Géotechnique Letters

Online ISSN: 2045-2543
a shows the mean magnitude of fluctuations, )u*, and the mean magnitude of normalized fluctuations, )V*, as a function of &! (taken from ! = 0). The mean displacement fluctuation increases monotonically throughout the test, from 0.066 (for &! = 10-3 ) to 6 mm (for &! = 0.25), this final value corresponding to about 10% of the average displacement of the grains between the beginning and the end of the shear test. Note that the smallest displacement fluctuation is well above the accuracy of Tracker. As far as the mean magnitude of normalized fluctuations )V* is concerned, a decrease of )V* from 3 to 2 is observed; this implies that the average fluctuation of local shear strain )u*/)d* is two to three times larger than the global strain window &!. b suggests that there is a direct link
Probability density function (pdf) of normalized fluctuations V" projected on the major principal direction " of strain for two different strain windows, "! = 10-3 and "! = 2#10-1. The major principal direction is given by " = tan-1 [!/(#x $ #y)]. The inset shows the kurtosis of the pdf as a function of "!.
In a granular material, a macroscopically homogeneous deformation does not correspond to a homogeneous displacement field when looking at the individual grains. The deviation of a grain displacement from the value dictated by the continuum field (referred to as fluctuation) is likely to hold valuable information about the characteristic length(s) involved in grains' rearrangement, which is the principal mechanism of irreversible deformation for granular materials. This paper shows a selection of results from a series of shear tests on a 2D analogue granular material. We have followed the route opened by the pioneering work of Radja\"i and Roux (2002), and used the same framework to analyze our experimental data on displacement fluctuations. Digital Image Correlation has been used to measure and characterize the displacement fluctuations. The analysis of their spatial organization reveals the emergence of a minimum length scale that is in the order of 10 times the mean particle size.
Schematics of three grains (grains' centres indicated by i, j and k) with connecting capillary bridges. The space has been divided using Voronoi tessellation, and water volumes within the cells vary by mass flux through cell boundaries, as ˙ Θ ij and ˙ Θ ik .
Model parameters used in the simulations.
Schematics of contact angle dynamics: (a) Contact angle limits, shown for a water drop sliding along a tilted surface. (b) Contact angle dynamics during wetting and drainage.
A particle-water discrete element-based approach to describe water movement in partially saturated granular media is presented and tested. Water potential is governed by both capillary bridges, dominant at low saturations, and pressure of entrapped air, dominant at high saturations. The approach captures the hysteresis of water retention during wetting and drainage by introducing local evolution of liquid-solid contact angles at the level of pores and grains. Next, extensive comparisons against experimental data are presented. While this is made without involving any fitting parameters, the method demonstrates relative high success by achieving a correlation coefficient of at least 82%, and mostly above 90%.
Grain breakage affects a number of geotechnical engineering problems. In this research study, the breakage of an artificial, porous granular material (light-expanded clay aggregate (LECA)) has been studied in one-dimensional compression with both standard laboratory techniques and in situ X-ray tomography during loading. X-ray tomography has revealed that there is a wide distribution of internal porosity among LECA particles, and particle tracking has been used, for the first time, to give an objective measurement of each particle's life expectancy. Links between micro- and macro-scale quantities are discussed.
Computation times for analyses 
Offshore pipelines are often laid directly on the seabed, particularly in deep water. Cyclic changes in the temperature and pressure of the pipe contents can lead to cyclic growth and contraction of (planned) lateral buckles. This results in severe plastic deformation of the near-surface soil as it is progressively scraped away and accumulated into berms by the lateral sweeping motion of the pipe. Quantifying the resistance offered by the soil is important in design, notably for fatigue calculations. This paper presents a new computational approach for modelling pipe-soil interaction during events dominated by large transverse pipe displacements. The main novelty is that a three-dimensional (3D) structural model of the pipe interacts with a series of discrete two-dimensional (2D) (plane strain) soil domains. Each 'slice' of soil is analysed using a coupled Eulerian-Lagrangian finite-element technique, though other large-displacement numerical methods could be employed. The concept is similar to recently developed macro-element approaches, except that the pipe-soil interaction at each (assumed) planar cross-section is captured by a 2D continuum model rather than by a 2D userdefined spring model. This is more efficient than a full 3D analysis of the pipeline and surrounding soil, yet it allows results of comparable accuracy to be obtained. This paper illustrates the new methodology with a simple calculation.
This technical note uses bolt-reinforced jointed granite to investigate the relationship between the joint angle and the reinforcing effects of bolt. Three-dimensional-printed (3DP) bolts were manufactured with a real bolt as the model prototype, the dimensions and surface geometries were scaled down, and the mechanical properties were similar to those of the prototype bolt. The cylindrical granite specimens were prepared with each separated by a through-going inclined joint plane, the bolt and the granite were combined by way of bonding agent and nuts. The bearing capacity of the system, the loading status of the bolt, and the displacement/strain changing patterns on the surface of the granite were comprehensively investigated. The results indicated that granite tended to crack when the joint angle was small and that a higher joint angle weakened the bearing capacity of the system, even with the reinforcement of the bolt. For specimens with joint angles of 45 and 60°, a shearing tendency and strain concentration zones nearby the joint plane were clearly observed. This study provides a new methodology for analysing the anti-shearing behaviour of bolt in fractured hard rock.
