Evaluation of critical submergence for horizontal intakes derived from the channel

Horizontal intakes are one of the most important parts of hydraulic sets such as river for irrigation or reservoir for power generation and industrial purposes. Air entrainment by means of a free air-core vortex occurring at intake pipes is an important problem encountered in hydraulic engineering. Formation of the vortices in front of intake is the result of complex interaction between many parameters and cause operational problems for turbine or pump and reduction of coefficient of discharge. Intake submergence depth could result in formation of the vortices. Study of previous research showed in the most of the equations that proposed by researchers were not included the effect of distance from the bottom of canal C in the equations and this equations was presented only in two state C=0 and C=Di/2 as two separate equations. In this study, the equation for estimating critical submergence are developed using experimental data. At first, the equation of present study was determined using dimensional analysis (Buckingham theory) and nonlinear regression and in the next step the Artificial Neural Network model and the Genetic Programming model was used for checking the accuracy of the results. At first, the overcoming equation of space research using the theory of dimensional analysis was defined as Sc/Di=f(C/Di, Vi/U, Fr). Then this equation was determined by nonlinear regression and SPSS software. The proposed equation includes the effect of vertical distance of intake to bottom of canal, velocity and Froud number. In this equation, the value of RMSE and are 0.3165 and 0.9363 respectively. Ahmed et al. (2008) and Ayoubloo et al (2011) Research was used to validate this equation. All the results was compared, Ahmed et al research in compared with experimental results predicted the depth of critical submergence 8% more, Ayoubloo et al research 6.5% less and the proposed equation of this research predicted the value 0.5% more. However, The Artificial Neural Network in compared with experimental results predicted the depth of critical submergence 1.1% less, while the Genetic Programming model estimates the depth of critical submergence 1.63% more. Compare the results of Artificial Neural Network and Genetic Programming was showed that the error functions of both of them are superior than the proposed equation. while the Genetic Programming model estimates the depth of critical submergence 1.63% more. Compare the results of Artificial Neural Network and Genetic Programming was showed that the error functions of both of them are superior than the proposed equation.