International Journal of Geomechanics

Artificial neural network (ANN) models are developed in this study to correlate resilient modulus with routine properties of subgrade soils and state of stress for pavement design application. A database is developed containing grain size distribution, Atterberg limits, standard Proctor, unconfined compression, and resilient modulus results for 97 soils from 16 different counties in Oklahoma. Of these, 63 soils (development data set) are used in training, and the remaining 34 soils (evaluation data set) from two different counties are used in the evaluation of the developed models. A commercial software, STATISTICA 7.1, is used to develop four different feedforward-type ANN models: linear network, general regression neural network, radial basis function network, and multilayer perceptrons network (MLPN). In each of these models, the input layer consists of seven nodes, one node for each of the independent variables, namely moisture content $\left(w\right)$ , dry density $\left({\gamma }_d\right)$ , plasticity index (PI), percent passing sieve No. 200 $\left(P_{200}\right)$ , unconfined compressive strength $\left(U_c\right)$ , deviatoric stress $\left({\sigma }_d\right)$ , and bulk stress $\left(\theta \right)$ . The output layer consists of only one node—resilient modulus $\left(M_R\right)$ . After the architecture is set, the development data set is fed into the model for training. The strengths and weaknesses of the developed models are examined by comparing the predicted $M_R$ values with the experimental values with respect to the $R^2$ values. Overall, the MLPN model with two hidden layers was found to be the best model for the present development and evaluation data sets. This model as well as the other models could be refined using an enriched database.

The inelastic ${\text{MSDP}}_u$ (for Mises-Schleicher and Drucker-Prager unified) criterion has been developed over the last decade, and has been applied to simple problems for a wide range of geomaterials. This paper presents the approach taken to introduce this multiaxial criterion into a commercially available finite difference code using an elastoplastic framework. The paper begins with a brief review of the ${\text{MSDP}}_u$ formulation. Then, the main components of the elastoplastic model and the implementation procedure are described. The ensuing model, called ${\text{MSDP}}_u$ -EP (for ${\text{MSDP}}_u$ elastoplasticity model), is then validated against some existing and recently developed analytical solutions for the stress state around a circular opening. The newly developed model is further applied to evaluate the stress state near underground openings, using the geometry of a tunnel and of vertical backfilled stopes. A brief discussion follows.

The effect of foundation embedment on consolidation has not been considered previously in a systematic manner, although this is of particular interest for offshore foundations, where embedment is provided by skirts that enclose a compressible soil plug. For skirted foundations, critical uncertainties include what to assume in terms of the degree of drainage at skirt tip level, and the relative time scales of consolidation within the soil plug and beneath the foundation. In this paper, results from small strain finite-element analyses are used to quantify the immediate and time-dependent response of circular skirted foundations to uniaxial vertical loading. Foundations with frictionless and fully rough skirt-soil interfaces with varying ratio of embedment depth to foundation diameter are considered and the responses compared with those for surface foundations. The findings show that both skirt-soil interface roughness and embedment ratio have a significant effect on the consolidation response.

A constitutive model formulated within the multilaminate framework is described in this paper. The basic model, incorporating deviatoric and volumetric hardening, was developed for normally consolidated or slightly overconsolidated clay. This model has been extended to account for shear softening by using a nonlocal formulation. The nonlocal plasticity is based on weighted average of plastic strains taken from neighboring material points within a certain distance which is a function of an internal length parameter. This parameter is related to the mean grain size that restricts the element size. By using a softening scaling this restriction can be eliminated and the model is therefore capable of simulating geotechnical boundary value problems. The capability of the model for simulating strain softening behavior is shown by numerical simulation of biaxial tests. Finally, the ground response curve obtained from the proposed model for excavation of a tunnel in soil is discussed.

Design of waterfront retaining walls under seismic conditions is an important topic of research among the geotechnical engineering fraternity, and recently there have been studies in which the stability of rigid waterfront retaining walls has been assessed. However, an important aspect of seismic rotational stability of such walls is still missing from the literature archives. The present study shows the importance of rotational displacements for the design of the rigid waterfront retaining wall. Consideration has been made for the calculation of the hydrodynamic pressure as well as the seismic forces, both due to the seismic pressure and seismic wall inertia. These seismic forces have been calculated using the pseudodynamic approach. The free water condition has been considered in the analysis, and thus the hydrodynamic pressure has been considered to exist on the downstream face of the retaining wall as well, and a well-known expression approximating the effect of the inertia of the water due to the earthquake has been used for the estimation of this hydrodynamic pressure force. Simple expressions for the calculation of rotational displacement both during and after the earthquake have been proposed, and typical results have been obtained. It is observed that with an increase in the ratio of the water level to the total height of the wall from 0.50 to 1.00 the rotational displacement of the wall increases by about 110%. Similar trend of an increase in the value of the rotational displacement was observed for an increase in the values of the horizontal and vertical seismic acceleration coefficients. Also, the parametric study carried out in the analysis suggested that the rotational displacement is sensitive to other parameters such as the upstream water height, pore pressure ratio, soil, and wall friction angles. Due to nonavailability of the results in which rotational stability of the waterfront retaining wall under the seismic conditions has been studied, the results from the present analysis seem to bring out a unique approach.