Stress Path Tests Research Articles

Behavior of a Loosely Compacted Unsaturated Volcanic Soil

In this paper, the stress-strain relationsihp and volumetric behavior of a loosely compacted unsaturated decomposed volcanic soil (fill) were studied by conducting three series of triaxial stress path tests: (1) consolidated undrained on the saturated fill; (2) constant water content; and (3) a reducing suction under constant deviator stress on the unsaturated fill. The last two series of tests were designed to simulate the effects of undrained response and rainfall infiltration in initially unsaturated slopes, respectively. It was found that the saturated loose volcanic soil behaves like clay under isotropic compression but it resembles sand behavior when it was subjected to undrained shear. For isotropically consolidated unsaturated specimens sheared under a constant water content, a hardening stress-strain and a nonlinear shear strength–suction relationship are observed. At relatively high suctions, both angle of friction and apparent cohesion appear to be independent of suction. Volumetric contraction during shear is observed in this series of tests. On the other hand, anisotropically consolidated loose unsaturated specimens subjected to a reducing suction change from contractive to dilative behavior as the net mean stress increases. This observed volumetric behavior, unlike the shear strength, is stress path–dependent and cannot be explained by using the existing elastoplastic critical state theoretical framework extended for unsaturated soils.

Structure degradation and changes in the mechanical behaviour of a stiff clay due to weathering

The paper discusses the results of research into the causes of weathering of stiff Pleistocene clays located in the Montemesola Basin (TA, Italy), and on the effects of weathering on the clay's mechanical behaviour. The weathered clays are yellow-brown and overlie the original grey clays in the whole basin. The liquidity indexes of both the yellow and the grey clays, their fabric, as observed by means of scanning electron microscopy, as well as the yellow clay oxidation features suggest that drying has been the main weathering process in the basin. This is confirmed by the results of drying–wetting cycle tests carried out in the laboratory on undisturbed clay samples, which show the drying–wetting cycles to produce changes in the grey clay similar to those present in situ due to weathering. Results of oedometer, stress path and triaxial tests show that weathering has caused a degradation of the clay bonding and an associated reduction in the size of the clay state boundary surface. Both one-dimensional and iso-tropic compression data and the shear data indicate that the effects of weathering on the mechanics of the clay may be established according to the decrease in the clay stress sensitivity. Weathering also causes a decrease in the normalised shear stiffness, as observed by means of bender element tests, which is also related to a decrease in the stress sensitivity.

The effect of anisotropic elasticity on the yielding characteristics of overconsolidated natural clay

Quantitative application of elastoplastic theory to the yielding behaviour of natural soils has always been uncertain. Part of the reason is that the theory was developed for reconstituted materials with isotropic structure, in contrast to natural soils that are usually anisotropic. The approach considered in this study assumes that pre-yielding behaviour is governed by the theory of linear anisotropic elasticity and that yield loci in the mean effective stress ( p') – deviator stress (q) plane are aligned approximately along the coefficient of earth pressure (K0) line. The assumption of a rotated yield locus associated with anisotropic elastic behaviour within the state boundary surface indicates that the elastic wall within the state boundary surface is inclined. The form of the state boundary surface has been determined mathematically in terms of anisotropic elastic and Cam-Clay soil parameters. Stress path tests were conducted on samples of Belfast Upper Boulder Clay removed from a depth of 28 m below ground surface. Good agreement was found between predicted and measured yield loci. The study also examined the influence of subsequent isotropic compression on the yielding characteristics of the natural clay. The indications are that the anisotropy developed during deposition disappears when the sample is loaded to a stress level at least twice the stress generated during the original deposition process. The methods developed in the paper have also been applied to test results reported previously on Winnipeg clay, and good agreement was obtained.Key words: sampling, anisotropy, elasticity, suction, stiffness, yielding.

Stress Path Testing for Proper Characterization of Unbound Aggregate Base Behavior

Realistic pavement stresses induced by moving wheel loads were examined in granular base layers, and the significant effects of rotation of principal stress axes were addressed for a proper characterization of unbound aggregate behavior. Granular material resilient moduli are commonly determined at the centerline of wheel loading without taking into account the effects of moving wheel loads and the constantly rotating field principal stress states. Differences that exist between field- and laboratory-applied stress states were identified by comparing the field stresses simulated using a nonlinear axisymmetric finite element program, GT-PAVE, with the stress states commonly used in the standard resilient modulus test procedure, AASHTO T294-94. Three sets of complete triaxial test data obtained from testing aggregates under various realistic in situ stress paths caused by moving wheel loading were analyzed. Eight different granular material modulus models were developed from the experimental test data to include the applied mean stress, the applied shear stress, and the slope of stress path loading. Because of the complex loading regimes followed in the laboratory tests, characterization models that simultaneously analyzed the static and dynamic components of the applied mean and deviator stresses produced a high degree of accuracy. Models that consider the stress path slope variations predict the best stress path dependency of aggregate behavior caused by moving wheel loads. Such advanced models that allow for the effects of principal stress rotation better describe the granular material behavior under the actual field loading conditions.

