State University Applied Research Laboratory Research Articles

Full-Scale Reynolds Number Testing of Rotor Hub Drag and Wake Turbulence

A 1:4.25-scale model of a generic helicopter rotor hub was tested at Reynolds numbers ranging from 1. 75× 106 to 7×106 at advance ratio of 0.2 in The Pennsylvania State University Applied Research Laboratory Garfield Thomas 48-inch diameter water tunnel. Measurements including drag and wake characteristics were performed up to full-scale Reynolds number with respect to an industry-representative helicopter rotor hub. In particular, the variation of drag and flow field with Reynolds number was characterized. Load measurements were conducted using an improved load cell design, with greater accuracy than in previous experiments. Wake velocity was measured using laser Doppler velocimetry at two downstream planes, yielding velocity statistics to the second order. Improved load measurement accuracy and wake velocity spatial resolution, at full-scale Reynolds number, provide a unique dataset for computational fluid dynamics validation as part of the Penn State Rotor Hub Flow Prediction Workshops and physical insight into rotor hub flows.

Visualization of a helicopter rotor hub wake

Experiments were conducted on a 1:17 model-scale helicopter rotor hub in the Pennsylvania State University Applied Research Laboratory 12-inch diameter Water Tunnel. Qualitative methods were used to investigate the interaction between the rotor hub and a canonical engine pylon upper surface. Oil paint visualization showed flow separation on the engine pylon and a high degree of interaction between the hub and pylon flows. Fluorescent dye injection revealed significant differences in the unsteady flow character between advancing and retreating sides of the rotor hub. Stereo particle image velocimetry was used in the wake of an isolated hub configuration to quantify the spatial and phase-dependent variations in wake velocity, revealing a wake bias towards the advancing side and, unexpectedly, a phase-averaged unsteadiness bias towards the retreating side. In general, a previously unavailable broad view of the rotor hub flowfield was obtained.

Scaling and Configuration Effects on Helicopter Rotor Hub Interactional Aerodynamics

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 to 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.

A comprehensive water tunnel test of a horizontal axis marine hydrokinetic turbine for model validation and verification

As interest in waterpower technologies has increased over the last few years, there has been a growing need for a public database of measured data for these devices. This would provide a basic understanding of the technology and means to validate analytic and numerical models. Through collaboration between Sandia National Laboratories, Penn State University Applied Research Laboratory, and University of California, Davis, a new marine hydrokinetic turbine rotor was designed, fabricated at 1:8.7-scale, and experimentally tested to provide an open platform and dataset for further study and development. The water tunnel test of this three-bladed, horizontal-axis rotor recorded power production, blade loading, near-wake characterization, cavitation effects, and noise generation. Additionally, preliminary comparisons are mode from unsteady CFD for the flow fields measured.

Investigation of two-dimensional numerical shock propagation using a weighted essentially non-oscillatory scheme

This work is focused on the computational simulation of realistic jet noise propagation in air. Consequently, in addition to accurate absorption, dispersion, wave steepening, etc., propagation of multidimensional shocks is necessary. Time-domain propagation of shocks requires some special technique to ensure stability and this is usually accomplished using artificial viscosity. In this investigation, instead of using artificial viscosity, a weighted essentially non-oscillatory (WENO) scheme is used. The WENO scheme is unique in that it uses a combination of multiple stencils to calculate derivatives. Shock propagation is achieved in this method through its use of smoothness indicators, which work to leave out any candidate stencils that contain a discontinuity in the calculation of derivatives. Its ability to handle a wide variety of acoustical conditions and the future potential of such a model is shown. [Work supported by ONR and the Pennsylvania State University Applied Research Laboratory Exploratory and Foundational Program.]

An approach to signal processing and condition-based maintenance for gearboxes subject to tooth failure

Condition vibration data from eleven gearboxes run to failure on the mechanical diagnostic test-bed at Penn State University Applied Research Laboratory was processed calculating the fault growth parameter (FGP) from the residual error signal. FGP, its revised version FGP1 and some other variables calculated from the residual error signal were analysed using the proportional-hazards modelling technique. Several statistical and replacement decision models were built based upon the observed condition data and ensuing failure events.

Volumetric imaging with an amplitude-steered array.

Volumetric acoustic imaging is desirable for the visualization of underwater objects and structures; however, the implementation of a volumetric imaging system is difficult due to the high channel count of a fully populated two-dimensional array. Recently, a linear amplitude-steered array with a reduced electronics requirement was presented, which is capable of collecting a two-dimensional set of data with a single transmit pulse. In this study, we demonstrate the use of the linear amplitude-steered array and associated image formation algorithms for collecting and displaying volumetric data; that is, proof of principle of the amplitude-steering concept and the associated image formation algorithms is demonstrated. Range and vertical position are obtained by taking advantage of the frequency separation of a vertical linear amplitude-steered array. The third dimension of data is obtained by rotating the array such that the mainlobe is mechanically steered in azimuth. Data are collected in a water tank at the Pennsylvania State University Applied Research Laboratory for two targets: a ladder and three pipes. These data are the first experimental data collected with an amplitude-steered array for the purposes of imaging. The array is 10 cm in diameter and is operated in the frequency range of 80 to 304 kHz. Although the array is small for high-resolution imaging at these frequencies, the rungs of the ladder are recognizable in the images. The three pipes are difficult to discern in two of the projection images; however, the pipes separated in range are clear in the image showing vertical position versus range. The imaging concept is demonstrated on measured data, and the simulations agree well with the experimental results.