A physical model based on transparent soil technology is presented as a novel laboratory tool for ground contamination research. The model aims to reproduce both unsaturated and saturated conditions within a Spanish soil formation known as Miga sand. A soil surrogate made of a water-absorbing polymeric gel was employed. Multi-phase flow from a leaking underground storage tank was simulated in both saturated and unsaturated conditions in order to illustrate the usefulness of the technique in simulating a hypothetical spill of a light non-aqueous phase liquid under well-controlled test conditions. The plume observed was found to be spatially variable. The technique permits, for the first time, the simulation of specific grain size distributions in a transparent soil surrogate and visualization of plume development in three dimensions. Preliminary results indicate that the model is a potentially powerful tool for visualising pollutant transport in porous media. A brief description of the test setup, and a summary of the advantages, limitations, and prospective applications are presented.
Semi-empirical solutions for fractal-based hydraulic properties of unsaturated three-dimensional fracture networks are presented, in which three assumptions are made: (a) the permeability is scale dependent, (b) each fracture cuts through the model and (c) the difference between fractal dimensions for fracture aperture distribution and tortuous length of a fracture is <1·7. The results show that the relationship between the relative permeability of the gas phase and saturation agrees well with those reported in the literature. The fractal dimensions for fracture aperture distributions of the fluid and gas phases are between 2 and 3. The permeability of the saturated fluid phase is influenced by the fractal dimension for the tortuous length of a fracture and maximum fracture aperture. With the increase in saturation, the variation range in fractal dimension for fracture aperture distribution of the fluid phase for different fracture porosities decreases; however, the variation range increases for different ratios of maximum aperture to minimum aperture. The permeability of both phases is more sensitive to the fractal dimension for the tortuous length of a fracture than to the fracture porosity and/or the ratio of maximum aperture to minimum aperture.
Compacted coal fly ash–carbide lime (two industrial by-products) blends have great potential application in geotechnical earthworks such as sub-base and base of pavements, beds of pipelines and spread footings, embankments and in erosion control. Pozzolanic reactions occur between silica and alumina in amorphous phases (found in the fly ash) and Ca 2+ (existent in carbide lime) in an alkaline environment. However, pozzolanic reactions are notoriously slow at ambient temperatures, requiring long curing periods for certain applications. Therefore, this study aims to verify the influence of application of distinct salts [sodium chloride (NaCl), magnesium oxide and potassium permanganate] in the acceleration of unconfined compressive strength (q u) increase of coal fly ash–carbide lime blends. Results have shown that small amounts of certain salts allowed to accelerate q u increase of coal fly ash–carbide lime blends. Among the studied salts, sodium chloride presented the best results after 28 days of curing. Coincidently, sodium chloride is also relatively cheap and of great availability, making it convenient for extensive use.
A major UK initiative entitled Mapping the Underworld (MTU) is seeking to address the serious social, environmental and economic consequences arising from an inability to locate - accurately and comprehensively - buried utility service infrastructure without resorting to extensive excavations. MTU aims to develop and prove the efficacy of a multi-sensor device for accurate remote buried utility service detection, location and, where possible, identification. One of the technologies to be incorporated in the device is low-frequency vibro-acoustics, and a number of different vibro-acoustic methods for detecting buried infrastructure have been investigated. The latest developments in the vibro-acoustic location research are presented here. Three complementary methods are described, one of which involves direct excitation of the buried asset and the other two require no such direct access. All involve measurement of the ground surface vibration as a result of the excitation, whether of the ground or of the buried asset directly. Together, these techniques constitute a substantial step change in the way buried infrastructure can be detected using vibro-acoustic methods.
The brittle fracture mechanism of shale is an important issue to both borehole stability and hydraulic fracturing. A number of experimental studies on shale macroscopic mechanical properties have been carried out; however, the acoustic emission (AE) patterns of anisotropic shale are not fully understood. In this paper, a series of uniaxial compressive strength tests were conducted on cylindrical shale obtained by drilling at different orientations to the bedding plane. The anisotropic failure modes of shale samples have been monitored by real-time AE monitoring and stress–strain response, under ambient conditions and uniaxial stresses. The experimental results suggest that the pronounced bedding planes of shale have a significant influence on the mechanical properties and the AE responses. The AE counts and AE accumulative energy release curves clearly indicate the initiation and propagation of cracks during compression, and the ‘U’ shaped curves well reflect the anisotropic failure mechanism. The sounded foliation in the shale formation results in the anisotropy of elasticity, strength, AE counts and accumulated energy release from which the changes in strength dominate the entire failure pattern of shale samples. It is suggested that the AE activities monitored during deformation can provide good information to grasp the anisotropic failure mechanism of shale.
