Search

The large deformation mechanisms and predictions of soft rock tunnels have always been important but difficult to solve problems in the field of geotechnical engineering. A combined finite–discrete element numerical simulation method (FDEM) was used to study large deformation mechanism, classification and prediction. The deformations or failure mechanisms of tunnel surrounding rock were revealed, and the failure modes and displacement value prediction of surrounding rock with different strength-stress ratios were also investigated. The following conclusions were obtained: (1) Critical hysteresis damping can be adopted to simulate the progressive large deformation process of soft rock tunnels, which can obtain the final deformation of unsupported tunnels and avoid a dynamic response. (2) Under concentrated tangential stress, the surrounding rock undergoes X-shaped conjugate shear fracture, and this type of fracture network continues to propagate toward the depth of the surrounding rock until the model reaches a stable state; the reduction in tunnel cross-section is mainly caused by the macroscopic movement and volume expansion of rock fragments, and the latter is due to the generation of a large number of macroscopic voids. (3) As the strength-stress ratio decreases, the deformation or failure modes of the surrounding rock can be divided into four categories: elastic–plastic deformation, closed fracturing, shear dilation and broken expansion. (4) Finally, a prediction equation for isotropic and homogeneous soft rock unsupported tunnel deformation with general size driven by hydrostatic in situ stress is obtained, which indicates that with an increasing strength-stress ratio, the surrounding rock displacement decreases as an exponential function, with a correlation coefficient of R 2 = 0.997. In addition, the robustness and reliability of the prediction equation is verified. [ABSTRACT FROM AUTHOR]

• A new phenomenological tensile failure criterion with an anisotropic coefficient was proposed. • Direct tensile test of the Kangding slate was carried out to verify the accuracy of the new criterion. • The tensile strength data of eight typical layered rocks were collected and compared with the other six typical tensile strength criteria to verify the universality and superiority of the new criterion. • Based on the sensitivity analysis of tensile rate, the loading rate of 0.01 m/s was recommended for the samples with various foliation angles. • Layer thickness has little effect on tensile strength, but a significant effect on tensile failure modes. The tensile strength is a key parameter for the design of many geotechnical engineerings such as tunnel support, slope support, and petroleum drilling. The tensile strength in layered rocks, however, is anisotropic due to the existence of bedding planes. In this paper, to study the anisotropic tensile behaviors, a series of direct tensile tests were conducted using Kangding slate with five different foliation angles (β = 0°, 30°, 45°, 60°, and 90°). Based on the N-Z criterion, a new tensile failure criterion with the anisotropic coefficient was proposed. The tensile strength data of nine typical layered rocks were collected to evaluate the prediction ability of seven anisotropic tensile failure criteria including the new criterion. Embedding the new criterion into FDEM, the loading rate-dependent and layer thickness-dependent tensile mechanism of layered rock under direct tensile test were simulated. The experimental and numerical simulation results showed that the foliation angle significantly affected the tensile strength and failure modes of the slate. With the increase of foliation angle (β), the tensile strength of nine typical layered rocks, including Kangding slate, presented a non-linear increase trend. Compared with the other six typical anisotropic tensile failure criteria, the new criterion has the best prediction capability on the tensile strength of rocks with four different anisotropic degrees, and therefore the new criterion was recommended. The new criterion was embedded in FDEM to carry out direct tensile simulation of slate with different β. The numerical simulation results were in good agreement with the experimental results, further verifying the accuracy of the new criterion. The loading rate sensitivity analysis indicated that a tensile rate of 0.01 m/s was recommended to assure quasi-static loading of the specimen. Through the sensitivity analysis of layer thickness, it is concluded that layer thickness has very little influence on tensile strength, but has a very significant influence on the final failure modes of the specimens. [ABSTRACT FROM AUTHOR]

