No abstract

With particle image velocimetry (PIV), cross-correlation and optical flow methods have been mainly adopted to obtain the velocity field from particle images. In this study, a novel artificial intelligence (AI) architecture is proposed to predict an accurate flow field and drone rotor thrust from high-resolution particle images. As the ground truth, the flow fields past a high-speed drone rotor obtained from a fast Fourier transform-based cross-correlation algorithm were used along with the thrusts measured by a load cell. Two deep-learning models were developed, and for instantaneous flow-field prediction, a generative adversarial network (GAN) was employed for the first time. It is a spectral-norm-based residual conditional GAN translator that provides a stable adversarial training and high-quality flow generation. Its prediction accuracy is 97.21 % (coefficient of determination, or R2). Subsequently, a deep convolutional neural network was trained to predict the instantaneous rotor thrust from the flow field, and the model is the first AI architecture to predict the thrust. Based on an input of the generated flow field, the network had an R2 accuracy of 94.57 %. To understand the prediction pathways, the internal part of the model was investigated using a class activation map. The results showed that the model recognized the area receiving kinetic energy from the rotor and successfully made a prediction. The proposed architecture is accurate and nearly 600 times faster than the cross-correlation PIV method for real-world complex turbulent flows. In this study, the rotor thrust was calculated directly from the flow field using deep learning for the first time.

In this study, the effects of rotor–rotor interaction on wake characteristics were investigated experimentally for a twin-rotor configuration in axial descent. The wake velocities were measured at descent rates (descent speed/induced velocity at the rotor disk during hover) from 0.87 to 1.52, and the rotor–rotor interaction strength was controlled by adjusting the distance between the rotor tips. As the descent rate increased, the wake of the isolated rotor gradually entered the vortex ring state (VRS), where the flow established an extensive recirculation zone. Correlation analysis was performed to distinguish the rotor wake between tubular and VRS topologies. The flow states for the isolated rotor were classified into pre-VRS, incipient VRS, and fully developed VRS, depending on the probability of vortex ring formation. The results reveal that the effects of rotor–rotor interaction on the wake characteristics of twin rotors differ depending on the descent rate, distance between rotor tips, and wake region. In the outer region, the flow state of the rotor wake remains consistent with that of the isolated rotor, irrespective of the distance between rotor tips. Conversely, the strong rotor–rotor interaction changes the flow state in the inner region by disrupting the vortex ring structure, intensifying the wake asymmetry about the rotational axis. The thrust measurements show that under the VRS, as the two rotors get closer, the thrust coefficient increases until vortex ring disruption occurs, and then decreases after the vortex ring is disrupted.