Waterproof curtain construction method applied to upper-soft lower-hard stratum
By combining milling cutter modules and hobbing cutter modules in soft upper and hard lower strata, along with an LSTM neural network model and a hydraulic correction system, the problems of low efficiency and high leakage risk of traditional construction methods in hard rock strata were solved, and efficient and continuous water-stop curtain construction was achieved.
Patent Information
- Application Number
- CN202511149851.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-18
- Publication Date
- 2025-11-18
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional water-stop curtain construction methods suffer from insufficient penetration, wall discontinuity, high leakage risk, and low efficiency in soft-over-hard strata. In particular, equipment efficiency is reduced and real-time geological sensing capabilities are lacking in hard rock strata.
The system combines milling cutter and hobbing cutter modules with drilling acoustic sensors and torque monitoring modules. It uses an LSTM neural network model to adjust construction parameters in real time, dynamically match the formation interface, and ensures verticality and efficiency through high-frequency vibration mode and hydraulic correction system.
It enables continuous multi-depth operation with a single drill bit in soft upper and hard lower strata, improving drilling efficiency and construction quality, reducing leakage risk, and meeting the verticality requirements of demanding projects such as subway tunnels.
Smart Images

Figure CN120967988A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method for constructing a water-stop curtain in soft-over-hard soil strata, belonging to the field of civil engineering. Background Technology
[0002] In complex strata with a soft upper layer and a hard lower layer (shallow layer consisting of silty soil and sand, and deep layer consisting of moderately weathered rock or granite), groundwater is abundant and highly permeable, placing extremely high demands on the seepage prevention performance of the cutoff wall. Traditional techniques such as the three-axis mixing pile (SMW method) lack sufficient penetration in hard rock strata, easily leading to problems such as wall discontinuity and increased leakage risk, directly affecting the safety of the foundation pit and the stability of surrounding buildings.
[0003] in: SMW method: Due to the limited torque of the drill rod, the pile depth in gravel or rock layers (uniaxial compressive strength > 30MPa) is usually less than 30m, and the verticality deviation is > 0.5%, resulting in loose wall overlap.
[0004] TRD construction method: Although it can form a continuous wall of uniform thickness, the efficiency of the chainsaw cutting box drops sharply in hard rock, and the height of the equipment is limited, making ultra-deep (>50m) construction difficult.
[0005] Traditional construction methods rely on manual experience to adjust parameters, lacking real-time geological perception capabilities. For example, at the interface between soft and hard rock (such as clay-sandstone abrupt change layers), the drill bit is prone to deflection or jamming, leading to decreased efficiency (milling speed in hard rock sections is only 0.5-1.0 m / h) and abnormal tool wear (life <15 linear meters / set). Summary of the Invention
[0006] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a method for constructing a water-stop curtain in a soft-over-hard stratum, so as to solve the problem.
[0007] To achieve the above objectives, the present invention provides a method for constructing a water-stop curtain in a soft-over-hard stratum, comprising the following steps: Step 1: Divide the stratigraphic interface from top to bottom into soft soil layer, soft-hard interface layer, and hard rock layer; The operation was carried out using a milling tool module in the soft soil layer; The milling tool module is subjected to high-frequency vibration mode at the soft-hard interface layer before operation. In the hard rock layer, the milling tool module is switched to a hobbing tool module for operation; Step 2: Using a drilling acoustic sensor and torque monitoring module, the formation acoustic attenuation rate and torque monitoring value are collected in real time. The historical database in the LSTM neural network model is matched with the acoustic attenuation rate and torque monitoring value to identify the formation interface. Step 3: Based on the LSTM neural network model, the sound wave attenuation rate and the torque monitoring value are input in real time, and the optimal milling / hobbing speed, downforce, grouting pressure are dynamically output, as well as the milling tool module is switched to high-frequency vibration mode or the milling tool module is switched to hobbing tool module. Step 4: Periodically correct the deviation of the milling cutter module or the hobbing cutter module to ensure the verticality of the drilling operation.
