Search

Real-time prediction of the rock mass class in front of the tunnel face is essential for the adaptive adjustment of tunnel boring machines (TBMs). During the TBM tunnelling process, a large number of operation data are generated, reflecting the interaction between the TBM system and surrounding rock, and these data can be used to evaluate the rock mass quality. This study proposed a stacking ensemble classifier for the real-time prediction of the rock mass classification using TBM operation data. Based on the Songhua River water conveyance project, a total of 7538 TBM tunnelling cycles and the corresponding rock mass classes are obtained after data preprocessing. Then, through the tree-based feature selection method, 10 key TBM operation parameters are selected, and the mean values of the 10 selected features in the stable phase after removing outliers are calculated as the inputs of classifiers. The preprocessed data are randomly divided into the training set (90%) and test set (10%) using simple random sampling. Besides stacking ensemble classifier, seven individual classifiers are established as the comparison. These classifiers include support vector machine (SVM), k-nearest neighbors (KNN), random forest (RF), gradient boosting decision tree (GBDT), decision tree (DT), logistic regression (LR) and multi-layer perceptron (MLP), where the hyper-parameters of each classifier are optimised using the grid search method. The prediction results show that the stacking ensemble classifier has a better performance than individual classifiers, and it shows a more powerful learning and generalisation ability for small and imbalanced samples. Additionally, a relative balance training set is obtained by the synthetic minority oversampling technique (SMOTE), and the influence of sample imbalance on the prediction performance is discussed.

To address the inadequacies of traditional pipe-roof methods, the steel support cutting pipe method (SSCP) — a novel pipe-roof method that improves construction security and underground space usage — is proposed. To further explore the applications of SSCP, its design scheme ought to be optimized. The failure mode and mechanical behaviors of the SSCP were investigated through laboratory experiments. Subsequently, a series of finite element models (FEMs) was established to study the deformation characteristics. Further, the parameters of the steel support of the proposed structure were optimized using fuzzy mathematics. The results indicated the ultimate bearing capacity to be 366.8 kN, and the specimen began to yield when the external load reached 70% of the ultimate value. The longitudinal spacing of the steel supports, transverse steel support size, and vertical steel support size had significant effect on the vertical deformation of the steel support and the ground settlement. Finally, the optimal combination of steel supports for the SSCP structure was obtained.

Due to the fast growth of urban areas worldwide, the demand for tunnels in developed areas is increasing. The design and construction of those tunnels are complex because of their shallow depths and their interaction with existing aboveground and buried structures, which results in rather limited allowable ground deformations induced by the tunnel excavation and support. In tropical regions, residual porous soils near the surface are common. Those soils are highly deformable; thus, tunneling may induce large ground deformations that may damage nearby structures. The new Austrian tunneling method (NATM) and the sprayed concrete lining (SCL) technique are being widely employed in several big cities in tropical regions, but little research has been conducted to assess the induced ground deformations in residual soils, common in tropical areas. This paper provides insight into this issue. A well-documented metro tunnel in Sao Paulo, Brazil, in a residual red porous clay, was analyzed using 3D finite element method (FEM). The behavior of the residual red porous clay was approximated by an advanced constitutive soil model calibrated with triaxial tests on intact samples extracted at the site. Predictions of the tunnel deformations during construction matched the field data. The calibrated model was then used to explore the tunnel performance under different construction strategies. The influence of partial face excavation, unsupported span length, support stiffness and pipe roof umbrella were assessed. The numerical results showed that partial face excavation was effective to reduce ground deformations ahead of the face of the tunnel and to improve face stability; however, the settlements behind the face increased because of the delay in closing the primary lining. The installation of a stiffer liner closer to the face reduced the ground deformations significantly. The pipe roof umbrella was the most effective technique to reduce the ground deformations around the tunnel; however, the numerical results did not consider deformations that could be induced by the drilling and grouting operations. The results shown in this paper provide both qualitative and quantitative information about the ground deformations induced by NATM tunneling in residual porous soils, that could help designers and contractors choose the optimum support and construction methods to minimize ground deformations.

Accurate prediction of displacements associated with deep excavations is essential to ensure safety and stability of the excavation and to prevent any damage and distress to the adjoining infrastructures. This paper presents a numerical approach for prediction of ground displacements related to a zone-divided deep excavation construction executed in Shanghai soft clays based on a new elasto–plastic constitutive model (small-strain Shanghai model) that incorporates small strain stiffness. This model can describe the mechanical properties and structural and over-consolidated characteristics of natural clays. The model is implemented into a finite element analysis software. Numerical analysis on the deep excavation in Shanghai using zone-divided method is conducted. A comparison between monitored and simulated results of horizontal displacements along the diaphragm wall, the settlements in the surroundings, and the effects on the adjoining metro tunnel due to excavation construction is carried out. Special attention is paid to the stiffness degradation of representative elements in the ground. The simulated displacements show a good agreement with the monitored data. Overall, this study provides an integrated solution for predicting displacements related to deep excavation in soft clays.

Vacuum preloading has been widely used to improve soft soils in coastal areas of China. An increasing amount of evidence from field operations has shown that conventional vacuum preloading is prone to clogging in prefabricated vertical drains (PVDs) and demands a large volume of sand fills. In recent years, air-boosted vacuum preloading has been developed to overcome these limitations; however, this method still requires more data to verify its performance. In this study, a field test for air-boosted vacuum preloading was conducted, and a large-strain two-dimensional (2D) finite element (FE) model was developed and validated against the field test data. Then, a series of FE parametric analyses was performed to assess key factors, i.e. the air injection pressure, the injection spacing, and the characteristics of cyclic injection, which affect the performance of the air-boosted vacuum preloading. The results showed that the ground settlement and lateral displacement of the soils increased due to an increase in the injection pressure, a decrease in the injection spacing, or increases in the number and duration of the injection cycles. Based on the parametric analyses, an empirical formula for ground settlement prediction was proposed and compared with a case history reported in the literature, showing good agreement.

Investigation of dynamic properties of carbonate/calcareous soils is important in earthquake and offshore engineering as these soils are commonly encountered in large-scale projects related with energy geomechanics and land reclamation. In this study, the stiffness and stiffness anisotropy of two types of calcareous sands (CS) from the Western Australia and the Philippines were examined using bender elements configured in different directions in stress path setups. Stiffness measurements were taken on specimens subjected to constant p′ compression/extension and biaxial stress paths and additional tests were performed on three types of silica sands with different geological origins and particle shapes, which were used as benchmark materials in the study. Compared with the three brands of silica sands, the stiffness of the CS was found to be more significantly influenced by anisotropic loading; an important observation of the experimental results was that stress anisotropy had different weighted influences on the stiffness in different directions, thus influencing stiffness anisotropy. Comparisons were made between the specimens subjected to complex loading paths, and respected model parameters as suggested from published expressions in the literature. These comparisons further highlighted that calcareous soils have different responses in terms of stiffness, stiffness anisotropy and loading history, compared with that of silica-based sands.

