Mentor: Claudia Cherubini
Project Team: Laura Beckingham, Rory O’Keefe
A number of field transport experiments have been performed in unsaturated fractured rocks, but the interpretation of field measurements is often problematic because detailed characterization of the subsurface is difficult. Data from laboratory experiments are generally easier to interpret than data from the field since conditions are controlled and the system is easier to characterize. In addition, laboratory experiments integrate field studies by furthering the understanding of smaller scale mechanisms, which may affect processes at a larger scale.
Laboratory studies of fluid flow and solute transport in unsaturated fractures have been very limited. In particular, very few studies exist on the analysis of the influence of fracture parameters on the infiltration in unsaturated flow conditions. Thus, a primary goal has been to set up an experimental apparatus at the Hydraulic Laboratory of School of Civil Engineering to analyse unsaturated flow in a single synthetic fracture as a function of aperture, inclination and roughness.
By means of a program such as SYNFRAC different fracture planes have been generated having different parameters. A 3D printer has been used to print these profiles on two plastic sheets which simulate the two fracture planes. A range of different angles, apertures and type of roughness are currently been used to investigate their influence on flow in unsaturated conditions. Water flows through the fracture by means of a hydraulic system composed by an upstream and a downstream tank at variable and constant hydraulic head. A graduated ruler is positioned in the upstream tank to measure hydraulic head.
The downstream tank is divided into different sections in order to measure the hydraulic head variation in different sections by means of graduated rulers to be able to detect zones of preferential film flow and tensiometers are placed in nine different positions to monitor changes in water content along the fracture.
A range of different fracture inclinations, apertures and types of roughness are being used in order to investigate the influence of those parameters on flow in unsaturated conditions. The observed experimental data will be compared with the ones simulated numerically using different mathematical models in order to evaluate and test alternative conceptual models of hydrogeological processes that describe recharge phenomena in unsaturated fractures. The project consists in carrying out further tests and collecting experimental data for conceptual and mathematical modelling.
The fracture flow infiltration experiments are aimed to determine how the pattern of water flow through fractures change with different fracture parameters such as saturation, aperture, roughness and inclination. By means of the Synfrac program, 15 fracture planes were created having different aperture and roughness’s. Then by means of a 3D printer, plastic samples were created that could then be moulded out of concrete for experimentation.
Before testing could be commenced, the experimental set up had to be altered based on the feedback from previous tests using the equipment. This included changing the water input method from a water reservoir directly at the fracture entrance to five drippers connected to a head tank inserted into the fracture at equally spaced locations. This allowed for a more realistic flow of water through the fracture that would be more representative of what would be encountered in the natural environment.
The experiments consists of analysing flow conditions in each different sample by varying the inclination and the degree of saturation. Water flow was from an upstream tank to a downstream tank that was divided into 5 sections according to the position of the drippers. This configuration allowed the detection of preferential flow conditions.
The experiments for three samples were completed for dry and fully saturated conditions at angles of 75°, 65°, 55° and 45° from the horizontal. The results have not yet been an analysed to determine trends in the data, however this is the next step in the project. More experiments are also planned for a remoulded sample, as the original mould was too poor a quality for the results to be reliable.