Fibre optic Distributed Acoustic Sensing (DAS) systems provide vibration response information comparable to accelerometers, geophones, and seismometers and may become widely used for infrastructure monitoring. DAS can be used to monitor earthquake activity, carbon sequestration, pipelines, and roadway/railway subgrade integrity, however little is known about the effect of soil type and burial method on DAS response. The objective of this paper is to present the results of a seven-month field study in which a DAS system was installed in different soil types (silty sand, clean sand, gravel, and a flowable fill) adjacent to an existing, decade-old DAS array. Impact tests were performed to evaluate DAS response in the different soil types and a portion of DAS array installed a decade prior. Signal-to-Noise Ratio (SNR) was used to compare performance of DAS response. Results of the monitoring program indicate that portions of the array in sand, gravel, and silty sand had good response with comparable SNR. A newer portion of array performed approximately five decibels better than the decade-old portion of DAS array, both in silty sand, with the old portion still performing well. These results may help build confidence with the geotechnical community regarding the longevity performance of DAS for infrastructure vibration monitoring.
This study aims to investigate the influence of joint thickness on seismic response across a filled fracture with strong nonlinear deformability. To simulate seismic attenuation of thicker joints subject to high-amplitude stress waves, the split Hopkinson pressure bar is utilised to generate normally incident P wave and the dry quartz sand is used to simulate the filled joints. Three joint thicknesses - that is 5, 10 and 15 mm, are studied under identical incident waves. The stress-strain response of the filling materials is described by Barton-Bandis model having different loading-unloading behaviours. The initial stiffness and the maximum allowable closure of the joints changing with the joint thickness are studied. The thicker joints result in lower initial stiffness and cause lower seismic wave transmission across the fracture. The high-amplitude stress strengthens the nonlinearity of the filling materials and increases the stiffness. Besides, the seismic attenuation factor Q, derived from the energy dissipation, is lower than that computed by the transmission coefficient due to the frequency filtering.
Having just relocated with my family from Sydney, Australia, to London, UK, I found it interesting to consider the analogy between travelling and this inaugural Géotechnique Letters themed issue. From our new house to the local grocery shop, I walk only 100 m; I travel by bus 1 km to the larger shopping centre and by train for furniture shops located nearly 10 km away. Of course, travelling more than 10 000 km from Sydney to London is quickest by aeroplane. The mode of transport I choose depends very much on the scale of the journey and, for longer journeys, I make use of many different modes of transport. By analogy, many large-scale geotechnical problems require us to embrace different research tools, each appropriate to a different length scale. However, soils are rich in features interacting across the scales: from asperity size to grain size, from the length of force chains to the thickness of shear bands, and from laboratory samples to the full engineering dimension. Clearly, we can gain much from a more accurate description of these features at the finer scales, but a more fundamental issue is to develop models appreciative of the cross-over between the scales – theories that can scale up as well as down.
Monitoring the curing process of mixtures produced by soil, fly ash and alkaline solution can be effectively performed by ultrasonic measurements of compression and shear wave velocities. These measurements detect the evolution of stiffness with time, as a result of alkali-activated ash reactions. Conventionally, ultrasonic testing relies on the use of a fixed input frequency, corresponding to the nominal frequency of the transducer, for which a maximum response is expected from the transducers. This procedure, however, has shown some limitations especially when applied to materials under curing, in which stiffness is changing with time. To overcome these limitations, a wide range of frequencies, from 24 to 200 kHz, was used at each of the selected curing periods for determining wave travel time, instead of a single frequency reading. The results show a clear increase in the optimum frequency with curing time, indicating that the ultrasound testing frequency should be adjusted to account for the stiffness increase due to curing. Additionally, the use of frequencies other than the nominal frequency of the ultrasonic transducers generated clearer signals, particularly at early curing periods.
Strip footing moment capacity when V/V 0 5 0?5: (a) adaptively refined mesh from FELA; (b) slip-line field  
Strip footing moment capacity when V/V 0 5 0?5: comparison of collapse mechanisms from (a) FELA and (b) slip-line solution  
Strip footing moment capacity when V/V 0 5 0?75: (a) adaptively refined mesh from FELA; (b) slip-line field  
Strip footing moment capacity when V/V 0 5 0?75: comparison of collapse mechanisms from (a) FELA and (b) slip-line solution  
Collapse of vertical cut: (a) adaptively refined mesh from FELA; (b) slip-line field  
The numerical method known as finite-element limit analysis (FELA) is generally employed as a tool for obtaining lower and upper bounds on the exact collapse load of a perfectly plastic structure or continuum. Most applications of FELA in geotechnical engineering have focused on plane strain problems involving the classical Tresca and Mohr-Coulomb yield criteria, and considerable computational effort has been expended on the calculation of lower- and upper-bound solutions for particular problems. This paper discusses and demonstrates an alternative use of FELA - as a tool for ascertaining slip-line fields for plane strain problems. A simple but effective strategy for adaptive mesh refinement is a key feature of the process; it allows the layout of plastic regions, rigid regions and velocity discontinuities to be determined by inspection of the FELA mesh. The corresponding slip-line field can then be constructed numerically in the usual way. The examples presented are restricted to purely cohesive soil, but the same approach is applicable in principle to frictional or cohesive-frictional materials.