[0008] Preferably, the elastic modulus of soft soil is defined as ≤50MPa; the elastic modulus of hard rock is defined as ≥5GPa.
[0009] Preferably, control is based on an LSTM neural network model: The milling cutter module is used to mill at a speed of 2.5-3 m / h in the soft soil layer, with a downward pressure of ≤150 kN and a grouting pressure of 0.8-1.5 MPa. The high-frequency vibration mode in the soft-hard interface layer has a vibration frequency of 10-15Hz, an amplitude of 0.5-1.0mm, a descent speed of 0.8-1.0m / h, and a downward pressure of 150-180kN. The roller cutter module descends at a speed of 1.2–1.8 m / h in the hard rock layer, with a downward pressure of 180–300 kN and a grouting pressure of 2.5–3.0 MPa.
[0010] Preferably, within a 0.5m range of the hard rock layer, ultrafine cement grout is injected to fill the cracks in the hard rock layer.
[0011] Preferably, every 2m of drilling, the tilt sensor provides feedback on the deviation, and a hydraulic correction plate applies a lateral force of 50-80kN to correct the trajectory, ensuring that the verticality is not deviated.
[0012] Preferably, the reference torque value ranges for the soft soil layer, the soft-hard interface layer, and the hard rock layer are defined respectively based on historical data; when the torque monitoring value exceeds the reference torque value range of the current bottom interface by 30% and lasts for more than 5 seconds or exceeds the rated operating time of the drill bit, the risk assessment module is activated.
[0013] Preferably, a torque fluctuation value is generated based on the torque monitoring value; and the risk assessment module is divided into: The torque fluctuation value is less than 15% and the duration is less than 3 seconds, which is defined as a level 1 risk. The torque fluctuation value is between 15% and 30%, and the duration is between 3 and 5 seconds, which is defined as a level 2 risk; If the torque fluctuation value is greater than 30% and the duration is greater than 3 seconds, it is defined as a level 3 risk.
[0014] Preferably, when the risk assessment module assesses it as a Level 1 risk, no intervention is performed, and a normal construction log is generated.
[0015] Preferably, when the risk assessment module assesses the risk as Level II, drilling is paused, the grouting pressure is checked to see if it matches the current formation interface, the tilt sensor provides feedback on the deviation, the hydraulic correction plate applies lateral force to correct the trajectory, an anomaly warning is generated, and the operator is alerted to the anomaly.
[0016] Preferably, when the risk assessment module assesses the risk as Level 3, the machine is immediately shut down, a high-frequency vibration mode is initiated to attempt to escape the predicament, an emergency shutdown is generated, and the operator is alerted to the situation.
[0017] Beneficial effects This invention establishes a three-layer dynamic matching system of "soft soil-boundary-hard rock" to achieve continuous multi-depth operation of a single drill bit; it improves drilling efficiency in soft-hard boundary layers under high-frequency vibration mode; the first-level risk automatic filtering mechanism reduces most of the ineffective downtime; the second-level risk precise handling increases the proportion of effective working time; and the third-level risk effectively identifies risks and takes extrication actions. Attached Figure Description
[0018] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the formation interface of the present invention. Detailed Implementation
[0019] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.
[0020] Please see Figure 1 This invention provides a technical solution for the construction of a water-stop curtain system applied to a soft-over-hard stratum, comprising the following steps: Step 1: To resolve the contradiction of "hole collapse in soft soil layer 1 and inefficiency in hard rock layer 3" in composite strata using traditional single cutting tools, and to adapt to the sudden change in the elastic modulus of the strata, the strata interface is divided into soft soil layer 1, soft-hard interface layer 2, and hard rock layer 3 from top to bottom. A milling tool module is used to operate on the soft soil layer 1; the elastic modulus of the soft soil layer 1 is defined as ≤50MPa. The milling tool module is operated after being subjected to a high-frequency vibration mode in the soft-hard interface layer 2; the elastic modulus of the soft-hard interface layer 2 is defined as 50MPa~5GPa.