In this study, a two-dimensional Voronoi element based – explicit numerical manifold method (VE-ENMM) is developed as a new approach to investigate the acoustic emission characteristics during the failure process of intact rocks. The micro-granular structures of the intact rocks are approximated by the random Voronoi polygons and the interactions among rock grains are interpreted using a modified interface contact model. A new acoustic emission (AE) monitoring technique considering the clustering effect of micro-cracks is developed based on the explicit time integration scheme of numerical manifold method. With the newly developed approach, a series of numerical simulations are carried out. Firstly, numerical uniaxial compressive and Brazilian tests are conducted to validate the new approach on simulating micro-/macro-mechanical behaviors of granite samples. After that, the acoustic event properties, such as time evolution, spatial clustering and frequency-magnitude distribution of AE events during the failure process under uniaxial compression are studied in details. Simulation results show that the development of acoustic emissions is in good consistence with the damage evolution process of the numerical model, which may help to reveal the micro-fracturing mechanism of intact rocks in terms of seismic energy dissipation. To further illustrate the capability of the developed approach in monitoring the failure process as well as failure characteristics during TBM tunnel excavation, an additional numerical simulation is conducted and the fracturing behaviors as well as the related AE properties of the surrounding rock mass during TBM tunnelling are investigated. [ABSTRACT FROM AUTHOR]

When tunnelling in densely faulted zones of coal mines, serious deformation often occurs due to adverse geological conditions such as extremely broken and weak rock mass, high tectonic stress, fissure water. To investigate a more efficient support strategy for tunnels passing through densely faulted zones, a tunnel in China's Huainan mining area is taken as an experimental site to study an improved support method. Some key factors associated with the stability of tunnels in the densely faulted zone are analysed. Based on the analysis of failure characteristics of a previously excavated tunnel, a detailed procedure for the improved support method is proposed in which a long pipe pre-grouting method together with a new kind of grouting cable are applied. Finally, the improved support strategy is implemented in the experimental tunnel. The monitoring results suggest that serious deformation is avoided, and the improved compound support method is feasible for tunnels in densely faulted zones. [ABSTRACT FROM AUTHOR]

• A novel intelligent perception model of surrounding rock class was proposed. • TBM cutterhead torque, thrust, penetration rate and rotational speed were chosen to conduct the study. • A complete perception framework was presented from data preprocessing to model construction. • Feature importance and class imbalance were discussed. The perception of surrounding rock geological conditions ahead the tunnel face is essential for TBM safe and efficient tunnelling. This paper developed a perception approach of surrounding rock class based on TBM operational big data and combined unsupervised-supervised learning. In data preprocessing, four data mining techniques (i.e., Z-score, K-NN, Kalman filtering, and wavelet packet decomposition) were used to detect outliers, substitute outliers, suppress noise, and extract features, respectively. Then, GMM was used to revise the original surrounding rock class through clustering TBM load parameters and performance parameters in view of the shortcomings of the HC method in the TBM-excavated tunnel. After that, five various ensemble learning classification models were constructed to identify the surrounding rock class, in which model hyper-parameters were automatically tuned by Bayes optimization. In order to evaluate model performance, balanced accuracy, Kappa, F 1 -score, and training time were taken into account, and a novel multi-metric comprehensive ranking system was designed. Engineering application results indicated that LightGBM achieved the most superior performance with the highest comprehensive score of 6.9066, followed by GBDT (5.9228), XGBoost (5.4964), RF (3.7581), and AdaBoost (0.9946). Through the weighted purity reduction algorithm, the contributions of input features on the five models were quantitatively analyzed. Finally, the impact of class imbalance on model performance was discussed using the ADASYN algorithm, showing that eliminating class imbalance can further improve the model's perception ability. [ABSTRACT FROM AUTHOR]