This paper presents a novel adaptive watershed algorithm for the segmentation of soil particles in X-ray three-dimensional microfocus computer tomography images. An erosion-filling-dilation technique is also proposed, which is particularly suitable for processing images of highly porous particles. An s-factor is introduced to improve segmentation quality. In the inverse Euclidean distance map of the binary image, the topography in the zone around the local minima defined by this factor is modified. By using an appropriate value of s, the catchment basins and thus the watershed lines can be more realistically defined and it therefore resolves the problem of ill-segmentation. The proposed methods are applied to assemblages composed of highly porous particles with various size distributions, from uniformly graded to well graded or gap graded. The value of s can be decided based on visual examination of the segmentation quality on a sub-volume of the entire scanned domain. It is found that an s-factor that ranged from 0·5 to 0·7 is applicable to all of the studied cases, which enables a balance between under- and over-segmentation. Particulate-scale information including particle size distribution and particle shape characteristics is evaluated from the processed images and the results are compared with other available methods in the literature. Promising results are obtained.
This technical note presents a simplified method for quantifying the added mass of the soil that is mobilised as part of the failure mechanism around a foundation during rapid loading, and the resulting additional soil resistance. This note focuses on the solutions for an embedded plate anchor, which is a potential foundation system for offshore floating facilities. In current practice, only the shear strength of the soil surrounding the foundation is considered in calculations of the ultimate bearing capacity. However, the solutions presented in this technical note show that the added mass of the soil involved in the failure mechanism around the foundation can result in a significant increase in ultimate bearing capacity during extreme dynamic wave loading events. These lead to snatch loads transmitted to the anchoring and mooring system, which are high but brief. The technical note provides a general approach applicable to all foundation types, and illustrates the effect of the added mass term and the additional capacity for an embedded plate anchor with typical input conditions.
Analysis of pullout data
(a) Image of root-inspired, 3D-printed anchor models; (b) and (c) side and top view schematics of root-inspired anchor models, respectively. Labelling nomenclature consists of first the number of branches followed by the interior branching angle (e.g. n3a30 is equivalent to three branches with a branching angle of 30°)
Using additive manufacturing, computer-assisted models can be rapidly transformed into physical objects. When combined with X-ray imaging, minimally attenuating three-dimensional (3D) printed objects can allow for enhanced insight into soil processes, whether this is through the design of novel testing devices, prototypes or soil surrogates. In this study, this methodology is exploited to investigate the pullout mechanics of bio-inspired anchors. The purpose of this paper is twofold: (1) to demonstrate the value of the union of 3D printing and X-ray computed tomography (CT) technologies; and (2) to evaluate the effect of geometry on the pullout behaviour of plant root-inspired anchor elements. The results of the experimental investigations demonstrate how X-ray CT was used to effectively capture the evolution of soil behaviour during the uplift of three unique 3D printed prototypes. The analysis of the soil response was completed from both a global perspective, through an interpretation of the force-displacement data, and a local perspective, through an image correlation investigation of image sets at sequential displacement increments. The results provide insight into not only the means to design a more optimised anchor element, but also the efficacy of additive manufacturing-assisted X-ray tomography in investigating geotechnical problems.
Comparisons are made between the measured and computed axial and lateral responses of piles supporting a viaduct bridge in Singapore. The computed values were obtained from programs that use simplified boundary-element analyses for axial and lateral pile response, combined with input free-field ground movements obtained from approximate closed-form solutions. Parallel twin tunnels were constructed adjacent to the piles and measurements of axial force and bending moment in two of the piles that were instrumented have been reported. The agreement between measured and computed behaviour is generally good.
Soil water adsorption strength represents the changing rate of adsorptive water content with regard to soil suction or water potential changes, dictating the magnitude of the soil water-retention curve at the dry end. Non-isothermal conditions are frequently encountered in energy and environmental geotechnics, posing the requirement to assess temperature effects on soil water adsorption strength. Yet, it remains challenging in assessing the temperature effects at the high suction range due to limitations in available experimental techniques. Here, the grand canonical Monte Carlo (GCMC) simulation was explored as a method to address this challenge. A series of GCMC simulations has been performed to assess the temperature effect on the external surface adsorption of three representative soil minerals – that is, K-muscovite, Na-montmorillonite and α-quartz. Molecular simulation results preliminarily reveal that the water adsorption strength of the external surface of these soil minerals only demonstrates marginal dependence on temperature. This observation is consistent with experimental results, substantiating the feasibility of the proposed method.
Previous studies on tunnel interaction mainly focused on parameters such as cover depth, pillar depth and construction sequence with little attention given to the influence of tunnel shape. With discontinuous hoop stress distribution along the tunnel lining, non-circular tunnels are expected to respond to the construction of new tunnels nearby differently from circular tunnels. This paper investigates the effect of tunnel shape on multi-tunnel interaction by means of physical and numerical modelling. Two, three-dimensional (3D) centrifuge model tests were carried out simulating the existing circular and horseshoe-shaped tunnels. The test results were back analysed using PLAXIS 3D. An advanced hypoplastic constitutive model with small-strain stiffness was adopted for the numerical back analysis. Centrifuge test results show that the existing horseshoe-shaped tunnel experienced similar to 20% more vertical elongation in the mid-plane than the existing circular tunnel. The measured and computed bending strains at the invert of the horseshoe-shaped tunnel were three times larger than those at the invert of the existing circular tunnel in the transverse direction. However, hoop stress along the tunnel lining in the existing circular tunnel was larger than that in the horseshoe-shaped tunnel.
