arXiv subjects

Siyao Shao

Publications and source records attributed to Siyao Shao.

12 recordsLinked to original sources

arXiv2023

Hopf Physical Reservoir Computer for Reconfigurable Sound Recognition

The Hopf oscillator is a nonlinear oscillator that exhibits limit cycle motion. This reservoir computer utilizes the vibratory nature of the oscillator, which makes it an ideal candidate for reconfigurable sound recognition tasks. In this paper, the capabilities of the Hopf reservoir computer performing sound recognition are systematically demonstrated. This work shows that the Hopf reservoir computer can offer superior sound recognition accuracy compared to legacy approaches (e.g., a Mel spectrum + machine learning approach). More importantly, the Hopf reservoir computer operating as a sound recognition system does not require audio preprocessing and has a very simple setup while still offering a high degree of reconfigurability. These features pave the way of applying physical reservoir computing for sound recognition in low power edge devices.

cs.SD↗

arXiv2021

Probing into gas leakage characteristics of ventilated supercavity through bubbly wake measurement

The stability of ventilated supercavitation is strongly influenced by gas leakage characteristics of the cavity. Here we conduct a systematic investigation of such characteristics under different closure conditions including re-entrant jet (RJ), quad vortex (QV), twin vortex (TV), and pulsating twin vortex (PTV). Using high-speed digital inline holography (DIH), all the individual bubbles shed from the cavity are imaged downstream and are used to quantify the instantaneous gas leakage from the cavity. In general, the supercavity gas leakage exhibits significant fluctuations under all closure types with the instantaneous leakage rate spiking up to 20 times of the ventilation input under RJ and QV closures. However, the magnitude and occurrence rate of such excessive gas leakage vary substantially across different closures, tunnel speeds, and ventilation conditions. Particularly, as the supercavity transitions from RJ, to QV, TV, and PTV with increasing ventilation or decreasing tunnel speed, the relative excessive gas leakage decreases sharply from RJ to QV, plateaus from QV to TV, and drops again from TV to PTV. Correspondingly, the occurrence frequency of such excessive leakage first exhibits a double peak distribution under RJ, migrates to a single peak mode under QV and TV, and eventually transitions to a distribution with a broadened peak at a higher frequency under PTV. These trends can be explained by the flow instabilities associated with three distinct gas leakage mechanisms. Subsequently, two metrics are introduced to quantify the relative change of ventilation needed to compensate for the change of extra gas loss and the predictability of the occurrence of excessive gas leakage, respectively. Based on these metrics, we suggest that the supercavity operating under TV closure with moderate ventilation is optimal for ventilation-based controls of supercavity stability.

physics.flu-dyn↗

arXiv2021

Insights into ventilation hysteresis shift due to flow unsteadiness in ventilated supercavitation

Understanding ventilation strategy of a supercavity is important for designing high-speed underwater vehicles wherein an artificial gas pocket is created behind a flow separation device for drag reduction. Our study investigates the effect of flow unsteadiness on the ventilation requirements to form (CQf) and collapse (CQc) a supercavity. Imposing flow unsteadiness on the incoming flow has shown an increment in higher CQf at low free stream velocity and lower CQf at high free stream velocity. High-speed imaging reveals distinctly different behaviors in the recirculation region for low and high freestream velocity under unsteady flows. At low free stream velocities, the recirculation region formed downstream of a cavitator shifted vertically with flow unsteadiness, resulting in lower bubble collision and coalescence probability, which is critical for the supercavity formation process. The recirculation region negligibly changed with flow unsteadiness at high free stream velocity and less ventilation is required to form a supercavity compared to that of the steady incoming flow. Such a difference is attributed to the increased transverse Reynolds stress that aids bubble collision in a confined space of the recirculation region. CQc is found to heavily rely on the vertical component of the flow unsteadiness and the free stream velocity. Interfacial instability located upper rear of the supercavity develops noticeably with flow unsteadiness and additional bubbles formed by the distorted interface shed from the supercavity, resulting in an increased CQc. Further analysis on the quantification of such additional bubble leakage rate indicates that the development and amplitude of the interfacial instability accounts for the variation of CQc under a wide range of flow unsteadiness. Our study provides some insights on the design of a ventilation strategy for supercavitating vehicles in practice.