Among the geotechnical in situ tests, the dynamic penetration test (DPT) is commonly used around the world. However, DPT remains a rough technique and provides only one failure parameter: blow count or cone resistance. This paper presents an improvement of the dynamic cone penetration test (DCPT) for soil characterisation based on the wave equation theory. Implemented on an instrumented lightweight dynamic penetrometer driving with variable energy, the main process of the test involves the separation and reconstruction of the waves propagating in the rods after each blow and provides a dynamic cone load-penetration (DCLT) curve. An analytical methodology is used to analyse this curve and to estimate additional strength and deformation parameters of the soil: dynamic and pseudo-static cone resistances, deformation modulus and wave velocity. Tests carried out in the laboratory on different specimens (wood, concrete, sand and clay) in an experimental sand pit and in the field demonstrated that the resulting DCLT curve is reproducible, sensitive and reliable to the test conditions (rod length, driving energy, etc.) as well as to the soil properties (nature, density, etc.). Obtained results also showed that the method based on shock polar analysis makes it possible to evaluate mechanical impedance and wave velocity of soils, as demonstrated by the comparisons with cone penetration test (CPT) and shear wave velocity measurements made in the field. This technique improves the method and interpretation of DPT and provides reliable data for shallow foundation design.

The node-based smoothed finite element method (NS-FEM) is shortly presented for calculations of the static and seismic bearing capacities of shallow strip footings. A series of computations has been performed to assess variations in seismic bearing capacity factors with both horizontal and vertical seismic accelerations. Numerical results obtained agree very well with those using the slip-line method, revealing that the magnitude of the seismic bearing capacity is highly dependent upon the combinations of various directions of both components of the seismic acceleration. An upward vertical seismic acceleration reduces the seismic bearing capacity compared to the downward vertical seismic acceleration in calculations. In addition, particular emphasis is placed on a separate estimation of the effects of soil and superstructure inertia on each seismic bearing capacity component. While the effect of inertia forces arising in the soil on the seismic bearing capacity is non-trivial, and the superstructure inertia is the major contributor to reductions in the seismic bearing capacity. Both tables and charts are given for practical application to the seismic design of the foundations.

Geological disasters such as slope failure and landslides can cause loss of life and property. Therefore, reproducing their evolution process is of great importance for risk assessment and mitigation. The recently developed SIMSAND critical state sand model combined with the smoothed particle hydrodynamics (SPH) method is adopted in this work to study slope failure under large deformations. To illustrate the efficiency and accuracy of the SIMSAND-SPH approach, a series of slope collapse studies using the discrete element method (DEM) considering various particle shapes (i.e. spherical, tetrahedral and elongated) is adopted as benchmarks. The parameters of the SIMSAND model are calibrated using DEM triaxial tests. In comparison to the DEM simulations, the runout distance and final slope height are well characterized with the SIMSAND-SPH approach with less computational cost. All comparisons show that the SIMSAND-SPH approach is highly efficient and accurate, which can be an alternative numerical tool to simulate real scale granular flow.

The thermal conductivity of the soil skeleton λs is an extremely important parameter from the point of view of the correct assessment of soil overall/effective conductivity. This work introduces the concept of “local thermal conductivity fluctuation” which characterizes the microscale variation of conductivity within the solid phase. It is proposed to link the “local fluctuation” of thermal conductivity λ with the soil texture – the information that is available at the scale of engineering applications. It was possible to relate the skeleton thermal conductivity with the grain size distribution of the soil. Finally, based on a large series of numerical simulations, the paper provides four triangle diagrams (at different organic matter contents: 0%, 2%, 4% and 6%) relating the value of λs with volume fraction of individual soil separates. This result is extremely important from the practical point of view. One can quickly evaluate λs value provided that information on the grain size distribution and organic matter content is available.

The soil–atmosphere interaction was investigated through laboratory testing, field monitoring and numerical monitoring. In the laboratory, the soil water evaporation mechanisms were studied using an environmental chamber equipped with a large number of sensors for controlling both the air parameters and soil parameters. Both sand and clay were considered. In case of sand, a dry layer could be formed during evaporation in the near surface zone where the suction corresponded to the residual volumetric water content. The evaporative surface was situated at a depth where the soil temperature was the lowest. In case of clay, soil cracking occurred, changing the evaporative surface from one-dimensional to three-dimensional nature. The suctionbased evaporation model was adapted to take these phenomena into account by adopting a function of dry layer evolution in the case of sand and by adopting a surface crack ratio and a retative humidity ratio in the case of clay. In the field, the volumetric water content, and the suction as well as the runoff were monitored for an embankment constructed with lime/cement treated soils. It appeared that using precipitation data only did not allow a correct description of the variations of volumetric water content and suction inside the soils, the consideration of water evaporation being essential. It was possible to use a correlation between precipiration and runoff. The hydraulic conductivity was found to be a key parameter controlling the variations of volumetric water content and suction. For the numerical modelling, a fully coupled thermohydraulic model was developed, allowing analyzing the changes in temperature, volumetric water content and suction of soil, with the upper boundary conditions at the interface between soil and atmosphere determined using meteorological data. Comparison between simulations and measurements showed the performance of such numerical approach.

Discrete element method (DEM) has been intensively used to study the constitutive behaviour of granular materials. However, to what extent a real granular material can be reproduced by virtual DEM simulations remains unclear. This study attempts to answer this question by comparing DEM simulations with typical features of experimental granular materials. Three groups of models with spherical and clumped particles are investigated from four perspectives: (i) deviatoric stress and volumetric behaviour; (ii) critical state behaviour; (iii) stress-dilatancy relationship; and (iv) the evolution of principal stress ratio against axial strain. The results demonstrate that DEM with spherical or clumped particles is capable of qualitatively describing macroscopic deviatoric stress responses, volumetric behaviour, and critical state behaviour observed in experiments for granular materials. On the other hand, some qualitative deviations between experiments and the investigated DEM simulations are also observed, in terms of the stress-dilatancy behaviour and principal stress ratio against axial strain, which are proven to be critical for constitutive modelling. The results demonstrate that DEM with spherical or clumped particles may not necessarily fully capture experimental features of granular materials even from a qualitative perspective. It is thus encouraged to thoroughly validate DEM with experiments when developing constitutive models based on DEM observations.

As mines go deeper, mine designs become more fragile and effective rock support becomes a strategic element for ground control to facilitate timely construction and cost-effective access for uninterrupted production. This article focuses on the design of integrated support systems for brittle ground when large displacements due to gradual bulking of stress-fractured rock or sudden violent bulking during rockbursts are induced by static and dynamic loading. It provides an overview of support design principles for a rational approach to ground control in deep mines when large deformations are anticipated near excavations. Such designs must not only account for load equilibrium but also for deformation compatibility. Most importantly, the design approach must account for the fact that the support’s displacement capacity is being consumed as it is deformed after support installation. It is therefore necessary to design for the remnant support capacity, i.e. the capacity remaining when the support is needed. Furthermore, if the support capacity can be consumed, it can also be restored by means of preventive support maintenance (PSM). The PSM concept for cost-effective ground control is introduced and illustrated by quantitative and operational evidence. Contrary to other design approaches, the deformation-based support design (DBSD) approach provides the capacity of an integrated support system as a function of imposed displacements. Reduction in this support capacity due to mining-induced deformation renders excavations increasingly more vulnerable if located within the influence of active mining and seismic activity. Because deformation measurements are robust indicators of the decay in support capacity, scanning and other displacement monitoring technologies enable measurements to verify the DBSD approach, to assess the remnant safety margin of the deformed support, and to make operational support maintenance decisions.