Preliminary results of water tunnel acoustic measurements using a time reversal mirror

A concept demonstration test was recently conducted at the Garfield Thomas Water Tunnel at the Penn State University Applied Research Laboratory using a time reversal mirror to acquire acoustic measurements of a cavitation noise source. The objective of the test was to demonstrate improved acoustic measurement capabilities compared to conventional underwater acoustic measurement techniques. A preliminary analysis of the data indicates an increased signal-to-noise ratio resulting from the time-reverse playback of recorded data compared to real-time measurements made using either omnidirectional or sum-beam sensors. Experimental results are presented, including a novel method for source localization based on the difference between the time-forward and time-reverse playback of recorded data. [Work supported by ONR Code 334, Naval Surface Warfare Center Carderock Division, and the Applied Research Laboratory.]

Review of ultrasonic transducer calibration techniques at Penn State University Applied Research Laboratory

Recent emphasis on development of higher-frequency transducers for use in medical ultrasound and sonar applications has necessitated the creation of a reliable means of evaluation of these transducers. This work provides an overview of recent progress in the development of a facility for this purpose at the Pennsylvania State University Applied Research Laboratory. [Research funded by ONR.]

Performance of a two-stack annular prime mover

In a previous study, numerical investigations of a two-stack annular thermoacoustic prime mover were made [H. T. Lin et al., J. Acoust. Soc. Am. 102, 3071(A) (1997)]. The conditions for onset were predicted. The subject of this paper is an experimental investigation of a two-stack annular prime mover. The prime mover, modeled using DeltaE and a matlab program, is constructed using an annular aluminum resonator, copper fin heat exchangers, and two parallel plate, glass stacks. The relative position of the two stacks is adjustable within the resonator. Fluid-cooled ambient heat exchangers, and cartridge-heater powered hot heat exchangers are used to apply independent temperature differences across the two stacks. The sound field and temperature within the resonator are measured with a series of microphones and thermocouples distributed along the resonator. The performance of the prime mover is investigated as a function of relative stack position, temperature differences applied to the stacks, and heater power levels. [Work supported by the Office of Naval Research and The Pennsylvania State University Applied Research Laboratory.]

Modeling the effects of refraction and diffraction on wave propagation in a 3D, turbulent atmosphere medium

In ocean acoustic modeling, it has been established that many nonturbulent, deep-water propagation problems can be treated by the ‘‘N×2-D’’ approximation, where 2-D ‘‘slices’’ are computed along N radials. In the atmosphere, the validity of the N×2-D approximation is questionable because of ever-present, multiscale turbulence. In this paper, a new approach that allows the inclusion of multiscale turbulence in practical calculations is presented. The approximations that are physically motivated by the relative contributions of refraction and diffraction in multiscale turbulence are discussed. It is shown, for example, that for horizontal atmospheric structures larger than 10’s of wavelengths, horizontal diffraction is negligible, and the N×2-D approximation suffices. For smaller turbulence structures, horizontal diffraction is a dominant effect and must be accounted for. The theoretical/numerical method presented allows efficient calculations for both large- and small-scale 3-D atmospheric structure. Numerical examples are given to illustrate the phenomena involved. [Work supported by DARPA and the Pennsylvania State University Applied Research Laboratory.]

Strategies for computing sound propagation through realistic three-dimensional turbulence

For 3-D sound propagation calculations, if computation time were not an issue, it would be straightforward to extend 2-D parabolic equation (PE) models to three dimensions. However, for any 3-D propagation model, computation time is a crucial factor on any personal computer presently available. Hence, in this paper, some computational approaches are discussed that allow one to treat realistic 3-D turbulence in an approximate but physically reasonable way. The computational methods preserve the 3-D statistical properties of the turbulence without generating an unacceptable computational burden. The proposed approaches are outlined and numerical examples are given for sound propagation through meteorologically realistic turbulence fields. [Work supported by DARPA and the Pennsylvania State University Applied Research Laboratory.]

The quantitative significance of images reconstructed by acoustical holography

Acoustical holography is being considered for quantitative studies of acoustic source distributions. We examine the quantitative significance of images reconstructed using the holographic technique described by R. L. Cohen in Pennsylvania State University Applied Research Laboratory Tech. Memo. 76-05, Jan 1976, for sources distributed over a plane surface. The image is formed by measuring the magnitude and phase of the sound pressure due to the sources on a parallel hologram plane and then calculating values of a Helmholtz integral of these pressures for field points located on the original source plane. Values of this integral form the reconstructed image. If the source and hologram planes are many wavelengths apart, this image represents a pressure distribution over the source plane which is nearly identical to the radiating, low wavenumber portion of the original pressure distribution on the source plane. The original pressure distribution may contain nonradiating high wavenumber components which are attenuated in the reconstructed image. To provide insight for understanding reconstructed images, the radiating portion of the pressure distribution is exhibited for point sources and for a finite circular membrane vibrating in an infinite rigid baffle. The radiating portions of these distributions differ considerably from the total pressure distribution for these sources.