Contribution à l'auscultation des chaussées mixtes par méthodes GPR et électrique

Quebec's infrastructures are ageing. A reliable, quick and economical assessment of the state of urban roads would help to plan civil engineering. This M.Sc. thesis aims to investigate and develop the potential of Ground Penetrating Radar (GPR) and electrical methods to assess pavement condition subject to wetting, freezing and drying cycles. The objectives are to display defects and estimate water content in the materials used for the mixed pavement construction; these consist in layers of asphalt (top), concrete (middle) and grade and sub-grade (bottom). Prior studies showed a valuable potential of joint application of GPR and electrical resistivity to assess mixed pavement condition. These studies were mainly based on numerical modeling and some site tests using both methods. It was concluded that a better evaluation of the potential of the coupled methods would come from the assessment of physical properties and validation from geotechnical evidences. At first, we characterized the context. Structural defects and water content implied modification of physical properties of road materials. After running numerical modeling, we carried out an experiment in one of the streets in Montreal (Canada) where pavement needed complete rehabilitation. GPR surveys were carried out before the pavement was destroyed and removed, and subsequently a resistivity imaging was carried out on the subgrade layer. We also sampled blocks of asphalt and concrete along the street from the demolition rubble in order to study their physical properties in the lab. Soil samples were taken from trenches to measure water content. The grade and sub-grade were mainly a mix of clayey-silt and gravel, as the embedding ground is grey clay. We measured the complex dielectric constants of asphalt and concrete in the range 50 MHz-900 MHz with a recently developed dielectric probe sampling a volume of 30 cm3. Resistivity measurements were also performed on large cores of asphalt and concrete using an in-house system. Measurements were carried out on dry and wet samples, which provided a range of values encompassing the climatic conditions for the whole year. Finally the physical property and geotechnical data were used as input to numerical modeling and the modeled responses compared to the survey data. The predicted data fit nicely with the observed data. This work aims to provide further information for a better calibration of geophysical devices in pavement condition assessment. Moreover, it gives a wide range of information, which could become a base for further and more specific works.