Integrating local data into geotechnical design through a predictive probabilistic approach
Integrating local data into geotechnical design through a predictive probabilistic approach
Site investigation and assessment of the soil and rock characteristics are essential parts of the geotechnical design process. These characteristics can be assessed from a variety of in situ tests. In France, the Ménard pressuremeter test is widely used. This in-situ testing method is used to determine the stress-strain relationship of the tested soil and to evaluate, among others, the Ménard limit pressure (plM) which is used to calculate the bearing capacity of a foundation. In this paper, we show how different predictive models can be used to infer a distribution, with depth, of the plM parameter for a new construction project, based on the historical data available in the vicinity. We then demonstrate how this information is used throughout the process of site investigations and geotechnical design. Before the site investigations, this approach allows for a preliminary overview of the mechanical characteristics of the soils, and for better targeted field investigations. This method is then used in the interpretative report, to ensure better reliability of the geotechnical design and construction recommendations.