Investigation of Energy Dissipation Rate of Stepped Vertical Overfall (SVO) Spillway Using Physical Modeling and Soft Computing Techniques

• Published in: KSCE journal of civil engineering Vol. 26; no. 12; pp. 5067 - 5081
• Main Authors: Baharvand, Saman, Rezaei, Reza, Talebbeydokhti, Nasser, Nasiri, Roohollah, Amiri, Seyed Mehrab
• Format: Journal Article
• Language: English
• Published: Seoul Korean Society of Civil Engineers 01.12.2022; Springer Nature B.V
• Subjects: Algorithms; Civil Engineering; Critical depth; Dimensional analysis; Dimensionless numbers; Energy dissipation; Energy exchange; Engineering; Free overfall; Froude number; Geometry; Geotechnical Engineering & Applied Earth Sciences; Hydraulic design; Hydraulic Engineering; Industrial Pollution Prevention; Parameters; Regression models; Soft computing; Spillways; Support vector machines; Training; Water depth; Water wheels; Physical modeling; Support vector regression; Energy dissipation; Stepped spillway; Random forest regression
• ISSN: 1226-7988; 1976-3808
• DOI: 10.1007/s12205-022-1870-y

In the present study a physical model of a stepped vertical overfall (SVO) spillway is proposed and designed as a novel combination of a free overfall spillway with horizontal steps. First, the hydraulic design characteristics of the proposed spillway were discussed using a laboratory-scaled model. Effective parameters on the energy dissipation rate were defined as the relative critical depth, Froude number, number of steps, and dimensionless steps’ geometry parameter using dimensional analysis. The energy dissipation rate of the stepped vertical overfall spillway is measured using a waterwheel laboratory setup. Different geometry and hydraulic scenarios were used to assess the energy dissipation rate variation of the proposed spillway. Furthermore, Support Vector Regression and Random Forest Regression methods were used to estimate the energy dissipation of the proposed structure. Investigating the energy dissipation rate of 27 geometry scenarios with the available range of discharge revealed that the energy dissipation rate against the water’s relative depth inside the SVO spillway follows a gradually increasing trend ranging between 88.53% to 98.06%. Also, random forest regression algorithm showed more accurate prediction performance than support vector regression approach with RMSE = 0.128 and R 2 = 0.99 in training stage and RMSE = 0.115 and R 2 = 0.99 in testing stage. The support vector regression model estimated the proposed spillway’s energy dissipation rate with an accuracy of RMSE = 0.67 and R 2 = 0.88 in training stage and RMSE = 0.61 and R 2 = 0.9 in testing stage.