacing in weak interlayer areas, and the susceptibility of single sensor thresholds to interference leading to false triggering.

[0109] Furthermore, in another embodiment, the method for slope protection construction of open-cut tunnels according to the present invention is as follows:

[0110] (1) During the prestress adjustment process of the hydraulic servo anchor, when the distributed fiber optic sensor detects that the slope strain rate exceeds 0.03% / h, the adjustment step size is increased to 12-15kN (each adjustment ≤5kN, interval ≥10 minutes); when the strain rate is less than 0.01% / h, the adjustment step size is reduced to 5-8kN.

[0111] (2) The anchoring section of the hydraulic servo anchor is equipped with a dual hydraulic cylinder cooperative control system. The main hydraulic cylinder applies axial tension, and the auxiliary hydraulic cylinder applies lateral pressure at an angle of 30°-45° with the rock stratum. The pressure ratio of the main and auxiliary hydraulic cylinders is dynamically controlled by the real-time calculated rock stratum strain gradient, and the range is 1:0.3-0.5.

[0112] (3) The surface of the anchor rod is machined with a continuous thread structure (pitch 15-20mm, depth 2-3mm), the thread area covers the entire length of the grouting section, and grout penetration holes with a diameter of 3-5mm are opened at the root of the thread, with a hole spacing of 50-80mm;

[0113] (4) When the pressure sensor at the head of the anchor bolt detects that the fluctuation of the contact stress exceeds 20% of the initial value, the anchor bolt spacing densification mechanism is automatically triggered, reducing the grid spacing to 1.5-1.8m×1.5-1.8m.

[0114] Specifically, the adjustment step size for the anchor bolt prestress can be selected as 12-15 kN or 5-8 kN, depending on the slope strain rate monitored by the distributed fiber optic sensor. When the strain rate is 0.03% / hour, the step size can be 12-15 kN; when the strain rate is 0.01% / hour, the step size can be 5-8 kN. The hydraulic servo system can use an Enerpac RC-106 pump station (rated pressure 70 MPa, flow rate 2.3 L / min, suitable for HRB400 grade anchor bolts with a prestress loading range of 0-200 kN), and the anchor bolt body can be made of HRB400 grade threaded steel. The anchor bolt installation angle can be selected as 15-25 degrees; for example, when the rock dip angle is 40 degrees, the installation angle can be set to 20 degrees. The adjustment process must respond within 3 minutes. For example, when the strain rate in a certain area is 0.035% / hour, increasing the step size to 14 kN increases the anchor bolt prestress from 50 kN to 64 kN. Compared to static loading of anchor bolts, this solution dynamically adjusts the step size through strain rate. During functional testing, the adjustment delay needs to be verified by simulating strain input data; when an input rate of 0.03% / hour is used, the system should complete the step size switch within 3 minutes.

[0115] The main hydraulic cylinder applies axial tension, while the auxiliary hydraulic cylinder applies lateral pressure at a 30-45 degree angle to the rock strata strike. The pressure ratio between the main and auxiliary hydraulic cylinders can be selected as 1:0.3-0.5; for example, if the main cylinder pressure is 20 MPa, the auxiliary cylinder pressure can be set to 6-10 MPa. The main hydraulic cylinder can be a Bosch Rexroth HED type, and the auxiliary hydraulic cylinder can be a Parker CDR series. The main cylinder is aligned with the anchor bolt axis, and the auxiliary cylinder is installed at an angle via a universal joint. The direction is measured in real-time by a total station; for example, if the rock strata strike is detected to be 35 degrees east of north, the auxiliary cylinder direction is adjusted to 30-40 degrees east of north.

[0116] For example, in a section of a tunnel with well-developed granite fissures, the slope height is 15m, the rock strata dip angle is 40°, fissures are well-developed (average spacing 0.5-1.2m), and seepage channels exist. It is necessary to dynamically control the stiffness of the anchor bolt support to prevent shear slippage in the fissure zone. The hydraulic cylinder deployment is as follows: the main hydraulic cylinder is aligned with the anchor bolt axis, with an initial tension of 50kN; the auxiliary hydraulic cylinder is installed at a north-west angle of 55°, with an initial lateral pressure of 15kN (pressure ratio 1:0.3). Adjustment phase 1 (2 hours after excavation): Real-time strain gradient ∇ε = 0.018% / m → Pressure ratio maintained at 1:0.3, auxiliary cylinder pressure 15kN; Phase 2 (6 hours after excavation): Real-time strain gradient ∇ε = 0.032% / m → Pressure ratio adjusted to 1:0.4, auxiliary cylinder pressure increased to 20kN; Phase 3 (12 hours after excavation): Real-time strain gradient ∇ε = 0.048% / m → Pressure ratio adjusted to 1:0.5, auxiliary cylinder pressure increased to 25kN. Anchor head displacement was measured using a total station, with a maximum slippage of 0.4mm (compared to 1.2mm for the traditional static anchor control group); at a grouting pressure of 1.2MPa, the grout diffusion radius along the permeation hole reached 2.1m (measured using the dye tracer method); core drilling showed a fracture filling rate >85%, and the rock mass shear strength increased from 0.8MPa to 1.6MPa (ASTM D3080). Compared to a single hydraulic cylinder anchor system, this solution enhances shear resistance through dual-cylinder synergy, and dual-cylinder control reduces the slippage at the anchor interface from 1.2 mm to below 0.5 mm.

[0117] The anchor bolt body can be machined with continuous threads using cold rolling or hot rolling processes (pitch 18±2mm, depth 2.5±0.5mm). Grout penetration holes (diameter 4±1mm, hole spacing 60±10mm) are laser-cut at the thread root. The machining accuracy conforms to GB / T 1499.2-2018 standard. The laser cutting equipment can be a Trumpf TruLaser 5030 (accuracy ±0.05mm). The pitch can be selected as 15-20mm, and the depth as 2-3mm. The threaded area covers the entire length of the grouting section; for example, if the grouting section is 8 meters long, the thread machining range is 0-8 meters. Grout penetration holes can be made at the thread root, with a diameter of 3-5mm and a hole spacing of 50-80mm. The penetration holes can be machined using laser CNC machining, with the hole position error controlled within ±0.1mm, and the inner wall roughness Ra≤1.6 micrometers. For example, a certain anchor bolt has a permeation hole diameter of 4 mm and a hole spacing of 60 mm, resulting in a grout diffusion radius of 1.8 meters after grouting. Compared to ordinary threaded anchor bolts, this solution optimizes grout diffusion through permeation holes. Core sampling during drilling revealed that the grout filling rate increased from 70% to over 85%.

[0118] A pressure sensor, such as the Kistler 9067, can be installed on the anchor head to detect contact stress fluctuations. The fluctuation threshold can be set to 20% of the initial value; for example, if the initial contact stress is 50 kN, a fluctuation exceeding 10 kN triggers reinforcement. After reinforcement, the anchor grid spacing can be reduced to 1.5-1.8 m × 1.5-1.8 m, for example, from 2.0 m × 2.0 m to 1.7 m × 1.7 m. Reinforcement must be completed within 30 minutes. In one experiment, when the contact stress fluctuation in a certain area reached 32%, the fluctuation dropped to 8% after the new anchor was driven in. Compared to manual periodic inspections, this solution automatically triggers reinforcement based on real-time data. The experimental subject was a weak interlayer area, and after reinforcement, the displacement rate decreased from 0.6 mm / day to 0.3 mm / day.

[0119] Taking the slope engineering at section K18+750 of a certain tunnel as an example, with a rock dip angle of 40°, the specific implementation steps are as follows:

[0120] (1) Dynamic adjustment of hydraulic servo anchor bolt

[0121] The anchor rod body is made of HRB400 threaded steel (28mm in diameter), with continuous threads rolled on the surface (18mm pitch, 2.5mm depth), and grout penetration holes (4mm in diameter, 60mm spacing) are laser-drilled at the root of the thread.

