| description abstract | A postprocessing method has been developed to enable the extraction of quantifiable data from images captured from within the rotating frame of reference of a Pelton turbine. The turbine tested was the reference Pelton runner, designed at the Waterpower Laboratory, Norwegian University of Science and Technology (NTNU). The method relies on interpolation to three-dimensional (3D) map the inner hydraulic surface of the bucket. Interpolation has been conducted with two different schemes, i.e., Barycentric triangular and biharmonic spline, where the latter showed significant increase in accuracy. The 3D mapping provides the world coordinates of the pixels within the bucket and enables the tracking of the water front as it propagates through the bucket. The method has been described and the uncertainties have been estimated in the order of 0.4 mm for most of the hydraulic surface. The results follow expected, and previously observed, behavior and show great promise with regard to validation of numerical simulations. Results obtained by the method will be of great interest for the computational fluid dynamics (CFD) community as it can be used as direct validation data for flow propagation found with numerical methods. The method relies heavily on manual input due to the high noise and low contrast of the available images, which causes an increase in both uncertainty and time consumption. Suggestion for reducing uncertainty and time consumption are presented and will be implemented in future publications. | |