This letter presents preliminary results from a series of laboratory tests that demonstrate the potential for aerosols to deliver bulk liquids in dry or partially saturated sand and gravel. The tests were carried out using soil columns of 50 mm internal diameter and 400 mm length. The soil columns were tested dry or partially saturated with distilled water with maximum moisture contents of 4-5%. Methanol was chosen as the bulk liquid in the first series of tests. Liquid methanol was injected as an aerosol using medical jet nebulisers. The tests results demonstrated that droplet sizes of between approximately 0.5 and 5 mu m of methanol were transported through the soil column with the remaining aerosol retained either on the surface of the grains or dissolved in aqueous menisci between soil grains. A second series of tests using only distilled water as both pore liquid and aerosol revealed hydrophobic behaviour during the initial wetting of dry soil samples. This initial dry state was followed by an almost total retention of aerosols, which eventually reached a plateau of between 70 and 80% mass of aerosol retained. This behaviour suggests the droplets were primarily deposited at a finite number of interparticle menisci.
In recent years, agglomerates have been used to investigate the fracture of soil particles and rocks using discrete-element modelling (DEM). Hexagonal closed packing (HCP) and the radius expansion methods are the two most popular algorithms used to produce agglomerates. However, the HCP agglomerates created are highly anisotropic. Agglomerates created by the radius expansion method are too porous and are therefore not dense enough to represent rock. This paper proposes a new method to form dense random packing agglomerates. An agglomerate model that is isotropic and nearly stress-free (there is a small degree of particle overlap) is presented. To achieve agglomerates that mechanically behave like rock, the method produces smaller and smaller particles to fill smaller and smaller voids. The packing density achieved is governed by the ratio of the radius of the largest and smallest spheres. Numerical simulations of diametral crushing tests have been performed on the model agglomerates. Analysis reveals that as the density of the agglomerate increases, the isotropy of the particle improves and the variability of the strength of agglomerates formed from the same size distribution of spheres is reduced.
Geotechnical properties of the studied soil 
Water contents and suctions of samples 
Lime treatment is a common technique of improving the workability and geotechnical properties of soils. In this study, the aggregate size effects on the water retention capacity and microstructure of lime-treated soil were investigated. Two soil powders with different maximum aggregate sizes (D max = 0·4 and 5 mm) were prepared and stabilised by 2% lime (by weight of dry soil). Soil samples were prepared by compaction at dry side of optimum water content (w = 17%) with a dry density of 1·65 Mg/m 3. Suction and pore size distribution were determined after different curing periods. The results obtained show that: (a) the treated soil with smaller D max presents relatively smaller modal sizes and lower frequency of macropores (10–330 μm); (b) lime addition effectively improves the soil water retention capacity and decreases both the modal sizes of macro-and micropores gradually over time. Moreover, a higher air entry value and larger water retention capacity were also observed for a smaller D max value, in agreement with the pore size distributions.
A roadmap is presented to transition seamlessly from an image to a predictive computational model for granular materials. So far, constitutive modelling in granular materials has been based on macroscopic experimental observations. Here, the point of departure is the basic granular scale where kinematics, contact forces and fabric control the macroscopic mechanical behaviour of the material. New computational and analytical tools are presented that allow for more accurate measurement of kinematics and inference of contact forces, directly from imaging tools (e. g. high-energy tomography). These grain-scale data are then used to construct powerful multiscale models that can predict the emergent behaviour of granular materials, without resorting to phenomenology, but can rather directly unravel the micro-mechanical origin of macroscopic behaviour. The aim of these tools is to furnish a 'tomography-to-simulation' framework, where experimental techniques, imaging procedures, and computational models are seamlessly integrated. These integrated techniques will help define a new physics-based approach for modelling and characterisation of granular soils in the near future.
The rising of groundwater level induced by intense rainfall has significant influence on the slope stability. A new drainage method for landslide mitigation, namely the air-injection method, is proposed in this paper to control the quick rising of the groundwater under heavy rainfall to reduce the risk of landslides in a situation of emergency. A large-scale slope model is established in the laboratory to study the influence of the air-injection method on the groundwater level variations. The test results indicate that air injection divides the slope into three zones and forms a low permeability unsaturated zone to prevent the upstream groundwater flowing to the downstream slope. Two phases are identified during the air injection, (a) a formation phase of the unsaturated zone and (b) a steady-state phase of the unsaturated zone, and the unsaturated zone continues to intercept the upstream groundwater flow within a certain period of time after air injection is ceased. Furthermore, increasing the injection pressure appropriately can enhance the effectiveness of the air-injection method in intercepting groundwater.
Experimental work on remoulded natural and lime-stabilised samples of an alluvial silty soil has been conducted. Microstructure analyses were performed by mercury intrusion porosimetry (MIP) on stabilised soil compacted at different initial water contents and cured for different times. Some results are reported, and interpreted referring to different microstructure levels. The stabilised samples showed an aggregate fabric evolving with curing time as a function of the initial water content. From the constricted pore size distributions, investigated by means of cycles of intrusion, two main fabric changes were observed after lime addition - dimensional growth of the aggregates in the short term and the formation of stable bonding compounds in the smaller pore range for longer curing times.