[0122] When the distributed fiber optic sensor detects a strain rate of 0.035% / h in a certain area, the trigger step size is adjusted to 14kN, the hydraulic system (Enerpac RC-106) outputs a pressure of 22MPa, and the anchor bolt prestress is increased from 50kN to 64kN.

[0123] The lateral pressure direction of the auxiliary hydraulic cylinder (Parker CDR series) is matched with the strike of the rock strata (35° east of north), and the pressure ratio is 1:0.4 (20MPa for the main cylinder and 8MPa for the auxiliary cylinder).

[0124] (2) Contact stress fluctuation and densification mechanism

[0125] A pressure sensor (Kistler 9067) was installed at the head of the anchor bolt. The initial contact stress was 52kN. During construction, a fluctuation range of 45-62kN (fluctuation amplitude of 32.7%) was detected, triggering an encryption command.

[0126] The anchor bolt grid spacing was reduced from 2.0m×2.0m to 1.7m×1.7m, and the contact stress fluctuation amplitude was reduced to ±8% after the new anchor bolts were driven in.

[0127] (3) Verification of grout penetration and grouting effect

[0128] The grouting material is PO 42.5 cement grout with 3% nano silica (water-cement ratio 0.45). When the grouting pressure is 1.0 MPa, the diffusion radius of the grout through the permeation pores reaches about 1.8 m (measured by dye tracer method). Core sampling test shows that the grout fills the fracture rate ≥85%, and the shear strength of the rock mass increases from 0.8 MPa to 1.5 MPa (ASTM D3080 standard).

[0129] This implementation plan, based on graded adjustment of strain rate steps, synergistic enhancement of shear resistance by dual hydraulic cylinders, and optimized grout diffusion through thread and permeation hole design, effectively solves the technical problems of stress lag caused by mismatch between anchor prestress adjustment and deformation rate; anchoring force loss caused by slippage at the anchor-soil interface; and insufficient grout diffusion range.

[0130] Furthermore, in another embodiment, the method for slope protection construction of open-cut tunnels according to the present invention is as follows:

[0131] 1) The ground data acquisition terminal inputs the strain increment and seepage rate data monitored by the distributed optical fiber sensor, as well as the contact stress data of the anchor head pressure sensor, into the pre-trained Bayesian network model to calculate the slope instability probability in real time.

[0132] 2) When the instability probability of the Bayesian network output is ≥85%, the following operations are triggered simultaneously:

[0133] a. The adjustment step of the hydraulic servo anchor bolt is increased to 12-15kN;

[0134] b. The grouting pressure of the high-pressure grouting machine is increased to 1.2-1.4 MPa;

[0135] c. The opening ratio of HDPE perforated drainage pipes is increased to 6-8%, and the hole diameter is adjusted to 8-10mm;

[0136] 3) The Bayesian network model is trained based on historical slope engineering data. The input parameters include strain increment (0.04-0.06%), seepage rate (4-6 mm / h), and standard deviation of contact stress fluctuation (0-1.0 kN). The output is the instability probability value. The model training accuracy is ≥90%.

[0137] (4) When the probability of instability is less than 60% and the displacement rate is less than 0.5 mm / d for 3 consecutive days, restore the anchor bolt step length to 8-12 kN, reduce the grouting pressure to 0.8-1.0 MPa, and adjust the drainage pipe opening rate back to 3-5%.

[0138] Specifically, the strain increment range monitored by the distributed fiber optic sensors can be selected as 0.04-0.06%, the seepage rate range as 4-6 mm / h, and the standard deviation range of contact stress fluctuation as 0.5-1.0 kN. The data can be input into a pre-trained Bayesian network model, with the model instability probability threshold set to 85% and the recovery threshold set to 60%. The distributed fiber optic sensors can be of the MOI SM125 type, the contact stress sensors can be of the Kistler 9067 type, and the ground data acquisition terminal can be of the Campbell CR6 type. The Bayesian network model training can be implemented using the pgmpy library in Python, and the running platform can be an industrial computer (such as the Advantech UNO-2184G). The grouting material can be PO 42.5 silicate cement, and the nano-silica can be of the Evonik AEROSIL 200 type. The distributed fiber optic sensors are pre-embedded along both sides of the tunnel axis slope, 0.5-1.0 m from the slope edge, at a depth of 0.6-0.7 times the slope height. The data acquisition terminal is installed in a temporary monitoring station at the top of the slope and connected to the sensors via fiber optic cables. The instability probability threshold of 85% was determined through historical data analysis: data from one hour prior to 20 landslide events were analyzed, and the instability probability was ≥85% in 17 of those events.

[0139] Compared to traditional monitoring systems, this implementation scheme fuses multi-parameter data through a Bayesian network, rather than relying on a single threshold for judgment. In some cases, traditional methods only rely on displacement rate > 0.6 mm / d to trigger an alarm, while this implementation scheme requires simultaneous satisfaction of strain, seepage, and contact stress conditions.

[0140] The anchor bolt adjustment step length can be selected as 12-15kN, the grouting pressure as 1.2-1.4MPa, and the drainage pipe opening ratio as 6-8%. During the recovery phase, the anchor bolt step length is adjusted back to 8-12kN, the grouting pressure is reduced to 0.8-1.0MPa, and the opening ratio is adjusted back to 3-5%. The hydraulic servo system can use an Enerpac RC-106 pump station (rated pressure 70MPa, flow rate 2.3L / min, suitable for HRB400 grade anchor bolts with a prestressing loading range of 0-200kN), the grouting machine can use a Sany Heavy Industry SYC-2000 model, and the drainage pipe can use Liansu PE100 grade HDPE pipe. The hydraulic pump station is deployed on the side of the construction platform, the grouting machine is placed 5-10m behind the slope, and the drainage pipe is laid along the toe of the slope. The anchor bolt step length adjustment delay must be controlled within 3 minutes; for example, after inputting excessive data, the hydraulic pump station pressure increases from 20 MPa to 24 MPa. Grouting pressure adjustment is automatically completed by the pressure control system of the high-pressure grouting machine, and the opening rate of the drainage pipe is achieved by replacing the pipe section with pre-made holes.

[0141] When the instability probability is <60% and the displacement rate is <0.5 mm / day for three consecutive days, the anchor bolt step length can be restored to 8-12 kN, the grouting pressure can be reduced to 0.8-1.0 MPa, and the drainage pipe opening ratio can be adjusted back to 3-5%. The model training data can include more than 50 sets of historical slope engineering records, with a training accuracy requirement of ≥90% and a mean square error (MSE) ≤0.005. Validation methods can include cross-validation, for example, dividing the data into a training set (40 sets) and a test set (10 sets) and calculating the predicted displacement rate error. If the error exceeds 0.1 mm / day, the model needs to be retrained. Compared to traditional manual review, this scheme automatically evaluates model performance through quantitative indicators. Experimental results show that the Bayesian network model has a prediction error of ±0.08 mm / day, which is 47% lower than the linear regression model error (±0.15 mm / day).

[0142] Taking the slope engineering at section K21+300 of a certain tunnel as an example, the rock strata are interbedded sandstone and shale, and there are seepage channels. The specific implementation steps are as follows:

[0143] (1) Bayesian network model construction and training

[0144] Data collection: Collect 50 sets of historical slope engineering data, including strain increment (0.02-0.08%), seepage rate (2-8mm / h), standard deviation of contact stress fluctuation (0.5-1.2kN), and corresponding slope status (stable / unstable).

[0145] Model training: A Bayesian network was constructed using the pgmpy library in Python, and the node dependency relationship was defined as "strain increment → seepage rate → contact stress fluctuation → instability probability". A directed acyclic graph was used to represent the node dependency relationship. The K2 algorithm was used for structure learning, and the number of parent nodes was constrained to not exceed 3 based on prior knowledge. The rationality of the final network structure was verified by chi-square test (p value > 0.05).

