Bulletin of the American Physical Society
74th Annual Meeting of the APS Division of Fluid Dynamics
Volume 66, Number 17
Sunday–Tuesday, November 21–23, 2021; Phoenix Convention Center, Phoenix, Arizona
Session M20: Experimental Techniques: Quantitative Flow Visualization and Data Analysis II |
Hide Abstracts |
Chair: Adam Nickels, Penn State Room: North 221 AB |
Monday, November 22, 2021 1:10PM - 1:23PM |
M20.00001: Effect of the boundary conditions, temporal, and spatial resolution on the pressure from PIV for an oscillating flow Nazmus Sakib, Alexander Mychkovsky, James T Wiswall, Randy D Samaroo, Barton L Smith The pressure field of an impinging synthetic jet has been computed from time-resolved, three-dimensional, three-component (3D-3C) particle image velocimetry (PIV) velocity field data using a Poisson equation-based pressure solver. The pressure solver used in this work can take advantage of the temporal derivative of the pressure to enhance the temporal coherence of the calculated pressure field for time-resolved velocity data. The reconstructed pressure field shows sensitivity to the implementation of the boundary conditions, as well as to the spatial and temporal resolution of the PIV data. The pressure from a 3D Poisson solver that does not consider the temporal derivative of the pressure shows high random error. Invoking the temporal derivative of the pressure eliminates this high-frequency noise, however, the calculated pressure exhibits an unphysical temporal drift. This temporal drift is affected by both the temporal resolution of the PIV data and the spatial resolution of the PIV vector field, which was systematically evaluated by downsampling the instantaneous data and increasing the interrogation window size. It was observed that decreasing the temporal resolution increased the drift, while decreasing the spatial resolution decreased the drift. |
Monday, November 22, 2021 1:23PM - 1:36PM |
M20.00002: 3D Particle Image Velocimetry Uncertainty Quantification Rozhin Derakhshandeh, Sayantan Bhattacharya, Pavlos P Vlachos 3D Particle Image Velocimetry (PIV) is a non-invasive flow measurement technique, which has been widely used to study complex three-dimensional flow structures and resolve the three-component velocity fields. Accordingly, it can be used as a reliable validation data set for numerical studies as well as for design purposes in the industry, such as for Turbines, Jets, airplanes, etc. These applications intensify the importance of having an accurate measurement for 3D PIV and make uncertainty quantification a crucial factor in assessing the statistical significance of measurements in the experiment. However, this problem is non-trivial due to the complexity of possible uncertainty sources, their combination, and propagation through the measurement chain. For example, in Tomographic PIV, although there have been notable improvements in calibration error correction and object reconstruction, predicting the uncertainty bounds for each velocity vector measurement is still an open problem. Since both 3D PIV and conventional planar PIV are correlation-based methods, the existing 2D PIV uncertainty methods can be potentially extended to quantify the 3D PIV measurement uncertainty. However, the applicability and performance of such approaches have not been tested in 3D PIV measurements. In the current work, we build upon two existing direct 2D uncertainty estimation methods, namely Image Matching (IM) and Moment of Correlation (MC), and make necessary implementation and algorithmic modifications to quantify the uncertainty in 3D velocity fields. The resulted uncertainty measurements are tested with artificial Poiseuille data, and it has been validated by experimental laminar Poiseuille data. |
Monday, November 22, 2021 1:36PM - 1:49PM |
M20.00003: Treatment of Motion Blur in High-Speed PIV using Deep Learning Jeong Suk Oh, Hoonsang Lee, Wontae Hwang A technique based on deep learning that can reduce errors caused by motion blur for high-speed PIV is proposed. Synthetic images had the Monte-Carlo method (MCM) applied to them, in order to assess the error caused by blurry images of tracers. Longer particle streaks resulted in increased displacement errors (reaching 0.2 – 0.5 pixels) and outlier frequency (sometimes exceeding 8%). A new deblur filter was developed utilizing a generative adversarial network (GAN) with 1 million synthetic images. The filter, or generator, was verified using MCM data that was not learned. The outlier frequency was reduced to approximately 5%, and displacement error decreased below 0.25 pixels. This generator was applied to real blurry PIV images of a synthetic jet and significantly reduced the number of outlier vectors. |
