International Journal of Geomechanics

Machine learning (ML) has emerged as a powerful tool for prediction of systems behavior in many engineering disciplines. A few applications of ML techniques are available in geotechnical engineering and other fields of civil engineering. The existing review studies on the application of ML techniques in conventional geotechnical engineering and earthquake geotechnics are available in the broader areas but not specific to liquefaction phenomenon. Studies exist on the liquefaction potential of cohesionless soils using ML techniques with a varying degree of success. More studies are needed to formalize the use of ML techniques in the seismic liquefaction assessment. In this review, an attempt is made to critically review the existing literature on ML techniques as applied to the liquefaction analysis. The published literature from 1994 to 2021 has been collected, critically reviewed, and presented systematically in the form of easy to understand tables and figures. The tables are labeled based on the data requirement in ML techniques, conventional methods, and in-situ tests. A summary table highlights the relative importance of input variables in the dataset required for the liquefaction potential assessment. Limitations of conventional methods and existing ML models, the importance of a large database, relative comparison of the four developed ML models, and the requirement of probabilistic analysis are included. A gap in the existing literature is outlined and followed by a way forward for future research. It is concluded that there is a need to update the liquefaction database and modify existing ML algorithms so that they become computationally efficient and reliable for liquefaction potential assessment.

Engineered fills, such as roads, earthen dams, embankments, and earthen sites, are enhanced by compaction to increase strength and decrease compressibility. Static and dynamic compaction are the most popular and important methods in geotechnical engineering practice with different compaction mechanisms. However, the significant differences in the physical and mechanical properties between both compaction methods that have been reported in the literature were limited to clay or sand. Limited attempts have been made to quantify the variations in silt’s mechanical attributes that were caused by the sample preparation method. A series of consolidated drained triaxial shear tests and some mercury intrusion porosimetry (MIP) tests were constructed to interpret this. The results showed that static and dynamic compaction had a significant influence on the mechanical properties of the silt coupled with the molding water content. The shear strength of the dynamically compacted samples was higher than those of the statically compacted samples, as was cohesion (c), and the sample preparation method had little effect on the internal friction angle (φ). Furthermore, samples that were compacted at optimum water content (w_ot) had greater strength than those compacted on the dry or wet side of the optimum. Compared with the static compaction sample, the pore size distribution curve of the dynamic compaction sample shifted to the left with the peak pore size, distribution density, and the interaggregate pores proportion decreased. The strength discrepancy could be traced to differences in the silt structure features, for example, pore size distribution and particle orientation between statically or dynamically compacted specimens.

The undrained bearing capacity of ring foundations has received significant attention from researchers, but most of the reported studies do not reflect the effect of consolidation due to preloading. In this work, we investigated the postconsolidation undrained vertical capacity of ring foundations resting on nonhomogeneous clays using finite differential analyses and the modified Cam-clay constitutive model. The results are presented in nondimensional form and were compared with data available in the literature. The coupling influences of the internal opening of the foundation, the magnitude of the preloading, the duration of the consolidation, and the initial overconsolidation ratio of the clay were all considered. A methodology based on approximating expressions that captured the numerical results well is proposed as a way of predicting the undrained capacity gain of ring foundations for any degree of preloading and consolidation.

Jack-up rigs are important equipment for drilling and workover in offshore oil and gas fields. Their operation site is often close to long-running jacket platforms. Spudcan penetration and extraction may have a significant influence on the lateral bearing capacity of adjacent piles of platforms. In this study, centrifuge model tests for the lateral bearing capacity of piles in sand were performed, in which model piles included an isolated pile and two spudcan-affected adjacent piles. Based on the test results of the isolated pile, the lateral soil resistance–displacement (p–y) curve for an isolated pile in sand was modified. By making a comparison of the test results of the three piles, the influence of spudcan penetration and extraction on lateral bearing capacities, bending moments, and the soil resistance of adjacent piles was analyzed. The analysis showed that the lateral bearing capacity of an adjacent pile had 8.9% reduction compared with that of an isolated pile. According to the modified p–y curve of isolated piles, a p–y curve model for a pile affected by spudcan penetration and extraction was proposed by introducing the attenuation coefficient of the sand strength. It provides an option for calculating the lateral bearing capacity of adjacent piles affected by spudcan penetration and extraction in the field of engineering.