[0146] The prior probabilities of the root nodes (strain increment, seepage rate) were determined based on historical data statistics; the conditional probability table (CPT) of intermediate nodes was optimized using maximum likelihood estimation (MLE) combined with Laplace smoothing (α=0.1). The training objective was to minimize the cross-entropy loss, using the Adam optimizer (learning rate 0.001) for 500 training rounds, with an early stopping mechanism (stopping when the validation set loss did not decrease for 20 consecutive rounds). Five-fold cross-validation was used, and the evaluation metrics included accuracy (92%), mean squared error (0.0048), and ROC-AUC (0.95). Sensitivity analysis showed that the seepage rate had the greatest impact on the instability probability (weight 0.45), followed by strain increment (0.30) and contact stress fluctuation (0.25). Limit tests showed that when the seepage rate > 8 mm / h, the model achieved 98% accuracy in predicting an instability probability ≥ 85%.

[0147] Model deployment: The trained model is integrated into the ground data acquisition terminal (industrial PC, model Advantech UNO-2484G), with a real-time calculation frequency of 1 time / minute.

[0148] (2) Real-time linkage control

[0149] When the monitored strain increment of 0.05%, seepage rate of 5.8 mm / h, and standard deviation of contact stress fluctuation of 0.9 kN in a certain area are detected, the Bayesian network calculates an instability probability of 88%.

[0150] Trigger linkage command:

[0151] a. The hydraulic servo anchor bolt step length was increased from 10kN to 14kN (the pressure of the Enerpac RC-106 pump station was increased to 24MPa).

[0152] b. Adjust the grouting pressure from 1.0MPa to 1.25MPa (Sany SYC-2000 grouting machine);

[0153] c. The opening rate of HDPE drainage pipe (Liansu PE100) has increased from 5% to 7%, and the hole diameter has been increased to 9mm;

[0154] Adjusted 24-hour monitoring data: strain rate decreased to 0.02% / h, seepage rate decreased to 3.5mm / h, and instability probability dropped to 58%.

[0155] Parameter recovery mechanism

[0156] When the displacement rate is 0.4 mm / d, 0.3 mm / d, and 0.4 mm / d for three consecutive days, and the instability probability stabilizes at 55%, a recovery operation is performed.

[0157] The anchor bolt increment was adjusted back to 10kN;

[0158] The grouting pressure dropped to 0.9 MPa;

[0159] The opening ratio of the drain pipe was adjusted to 4%, and the diameter of the hole was reduced to 7mm.

[0160] Within 48 hours after restoration, the slope deformation rate remained stable, and the standard deviation of contact stress fluctuation was ≤0.6kN.

[0161] This implementation method integrates multi-parameter data using a Bayesian network model, reducing the false trigger rate from 15% in traditional methods to below 5%. The linkage control shortens the support response time to 3 minutes and reduces the reinforcement blind zone area by 60%. The parameter recovery mechanism avoids over-support, saving approximately 1 ton of grouting material per 100 linear meters. Model validation, through cross-error control, improves prediction accuracy by 47% and enhances long-term stability, effectively addressing the technical problems of low collaborative efficiency caused by independent multi-parameter control in traditional methods and parameter adaptation lag under seepage-stress coupling.

[0162] Furthermore, in another embodiment, the method for slope protection construction of open-cut tunnels according to the present invention is as follows:

[0163] (1) In the preparation process of the silicate cement slurry, nano-silica is incorporated in two stages:

[0164] a. In the first stage, 1-2% by mass of nano-silica and water-reducing agent (polycarboxylic acid based, dosage 0.1-0.3%) are added together to the mixing water and ultrasonically dispersed for 3-5 minutes (power 200-400W).

[0165] b. In the second stage, add the remaining 1-2% nano-silica, mix it with silicate cement (PO 42.5), and continue stirring for 2-3 minutes. The total stirring time is ≥5 minutes.

[0166] (2) A medium-coarse sand transition layer (thickness 50-80mm, particle size 0.5-1mm) is added to the bottom of the gravel filter layer of the drainage blind ditch, and geotextile (specification ≥200g / m²) is laid between the sand layer and the gravel layer.

[0167] (3) Within 24 hours after grouting, the drainage blind ditch shall be backwashed with a water pressure of ≤0.4MPa and a flushing duration of 10-20 minutes until the drainage pipe has a water flow capacity of ≥5L / (m·s);

[0168] (4) When the hydraulic servo anchor adjustment step length is ≥12kN, 0.2-0.5% of early strength agent (triethanolamine) is injected into the slurry delivery pipe in real time through a metering pump (such as Prominent Gamma / X) to shorten the initial setting time of the slurry to 45-60 minutes.

[0169] Specifically, the preparation of silicate cement slurry can be carried out in two stages by incorporating nano-silica. In the first stage, 1-2% by mass of nano-silica, such as Evonik AEROSIL 200, can be added together with the water-reducing agent. The water-reducing agent can be a polycarboxylate-based agent, with a dosage of 0.1-0.3%, such as BASF MasterGlenium SKY 8233. After mixing, it can be treated with an ultrasonic disperser for 3-5 minutes, with a power of 200-400W, for example, 300W. In the second stage, the remaining 1-2% nano-silica can be added, mixed with PO 42.5 silicate cement, and stirred for another 2-3 minutes, with a total stirring time of no less than 5 minutes. A twin-shaft mixer, such as SICOMA MAO 4500, with a speed of 60 rpm can be used for mixing. Compared to the single-stage doping method, this method improves the dispersion uniformity by doping in stages. The compressive strength of the slurry with 3% nano silica reaches 18.5 MPa, which is 50% higher than that of the undoped group.

[0170] The gravel filter layer of the drainage ditch can be constructed using a three-tiered structure. The lower layer can be gravel with a particle size of 8-12 mm, the middle layer 12-18 mm, and the upper layer 18-22 mm, with each layer 30-40 mm thick. A medium-coarse sand transition layer can be added at the bottom of the filter layer, with a thickness of 50-80 mm and a particle size of 0.5-1 mm, such as quartz sand (SiO2 content ≥95%). Geotextile can be laid between the sand and gravel layers, with a specification of 200 g / m² and a permeability coefficient ≥0.1 cm / s. HDPE perforated pipes can be used for drainage, such as Liansu PE100 grade, with a diameter of 140-160 mm and an open area ratio of 3-8%. Compared to a single-tiered filter layer, this scheme reduces the risk of clogging through a three-tiered structure and a sand transition layer, and the mixed-tiered filter layer ensures a stable drainage flow rate of over 5 liters / (m·s).

[0171] Within 24 hours after grouting, the drainage blind ditch can be backflushed. The flushing water pressure can be below 0.4 MPa, and the duration can be around 10 minutes, for example, using a Grundfos CR 45 multistage centrifugal pump. After flushing, the water flow capacity of the drainage pipe needs to be tested; the threshold can be set to 5 liters / (m·s), for example, using a Keyence FD-Q series flow meter for real-time monitoring. If the water flow capacity does not meet the standard, flushing must be repeated until it is qualified. This solution ensures long-term performance through high-pressure flushing and quantified thresholds, reducing the probability of filter layer clogging by 70% after flushing.

[0172] When the hydraulic servo anchor bolt adjustment step length is ≥12 kN, 0.2-0.5% of an accelerator can be added to the grout. Triethanolamine-based accelerators, such as Sika TEA-25, can be used. After addition, the initial setting time of the grout can be shortened to 45-60 minutes; for example, with 0.3% accelerator added, the initial setting time is 52 minutes. The addition process can be controlled by a metering pump, such as the Prominent Gamma / X type, with an error not exceeding ±0.2%. For example, when the anchor bolt step length is increased to 14 kN, 0.3% accelerator is added simultaneously, and the grout diffusion radius under dynamic water conditions reaches 1.5 meters or more. Compared to a fixed accelerator dosage, this scheme achieves dynamic control through step length linkage, increasing the grout compressive strength by 30% after the accelerator is added.

