Abstract
When cross-flow turbines are deployed in a vertical orientation, designs have considered a cantilevered rotor - with bearings and seals that can withstand the moment from the loads - or a support structure (a “cage frame”) that allows smaller bearings to be positioned above and below the turbine rotor. One advantage of the latter arrangement is its compatibility with water lubricated bearings which eliminates the cost of seals as well as their associated parasitic losses. Although structurally advantageous, the hydrodynamic effects of the cage frame on turbine performance are unknown, making it difficult to evaluate the costs and benefits of this arrangement.
We conducted an investigation into the performance effects from a cage frame using scale-model experiments in a laboratory flume. We began with the case of a single-bladed turbine and one cylindrical support column positioned 1.5 cylinder diameters outside of the blade sweep. By varying the azimuthal position of the support column and collecting data across a range of tip-speed ratios, we identified phase-dependent, as well as time-averaged, variations in blade-level performance. Initial results (see figure) show that the time-averaged coefficient of performance, or efficiency, significantly decreases if the support column is positioned directly upstream, as well as the subsequent 20 degrees of the blade sweep. However, relative to the case without the support column, the performance unexpectedly increases if the support column is in the first half of the upstream sweep (0-90 degrees in the figure). Given the asymmetrical hydrodynamics that occur in the upstream sweep, it is important to highlight the asymmetrical trends in performance for positions within this range, although attributing these trends to certain blade-level hydrodynamics requires further experimentation with flow visualization. If the support is positioned downstream, performance marginally increases near 0 and 180 degrees and slightly decreases for positions directly downstream. When considering phase-average performance, the decrease in time-averaged performance for a support column directly upstream can be attributed to the blade passing through the column wake at the azimuthal position where it would normally produce maximum power. For upstream cases with improved performance, the trends are more subtle, with phase-average increases and decreases relative to the case without a support column that, in aggregate, slightly increase efficiency. Additional experiments consider the effects of different support column diameters and performance implications for multi-bladed turbines.
Figure: (a) The peak, time-average coefficient of performance, CP, of a single-bladed cross-flow turbine as a function of support column position (blue markers) relative to the case without support structure (black line). Azimuthal position refers to the location of the support structure. (b) The corresponding coefficient of thrust, CT, at peak performance. (c) The corresponding tip-speed ratio, λ, at peak time-average performance with U∞ indicating the direction of the freestream flow.