view:18448 Last Update: 2021-3-1
Masoud Akbari, Farzin Salmasi, Hadi Arvenaghi, Masoud Karbasi, Davood Farsadi Zade
Effective Geometrical parameters on discharge coefficient of gated piano key weir
پارامترهای هندسی موثر بر ضریب دبی سرریز کلید پیانویی دریچهدار
The Piano Key (PK) weir is a new type of a long crested weir that improves the capacity of discharge through the constant upstream head by increasing the length of the crest. Due to the resulting high discharge capacity of these types of weir, reservoir capacity is increased and dams are stabilised against flooding. The PK weir has been developed by Lempérière, HydroCoop, France and Blanc, Hydraulic and Environmental Laboratory, University of Biskra, Algeria. The first PK weir was built in 2006 on the Goulours dam in France, with the second being installed in 2008 on Saint-Marc dam in France. Due to the geometrical complexity of these weirs, many studies have been conducted on them in recent years. A review of pertinent literature shows that previous studies regarding PK weirs have not investigated the effect of an additional gate on each of the inlet keys. This study is the first to examine such an effect. It also investigates other parameters of PK weirs, using multiple linear and non-linear regressions methods and with suggesting the relation for discharge coefficient (Cd). Thus, this study is distinguished from other studies in this area. For first time, the adding of a gate in the inlet keys of PK weir with the hope of improving the hydraulics performance of this type of weir was investigated in the hydraulic laboratory, Department of Water Engineering, University of Tabriz, Iran. The experiments were conducted on a 10 m long and 0.933 m wide horizontal rectangular flume. The Gated Piano Key (GPK) weir was installed on a 1 cm high ramp. In this study, a four-cycle GPK weir was used. The water circulation system included an underground tank equipped with a 100 hp. pump which provided a steady supply of water into a head tank. The upstream tank (head tank) provided water flow into the flume. The weir returned extra water from the head tank into the underground tank, thereby ensuring the discharge in the rotation system would be a fixed amount. The water flowed downstream through the flume, was collected in the collection tank, and then flowed into the underground tank to be pumped again into the head tank. The discharge rate was measured using an ultrasonic flow meter with 1% precision. Discharge measuring sensors were installed on a 10 inch pipe which supplied water to the flume. The discharge ranges were between 10 and 130 liters per second. In order to reduce the turbulence of the input water flow from the pipe into the laboratory flume, the water travelled across a porous medium space created by sequential lattice plates placed along the first three meters of the laboratory flume. By placing the Styrofoam on the water surface oscillations were prevented and in this way the height of water was stabilized in the upstream of the GPK weir. The water depth was measured by a point gauge moveable along the flume, with a precision of 0.1 mm. After dimensional analysis for diagnosing the effective parameters in the studied phenomenon and determining the dimensionless ratios, the effective parameters on the GPK weir discharge coefficient (Cd), such as gate dimensions (b and d), gate insertion depth in the inlet key (Hgate), the ratio of the inlet key width to the outlet key width (Wi/Wo) and the head over the GPK weir crest (H) were investigated. From 188 experiments for both PK and GPK weirs, 156 data set was relevant to the GPK weir. In addition, two relationships were presented for the GPK weir discharge coefficient using multiple linear and non-linear regressions models in MATLAB 2018 software. From this data, 70% and 30% were used as the training data and model testing data, respectively. Training and test data were used for training process of models and evaluation of models accuracy, respectively. The results of this study showed that H, Wi/Wo, Hgate and b and d, had the greatest effect on the GPK weir discharge coefficient, respectively. To investigate the accuracy of models suggested in this study, the Statistical Criteria have been used. The nonlinear regression method with RMSE = 0.035, R2 = 0.961 and MARE = 4.562% yielded better results than the linear regression method with RMSE = 0.061, R2 = 0.865, and MARE = 9.49%.