A new, alternative approach in a triaxial set-up to fully determine small-strain cross-anisotropic elastic parameters is proposed, which requires neither radial strain sensors nor a radial loading process. This was made possible by additionally measuring pore water pressure responses during undrained axial loading, and inputting them into theoretical expressions derived based on cross-anisotropic elasticity theory. Although the method is intended for saturated soils, slight undersaturation in the system is accounted for by incorporating the pore water pressure coefficient B into the parameter deduction process. Being free from the cumbersome radial sensors and ultra-precise cell pressure controlling devices/operations, the proposed method is applicable even in less specialised laboratories. Four different deduction routes of elastic parameters were applied to results from tests on two reconstituted fine-grained soils. The horizontal Young's modulus obtained with the newly proposed method is consistent not only with those based on radial measurements, but also with those directly measured by axial loading on horizontally cut specimens. A challenge is still open for precisely determining Poisson's ratios, which exhibited only partial matching between different measurement methods. Overall, however, the proposed method is consistent with those adopting full local instrumentation.
Biochar is a solid carbon material produced by heating organic biomass in the absence of oxygen through a process known as pyrolysis. Scientists and policy makers have recognised its potential role in carbon sequestration, reducing greenhouse gas emissions, renewable energy, waste mitigation and as soil amendment. To date, most studies related to biochar application to soil have focused mainly on the agronomic benefits, while little attention is paid on its effects on soil properties from geotechnical point of view. Considering that biochar can be produced from biomass, its use would fit satisfactorily the current trends of recycling wastes for ground improvement and liquefaction mitigation. To address this, the effect of adding various proportions by weight of biochar, derived from poultry litter, on the liquefaction resistance of sand is investigated through undrained cyclic triaxial tests. The findings suggest that, contrary to other earlier published reports using another type of biochar, adding poultry litter biochar to sand decreases its liquefaction resistance in a similar manner as adding non-plastic fines. The results highlight the fact that the type of biochar used as soil amendment is important in determining whether the liquefaction resistance of the host sand will improve or not.
This note presents a new pressure-controlled elliptical cavity expansion model for elastic-perfectly plastic soil. The model accounts for the non-asymmetric characteristics of elliptical cavity expansion, which differ from conventional asymmetric cylindrical cavity expansion. A closed-form solution for the pressure-expansion relationship is used to predict the uplift capacity of strip anchors in undrained clay. Theoretical variations of the strip anchor breakout factor are compared with conventional cylindrical cavity expansion theory and laboratory test results and upper and lower bound finite-element limit analysis solutions from the literature. The results show that a pressure-controlled elliptical cavity expansion model can predict the uplift capacity of strip anchors in undrained clay more accurately than the conventional cavity expansion method at larger embedment depths (H/B > 3). The proposed solution improves on the conventional cavity expansion theory and provides a new method for solving the non-asymmetric cavity expansion problem.
Equations that can be used to predict the trajectory of a plate anchor under pull-out load are derived from a simple combined loading yield surface assuming associated flow. They are used to directly predict the behaviour of plate anchors with different padeye offsets and under loads of varying inclinations. It is demonstrated that there is an optimal padeye offset that depends on the load inclination and causes the anchor to increase its embedment, and hence its capacity, under pull-out. The equations represent an elegant and efficient method to assist in the design of plate anchors, notably in optimising the padeye location.
This paper reports the results from field tests on a 1/15 th scale recently developed fish anchor. The tests were conducted at three locations in the Swan River, Perth. Two series of tests were performed from the Burswood and Maylands jetties with water depths between 1.1 and 1.9 m. The final series of tests were undertaken in deeper waters of 2.6 m from a barge. The riverbed at the Burswood Jetty and barge test location consisted of soft clay, and that at the Maylands Jetty comprised sandy silt. The tip embedment depths of the scaled fish anchor, with dry weight of 0.304 kN and impact velocity of 5.89∼9.55 m/s, in soft clay were 1.17∼2.40 times the anchor length. For similar impact velocities, the tip embedment depths in sandy silt were 30 ∼ 60% shallower than those in soft clay. By comparing the field test data in clay, the fish anchor achieved normalised embedment depths similar to those of the torpedo and OMNI-Max anchors under half or less impact velocity. Most importantly, the field tests confirmed the diving behaviour of the fish anchor under loading with mudline inclination of 20° and 25°, with the second peak dictated the capacity. The ultimate capacity was 5∼7 times the anchor submerged weight in water.