Yielding characteristics of unsaturated compacted soils

The elasto-plastic constitutive model for unsaturated soils postulates the existence of a loading collapse yield locus (LC) and a suction increase yield locus (SI) in the p:s stress plane. The characteristics of the LC yield locus have been reasonably well established in recent years but there is a much greater degree of uncertainty in relation to the SI yield locus. In the work reported in this paper a number of stress path tests in the p:s plane were conducted on samples of unsaturated compacted kaolin in a modified triaxial apparatus. Two series of tests were conducted on samples compacted to an identical initial condition so as to ensure a uniform soil structure prior to testing. Samples tested in Series 1 were taken through three stages: wetting at constant mean net stress, drying at constant mean net stress and a combination of isotropic compression and wetting. Samples tested in Series 2 were taken to pre-selected values of suction prior to isotropic compression. Data obtained from these tests indicate a possible form of single yield locus for unsaturated compacted soil. The data also indicate that the position of the yield locus, initially imposed by the compaction process will be influenced by the subsequent wetting process. Copyright © 2000 John Wiley & Sons, Ltd.

Undrained anisotropy ofK0-consolidated silt

Direct measurements of the initial shear strength and yielding anisotropy of a dense, K0-consolidated silt are described and interpreted within a bounding-surface framework. The experiments were performed using the Imperial College large hollow cylinder apparatus, in which samples were sheared undrained with a range of orientations, α, of the major principal stress, σ1, following initial K0consolidation. The interpretation is aided by data from oedometer and triaxial stress path tests. Strongly anisotropic stiffness, yielding, undrained strength, and mobilized angle of shearing resistance, ϕ', characteristics are revealed. The effects of drained and undrained stress changes applied to the samples after K0consolidation are also described.Key words: anisotropy, hollow cylinder, K0consolidation, silt, bounding surface.

The Effects of Varying Centerline Tube Sampling Disturbance on the Behavior of Reconstituted Clay

Abstract In order to simulate varying degrees of disturbance of soil at the centerline of a tube sampler, triaxial stress and strain path tests have been carried out on normally consolidated and overconsolidated reconstituted London clay specimens. The tests were carried out using an automated system capable of controlling both stresses and deformations imposed on specimens. The specimens were instrumented with local axial and radial strain and mid-plane pore pressure measuring devices. Test results indicate that the most pronounced effects due to imposed tube sampling strains in reconstituted normally consolidated London clay are significant reductions of mean effective stress and undrained small-strain stiffness, which are accompanied by reductions in the excess pore pressure generated during shearing. The higher the magnitudes of tube sampling strains, the greater the change in behavior. In over-consolidated reconstituted London clay, however, laboratory simulation of tube sampling strains together with deviatoric stress relief (i.e., “ideal” sampling) causes little change in effective stress, stiffness, and strength.

Stress–strain behaviour of a loosely compacted volcanic-derived soil and its significance to rainfall-induced fill slope failures

The stress–strain behaviour of a soil is crucial to the recognition of the mechanism of slope failure. Triaxial tests, composed of isotropically consolidated drained (ICD) and undrained compression (ICU) tests and anisotropically consolidated constant deviatoric stress path (CQD) tests, were carried out with the aim of investigating the stress–strain behaviour of loosely compacted volcanic-derived soils. The fact that the critical states are the same for ICU, ICD and CQD tests may show that the critical state is independent of the above three stress paths. The critical state line, as defined in critical state soil mechanics, is obtained from the e–log p′ and q– p′ plots based on the results of the above tests. The initial state of the consolidated specimens at initiation of failure may be classified as dilative or contractive in the light of the locations of the soil state relative to the critical state line. For contractional soils, the increased pore water pressure generated by rainfall infiltration leads to a contractive failure in a drained manner, giving rise to high excess pore water pressure. The excess pore water pressure caused by contraction cannot be dissipated instantly, resulting in a decrease in the shear resistance of the soil. The failure process is rapid. The failed soil mass is prone to flow after failure under the action of gravity due to its high moisture content and inflow of surface runoff and rainwater. For dilational soils, the increased pore water pressure resulting from infiltration leads to dilation, which reduces pore water pressure and thus increases the shear resistance of the soil. However, continued rainfall infiltration may be able to equilibrate the reduction in pore water pressure caused by dilation and, therefore, the dilation or displacement can continue. In Hong Kong, volcanic-derived soil is characterized by high permeability. Both the high permeability of volcanic-derived soil and a shallow failure surface make it possible for the reduction in pore pressure to equilibrate relatively quickly. Therefore, the failure is also rapid, at least for poorly compacted fill slopes.

