Search

Physical building vulnerability to debris flows is defined as the potential damage degree of buildings for a given debris flow intensity. In this paper, the physical characteristics of both debris flow intensity and building response are considered. Uncertainties in building capacity and debris flow intensity are explicitly quantified to evaluate the damage probability of a typical reinforced concrete building subjected to debris flow impact. Four damage states with clear failure mechanisms are defined using multi-source information from field observations, numerical simulation, and expert experience. Two series of fragility models are proposed based on practical debris flow impact pressure models. Several debris flow intensity measures are investigated. A better indication can be provided using the intensity measure that represents a specific failure mechanism; e.g., impact force (hv2) for force-dominated failures or overturning moment (h2v2) for moment-dominated failures, where h and v are debris flow depth and velocity, respectively. The corresponding fragility surfaces best express potential building damage. The intensity thresholds in the proposed fragility curves are consistent with those in empirical vulnerability curves. The methodology presented in this paper promotes vulnerability assessment using physics-based modeling, leading to a more reliable evaluation of building damage caused by debris flows.

Earthquakes frequently induce landslides and other natural disasters that would have a huge impact on human life and properties. In geotechnical engineering, evaluation of the seismic stability of earth slopes has been attracting a substantial amount of research interest. In this regard, the Newmark permanent displacement provides a simple yet effective index of slope co-seismic performance. The traditional Newmark method involves many assumptions and the displacement results thereby calculated are subject to various degrees of uncertainty. In this paper, a modified rotational sliding block model considering depth-dependent shear strength and dynamic yield acceleration is established. The seismic critical slip surface is analysed through a pseudo-static approach, where the failure volume is larger than that in the static condition. The dynamic yield acceleration is updated by considering the instantaneous movement of the sliding mass in each time-step. The parametric sensitivity of soil shear strength, slope geometry, and Arias intensity to the seismic displacement is also analysed. Results show that the internal friction angle and the cohesion have equal effects on the permanent displacement. On a logarithmic scale, the displacement approximately linearly correlates with Arias intensity. Furthermore, the underlying uncertainty of the ground motion is introduced to obtain the probabilistic distribution of the seismic slope displacement. The uncertainty of earthquake time history details has considerable influence on the permanent displacement results. Under the specific allowable displacement, the probability of failure increases exponentially with seismic intensity.

Several piezocone penetration test (CPTu)-based soil behaviour classification systems (SBCs) have been developed for standard sites, where clays, silt, and sand dominate. However, problems can occur when applying the SBCs to offshore sites, where the marine soils may be decomposed from rocks or mixed with artificial fills. This study evaluates the accuracy of six CPTu-based SBCs for marine soils at a site offshore Hong Kong based on 16 CPTu soundings with 25 367 data points by comparing them with composition-based SBCs from borehole records in the vicinity of each sounding. The soil types are determined from six common CPTu-based SBCs. The interpretation of CPTu data is first performed to generate soil type variables comparable to borehole data, followed by a cross-validation study. The soil classification performance of each SBC is quantified by the weighted kappa coefficient and the Kendall correlation coefficient between the soil types generated by the CPTu-based and composition-based SBCs. The classification accuracy for each soil type is also evaluated via the root mean squared error and the mean absolute error. The classified soil types from the CPTu data are associated with a median degree of consistency, indicating the need for combining CPTu-based and composition-based SBCs for marine soil classification.

The effects of the first four statistical moments defining the statistical characteristic of elastic modulus on the probabilistic foundation settlement are investigated in this study. By combining the Hermite probability model and spectral representation method, a method to simulate nonGaussian homogenous fields based on the first four statistical moments is proposed. Linear elastic finite element models are employed to study the total settlement and the differential settlement of a shallow foundation. Probabilistic measurements of total–differential settlement obtained by the Monte Carlo simulations are presented. For the cases considered, the effects of skewness and kurtosis defining the probabilistic characteristic of elastic modulus on the total–differential settlement of a probabilistic foundation are illustrated. The computed results show that the value of skewness has a more significant effect on the probabilistic foundation settlement than kurtosis, and the case with the smallest skewness is observed as the most critical one.

Natural soils often exhibit an anisotropic fabric pattern as a result of soil deposition, weathering or filling. This study aims to investigate the effects of spatially variable anisotropic soil fabric in a slope on its safety factor and failure mechanisms, and to identify the critical fabric orientation that is most unfavorable to the slope stability. The strength properties of colluvium (i.e., cohesion and friction angle) are modeled as random fields under two conditions (i.e., independent and negatively correlated). The study reveals that there exists a critical fabric orientation at 30° at which the mean factor of safety is the lowest and the probability of failure is the highest. The negative cross-correlation between soil shear strength properties leads to a significantly lower probability of failure, compared to the independent case. The highest proportion of deep failure mechanism is also identified at the same fabric orientation of 30°. The identified critical fabric orientation is gentler than the slope inclination. This study suggests that the conventional understanding that stratification parallel to the slope surface appears to be the most unfavorable condition leading to the lowest safety level does not hold for spatially varying soils.