Simulated breakout factor N γ for strip anchors of differing width and embedment ratios. The error bars denote the standard deviation of the measured breakout factor considering ten tests (see text) 
Microstructure evolution during the anchor uplift for an anchor with an embedment ratio H/B = 1 and a width ratio B/d = 30. (a) Contact network before anchor motion (see text). (b) Total force F, force on the top face F top and force on the bottom face F bottom of the anchor during uplift, as a function of the anchor vertical displacement δy. (c) Contact network at failure (see text). (d) Displacement field of individual grains; each line connects the initial position (blue) of a grain to its position when failure occurs (red). In (c) and (d), the black lines represent the frustum boundary, with an inclination of 15·5° from vertical. Note that (a), (c), (d) only represent a small part of the simulated test chamber 
Microstructure evolution during the anchor uplift for an anchor with an embedment ratio H/B = 1 and a width ratio B/d = 10 (see Fig. 6 caption) 
This letter investigates the uplift capacity of plate anchors in granular soils. Simulations based on a discrete-element method are used to measure the uplift capacity of anchors of differing widths to embedment B/H and width to grain-size B/d ratios. Results confirm that the uplift capacity of anchors with a large B/d ratio is well described by existing models developed from continuum mechanics, with no grain-size effect. In contrast, results reveal a strong deviation from these models for anchors with relatively small B/d ratios. A semi-empirical model is introduced that captures this strong grain-size effect. This model is further supported by a micro-mechanical analysis, indicating that anchor uplift capacities are not only governed by a frustum mechanism predicted by continuum mechanics but also involve the mobilisation of grains surrounding this frustum. These results and model are particularly important to rationalise uplift capacities measured in small-scale experiments, typically involving small ...
Model test database of horizontal strip anchor uplift resistance
Comparison with other solutions: (a) with given effective friction angles; (b) with peak friction and dilation angles calculated by Bolton (1986) correlations
This letter presents an analytical cavity expansion theory-based method for predicting peak uplift resistance of shallow horizontal strip anchors buried in sand. Based on an analytical two-dimensional stress solution for loading analysis around a cylindrical cavity, the method was developed by assuming that the peak anchor uplift resistance can be approximated by the cavity breakout pressure. In the new cavity expansion model, the ultimate failure is reached once the plastic zone develops to the ground surface, and the biaxial state of in-situ ground stresses is taken into account. A database consisting of 75 model tests on shallow strip anchors in sands was compiled to valid the new method. The predicted results and measured data are in reasonable agreement, with a mean over-prediction of the peak uplift resistance by 1.6%. The reliability of the new solution was also checked by comparing with other commonly used analytical solutions. It is shown that the present solution can provide a simple analytical tool for predictions of the peak uplift resistance of strip anchors in sand while a sliding-block failure mechanism dominates.
This paper reports the results from dynamic finite-element analyses undertaken to provide insight into the behaviour of torpedo anchors during pullout in non-homogeneous clay. The results were validated against field data prior to undertaking a parametric study, exploring the relevant range of parameters in terms of installation effect, impact velocity, anchor geometry, padeye position, pullout angle. Installation method (e.g. quasi-static and dynamic installation) has shown significant influence on the vertical pullout capacity, but that influence reduced to marginal for lateral pullout. The inclined pullout capacity was dictated by the anchor weight and fully mobilised anchor-soil contact surface area and consequently the capacity increased with increasing the number of fins. Due to the lower weight and total surface area, the 45° (at the padeye) pullout capacity of the finless anchor was 29% of that of the considered four-fin anchor. The results showed that this percentage can be increased to 83% by shifting the padeye location of the finless anchor to 0·56 times the anchor shaft length from the top. Anchor capacity under inclined pullout loading was presented as failure envelopes expressed in terms of dimensionless vertical and horizontal components, which lied between the envelops developed for piles and embedded foundations.
The pull-out capacity of anchors inspired by tree roots was evaluated using model tests to investigate the effect of anchor geometry on pull-out capacity. Experimental pull-out tests were performed on anchors with fractal geometries buried at shallow depths in a model granular medium. The results show that an existing pull-out model for plain shape anchors can be extended to fractal shape anchors by introducing an additional shape factor. The values of this shape factor reveal that fractal anchors can achieve the same pull-out capacity as plain shape anchors but the amount of material needed for their construction is much less. As a consequence, the use of fractal shape anchors may be a cost-effective solution in a number of geotechnical applications.
Properties of silica sand 
Centrifuge test programme and results (prototype scale) 
Framework representation to describe the behaviour of a plate anchor-soil system: (a) and (b) monotonic loading; (c) and (d) cyclic loading 
This paper provides experimental evidence that shows that the drained cyclic capacity of a plate anchor in dry dense sand may be higher than the equivalent monotonic capacity. The experimental data show that when cyclic loading is low relative to the monotonic capacity, increases in the eventual capacity are observed; when the magnitudes of the cyclic loads are closer to the monotonic capacity, no increases in capacity are observed. These responses are explained in the paper using an elasto-plastic macro-element model extended with expandable bounding and memory surfaces that address the increase in strength or stiffness caused by changes in soil density and fabric when the anchor is subject to cyclic loading in dense sand.
The rigour of extracting friction angles, and eventually lower-bound bearing strength, in sandy beach settings through slope angles determined from digital images (visual spectrum) is explored. Digital images of topographic sand features using hand-held cameras, an unmanned aerial vehicle and a panchromatic satellite sensor are analysed to determine average slope angles using three-dimensional reconstruction. Greyscale gradients and shadows are utilised in the satellite images to extract slope estimates. The slope angles matched tilt table results of samples from the same locations at the Duck, NC, and Claytor Lake, VA, field sites. Direct shear testing of sample material suggest friction angles of 33° and 35°, respectively. The authors test a potential pathway to derive lower-bound bearing strength using these remotely sensed slope angles. Preliminary results are encouraging, but likely sensitive to the impact of moisture content, differences between the maximum and the observed slope angle and internal friction angles.