This paper presents the results of an experimental study on saturated silty clay–concrete interface shear behavior, and a shear strength prediction model is established for the silty clay–concrete interface. Based on three-dimensional (3D) scanning data of actual concrete pile surfaces, regular and irregular concrete surfaces with different roughnesses are replicated in the laboratory tests. The interface shear strength is found to increase as the roughness coefficient of concrete increases; the shear process of a regular interface shows a strain softening feature, while that for an irregular interface shows strain hardening. The effect of a normal stress increase on the residual strength of a regular interface is more obvious. The modified Grasselli roughness evaluation method is reasonably effective in characterizing both regular and irregular contact surfaces. In the framework of this method, a prediction formula for the estimation of peak strength of a clay–concrete interface is proposed, which performs favorably well in comparison with test results obtained from this and previous studies.

Stress concentration has commonly been underlined, whereas stress rotation has attracted little attention in deep mining. In this study, we aimed to introduce the phenomenon of stress rotation into the roof control of a deep longwall panel. Microseismic monitoring has revealed that roof activity presents an asymmetrical distribution, being more active on the tailgate side than on the maingate side. Front abutment stress presents a symmetrical distribution, which is inconsistent with that of the roof microseismicity. Such inconsistency implies that roof stability is not only related to stress magnitude, but is also sensitive to stress orientation. Accordingly, the characteristics of stress rotation were thoroughly analyzed. By taking a vertical plane parallel to the face-advance direction as the reference, the stress-rotation trajectory was divided into slow deviation, fast approach, and synergistic rotation stages, showing asymmetrical distribution on a stereogram. Moreover, the area sensitive to stress orientation was identified and overlaid on the stereogram with the rotation trajectory, revealing the mechanisms underlying the variation in roof stability in the face-length direction. Thus, the more-active roof microseismicity on the tailgate side is reasonably explained. Based on the overlaid images, a stress rotation–based method is proposed for improving roof stability. Its effectiveness has been proven by laser scanning results on the state of the hydraulic supports and the face wall.

The deep soil mixing (DSM) method is used to improve the mechanical and/or hydraulic properties of soils. The DSM method refers to an in situ mixing of soil, water, and hydraulic binder without excavation. After describing the specific microstructure of DSM materials compared with soils or concrete, four porosity–permeability analytical models from literature and based on microstructural parameters (i.e., porosity and pore-size distribution derived from mercury intrusion porosimetry) were applied to predict the water permeability (K) of soil–cement mixtures. After comparing the estimated permeabilities (K_est) with the experimental measurements of such a parameter (K_exp) using a triaxial cell, an indicator parameter A (=K_est/K_exp) was proposed to evaluate the ability of each model to fit the water circulation in DSM material. By calculating the average ratio 〈A〉 from the 24 materials in the studied database, three of the four applied models satisfactorily estimated the water permeability of the studied DSM materials. To reach an 〈A〉 value closer to 1, an extension of the Aït-Mokhtar's model, initially developed for cementitious mortars, was proposed by considering the specificities of the DSM materials. If previous models used only the contribution of nanopores less than 100 nm to the water flow, the extended model considered the contribution of all pore populations in the permeability calculation. Such an approach resulted in a significant improvement in the matching between the calculated and measured permeabilities.

Assessing the effect of rainfall on the short-term and long-term stability of earthen structures warrants studying the earth pressures of unsaturated soils under transient flow. This paper presents a variational approach applied to the limit equilibrium method for calculating the resultant force of active and passive earth pressures in unsaturated soils subject to transient flow conditions. The method is developed using the unified effective stress representation and utilizes a log-spiral slip surface. The latter eliminates the need to make any static assumptions in the derivation and leads to a more comprehensive failure mechanism than the planar failure mechanism used in most alternative methods. The proposed analytical solution accounts for the infiltration rate, batter angle, backslope inclination, and surcharge. For verification purposes, the results from the proposed formulations are compared against those from several existing analytical solutions. A set of parametric studies is performed to examine the effects of rainfall on active and passive earth pressures for different soil types. The results show that the soil type, mechanical parameters, and shape of the failure surface will significantly affect the earth pressure coefficient. The proposed formulations provide a valuable tool for evaluating the effect of rainfall on the behavior of earthen structures and forensic studies of failed earthen structures.