This paper presents a coupled air–water flow analysis to evaluate the performance of a three-layer capillary barrier for controlling water infiltration into and gas emission from a waste containment system in a high precipitation environment, and to optimize combinations of local soil layers for barrier construction. A multi-phase flow model is proposed considering the movements of the gas and water phases simultaneously. The governing partial differential equations are solved in COMSOL Multiphysics software. Several combinations of lean clay with sand (CL), clayey sand with gravel (SC), silty sand with gravel (SM), sandy silt (ML), and well-graded gravel with silt (GW–GM) are examined. The rates of percolation water and gas emission are used as indicators to compare the performance of different combinations. A fine-grained surface soil layer reduces both water infiltration and gas emission due to its low desaturation rate and high water-retention capacity. The coarse middle layer plays a critical role, promoting capillary effects and hindering water infiltration during rainfall as well as draining any infiltrated water or percolated gas.

The effects of coarse content on the microstructure, soil-water characteristics, shearing behavior, and critical state characteristics of three types of unsaturated coarse granular soil with different coarse contents are investigated in an integrated manner through an extensive laboratory test program. The tests were performed on samples with the same void ratio of 0.62. As the coarse content increases, the pore-size distribution of the soil changes from unimodal to bimodal. Accordingly, the soil-water characteristic curve (SWCC) changes from unimodal to bimodal. The whole drainage process of the bimodal SWCC is divided into four stages based on the role of the pore water. The contribution of suction to the peak state parameters and critical state parameters, except for the intercept of the critical state line in the stress plane (μ(s)) of a bimodal soil, reaches its maximum in the first drainage region. The pore water in the two pore series of a bimodal soil contributes to the shear strength differently during the desaturation process. Hence, the variations of the maximum dilatancy and the shear strength parameters with degree of saturation are highly nonlinear and μ(s) shows bimodal features. The contributions of suction to the shear strength parameters decrease with increasing coarse content.

The 12 May 2008 Wenchuan earthquake in China triggered numerous landslides. Loose landslide materials can easily evolve into deadly debris flows during wet seasons. During the period from 2008 to 2011, three separate large-scale debris flows occurred in the Pubugou Ravine near the epicentre of the earthquake that were among the largest repeated debris flows ever reported. Approximately 1.76 × 106 m3 of sediment was deposited during these three events. This paper aims to (i) analyze the movements of solid materials during the repeated debris flows, (ii) discuss the evolution of the initiation mechanisms of these debris flows, and (iii) evaluate the changing depositional morphology of the debris flow fans and examine the particle sizes of the debris flow materials. To achieve the above research objectives, timely field investigations were undertaken in the past 5 years after each of these debris flow events. Satellite images were used to delineate the boundaries of the initiation areas, the transportation channels, and the deposition zones of these debris flows. With the occurrence of the repeated debris flows, the hill slope deposits gradually evolved into channel deposits and the solid materials in the channels moved toward the gully mouth. Hence, channelized flows gradually became dominant. The debris fan materials of the repeated debris flows became coarser and coarser over time. The three debris flows were all characterized by coarse boulder fronts.

Dam safety has drawn increasing attention from the public. To ensure dam safety, it is essential to diagnose any dam distresses and their causes properly. The main objective of this paper is to develop a robust probability-based tool using Bayesian networks for the diagnosis of embankment dam distresses at the global level based on past dam distress data. A database of 993 distressed in-service embankment dams in China has been compiled, including general information on the dams, distresses, and causes. Based on the database, general characteristics of embankment dam distresses are studied using Bayesian networks, which can tackle not only the multiplicity of dam distresses and causes, but also the complex interrelations among them. Common patterns and causes of distresses are identified. The interrelations among the dam distresses and their causes are quantified using conditional probabilities determined based on the historical frequencies from the dam distress database. A sensitivity analysis is also conducted to identify and rank the most important factors that cause the distresses. With the prior information of common characteristics extracted from the database, Bayesian networks are further used to diagnose a specific distressed dam at the local level by combining global-level performance records and project-specific evidence in a systematic structure, which is presented in a companion paper.

