No abstract

The longitudinal variation of soil properties has a major influence for many types of structure, including pavements, buried pipes, raft foundations and railways, as it induces stresses and/or displacements that cannot be predicted when assuming soil homogeneity. A set of simple numerical models has been developed to describe how soil–structure interaction can be influenced by soil variability. These models include: (a) a description of the soil spatial variability, within the frame of geostatistics, where the correlation length of soil properties is the main parameter; and (b) a mechanical description of the soil–structure interaction, which depends on the structure resting on the ground. There are some differences between a (more or less) rigid raft on piles, a set of buried pipes with (more or less) flexible connections and a hyperstatic beam, but the basic principles of mechanics are similar in all these cases. Several very general conclusion are drawn. (a) Soil heterogeneity induces effects (differential settlements, bending moments, stresses and possible cracking) that cannot be predicted if homogeneity is assumed. (b) The magnitude of the induced stresses depends on three factors: the magnitude of the soil variability (i.e. its coefficient of variation); a soil–structure stiffness ratio (in some cases, where the mechanics are more complex, one can consider two stiffness ratios, as in buried pipes for example, when one has to account for the flexibility both of the pipes and of the connections); a soil–structure length ratio, which combines the soil fluctuation scale and a structural characteristic length (distance between supports, buried pipe length, etc). In all cases, a worst value, corresponding to the value leading (from a statistical point of view) to the (statistically) largest effects in the structure, can be found. The principal benefit of such an approach is to provide some new approaches for better considering phenomena such as the geometrical irregularities in the longitudinal profile of pavements or during the control of soil compaction of sewer trench filling. This kind of approach can also give experts new tools for better calibration of safety in soil–structure interaction problems, when the soil variability is an influential parameter. Some practical conclusions are drawn in this direction.

No abstract

This project aims to evaluate the effects of changes in the correlation length of the rigidity of the subgrade flexible pavement roughness expressed in terms of the International Roughness Index (IRI). The correlation length, which defines the structure of the dispersion of soil rigidity, has a direct effect on the wavelength of the induced deformations, and thus the value of the IRI. The IRI is evaluated from pseudoprofiles decomposed according to specific wavelengths. It aims: (1) to show, according to a numerical and statistical approach, the influence of the inherent spatial variability of soil properties on the road in terms of stress and displacement, and (2) to analyze the effects of ground deformation on the values of the IRI.

Spatial and geometrical variability of mechanical soil properties can induce differential settlements and load redistribution in hyperstatic structures. Therefore damage prevention requires specific attention to be paid to the global mechanisms of soil–structure interaction. A reduced-scale model of a hyperstatic three-support frame (scale 1:100) is installed on the CEA-CESTA centrifuge, up to 100g. Various configurations are studied, with different loadings, different structural stiffnesses, and different geometries of the soil layer. Strain gauges are fixed at various points so as to enable the retro-analysis of all components of forces at free ends. Displacements are recorded at several points on the structure and on the free surface. A numerical model of the frame (based on beam theory with elastic supports) is calibrated, first to determine an equivalent support stiffness, and second to quantify the effect of variations of the structural stiffness/soil modulus ratio on the structural response. A probabilistic approach based on Monte Carlo simulation is thus developed. It mixes the soil description, based on random field theory and enabling the description of spatial correlation of properties and the frame model identified above. The influence of (a) scattering in soil properties and (b) spatial correlation of soil properties—through correlation length or scale of fluctuation—is quantified in terms of displacements and bending moments in the frame. The approach reveals the important weight/influence of soil variability on differential settlement and load redistribution, particularly when the scale of fluctuation of soil and length of span have the same magnitude. The analysis makes it possible to evaluate the risk that the bending moment in any critical section may be larger than predicted by the usual engineer's model. Some conclusions are drawn to better estimate ‘characteristic bending moments’ in the frame, accounting for both the soil variability and the soil–frame interaction.