No abstract

Stone columns and geosynthetics have been widely used as effective stabilisation techniques to improve the load–settlement response of foundations. In view of this, an analysis of rigid pavements resting on geosynthetic-reinforced granular bed–stone column improved subgrade soil has been carried out. A detailed idealisation of the problem has been done using elements of Kelvin–Voigt body, Pasternak shear layer and Winkler springs, respectively, to simulate the behaviour of subgrade soil, granular fill and stone columns. A non-linear constitutive relationship for these elements has been considered. Governing differential equations for the soil–pavement system have been developed by modelling the geosynthetic as a rough elastic membrane and the pavement as a thin plate. These equations have been written in their finite-difference form and solved using the lower–upper (LU) decomposition method with the help of appropriate boundary conditions. A detailed parametric study to depict the influence of various factors affecting the pavement behaviour has been carried out. Quantitative evaluation of this influence has been done to enable the deflection and bending moment in the pavement and the tension mobilised in the geosynthetic layer to be obtained.

Abstract. Suction caissons are extensively used as anchors for o shore foundation structures. The uplift capacity of suction caisson is an important factor with respect to e ective design. In this paper, two recently developed AI techniques, i.e. Functional Network (FN) and Multivariate Adaptive Regression Spline (MARS), have been used to predict the uplift capacity of suction caisson in clay. The performances of the developed models are compared with those of other AI techniques: arti cial neural network, support vector machine, relevance vector machine, genetic programming, extreme learning machine, and Group Method of Data Handling with Harmony Search (GMDH-HS). The model's inputs include the aspect ratio of the caisson, undrained shear strength of soil at the depth of the caisson tip, relative depth of the lug to which the caisson force is applied, load inclination angle, and load rate parameter. The results of the above AI techniques are comparatively analysed via di erent statistical performance criteria: correlation coe cient (R), root mean square error, Nash-Sutcli e coe cient of e ciency, and log-normal distribution of ratio of the predicted load capacity to observed load capacity, with a ranking system to determine the best predictive model. The FN and MARS models are found to be comparably e cient which can outperform other AI techniques.

No abstract

No abstract