The deformation modulus of a rock mass is an important parameter to describe its mechanical behavior. In this study, an analytical method is developed to determine the deformation modulus of jointed rock masses, which considers the mechanical properties of intact rocks and joints based on the superposition principle. Due to incorporating the variations in the orientations and sizes of joint sets, the proposed method is applicable to the rock mass with persistent and parallel joints as well as that with non-persistent and nonparallel joints. In addition, an anisotropy index AIdm for the deformation modulus is defined to quantitatively describe the anisotropy of rock masses. The range of AIdm is from 0 to 1, and the more anisotropic the rock mass is, the larger the value of AIdm will be. To evaluate the proposed method, 20 groups of numerical experiments are conducted with the universal distinct element code (UDEC). For each experimental group, the deformation modulus in 24 directions are obtained by UDEC (numerical value) and the proposed method (predicted value), and then the mean error rates are calculated. Note that the mean error rate is the mean value of the error rates of the deformation modulus in 24 directions, where for each direction, the error rate is equal to the ratio of numerical value minus predicted value to the numerical value. The results show that (i) for different experimental groups, the mean error rates vary between 5.06% and 22.03%; (ii) the error rates for the discrete fracture networks (DFNs) with two sets of joints are at the same level as those with one set of joints; and (iii) therefore, the proposed method for estimating the deformation modulus of jointed rock masses is valid.

Trapdoor stability has been widely studied by many researchers in the field of tunneling engineering. A general question being frequently asked is that why most sinkholes have a near-perfect circular shape on the ground surface. This could be possibly explained by the current numerical study using finite element limit analysis under axisymmetric condition, where upper and lower bound solutions of active circular trapdoors are determined. The failure study of sinkholes and the associated failure mechanisms in this paper are for non-homogeneous clay with a linear increase of strength with depth under various cover depth ratios and dimensionless strength gradients. A design equation for predicting the stability solutions is also developed based on the novel three dimensional solutions using axisymmetry.

Effective thermal conductivity of soils can be enhanced to achieve higher efficiencies in the operation of shallow geothermal systems. Soil cementation is a ground improvement technique that can increase the interparticle contact area, leading to a high effective thermal conductivity. However, cementation may occur at different locations in the soil matrix, i.e. interparticle contacts, evenly or unevenly around particles, in the pore space or a combination of these. The topology of cementation at the particle scale and its influence on soil response have not been studied in detail to date. Additionally, soils are made of particles with different shapes, but the impact of particle shape on the cementation and the resulting change of effective thermal conductivity require further research. In this work, three kinds of sands with different particle shapes were selected and cementation was formed either evenly around the particles, or along the direction parallel or perpendicular to that of heat transfer. The effective thermal conductivity of each sample was computed using a thermal conductance network model. Results show that dry sand with more irregular particle shape and cemented along the heat transfer direction will lead to a more efficient thermal enhancement of the soil, i.e. a comparatively higher soil effective thermal conductivity.

Due to the complex diagenesis process, basalt usually contains defects in the form of amygdales formed by diagenetic bubbles, which affect its mechanical properties. In this study, a synthetic rock mass method (SRM) based on the combination of discrete fracture network (DFN) and finite-discrete element method (FDEM) is applied to characterizing the amygdaloidal basalt, and to systematically exploring the effects of the development characteristics of amygdales and sample sizes on the mechanical properties of basalt. The results show that with increasing amygdale content, the elastic modulus (E) increases linearly, while the uniaxial compressive strength (UCS) shows an exponential or logarithmic decay. When the orientation of amygdales is between 0° and 90°, basalt shows a relatively pronounced strength and stiffness anisotropy. Based on the analysis of the geometric and mechanical properties, the representative element volume (REV) size of amygdaloidal basalt blocks is determined to be 200 mm, and the mechanical properties obtained on this scale can be regarded as the properties of the equivalent continuum. The results of this research are of value to the understanding of the mechanical properties of amygdaloidal basalt, so as to guide the formulation of engineering design schemes more accurately.

As mines go deeper and get larger, mine designs become more fragile largely due to the response of the rock mass to mining. Ground control and rock support become important levers in the mine construction schedule, production performance, and excavation health. For example, in cave mines, the production footprint together with associated mine infrastructure are significant investments in a modern caving operation. This investment must be protected and maintained to reduce the risk of ground-related production disruptions. It is necessary to preserve the health of these excavations and their maintenance through an effective rock support design. Rock support thus becomes a strategic element in asset management. This article focuses on support design for brittle ground when displacements induced by stress-fracturing consume much of the support’s capacity. It deals with the functionality of the support in deforming ground. Several interlinked concepts are important when assessing excavation health. Designs must not only account for load equilibrium but also for deformation compatibility and capacity consumption. Most importantly, the support’s displacement capacity is being consumed when the rock mass is deformed after support installation. Hence, it is necessary to design for the support capacity remaining at the time when the support is needed. If support capacity can be consumed, it can also be restored by means of preventive support maintenance (PSM). This concept for cost-effective ground control is introduced and illustrated on operational evidence. Furthermore, how design can impact construction costs and schedule are discussed. Support is installed to provide a safe environment and preserve an operationally functional excavation. It also must assure senior management that investments in high quality support and its maintenance will substantially reduce delays and with it, costs. It is demonstrated that the use of ‘gabion-like’ support systems can achieve these goals. A technical summary of the ‘gabion panel’ support system design is presented.

The existence of squeezing ground conditions can lead to significant challenges in designing an adequate support system for tunnels. Numerous empirical, observational and analytical methods have been suggested over the years to design support systems in squeezing ground conditions, but all of them have some limitations. In this study, a novel experimental setup having physical model for simulating the tunnel boring machine (TBM) excavation and support installation process in squeezing clay-rich rocks is developed. The observations are made to understand better the interaction between the support and the squeezing ground. The physical model included a large true-triaxial cell, a miniature TBM, laboratory-prepared synthetic test specimen with properties similar to natural mudstone, and an instrumented cylindrical aluminum support system. Experiments were conducted at realistic in situ stress levels to study the time-dependent three-dimensional tunnel support convergence. The tunnel was excavated using the miniature TBM in the cubical rock specimen loaded in the true-triaxial cell, after which the support was installed. The confining stress was then increased in stages to values greater than the rock's unconfined compressive strength. A model for the time-dependent longitudinal displacement profile (LDP) for the supported tunnel was proposed using the tunnel convergence measurements at different times and stress levels. The LDP formulation was then compared with the unsupported model to calculate the squeezing amount carried by the support. The increase in thrust in the support was back-calculated from an analytical solution with the assumption of linear elastic support. Based on the test results and case studies, a recommendation to optimize the support requirement for tunnels in squeezing ground is proposed.

