Joint Posterior Estimation of Overconsolidation Ratio and Rigidity Index from Co-Located CPTu-DMT Soundings

Abstract

The accurate characterization of subsurface soil properties is a fundamental prerequisite for the safe and economical design of geotechnical structures. Among the various state parameters, the overconsolidation ratio and the rigidity index are particularly critical, as they govern the stress history, compressibility, and undrained shear strength of fine-grained soils. Traditional deterministic approaches typically rely on empirical transformation models applied independently to different in-situ tests, which fails to account for inherent geological variability, measurement errors, and transformation uncertainties. This paper presents a comprehensive methodological framework for the joint posterior estimation of the overconsolidation ratio and rigidity index by integrating data from co-located Cone Penetration Tests with pore pressure measurement and Flat Dilatometer Tests. By employing a rigorous Bayesian inference approach, the proposed framework fuses the complementary mechanistic sensitivities of the two sounding methods into a unified probabilistic structure. The methodology relies on Markov Chain Monte Carlo sampling techniques to derive the joint posterior probability distributions, completely characterizing the parameter uncertainties and their cross-correlations. The application of this framework to a well-characterized clay deposit demonstrates that the joint assimilation of co-located sounding data significantly reduces the statistical dispersion of the estimated parameters compared to prior estimates or single-source interpretations. The results underscore the necessity of adopting multiparameter, multi-tool probabilistic methods to achieve more reliable and resilient geotechnical site characterizations.