Self-cleansing drainage system design by incipient motion and incipient deposition-based models

Abstract

In sediment transport studies it is aimed to determine flow and channel properties which prevent deposition of sediment particles. Continuous deposition of solid particles in the flow changes channel cross-section. It, hence, causes change in the velocity and wall shear stress distribution in rigid boundary channels, and significantly affects the carrying capacity and hydraulic resistance of the channel. Additional costs arise for cleaning a channel under the effect of sediment deposition. Also, deposition of sediment in urban drainage channels and sewer systems causes sediment to be contaminated with toxic substances and it is important to keep in mind that pollutants are transported with the sediment deposited. Self-cleansing is a substantial aspect of drainage system design. It is desired to minimize the sediment deposition problem in the drainage systems. In the conventional self-cleansing design criteria, a single value of velocity or shear stress is used based on experience. In the recent decades rather than using a single value, this concept was further modified to develop self-cleansing models based on higher number of parameters considering flow, fluid, sediment and channel characteristics. However, self-cleansing models were mostly developed for circular channels, and none of the models considers the effect of the channel cross-section although some models take the effect of geometry of the cross-section into account by using pipe diameter, hydraulic radius and cross-section area as independent variables. In this study, experimental studies of sediment transport in rigid boundary channels are briefly reviewed. Available self-cleansing models in the literature are classified into two main groups of "bed sediment motion" and "non-deposition". It is found that there is no study in the sediment transport literature that cover the conditions of incipient motion, incipient deposition and non-deposition all together. In this study, experiments were conducted in five different cross-section channels namely; trapezoidal, rectangular, circular, U-shape and V-bottom in order to investigate incipient motion and incipient deposition of sediment in rigid boundary channels. In experiments four sand sizes were used as sediment. Experiments were performed considering the sediment motion cycle composed of decreasing velocity and increasing velocity half cycles. Within the experiments, non-deposition, incipient deposition, deposition and incipient motion conditions of sediment motion were examined. 180 experiments and accordingly 2609 observations were carried out for five different cross-section channels. Theoretical consideration of sediment threshold is performed considering shear stress and velocity approaches. The Shields and Yalin criteria are used for the shear stress approach, while the Yang, and Novak and Nalluri methods are considered for the velocity approach into analysis of experimental data. Available models in literature are evaluated and new models for incipient motion and incipient deposition of sediment in rigid boundary channels are developed in this study. The difference between incipient motion in loose and rigid boundary channels and also the effect of channel cross-section on the threshold condition are discussed herein. Analysis of incipient motion experimental data based on shear stress and velocity approaches and comparing the results with Shields, Yalin and Yang curves indicates that boundary dissimilarity of loose and rigid boundary channels is expected to affect the mobilization of sediment particles. In loose boundary channels flow needs higher shear stress and velocity to mobilize the particles in the bed. Analysis of incipient deposition experimental data based on shear stress and velocity approaches demonstrates that rectangular channel needs lower incipient deposition shear stress and velocity in comparison with other channels while U-shape and V-bottom channels need higher. However, rectangular channel has a different performance than other cross-sections that required lower shear stress and velocity. This indicates that rectangular channel has a higher efficiency of sediment transport as sediment particles deposit in lower velocities. A comparison of results for incipient motion and incipient deposition data shows that flow characteristics such as velocity, the dimensionless shear stress and the particle Froude number are higher in case of incipient deposition than incipient motion under the same hydraulic conditions. It is concluded that a critical range between the lower boundary-incipient motion and the upper boundary-incipient deposition can be defined as the threshold region of sediment motion in which sediment particles in the motion tend to deposit and stationary particles tend to move. Self-cleansing models are developed in this study for different cross-section channels. In order to make a general practical tool, a self-cleansing model considering the effect of cross-section by a shape factor is developed to determine the non-deposition particle Froude number for bed load sediment transport. The model is developed using experimental data of five different cross-section channels namely; trapezoidal, rectangular, circular, U-shape and V-bottom. For validation of the model, five sets of data available in literature are used. Validation of the model for each cross-section data taken from the literature shows its applicability that is independent of channel cross-section.