To analyze and predict the mechanical behaviors of deep hard rocks, some key issues concerning rock fracturing mechanics for deep hard rock excavations are discussed. First, a series of apparatuses and methods have been developed to test the mechanical properties and fracturing behaviors of hard rocks under high true triaxial stress paths. Evolution mechanisms of stress-induced disasters in deep hard rock excavations, such as spalling, deep cracking, massive roof collapse, large deformation and rockbursts, have been recognized. The analytical theory for the fracturing process of hard rock masses, including the three-dimensional failure criterion, stress-induced mechanical model, fracturing degree index, energy release index and numerical method, has been established. The cracking-restraint method is developed for mitigating or controlling rock spalling, deep cracking and massive collapse of deep hard rocks. An energy-controlled method is also proposed for the prevention of rockbursts. Finally, two typical cases are used to illustrate the application of the proposed methodology in the Baihetan caverns and Bayu tunnels of China.

A B S T R A C T This laboratory study examines the potential use of an anionic polyacrylamide (PAM)-based material as an environmentally sustainable additive for the stabilization of an expansive soil from South Australia. The experimental program consisted of consistency limits, sediment volume, compaction and oedometer cyclic swell–shrink tests, performed using distilled water and four different PAM-to-water solutions of PD = 0.1 g/L, 0.2 g/L, 0.4 g/L and 0.6 g/L as the mixing liquids. Overall, the relative swelling and shrinkage strains were found to decrease with increasing number of applied swell–shrink cycles, with an ‘elastic equilibrium’ condition achieved on the conclusion of four cycles. The propensity for swelling/shrinkage potential reduction (for any given cycle) was found to be in favor of increasing the PAM dosage up to PD = 0.2 g/L, beyond which the excess PAM molecules self-associate as aggregates, thereby functioning as a lubricant instead of a flocculant; this critical dosage was termed ‘maximum flocculation dosage’ (MFD). The MFD assertion was discussed and validated using the consistency limits and sediment volume properties, both exhibiting only marginal variations beyond the identified MFD of PD = 0.2 g/L. The accumulated axial strain progressively transitioned from ‘expansive’ for the unamended soil to an ideal ‘neutral’ state at the MFD, while higher dosages demonstrated undesirable ‘contractive’ states.

Numerous intersected rock fractures constitute the fracture network in enhanced geothermal systems. The complicated convective heat transfer behavior in intersected fractures is critical to the heat recovery in fractured geothermal reservoirs. A series of three-dimensional intersected fracture models is constructed to perform the flow-through heat transfer simulations. The geometry effects of dead-end fractures (DEFs) on the heat transfer are evaluated in terms of intersected angles, apertures, lengths, and the connectivity. The results indicate that annular streamlines appear in the rough DEF and cause an ellipse distribution of the cold front. Compared to plate DEFs, the fluid flow in the rough DEF enhances the heat transfer. Both the increment of outlet water temperature ΔTout and the ratio of heat production Qr present the largest at the intersected angle of 90° while decline with the decrease of the intersected angle between the main flow fracture (MFF) and the DEFs. The extension of the length of intersected DEFs is beneficial to heat production while enhancing its aperture is not needed. Solely increasing the number of intersected DEFs induces a little increase of heat extraction, and more significant heat production can be obtained through connecting these DEFs with the MFF forming the flow network.

Excessive structural forces generated inside tunnel linings could affect the safety and serviceability of tunnels, emphasizing the need to accurately predict the forces acting on tunnel linings during the preliminary design phase. In this study, an anisotropic soil model developed by Norwegian Geotechnical Institute (NGI) based on the Active-Direct shear-Passive concept (NGI-ADP model) was adopted to conduct finite element (FE) analyses. A total of 682 cases were modeled to analyze the effects of five key parameters on twin-tunnel structural forces; these parameters included twin-tunnel arrangements and subsurface soil properties: burial depth H, tunnel center-to-center distance D, soil strength s u A , stiffness ratio G u / s u A , and degree of anisotropy s u P / s u A . The significant factors contributing to the bending moment and thrust force of the linings were the tunnel distance and overlying soil depth, respectively. The degree of anisotropy of the surrounding soil was found to be extremely important in simulating the twin-tunnel construction, and severe design errors could be made if the soil anisotropy is ignored. A cutting-edge application of machine learning in the construction of twin tunnels is presented; multivariate adaptive regression splines and decision tree regressor methods are developed to predict the maximum bending moment within the first tunnel’s linings based on the constructed FE cases. The developed prediction model can enable engineers to estimate the structural response of twin tunnels more accurately in order to meet the specific target reliability indices of projects.

Surface settlements caused by twin tunneling are a major interest in underground space engineering. Most prevailing studies have investigated the surface settlements observed over twin tunnels constructed in clay. Yet, less attention is given to the twin tunnels excavated in silty sand, which is the focus of this study. A comprehensive field monitoring scheme was designed in the Hangzhou metro line six project to measure the surface settlements during twin tunneling by adopting earth pressure balance shields. The influence of four shield operational parameters, namely, face pressure, tail grouting pressure, penetration rate, and pitching angle on the surface settlements due to twin tunneling are explored. In general, the total settlement profiles observed over twin tunnels follow an inverted bimodal shape. The post-settlement observed over the 1st tunnel due to the 2nd tunnel excavation is non-trivial when the twin tunnels are spaced closely. Low face pressure and tail grouting pressure can induce excessive peak surface settlements during twin tunneling in silty sand.

Circumferential yielding lining is able to tolerate controlled displacements without failure, which has been proven to be an effective solution to large deformation problem in squeezing tunnels. However, up to now, there has not been a well-established design method for it. This paper aims to present a detailed analytical computation of support characteristic curve (SCC) for circumferential yielding lining, which is a significant aspect of the implementation of convergence-confinement method (CCM) in tunnel support design. Circumferential yielding lining consists of segmental shotcrete linings and highly deformable elements, and its superior performance mainly depends on the mechanical characteristic of highly deformable element. The deformation behavior of highly deformable element is firstly investigated. Its whole deforming process can be divided into three stages including elastic, yielding and compaction stages. Especially in the compaction stage of highly deformable element, a nonlinear stress–strain relationship can be observed. For mathematical convenience, the stress–strain curve in this period is processed as several linear sub-curves. Then, the reasons for closure of circumferential yielding lining in different stages are explained, and the corresponding accurate equations required for constructing the SCC are provided. Furthermore, this paper carries out two case studies illustrating the application of all equations needed to construct the SCC for circumferential yielding lining, where the reliability and feasibility of theoretical derivation are also well verified. Finally, this paper discusses the sensitivity of sub-division in element compaction stage and the influence of element length on SCC. The outcome of this paper could be used in the design of proper circumferential yielding lining.

