26 July 2019 in Journal of Fluid Mechanics
This paper presents a combined experimental and large-eddy simulation study to characterise the effect of aspect ratio on the near-wake structure of a square finite wall-mounted cylinder (FWMC). The cylinder aspect ratios (span $L$ to width $W$ ) investigated in the experiments were $1.4\leqslant L/W\leqslant 21.4$ and the oncoming boundary-layer thicknesses were $1.3W$ and $0.9W$ at a Reynolds number based on cylinder width of $1.4\times 10^{4}$ and $1.1\times 10^{4}$ , respectively. In complementary simulations, the cylinder aspect ratios investigated were 1.4, 4.3, 10 and 18.6. The cylinder wake structure was visualised in three-dimensional space using a vortex core detection method and decomposed to its oscillation modes using the spectral proper orthogonal decomposition (SPOD) technique. A parametric diagram is proposed to predict whether the time-averaged wake structure is a dipole or a quadrupole pattern, based on oncoming boundary-layer height and aspect ratio. Cellular shedding occurs when the aspect ratio is high with up to three shedding cells occurring across the span for aspect ratios $L/W>18$ . Each of these cells sheds at a distinct frequency, as evidenced by the spectral content of the surface pressure measured on the side face and the near-wake velocity. Amplitude modulation is also observed in the vortex shedding, which explains the amplitude modulation of the acoustic pressure emitted by square FWMCs. SPOD is shown to be a viable method to identify the occurrence of cellular shedding in the wake.
1 March 2019 in Journal of Sound and Vibration
1 September 2017 in The Journal of the Acoustical Society of America
Numerical simulations are used to investigate the noise generated by the passage of a rotor blade past a fixed object (the blade-passage effects), which was studied by simulating a three-bladed rotor that is supported by a vertical cylindrical tower. To isolate the blade-passage effects, no incoming wind was introduced in the simulation. The symmetric blade was set to zero pitch angle relative to the plane of rotation and two blade-tower distances were investigated. The sliding mesh method was used to simulate the rotation of the blades and Curle's acoustic analogy was used to predict the noise generated from the simulated flow data. Intense force fluctuations occur during the interaction on both the tower and the passing blade, and these are the primary sources of blade-passage noise. The contribution of the force fluctuations on the support tower to blade-passage noise, which previously had been ignored, was revealed to be more significant than that of the blades. The numerical model successfully predicts the noise spectra, which are validated by the very good agreement with experimental measurements. The simulations provide a framework to better understand blade–tower interaction noise in various applications.
23 August 2024 in Acoustics Australia
This paper examines and compares methods of separating tonal and broadband components of the noise generated by small rotors as commonly found on small unmanned aerial systems (sUAS). Time synchronous analysis (TSA) methods with varying averaging algorithms, such as ensemble averaging (EA), exponential weighted moving averaging (EWMA) and Kalman filter averaging (KFA), are compared against themselves and against a cross-correlation-based method. The decomposition methods are used on noise measurements of a small isolated rotor under static operation and edgewise flight conditions at 10 m/s in the UNSW anechoic wind tunnel (UAT). The best method for isolating tones is TSA-EWMA and for isolating the broadband spectrum of noise is the cross-correlation method, based on the spectral reconstruction of the experimental data.