Constructing meta-models and selecting a suitable deterministic analysis method are important to improve the computational efficiency and accuracy of the nonintrusive reliability analysis of a spatially varying soil slope. However, existing meta-models are not applicable to the slopes considering multiple parameters with high spatial variability. Moreover, it is difficult to identify the failure modes when the spatial variability is high by using deterministic analysis methods based on slip surface search. Therefore, a nonintrusive stochastic strength reduction finite-element method (SRFEM) is developed based on the multi-input convolution neural networks (CNNs) and ABAQUS 2016. The SRFEM developed based on ABAQUS is adopted as the deterministic analysis method to avoid the uncertain search for the critical slip surfaces of slopes with high spatial variability. A multi-input CNN is proposed to construct the meta-model to avoid the “curse of dimensionality” and replace the overmuch times of time-consuming finite-element simulations. It can fit the relationships between multiple spatially varying parameters and the factor of safety by processing different parameters with different streams of CNNs. Two illustrative examples show that the proposed method can accurately identify the failure modes of slopes with different degrees of spatial variability. The agreement of the reliability results based on the proposed method and the general random finite-element method (RFEM) shows the high accuracy of the proposed method. The time cost of the proposed method can be reduced to 6.0 × 10⁻³ times that of the general RFEM, verifying the high computational efficiency of the proposed method. The multi-input CNN also shows higher fitting accuracy and better interpretability than the single-stream CNN and the support vector machines (SVMs). The generalization ability, accuracy, and efficiency of the proposed method show its potential to carry out the reliability analyses of slopes with multiple spatially varying parameters.

To reduce pressure fluctuations in the excavation chamber during earth pressure balance (EPB) shield tunneling, it is necessary to inject foam into excavated soil to improve its compressibility. However, the effect of grain gradation on the undrained compressibility of foam-conditioned soil remains elusive. In this study, a series of one-dimensional undrained compression tests using a custom-designed large-scale compression device were carried out to study the role of grain gradation on the compressibility characteristics of foam-conditioned coarse-grained soils. The results demonstrated that the compressibility decreased with an increase in grain diameter (d₆₀) and uniformity coefficient (C_u) but with a decrease in the coefficient of curvature (C_c). In the unloading stage, with higher d₆₀ and C_c but lower C_u, the void ratio and pore pressure initially stayed constant with a larger range of total vertical stress (σ_v), but the effective vertical stress ${\sigma }_v^{\mathrm{\prime }}$ sharply decreased to zero at lower σ_v. With further unloading, pore pressure (u) and total lateral stress (σ_L) started to decline significantly at lower σ_v. With higher d₆₀ and C_c but lower C_u, there were lower lateral pressure and total lateral pressure coefficients, indicating a lower workability of muck. Although the foam injection ratio of 20% was not sufficiently high to effect compression dominated by air in the early loading stage, the soil showed air-governed behavior in the unloading stage when e/e_max became larger than e_t/e_max, which was approximately equal to 1.0, regardless of soil gradation.

Paraffin-based phase change materials (PPCMs) have been recognized as a potential tool to regulate soil temperature and reduce the freeze–thaw impact on soil. However, as a new filling material for geotechnical structures, such as dam core walls, the engineering applicability of PPCM-mixed soil (PS) needs to be further studied. In this paper, the influences of PPCM on the soil engineering properties (i.e., compaction characteristics, consolidation properties, mechanical behaviors, and permeability) and the PPCM leakage behavior were studied through a series of laboratory experiments. The results showed that the addition of PPCM results in an obvious decrease in the maximum dry density and optimum water content, a slight change in the compression coefficient, a significant reduction of the plasticity and cohesion, and an enhancement of the friction angle and impermeability. In addition, the PPCM in PS specimens will leak under the constant hydraulic gradient or vacuum saturation, and the leaked mass is positively related to the initial PPCM content. The PS specimens after 64-day seepage will still have a good impermeability. This paper is an important prerequisite for the practical applications of PS and provides a basis for its performance improvement.

Deeply embedded caissons with a significantly large diameter, widely used as bridge foundation elements, are generally subjected to the complex combinations of several environmental and operational forces. The undrained capacities of rigid caisson foundations embedded in soils with increasing strength with depth and under combined loading are investigated using finite-element analyses. Estimation of uniaxial capacities followed by biaxial failure envelopes is carried out to capture the effect of vertical load on ultimate transverse capacities. Two-dimensional failure envelope in H-M loading space is presented to define the combined ultimate loading state of the circular caissons where the shape of the envelope is illustrated by a series of closed form expressions. The trajectory of the V-H-M failure envelope obtained from numerical analysis is validated by the mechanisms of upper bound limit analysis. Finally an example problem is provided to showcase the applicability of the expressions proposed to execute the design approach.