Surface subsidence induced by underground mining is a typical serious geohazard. Numerical approaches such as the discrete element method (DEM) and finite difference method (FDM) have been widely used to model and analyze mining-induced surface subsidence. However, the DEM is typically computationally expensive, and is not capable of analyzing large-scale problems, while the mesh distortion may occur in the FDM modeling of largely deformed surface subsidence. To address the above problems, this paper presents a geometrically and locally adaptive remeshing method for the FDM modeling of largely deformed surface subsidence induced by underground mining. The essential ideas behind the proposed method are as follows: (i) Geometrical features of elements (i.e. the mesh quality), rather than the calculation errors, are employed as the indicator for determining whether to conduct the remeshing; and (ii) Distorted meshes with multiple attributes, rather than those with only a single attribute, are locally regenerated. In the proposed method, the distorted meshes are first adaptively determined based on the mesh quality, and then removed from the original mesh model. The tetrahedral mesh in the distorted area is first regenerated, and then the physical field variables of old mesh are transferred to the new mesh. The numerical calculation process recovers when finishing the regeneration and transformation. To verify the effectiveness of the proposed method, the surface deformation of the Yanqianshan iron mine, Liaoning Province, China, is numerically investigated by utilizing the proposed method, and compared with the numerical results of the DEM modeling. Moreover, the proposed method is applied to predicting the surface subsidence in Anjialing No. 1 Underground Mine, Shanxi Province, China.

In this study, a refined numerical model for segmental lining of a shield tunnel, which contains detailed models of reinforcement and connecting bolts, is established using finite element software. The model is first validated by the results from a full-scale model test. Then, based on the load-structure method, this numerical model is adopted to investigate the internal force distribution and the transverse deformation characteristics of the shield tunnel when it is subject to local soil loosening. The influence of loosening position, loosening range, and loosening extent on the mechanical response is extensively studied through comprehensive numerical analyses. The results show that the main influence of local soil loosening on the ring is to disturb the force balance and change the constraint conditions, thus changing the deformation pattern and force state. After the loosening occurs, the bending moment of the ring in the loosening range increases and the axial force decreases. The vertical convergence of the ring is the largest and the equivalent stiffness of the ring is the smallest when the local soil loosened at the haunch and the loosening range α is 90°. The vertical convergence of the ring increases with increasing of the loosening extent, and the equivalent stiffness decreases linearly with increasing of the loosening extent. The results can enhance our understanding of mechanical behaviors of segmental lining associated soil loosening, and will show a possible way for detecting soil loosening based on the measured deformation and internal forces.

The safety assessment of tunnel stability is critical to tunnel construction and requires accurate analysis to obtain a reliable prediction. Strength anisotropy is an important aspect of clay behavior, but it is mostly neglected in practical stability analyses. In this study, the effects of undrained strength anisotropy and strength nonhomogeneity on the stability of unlined circular tunnels in clays are investigated. The static approach of lower-bound (LB) analysis using finite-element and second-order cone programming is employed to examine the aforementioned effects. The anisotropic shear strength of clay is modeled by employing an elliptical yield function under plane strain conditions. A complete set of dimensionless parameters covering the cover depth ratios of tunnels, normalized overburden pressure ratios, normalized strength gradient ratios of clays, and anisotropic strength ratios, are systematically investigated. The new LB solutions indicate that the stability load factor of the problem has a nonlinear relationship with the cover depth ratio and the anisotropic strength ratio, and there exists a linear relationship with the normalized overburden pressure and the normalized strength gradient. Their influence on the predicted failure mechanism is parametrically evaluated. A statistically approximate stability equation of unlined circular tunnels in anisotropic and non-homogeneous clay is proposed for the first time, which contains four new stability factors, namely, constant undrained strength, linearly increasing strength gradient, undrained strength anisotropy, and soil unit weight, and it can serve as a fast and accurate tool for predicting the undrained stability of this problem in practice.

Particle morphology has great influence on mechanical behaviour and hydro/thermal/electrical conductivities of granular materials. Surface reconstruction and mesh generation are critical to consider realistic particle shapes in various computational simulations. This study adopts the combined finite- discrete element method (FDEM) to investigate single particle crushing behaviour. Particle shapes were reconstructed with spherical harmonic (SH) in both spherical and Cartesian coordinate systems. Furthermore, the reconstructed surface mesh qualities in two coordinate systems are investigated and compared. Although the efficiency of the two SH systems in reconstructing star-like shapes is nearly identical, SH in Cartesian coordinate system can reconstruct non-star-like shapes with the help of surface parameterisation. Meanwhile, a higher triangular mesh quality is generated with spherical coordinate. In single particle crushing tests, the low mesh quality produces more fluctuations on load–displacement curves. The particles with more surficial mesh elements tend to have a lower contact stiffness due to more contact stress concentrations induced by complexity of morphology features and more volumetric tetrahedral elements. The fracture patterns are also influenced by mesh quality and density, e.g. a particle with fewer mesh elements has a simpler fragmentation pattern. This study serves as an essential step towards modelling particle breakage using FDEM with surface mesh directly from SH reconstruction.

To date, the accurate prediction of tunnel boring machine (TBM) performance remains a considerable challenge owing to the complex interactions between the TBM and ground. Using evolutionary polynomial regression (EPR) and random forest (RF), this study develops two novel prediction models for TBM performance. Both models can predict the TBM penetration rate and field penetration index as outputs with four input parameters: the uniaxial compressive strength, intact rock brittleness index, distance between planes of weakness, and angle between the tunnel axis and planes of weakness (α). First, the performances of both EPR- and RF-based models are examined by comparison with the conventional numerical regression method (i.e., multivariate linear regression). Subsequently, the performances of the RF- and EPR-based models are further investigated and compared, including the model robustness for unknown datasets, interior relationships between input and output parameters, and variable importance. The results indicate that the RF-based model has greater prediction accuracy, particularly in identifying outliers, whereas the EPR-based model is more convenient to use by field engineers owing to its explicit expression. Both EPR- and RF-based models can accurately identify the relationships between the input and output parameters. This ensures their excellent generalization ability and high prediction accuracy on unknown datasets.

The paper discusses existing reinforcement, future reinforcement and new technologies for concrete pipes used in the sewage systems. Concrete pipes currently in use and under investigation are reviewed. Structural fibres, as the main reinforcement of concrete pipes, are known as an attractive alternative to the traditional steel bars. Steel, synthetic and basalt fibres have been considered. The latest research and mechanical properties of individual fibres are presented. Advances in fibre-reinforced concrete provide a new basis for the design of more efficient concrete pipes, especially those resistant to biological corrosion and with a longer service life. In the article, future non-corrosive reinforcement due to the reduction of steel reinforcement and corrosion protection linings has been proposed.