[0021] In the hard rock layer 3, the milling tool module is switched to the hobbing tool module for operation; the elastic modulus of the hard rock layer 3 is defined as ≥5GPa; This step overcomes the limitation of traditional layered construction requiring multiple devices by using a dynamic matching system of "formation-tool-parameters", achieving "one drill bit adaptable to all formations".
[0022] Step 2: Using existing drilling acoustic sensors and existing torque monitoring modules, the formation acoustic attenuation rate and torque monitoring values are collected in real time to identify the formation interface; wherein, the detection ends of the drilling acoustic sensors and torque monitoring modules are installed on the drilling tools (milling tool module and hobbing tool module) to detect the corresponding data; Step 3: To address the "lag" problem caused by the reliance on manual experience in traditional parameter adjustments and achieve real-time adaptation between formation changes and construction parameters, an LSTM neural network model is used as the basis. The acoustic attenuation rate and torque monitoring value are input in real time, and the optimal milling / hobbing speed, downforce, and grouting pressure are dynamically output. The model also switches the milling cutter module to high-frequency vibration mode or to a hobbing cutter module. Specifically, the LSTM neural network model contains historical engineering cases, forming a historical database covering acoustic attenuation rate and torque monitoring values for soft soil layer 1, soft-hard interface layer 2, and hard rock layer 3. When the parameters detected by the drilling acoustic sensor and torque monitoring module match the parameters in the LSTM neural network model, it is confirmed that the drill string has moved to a certain formation interface. In one embodiment, when the acoustic attenuation rate suddenly increases from 0.8 dB / m to 1.5 dB / m, and the torque monitoring value rises from 120 kN / m to 180 kN / m, the LSTM neural network model determines that it has entered the soft-hard interface layer 2, automatically switches to high-frequency vibration mode, and adjusts the downforce to 160 kN. When the acoustic attenuation rate and the torque monitoring value parameters match the formation interface in the model library, the parameters are automatically adjusted or the tool mode is switched, forming a "perception-decision-execution" closed loop.
[0023] Preferably, the milling cutter module operates at a milling speed of 2.5–3 m / h in the soft soil layer 1, with a downward pressure ≤150 kN and a grouting pressure of 0.8–1.5 MPa to prevent hole collapse. This increases the milling speed from the traditional 2.0–2.5 m / h to 2.5–3 m / h, improving the working efficiency of the soft soil layer 1 by 15%.
[0024] Preferably, the high-frequency vibration mode in the soft-hard interface layer 2 has a vibration frequency of 10-15Hz, an amplitude of 0.5-1.0mm, a descent speed of 0.8-1.0m / h, and a downward pressure of 150-180kN, so as to reduce the risk of drill string jamming while maintaining high-speed drilling.
[0025] Preferably, the roller cutter module descends at a speed of 1.2–1.8 m / h in the hard rock layer 3, with a downward pressure of 180–300 kN and a grouting pressure of 2.5–3.0 MPa, to break up the rock mass. The drilling speed in the hard rock layer 3 is increased from 0.5–1.0 m / h (traditional method) to 1.2–1.8 m / h, and the single-well cycle is shortened by more than 40%.
[0026] Preferably, ultrafine cement grout is injected within a 0.5m range of the hard rock layer 3 to prevent leakage from cracks in the hard rock (impermeable layer). Furthermore, the water-cement ratio of the ultrafine cement grout used is 0.8:1, and the grouting pressure is 1.5MPa.
