Anupam Sharma

While many existing machine learning studies have focused on augmenting Reynolds-averaged Navier–Stokes (RANS) turbulence models for steady or time-averaged unsteady flows, this paper takes a first step toward extending such augmentation to time-resolved unsteady flows. An unsteady field inversion and machine learning (FIML) method is developed, in which a temporally evolving correction field ([Formula: see text]) is incorporated into the production term of a RANS turbulence model. The inverse problem is solved by optimizing the spatiotemporal distribution of [Formula: see text] to minimize the regularized prediction errors. The resulting optimized [Formula: see text] field is then used to train a multilayer neural network that learns the time-dependent relationship between local flow features and [Formula: see text]. The approach is demonstrated using the unsteady flow over a NACA0012 airfoil undergoing dynamic stall. Results show that the unsteady FIML model, trained using only the time series of drag data at a given pitch rate, can accurately reproduce the spatiotemporal evolution of reference drag, lift, pitching moment, surface pressure, and velocity fields at both identical and different pitch rates. The unsteady FIML is integrated into the open-source DAFoam framework, enabling a pathway toward developing accurate and generalizable RANS turbulence models for time-resolved unsteady flows.

Novel ultrasonic bat deterrents based on aerodynamic whistles are proposed and investigated experimentally and numerically for their ability to generate ultrasound. The baseline deterrent, a single-whistle design inspired by Beeken [“Fluid ultrasonic generator,” U.S. patent 3,432,804 (1969)], is examined first. It consists of two resonating cavities/chambers. The whistle is “powered” by a regulated high-pressure air supply, and the performance of the whistle is examined for a range of supply pressures. Far-field acoustic measurements in the 20 Hz–50 kHz frequency range are made in an anechoic chamber. The noise measurements are supplemented with two- and three-dimensional unsteady Reynolds-averaged Navier–Stokes (uRANS) simulations to investigate the mechanisms of ultrasound generation. The Ffowcs Williams–Hawkings acoustic analogy is used with the three-dimensional uRANS results to predict the far-field radiation. The far-field acoustic predictions are in good agreement with the measurements in the anechoic chamber. The peak frequency (fundamental) of the radiated ultrasound of the baseline deterrent is approximately 23 kHz. Harmonics and a sub-harmonic of the fundamental tone are also observed. The numerical simulations show that the two resonating chambers of the baseline deterrent operate out-of-phase and Helmholtz resonance determines the whistling frequency over the range of supply air pressure considered. A six-whistle deterrent targeting a broad spectral coverage in the 20–50 kHz frequency range is designed, fabricated, and tested in the anechoic chamber. Each whistle in the deterrent is obtained by geometrically scaling the baseline whistle. Measurements show that the six-whistle ultrasound deterrent generates ultrasound with six dominant peaks in the designed frequency range when the supply air pressure exceeds 5 psig. The proposed ultrasound devices can be used for a variety of purposes including bat deterrence at wind turbines.