Brick sewers were designed as egg-shaped, pear-shaped, bell-shaped, vaulted, and even rectangular (sometimes with granite ceilings and floor slabs). In exceptional cases, circular sections were also made of brick. Efforts were made in order to ensure optimal flow conditions, and also that the cross-section was adapted to the shape of the rock mass pressure line. This is due to the fact that the most advantageous shapes for masonry collectors are shapes in which no tensile stresses will occur in any part of the cross-section under the influence of external loads. Nevertheless, sewage conduits degrade over time. The boundary conditions of their use also change, which affects the magnitude of mechanical and hydraulic loads. Further use of a sewer in such a case requires its renewal, and less frequently, modernization that results from the necessity to change its function. This is usually done by introducing a new conduit into the interior of the renovated or modernized sewer, which in literature is called a liner. The aim of the analysis was to determine the thickness of the liners that strengthen the structures of brick channels with an inverted egg cross-section and with dimensions of 1050 × 700 mm, which are intended for gravitational sewage systems. The analysis included the performance of variant static and strength calculations for the assumption that the conduit after its modernization will be replaced with a conduit operating in the pressure system, which is a very rare requirement. It was assumed that the best solution would be to use a CIPP (Cured In Place Pipe) liner.

The paper presents a numerical analysis of the behaviour of egg-shaped glass-reinforced plastic (GRP) panels during the grouting process when using short relining technology. The analysis was carried out for panels subjected to temperature changes. The temperature increase was caused by the heat of hydration of the grout. It was shown that temperature had a significant effect on the stresses occurring in the panels’ walls and also on their deformations. The analysis involved grout being added in a single stage and then in two stages for comparison. The distribution of stresses and deformations were examined for panels with different wall thicknesses that ranged from 12 to 20 mm. Extensive knowledge about the grouting process and the effect of temperature on the behaviour of GRP panels during the assembly stage when using short relining technology could make this non-disruptive technology more competitive with regards to the time of its implementation and its costs when compared to traditional methods.

Natural gas hydrates are mostly formed in low-permeability and fractured muddy sedimentary formations. Adding suitable nanoparticles to the drilling fluid system can improve its filtrate resistance and fracture plugging, and effectively weaken the invasion of drilling fluid into the reservoir. However, it is likely that nanoparticles promote hydrate formation and accumulation in wellbores and will induce accidents. Therefore, this study investigated the effect of hydrophilic silica nanoparticles with particle sizes of 30 nm, 60 nm, and 80 nm and concentrations of 0.5–4.0 wt.% on hydrate formation during upward migration of methane gas using a dynamic simulation system for hydrate formation in a wellbore. The experimental results show that under the condition of methane gas migration, hydrophilic silica nanoparticles inhibit hydrate formation. The inhibition effect increased with an increase in the particle size under a constant concentration, whereas it first increased and then decreased with increasing nanoparticle concentration under a constant particle size. The strongest inhibition effect was observed at a hydrophilic silica nanoparticle concentration of 2.0 wt.%. The influence of hydrophilic silica nanoparticles on hydrate formation may be mainly determined by their hydrophilic properties, heat and mass transfer, and gas migration in the wellbore. Our research indicates that hydrophilic silica nanoparticles can be added to hydrate drilling fluid systems if their concentration can be properly controlled.

Tunnel boring machines (TBMs) are widely used in tunnel engineering because of their safety and efficiency. The TBM penetration rate (PR) is crucial, as its real-time prediction can reflect the adaptation of a TBM under current geological conditions and assist the adjustment of operating parameters. In this study, deep learning technology is applied to TBM performance prediction, and a PR prediction model based on a long short-term memory (LSTM) neuron network is proposed. To verify the performance of the proposed model, the machine parameters, rock mass parameters, and geological survey data from the water conveyance tunnel of the Hangzhou Second Water Source project were collected to form a dataset. Furthermore, 2313 excavation cycles were randomly composed of training datasets to train the LSTM-based model, and 257 excavation cycles were used as a testing dataset to test the performance. The root mean square error and the mean absolute error of the proposed model are 4.733 and 3.204, respectively. Compared with Recurrent neuron network (RNN) based model and traditional time-series prediction model autoregressive integrated moving average with explanation variables (ARIMAX), the overall performance on proposed model is better. Moreover, in the rapidly increasing period of the PR, the error of the LSTM-based model prediction curve is significantly smaller than those of the other two models. The prediction results indicate that the LSTM-based model proposed herein is relatively accurate, thereby providing guidance for the excavation process of TBMs and offering practical application value.

A novel and effective artificial neural network (ANN) optimized using differential evolution (DE) is first introduced to provide a robust and reliable forecasting of jet grouted column diameters. The proposed computational method adopts the DE algorithm to tackle the difficulties in the training and performance of neural networks and optimize the four quintessential hyper-parameters (i.e. the epoch size, the number of neurons in a hidden layer, the number of hidden layers, and the regularization parameter) that govern the neural network efficacy. This approach is further enhanced by a stochastic gradient optimization algorithm to allow ‘expensive’ computation efforts. The ANN-DE is first trained using a prepared jet grouting dataset, then verified and compared with the prevalent machine learning tools, i.e. neural networks and support vector machine (SVM). The results show that, the ANN-DE outperforms the existing methods for predicting the diameter of jet grouting columns since it well balances training efficiency and model performance. Specifically, the ANN-DE achieved root mean square error (RMSE) values of 0.90603 and 0.92813 for the training and testing phases, respectively. The corresponding values were 0.8905 and 0.9006 for the optimized ANN, then, 0.87569 and 0.89968 for the optimized SVM, respectively. The proposed paradigm is bound to be useful for solving various geotechnical engineering problems regardless of multi-dimension and nonlinearity.

A series of ring shear tests were conducted to investigate the ultimate particle size distribution of a carbonate sand. The tests were carried out under different stress levels, on three types of specimens: 1) uniformly graded specimens made of dry natural sand 2) remoulded specimens of the crushed sand after first shearing to large strains 3) specimens made of natural sand grains but with the same grading as in (2). The first series of tests on type (1), carried out to very large strains, led to apparently stable gradings, distinct for each stress level. Only limited additional particle breakage could be induced by remoulding the specimens after shearing (type (2)) and subjecting them to more shearing. Tests on specimens created at the apparently stable gradings (type (3)) but from the intact sand particles however led to significantly greater breakage. For the three types a stable, fractal grading was achieved. Analyses of the soil particles’ shape showed that the aspect ratio, sphericity and circularity reach a steady value at large strains, in parallel to reaching a stable grading. The mobilized angle of shearing resistance however was not significantly different in the different types of samples, suggesting the final grading dominates the behaviour.

Recent years have witnessed several accidents associated with tunnel face failure in cobble-rich soil in the city of Chengdu, China. Due to its lack of cohesion, cobble-rich soil can be easily disturbed by shield tunneling. Based on the general conditions of the Chengdu Metro Line 1 project, the mechanisms of face failure of tunnels in cobble-rich soil driven with earth pressure balance (EPB) machines are studied. Specifically, we present results of tests carried out using a laboratory reduced-scale model of EPB tunneling operations in cobble-rich soil. The failure kinematics and limit face pressures are presented and analyzed. Then a three-dimensional (3D) discrete-element method (DEM) model, which is able to simulate the main EPB excavation processes is employed to gain further insight into the mechanisms of face failure in cobble-rich soil. Comparisons of these results with the observations based on previous studies are discussed. The results reveal a fundamentally different tunnel-face failure mechanism in cobble-rich soil in contrast with that in clayey or sandy soils. It shows that the ground movement during face failure is sudden in cobble-rich soil, which is different from the progressive mechanism in frictional–cohesive materials. The observed sinkhole at surface takes the shape of an oval, and the failure zone behind tunnel face extends almost as far as that ahead of the face, which is different from the observations in previous studies. The failure zone is found to be wider than that of sandy soils in both the transverse and longitudinal directions.