Based on prior information on common characteristics of dam distresses extracted from the dam distress database described in a companion paper, this paper attempts to extend the technique of Bayesian networks to the diagnosis of a specific distressed dam. The diagnosis is conducted by combining two sources of information, i.e., global-level knowledge from the database and project-specific evidence. Based on results of the diagnosis, key distress factors for a specific dam can be identified and suitable remedial measures can be suggested. Further, the Bayesian network analysis is conducted to evaluate the effectiveness of the adopted remedial measures. A case study on the diagnosis of a distressed embankment dam, Chenbihe Dam, with seepage problems is presented to illustrate the methodology. In this case study, the observed leakage rates, seepage exit locations, and boundary conditions of the embankment are used as project-specific evidence.

Unsaturated hydraulic conductivity is the primary soil parameter required when performing seepage analyses for unsaturated–saturated soil systems. Unsaturated hydraulic conductivity is also one of the most difficult parameters to measure because of the time involved and the limited suction measurement range (e.g., 0∼1500 kPa in a test using the steady-state method). In this study, a new wetting front advancing method was developed for measuring unsaturated hydraulic conductivity. The wetting front advancing method simulates and monitors a soil wetting process through a large-scale soil column. A new interpretative procedure was developed to calculate the unsaturated hydraulic conductivity based on the monitored water content, suction, and wetting front advancing velocity. The proposed technique is used to measure the unsaturated hydraulic conductivities of five soils, which vary from gravel to clay. The results indicate that the proposed technique is time-saving (i.e., requires several days for a complete test) and is applicable over wide ranges of suctions and unsaturated hydraulic conductivities. The measured unsaturated hydraulic conductivity using the wetting front advancing method is similar to that obtained using the instantaneous profile method, with the latter covering narrower ranges of soil suction and hydraulic conductivity.

Cracks are prevalent in near-ground-surface soils and provide preferential pathways for fluid flow. Cracks increase water infiltration or contaminant solute transport into soils. This paper studies the development of a permeability tensor and a representative elementary volume (REV) for saturated cracked soils. First, a method of generating random crack networks based on the statistical parameters of the crack geometry is presented. Then, the permeability tensor and REV for a crack network are studied by modeling water flow through the generated random crack networks. Finally, the permeability tensor for a cracked soil is obtained by combining the crack network and the soil matrix, and the properties of the permeability tensor and the REV for the cracked soils are investigated. Results show that the permeability tensor can be used to represent the permeability anisotropy of a cracked soil. An REV can be readily established when the crack network is relatively dense. An REV can be defined and the contribution of the crack network to permeability is small when a crack network exists in homogeneous sands. An REV is harder to establish and the hydraulic conductivity is dominated by the random crack network when a sparse crack network exists in clays.

The microporosity structure of soil provides important information in understanding the shear strength, compressibility, water-retention ability, and hydraulic conductivity of soils. It is a soil characteristic that depends on sample preparation method and wetting–drying history. A comprehensive study of the microporosity structure of a lean clay with sand was conducted in this research to investigate variations of the microporosity structure during compaction, saturation, and drying processes. Scanning electron microscopy was used to observe the microporosity structure of soil sample surfaces. Mercury intrusion porosimetry was used to measure the microporosity structure quantitatively by showing the relationship between cumulative pore volumes and pore radius. The experimental results show that a dual-porosity structure (i.e., interaggregate pores and intra-aggregate pores) forms during the compaction process. The interaggregate pores are compressible and the associated volume is closely related to the final void ratio of the compacted sample. Changes to interaggregate pores is dominant during compaction, but changes to intra-aggregate pores is dominant during saturation and drying. Based on the experimental results, a dual-porosity structure model was developed by relating the pore-size distribution to the void ratio. Consequently, the pore-size distribution at any void ratio can be predicted.

The soil specimen in a centrifuge model is subject to spatial variability depending on the method of sample preparation and the stress field induced by the centrifugal acceleration, even though it is intended to be uniformly prepared. In contrast to extensive measurements for studying the variability of in situ soil properties, soil variability in centrifuge models, especially that which is based on data at very close sampling distances, is less understood. In this paper, the variability of soil density in two centrifuge models is presented. Random field theory is adopted to characterize the spatial soil variability in the two centrifuge models. The importance of taking spatial variability parameters as a model similarity requirement in centrifuge model design is illustrated and discussed. It is demonstrated that, although centrifuge models of different sizes can be designed to simulate the same prototype, the prototypes these models actually represent are not identical in terms of soil spatial variability. To achieve similarity in spatial variability between a centrifuge model and its prototype, one may need to control either the point coefficient of variation or the scale of fluctuation of the model soil so that the coefficients of variation of the spatially averaged soil property in the model and the prototype are the same.