[0173] Taking the slope engineering at section K24+150 of a certain tunnel as an example, the groundwater level is high and the dynamic water pressure is significant. The specific implementation steps are as follows:

[0174] (1) Preparation of nano-silica slurry

[0175] First stage: Add 1.5% nano silica (Evonik AEROSIL 200) and 0.2% polycarboxylate superplasticizer (BASF MasterGlenium SKY 8233) to the mixing water and treat it with an ultrasonic disperser (300W power, 40kHz frequency) for 4 minutes;

[0176] Second stage: Mix PO 42.5 silicate cement with the remaining 1.5% nano silica, and stir for 3 minutes at 60 r / min using a twin-shaft mixer (model SICOMA MAO 4500) until the slurry fluidity reaches 260 mm (slump test ASTM C143).

[0177] Performance verification: The slurry with 3% nano silica had an initial setting time of 2.5 hours (ASTM C191 standard) and a compressive strength of 20.8 MPa after 28 days (16.5 MPa for the control group without silica).

[0178] (2) Optimized construction of drainage blind ditches

[0179] A medium-coarse sand transition layer (60mm thick, 0.5-1mm particle size, quartz sand SiO2 content ≥95%) is laid at the bottom of the blind drain, and geotextile (250g / m², permeability coefficient ≥0.1cm / s) is laid on top of the sand layer.

[0180] Backwashing was performed using a Grundfos CR 45 multistage centrifugal pump at a water pressure of 0.55 MPa for 18 minutes, increasing the drain pipe's flow rate from 3.8 L / (m·s) to 5.3 L / (m·s).

[0181] (3) Synergistic control of early strength agents

[0182] When the anchor bolt adjustment step length is increased to 14kN (corresponding to grouting pressure of 1.25MPa), 0.3% triethanolamine early strength agent (Sika TEA-25) is added to the grout, and the initial setting time of the grout is shortened to 52 minutes.

[0183] Core sampling showed that after the early strength agent was added, the diffusion radius of the grout under dynamic water conditions (seepage rate 6.5 mm / h) reached 1.5 m or more, and no segregation phenomenon was observed.

[0184] The phased incorporation of nano-silica in this method increases the compressive strength of the slurry by 50% and reduces the permeability coefficient to 5×10⁻. 8 cm / s. The three-stage filter layer and sand transition layer design increases the drainage blind channel water flow capacity compliance rate from 70% to 95%. Backwashing effectively removes silt, maintaining a water flow capacity of over 5 liters / (m·s). Dynamic addition of early-strength agent shortens the initial setting time of the grout to 52 minutes, achieving a filling rate of over 85% under dynamic water conditions. This solves the technical problems of insufficient grout penetration distance under high seepage conditions; drainage blockage caused by unreasonable filter layer gradation; and low curing strength of dynamic water grouting.

[0185] Furthermore, in another embodiment, the method for slope protection construction of open-cut tunnels according to the present invention is as follows:

[0186] (1) During the slope excavation process, the rock temperature data (accuracy ±0.5℃) is monitored in real time by distributed optical fiber sensors and fused with strain and seepage data to establish a temperature-strain-seepage coupling model.

[0187] (2) An abnormal temperature rise may be caused by accelerated groundwater runoff or frictional heating of the rock mass, leading to a decrease in the shear strength of the soil. When the temperature rises abnormally, a thermal infrared imager (accuracy ±0.1℃) should be activated simultaneously. If the temperature gradient is >0.5℃ / m, it is determined to be frictional heating; otherwise, it is due to accelerated groundwater runoff. When the temperature rises abnormally (ΔT≥3℃ / 24h, baseline value is the average value of monitoring over the 30 days before construction ±1℃, window period is 7 days) and is accompanied by a strain rate >0.02% / h, the following emergency measures should be triggered:

[0188] a. Increase the hydraulic servo anchor bolt adjustment step size to 15kN, and refine the anchor bolt grid spacing to 1.5m×1.5m;

[0189] b. Add 0.5-1.0% of an expanding agent (calcium sulfoaluminate) to the grout to achieve a volume expansion rate ≥5% after solidification;

[0190] c. The opening rate of HDPE perforated pipes for drainage blind drains is increased to 8-10%, and the hole diameter is increased to 10-12mm;

[0191] (3) The temperature-strain-seepage coupling model is constructed based on finite element analysis (FEA). The input parameters include the thermal conductivity of the rock mass (1.2-2.5W / m·K), porosity (15-25%), and radius of curvature of the seepage path (0.5-2.0m). The output is the potential thermally induced slip risk level (low / medium / high).

[0192] (4) After the emergency measures are implemented, if the temperature recovers to the baseline value ±1℃ and the strain rate <0.01% / h within 48 hours, the anchor bolt step length, grouting pressure and drainage pipe opening ratio shall be gradually restored to the initial set value.

[0193] Specifically, distributed fiber optic sensors can monitor rock layer temperature data in real time with an accuracy of ±0.5℃, such as the distributed temperature sensing system from LIOS Sensing GmbH. Temperature data can be fused with strain increments (0.02-0.08%) and seepage rates (2-10 mm / h) for analysis. The coupled model can be built using finite element analysis software, such as COMSOL Multiphysics. Input parameters can include rock thermal conductivity (1.2-2.5 W / (m·Kelvin), porosity (15-25%), and seepage path curvature radius (0.5-2.0 m). Model output can be categorized into low, medium, and high risk levels; for example, a temperature change rate ΔT ≥ 3℃ / 24 hours is considered high risk. Sensors can be pre-embedded along the slope, and data acquisition terminals can be installed in monitoring stations at the top of the slope. Compared to traditional temperature monitoring, this solution improves risk prediction capabilities through a multi-parameter coupled model, achieving a 90% match between model-predicted risk levels and actual landslide events.

[0194] Emergency operation can be triggered when the temperature rises abnormally by ΔT ≥ 3℃ / 24 hours and the strain rate is 0.02% / hour. The hydraulic servo anchor adjustment step can be increased to 15 kN, and the anchor grid spacing can be tightened to 1.5m × 1.5m, for example, adjusting from 2.0m × 2.0m to 1.5m × 1.5m. 0.5-1.0% of an expanding agent, such as calcium sulfoaluminate (BASF MasterLife CSA 30), can be added to the grout, resulting in a volume expansion rate of 5-7% after curing. The opening rate of the HDPE perforated pipe in the drainage blind ditch can be increased to 8-10%, and the pore diameter can be enlarged to 10-12 mm, for example, using Liansu PE100 grade pipes. Emergency operation must be initiated within 30 minutes; for example, if the temperature in a certain area rises by 3.5℃ in 24 hours, the grouting pressure should be increased from 1.0 MPa to 1.3 MPa. Compared to traditional grouting reinforcement, this solution enhances timeliness through temperature-strain linkage control, and emergency measures shorten the temperature recovery time from 72 hours to 36 hours.

[0195] After emergency measures are implemented, if the temperature recovers to baseline ±1℃ and the strain rate is below 0.01% / hour within 48 hours, parameters can be gradually restored. For example, the anchor bolt step length can be gradually reduced from 15 kN to 12 kN and then to 8 kN, the grouting pressure can be reduced from 1.3 MPa to 0.9 MPa, and the drainage pipe opening rate can be restored from 10% to 5%. Preventive maintenance can include increasing the flushing frequency of drainage blind ditches to twice a week before the rainy season, with a water pressure of 0.6-0.8 MPa, for example, using a Grundfos CR 45 multistage centrifugal pump. Temporary diversion channels can be added to the slope top intercepting ditch, with a depth of 200-300 mm and a slope of 5-6%, for example, using galvanized steel plates welded on-site. Compared to regular manual inspections, this solution uses dynamic adjustments based on quantitative indicators, and the diversion channels reduce surface runoff infiltration by 60%.