The failure mechanism and failure mode of anisotropic rock masses are very different from those of isotropic rock masses. For unsupported tunnels, the failure modes of anisotropic rock masses can be summarized as composite failure, V-shaped notch failure and spalling of bedding planes. Different failure modes are controlled by the mechanical parameters of the rock masses, the environmental occurrences and the excavation conditions. The combined finite-discrete element method (FDEM) was employed to investigate the failure mechanism and failure process of layered rock masses and the influence of different factors on the failure mode. The study results indicate that (1) composite failure is the basic failure mode of the layered rock masses and that it is caused by the combined action of shear-slip fractures F1 parallel to the bedding plane, tensile fractures F2 perpendicular to the bedding plane, and conjugate shear fractures F3; the rotation movement of slab-like rock fragments induced by fractures F1 and F3 into the tunnel causes significant deformation of the surrounding rock masses; (2) on the basis of composite failure, when fractures F1 disappear and fractures F2 obliquely intersect with the bedding plane, V-shaped notch failure occurs; furthermore, spalling of the bedding plane occurs when the propagation of fractures F3 is blocked at the bedding plane; and (3) with increasing rock mass strength, lateral pressure coefficient of in situ stress and layer thickness, or decreasing tunnel span, the surrounding rock masses change from composite failure to V-shaped notch failure and then to spalling of the bedding plane. However, the effect of the elastic modulus on the failure mode of anisotropic rock masses is weak, and the influence of the dip angle of the rock layer on the failure mode is complicated. [ABSTRACT FROM AUTHOR]

[Display omitted] • Real-time generated data is used to predict tunnel deformation with the optimized ML. • PSO variants are thoroughly compared and discussed in tunnel deformation prediction. • HPSO-TVAC-XGBoost (100) is developed to provide a relatively accurate prediction. • A series of efficient TBM jamming treatments is summarized and developed. Tunnel deformation is closely related to construction safety and requires some specialists to monitor it for an extended period. So, seeking a timely and convenient method to assess the tunnel deformation is urgently necessary. Tunnels built with the TBMs own the mass of mutual data generated between rock and machine, which is seldom used to investigate and predict tunnel deformation. Machine learning (ML) methods are adopted here to forecast tunnel deformation using mutual rock and machine data. Then, XGBoost and particle swarm optimization (PSO) constitute a robust prediction model (PSO-XGBoost) in which PSO is used to optimize the XGBoost's hyperparameters. Besides, the swarming number (50, 100, 150, 200, and 250) of PSO is also considered and compared in detail. Then, the preliminarily established PSO-XGBoost can outperform PSO-ET, PSO-RF, PSO-SVR, PSO-AdaBoost, PSO-KNN, and PSO-MLP models. Furthermore, to improve the model's accuracy, the training database's feature inputs are enlarged, and the original PSO is also compared and replaced sequentially with the other PSO variants. Such comparisons between the PSO variants about geological issues haven't been carried out yet. Finally, the more precise HPSO-TVAC-XGBoost (100) prediction model is constructed and proposed (training dataset: VAF = 0.9972, RMSE = 0.228; testing dataset: VAF = 0.8846, RMSE = 0.9583). Then, this model is ultimately applied to assess the later risk of tunnel deformation in a fault-fractured tunnel zone where a TBM jamming case happened. And its prediction results agree well with the monitoring values. In addition, a series of efficient TBM jamming treatment measures is briefly stated and summarized in the three key steps: tightly set rigid supporting structure, chemical grouting, and stress release groove excavation, which guides similar projects. [ABSTRACT FROM AUTHOR]

Rockburst has become a main factor threatening the safe construction of deep-buried tunnels. To monitor microseismic (MS) event and comprehend evolution process of rockbursts, a MS monitoring system was established in a deep-buried tunnel excavated by drilling and blasting. The mechanisms of three typical rockburst cases were investigated through the analysis of the MS sequence and development process. The results showed that the MS sequence of first rockburst case was a main-aftershock type, corresponding to a strain rockburst that occurred in intact surrounding rocks. Most of the MS events were failure in tensile, with few MS events presenting mixed failure mode. For the second rockburst case, the MS sequence was a fore-main-aftershock type, corresponding to a strain-structural slip rockburst that occurred in surrounding rocks with a considerable number of small-scale structural planes. The MS events in tensile failure were predominated, and a few MS events were failure in mixed and shear. The MS sequence of the third rockburst case was a new type observed in the present study, i.e. main shock type, corresponding to a strain-structural slip rockburst. It occurred in high strength surrounding rocks with rigid structural planes, which easily induced collapse in the rockburst process. The MS events in tensile failure were predominated. The precursor phenomena of former two cases of the rockburst can be identified through the analysis of MS sequence and source parameters. While before the third rockburst case occurred, no or a few MS events were obtained. It is difficult to predict this type of rockburst based on source parameters. For this case, destress blasting method was adopted. The number of high-energy MS events decreased abruptly, and no intense rockbursts occurred. This method can effectively release the concentrated stress of the surrounding rocks and reduce the risk of rockbursts. The results of this study could provide guidance and reference for the early warning and prevention of rockbursts in deep-buried tunnels. • The temporal and spatial evolution process of microseismicity of three typical rockbursts were studied. • The evolution mechanism of three typical rockbursts was revealed. • Mitigation measures were adopted to reduce the risk of rockbursts. [ABSTRACT FROM AUTHOR]

