Comparison Between 6-DOF UDF and Moving Mesh Approaches in CFD Methods for Predicting Cross-Flow PicoHydro Turbine Performance

Abstract

Many studies have been conducted on improving the performance of small-scale (micro-hydro and pico-hydro) cross-flow turbines using computational fluid dynamics (CFD) methods. In parallel with the development of technology, CFD methods are becoming increasingly sophisticated and are getting closer to real conditions in the field, with various developing features, progressing from the steady-state, transients with moving references, and moving mesh approaches, to using user-defined functions (UDF). This study compared the moving mesh method with the six degrees of freedom (6-DOF) UDF method for simulation of cross-flow turbines at the pico scale. The two simulations were treated as similarly as possible, outside of the dynamic approach, and were run in the 2D domain with 9.5 m of inlet head, using shear stress transport turbulence modelling. 6-DOF was found to have a much smaller deviation from experimental results, of about 6.8%, than the moving mesh results, at about 12.4% deviation. The deviation in the moving mesh results was assessed to be mainly caused by overcalculated turbulence conditions inside the rotating domain in the moving mesh calculation. The overcalculated turbulence condition makes the prediction of second stage energy absorption too low. The 6-DOF method is therefore more accurate than the moving mesh method for predicting the performance of cross-flow turbines at the pico scale. However, the moving mesh method is still an alternative because it has some advantages in terms of effectiveness of resource usage and rate of convergence.