[0196] This implementation plan's temperature-seepage-strain coupling model improves the accuracy of thermal landslide early warning to 90%, a 35% improvement compared to traditional single-parameter monitoring. Emergency response measures ensure that grout filling rates in temperature anomaly areas exceed 85%, and anchor bolt shear strength is increased by 25%. The parameter recovery mechanism avoids over-support, saving approximately 0.3 tons of expansion agent per 100 linear meters. Preventative maintenance stabilizes the drainage system's flow capacity at over 5 liters / (m·s), reducing peak seepage rates during the rainy season by 50%, effectively addressing the risk of thermal deformation caused by temperature anomalies—a problem that traditional monitoring methods do not cover.

[0197] Taking the slope engineering at section K27+600 of a certain tunnel as an example, the rock strata are in a granite fissure development zone, which poses a risk of geothermal anomalies. The specific implementation steps are as follows:

[0198] (1) Construction of temperature-strain-percolation coupling model

[0199] Data acquisition: Distributed fiber optic sensors (model LIOS Sensing GmbH) monitor temperature (resolution 0.1℃), strain (accuracy ±5με), and seepage rate (accuracy ±0.2mm / h) in real time.

[0200] Model simulation: A three-dimensional thermal-fluid-solid coupled model was established using COMSOL Multiphysics software. The input parameters included the thermal conductivity of granite (1.8 W / m·K), porosity (18%), and radius of curvature of the seepage path (1.2 m).

[0201] Risk classification: The model outputs risk level thresholds (low: ΔT < 1℃ / 24h; medium: ΔT = 1-3℃ / 24h; high: ΔT ≥ 3℃ / 24h).

[0202] (2) Emergency response to abnormal temperature

[0203] Monitoring revealed that the temperature in a certain area rose from 25℃ to 28.5℃ within 24 hours (ΔT=3.5℃), with a strain rate of 0.025% / h and a seepage rate of 6.8mm / h. The model determined it to be a high-risk area.

[0204] Emergency operations were triggered: the anchor bolt step length was increased from 10kN to 15kN (Enerpac RC-106 pump station pressure increased to 26MPa), and the grid spacing was increased to 1.5m×1.5m; 0.8% calcium sulfoaluminate expansion agent (BASF MasterLife CSA 30) was added to the grout, and the volume expansion rate after curing was 6.2% (ASTM C878 standard); the opening ratio of HDPE drainage pipe (Liansu PE100) was increased from 5% to 9%, and the pore diameter was expanded to 11mm;

[0205] Thirty-six hours after the emergency measures were implemented, the temperature dropped to 26.2℃, the strain rate was 0.008% / h, and the seepage rate decreased to 3.1mm / h.

[0206] (3) Parameter recovery and effect verification

[0207] Initial parameters were restored in stages: Stage 1 (24 hours): the anchor bolt step length was adjusted back to 12kN, and the grouting pressure was reduced from 1.3MPa to 1.0MPa; Stage 2 (24 hours): the drainage pipe opening rate was restored to 5%, and the hole diameter was reduced to 8mm.

[0208] Core sampling after restoration showed that the grout filling rate in the area where the expansion agent was added was >90%, and the rock mass compressive strength increased to 28.5 MPa (compared to 22.1 MPa in the area where the expansion agent was not added).

[0209] Furthermore, in another embodiment, the method for slope protection construction of open-cut tunnels according to the present invention is as follows:

[0210] (1) After construction is completed, a slope stability classification and early warning mechanism will be established based on long-term monitoring data from distributed fiber optic sensors:

[0211] a. When the displacement rate remains stable at 0.3-0.5 mm / d for 7 consecutive days, it is determined to be a Level 1 warning, and a weekly manual inspection is initiated to check whether the prestress loss of the anchor bolts exceeds 10%;

[0212] b. When the daily displacement rate is >0.6mm / d or the seepage rate is >6mm / h, it is judged as a level II warning, and local grouting (grouting pressure 0.8-1.0MPa) and anchor bolt prestress compensation (step length 5-8kN) should be implemented immediately.

[0213] c. When the displacement rate is greater than 0.8 mm / d and the seepage rate is greater than 8 mm / h for 3 consecutive days, it is judged as a level 3 warning. The anchor mesh is fully densified to 1.5m×1.5m, and temporary drainage wells (500mm in diameter and 6-8m in depth) are added.

[0214] (2) The manual inspection shall record the following parameters: the fluctuation range of the contact stress at the anchor head, the width of the crack in the grouting body (≤2mm), and the drainage capacity of the blind channel (≥4L / (m·s)). The subsequent construction parameters shall be dynamically adjusted based on the inspection results.

[0215] (3) The grouting material is quick-setting silicate cement grout (initial setting time ≤ 1 hour, with 4-6% nano silica added), and the grouting hole position is preferentially selected from the node with the largest strain increment;

[0216] (4) Warning cancellation conditions: After maintenance, the displacement rate is <0.3mm / d and the seepage rate is <2mm / h for 7 consecutive days, and the prestress loss of the anchor bolt is ≤5%.

[0217] Specifically, the displacement rate thresholds monitored by the distributed fiber optic sensors can be set to three levels: Level 1 warning is a displacement rate of 0.3-0.5 mm / day for 7 consecutive days; Level 2 warning is a single-day displacement rate > 0.6 mm / day or seepage rate > 6 mm / hour; Level 3 warning is a displacement rate > 0.8 mm / day and seepage rate > 8 mm / hour for 3 consecutive days. The sensors can be MOI SM125 models, and the data acquisition terminals can be Campbell CR6 models. The sensors can be pre-buried along the slope, and the data acquisition terminals can be installed in the monitoring station at the top of the slope, connected via fiber optic cables. For example, if a displacement rate of 0.4 mm / day is detected for 7 consecutive days in the tunnel section, it is determined as a Level 1 warning, triggering a weekly manual inspection. Compared to single-threshold alarms, this solution achieves differentiated management through tiered warnings. Parameter settings are based on historical data analysis; for example, in a certain area, the displacement rate before 10 landslides all exceeded 0.6 mm / day.

[0218] Manual inspections should check the prestress loss of anchor bolts, the width of cracks in the grouting body, and the drainage capacity of blind channels. The prestress loss threshold for anchor bolts can be set to 10%; for example, if the initial prestress is 50 kN, a loss exceeding 5 kN requires compensation. The crack width threshold for the grouting body can be set to ≤2 mm, measured using a crack width meter (such as the Hilti PS 200). The drainage capacity threshold can be set to ≥4 liters / (m·s), detected using a flow meter (such as the Keyence FD-Q series). Inspection results can be recorded in a digital log, for example, by on-site input using a tablet computer (such as an iPad Pro). Compared to conventional manual inspections, this solution guides maintenance through quantitative indicators. For example, if an inspection finds that three anchor bolts have a prestress loss of 12%, compensation operations are immediately triggered.

[0219] After a Level II warning is triggered, localized grouting can be performed in the area with the largest strain increment. The grouting pressure can be selected as 0.8-1.0 MPa, and the grouting holes can be spaced 1.3-1.7 meters apart. The grout can be quick-setting silicate cement with an initial setting time ≤1 hour, incorporating 4-6% nano-silica (such as Evonik AEROSIL 200). The anchor bolt compensation step can be selected as 5-8 kN; for example, if the prestress of an anchor bolt is to be compensated from 44 kN to 51 kN, an Enerpac RC-106 hydraulic pump station can be used for adjustment. Drainage blind ditch repair includes removing silt and backflushing; the water pressure can be selected as 0.5 MPa, lasting 15 minutes. Compared to traditional grouting reinforcement, this scheme uses data to locate priority areas, achieving a grout diffusion radius of 1.5 meters or more after grouting, and a crack filling rate >85%.

[0220] The warning is lifted if the displacement rate is <0.3 mm / day for 7 consecutive days, the seepage rate is <2 mm / hour, and the anchor prestress loss is ≤5%. For example, if a slope is considered stable after maintenance, with a displacement rate of 0.2-0.28 mm / day, a seepage rate of 1.5-1.8 mm / hour, and contact stress fluctuation ≤3% for 7 consecutive days, it is considered stable. Verification methods include core drilling to check the grout filling rate (≥85%) and pull-out tests to verify the anchor pull-out strength (≥110% of the design value). For example, randomly sampling 10 anchors yields a pull-out strength range of 418-435 kN (design value 400 kN), with a 100% pass rate. Compared to manual sampling, this method ensures objectivity through quantitative indicators. Statistical data shows that the recurrence rate within 30 days after the warning is lifted is <3%.