[Display omitted] Tunnels excavated in soft rock under high in situ stress are usually accompanied by large time-dependent deformation, slow convergence, and long periods of increased support load due to the strong rheological properties of the surrounding rock. This paper presents a time-dependent solution for deeply buried supported circular tunnels excavated in viscoelastic-plastic strain-softening rock masses, in which both the whole process of longitudinal excavation and the sequential installation of two elastic linings are considered. In the solution, the surrounding rock obeys a Burger-creep viscous elastic–plastic strain-softening model and two stages, referring to the "tunnel face influence stage" and the "creep stage", are divided to derive the explicit expressions of the time-dependent tunnel convergences and support loads, in which the former stage accounts for the effect of stress release based on the elastic–plastic analysis, and the latter concentrates on the long-term response induced by the viscous behavior of the ground. To account for the different rock qualities, tunnel face advance rates, and lining tunnel support locations, the elastic–plastic strain-softening solution in the former stage is divided into three cases and the viscos-elastic–plastic solution in the creep stage includes two sub-solutions depending on the evolutions of stress paths being considered or not. For validation purposes, the provided analytical solutions are compared with the numerical results by finite-difference simulations and previous solutions. Following that, parametric studies are conducted to compare the stresses, deformations, and support loads under different stages and solutions. Finally, to further verify the applicability of the solutions in practical tunnels, the convergences, support loads and lining stresses of a typical tunnel cross-section in the Saint Martin La Porte access adit are successfully predicted. Overall, the proposed solutions in this study provide a profitable idea for deriving various viscoelastic-plastic or elasto-viscoplastic explicit solutions that can simultaneously account for both the rock rheology and tunnel face effect. [ABSTRACT FROM AUTHOR]

Abstract A Voronoi-element based explicit numerical manifold method is adopted to investigate the evolution process and failure characteristics of the excavation damaged zone around a deeply buried TBM tunnel. An explicit time integration scheme of numerical manifold method is adopted as the numerical platform, based on which a Voronoi polygon assemblage is generated to approximate the blocky structure of the surrounding rock mass resulting from discrete fracture networks. A new mathematical mesh generation method is developed to better adapt to the Voronoi polygons and an excavation algorithm including both in-situ stress equilibrium and TBM excavation modelling techniques is proposed to capture the mechanical response of the TBM tunnelling process. With the newly developed VE-ENMM approach, the progressive failure process of the surrounding rock mass in a headrace tunnel of Yindajihuang Project is simulated and the formation mechanism of the excavation damaged zone is explored. Simulation results show that a large extent of excavation damaged zone is easily formed in the highly fractured rock mass due to the high in-situ stress in deep grounds. Besides, relatively large convergence displacements are predicted around the tunnel surface, leading to a high risk of TBM shield jamming as in the practical situation. However, the development of damaged zone is effectively restrained and the convergence displacements are well controlled by an early installation of the tunnel lining segments, thus preventing the occurance of TBM shield jamming. Additional simulations are conducted to study the effects of the lateral in-situ stress ratio on the excavation damaged zone. It is inferred that the increase of lateral stress ratio facilitates the fracturing of the surrounding rock mass, which results in a larger extent of the damaged zone and more severe squeezing deformation compared with those under lower lateral stresses. [ABSTRACT FROM AUTHOR]

