EXPERIMENTAL INVESTIGATION OF FLOW OVER TRENCH WEIR WITH FLAT BARS

Abstract

Boulder streams carry a large amount of sediments which usually ranges from sands and gravels to boulders and even large rocks, and it can include trees and other debris. The sediment load during floods is boosted by the landslides. Conventional types of raised-crest weirs for diverting water from the boulder streams for its use in generation of hydropower, irrigation and water supply schemes etc. are not well suited as the afflux brings a remarkable change in the flow conditions. The most suitable weir adopted in such scenario is the trench weir, which is simply a trench built across the stream below its bed level. The top of the weir is covered with bottom rack bars. Water while flowing over it, passes through the bottom racks and enters into the trench and being collected in an intake well located at either of the banks at the end of the weir. This type of weir has a definite advantage as it does not affect the general bed level of the stream. However,, due to being below the bed of the river, the, bed sediment load of size less than the clear spacing of the rack bars enters into the trench. Thus post-monsoon clearance of the trench is obligatory. The hydraulic design of a trench weir deals with fixing the dimensions of the bottom rack, trench, intake well, flushing arrangement, protection works upstream and downstream of the trench etc. Bottom racks are generally made of bars placed either parallel or perpendicular to the flow.. Mostkow (1957) analysed the flow over longitudinal bottom rack of circular bars by assuming the specific energy of flow to be constant all over the rack and expressed the diverted discharge into the trench as function of specific energy. Computation of diverted discharge into the trench and water surface profile over the rack and in the trench have extensively been studied (Noseda 1956; White et al. 1972; Ranga Raju 1977; Subramanya and Sengupta 1981; Subramanya and Shukla 1988; Subramanya 1990, 1994, 1997; Brunella et al. 2003; Ahmad and Mittal 2004; Ghosh and Ahmad 2006; Drobir et al. 1981; Righetti and Lanzoni 2008). From the structural consideration, the flat bars are preferred over the round bars due to their more flexural rigidity. However, information related to the hydraulics of bottom racks consisted of flat bars is scarce. Further, effect of various parameters on the discharging capacity of bottom racks is also very little. The mechanics of movement of sediments, flowing v in the channel as bed load, over the rack is also not studied so far. Information related to the amount of sediment encroached in the trench while moving on the rack is also not available. Keeping in view these points, the present work is aimed at investigation of flow and discharge characteristics of trench weir using flat rack bars under free and submerged flow conditions through analytical and experimental considerations. This study is also intended to investigate the mechanics of movement of sediments, flowing as bed load in the approach channel, over the racks and to quantify the sediments encroached into the trench. The experiments were carried out in the Hydraulics Laboratory of the Department of CIVILENGINEERING, Indian Institute of Technology Roorkee, India. A rectangular channel of 17.00 m length, 0.50 m width and 0.64 m depth was used as the approach channel. A trench of 0.50 m long, 0.30 m wide and 0.34 m deep with 1/12.5 bed slope was provided in the bed of the channel at a distance of 10.6 m from the upstream end of the approach channel. Bottom rack consisting of flat bars was placed over the trench. The trench was opened to an intake well of size 1.00 m length, 1.00 m width and 0.91 m depth. Diversion channel of length 3.90 m, width 0.30 m and 0.72 m deep was taken from the intake well. A gate was provided at the head of diversion channel to increase the depth of water in the intake well required for the submergence study of the trench weir. At the trench, a glass panel of size 0.95 m x 0.95 m was provided in place of right brick wall to visualize the flow in the trench and over the trench. Five bottom racks consisting of mild steel flat bars of 5.00 mm thickness; 335.00 mm length; and 19.10, 25.40, 31.80, 38.10 and 50.80 mm width were used. To study the movement of sediment particle over the rack and encroachment of these particles, in trench, three sizes of sediment of mean sizes 1.67 mm, 4.0 mm and 7.1 mm were used in the experiment. The experiments were conducted to study the discharge characteristics of the trench weir for both free and submerged flow conditions. Experiments were performed for free flow conditions for bed slopes of approach channel S =1/1250, 1/ 847, 1/ 500 and 1/200; rack slopes sr = 1/15, 1/5 and 1/3; thickness of flat bars t = 5.00 mm; width of flat bars w = 19.10 