Stress Paths in Relation to Deep Excavations

The mechanical behavior of many soils such as stiff clays depends on their current effective-stress states and stress history. For improving design and analysis of soil-structure interaction associated with deep excavations in these soils, it is important to understand effective-stress changes around excavations caused by both horizontal and vertical stress relief. In this paper, total and effective-stress variations adjacent to a diaphragm wall during construction of a 10–m-deep excavation in stiff fissured clay are reported and discussed. Interpreted field stress paths are compared with some relevant laboratory triaxial stress path tests, which simulate the horizontal and vertical stress relief in the field at an appropriate stress level. The interpreted field effective-stress paths in front of the wall are found to be similar to laboratory stress paths determined in undrained extension tests. Field stress paths behind the wall do not correspond particularly well with those from laboratory undrained compression tests, except when the stress state approaches active failure. The conventional undrained assumption does not seem to hold for the soil located immediately behind the wall during a relatively rapid excavation in the stiff clay.

Analysis of rainstorm-induced slide-debris flows on natural terrain of Lantau Island, Hong Kong

Lantau Island, the largest outlying island of the territory of Hong Kong, experienced a severe rainstorm on 4–5 November 1993, which induced >800 slope failures on natural terrain there. Detailed field investigations were carried out to study the failure modes, in relation with various influencing factors. It was found that the occurrence of slide-debris flows has a close relationship with bedrock geology, slope gradient, vegetation cover and micro landform. The failure modes of slide-debris flows may be classified into translational slides and rotational slides, and the former are predominant. Analysis of the hydrological response of colluvial slopes during the rainstorm indicated that the majority of the failures were caused by the development of a perched water table in the thin surface layer of colluvium of volcanic origin due to infiltration during the heavy rain. Undisturbed soil samples from south Lantau have been subjected to anisotropically consolidated undrained compression tests at comparatively low stress levels. Constant deviatoric stress path tests (CQD) simulating the stress path in the field at in situ stress levels have been performed to investigate soil behavior. The CQD test results indicate that the material of slopes at undisturbed state is brought to dilation because of the increase in pore water pressure caused by infiltration of rain water. For a translational slide, the displacement, resulting from dilation, may destroy cohesion along the failure surface and locally within the interior of the slide. The surplus water during the intense rainstorm was able to equilibrate the reduction in pore pressure caused by dilation, and the dilation and displacement may be further increased. The strain-softening after significant strains triggered debris flow mobilization. However, for a rotational slide, the increase in pore water pressure caused by surplus water infiltration during the intense rainstorm could not equilibrate the reduction in pore pressure caused by dilation, much or even all of the sliding block could not mobilize into a debris flow.

The Cilfynydd flow slide of December 1939

Abstract This paper describes an investigation of the 1939 flow slide in colliery spoil at Cilfynydd Common, South Wales. The investigation concentrated on the geomorphological features and other aspects that were to be destroyed by a proposed reclamation scheme. The overall aim of the work was to compile a database of factual information on ground and material conditions at the flow slide site. The work involved geomorphological mapping, a site investigation which involved 18 trail pits and a detailed programme of laboratory testing including fabric analyses. The results of the study indicated distinctly different fabrics for materials from different locations in the failed mass of the landslide. In triaxial stress path tests excess pore water pressures were generated very rapidly during shearing, indicating the susceptibility of slopes formed of colliery spoil to fail by flow slide.