To investigate the mechanism of tunnel supports for large squeezing deformation hazards, the explicit cohesive element-based numerical manifold method (ECo-NMM) is extended. First, for modeling the compound effect of the artificial supports and tunnel face, a layer-based tunnel support algorithm is proposed to simulate the interaction between the supports, tunnel face and surrounding rock mass and represent the geometric and mechanical properties of the tunnel supports. Then, to investigate the influence of tunnel support timing, a mapping algorithm is proposed to correlate simulated processes with real excavation processes, based on which the installation timing of tunnel supports is characterized. In addition, a reconstruction algorithm is proposed to illustrate 2D simulation results in 3D. To validate the accuracy and applicability of this improved method, numerical cases are conducted using the developed method, and the predicted results are compared with analytical solutions, test observations and original numerical results. Finally, this improved method is applied to investigate the influence of the tunnel support timing, and the predicted results indicate that the squeezing stress acting on the tunnel supports first decreases and then increases with the continuous increase in the delay in the tunnel support, based on which the tunnel support timing is optimized. [ABSTRACT FROM AUTHOR]

In this study, an enhanced numerical manifold method (NMM) simulator is developed to determine the combined impacts of reinforcement range and grouting timing on the effectiveness of grouting reinforcement for preventing the large deformation geo-hazard during the tunnel excavation through a fault rupture zone. To accomplish this objective, the limitations of the current two-dimensional simulation method, which lacks connection to the actual tunnel excavation process, are overcome by proposing a novel softening path function. Using this novel softening path function, each simulation step is given a specific physical meaning, establishing a connection to the actual tunnel excavation process. This improvement enables more accurate and analyzable simulation outcomes. To verify the accuracy of the proposed softening path function, a series of simulations of tunnel excavation are conducted using this softening path function. These results align closely with the results obtained by other reliable numerical methods and previous studies. After validation, several sets of numerical cases are conducted, considering 6 different geo-stress conditions, 10 different grouting reinforcement ranges, and 7 different grouting timings. Based on these simulation results, a critical value for reinforcement range, which equals to the diameter of tunnel face, is identified, and it is founded that delaying the grouting timing can increase the slurry migration range up to 94.6%. Furthermore, delaying the grouting timing can also reduce the convergence displacement of reinforced surrounding rocks after excavation up to 18.2% within a geo-stress range of 10 to 20 MPa. • A novel softening path function with clear physical meanings for tunnel excavation simulations is proposed. • A critical value of the grouting reinforcement range is identified. • Influences of grouting timing on the slurry migration range are revealed. • Influences of grouting timing on the convergence displacement of tunnels are revealed. [ABSTRACT FROM AUTHOR]

• Aquifer drainage has a vital influence on the inclined TBM tunnel. • Bending moment and compression coupling in tunnel longitudinal direction. • Internal forces and stresses are greatest at the sidewall. • Noticeable bending at the junction of the two strata after drainage. • Severe stress concentrations and shear occured at the stratigraphic junction. Drainage of deep aquifers can cause stratum settlement, which has significant effects on the architectural structures within them. Especially for inclined TBM tunnels, the settlement may lead to the longitudinal differential deformation and damage of segments. However, most previous studies on the effects of aquifer drainage settlement on deep structures have focused on the vertical shaft in coal mines, and little attention has been paid to inclined TBM tunnels in deep sandstone aquifers. Through the physical model test and numerical simulation, in this study, the mechanical characteristics and deformation behavior of an inclined TBM tunnel in the deep water-bearing sandstone and impermeable mudstone before and after the aquifer drainage were investigated. The distribution and variation of stresses, internal forces, and deformation on tunnel segments before and after drainage were systematically analyzed along the tunnel longitudinal and circumferential directions. Results show that the whole tunnel is always under compression and mainly subjected to the coupling effect of bending moment and axial compression in the longitudinal direction. After drainage, the longitudinal and circumferential stresses as well as circumferential axial force increase remarkably at the tunnel sidewall but decrease at the tunnel crown and invert. While for the circumferential bending moment, by contrast, it increases at the three positions. In addition, the tunnel experiences a noticeable bending at the junction of the two strata after drainage due to the synchronous settlement with the sandstone, resulting in severe stress concentrations and the formation of the longitudinal shear band at the stratigraphic junction. [ABSTRACT FROM AUTHOR]