Anisotropic behaviour of a saturated clayey loess has been investigated by means of triaxials and oedometers. It was found that the vertically cored samples reached well-defined gross yield points outside the horizontally cored samples in oedometer and isotropic compression. The undrained and drained shear stress of vertically cored samples increases more quickly than that of horizontally cored samples though they finally reach the same values. In the volumetric plane, the critical state lines (CSLs) for vertically and horizontally cored samples were obviously distinguished. For the horizontally cored samples, the CSL is with horizontal asymptotes at low stress levels and is distinctly lower than that of vertically cored samples. At high stress levels they converge due to the elimination of anisotropy. The anisotropic effects identified for soils retrieved from different depths are similar but slowly decrease as the depth increases. Through normalising the stress paths with a reference of intrinsic behaviour of the clayey loess, vertically and horizontally cored samples show a significant difference.
Constant-K stress paths applied to the reconstituted specimens  
Relationship between the volumetric strain ε v and the mean effective stress p′ during anisotropic consolidation with different initial water contents  
Initial yield curves of two series tests during anisotropic consolidation with different initial water contents  
Triaxial anisotropic consolidation tests were performed on Wenzhou clay reconstituted at two initial water contents to investigate the effect of deviatoric stress on volume change during compression. The laboratory test results show that volumetric strain of reconstituted clays under different stress paths is significantly affected by the effective principal stress ratio defined as the ratio of minor effective principal stress to major effective principal stress. The volumetric change during anisotropic consolidation can be considered as functions of the mean effective stress and the deviatoric stress. The initial water content is also found to induce the difference in the volumetric strain of reconstituted clays.
Some previously published triaxial strength data of Delabole slate and Angers schist are re-analysed. The strength data are grouped according to the orientation of the anisotropy plane relative to the applied major principal stress, and Bayesian regression is used to model peak strength as a linear function of confining pressure for each group. The analysis indicates that not only the magnitude but also the variance of strength and the effect of confining pressure on strength are orientation dependent. These dependencies have profound ramifications for the implementation of a geotechnical limit state design framework.
A cross-anisotropic formulation for elasto-plastic constitutive models based on a non-uniform scaling of the stress tensor is described. Taking advantage of the material symmetries characterising cross-anisotropy, only two scaling factors, one for the normal stress components and one for the shear stress components, are required. It is shown that the formulation can be easily introduced in already implemented models with minor modifications. The performance of this formulation is investigated by reproducing the strength variation of anisotropic rocks in triaxial tests. The numerical simulation of an unsupported excavation is also presented to show the effect of different scaling factors and bedding plane orientations.
Anisotropic behaviour and three-dimensional (3D) strength are two significant properties of soils. This paper incorporates the spatially mobilised plane (SMP) criterion into the anisotropic critical state model and a semi-analytical solution for drained cylindrical cavity expansion problem is derived based on the modified model. The problem is formulated by transforming equilibrium equation, constitutive equations and strain displacement relation into a set of seven first-order ordinary differential equations concerning three transformed stresses, anisotropic parameters and specific volume. The present solution is expected to provide more rational predictions than other solutions as both the 3D strength and rotational hardening are considered in the proposed solution. Case analyses regarding distributions of seven basic variables, the effective stress and anisotropic parameter path around the cavity are conducted and compared with other solutions. The results show that neglect of 3D strength and rotation hardening would result in unreasonable predictions for geotechnical problems.
This contribution proposes a numerical microstructural modeling approach to investigate stress-induced seismic velocity changes on anisotropic rocks. By introducing pre-existing cracks with preferential orientations in bonded-particle assemblies, the transverse isotropic structure of the Whitby Mudstone is simulated. Using power-law distributed aperture and calibrated micro-properties, we successfully reproduce stress-dependent velocity changes on Whitby Mudstones with different anisotropic angles in relation to the applied loads. The proposed model also duplicates the directional dependence of wave speed with respect to the bedding plane as expected theoretically. The numerical models show that velocity increase results from the closure of pre-existing cracks due to load increase. Direct relations are established between velocity changes and opened crack density (or crack closure), which displays a similar tendency compared with theoretical predictions. This relation can be used to quantify the micromechanisms behind the velocity changes. The proposed model provides the ability to directly examine the micro-processes underlying velocity changes.
Discrete element modelling (DEM) of a face-centred cubic assembly of spherical particles has been used to study the influence of anisotropic stress states on the shear wave velocity of a granular material. The shear waves were generated and detected in a way equivalent to the use of bender elements in laboratory testing. Comparisons are presented between the discrete element simulations and analytical and empirically derived methods of relating stiffness to the degree of confining stress anisotropy. The results confirm previous empirical observations that wave velocity is strongly influenced by the stresses in the direction of propagation and in the direction of oscillation of the shear wave. The wave velocity is, however, largely independent of the stress orthogonal to the plane containing the wave motion.
Top-cited authors
C.W.W. Ng
  • The Hong Kong University of Science and Technology
Gioacchino Viggiani
  • Université Grenoble Alpes
Edward Andò
  • École Polytechnique Fédérale de Lausanne
Lyesse Laloui
  • École Polytechnique Fédérale de Lausanne
Stephen A. Hall
  • Lund University