A Systematic Validation of a Francis Turbine Under Design and Off-Design Loads

Abstract

Computational fluid dynamic (CFD) techniques have played a significant role in improving the efficiency of the hydraulic turbines. To achieve safe and reliable design, numerical results should be trustworthy and free from any suspicion. Proper verification and validation (V&V) are vital to obtain credible results. In this work, first we present verification of a numerical model, Francis turbine, using different approaches to ensure minimum discretization errors and proper convergence. Then, we present detailed validation of the numerical model. Two operating conditions, best efficiency point (BEP) (100% load) and part load (67.2% load), are selected for the study. Turbine head, power, efficiency, and local pressure are used for validation. The pressure data are validated in time- and frequency-domains at sensitive locations in the turbine. We also investigated the different boundary conditions, turbulence intensity, and time-steps. The results showed that, while assessing the convergence history, convergence of local pressure/velocity in the turbine is important in addition to the mass and momentum parameters. Furthermore, error in hydraulic efficiency can be misleading, and effort should make to determine the errors in torque, head, and flow rate separately. The total error is 9.82% at critical locations in the turbine. The paper describes a customized V&V approach for the turbines that will help users to determine total error and to establish credibility of numerical models within hydraulic turbines.