ANALYSIS OF HYPORHEIC FLOW INDUCED BY A BAR IN A GRAVEL STREAM: AN EXPERIMENTAL STUDY

Pipelines are one of the safest means to transport crude oil, but are not spill-free. This is of concern in North America, due to the large volumes of crude oil shipped by Canadian producers and the lengthy network of pipelines. Each pipeline crosses many rivers, supporting a wide variety of human activities, and rich aquatic life. However, there is a knowledge gap on the risks of contamination of river beds due to oil spills. This thesis addresses this knowledge gap by focussing on mechanisms that transport water (and contaminants) from the free surface flow to the bed sediments, and vice-versa. The work focuses on gravel rivers, in which bed sediments are sufficiently permeable that pressure gradients caused by the interactions of flow with topographic elements (gravel bars), or changes in direction induce exchanges of water between the free surface flow and the bed, known as hyporheic flows. The objectives of the thesis are: to present a new method to visualize and quantify hyporheic flows in laboratory experiments; to conduct a novel series of experiments on hyporheic flow induced by a gravel bar under different free surface flows. The new method to quantify hyporheic flows rests on injections of a solution of dye and water. The method yielded accurate flow lines, and reasonable estimates of the hyporheic flow velocities. The present series of experiments was carried out in a 11 m long, 0.39 m wide, and 0.41 m deep tilting flume. The gravel had a mean particle size of 7.7 mm. Different free surface flows were imposed by changing the flume slope and flow depth. Measured hyporheic flows were turbulent. Smaller free surface flow depths resulted in stronger hyporheic flows (higher velocities, and deeper dye penetration into the sediment). A significant finding is that different free surface flows (different velocities, Reynolds number, etc.) produce similar hyporheic flows as long as the downstream hydraulic gradients are similar. This suggests, that for a specified bar geometry, the characteristics of the hyporheic flows depend on the downstream hydraulic gradients, and not or only minimally on the internal dynamics of the free surface flow.

Monitoring of Dynamic Bridge Behaviour Using Digital Image Correlation

Bridges are a critical part of North America’s transportation network that need to be assessed frequently to inform bridge management decision making. Visual inspections are usually implemented for this purpose, during which inspectors must observe and report any excess displacements or vibrations. Unfortunately, these visual inspections are subjective and often highly variable and so a monitoring technology that can provide quantitative measurements to supplement inspections is needed. Digital Image Correlation (DIC) is a novel monitoring technology that uses digital images to measure displacement fields without any contact with the bridge. In this research, DIC and accelerometers were used to investigate the dynamic response of a railway bridge reported to experience large lateral displacements. Displacements were estimated using accelerometer measurements and were compared to DIC measurements. It was shown that accelerometers can provide reasonable estimates of displacement for zero-mean lateral displacements. By comparing measurements in the girder and in the piers, it was shown that for the bridge monitored, the large lateral displacements originated in the steel casting bearings positioned above the piers, and not in the piers themselves. The use of DIC for evaluating the effectiveness of rehabilitation of the LaSalle Causeway lift bridge in Kingston, Ontario was also investigated. Vertical displacements were measured at midspan and at the lifting end of the bridge during a static test and under dynamic live loading. The bridge displacements were well within the operating limits, however a gap at the lifting end of the bridge was identified. Rehabilitation of the bridge was conducted and by comparing measurements before and after rehabilitation, it was shown that the gap was successfully closed. Finally, DIC was used to monitor the midspan vertical and lateral displacements in a monitoring campaign of five steel rail bridges. DIC was also used to evaluate the effectiveness of structural rehabilitation of the lateral bracing of a bridge. Simple finite element models are developed using DIC measurements of displacement. Several lessons learned throughout this monitoring campaign are discussed in the hope of aiding future researchers.

Estimation of Sample Size and Power For Quantile Regression

Quantile regression (QR) was first introduced by Roger Koenker and Gilbert Bassett in 1978. It is robust to outliers which affect least squares estimator on a large scale in linear regression. Instead of modeling mean of the response, QR provides an alternative way to model the relationship between quantiles of the response and covariates. Therefore, QR can be widely used to solve problems in econometrics, environmental sciences and health sciences. Sample size is an important factor in the planning stage of experimental design and observational studies. In ordinary linear regression, sample size may be determined based on either precision analysis or power analysis with closed form formulas. There are also methods that calculate sample size based on precision analysis for QR like C.Jennen-Steinmetz and S.Wellek (2005). A method to estimate sample size for QR based on power analysis was proposed by Shao and Wang (2009). In this paper, a new method is proposed to calculate sample size based on power analysis under hypothesis test of covariate effects. Even though error distribution assumption is not necessary for QR analysis itself, researchers have to make assumptions of error distribution and covariate structure in the planning stage of a study to obtain a reasonable estimate of sample size. In this project, both parametric and nonparametric methods are provided to estimate error distribution. Since the method proposed can be implemented in R, user is able to choose either parametric distribution or nonparametric kernel density estimation for error distribution. User also needs to specify the covariate structure and effect size to carry out sample size and power calculation. The performance of the method proposed is further evaluated using numerical simulation. The results suggest that the sample sizes obtained from our method provide empirical powers that are closed to the nominal power level, for example, 80%.