Monday, November 22, 2021 1:49PM - 2:02PM |
M20.00004: Analysis of resolution and noise propagation in volumetric PIV measurements of kinetic energy, impulse, and their derivatives Derek J Li, Leah R Mendelson Vortex kinetic energy (KE) and impulse are typical quantities derived from fluid velocity data obtained via volumetric Particle Image Velocimetry (PIV). 3D velocity fields enable direct calculation of these quantities, but PIV measurement also filters the true flow field and inextricably involves noise. In this study, we consider the error in KE and impulse computations in relation to the level of noise, the resolution of the flow feature, smoothing, origin selection, and the field of view. Using a synthetic Hill's vortex, we show that error, for both KE and impulse, is primarily a function of feature resolution. Given a desired percentage of error, we identify the minimum resolution required as a practical reference for experiment design. Further, our results show that typical PIV post-processing is effective in reducing error in KE, while impulse computation is robust, with low unfiltered errors despite high noise levels. Temporal smoothing also has a significant effect on these quantities and their derivatives, which are needed for many formulations of wake power and forces. This study affords us practical recommendations when the computation of such quantities is needed, such as to assess the locomotive performance of swimming and flying organisms. |
Monday, November 22, 2021 2:02PM - 2:15PM |
M20.00005: Neural Network Approach to Traditional Particle Image Velocimetry Ryan J Sirimanne, Michael H Krane, Adam Nickels Particle Image Velocimetry (PIV) has become a standard tool in fluid dynamics to achieve planar and volumetric velocity measurements. Since its inception, calculation of velocity vectors have most commonly been achieved by tracking particles seeding a particular flow with spatial cross-correlation methods to calculate particle shifts in a certain time window. While this method of velocity vector calculation can be reliable in planar applications, measurement uncertainty, algorithm robustness, and scientist in the loop decision making have remained areas of concern. This work presents an outline for a PIV neural network architecture. Rather than training the neural network to identify and exploit physics, this work seeks to substitute a neural network into the contemporary PIV pipeline as a robust alternative to spatial cross correlations. The neural network is trained to track particle shifts over sub-regions of a particle image field with the added benefit of a more holistic treatment of uncertainty via probability. |
Monday, November 22, 2021 2:15PM - 2:28PM |
M20.00006: High speed PIV measurements in water hammer Roberto Capanna, Philippe M Bardet An experimental study addressing the challenge to measure relaxation coefficient of very fast phenomena such as water hammers is presented. An acrylic projectile containing water is accelerated and impacts a metal wall creating a water hammer. State of the art laser measurements techniques will be deployed in order to achieve such goal. A compressed air custom built cannon is used to accelerate the projectile and create the impact leading to the water hammer. First experimental results with Time Resolved High Speed PIV measurements (100kHz) are presented and discussed with focus on the future development for the presented facility. |
Monday, November 22, 2021 2:28PM - 2:41PM |
M20.00007: Ray-tracing 3D PIV/PTV for sparse particle distribution with a light field camera Liu Hong, Leonardo P Chamorro We will discuss a rapid 3D reconstruction of sparse particles distribution especially suitable for micro-scale interrogation volumes. Lighting difficulties and space constraints limit micro-scale particle image velocimetry (PIV) and particle tracking velocimetry (PTV) using a standard approach. Also, multiplicative algebraic reconstruction (MART) and simultaneous algebraic reconstruction technique (SART) used for light field reconstruction require large memory and substantial computational power to process the data. The use of microlens arrays (MLA) into the optical train of a microscope, 5D light array, which describes pixel position and direction of light, can provide a robust calibration method. Ray tracing and cloud point classification algorithm is used to obtain the spatial position of particles based on device parameters quickly in the order of a couple of seconds. We will discuss uncertainty using simulated setups with known particle positions and comparison with refocused positions. We discuss the 3D PIV approach to characterize the flow in a simple micro channel. |