Occurrence of the so-called residual force of an unknown value significantly disturbs interpretation of static load tests performed on piles equipped with additional measuring instruments. Screw displacement piles are the piling technology in which the residual force phenomenon is very common. Its formation mechanism is closely related to the installation method of this type of piles, which initiates generation of negative pile skin friction without any additional external factors. Knowledge of the value and distribution of a residual force (trapped in a pile shaft before starting the load test) is a necessary condition for the proper interpretation of instrumented pile test results. In this article, a clear and easy-to-use method of residual force identification, based on the analysis of shaft deformations recorded during pile unloading is presented. The method was successfully verified on two pile examples and proved to be effective and practical.

The paper will analyse and review the experience to date in determining the impact range of implementation of deeply founded structures on the displacement of the subsoil in the vicinity. With the background of these experiences, primarily empirical, the present possibilities of using numerical modelling to forecast the displacements of the terrain surface in various stages of works, that is, execution of deep excavation support systems, excavation-deepening phases with successive adding of struts, construction of underground levels and erection of the above-ground part of the building, will be presented. Based on the results of own research, conclusions on the use of 3D numerical models in spatial shaping and designing the structure of underground parts of new buildings erected in dense urban development will be presented. The characterised 3D numerical models were verified, taking into account the actual results of geodetic measurements of the completed buildings. Determining the range and forecasting the displacements of the subsoil are necessary for the design and implementation of investments due to the need to ensure the safety of erection and use of a new building and the buildings located within the area of influence.

Discrimination of seismicity distributed in different areas is essential for reliable seismic risk assessment in mines. Although machine learning has been widely applied in seismic data processing, feasibility and reliability of applying this technique to classify spatially clustered seismic events in underground mines are yet to be investigated. In this research, two groups of seismic events with a minimum local magnitude (ML) of −3 were observed in an underground coal mine. They were respectively located around a dyke and the longwall face. Additionally, two types of undesired signals were also recorded. Four machine learning methods, i.e. random forest (RF), support vector machine (SVM), deep convolutional neural network (DCNN), and residual neural network (ResNN), were used for classifying these signals. The results obtained based on a primary dataset showed that these seismic events could be classified with at least 91% accuracy. The DCNN with seismogram images as the input reached the best performance with more than 94% accuracy. As mining is a dynamic progress that could change the characteristics of seismic signals, the temporal variance in the prediction performance of DCNN was also investigated to assess the reliability of this classifier during mining. A cascaded workflow consisting of database update, model training, signal prediction, and results review was established. By progressively calibrating the DCNN model, it achieved up to 99% prediction accuracy. The results demonstrated that machine learning is a reliable tool for the automatic discrimination of spatially clustered seismicity in underground mining.

Real-time perception of rock mass information is of great importance to efficient tunneling and hazard prevention in tunnel boring machines (TBMs). In this study, a TBM–rock mutual feedback perception method based on data mining (DM) is proposed, which takes 10 tunneling parameters related to surrounding rock conditions as input features. For implementation, first, the database of TBM tunneling parameters was established, in which 10,807 tunneling cycles from the Songhua River water conveyance tunnel were accommodated. Then, the spectral clustering (SC) algorithm based on graph theory was introduced to cluster the TBM tunneling data. According to the clustering results and rock mass boreability index, the rock mass conditions were classified into four classes, and the reasonable distribution intervals of the main tunneling parameters corresponding to each class were presented. Meanwhile, based on the deep neural network (DNN), the real-time prediction model regarding different rock conditions was established. Finally, the rationality and adaptability of the proposed method were validated via analyzing the tunneling specific energy, feature importance, and training dataset size. The proposed TBM–rock mutual feedback perception method enables the automatic identification of rock mass conditions and the dynamic adjustment of tunneling parameters during TBM driving. Furthermore, in terms of the prediction performance, the method can predict the rock mass conditions ahead of the tunnel face in real time more accurately than the traditional machine learning prediction methods.

A bearing capacity evaluation for the surface strip foundation on a working platform modelled on a two-layered substrate is considered in the study. The upper layer is assumed as man-made and well-controlled and thus non-variable. The lower layer modelling natural cohesive soil is subjected to spatial variability of undrained shear strength. The random failure mechanism method (RFMM) is used to evaluate the bearing capacity. This approach employs a kinematic assessment of the critical load and incorporates the averaging of three-dimensional (3D) random field along dissipation surfaces that result from the failure geometry. A novel version of the approach considering an additional linear trend of undrained shear strength in the spatially variable layer is proposed. The high efficiency of the RFMM algorithm is preserved. The influences of foundation length, trend slope in the spatially variable layer, fluctuation scales, and thickness of the homogenous sand layer on the resulting bearing capacity evaluations are analysed. Moreover, for selected cases, verification of the RFMM based assessment obtained using random finite difference method (RFDM) based on 3D analysis is provided. Two types of analyses are performed using RFDM based on associated and non-associated flow rules. For associated flow rule which corresponds to RFDM, the RFMM is conservative and efficient and thus it seems preferable. However, if RFDM employs non-associated flow rule (much lower dilation angle for sand layer), the efficient RFMM is no longer conservative. For this situation, a combined approach that improves the efficiency of the numerical method is suggested.

The paper focuses on the possibilities of rainwater flow control in an innovative rainwater system which is equipped with a retention canals system. Sewage retention canal is a modern solution that provides effective retention of excess rainwater by using a capacity of sewer pipes and manholes. The retention is possible by using special damming partitions which have flow openings. The hydraulic working of the traditional rainwater system and the innovative rainwater system were compared with each other. The analysis was based on the results obtained from simulations using hydrodynamic modeling. Maximum possible values of rainwater outflow intensity from outlet nodes for the traditional rainwater system and the innovative rainwater system were discussed. On the basis of the analysis it was shown that the innovative rainwater system outweighs the classic rainwater one. It discharges two functions: transports and simultaneously retains excess rainwater in canals.

