Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Civil and Environmental Engineering

Committee Chair

Parola, Arthur C.

Author's Keywords

Bedload transport; Impact sensor; Indirect method; Particle size distribution; Calibration; Stream restoration


Particle size determination; Stream restoration


A multi-channel Piezoelectric Bedload Impact Sensor (PBIS) is developed to estimate mass and particle-size distribution of bedload in low to moderate slope natural streams. The PBIS’s stand-alone design with a sufficiently large data memory facilitates continuous long-term monitoring for low-scale bedload measurements. The design concept of PBIS is based on a hypothesis that particle collision energy on the PBIS plate increases with hydraulic energy and particle size. Thus, particle size can be differentiated by the number of impulses registered in four different threshold channels. The feasibility of PBIS was evaluated by developing a calibration model based on laboratory flume experiments. Two different types of experiment were conducted: (1) individual particle experiment and (2) multi-size particle experiment. The individual particle experiment results indicated that hydraulic condition affected the mode of bedload particle motion. The mean impulse rate per particle is expressed as a function of bed shear stress, t. In addition, the represented particle sieve size range of each threshold channel was determined based on the fractional impulse ratio per unit mass by the particle sieve size class. The multi-size particle experiment results indicated that multi-particle interaction and signal interference from consecutive particle impacts on the PBIS plate caused a notable reduction of registered impulses in channel 1 from approximately 700 impulses per minute. The bias between individual and multi-size particle experiment results caused by multi-particle effects were expressed by a function of bed shear stress, t, and mean impulse rate, Rj, in each PBIS threshold channel. The adjusted calibration coefficient is a coefficient of linear equation to convert registered impulses to mass of particles retained in the represented particle sieve size classes of each channel. The adjusted calibration coefficients for each channel were estimated using the two-dimensional response surface methodology (RSM) with two variables, bed shear stress, t, and mean impulse rate, Rj. This study was enough to show the feasibility of the multi-channel PBIS to obtain mass and particle-size distribution of small gravel bedload. However, many issues associated with the calibration model are still remained beyond this study. First of all, the calibration model was developed based on laboratory flume experiments conducted in a well-controlled small-scale environments. Second, from the comparison between the actual and the estimated values, it was found that two inherent error factors which can cause overestimate are imbedded in the linear calibration model of PBIS. Third, two major assumptions for the model, an equal fractional bedload particle-size distribution and the law of large number, always have a distinction from chaotic phenomena in natural bedload transport. Because of that, the most preferential request will be the field application.