Statistical characterization and stochastic simulation of load-displacement behavior of shallow footings

This paper focuses on the statistical characterization and stochastic modeling of the load-displacement behavior of full-scale quasi-square footings. The study relies on a large and comprehensive field database of 66 full-scale footings subjected to vertical loading on 4 different soil types (sands, silts, intact clays, fissured clays) with CPT cone resistances available for each site. The performance of 2 analytical load-displacement models in replicating empirical field data is assessed comparatively through statistical modeling of load-displacement data. Correlation between model parameters is investigated and replicated using copula theory. Following the identification of best-performance models for each soil type, load-displacement curves are simulated stochastically. The influence of soil type on statistical properties and modeling details are discussed. Best-performing configurations of modeling options are provided for practical implementation in geotechnical limit state design.