The chemical grout injection method is performed in concrete fractures in tunnels for waterproofing. Liquid polymers such as polyurethane, epoxy or acrylic are injected in the fracture with a pump.
Presently, the professional performing an injection has no guidelines or standard procedure to develop an injection protocol that ensure an adequate and permanent sealing of the fracture. Therefore, the success of this method relies on the professional experience. This thesis aimed to define the injection parameters from studying the hydraulic characteristics of a fracture in concrete fracture during injections. This main objective can be divided in three subobjectives. Initially, it was necessary to develop multiple methods to establish the different apertures of a concrete fracture. Secondly, the influence of the injection parameters and the fracture properties on the resin propagation were defined. Thirdly, a method was developed to establish the injection parameters from the fracture hydraulic properties.
The first step was to study the flow when injections were performed in concrete physical models. Water-glycerol mixtures were injected by varying the injection parameters. The results showed that the dynamic viscosity had an important influence on the liquid propagation for dry fractures. For moist fractures, the liquid propagation was irregular and partially guided by the dynamic viscosity. The boundary conditions and the initial fracture saturation level (dry or moist) influenced the duration needed to reach fracture saturation. With pressure gauges installed in the fracture, it was possible to establish the duration needed to reach fracture saturation. Important head loss was noted the pump outlet and the injection point in the model. It would have been an error to assume that the entry pressure in the model is the same as the pump outlet pressure. The head loss between both point is influenced by the tubing diameter. A numerical model was developed with COMSOL Multiphysics software to validate the results obtained with the physical models and establish the effect of the aperture and roughness on the injection pressure. The results showed that the aperture has a higher effect on the injection pressure than the roughness.
The second step was to enhance the numerical model built on COMSOL Multiphysics with the implementation of multiple types of fracture aperture and a time-dependent dynamic viscosity. Three types of apertures were used: Gaussian spatially correlated aperture, Gaussian nonspatially correlated aperture and a constant aperture. Gaussian spatially correlated aperture was able to simulate the propagation anisotropy. Multiple reproductions of the simulations were performed to obtain the average injection pressure for all the aperture types. Despite the fact the average pressure was the same for all the aperture type, an important variation was observed between the highest and lowest injection pressure for the Gaussian spatially correlated aperture. This variation can be explained by the aperture distribution around the injection point. A time-dependent dynamic viscosity was implemented to establish its effect on the grout flow when the polymerization began. When the liquid polymerization was considered, the results showed a significant increase of the injection pressure and the head loss during flow fracture while the liquid propagation decreased.
The last step was to conceive and realise an in-situ test for the hydraulic characterization of a concrete fracture and established the parameters to effectively perform a chemical grout injection in a construction joint of an experimental site. With pressure gauges installed in the joint and water-glycerol mixtures, it was possible to establish the hydraulic aperture and hydraulic conductivity of the studied fracture area of the joint. The injection volume and duration needed to seal the studied fracture area with chemical grout were determined from the in-situ test. A chemical grout injection was performed based on the in-situ test results. Core sampling was performed to validate the full propagation of the grout in the studied fracture area and establish the mechanical aperture and its distribution in space with X-ray microtomography (CT scan).
To conclude, this thesis allowed to define the components of the injection protocol that influence the resin flow in a concrete fracture. The dynamic viscosity, the fracture aperture, the injection equipment, the boundary conditions, and the initial saturation state influenced the pressure distribution in time, the head loss, and the resin propagation. Those components need to be considered when an injection protocol is developed to ensure a fully sealed fracture.
A total of three journal papers with pairs revision and three conference papers has been produced. This thesis contains the journal papers and one conference paper.
| Date | 9 Apr 2024 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | François Duhaime (Supervisor) |
|---|
El Mekari, K. (Author),
Duhaime (Supervisor),
9 Apr 2024Student thesis: Doctoral thesis › Doctorate in Engineering: Engineering