The tunnel boring machine (TBM) is widely recognized for its advantages of speed, efficiency, and environmental friendliness; however, the observation and evaluation of the excavated rock surface are rendered difficult because of the TBM design defects. To improve construction efficiency and safety, it is crucial to study the real-time estimation of tunnel-excavated rock strength. In this study, the TBM data from Chongqing Metro Line 9 of the Zhongcong Tunnel showed that the thrust force (force normal to the rock surface) F_N is proportional to the cutting depth per revolution P, and the torque of the cutting rock T_R/thrust force F_N is proportional to P^0.5. According to the engineering practice of the Zhongcong Tunnel and previous studies, we developed a real-time estimation method for tunnel-excavated rock strength based on three TBM operating parameters: F_N, T_R, and P. A statistical analysis of the construction data of 17 TBM rock tunnels showed that the two dimension parameters γ₁ and γ₂ of the estimation formula are correlated with the cutter diameter d/tunnel diameter D and number of cutters n. The results of the fitting analysis were used to obtain the calculation methods for γ₁ and γ₂. Subsequently, the proposed formula for rock strength estimation was applied to both the Zhongcong Tunnel (soft rock) and the Maiko Tunnel (hard rock), and the generalizability and reliability of the estimation formula were verified. The results provide a more practical and feasible approach to estimating excavated rock strength in real time. Meanwhile, the proposed method has an important application value as it can improve the efficiency and safety of TBM tunnel construction.

This study evaluated the strip foundation bearing capacity and settlement on sandy soil reinforced with scaled geocomposite and geotextile by laboratory model testing. Therefore, several reinforcement compounds were used in different layers (one to four), including identical and various vertical distances from each other and the footing base. The geocomposite used in this study consisted of geogrid and geotextile layers. The most crucial consideration in this research was the application of scaled reinforcement with a factor of 7.5 to create more precise laboratory models. Based on the results, all models failed in the lower settlement, which in most cases was 7% of the width of the footing. Test results indicated that the maximum increasing factor in bearing pressure of foundation in the case of ultimate bearing was 3.4, which could be related to using four layers of geocomposite at optimal distances. The percentage of decreasing footing settlement at the maximum value was about 86% in the two-layer geocomposite state. Its value remained constant with increasing number of layers. Percentage variation in bearing capacity value indicated that geotextile had a better performance in increasing the bearing capacity at the high settlement level. It was close to the values of the geocomposite.

The calculation of the settlement of a peat soil foundation is based mainly on the formula obtained by data fitting or simply modifying the universal calculation formula. The deformation mechanism reflected by these calculation methods is indistinct. Therefore, several model tests were carried out using the embankment structure of the Dali–Lijiang expressway as a prototype, and the settlement was analyzed with the one-way consolidation deformation and in situ settlement of other researchers. First, the S − logt (S is the deformation) curve of peat soil showed no obvious inverse “S” change, which is different from that of soft soil or mucky soil. According to the S − t curve, the deformation process of peat soil can be divided into three stages. Second, the relationship between the compression modulus and load was analyzed and combined with the settlement calculation results. The compressive modulus between 12.5 and 25 kPa [E_{n(0.125–0.25)}] was proposed to be used when calculating the settlement of peat soil foundations by the layerwise summation method. Moreover, the compression modulus of the three consolidation stages was proposed to be used to calculate the settlement of peat soil foundations by the layered sum method. At the same time, the influence of confinement on the compression modulus of the third consolidation stage should be considered and corrected. The accuracy of the settlement calculation method proposed in this paper was higher than the accuracy of the standard layerwise summation method, regardless of whether the horizontal limiting conditions of the standard method are modified.

The tensile strength of unsaturated soils is essential to assess the safety and reliability of earthen structures over a prolonged period. This specific property of unsaturated soils plays a significant role in desiccation cracking and tensile failure of earth structures. The tensile strength of unsaturated soils can be measured through various direct and indirect testing methods in the laboratory and through different models available in the literature. These models can rapidly predict the tensile strength of unsaturated soils over a wide suction range. However, a practical framework to capture the tensile strength behavior with variations in soil type, tensile strength testing methodologies, and sample preparation methodologies has not been probed. In the present study, a novel tensile strength framework for both coarse- and fine-grained soils is proposed to determine the tensile strength characteristic curve (TSCC) over the entire saturation range from the respective soil water retention curves. A critical assessment of the variation of the parameters for the TSCC model for various types of soils, tensile strength testing methods, and sample preparation methodologies was also performed in this study. Further, the proposed framework was validated with several experimental data sets available in the literature for both coarse- and fine-grained soils. A good agreement was found between the experimental data and predicted results in most cases, with variation in soil type, sample preparation, and testing method. Therefore, the proposed model can be effectively used to determine the tensile strength of unsaturated coarse- and fine-grained soils over the entire saturation range and can be widely utilized for various engineering assessments in real-field situations.