[0221] This implementation plan improves resource allocation efficiency by 40% through a tiered early warning mechanism, reducing the frequency of manual inspections to once a week for Level 1 early warnings. Quantified inspection standards shorten maintenance response time to within 2 hours, and reduce the anchor bolt breakage rate from 8% to below 2%. Local grouting and anchor bolt compensation operations save approximately 0.8 tons of cement per 100 linear meters in material costs. Multi-parameter release conditions reduce the misjudgment rate from 15% using traditional methods to 5%, significantly improving long-term stability. It effectively solves the technical problems of easy misjudgment using a single displacement rate criterion; lack of control over the recurrence of hidden deformations; and insufficient timeliness of maintenance response.

[0222] Taking the slope engineering at section K30+900 of a certain tunnel as an example, after construction, the stability maintenance phase begins, and the specific implementation steps are as follows:

[0223] (1) Long-term monitoring and early warning classification

[0224] Data from the distributed fiber optic sensor (MOI sm125): Displacement rate was 0.4 mm / d and seepage rate was 3.2 mm / h from day 1 to day 7, which was considered as no warning; on day 8, the flow rate suddenly increased to 0.7 mm / d and the seepage rate was 7.1 mm / h, triggering a level 2 warning.

[0225] Manual inspection revealed that the prestress of three anchor bolts decreased by 12% (from 50kN to 44kN); the drainage capacity of the blind channel at the toe of the slope decreased to 3.5L / (m·s).

[0226] (2) Level II Early Warning Response Operation

[0227] Local grouting: Drill grouting holes (42mm diameter, 1.2m depth) in the area with the largest strain increment (coordinates X:125.34, Y:43.21), inject quick-setting grout (water-cement ratio 0.4, with 5% nano-silica added), grouting pressure 0.9MPa, initial setting time 50 minutes;

[0228] Anchor bolt compensation: Apply a step size of 7kN to the prestressed anchor bolts to restore them to 51kN (hydraulic pump station pressure 21MPa).

[0229] Repair of blind drains: Remove silt and backflush (water pressure 0.5MPa, 15 minutes), restoring water flow capacity to 4.8L / (m·s);

[0230] (3) Verification of warning cancellation

[0231] Post-maintenance monitoring data: Displacement rate 0.2-0.28 mm / d, seepage rate 1.5-1.8 mm / h; anchor prestress fluctuation ≤3% (48-52 kN);

[0232] Manual re-inspection confirmed that there were no visible cracks in the grouting body and the drainage blind channel's water flow capacity was stable at above 4.5L / (m·s), and the warning was lifted.

[0233] Furthermore, in another embodiment, the method for slope protection construction of open-cut tunnels according to the present invention is as follows:

[0234] (1) Dynamically adjust anchor bolt support parameters according to different geological conditions:

[0235] a. Water-rich sand layer area: The length of the anchor bolt is increased to 1.3-1.5 times the excavation depth, the initial prestress is increased to 55-65kN, and the grouting pressure is set to 1.2-1.5MPa (seepage rate 7.2mm / h).

[0236] b. Soft interlayer area: The anchor rod grid spacing is increased to 1.5m×1.5m, the grouting hole depth is increased to 0.6-0.7 times the design height of the slope, and 4-8% nano silica is added to improve the impermeability of the grout;

[0237] c. In areas with well-developed rock fractures: the anchor bolt installation angle is adjusted to be orthogonal to the fracture direction (error ≤ ±5°), and the spacing of the grout penetration holes is reduced to 30-50mm;

[0238] (2) During the rainy season or when the groundwater level is rising, implement the following preventative maintenance measures:

[0239] a. Increase the opening rate of the HDPE perforated pipe for drainage blind drain to 8-10%, enlarge the hole diameter to 10-12mm, and perform backflushing once a week (water pressure 0.6-0.8MPa, lasting 20-30 minutes).

[0240] b. The frequency of anchor bolt prestress monitoring is increased from once a day to three times a day. When a single monitoring shows a prestress loss of ≥5%, a compensation step of 5-8kN is immediately applied.

[0241] c. Install a temporary diversion channel (200-300mm deep, slope ≥5%) in the intercepting ditch at the top of the slope to prevent surface runoff from seeping into the slope;

[0242] (3) The matching relationship between the amount of nano-silica and the grouting pressure is as follows:

[0243] a. When the grouting pressure is 0.8-1.0 MPa, add 4-5% nano-silica; (or replace with an equal amount of fly ash).

[0244] b. When the grouting pressure is 1.2-1.5 MPa, add 6-8% nano-silica; (or replace with an equal amount of fly ash).

[0245] (4) Construction records and effect verification:

[0246] a. Record the support parameters, grouting pressure, and seepage rate of each slope level in real time and generate a digital log;

[0247] b. After each level of slope protection is completed, the grout filling rate is tested by core drilling (≥85% is acceptable), and the pull-out force of the anchor rod is verified by pull-out test (≥110% of the design value).

[0248] Specifically, in water-rich sandy areas, the anchor bolt length can be selected as 1.3-1.5 times the excavation depth. For example, when the excavation depth is 10 meters, the anchor bolt length can be 13-15 meters. The initial prestress can be set at 55-65 kN, and the grouting pressure can be selected at 1.2-1.5 MPa (seepage rate of about 7.2 mm / h). The anchor bolt body can be made of HRB400 grade threaded steel, and the hydraulic system can be an Enerpac RC-106 pump station (rated pressure 70 MPa, flow rate 2.3 L / min, suitable for HRB400 grade anchor bolts with a prestress loading range of 0-200 kN). The anchor bolt can be driven perpendicular to the slope surface, with the anchoring section extending into stable strata (such as moderately weathered rock strata). In areas with weak interlayers, the anchor bolt grid spacing can be increased to 1.5m x 1.5m, and the grouting hole depth can be set to 0.6-0.7 times the slope height. For example, when the slope height is 12m, the grouting hole depth can be 7.2-8.4m. Compared with traditional anchor bolt support, this scheme dynamically adjusts parameters according to geological conditions. After construction in a water-rich sand layer area, when the grouting pressure is around 1MPa, the diffusion radius is 1.5~3.5m, and the anchor bolt pull-out resistance is 115% of the design value.

[0249] During the rainy season or periods of rising groundwater levels, the opening rate of HDPE perforated pipes in drainage blind drains can be increased to 8-10%, and the aperture diameter can be enlarged to 10-12 mm, for example, using Liansu PE100 grade pipes. Backflushing frequency can be increased to once a week, with a water pressure of 0.6-0.8 MPa and a duration of 20-30 minutes, for example, using a Grundfos CR 45 multistage centrifugal pump. Anchor bolt prestress monitoring frequency can be increased from once a day to three times a day, for example, using a Kistler 9067 pressure sensor to collect data in real time. Temporary diversion channels can be added to the slope-top intercepting ditch, with a depth of 200-300 mm and a slope of 5-6%, for example, using galvanized steel plates for welding installation. Compared to conventional drainage system inspections, this implementation plan, through quantitative threshold control, keeps the peak seepage rate below 5 mm / hour during the rainy season of a specific project.

[0250] The dosage of nano-silica (or fly ash) can be dynamically adjusted according to the grouting pressure. When the grouting pressure is 0.8-1.0 MPa, 4-5% nano-silica (such as Evonik AEROSIL 200) is added; when the pressure is 1.2-1.5 MPa, 6-8% is added. Grout preparation can be carried out in two stages: the first stage involves adding the nanomaterials and a water-reducing agent (such as BASF MasterGlenium SKY 8233) to water and ultrasonically dispersing for 3-5 minutes (300W power); the second stage involves mixing with cement for 2-3 minutes. For example, in a certain project with a grouting pressure of 1.3 MPa, adding 7% nano-silica resulted in a grout compressive strength of 20.8 MPa. Compared to traditional grouting material ratios, this scheme optimizes performance through pressure-dosage linkage, reducing the grout permeability coefficient to 5×10⁻⁻⁻⁶ with the addition of 7% nano-silica. 8 cm / s.

