Banca de DEFESA: María Camila Olarte Garzón

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : María Camila Olarte Garzón
DATE: 28/04/2023
TIME: 14:30
LOCAL: Online-Teams
TITLE:

MODELLING OF THE HYDROMECHANICAL BEHAVIOUR OF UNSATURATED EXPANSIVE SOILS

KEY WORDS:

HYDROMECHANICAL BEHAVIOUR, EXPANSIVE SOILS, UNSATURATED MEDIA, MATHEMATICAL MODELLING

PAGES: 104
BIG AREA: Engenharias
AREA: Engenharia Civil
SUBÁREA: Geotécnica
SUMMARY:

Expansive soils are materials composed mainly of active minerals, such as montmorillonite, which tend to undergo volumetric changes due to moisture content migration. Different geotechnical engineering practices, such as the design of a dam core or the compaction of the subgrade, require a thorough analysis and prediction of the expansion phenomenon under unsaturated transient conditions. The main problem in evaluating the mechanical response of this type of soil lies in the nonlinear nature of most of the unsaturated properties, which requires sophisticated tools that include parameters that are difficult to obtain. This study performs a mathematical analysis to understand the hydromechanical behaviour of expansive soils under unsaturated transient flow. The implementation of two mathematical models based on the Richards (1931) equation using a simplified coupling is used to evaluate the approach. Based on experimental results obtained from soil-column tests, the behaviour of the resulting model under different suction and moisture content conditions is explored. Four cases of analytical solutions are established to build a model-specific response framework, each with restrictive assumptions tailored to an initial and two boundary conditions. In this case, the hydromechanical predictions in unsaturated flow are adjusted considering the variations of the void ratio along the expansion process to account for the impact of volumetric deformations on the model parameters. Finally, a parametric calibration with literature results is performed to evaluate the parametric sensitivity and the range of volumetric variation in which the model is constrained. The results indicate a high statistical proximity between the numerical estimation and the experimental data, with correlation factors higher than 97%, both in the wetting and drying paths of the void ratio - log suction curve. Likewise, the predictions of the analytical solutions for cases 1 and 3 are consistent for each parameter analyzed and work for unimodal and bimodal Soil Water Retention Curves.

BANKING MEMBERS:
Presidente - 1805002 - ANDRE LUIS BRASIL CAVALCANTE
Externo à Instituição - JUAN CARLOS RUGE CARDENAS - UMNG
Interno - 2161425 - JUAN FELIX RODRIGUEZ REBOLLEDO
Interno - 3132829 - RAFAEL CERQUEIRA SILVA