mm, 25.40 mm, 31.80 mm, 38.10 mm and 50.80 mm; clear spacing of rack bars s = 2.00 mm, 5.00 mm and 9.00 mm and for each set of these parameters, five different discharges. For submerged flow conditions, different submergence was created by raising the water level in the intake well by regulating the gate provided at the head of the diversion channel. Water level in the well was measured and the submergence depth Ys, which was equal to depth of water in the intake well above the bed level of approach channel at the upstream of the trench, was calculated. Experiments were performed for S =1/1250, 1/ 847, 1/ 500 and 1/200; sr= 1/15, 1/5 vi and 1/3; t = 5.00 mm; w = 19.10 mm, 25.40 mm, 31.80 mm, 38.10 mm and 50.80 mm; s = 2.00 mm, 5.00 mm and 9.00 mm and for each set of these parameters, one maximum discharge in the approach channel at different submergences was taken. Experimental data collected in the present study were analyzed to investigate the discharge characteristics of the trench weir under free and submerged flow conditions. Various parameters affecting the discharge characteristics of trench weir under free and submerged flow conditions were identified and relevant parameters were used to propose relationships for the coefficient of discharge under different flow conditions. Flow characteristics over the rack and in the trench were also investigated. Part of the bed load of the stream, which enters into the trench is studied and relationship is developed to calculate the quantity of sediment entering into the trench. Based on the present and previous studies, guidelines for the hydraulic design of trench weir have also been discussed. This study indicates that aspect ratio of trench size, i.e., ratio of depth and width of trench affects the diverted discharge. The diverted discharge increases with increase in aspect ratio, however, as the aspect ratio approaches to unity, the diverted discharge, becomes almost constant and any further increases in aspect ratio does not affect the diverted discharge. The coefficient of diverted discharge Cd under free flow conditions for partial withdrawal decreases with an increase of rack slope with other parameters kept constant. The coefficient of diverted discharge also decreases with increase of clear spacing. For the low width (flatness) of bars, the separated flow does not rejoin the bars and enough space is available behind the bars for the formation of vortex, which results in more dissipation of energy. An increase in width of bars decreases the space for the vortex formation resulting slight decrease in energy loss and increase of Cd. However, with further increase in width, the separated flow rejoined the bars and increases the energy loss by dissipation of energy due to friction on the face of the bars, onwards, Cd decreases with increase of flatness of bars. The coefficient of diverted discharge decreases with increase of specific energy and depth of flow and does not change with Froude number and dimensionless specific energy of the approach flow in the range of data studied. However, it decreases with increase of Reynolds number of the approach flow. The discharge characteristic of rack under free flow condition for partial and complete withdrawal of discharge is found different. Under partial withdrawal, the trench is completely filled with water with vortex flow in the trench, which results in non-existence of free water surface in the trench. However, for complete withdrawal, atmospheric pressure exists on the both sides of the water jet, which results in higher value of coefficient of diverted discharge. vii Out of the total collected data in the present study, around 80% data were used for the calibration and remaining 20% data for the validation. An equation for estimating the coefficient of diverted discharge for partial withdrawal under free flow conditions relating it with rack porosity, rack slope, and flatness of bars has been proposed using the 432 data sets collected in the present study. The coefficient of determination (R2) of the proposed equation is 0.87. The remaining 166 data sets, not used in the derivation of equation, were used to validate the proposed relationship for Cd. The observed and computed values of Cd and diverted discharge using proposed equation for the test data are compared graphically which revealed that the computed Cd and diverted discharge into trench Qd have a maximum error of ±10% which is considered as satisfactory. Under complete withdrawal and free flow conditions, the coefficient of diverted discharge increases with decrease of wetted length of rack. Out of 172 data sets collected in the present study for complete withdrawal, 132 data sets were used to develop an equation for Cd for complete withdrawal relating it with Cd for partial withdrawal and dimensionless wetted rack length. The