Piles in a pile group subjected to torsion simultaneously mobilize lateral and torsional resistances. Hence, complicated pile–soil–pile interaction effects and load deformation coupling effects occur in the pile group. In this study, a series of centrifuge model tests were carried out to investigate these effects in three-diameter spaced 1 × 2, 2 × 2, and 3 × 3 pile groups subjected to torsion in both loose and dense sands. The test results showed that the effect of horizontal movement of a pile on lateral behaviors of its adjacent piles is significant in 3 × 3 pile groups and such effect varies with group configuration and pile position. The p-multiplier concept can be used to quantify the effect and values for the p-multiplier are suggested. The effect of lateral movement of a pile on the torsional resistances of its adjacent piles and the effect of torsional movement of a pile on the lateral resistances of its adjacent piles were found to be minor in these tests. For an individual pile in a pile group subjected to torsion, the mobilized lateral resistance was found to substantially increase the torsional resistance of the pile. Such a coupling effect is quantified by a coupling coefficient, β, which describes the contribution of subgrade reaction to the increase of torsional shear resistance.

Due to the presence of uncertainties, errors inevitably arise with the estimations of pile settlement. To properly consider serviceability requirements in limit state design, it is necessary to characterize the performance of commonly used settlement prediction models. In this work, information from 64 cases of long driven steel H-piles from field static loading tests in Hong Kong is utilized to evaluate the errors of three settlement prediction models for single piles: two elastic methods and a nonlinear load–transfer method. Commonly adopted soil parameters recommended in two Hong Kong design guidelines are used to reflect the uncertainty arising from evaluation of soil properties. The model error is represented by a bias factor. A conventional statistical analysis was first conducted to study the variability of model bias. A regression analysis method was then proposed as a supplemental analysis of model bias when only limited test data were available or when the measured settlement data distribute in a large range. Both methods result in very similar mean biases. The mean bias of each prediction model tends to vary with the load level and the bearing stratum at the pile toe; while the coefficient of variation of model bias only varies in narrow ranges.

The Gouhou Dam was a concrete-faced rockfill dam built in a steep canyon that collapsed in 1993 due to internal erosion during the initial reservoir filling. In this paper, the process of water infiltration into the originally unsaturated rockfill dam is studied using three-dimensional saturated–unsaturated seepage theory. The three-dimensional characteristics of seepage through the dam bounded by steep abutments, the effect of material anisotropy, and the effect of rockfill stratifications are studied. The three-dimensional results are compared with those from two-dimensional analyses. The three-dimensional simulations show that seepage water flows faster and the hydraulic gradients are greater near the abutment boundary in the dam. As such, the evolution of the seepage failure in the three-dimensional cases is faster than that in the two-dimensional analyses, and the two-dimensional analyses will underestimate the risk of seepage failure, particularly near the abutment boundary. If the materials in the dam were uniform, the reservoir water would infiltrate into the dam along a downward flow path towards the riverbed, and not exit from the surface on the downstream slope. Increasing the horizontal coefficient of permeability of the rockfill increases the infiltration velocity, but the material anisotropy does not appreciably change the infiltration pattern. Stratifications in the rockfill, however, cause the seepage water to advance more quickly in the horizontal direction along the interface between the sandwich layer and the rockfill, thus increasing the possibility of seepage failure.Key words: seepage, seepage failure, rockfill dam, unsaturated soils, wetting front, numerical analysis.

The failure of a pile is always defined by a certain failure criterion. Several different failure criteria are commonly used, and the pile capacity values associated with each of these failure criteria can be considerably different. For the sake of international harmonization, it is necessary to calibrate the reliability levels associated with various failure criteria and factors for loads and resistances. This paper aims to evaluate the effects of failure criteria and factors for loads and resistances on the reliability of single bored piles. The bias arising from failure criteria is described by a bias factor, which can easily be accommodated in a reliability analysis. A comprehensive database of static load tests of bored piles is utilized to evaluate the bias associated with several failure criteria. Five limit-state design codes for piles are investigated to illustrate the effect of the bias from failure criteria, the effect of factors for loads and resistances, and their combined effect. The results indicate that the bias from failure criteria has a significant influence on the reliability of piles. Similarly, the use of different factors for loads and resistances in various design codes can also cause considerable differences in the calculated reliability. As a result of these effects, the actual reliability levels of any two design codes, assuming the same nominal target reliability index, can differ considerably.Key words: bored piles, pile capacity, failure criterion, reliability analysis, load factors, resistance factors.