[0251] During construction, support parameters, grouting pressure, and seepage rate can be recorded in real time, generating a digital log, which can be entered on-site using an industrial tablet (such as an iPad Pro). After each level of slope protection is completed, the grout filling rate can be tested by core drilling, requiring ≥85%. For example, core sampling in a weak interlayer area showed a filling rate of 88%. Anchor pull-out strength verification can be performed through pull-out tests, for example, using a hydraulic jack (Enerpac P392 type) to load to 110% of the design value (e.g., 440 kN). Randomly sampling 10 anchors showed pull-out strength ranging from 418-435 kN (design value 400 kN), with a 100% pass rate. Compared to traditional sampling inspections, this solution ensures construction quality through quantitative indicators. Statistical data shows that the anchor breakage rate has decreased from 8% using traditional methods to below 2%.

[0252] The geological condition adaptation parameters of this implementation method increase the pull-out resistance of anchor bolts by 15% and stabilize the grout filling rate at over 85%. Rainy season maintenance measures maintain drainage capacity at over 5 liters / (m·s), reducing the risk of seepage erosion by 60%. Dynamically matching the grouting pressure with the dosage of nano-silica increases the grout compressive strength by 50%. The construction record and verification system improves the quality pass rate from 90% to 98%, significantly enhancing long-term stability. This method can solve the challenges of slope stability control under complex geological conditions (water-rich layers, weak interlayers, fractured zones) and seasonal hydrological changes (rainy season).

[0253] For example, taking the slope engineering at section K33+200 of a tunnel as an example, the geological conditions are characterized by alternating water-rich sand layers and weak interlayers. The specific implementation steps are as follows:

[0254] Construction in water-rich sandy layer area (seepage rate approximately 7.2 mm / h)

[0255] The anchor bolt length was set at 1.4 times the excavation depth (10m excavation depth, 14m anchor bolt length), with an initial prestress of 60kN and a grouting pressure of 1.3MPa. The grout contained 7% nano silica (Evonik AEROSIL 200), with a water-cement ratio of 0.4. After grouting, the grout diffusion radius reached more than 1.5m. Core drilling showed that the grout filling rate in the sand layer area was 88%, and the anchor bolt pull-out force was 115% of the design value (standard HRB400 threaded steel, pull-out force ≥400kN).

[0256] The control group used traditional static anchor bolt support (anchor bolt length 10m, fixed prestress 50kN, point monitoring + conventional grouting).

[0257] Compared with traditional static anchor support, in the same water-rich sand layer area (excavation depth 10m), the technical anchor breakage rate of this solution is 4.2%, and the displacement rate is stable at 0.3mm / d; the traditional method has a breakage rate of 12.5% ​​and a peak displacement rate of 0.8mm / d.

[0258] Optimization of weak interlayer regions

[0259] The anchor mesh spacing was increased to 1.5m × 1.5m, and the grouting hole depth was 0.65 times the slope height (slope height 12m, grouting hole depth 7.8m); after the grout containing 8% nano-silica solidified, the permeability coefficient decreased to 5 × 10⁻⁻⁻⁶. 8 cm / s (ASTM D5084 standard), the shear strength of the weak interlayer increased from 0.5MPa to 1.2MPa; monitoring data: the displacement rate stabilized at 0.2mm / d 30 days after construction, and the seepage rate was ≤2mm / h.

[0260] Rainy season preventative maintenance

[0261] The opening rate of HDPE perforated pipes for drainage blind drains was increased to 9%, with a hole diameter of 11mm. The frequency of backwashing was increased to twice a week (water pressure 0.7MPa, 25 minutes each time). Anchor bolt prestress monitoring revealed a daily loss of 6% (from 60kN to 56.4kN), which was immediately compensated by a step of 7kN to restore it to 63.4kN. The slope top diversion channel (depth 250mm, slope 6%) effectively intercepted surface runoff, and the peak slope seepage rate during the rainy season was ≤5mm / h.

[0262] Effect verification and recording

[0263] Digital log records: Grouting pressure for each slope level is 1.2-1.5 MPa, nano-silica content is 7-8%, and seepage rate is 2-5 mm / h;

[0264] Pull-out test results: 10 anchor rods were randomly selected for testing, and the pull-out force ranged from 418 to 435 kN (design value 400 kN), with a pass rate of 100%.

[0265] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. It can be applied to various fields suitable for the present invention. Further modifications can be readily implemented by those skilled in the art.

Claims

1. A method for slope protection construction of open-cut tunnels, characterized in that, Includes the following steps: 1) Before the slope is excavated, distributed optical fiber sensors are pre-embedded on both sides of the slope along the tunnel axis. The depth of the distributed optical fiber sensors is 0.6-0.7 times the design height of the slope. The horizontal spacing between adjacent sensors is 2.5-3.5m and the vertical spacing is 1.8-2.2m. The sensors are connected to the ground data acquisition terminal. 2) When the distributed fiber optic sensor detects that the slope strain increment reaches 0.04-0.06% and the seepage rate reaches 4-6 mm / h, the hydraulic servo anchor support system is activated. The installation angle of the hydraulic servo anchor is 15°-25°, the anchor length is 1.1-1.4 times the excavation depth, the anchor spacing is 1.8-2.2m×1.8-2.2m grid arrangement, the initial prestress is set to 45-55kN, and the step size is dynamically adjusted according to the real-time monitored strain rate of 0.01-0.03% / h, with a step size of 5-15kN. 3) After each slope is excavated to the design elevation, a high-pressure grouting machine is used to seal the slope surface with grout. The grouting pressure is dynamically adjusted according to the seepage rate: when the seepage rate is 4-6 mm / h, the pressure is 0.8-1.0 MPa; when the seepage rate is >6 mm / h, the pressure is increased to 1.2-1.4 MPa. The spacing between grouting holes is 1.3-1.7 m, and the grouting material is silicate cement slurry mixed with 2-8% nano silica. 4) Lay longitudinal drainage blind ditches along the toe of the slope, lay a three-grade gravel filter layer at the bottom of the ditch, install HDPE perforated drainage pipes on top of the filter layer, and wrap the pipes with double layers of geotextile. 5) Continuously monitor slope deformation data using distributed fiber optic sensors. When the displacement rate is <0.5mm / d and the seepage rate is <2mm / h for 3 consecutive days, terminate the hydraulic servo anchor prestress adjustment in stages: in the first stage, reduce the step size to 5kN and maintain it for 24 hours; in the second stage, completely stop the adjustment and continue monitoring for 48 hours.

2. The method for slope protection construction of open-cut tunnels according to claim 1, characterized in that: 1) In areas with weak interlayers, the deployment depth of distributed fiber optic sensors is increased to 0.8-0.9 times the design height of the slope, the horizontal spacing is reduced to 2.0-3.0m, and the vertical spacing is reduced to 1.5-2.0m; 2) When pre-embedding distributed fiber optic sensors, stainless steel fixing clips are installed every 0.5m along the sensor axis, and an epoxy resin bonding layer with a thickness of 10-15mm is filled between the clips and the slope soil. 3) The ground data acquisition terminal performs moving average filtering on the monitoring data of the distributed fiber optic sensors. The filtering window time is 10-15 minutes, and outliers outside the ±3σ range are removed. 4) When three adjacent distributed fiber optic sensors on the same monitoring profile simultaneously detect a strain increment ≥ 0.04% and a seepage rate ≥ 4 mm / h, the start-up conditions of the hydraulic servo anchor bolt support system are triggered.