[0027] Step Four: To address the rework issues caused by accumulated deviations in traditional methods and ensure the verticality of the cut-off wall meets the stringent requirements of projects such as subways and tunnels, the milling cutter module or the hobbing cutter module is periodically corrected to guarantee the verticality of the drilling operation. In one embodiment, every 2 meters of drilling, the deviation is fed back by an existing tilt sensor, and a lateral force of 50–80 kN is applied by an existing hydraulic correction plate to correct the trajectory. This ensures that the verticality deviation is ≤0.3% (compared to ≥0.5% in traditional methods), meeting the stringent requirements of projects such as subway tunnels.
[0028] Furthermore, to address the limitations of traditional "post-event remediation" risk management and achieve "in-process control" to reduce accident losses, baseline torque value ranges are defined for the soft soil layer 1, the soft-hard interface layer 2, and the hard rock layer 3 based on historical data. When the monitored torque value exceeds the baseline torque value range of the current underlying interface by 30% for more than 5 seconds or exceeds the rated operating time of the drill bit (in one embodiment, the rated operating time is 2 hours), the risk assessment module is activated. This dual triggering condition—a torque monitoring value exceeding the limit by 30% for more than 5 seconds or the rated operating time of the drill bit exceeding the limit—avoids missed reports. It is worth noting that the formation interface is matched using both the acoustic attenuation rate and the torque monitoring value. In one embodiment, when the monitored torque value exceeds the baseline torque value range of the soft soil layer 1 by 30%, while the acoustic attenuation rate is still within the acoustic attenuation rate parameter of the soft soil layer 1 covered by the LSTM neural network model, the risk assessment module is activated. In another embodiment, when the torque monitoring value exceeds the reference torque value range of soft soil layer 1 by 30%, and the acoustic attenuation rate is also not within the acoustic attenuation rate parameter of soft soil layer 1 covered by the LSTM neural network model, the risk assessment module will not be activated.
[0029] Furthermore, a torque fluctuation value is generated based on the torque monitoring value; and the risk assessment module is divided into: A torque fluctuation value less than 15% and a duration of less than 3 seconds is defined as a Level 1 risk. When the risk assessment module assesses a situation as a Level 1 risk, no intervention is performed, and a normal construction log is generated. Non-abnormal fluctuations are automatically filtered out through Level 1 risk assessment, reducing the number of downtimes.
[0030] The torque fluctuation value is between 15% and 30%, and the duration is between 3 and 5 seconds, which is defined as a level 2 risk. When the risk assessment module assesses it as a level 2 risk, drilling is paused, the grouting pressure is checked to see if it matches the current formation interface, and the deviation is fed back by the tilt sensor. The hydraulic correction plate applies lateral force to correct the trajectory, generates an abnormal warning, and reminds the operator to pay attention.
[0031] If the torque fluctuation value is greater than 30% and the duration is greater than 3 seconds, it is defined as a level 3 risk. When the risk assessment module assesses it as a level 3 risk, it immediately stops the machine, starts a high-frequency vibration mode to attempt to get out of trouble, generates an emergency stop, and alerts the operator.
[0032] The torque fluctuation value is calculated as follows: (Previous torque monitoring value - Current torque monitoring value) / Current torque monitoring value × 100%.
[0033] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the scope of the invention. No reference numerals in the claims should be construed as limiting the scope of the claims. Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.
Claims
1. A method for constructing a water-stop curtain in a soft-over-hard stratum, characterized in that: Includes the following steps: Step 1: Divide the stratigraphic interface from top to bottom into soft soil layer, soft-hard interface layer, and hard rock layer; The operation was carried out using a milling tool module in the soft soil layer; The milling tool module is subjected to high-frequency vibration mode at the soft-hard interface layer before operation. In the hard rock layer, the milling tool module is switched to a hobbing tool module for operation; Step 2: Using a drilling acoustic sensor and torque monitoring module, the formation acoustic attenuation rate and torque monitoring value are collected in real time. The historical database in the LSTM neural network model is matched with the acoustic attenuation rate and torque monitoring value to identify the formation interface. Step 3: Based on the LSTM neural network model, the sound wave attenuation rate and the torque monitoring value are input in real time, and the optimal milling / hobbing speed, downforce, grouting pressure are dynamically output, as well as the milling tool module is switched to high-frequency vibration mode or the milling tool module is switched to hobbing tool module. Step 4: Periodically correct the deviation of the milling cutter module or the hobbing cutter module to ensure the verticality of the drilling operation.