The main purpose of blasting operation is to produce desired and optimum mean size rock fragments. Smaller or fine fragments cause the loss of ore during loading and transportation, whereas large or coarser fragments need to be further processed, which enhances production cost. Therefore, accurate prediction of rock fragmentation is crucial in blasting operations. Mean fragment size (MFS) is a crucial index that measures the goodness of blasting designs. Over the past decades, various models have been proposed to evaluate and predict blasting fragmentation. Among these models, artificial intelligence (AI)-based models are becoming more popular due to their outstanding prediction results for multi-influential factors. In this study, support vector regression (SVR) techniques are adopted as the basic prediction tools, and five types of optimization algorithms, i.e. grid search ( GS ), grey wolf optimization (GWO), particle swarm optimization (PSO), genetic algorithm ( GA ) and salp swarm algorithm ( SSA ), are implemented to improve the prediction performance and optimize the hyper-parameters. The prediction model involves 19 influential factors that constitute a comprehensive blasting MFS evaluation system based on AI techniques. Among all the models, the GWO-v-SVR-based model shows the best comprehensive performance in predicting MFS in blasting operation. Three types of mathematical indices, i.e. mean square error (MSE), coefficient of determination (R2) and variance accounted for (VAF), are utilized for evaluating the performance of different prediction models. The R2, MSE and VAF values for the training set are 0.8355, 0.00138 and 80.98, respectively, whereas 0.8353, 0.00348 and 82.41, respectively for the testing set. Finally, sensitivity analysis is performed to understand the influence of input parameters on MFS. It shows that the most sensitive factor in blasting MFS is the uniaxial compressive strength.

PetroChina’s Karamay heavy oil reservoirs mostly comprise Jurassic deposits of braided fluvial facies that have a low porosity, extremely high bitumen viscosity, and strong heterogeneities within the reservoir. Steam-assisted gravity drainage (SAGD) dilation start-up has been proven to be an efficient reservoir geomechanical stimulation method as it shortens the steam circulation time, reduces steam usage, and increases the oil production rates. This method worked well on relatively good-quality oil sand reservoirs. However, for reservoirs with heterogeneous reservoir properties along the SAGD well, field well production data indicated the presence of non-uniform steam chambers in post-dilation SAGD wells. Enhanced dilation methods, such as vertical well assisted SAGD dilation (VW-SAGD-Dilation) are have been used, and field pilot tests have yielded promising results. This reservoir stimulation process utilizes one or more vertical wells beside the SAGD well to initiate the formation of dilation zones that connect to the dilation zone formed during the conventional SAGD dilation start-up. Compared with the normal SAGD dilation start-up, this composite dilation process creates a uniform dilation zone along the SAGD well, thus improving the thermal conformance and well productivity. The vertical well and SAGD well are connected by this composite dilation zone, which enables the combined use of the SAGD process and cyclic steam stimulation process. This paper presents our understanding and numerical case studies of this enhanced dilation strategy. The well production data from the post-dilation wells were positive (uniform steam chamber growth and higher oil rate), which highlights the promising potential of using geomechanical dilation as a reservoir stimulation method to create a large stimulated reservoir volume in heterogeneous super-heavy oil reservoirs.

The spatial information of rockhead is crucial for the design and construction of tunneling or underground excavation. Although the conventional site investigation methods (i.e. borehole drilling) could provide local engineering geological information, the accurate prediction of the rockhead position with limited borehole data is still challenging due to its spatial variation and great uncertainties involved. With the development of computer science, machine learning (ML) has been proved to be a promising way to avoid subjective judgments by human beings and to establish complex relationships with mega data automatically. However, few studies have been reported on the adoption of ML models for the prediction of the rockhead position. In this paper, we proposed a robust probabilistic ML model for predicting the rockhead distribution using the spatial geographic information. The framework of the natural gradient boosting (NGBoost) algorithm combined with the extreme gradient boosting (XGBoost) is used as the basic learner. The XGBoost model was also compared with some other ML models such as the gradient boosting regression tree (GBRT), the light gradient boosting machine (LightGBM), the multivariate linear regression (MLR), the artificial neural network (ANN), and the support vector machine (SVM). The results demonstrate that the XGBoost algorithm, the core algorithm of the probabilistic N-XGBoost model, outperformed the other conventional ML models with a coefficient of determination (R2) of 0.89 and a root mean squared error (RMSE) of 5.8 m for the prediction of rockhead position based on limited borehole data. The probabilistic N-XGBoost model not only achieved a higher prediction accuracy, but also provided a predictive estimation of the uncertainty. Thus, the proposed N-XGBoost probabilistic model has the potential to be used as a reliable and effective ML algorithm for the prediction of rockhead position in rock and geotechnical engineering.

Recent works have shown that delayed events of particle crushing are partially responsible of creep deformation in granular materials, and that Stress Corrosion Cracking promoted by high humidity within particles is the source of this mechanism. A number of experimental studies have focused on creep behaviour of water saturated samples and wetting-deformation after soaking dry material. However, there are few evidences of the effect of varying total suction in time-dependent deformation of partially saturated crushable material, and this mechanism have been rarely considered in constitutive models. The aims of this paper are to present experimental evidence of the effect of total suction on compressibility and creep of sandy sized samples from crushed rock, and to propose a simple one-dimensional elasto-plastic modelling approach based on the enhancement of an existing model. Oedometric compression tests at different total suctions are presented. The results show that compressibility and creep strains increase with both stress and humidity. The model proposed uses a time-dependent hardening law coupling suction with the amount of particle breakage. Based on preliminary calibrations, the model captures the effect of suction and time-dependent behaviour over a large range of total suction.

Excessive ground surface settlement induced by pit excavation (i.e. braced excavation) can potentially result in damage to the nearby buildings and facilities. In this paper, extensive finite element analyses have been carried out to evaluate the effects of various structural, soil and geometric properties on the maximum ground surface settlement induced by braced excavation in anisotropic clays. The anisotropic soil properties considered include the plane strain shear strength ratio (i.e. the ratio of the passive undrained shear strength to the active one) and the unloading shear modulus ratio. Other parameters considered include the support system stiffness, the excavation width to excavation depth ratio, and the wall penetration depth to excavation depth ratio. Subsequently, the maximum ground surface settlement of a total of 1479 hypothetical cases were analyzed by various machine learning algorithms including the ensemble learning methods (extreme gradient boosting (XGBoost) and random forest regression (RFR) algorithms). The prediction models developed by the XGBoost and RFR are compared with that of two conventional regression methods, and the predictive accuracy of these models are assessed. This study aims to highlight the technical feasibility and applicability of advanced ensemble learning methods in geotechnical engineering practice.

The mining of underground deposits causes the inflow of water to workings and the necessity of pumping them to the surface. The mining plant of KGHM Polska Miedź S.A. extracts copper ore in plant branches with different hydrogeological conditions. The inflowing water into the workings is characterised by variable mineralisation, which depends on the location of the branch. In the south-western part of the deposit, a low-mineralised stream with a relatively high flow rate can be observed, while the outflow of highly saline waters occurs in the north-eastern branch. Despite the activities undertaken that aim at using the pumped-off mine waters industrially, it is necessary to deposit them into the Odra River. Reducing the environmental impact on the Odra River is one of KGHM's goals, which is being implemented by stabilising its salt concentration at a safe level. The paper presents the results of a 3D simulation of brine plume propagation based on a numerical model of advection–diffusion and turbulent flow. Bathymetric data from a section of the river approximately 500 m long and point data from an Odra water quality test were used to develop and validate the model. The paper discusses the types of factors that minimise the impact of brine discharge. The developed model will be used in the future to propose solutions that accelerate the mixing of mine waters with the waters of the Odra River.