Selecting the rate of loading for drained stress path triaxial tests

Despite certain restrictions on its validity, the use of a triaxial apparatus is still a commonly used laboratory test method to determine soil behaviour. Solutions were obtained many years ago for evaluating mean excess pore pressure values in soils subject to continuous rates of loading. They have been used to predict mean excess pore pressures in consolidation tests with constant rates of loading. Such solutions may be suitable for predicting excess pore pressures in soil specimens, subject to continuous loading, in drained triaxial stress path tests. This note proposes the use of these solutions to select the rate of loading for continuously loaded, drained stress path triaxial tests. The analytical solutions of the equations in the known mathematical model may be used to predict the average excess pore pressure in a triaxial sample, subjected to time-dependent loading, at any time during a continuous triaxial stress path. This requires an estimate of the imposed excess pore pressure. A calculation method is given for this pressure, together with a proposed method for predicting deviation from the required effective stress path. As an illustrative example, a prediction of the effective stress path, during a typical drained triaxial stress path test, is presented.

Anisotropy of G0 shear stiffness in Gault Clay

The inclusion of anisotropic stiffness parameters in sophisticated constitutive models necessitates their determination in the field and laboratory. Hitherto, anisotropy of small-strain stiffness of clays has occasionally been examined in the laboratory by measurements on specimens sampled at different orientations. The authors have developed a device to propagate and receive horizontal shear waves with both vertical and horizontal polarization through 100 mm triaxial samples. The device incorporates bender elements embedded in the pads of a horizontal belt. In tests on both reconstituted and natural samples of Gault Clay, each of the transmission velocities Vs(xh), Vs(hx) and Fs(hh) has been measured during stress path tests. The ratio G0(hh)/G0(vh) has been shown to be highly dependent on stress state. Laboratory data on natural samples show results broadly consistent with in situ results. Pour pouvoir inclure des paramètres de rigidité anisotrope dans des modèles constitutifs perfectionnés, il faut les déterminer sur le terrain et en laboratoire. Jusqu'à présent, on a parfois examiné l'anisotropie de la rigidité d'argiles aux petites contraintes en mesurant des échantillons prélevés à diverses orientations. Les auteurs ont mis au point un dispositif qui permet la propagation et la réception d'ondes de cisaillement horizontals avec polarisation verticale et horizontal dans des échantillons de compression triaxiale de 100mm. Le dispositif comprend des elements de flexion noyés dans les plaquettes d'une bande horizontal. Dans des essais de parcours de contrainte sur des échantillons reconstitués et naturels d'argile de Gault, on a mesuré les vitesses de transmission Vs(vh), Vs(hv) et Vs(hh). On a montré que le rapport G0(hh)/G0(vh) depend en grande partie de l'etat de contrainte. Les résultats obtenus en laboratoire sur des échantillons naturels concordent assez bien avec les résultats obtenus sur le terrain.

Stiffness determination and deformation analysis for a trial excavation in Oxford Clay

Block samples of the heavily overconsolidated Oxford Clay were retrieved from a 10m deep, instrumented, trial excavation at Elstow, Hertfordshire, and from a nearby brick pit. Triaxial stress path tests were conducted on 100 mm diameter specimens to determine appropriate stiffness parameters for numerical modelling of the excavation. Each test followed a succession of constant p' (mean effective normal stress) and constantq (deviator stress) paths in the extension region. Small axial and radial strains were measured locally using a proximity transducer system, permitting stiffness parameters for the coupled deviatoric and volumetric behaviour to be derived and plotted against changes of strain energy for modelling purposes. The maximum shear stiffness, G0, was also determined using bender elements. During the multistage tests, stiffness behaviour at small strains appeared to be governed by the recent strain history and recent stress history effects were not as prominent as those reported in the literature for other clay soils. Finite element modelling of the excavation was carried out, employing a non-linear cross-anisotropic elastic model in which stress path dependent stiffnesses were specified. Predicted lateral displacements were modestly overestimated by comparison with field measurements, but vertical displacements were generally underpredicted and to a greater degree. Des blocs d'échantillonnage d'argile d'Oxford fortement préconsolidée ont été prélevés dans une fouille de recherche instrumentée, profonde de 10m, à Elstow dans le Hertfordshire, ainsi que dans une argilière voisine. On a effectué des essais de parcours de contraintes triaxiales sur des échantillons de 100 mm de diamètre pour déterminer les paramètres de rigidité convenant à une modélisation numérique de la fouille. Chaque essai a suivi une succession de parcours de contrainte constante p' (contrainte intergranulaire normale moyenne) et de contrainte constante q (contrainte déviatrice) dans la zone d'allongement. On a mesuré localement les petites contraintes axiales et radiales à l'aide d'un réseau de capteurs de proximité en vue de calculer les paramètres de rigidité pour le comportement dévuatorique et volumétrique jumelé et de les mettre en corrélation avec les variations d'énergie de contrainte à des fins de modélisation. In a également déterminé la rigidité maximale au cisaillement G0 à l'aide d'é1éments de flexion. Dans les essais à étages multiples, la rigidité aux petites contraintes a semblé &circtre dictée par les contraintes subies récemment, et les effets des tensions récentes ne semblent pas aussi importants que ceux dont font état d'autres travaux portant sur d'autres sols argileux. On a effectué une mod6lisation de la fouille par é1éments finis en utilisant un mod&gravele élastique trans-anisotrope non linéaire où l'on a précisé des rigidités tributaires des parcours de contrainte. Les déplacements latéraux prédits ont été 1égèrement supérieurs aux déplacements mesurés sur le terrain, tandis que les déplacements verticaux ont W généralement sous-estimés et dans de plus grandes proportions.