2. The method for constructing a water-stop curtain in a soft-over-hard stratum according to claim 1, characterized in that: The elastic modulus of soft soil is defined as ≤50MPa; the elastic modulus of hard rock is defined as ≥5GPa.
3. The method for constructing a water-stop curtain in a soft-over-hard stratum according to claim 2, characterized in that: LSTM neural network model-based control: The milling cutter module is used to mill at a speed of 2.5-3 m / h in the soft soil layer, with a downward pressure of ≤150 kN and a grouting pressure of 0.8-1.5 MPa. The high-frequency vibration mode in the soft-hard interface layer has a vibration frequency of 10-15Hz, an amplitude of 0.5-1.0mm, a descent speed of 0.8-1.0m / h, and a downward pressure of 150-180kN. The roller cutter module descends at a speed of 1.2–1.8 m / h in the hard rock layer, with a downward pressure of 180–300 kN and a grouting pressure of 2.5–3.0 MPa.
4. The method for constructing a water-stop curtain in a soft-over-hard stratum according to claim 1, characterized in that: Within a 0.5m range of the hard rock layer, ultrafine cement grout is injected to fill the cracks in the hard rock layer.
5. The method for constructing a water-stop curtain in a soft-over-hard stratum according to claim 1, characterized in that: Every 2 meters of drilling, the tilt sensor provides feedback on the deviation, and a hydraulic correction plate applies a lateral force of 50-80 kN to correct the trajectory, ensuring that the verticality is not deviated.
6. The method for constructing a water-stop curtain in a soft-over-hard stratum according to claim 1, characterized in that: Based on historical data, the baseline torque value ranges for the soft soil layer, the soft-hard interface layer, and the hard rock layer are defined respectively; when the torque monitoring value exceeds the baseline torque value range of the current bottom interface by 30% and lasts for more than 5 seconds or exceeds the rated operating time of the drill bit, the risk assessment module is activated.
7. A method for constructing a water-stop curtain in a soft-over-hard stratum according to claim 6, characterized in that: Torque fluctuation values are generated based on the torque monitoring values; and the risk assessment module is divided into: The torque fluctuation value is less than 15% and the duration is less than 3 seconds, which is defined as a level 1 risk. The torque fluctuation value is between 15% and 30%, and the duration is between 3 and 5 seconds, which is defined as a level 2 risk; If the torque fluctuation value is greater than 30% and the duration is greater than 3 seconds, it is defined as a level 3 risk.
8. A method for constructing a water curtain in soft-over-hard strata according to claim 7, characterized in that: When the risk assessment module assesses it as a Level 1 risk, no intervention is performed, and a normal construction log is generated.
9. A method for constructing a water-stop curtain in a soft-over-hard stratum according to claim 7, characterized in that: When the risk assessment module assesses the risk as Level II, drilling is paused, the grouting pressure is checked to see if it matches the current formation interface, the tilt sensor provides feedback on the deviation, the hydraulic correction plate applies lateral force to correct the trajectory, an anomaly warning is generated, and the operator is alerted to the anomaly.
10. A method for constructing a water-stop curtain in a soft-over-hard stratum according to claim 7, characterized in that: When the risk assessment module determines the risk level to be Level 3, the machine will be shut down immediately, a high-frequency vibration mode will be initiated to attempt to extricate itself from the predicament, an emergency shutdown will be generated, and the operator will be alerted to the situation.
Citation Information
Cited By
Intelligent milling construction method for waterproof curtain based on three-layer closed-loop control system
CN121539005A