sensitivity analysis of influence factors on discharge coefficient of short throat shaft spillway by utilizing artificial neural network.

The shaft spillway is simply a closed conduit in which the flood flow is carried rapidly from a high to a low elevation. This spillway can be used advantageously at dam sites in narrow canyons where the abutments rise steeply or where a diversion tunnel or conduit is available as the downstream leg. Another advantage of this type of spillway is that near-maximum capacity is attained at relatively low heads. This characteristic makes the spillway ideal for use where the maximum spillway outflow is limited. In this study by utilizing nearly 300 data that comes from Laboratory tests of different conditions of morning glory spillway, hydraulic parameters such as Froude number, submergence ratio, dimensionless parameters like effective of vortex breaker shape and number of vortex breaker received, then by use of these parameters accurate equation for discharge coefficient received. Using these parameters in the artificial neural network defines each parameter's impact and importance. The criterion for choosing the most important factor in this study is a root-mean-square error (RMSE). The results show that the number of vortex breakers has more effect on the discharge coefficient, and by omitting this parameter, a maximum error (2.154) occurred.