A method for an a priori estimation of the frost heave ratio for clayey soils, considering the degree of their water saturation upon a closed system of freezing, is proposed. The proposed approach involves the introduction of a new parameter in the mechanics of frozen soils, namely, the frost heave index of clayey soils, which characterizes the degree of frost heaving of soil relative to the maximum height of the frost heave of soil that is fully saturated with water. An introduced parameter is used in the final formula to estimate the frost heave ratio of clayey soils. Laboratory experiments were performed to determine the height of frost heaving of clayey soils with a variation in their moisture and porosity to verify the proposed approach. A comparison of the experimental and calculated results showed a suitable convergence sufficient for the practical use of the formula to determine the frost heave ratio in a closed system of freezing. An example of a calculation to justify antiheaving measures using the proposed estimation of the frost heave ratio is given.

The response surface method (RSM) is used in slope reliability evaluation. Abundant sample points are required to establish an accurate response surface modeling. Therefore, the extreme computational effort is time-consuming, especially when the strength reduction method (SRM) is adopted in deterministic analysis modeling. In this study, a supplementary sampling method (SSM) is proposed based on the theoretical relationship between the safety factors and strength parameters. The proposed SSM can generate any number of supplementary sample points to establish the response surface modeling. Combined with SRM, the SSM-based RSM for slope reliability analysis is introduced in detail. Three slope examples are examined using the proposed SSM. The results show that, compared with other reliability methods, the SSM-based RSM can not only obtain a more accurate failure probability but also requires fewer sample points. The computational efficiency of the RSM is a three-fold improvement compared to previous methods.

This paper proposes an analytical method based on large-deformation theory to predict the mechanical responses of the existing pipeline. The hypothesis that the soil and structure can be contacted or separated is proposed. The method primarily obtains the pipeline’s horizontal displacement and axial force and predicts the soil gap formation potentially beneath the pipeline. Moreover, the deformation and axial force of the pipeline obtained by our method agree well with the in situ data and measured data from the centrifuge test. Compared with the other methods, the settlements and internal forces have the same characteristics except for the axial force. The results based on the small deformation theory are larger because of ignoring the axial displacement and axial force. The axial force can offset parts of the vertical load acting on the pipeline. When using the large-deformation theory without considering the horizontal displacement, the phenomenon of pipelines being only in tension is not in conformity with reality. In our proposed method, the axial force is compressive at the centerline and tensile at the hogging zone, which is consistent with the centrifuge test. It indicates that our method has extensive applications. Besides, the effects of the proportionality coefficient of vertical displacement to horizontal displacement h, coefficient of subgrade reaction k, equivalent bending stiffness EI, and maximum settlement of the underlying soil $S_{max}^2$ are studied. The settlements and axial forces of the existing pipeline increase with the increase of k, $S_{max}^2$ and decrease of h, EI. In contrast, the axial force increases with the increased EI.

Expansive soils are extremely susceptible to swelling and shrinkage associated with seasonal changes in their moisture content. This can cause serious damage to structures constructed on expansive soils. Stabilization using additives is widely used to mitigate the concerns of expansive soils by controlling the changes in volume, thereby increasing the strength and durability of the soils. Multistabilizers, including a low percentage of conventional additives with nonconventional additives, can reduce the costs of stabilization. The performance of the stabilized soil can be enhanced by admixture optimization to ascertain that the agent’s ratios are quite adequate to meet the optimal levels of modification. Unfortunately, despite the development of advanced techniques for investigating the stabilization of expansive soil, limited studies are devoted to the optimization of admixture to soils. Hence, in this research, an approach based on rough set theory, a mathematical data mining approach, is presented for multiagent optimization for efficient treatment of expansive soils. The multiadditives used in this study are almond shell ash (ASA), volcanic rock dust (VRD), and portland cement (PC). The differential free swell test, the California bearing ratio test, and the unconfined compressive strength test were applied for assessing the characteristics of the stabilized soil specimens. The obtained results and the confirmatory tests demonstrated that a combination of 20% ASA and 20% VRD with 8% PC provides the optimum results in the properties of expansive soil. The microstructural analysis denoted an enhanced modification of the soil engineering properties. The outcomes of this research confirm that the rough set optimization procedure is highly suitable for application in multiagent optimization for the efficient treatment of expansive soils.