physics.flu-dyn↗

arXiv2020

Droplet evaporation residue indicating SARS-COV-2 survivability on surfaces

SARS-CoV-2 survives and remains viable on surfaces for several days under different environments as reported in recent studies. However, it is unclear how the viruses survive for such a long time and why their survivability varies across different surfaces. To address these questions, we conduct systematic experiments investigating the evaporation of droplets produced by a nebulizer and human-exhaled gas on surfaces. We found that these droplets do not disappear with evaporation, but instead shrink to a size of a few micrometers (referred to as residues), persist for more than 24 hours, and are highly durable against changes of environmental conditions. The characteristics of these residues change significantly across surface types. Specifically, surfaces with high thermal conductivity like copper do not leave any resolvable residues, while stainless steel, plastic, and glass surfaces form residues from a varying fraction of all deposited droplets at 40% relative humidity. Lowering humidity level suppresses the formation of residues while increasing humidity level enhances it. Our results suggest that these microscale residues can potentially insulate the virus against environmental changes, allowing them to survive inhospitable environments and remain infectious for prolonged durations after deposition. Our findings can also be extended to other viruses transmitted through respiratory droplets (e.g., SARS-CoV, flu viruses, etc.), and can thus lead to practical guidelines for disinfecting surfaces and other prevention measures (e.g., humidity control) for limiting viral transmission.

physics.med-ph↗

arXiv2020

Risk assessment of airborne transmission of COVID-19 by asymptomatic individuals under different practical settings

The lack of quantitative risk assessment of airborne transmission of COVID-19 under practical settings leads to large uncertainties and inconsistencies in our preventive measures. Combining in situ measurements and numerical simulations, we quantify the exhaled particles from normal respiratory behaviors and their transport under elevator, small classroom and supermarket settings to evaluate the risk of inhaling potentially virus-containing particles. Our results show that the design of ventilation is critical for reducing the risk of particle encounters. Inappropriate design can significantly limit the efficiency of particle removal, create local hot spots with orders of magnitude higher risks, and enhance particle deposition causing surface contamination. Additionally, our measurements reveal the presence of substantial fraction of crystalline particles from normal breathing and its strong correlation with breathing depth.

physics.med-ph↗

arXiv2020

Machine learning shadowgraph for particle size and shape characterization

Conventional image processing for particle shadow image is usually time-consuming and suffers degraded image segmentation when dealing with the images consisting of complex-shaped and clustered particles with varying backgrounds. In this paper, we introduce a robust learning-based method using a single convolution neural network (CNN) for analyzing particle shadow images. Our approach employs a two-channel-output U-net model to generate a binary particle image and a particle centroid image. The binary particle image is subsequently segmented through marker-controlled watershed approach with particle centroid image as the marker image. The assessment of this method on both synthetic and experimental bubble images has shown better performance compared to the state-of-art non-machine-learning method. The proposed machine learning shadow image processing approach provides a promising tool for real-time particle image analysis.

eess.IV↗

arXiv2019

Machine learning holography for measuring 3D particle size distribution

Particle size measurement based on digital holography with conventional algorithms are usually time-consuming and susceptible to noises associated with hologram quality and particle complexity, limiting its usage in a broad range of engineering applications and fundamental research. We propose a learning-based hologram processing method to cope with the aforementioned issues. The proposed approach uses a modified U-net architecture with three input channels and two output channels, and specially-designed loss functions. The proposed method has been assessed using synthetic, manually-labeled experimental, and water tunnel bubbly flow data containing particles of different shapes. The results demonstrate that our approach can achieve better performance in comparison to the state-of-the-art non-machine-learning methods in terms of particle extraction rate and positioning accuracy with significantly improved processing speed. Our learning-based approach can be extended to other types of image-based particle size measurements.

physics.app-ph↗

arXiv2019

Internal flows of ventilated partial cavitation

Our study provides the first experimental investigation of the internal flows of ventilated partial cavitation (VPC) formed by air injection behind a backward-facing step. The experiments are conducted using flow visualization and planar particle image velocimetry (PIV) with fog particles for two different cavity regimes of VPC, i.e., open cavity (OC) and two-branch cavity (TBC), under various range of free stream velocity (U) and ventilation rates (Q). Our experiments reveal similar flow patterns for both OC and TBC, including forward flow region near the air-water interface, reverse flow region, near-cavitator vortex, and internal flow circulation vortex. However, OC internal flow exhibits highly unsteady internal flow features, while TBC internal flow shows laminar-like flow patterns with a Kelvin-Helmholtz instability developed at the interface between forward and reverse flow regions within the cavity. Internal flow patterns and the unsteadiness of OC resemble those of turbulent flow separation past a backward-facing step (BFS flow), suggesting a strong coupling of internal flow and turbulent external recirculation region for OC. Likewise, internal flow patterns of TBC resemble those of laminar BFS flow, with the presence of unsteadiness due to the strong velocity gradient across the forward-reverse flow interface. The variation of the internal flow upon changing U or Q is further employed to explain the cavity regime transition and the corresponding change of cavity geometry. Our study suggests that ventilation control can potentially stabilize the cavity in the TBC regime by delaying its internal flow regime transition from laminar-like to highly unsteady.

