Zonation of cavitation hazard in the chute spillway of Surk dam with Nearest Neighbor Classification Algorithm

Cavitation is one of the failure factors of spillways, which requires risk zoning to control this phenomenon. In this research, to obtain a method for zoning the risk of cavitation, the spillway information of Surk dam in Chaharmahal Bakhtiari province  was used. In the modeling process, first, the geometric model of the overflow was constructed and after meshing and applying boundary conditions, flow analysis was done. The cavitation index was calculated in 18 sections according to the values of flow velocity and height, chute slope, and other necessary parameters. The results of Flow-3D software for qualitative assessment of the cavitation risk situation in Surk dam spillway are of appropriate accuracy; Thus, the RMSE error of pressure 0.26×10-2 pascal and velocity 0.23×10-2 m/s was obtained compared to the laboratory results. Also, parameters affecting the reduction of cavitation such as roughness and aeration were investigated. The results showed that there is a possibility of cavitation and damage caused at a distance of 70 to 95 meters from the crest of spillway. The results of the sensitivity analysis showed that the use of a uniform roughness of 2.5 mm and aeration during the chute increases the cavitation index. This roughness moves the cavitation areas to the downstream sections of the spillway. Also, by creating a roughness of 1.5 mm in two end sections 99.75 and 105  meters from  the crest of spillway, the results of the nearest neighbor algorithm (NNA) showed a more critical state than the Flow-3D model. By applying a roughness of 2.5 mm, in the two end sections of 42 and 89.25 meters from the crest of spillway, the NNA showed a more critical state than the Flow-3D model, which means that these areas are more vulnerable to the cavitation phenomenon.