coefficient of determination (R2) is 0.87. Computed values of Cd and Qd using proposed equation are within +10% of the observed one. A procedure is described in the thesis for the computation of diverted discharge into the trench by identifying the flow condition (i.e. partial or complete). Under submerged flow condition, the trench was completely filled- with water with no air entrainment due to higher water level in the intake well. The diverted discharge decreases with an increase in the submergence. A relationship is proposed for the computation of diverted discharge under submerged flow conditions which compute computes the discharge with a maximum error of *10%. Flow pattern over rack and in the trench were also studied. Stream lines of the flow over the rack were traced out by injecting the dye upstream of the rack at various depths. Analogy to the flow over sudden drop in the channel, the depth of the flow decreases downstream due to accelerated flow. However, the flow depth at the end of the racks was more due to curvature effect of the streamlines. The curvature of the streamlines indicates that the diverted discharge per unit length in the upper part of the rack is more and decreases along the rack. Water surface profiles were also measured for different flow conditions. Brunella et al.'s (2003) equation overestimates the water surface profile over the rack. The computed water surface profile based on constant specific energy compares well with observed ones; however, this is limited to the water surface profile over the rack. Measurements of three-dimensional velocities at the viii outlet of the trench indicate vortex flow in the trench with more concentration of flow near the boundary of the trench. Loss of energy of flow over the rack, i.e., difference of upstream energy head and downstream energy head of the rack is also studied in this thesis. Analysis of data reveals that for free flow condition, there is practically no loss of energy of the flow over the rack. However, as the flow changes from partial to complete withdrawal condition, the flow over the rack strikes the downstream edge of the rack, which results in formation of hydraulic jump and creation of turbulence which are associated with loss of energy. The experimental data collected in the present study were analyzed to study the movement of the sediment particle over the bottom rack and also to quantify the sediment encroached into the trench. The path of movement of the sediment particles of the different sizes over the rack were traced out by capturing the moving particle by high resolution camera. The jump length for large size sediments over the rack was more than the small size sediments due to its high initial velocity compared to the small size particle velocity. The trajectory of the sediments movement and stream lines of flow over the rack were also different. For the same geometrical and flow parameters, a sediment particle traveled more than water particle due to its high inertia. Effect of opened length of the rack on the jump length of three sizes of sediment was also studied by varying the opened length of the rack. It is found that the jump length of the sediment particle over a bottom rack is independent of the rack length subjected the rack length is more than the jump length. If the rack length is kept equal to the jump length of small size sediment, the jump lengths of the three sizes of sediments get increased. Due to the increase in jump length of the sediment particles, they jump beyond the rack opening without getting encroached in the trench. It reveals that amount of sediment entering into the trench can be reduced by providing rack length less than the jump length of the sediment. Data collected in the present study related to sediment entering into the trench for rack length more than the jump length of sediment over the rack were analyzed and a relationship is proposed to estimate the percentage of incoming sediment in the approach channel encroached into the trench, (i.e. (quantity of incoming sediments in the approach channel — quantity of sediments collected at the end of approach channel) x 100 / quantity of incoming sediments in the approach channel)). For ratio of size of sediment and clear spacing of rack less than 0.75; the total incoming sediment is encroached into the trench. However, for the ratio equal or greater than 0.75, the percentage encroached sediment into the trench depends on the diverted discharge. The proposed equation can be used to estimate the percentage of encroached ix sediment into the trench, which computes the percentage encroached sediment within ±3% of the observed ones. Based on the present and previous studies, a guideline for the hydraulic design of trench weir has been presented, which can be used to fix the various dimensions of the trench weir. x