3. The method for slope protection construction of open-cut tunnels according to claim 2, characterized in that: 1) During the prestress adjustment process of the hydraulic servo anchor, when the distributed fiber optic sensor detects that the slope strain rate exceeds 0.03% / h, the adjustment step size is increased to 12-15kN; when the strain rate is less than 0.01% / h, the adjustment step size is decreased to 5-8kN. 2) The anchoring section of the hydraulic servo anchor is equipped with a dual hydraulic cylinder cooperative control system. The main hydraulic cylinder applies axial tension, and the auxiliary hydraulic cylinder applies lateral pressure at an angle of 30°-45° to the rock stratum. The pressure ratio of the main and auxiliary hydraulic cylinders is dynamically controlled by the real-time calculated rock stratum strain gradient, and the range is 1:0.3-0.

5. 3) The surface of the anchor rod is machined with a continuous thread structure, the thread pitch is 15-20mm, the thread depth is 2-3mm, the thread area covers the entire length of the grouting section, and grout penetration holes with a diameter of 3-5mm are opened at the root of the thread, with a hole spacing of 50-80mm. 4) When the anchor head pressure sensor detects that the contact stress fluctuation exceeds 20% of the initial value, the anchor spacing densification mechanism is automatically triggered, reducing the grid spacing to 1.5-1.8m × 1.5-1.8m.

4. The method for slope protection construction of open-cut tunnels according to any one of claims 1 to 3, characterized in that: 1) The ground data acquisition terminal inputs the strain increment, seepage rate and standard deviation of contact stress fluctuation at the anchor head monitored by the distributed optical fiber sensor into a Bayesian network model trained based on historical slope engineering data. The model input nodes are defined as: strain increment → seepage rate → contact stress fluctuation → instability probability. The mean square error of the output instability probability value is ≤0.

005. 2) When the instability probability output by the Bayesian network is ≥85% or the real-time strain rate is <0.05% / h, the following operations are triggered synchronously: a. The adjustment step of the hydraulic servo anchor bolt is increased to 12-15kN; b. The grouting pressure of the high-pressure grouting machine is increased to 1.2-1.4 MPa; c. The opening ratio of HDPE perforated drainage pipes is increased to 6-8%, and the hole diameter is adjusted to 8-10mm; 3) The Bayesian network model is trained based on historical slope engineering data. The input parameters include strain increment, seepage rate, and standard deviation of contact stress fluctuation. The output is the instability probability value. 4) When the instability probability is <60% and the displacement rate is <0.5mm / d for 3 consecutive days, restore the anchor bolt step length to 8-12kN, reduce the grouting pressure to 0.8-1.0MPa, and adjust the drainage pipe opening rate back to 3-5%.

5. The method for slope protection construction of open-cut tunnels according to claim 4, characterized in that: 1) In the preparation process of the silicate cement slurry, nano-silica is incorporated in two stages: a. In the first stage, 1-2% by mass of nano-silica and water-reducing agent are added together to the mixing water and ultrasonically dispersed; b. In the second stage, add the remaining 1-2% nano-silica, mix it with silicate cement, and continue stirring. 2) A medium-coarse sand transition layer is added to the bottom of the gravel filter layer of the drainage blind ditch, and geotextile is laid between the sand layer and the gravel layer. 3) Within 24 hours after grouting, backwash the drainage blind ditch with a water pressure of ≤0.4MPa and a flushing duration of 10-20 minutes until the drainage pipe has a water flow capacity of ≥5L / (m·s). 4) When the hydraulic servo anchor bolt adjustment step length is ≥12kN, 0.2-0.5% of early strength agent is added to the grout simultaneously to shorten the initial setting time of the grout to 45-60 minutes.

6. The method for slope protection construction of open-cut tunnels according to claim 5, characterized in that: 1) During the slope excavation process, the temperature data of the rock strata is monitored in real time by distributed optical fiber sensors, and the data is fused and analyzed with strain and seepage data to establish a temperature-strain-seepage coupling model. 2) When the temperature rises abnormally and is accompanied by a strain rate > 0.02% / h, the following emergency measures shall be triggered: a. Increase the hydraulic servo anchor bolt adjustment step size to 15kN, and refine the anchor bolt grid spacing to 1.5m×1.5m; b. Add 0.5-1.0% of an expanding agent to the grout so that the volume expansion rate of the grout after solidification is ≥5%; c. The opening rate of HDPE perforated pipes for drainage blind drains is increased to 8-10%, and the hole diameter is increased to 10-12mm; 3) The temperature-strain-seepage coupling model is constructed based on finite element analysis. The input parameters include the thermal conductivity of the rock mass, porosity, and radius of curvature of the seepage path. The output is the potential thermally induced slip risk level. 4) After the emergency measures are implemented, if the temperature recovers to the baseline value ±1℃ and the strain rate is <0.01% / h within 48 hours, the anchor bolt step length, grouting pressure and drainage pipe opening ratio shall be gradually restored to the initial set values.

7. The method for slope protection construction of open-cut tunnels according to claim 6, characterized in that: 1) After construction, based on long-term monitoring data from distributed fiber optic sensors, a graded early warning mechanism for slope stability is established: when the displacement rate is stable at 0.3-0.5 mm / d for 7 consecutive days, it is judged as a Level 1 warning, and a weekly manual inspection is initiated to check whether the prestress loss of the anchor bolts exceeds 10%; when the daily displacement rate is >0.6 mm / d or the seepage rate is >6 mm / h, it is judged as a Level 2 warning, and local grouting and anchor bolt prestress compensation are immediately implemented; when the displacement rate is >0.8 mm / d and the seepage rate is >8 mm / h for 3 consecutive days, it is judged as a Level 3 warning, and the anchor bolt grid is fully densified to 1.5m×1.5m, and temporary drainage wells are added. 2) The following parameters need to be recorded during manual inspection: the fluctuation range of contact stress at the anchor head, the width of cracks in the grouting body, and the water-carrying capacity of the drainage blind channel. The subsequent construction parameters should be dynamically adjusted based on the inspection results. 3) The grouting material is quick-setting silicate cement grout with an initial setting time of ≤1 hour, and 4-6% nano silica is added. The grouting hole positions are preferentially selected from the nodes in the area with the largest strain increment. 4) Warning cancellation conditions: After maintenance, the displacement rate is <0.3mm / d and the seepage rate is <2mm / h for 7 consecutive days, and the prestress loss of the anchor bolt is ≤5%.

8. The method for slope protection construction of open-cut tunnels according to claim 7, characterized in that: 1) Dynamically adjust anchor bolt support parameters according to different geological conditions: In water-rich sandy areas: the length of anchor bolts is increased to 1.3-1.5 times the excavation depth, the initial prestress is increased to 55-65kN, and the grouting pressure is set to 1.2-1.5MPa; In weak interlayer areas: the anchor mesh spacing is increased to 1.5m×1.5m, the grouting hole depth is increased to 0.6-0.7 times the design height of the slope, and 2-8% nano silica is added to improve the impermeability of the grout; In areas with well-developed rock fractures: the anchor bolt installation angle is adjusted to be orthogonal to the fracture direction, and the spacing of the grout penetration holes is reduced to 30-50mm; 2) During the rainy season or periods of rising groundwater levels, implement the following preventative maintenance measures: Increase the opening rate of the HDPE perforated pipe for drainage blind drain to 8-10%, enlarge the hole diameter to 10-12mm, and perform backflushing once a week with a water pressure of 0.6-0.8MPa for 20-30 minutes. The frequency of anchor bolt prestress monitoring has been increased from once a day to three times a day. When a single monitoring shows a prestress loss of ≥5%, a compensation step of 5-8 kN is immediately applied. A temporary diversion channel is added inside the intercepting ditch at the top of the slope. The depth of the diversion channel is 200-300mm and the slope is ≥5% to prevent surface runoff from seeping into the slope. 3) The matching relationship between the amount of nano-silica and the grouting pressure is as follows: When the grouting pressure is 0.8-1.0 MPa, 4-5% nano-silica is added; when the grouting pressure is 1.2-1.5 MPa, 6-8% nano-silica is added.

Images