Torsional load arising from horizontal load eccentricity has a significant effect on the ultimate horizontal load carrying capacity of tripod bucket foundations in clay. In this paper, the effect of torsional load on the failure mechanism and ultimate horizontal capacity of tripod bucket foundations in undrained clay was studied using 3D finite-element analyses. Three single buckets were connected to form a rigid tripod bucket foundation system. Two different shear strength profiles, that is, uniform and linearly increasing with depth were considered. A detailed parametric study was conducted by varying the bucket spacing and embedment depth to explore the effects of different governing parameters. All the analyses were carried out for smooth and rough interface conditions. The influence of torsion on horizontal capacity is presented in the form of combined horizontal-torsional (H-T) failure envelopes. Reductions in ultimate horizontal capacity with increasing load eccentricity were quantified for the ease of practice. Finally, an expression for estimating combined H-T capacity is provided as a function of normalized spacing and embedment. Additional analyses were also performed to evaluate the effect of moment arising from vertical eccentricity on the combined H-T capacity.

Abstract

In pile-supported embankments, particle breakage commonly has a significant influence on soil properties, governing the load transfer induced by the soil arching effect. Thus, the influence of particle breakage on the soil arching effect in pile-supported embankments needs to be considered, especially through the operation period. Based on the three-dimensional discrete element method simulation on a pile-supported embankment model, the influence of particle breakage on the evolution of the soil arching effect is unraveled, including the initial, arching, static load, cyclic load, and after cyclic load states. The results indicate that particle breakage leads to a more significant degradation of the soil arching effect under the static load. Three stages for the evolution of the soil arching effect can be identified in the case of particle breakage under the cyclic load: the destruction of the original soil arching, the relaxation of the fragment, and the reformation of a new soil arching. These three stages are sequential due to the successive particle breakage above the pile and the primary and secondary independent particle breakage above the subsoil from a microscopic perspective. The new soil arching is less effective than the original one. In addition, the accumulative settlement on the embankment surface increases under cyclic load due to particle breakage, while the differential settlement on the embankment surface changes slightly.

Practical Applications

In this study, the evolution of the soil arching effect in the pile-supported embankment was investigated through two cases with or without particle breakage. For a practical scenario, the particle breakage or the crushability index of the fill can be approximately evaluated based on the mineralogy of the soil particle. Under a given design load, the occurrence of the particle breakage can be assessed accordingly by an experimental method or numerical modeling. When particle breakage occurs, the soil arching effect is weakened, resulting in a decreasing load transferred to the pile cap. In other words, particle breakage is beneficial for the bearing capacity of piles. Nevertheless, when particle breakage occurs, the settlement on the fill surface increases, which is not favorable to the superstructures. In the engineering practice, these two issues should be jointly considered in pile-supported embankments with possible soil particle breakage, according to the requirement and desire of the design.

Experimental studies on different materials demonstrate a wide variety of strength behavior. Although the classical failure/yield criteria might be suitable for special cases, they cannot accurately cover the whole range. Thus, the need to employ a unifying model, capable of comprising various criteria seems inevitable. This paper develops a flexible Lode-dependent failure and yield criterion that allows the generation of various shapes in both deviatoric and meridian planes. The model, thus, can cover all the classical criteria including von Mises, Drucker–Prager, Tresca, Mohr–Coulomb, Matsuoka–Nakai, Lade–Duncan, and other recently developed failure/yield criteria. Moreover, the model can consider tensile strength that makes it appropriate for cohesive materials as well. After defining the model, the condition for its convexity and differentiability is discussed. Then, the results derived from the developed model are compared with some of the well-known criteria and different sets of experimental data to attest to its accuracy and applicability in predicting the strength and yield of different materials.

Characterizing the mechanical behavior of jointed rocks is important to understand the behavior of structures in rock masses. Jointed rocks can be composed of persistent and nonpersistent joints where the impact of nonpersistent joints requires careful consideration for an accurate rock mass mechanical characterization. Most previous investigations into nonpersistent jointed rocks focused on joints with smooth surfaces, and a few experimental studies focused on nonpersistent rough joints and nothing specific has been reported numerically. Therefore, this study investigated several synthetic jointed rocks with nonpersistent rough joints numerically under uniaxial compressive and shear loadings. The PFC2D-based synthetic rock mass (SRM) approach was adopted to assess the impact of bridge angle (γ) and length (L), joint roughness coefficient (JRC), and normal stress (σ_n) on the shear strength (τ_n) and cracking in jointed rocks with nonpersistent rough joints. In addition, the impacts of γ, L, JRC, and joint inclination (θ) on the uniaxial compressive strength (UCS or σ_cm), elastic modulus (E_m), and failure pattern in the jointed blocks were examined numerically. First, several numerical models were developed and verified by the laboratory data, followed by an extensive parametric study to assess the effects of the defined parameters further. The effects of JRC and σ_n on τ_n were more pronounced than γ and L due to the formation of interlocking cracks, which could cause significant shear resistance during shear loading. In addition, the numerical results under axial loading revealed that an increase in θ could reduce the deformation modulus and the value of the other parameters, in particular the JRC, could lead to an increase in the strength of jointed samples.