The effects of vertical load on the lateral resistance of single piles were initially reviewed to facilitate the interpretation of the test results of pile groups. Then, 18 different lateral load tests were carried out in the centrifuge on the 3 × 3 and the 4 × 4 fixed-head battered pile groups to investigate the effects of vertical load on the group lateral resistance. Vertical dead loads ranging from approximately 20 to 80% of the vertical ultimate group capacity Puv were applied. Based on these tests, the effects of vertical dead load on the lateral resistance of the battered pile groups are found to depend on pile arrangement, pile inclination, and soil density. The lateral resistances of the 3 × 3 pile groups do not appear to vary considerably with the vertical dead loads in the range of the vertical loads studied. For the 4 × 4 pile groups however, the lateral resistances at vertical loads of approximately 50 and 80% Puv may be 26-29% and even 40% higher than that at the 20% Puv dead load. It may be inferred that designs based on standard lateral load tests with small vertical dead loads would be on the safe side. Three mechanisms for vertical load effects are discussed in terms of axial tension and compression failures, influence of pile inclination, and initial subgrade reaction caused by vertical loading. Preliminary numerical analyses are also performed to simulate the responses of some of the battered pile groups.Key words: pile group, battered pile, lateral resistance, load test, pile-soil interaction, centrifuge test.

Many high-rise buildings, bridges, and transmission towers are constructed on steep slopes in Hong Kong and are supported by large-diameter piles. These structures may be subjected to large lateral loads, such as those caused by typhoons, earthquakes, and high-speed vehicles. The margin of safety of the slope may decrease as a result of stresses transferred from the piles to the slope. To minimize the transfer of lateral load from the buildings to the shallow depths of the slope, an annulus of compressible material (sleeving) is sometimes formed between the piles and the adjacent soils. In this paper, a three-dimensional analysis is carried out to investigate the effects of unsleeved and sleeved single piles and pile groups on the stability of a cut slope. Mechanisms of load transfer from the piles to the slope are studied. The stability of the slope is evaluated using the strength reduction technique. The evolution of slope failure is examined and the factors of safety for both initiation of instability and global failure of the slope are identified from the numerical analyses. The sleeving technique is found to be capable of significantly reducing the stresses in the shallow depths of the slope in front of the piles, thus improving the local stability of the slope, but offers limited benefit with respect to global stability.Key words: laterally loaded pile and pile group, sleeving, slope stability, three-dimensional analysis, load transfer mechanism, factor of safety.

This paper presents some possible interpretations of the physical meaning of the lumped, toe, and skin Case damping factors, jcL, jct, and jcs, respectively, which are extensively utilized in the dynamic analysis of pile driveability and capacity. A single degree of freedom model is employed to relate the Case damping to the hysteretic damping ratios of soil and pile materials. This relation and the damping ratios of soils and piles show that the Case damping factors for piles in sandy and clayey soils may overlap at all strain magnitudes. Coupling of pile toe and skin resistance is analyzed, and the jcLfactor is found to be a function of the skin and toe resistance ratio of the pile. Consequently, the jcLfactor is an important indicator with which the skin friction and toe resistance of piles can be separated. A database of 133 cases of dynamic pile tests in Florida has been used to substantiate the analyses and interpretations. The effects of the assumptions made in this paper are also discussed.Key words: pile dynamics, wave equation, Case damping, hysteretic damping, bearing capacity, skin friction.

Extensive centrifuge model tests were carried out to investigate optimum compaction density zones within loess embankments 30 and 63.8 m high. In these tests, water infiltration into the slopes and upstream storm water ponds were modelled to simulate the most unfavourable working conditions. Displacement contours in the cross sections were measured and crest settlements were monitored during the tests. Finally, the optimum compaction zoning was analysed based on the comprehensive evaluation of the total crest settlement, the postconstruction settlement, and the slope stability of the embankments. It was indicated that a properly designed high loess embankment could perform satisfactorily with the lower part compacted looser than the conventional code requirements. In terms of total crest settlement and embankment cracking, the degree of compaction K of loess fills should not be less than 0.85 for 30 m high embankments and not less than 0.90 for 63.8 m high embankments. In terms of the postconstruction settlement (

Une methode par elements finis tridimensionnelle non lineaire a ete utilisee pour analyser le comportement general contrainte-deformation de barrages en enrochements de grande hauteur dans des vallees etroites. L'influence de la pente des appuis sur l'effet de voute qui prend place dans le sol des barrages est largement discutee, de meme que le potentiel de fracturation hydraulique dans le noyau et la fissuration dans les zones de contact du noyau. On a aussi analyse et compare le comportement de barrages de differentes hauteurs. On suggere des ameliorations pour la preparation des fondations des noyaux et pour la protection vis-a-vis de la fissuration.