The development and evaluation of a constitutive model for the prediction of ground movements in overconsolidated clay

The stress—strain response of overconsolidated clay depends both on its current state and on the loading history followed to reach that state, in particular the relative directions of the current and previous loading paths. A constitu-tive soil model is developed which predicts this behaviour by allowing elasto-plastic deformations controlled by two nested kinematic hardening surfaces inside a conventional Modified Cam-clay state boundary surface. This relatively straightforward model requires only eight para-meters, each with a rational basis, and which can be determined from a small number of well-controlled stress path tests. Predictions of soil behaviour using this model are compared with data from triaxial stress path tests. The close agreement confirms that the essential features of soil behaviour are predicted by the model. The wider implications of the use of the model in geotechnical analysis are illustrated by compar-ing predictions made using the model (in conjunction with finite element analysis) with data from a specially commissioned series of centrifuge tests of a circular foundation loaded on overconsolidated clay. The stress history of the soil was carefully controlled in the experiments and was replicated in the course of the analyses. The computations reproduced the main characteristics of the observed ground movement, in particular the surface profile. In contrast, conventional constitutive models of soil behaviour show very poor predictions. This demonstrates the importance of using a model that simulates the behaviour of soil over a wide range of strain increments and with changes in load path direction. La réaction tension/déformation de ĺargile préconsolidée dépend à la fois de ĺétat de ĺargile et des charges antérieures, notamment des directions relatives de la charge présente et des charges précédentes. ĺexposé décrit un mod&gave;le de sol constitutif qui prédit ce compor-tement en permettant des déformations élastoplastiques engendrées par deux surfaces ciné matiques de durcissement emboîtées dans une surface limite classique ´argile du Cambrien modifiée. Ce modèle relativement simple ńexige que huit paramètres, chacun avec une base rationnelle, et pouvant être déterminés à pardr ´un petit nombre ´essais bien contrôlés des parcours de tension. Les prédictions du comportement du sol faites à ĺaide de ce modèle sont comparées aux données fournies par des essais triaxiaux des parcours de tension. La bonne corrélation entre les deux ensembles de valeurs confirme que le modèle pent prédire les caractéristiques essentielles du comportement du sol. ĺexposé illustre les autres applications possibles de ce modèle à ĺanalyse géotechnique en comparant les prédictions faites à ĺaide du modèle (en conjunction avec une analyse des é1éments finis) aux données fournies par une série spéciale ´essais centrifuges ´une fonda-tion circulaire chargée sur de ĺargile prĺcon-solidée. On a soigneusement documenté les tensions successives dans les essais pour pouvoir les reproduire dans les analyses. Les calculs ont reproduit les principales caractéristiques du mouvement de sol observé, notamment le profil de surface. Par contraste, les modéles constitu-tifs classiques du comportement du sol donnent de très mauvaises prédictions. Cela montre qúil est important ´utiliser un modèle qui simule le comportement du sol sur une vaste gamme de tensions et avec des changements de direction des charges.

Some mechanical properties of reconstituted Boom clay

The mechanical behaviour of reconstituted normally consolidated Boom clay was examined in a series of laboratory triaxial stress path tests. The aim was to establish some basic characteristics of this soil. The compressibility of the reconstituted Boom clay was found to be moderate, corresponding to the soils of the same plasticity. The results indicated also that the destructured Boom clay exhibited a brittle behaviour. The undrained secant stiffness was found to vary with strain level and also to be dependent on the consolidation pressure.