• Deep-buried TBM-excavated coal mine roadways were supported by rockbolts. • Support schemes of installing inclined and perpendicular rockbolts were compared. • Stress concentration and damage of surrounding rocks were quantitatively analyzed. • Loads applied on rockbolts were monitored by self-developed instrumented rockbolts. There is an increasing number of TBMs applied in coal mine roadway excavation projects due to the higher safety, faster tunnelling speed and fewer labor works. Nevertheless, the experiences gained from TBM tunnelling projects of transportation or water conveyance tunnels are not well applicable in underground coal mines because of their special engineering and geological conditions. This paper proposed a case study of the excavation disturbance behaviors and support design of TBM-excavated roadway in Zhangji coal mine in Anhui Province, China. Mechanical properties of roadway surrounding rocks were obtained based on parameters of intact rocks and Rock Quality Designation (RQD) values of rock masses. The Universal Discrete Element Code (UDEC) software was used to establish numerical model of the roadway and micro-parameters are also calibrated. The excavation disturbance behaviors of unsupported roadway were firstly simulated. Moreover, the excavation disturbance behaviors and rockbolts loading characters under two different support schemes were studied and compared based on simulation and monitoring results. In this case, simulation results are in good agreement with monitoring results. Research results and experiences obtained from the working site suggest that installing rockbolts perpendicularly to rock surface can better satisfy the requirements of TBM tunnelling in weak grounds of coal mine roadways. [ABSTRACT FROM AUTHOR]

• A novel algorithm for modeling the reinforcement effect after the slurry coagulation was proposed. • A higher computation efficiency for simulating fluid-solid coupling processes under the framework of NMM was achieved. • The effect of the grouting reinforcement in a deeply buried rock tunnel was studied. In this study, a numerical manifold method (NMM) based simulator, which is able to simulate the whole process of grouting reinforcement, including the migration and coagulation process of injected slurries in rock fractures networks, is developed for evaluating the effectiveness of grouting reinforcement in deep rock tunnels. To implement the migration process, the grouting hole generation algorithm, the Bingham fluid model, the migration network searching algorithm as well as the framework for fluid–solid (FS) coupling process are introduced into the zero-thickness cohesive element (ZE) based NMM (Co-NMM). To reflect the reinforcement effect of the slurry coagulation after the migration process, a bonding reinforcement algorithm based on the ZE model is proposed. To improve the computational efficiency, an explicit integration scheme is introduced. To validate and illustrate the capability of the developed method, an uncoupled and a coupled numerical examples are conducted and the simulation results are in good accordance with the analytical solutions or laboratory observations. In addition, a series numerical direct tensile test and uniaxial compressive test are conducted to verify the capability of the proposed bonding reinforcement algorithm. Finally, to demonstrate the capability of the developed simulator in simulating problems with large scale, grouting reinforcement in a deep rock tunnel model is conducted and the predicted results show that the convergence deformation and fragmentation-bulking phenomenon induced by excavation in the deep rock tunnel are effectively controlled through the grouting reinforcement. [ABSTRACT FROM AUTHOR]

For tunnelling in a deep soft rock mass, rock rheology and the advancing process significantly affect the induced pressures on the support system and the long-term safety of the tunnel. In this study, coupled analytical solutions, which take both the rock rheology and tunnel face advancement effects into consideration, are proposed to predict the mechanical behavior of deep-buried circular lined tunnels in a soft rock mass. To account for different rheological behaviors of rocks, five types of viscoelastic models are adopted in the theoretical derivations. The analytical solutions are validated by comparing the predicted results with corresponding existing solutions and those predicted by finite difference simulations. According to parameter degradation, the proposed solutions can be simplified to the classic viscoelastic solution without consideration of tunnel face advancement and the elastic solution with consideration of tunnel face advancement. Based on the proposed coupled solutions, the characteristics of convergent displacements of unlined tunnels with or without consideration of the tunnel face advancement, are studied. In addition, for lined tunnels, the influences of the initial stress release parameter, influence radius of the tunnel face, excavation rate and liner installation time on rock stress and displacement and support pressure are systematically investigated. Finally, the proposed solution is successfully applied to analyze the time-dependent behavior of the Lyon-Turin Base Tunnel, and good agreements are achieved between the theoretically predicted values and corresponding field monitoring data. [ABSTRACT FROM AUTHOR]