Skip to Main content Skip to Navigation

Etude probabiliste de la stabilité d'un barrage en remblais. Prise en compte des données in-situ

Abstract : Uncertainties of soil properties are widely encountered in the field of geotechnical engineering especially for earth dams which are constructed with earthen materials. In recent years, there is an increasing need, motivated by the deficiencies of the traditional deterministic approach or guided by the national regulations such as in France, of accounting for these uncertainties for a safe assessment of large dams particularly in the framework of risk analysis studies. However, probabilistic analyses are still complex and not so easy to implement in practice due to the limited number of in-situ measurements, expensive computation efforts and lack of implementation of reliability methods in commercial simulation tools. Moreover, most of the previous studies are based on academic cases and hypothetic data.This work attempts to deal with the aforementioned issues by providing a probabilistic analysis study for the stability of a real earth dam using available field data. This study includes the following main elements: (1) definition of the soil variability by using the available measurements; (2) development of the deterministic models; (3-4) dam probabilistic analyses using the random-variables and random-fields approaches; (5) three-dimensional reliability analysis of the considered dam. Advanced reliability methods, such as the adaptive surrogate modelling, are introduced for the studied earth dam problem. This allows accurately estimating the dam failure probability and the safety factor statistics with a significantly reduced calculation time. In addition, some issues, that remain unknown or unclear in the field of the dam probabilistic analysis, are discussed (e.g. global sensitivity analysis of the soil hydraulic and shear strength parameters; performance survey of five reliability methods; simulation/comparison of three different kinds of random fields: generic (unconditional-stationary), conditional and nonstationary). The presented work, based on real measurements, could be a good supplement to the existing probabilistic studies of geo-structures. Readers will find useful information from the obtained results in order to better solve the practical geotechnical problems in a probabilistic framework.
Complete list of metadatas

Cited literature [242 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Tuesday, July 7, 2020 - 11:13:41 AM
Last modification on : Wednesday, October 14, 2020 - 4:18:52 AM


Version validated by the jury (STAR)


  • HAL Id : tel-02891918, version 1




Xiangfeng Guo. Etude probabiliste de la stabilité d'un barrage en remblais. Prise en compte des données in-situ. Mechanics of materials [physics.class-ph]. Université Grenoble Alpes [2020-..], 2020. English. ⟨NNT : 2020GRALI017⟩. ⟨tel-02891918⟩



Record views


Files downloads