Determination of the correlation length of geophysical data and its application to pile foundation design : How many cpt test are needed to caracterize the soil, have a realistic probability of failure ?
Files
DamiriMariam_84551900_2526.pdf
UCLouvain restricted access - Adobe PDF
- 10.6 MB
Details
- Supervisors
- Faculty
- Degree label
- Abstract
- The spatial variability of soil properties has a significant influence on the performance and reliability of geotechnical structures, particularly pile foundations. In current engineering practice, soil characterization is commonly based on a limited number of in situ tests and deterministic design approaches, which often neglect spatial uncertainty and may lead either to overly conservative designs or to an unrealistic assessment of failure risk. Recent advances in geophysical investigation techniques provide continuous subsurface information over large areas, offering new opportunities to better capture soil heterogeneity. However, the integration of geophysical data with geotechnical measurements, such as cone penetration tests (CPT), remains challenging. In particular, the determination of an adequate number of CPTs required to reliably characterize the soil and to achieve a realistic probability of failure is still an open question. This study investigates the use of geostatistical concepts, with a particular focus on correlation length, to quantify the spatial variability of geophysical data and to assess its implications for pile foundation design. By linking soil spatial structure to the density of geotechnical investigations, the proposed framework aims to improve the representation of uncertainty in soil properties and to evaluate the influence of site investigation strategies on foundation reliability. The results highlight the critical role of correlation length in soil characterization and demonstrate its impact on the convergence of soil parameters, the estimation of failure probability, and the overall reliability of pile foundation design. This work contributes to a more rational and probabilistic approach to geotechnical design, supporting informed decisions regarding site investigation efforts and foundation safety.