Scaling and Configuration Effects on Helicopter Rotor Hub Interactional Aerodynamics

Abstract

A 1:17 scale model of a notional rotor hub of a large helicopter was tested in the Pennsylvania State University Applied Research Laboratory 12 in. test-section water tunnel. Objectives of the experiment were to quantify the effects of Reynolds number, advance ratio, and hub geometry configuration on the drag and shed wake of the rotor hub. A range of flow conditions was tested, with hub-diameter-based Reynolds numbers ranging from 1.0 × 10 6 to 2.6 × 10 6 and advance ratios ranging from 0.2 to 0.6, as well as nonrotating cases. Five hub geometry configurations were tested with various combinations of components including blade stubs, spiders, scissors, the swashplate, pitch links, and beanie fairing. Measurements included the steady and unsteady hub drag and particle image velocimetry at two downstream locations. Results include time-averaged and phase-averaged analyses of the unsteady drag and wake velocity. A strong dependence of the steady and unsteady hub drag and wake on the advance ratio, Reynolds number, and configuration was observed, demonstrating the importance of adequate Reynolds number scaling for model helicopter rotor hub tests.

Characterization of a canonical helicopter hub wake

“…All testing was performed at a freestream speed (U∞) at the test section inlet of 9.9 m/s. The models were rotated at 30 Hz to set the advance ratio (µ = U∞/ωR, where ω is the shaft angular velocity and R is the rotor blade radius) at 0.2, which is consistent with previous studies (Roesch & Dequin, 1985;Phelps & Berry, 1987;Reich et al, 2014aReich et al, , 2015Reich et al, , 2017Reich et al, , 2018. Note that ratios were being scaled to be consistent with Reich et al (2014a) (i.e.…”

“…4 to illustrate the general orientation and phase variation within a hub revolution. The streamwise spatial distribution of vectors was converted to a relative phase position assuming Taylor's frozen turbulence hypothesis (similar to Reich et al, 2014aReich et al, , 2017, which is valid given that the streamwise velocity fluctuations were less than 5% of the free-stream speed. This assumes the mean flow dominates advection and lag times per vector spacing were estimated by dividing the vector spacing (dx) by the average convection velocity for a given z-location.…”

“…Phase-averaged PIV (2D and stereo-) is a standard approach for rotor-wake measurements (2D - Reich et al, 2014aReich et al, , 2017stereo -Raffel et al, 2004;Raghav & Komerath, 2015;Reich et al, 2014aReich et al, , 2018. Note that given the strongly three-dimensional flow-field, stereo-PIV is generally preferred to mitigate the influence of out-of-plane particle motions.…”

“…While additional high-Reynolds number long-age wake surveys (Schmitz et al, 2017;Metkowski et al, 2018) have been performed since Reich et al (2014a), the wake survey data have not been published. Reich et al (2017Reich et al ( , 2018 have wake survey results, but Reich et al (2017) primarily examined variations with additional model features (pitch links and beanie fairing), which makes it difficult to compare results due to the changing model complexity. Reich et al (2018) used stereo-PIV to characterize the wake behind a configuration similar to Reich et al (2014a).…”