Monday, November 22, 2021 2:41PM - 2:54PM |
M20.00008: Kalman filter-based volumetric PTV particle tracking Rudra Sethu Viji, Javad Eshraghi, Sayantan Bhattacharya, Pavlos P Vlachos Volumetric Particle Tracking Velocimetry (3D-PTV) is a non-invasive optical technique that measures fluid flow velocity by tracking the motion of neutrally buoyant tracer particles seeded into the flow. Each particle's images from different cameras are triangulated to find the 3D positions, which are subsequently tracked across frames to estimate the Lagrangian trajectory. This process is especially challenging at higher particle concentrations leading to overlapping particle images and erroneous reconstructions. Shake-The-Box has overcome this issue for time-resolved data by imposing the temporal continuity of tracks using a Wiener filter-based track predictor. However, optimal Wiener filter predictions require stationarity of the track evolution process and knowledge of noise statistics. On the other hand, a Kalman filter's Bayesian approach works on an uncertainty-informed iterative estimation methodology. Taking into consideration the recently proposed comprehensive 3D-PTV uncertainty estimation model we integrate a Kalman filter-based (KF) prediction, in place of the Wiener filter, to update the trajectory at each step. The current framework is tested for a synthetic vortex ring case at different noise levels. |
Monday, November 22, 2021 2:54PM - 3:07PM |
M20.00009: PINN-enhanced particle tracking velocimetry Shengze Cai, George E Karniadakis Particle tracking velocimetry (PTV), capable of providing a non-intrusive way for global velocity measurement, has been playing an important role in experimental fluid mechanics. However, due to the limitation of the particle matching algorithms, the resulting velocity field by PTV is generally sparse and noisy. In this work, we propose to use physics-informed neural networks (PINNs) to deal with this problem. The PINN-enhanced PTV approach assimilates the particle tracking vectors and the governing equations of the investigated flow, and then returns a temporally- and spatially-continuous flow field. In this context, the resulting velocity field is physically-interpretable and other physical quantities, including pressure and vorticity, can be inferred simultaneously. Moreover, we demonstrate that by using the PINN-enhanced PTV, we can downsample the velocity measurements in space and time to an extremely sparse level. We also show that with the boundary conditions imposed, the PINN-enhanced PTV is able to extend the flow field well beyond the observation domain. The proposed method is evaluated on various flow topologies (horseshoe vortex flow, wake flow and jet flow) and different experimental setups (2D PTV and Tomo-PTV). We demonstrate that the PINN-enhanced method is promising to become a standard way for processing the PTV data. |
Monday, November 22, 2021 3:07PM - 3:20PM |
M20.00010: Field-scale volumetric particle tracking system for snow settling measurement Nathaniel Bristow, Jiaqi Li, Michele Guala, Jiarong Hong Snow settling dynamics are a crucial parameter in our predictive capability of important processes such as cloud lifetime and ground snow accumulation. Recent studies have investigated the particle-turbulence interactions that modulate these dynamics for the first time in the field (Nemes et al. JFM, 2017; Li et al., JFM, 2021), utilizing large-scale particle tracking velocimetry (PTV), but have thusfar been limited to 2D measurements. Volumetric PTV can provide richer information of snow settling dynamics for longer time scales than in 2D, as particles do not leave the measurement region as frequently due to turbulent motions. Herein we present a field-scale volumetric PTV system using four cameras that is designed to be low-cost and enable flexible deployment for various arrangements to capture motion in different 3D volumes, depending on the resolution desired and snowfall density. A drone-based method for calibrating such a system was developed that can be readily extended to measurements over sample volumes at different scales, including regions of interest far above the ground. The performance of the system is evaluated using artificial snow particles over a sample volume of 3x3x3 m^{3} prior to field deployments with natural snow. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700