This study proposed an analytical approach for predicting the end-bearing capacity of driven piles that are subjected to axial loads in saturated and unsaturated soils. This is a generalized approach in which the stress characteristic method is employed successfully for both saturated and unsaturated soils. An iterative technique computer code was developed for the proposed analytical approach extending the finite difference method to develop solutions with the aid of MATLAB (version 2019a) that provides graphical output to visualize the results. The results from the proposed approach were compared against measurements for 13 pile load tests that include 11 in saturated soils and 2 in unsaturated soils, with good agreement. In addition, numerical analyses were performed using ABAQUS (version 6.14) to simulate the driven pile penetration and pile loading by employing the arbitrary Lagrangian–Eulerian adaptive mesh methods. The comparisons between the numerical predictions and measurements from a published model pile test suggested that the ultimate bearing capacity is well predicted by the finite-element model in comparison to the proposed analytical method. However, the proposed analytical method was simple for use in engineering practice applications to estimate the end-bearing capacity of pile foundations in both saturated and unsaturated soils.

Excavation-induced unloading effects and dewatering-induced groundwater seepage inevitably result in basal soil reverse consolidation in deep excavation. However, this reverse consolidation process is rarely considered because most previous analytical methods were developed based on total stress analyses. This study proposes a one-dimensional reverse consolidation model for basal soil of deep excavation. Based on the consolidation theory proposed by Terzaghi, governing equations of soil reverse consolidation caused by excavation and dewatering were separately established. The continuous drainage boundary was introduced to describe the construction processes. The reverse consolidation responses of basal soil were obtained by superimposing analytical solutions of the excess pore-water pressures that resulted from excavation and dewatering. The proposed model was verified by existing solutions and a well-documented excavation case history. Moreover, the reverse consolidation characteristics of basal soil were investigated by parametric analyses. Results indicate that the excavation-induced variations in pore-water pressure decreased with increasing excavation depth. The final pore-water pressure and effective stress were predominantly affected by excavation duration rather than interval distribution. In addition, a smaller coefficient of consolidation led to lower pore-water pressure and greater effective stress at a given excavation depth.

Abstract

Hydraulic hysteresis and mechanical behavior [i.e., void ratio (e) dependency] are fundamental in unsaturated soil behaviors. Applications of the upper bound theorem to account for these aspects in unsaturated soil mechanics have never been reported in geotechnical problems, such as bearing capacity, to the best of the author’s knowledge. Therefore, this paper presented modifications to and the development of a computational limit analysis method [discontinuity layout optimization (DLO)], which considered the influence of hydraulic hysteresis, mechanical behavior, and changes in soil fabric on strength. The modified approach unsaturated DLO (UNSAT-DLO) was utilized to carry out a parametric study for a bearing capacity problem, where the numerical results were compared with 1g physical modeling tests that were conducted for a surface strip footing that was placed on sandy soil. The comparison demonstrated that the UNSAT-DLO approach could capture the nonlinearity of the experimental results within a small range of suction (s) profiles. In addition, the bearing capacity scatter-induced hydraulic hysteresis, mechanical behavior, and soil fabric changes, even at approximately the same applied s, were reasonably well defined by the UNSAT-DLO approach, because the parameters were carefully assigned. In addition, differences in the failure mechanisms due to the hysteresis were investigated to provide insights into distinct changes in the failure modes due to the hydraulic hysteresis and mechanical behaviors.

Practical Applications

The conventional design of shallow foundations was based on dry or saturated states, according to the saturation conditions, which led to a conservative design. Due to the drying and wetting cycles, soils are frequently found in unsaturated conditions. The additional strength that was imparted to soils, which was attributed to the effect of partial saturation, led to more stable geotechnical structures. The literature confirmed that suction (s) had a significant effect, and a ≤10-fold increase in bearing capacity was obtained when unsaturated cases were compared with the fully saturated conditions. Further evidence of this was the stability of vertical cuts (trenches), embankments, retaining walls, and railways. With a better understanding of the mechanics of unsaturated soils, the design of the geotechnical structures could impart additional strength due to partial saturation. The utilized (UNSAT-DLO ) approach, which accounted for the effect of s and saturation, exhibited good agreement with the experimental data. In addition, it considered hydraulic and mechanical hysteresis. Therefore, the approach was a robust tool that could help with safe structures and more economical designs.

