Reconciling Predicted Hydro Turbine Results To Test Stand Performance

A new hydropower design was tested by the customer, with actual test stand results significantly lower than the customer’s predicted performance.  MSI was tasked with resolving the discrepancy, either by determining the flaw in the testing apparatus/procedure, or by correcting the computer model to better represent actual performance.

Two computational fluid dynamics (CFD) programs were used, StarCCM+ and ANSYS CFX, to compare results both to each other and to the customer’s CFD and test stand results.  MSI’s analyses included an assessment of the torque produced by the runner on all rotating surfaces, and the head loss and head loss coefficients for all flowpath components.  Each model consisted of 5.7 million nodes and 13 million elements.  Due to MSI’s considerable computational resources, the results were quickly generated and available for analysis and comparison.

The hydro turbine test setup was reviewed for any configuration issues or inconsistencies with the modeling, and found to be quite satisfactory.

MSI’s dual CFD results both correlated extremely well with the test results, prompting additional consideration of the customer’s CFD analysis.  Further review of the customer’s CFD model revealed the flowpath boundary was too constrained, and did not account for the drag created by a major rotating surface.  With this correction made, MSI was able to reconcile the results and suggest modifications to the design to significantly improve actual performance.

For additional details about this study, please read the Hydrovision paper on the topic here.