Effect of intermediate principal stress on the deformation response of sand

An experimental investigation of the relevance of the intermediate principal stress (σ2) on the deformation response of a sand is presented. The effects of σ2 are conveniently studied through the nondimensional stress parameter b = (σ2 – σ3)/(σ1 – σ3). A series of stress path tests was performed on Ottawa sand specimens in a hollow cylinder torsional shear device. The experimental program includes shear loading at different values of b, and special b tests, in which b was continuously varied at different stress directions. It is shown that the b value may have a significant influence on the stress–strain response of sand, depending on the loading conditions. Key words: hollow cylinder tests, generalized stress paths, sand, stress–strain behaviour, intermediate principal stress.

Behavior of Shaft-Sand Interface from Local Measurements

The load-displacement response of axially loaded structural inclusions in soil, which transfer the load to the soil along their shafts, is of growing interest in geotechnical engineering. The load-displacement response of a shaft interface is characterized by nonlinear and inelastic behavior. Surface roughness of the inclusion and stresses and deformation characteristics (stress-strain response, dilation, or contraction) of the soil element surrounding the inclusion are significant aspects of the interface mechanism. Localized shear displacement at the soil-shaft interface necessitates use of a constitutive model specifically developed for the interface. To verify theoretical models and modify and improve them, laboratory tests are performed. In recent years, there has been increasing emphasis in measurement of small-magnitude local strains to define stiffness at low strains similar to those encountered in the field in stress-path testing. To verify the theoretical model of the shaft-soil interface, a special test rig was developed that uses a cylindrical soil specimen with an inclusion in its center. The specimen represents a soil element surrounding a structural inclusion. The shaft-sand interface was investigated by small-magnitude local measurement of interface strains and thereby interface stresses and displacements during axial loading of the inclusion. The results are compared with those inferred from global measurements of interface variables. The advantages of the small-magnitude local strain measurements in determining the interface model parameters are presented. The use of the measured quantities in an elasto-plastic interface model is demonstrated by capturing the effects of confining stresses and boundary conditions of the soil specimen surrounding an axially loaded inclusion as well as the surface roughness of the inclusion.

Behavior of Shaft-Sand Interface from Local Measurements

The load-displacement response of axially loaded structural inclusions in soil, which transfer the load to the soil along their shafts, is of growing interest in geotechnical engineering. The load-displacement response of a shaft interface is characterized by nonlinear and inelastic behavior. Surface roughness of the inclusion and stresses and deformation characteristics (stress-strain response, dilation, or contraction) of the soil element surrounding the inclusion are significant aspects of the interface mechanism. Localized shear displacement at the soil-shaft interface necessitates use of a constitutive model specifically developed for the interface. To verify theoretical models and modify and improve them, laboratory tests are performed. In recent years, there has been increasing emphasis in measurement of small-magnitude local strains to define stiffness at low strains similar to those encountered in the field in stress-path testing. To verify the theoretical model of the shaft-soil interface, a special test rig was developed that uses a cylindrical soil specimen with an inclusion in its center. The specimen represents a soil element surrounding a structural inclusion. The shaft-sand interface was investigated by small-magnitude local measurement of interface strains and thereby interface stresses and displacements during axial loading of the inclusion. The results are compared with those inferred from global measurements of interface variables. The advantages of the small-magnitude local strain measurements in determining the interface model parameters are presented. The use of the measured quantities in an elasto-plastic interface model is demonstrated by capturing the effects of confining stresses and boundary conditions of the soil specimen surrounding an axially loaded inclusion as well as the surface roughness of the inclusion.

Collapse behavior of sand: Discussion

The test data that are presented in paper are of most interest. The two constant deviator stress path tests may be compared with tests done by Lade et al. (1988) and Lade and Pradel (1990) for very loose sand under undrained conditions and by Chu (1991) and Chu et al. (1993) for medium loose to dense sand under non-undrained (or strain path controlled) conditions. It should be noted that testing procedures used in and test results obtained from these tests are verv similar. All above-mentioned tests were conducted uhder a constant deviator load and decreasing effective confining stress conditions. The prefailure instability observed in these tests appears to be same type. Although authors called instability observed in constant deviator stress path tests collapse behaviour, word drained does not carry same meaning as in a drained conventional compression test. This is because pore-water pressure of sample was controlled during tests, and thus volume change of sample is not totally free. As the volume of sample for constant deviator stress (q = 125 kPa) test was essentially unchanged and remained at 0.804 before sample collapse, this test was actually conducted under an undrained condition. Therefore, 'paper by S. Sasitharan, P.K. Robertson, D.C. Sego, and N.R. Morgenstern. 1993. Canadian Geotechnical Journal, 30: 569-577.