The limit equilibrium method that considered the concept of lateral earth pressures exerted on retaining walls has been applied to determine a closed-form solution for the bearing capacity in a foundation on cohesive soils that were improved by a granular trench (column). The improvement factor (IF) as a design criterion was defined by the bearing capacity ratio with the trench to that without it. The terms of the obtained bearing capacity expression and values of the IF were explored by varying a wide range of width ratios (W/B) where the width of the trench was W, and the loading area was B. The results showed that the IF increased significantly with an increase in W/B and further improved with increases in the friction angle (ϕ) of the trench and decreases in C_u/γB, where the undrained cohesion was C_u and the unit weight of the cohesive in situ soil was γ. Moreover, the improving rate increased with increasing ϕ and for W/B<1 was more effective. In addition, the optimum depth of the trench (H) depended on W/B and ϕ. When W/B increased ≤1, H/B linearly increased from 0.5 to a certain value regarding the value of ϕ and afterward remained approximately constant. In this study, the effect of γ on the bearing capacity was highlighted. To ensure the improvement in the bearing capacity, the minimum required values of ϕ were determined through lower bound limit analysis.

The field measurements and numerical simulations were conducted to evaluate the response of the nearby metro tunnels to the construction of diaphragm walls in Shenzhen, China. Results showed that the diaphragm wall construction significantly affects the metro tunnels’ horizontal deformation. The diaphragm wall construction was numerically modeled to perform a parametric study. The results showed that the concreting stage could reduce the differential deformation of the tunnel structure, thus effectively reducing the internal force of the tunnel structure. The synchronous interval construction of the diaphragm wall was more conducive to controlling deformation than the single-wall constructs in sequence. Further, the horizontal displacement of the metro tunnels at different locations on the diaphragm wall was evaluated. Finally, soil strengthening is discussed as an effective measure for protecting metro tunnels from being affected by diaphragm wall construction. This study helps assess the potential response of metro tunnels to adjacent diaphragm wall construction for new deep excavation projects or future workings.

In this study, a field test on the consolidation of sand-doped improved soft soil under surcharge-vacuum preloading is carried out. By analyzing the vacuum pressure transmission efficiency, the settlement rate, and the dissipation rule of excess pore-water pressure of sand-doped and original soft soil foundation in the consolidation process and comparing the indices of vane shear strength, water content, and void ratio before and after consolidation, the technical superiority and reinforcement effect of the sand mixing method for improving soft soil are comprehensively evaluated. Considering the fact that the vacuum pressure decreases linearly along the depth and the surcharge preloading is applied linearly, an assumption of equal volumetric strain is introduced to reflect the lateral deformation, and an axisymmetric nonlinear radial consolidation model of sand-doped soft soil is established. The model accuracy is verified by model degradation research and a field test. The results show that the sand-doped method (20% sand content) can improve the transmission efficiency of vacuum pressure, reduce the loss of vacuum pressure along the prefabricated vertical drains, accelerate the dissipation rate of excess pore-water pressure, increase the final settlement and shear strength, reduce the final void ratio and water content, and improve the reinforcement effect of deep soil. A nonlinear consolidation solution of the sand-doped soil with linear surcharge preloading considering the actual negative boundary conditions can reflect the consolidation behavior more accurately. If lateral deformation and linear surcharge preloading are not considered, the consolidation rate of soil will be overestimated.

In this research, creep behavior during shearing (deviatoric creep) was studied through a series of constant water content triaxial tests for unsaturated soil under different stress histories. The test samples were first compressed to a net confining pressure of 400 kPa and three different overconsolidation ratios (OCRs = 1, 2, 4) were obtained by unloading the confining pressure from 400 to 200 and 100 kPa, respectively. The samples were then dried with different sections (0, 150, 300, and 500 kPa) by the axis-translation technique. After that, the unsaturated samples were sheared until the deviatoric stress ratio reached around 1.2, and the axial stress was maintained during shearing until failure or reaching 200 h. The responses of samples, including the time, volumetric strain, axial strain, degree of saturation, and suction, were monitored during the triaxial testing. The test results are presented and thoroughly discussed in this paper, including the relationships between the deviatoric creep effect and stress history effect under different suctions.