physics.flu-dyn↗

arXiv2019

Machine Learning Holography for 3D Particle Field Imaging

We propose a new learning-based approach for 3D particle field imaging using holography. Our approach uses a U-net architecture incorporating residual connections, Swish activation, hologram preprocessing, and transfer learning to cope with challenges arising in particle holograms where accurate measurement of individual particles is crucial. Assessments on both synthetic and experimental holograms demonstrate a significant improvement in particle extraction rate, localization accuracy and speed compared to prior methods over a wide range of particle concentrations, including highly-dense concentrations where other methods are unsuitable. Our approach can be potentially extended to other types of computational imaging tasks with similar features.

eess.IV↗

arXiv2019

Effect of mounting strut and cavitator shape on the ventilation demand for ventilated supercavitation

The present work reports behaviors regarding the formation and collapse of a ventilated supercavity while varying the cavitator shapes, including triangle, disk, and cone and varying mounting struts. Three cavitators with the same frontal area are fabricated with 3D printing and mounted on a forward facing model (FFM). The ventilation requirements to generate (C_Qf) and sustain (C_Qc) a supercavity are tested over a wide range of Froude number (Fr) for each cavitator and compared with backward facing model(BFM). Compared to the triangle and disk cavitators, the cone-shaped cavitator requires the least amount of air to generate a supercavity in nearly all of the tested flow regime except very high Fr. The C_Qc of disk FFM is lower than that of its BFM counterpart at small Fr and exceeds the BFM C_Qc with further increase of Fr. The cone cavitator has the smallest C_Qc among all the cavitators across the range of Fr in our experiments. Simultaneous internal pressure and cavity dimension measurements are conducted to elucidate the cavity sustenance behaviors. The cone-generated cavity yields a significantly smaller maximum diameter and a shorter half-length. Cavity geometric information and cavity pressure measurements with high-speed imaging of re-entrant jet are employed to estimate the re-entrant jet momentum under different Fr for disk and cone cavitators. The estimated re-entrant jet momentum shows reasonable match with the ventilation air momentum under C_Qc in lower Fr for both cavitator cases, with the disk cavitator case yielding significantly stronger re-entrant jet, providing support to the re-entrant jet mechanism governing on the cavity collapse. Our study sheds some light on the cavitator design and ventilation strategy for a supercavitating vehicle in practice.

physics.flu-dyn↗

arXiv2019

Numerical study of the behaviors of ventilated supercavities in a periodic gust flow

We conducted a numerical simulation of ventilated supercavitation from a forward-facing cavitator in unsteady flows generated by a gust generator under different gust angles of attack and gust frequencies. The numerical method is validated through the experimental results under specific steady and unsteady conditions. It has been shown that the simulation can capture the degree of cavity shape fluctuation and internal pressure variation in a gust cycle. Specifically, the cavity centerline shows periodic wavelike undulation with a maximum amplitude matching that of the incoming flow perturbation. The cavity internal pressure also fluctuates periodically, causing the corresponding change of difference between internal and external pressure across the closure that leads to the closure mode change in a gust cycle. In addition, the simulation captures the variation of cavity internal flow, particularly the development internal flow boundary layer along the cavitator mounting strut, upon the incoming flow perturbation, correlating with cavity deformation and closure mode variation. With increasing angle of attack, the cavity exhibits augmented wavelike undulation and pressure fluctuation. As the wavelength of the flow perturbation approaches the cavity length with increasing gust frequency, the cavity experiences stronger wavelike undulation and internal pressure fluctuation but reduced cavitation number variation.

physics.flu-dyn↗

arXiv2019

Measurement of 3D bubble distribution using digital inline holography

The paper presents a hybrid bubble hologram processing approach for measuring the size and 3D distribution of bubbles over a wide range of size and shape. The proposed method consists of five major steps, including image enhancement, digital reconstruction, small bubble segmentation, large bubble/cluster segmentation, and post-processing. Two different segmentation approaches are proposed to extract the size and the location of bubbles in different size ranges from the 3D reconstructed optical field. Specifically, a small bubble is segmented based on the presence of the prominent intensity minimum in its longitudinal intensity profile, and its depth is determined by the location of the minimum. In contrast, a large bubble/cluster is segmented using a modified watershed segmentation algorithm and its depth is measured through a wavelet-based focus metric. Our processing approach also determines the inclination angle of a large bubble with respect to the hologram recording plane based on the depth variation along its edge on the plane. The accuracy of our processing approach on the measurements of bubble size, location and inclination is assessed using the synthetic bubble holograms and a 3D printed physical target. The holographic measurement technique is further implemented to capture the fluctuation of instantaneous gas leakage rate from a ventilated supercavity generated in a water tunnel experiment. Overall, our paper introduces a low cost, compact and high-resolution bubble measurement technique that can be used for characterizing low void fraction bubbly flow in a broad range of applications.

physics.app-ph↗