Bulletin of the American Physical Society
69th Annual Meeting of the APS Division of Fluid Dynamics
Volume 61, Number 20
Sunday–Tuesday, November 20–22, 2016; Portland, Oregon
Session A31: Experimental Techniques - 2D Particle Velocimetry |
Hide Abstracts |
Chair: Ronald Adrian, Arizona State University Room: F152 |
Sunday, November 20, 2016 8:00AM - 8:13AM |
A31.00001: On the rms errors and dynamic ranges of triple- and quadruple-pulse particle tracking velocimetry (PTV) Liuyang Ding, Ronald Adrian, Sivaram Gogineni Multi-pulse PTV extends conventional dual-pulse PTV by fitting a polynomial to particle locations measured from three or four pulses in a burst, aiming at more accurately resolving a particle short-period trajectory. Particle velocity and acceleration are then evaluated at an optimal time minimizing rms errors. Numerical simulations were performed to completely study the behaviors of position, velocity, and acceleration rms errors of triple- and quadruple-pulse PTV in a 4-D space spanned by four dimensionless variables -- normalized time, normalized displacement, normalized particle locating noise, and acceleration factor. We compared three analysis methods -- 3-pulse with quadratic fitting, 4-pulse with cubic fitting and 4-pulse with quadratic least-square fitting. In addition, generalized definitions of dynamic spatial range (DSR) and dynamic velocity range (DVR) are proposed for multi-pulse analyses. We calculated DSR ratios and DVR ratios between the multi-pulse and 2-pulse under various flow conditions and noise levels. It is found that the DSR and DVR could be improved by up to 100 times and 10 times, respectively, when the particle trajectory is strongly curved, deceleration is pronounced, and particle locations are accurately determined. [Preview Abstract] |
Sunday, November 20, 2016 8:13AM - 8:26AM |
A31.00002: Uncertainty Quantification and Statistical Convergence Guidelines for PIV Data Matthew Stegmeir, Dan Kassen As Particle Image Velocimetry has continued to mature, it has developed into a robust and flexible technique for velocimetry used by expert and non-expert users. While historical estimates of PIV accuracy have typically relied heavily on ``rules of thumb'' and analysis of idealized synthetic images, recently increased emphasis has been placed on better quantifying real-world PIV measurement uncertainty. Multiple techniques have been developed to provide per-vector instantaneous uncertainty estimates for PIV measurements. Often real-world experimental conditions introduce complications in collecting ``optimal'' data, and the effect of these conditions is important to consider when planning an experimental campaign. The current work utilizes the results of PIV Uncertainty Quantification techniques to develop a framework for PIV users to utilize estimated PIV confidence intervals to compute reliable data convergence criteria for optimal sampling of flow statistics. Results are compared using experimental and synthetic data, and recommended guidelines and procedures leveraging estimated PIV confidence intervals for efficient sampling for converged statistics are provided. [Preview Abstract] |
Sunday, November 20, 2016 8:26AM - 8:39AM |
A31.00003: Uncertainty Estimation for 2D PIV: An In-Depth~Comparative Analysis~ Aaron Boomsma, Syantan Bhattacharya, Dan Troolin, Pavlos Vlachos, Stamatios Pothos Uncertainty~quantification methods have recently~made great strides in accurately predicting uncertainties for planar PIV, and several different approaches are now documented.~ In the present study, we provide an~analysis of these methods across~different~experiments and different PIV processing codes.~ To assess the performance of said methods, we follow the approach of Sciacchitano et al. (2015) and~utilize two PIV measurement systems with overlapping fields of view---one acting as a reference~(which is validated using simultaneous LDV measurements)~and the other as a measurement system,~paying close attention to the effects of interrogation window overlap~and bias errors~on the analysis.~~A total of three experiments were performed: a jet flow and a cylinder in cross flow at two Reynolds numbers. In brief, the standard coverages (68{\%} confidence interval) ranged from approximately 65{\%}-77{\%} for PPR and MI methods, 40{\%}-50{\%} for image matching methods. We present an in-depth~survey~of both~global (e.g., coverage and error histograms) and local (e.g., spatially varying statistics) parameters to examine the strengths and~weaknesses~of each method~in monitor their responses to different regions of the experimental flows.~~ ~ [Preview Abstract] |
Sunday, November 20, 2016 8:39AM - 8:52AM |
A31.00004: Adaptive spectral filtering of PIV cross correlations Matthew Giarra, Pavlos Vlachos Using cross correlations (CCs) in particle image velocimetry (PIV) assumes that tracer particles in interrogation regions (IRs) move with the same velocity. But this assumption is nearly always violated because real flows exhibit velocity gradients, which degrade the signal-to-noise ratio (SNR) of the CC and are a major driver of error in PIV. Iterative methods help reduce these errors, but even they can fail when gradients are large within individual IRs. We present an algorithm to mitigate the effects of velocity gradients on PIV measurements. Our algorithm is based on a model of the CC, which predicts a relationship between the PDF of particle displacements and the variation of the correlation's SNR across the Fourier spectrum. We give an algorithm to measure this SNR from the CC, and use this insight to create a filter that suppresses the low-SNR portions of the spectrum. Our algorithm extends to the ensemble correlation, where it accelerates the convergence of the measurement and also reveals the PDF of displacements of the ensemble (and therefore of statistical metrics like diffusion coefficient). Finally, our model provides theoretical foundations for a number of "rules of thumb" in PIV, like the quarter-window rule. [Preview Abstract] |
Sunday, November 20, 2016 8:52AM - 9:05AM |
A31.00005: Application of photogrammetry to transforming PIV-acquired velocity fields to a moving-body coordinate system Pourya Nikoueeyan, Jonathan Naughton Particle Image Velocimetry is a common choice for qualitative and quantitative characterization of unsteady flows associated with moving bodies (e.g. pitching and plunging airfoils). Characterizing the separated flow behavior is of great importance in understanding the flow physics and developing predictive reduced-order models. In most studies, the model under investigation moves within a fixed camera field-of-view, and vector fields are calculated based on this fixed coordinate system. To better characterize the genesis and evolution of vortical structures in these unsteady flows, the velocity fields need to be transformed into the moving-body frame of reference. Data converted to this coordinate system allow for a more detailed analysis of the flow field using advanced statistical tools. In this work, a pitching NACA0015 airfoil has been used to demonstrate the capability of photogrammetry for such an analysis. Photogrammetry has been used first to locate the airfoil within the image and then to determine an appropriate mask for processing the PIV data. The photogrammetry results are then further used to determine the rotation matrix that transforms the velocity fields to airfoil coordinates. Examples of the important capabilities such a process enables are discussed. [Preview Abstract] |
Sunday, November 20, 2016 9:05AM - 9:18AM |
A31.00006: Postage-Stamp PIV: Small Velocity Fields at 400 kHz for Turbulence Spectra Measurements Steven Beresh, John Henfling, Russell Spillers Time-resolved particle image velocimetry (TR-PIV) recently has been demonstrated in high-speed flows using a pulse-burst laser at repetition rates reaching 50 kHz. However, the turbulent behavior at all but the largest scales occurs at still higher frequencies. Pulse-burst PIV can be achieved at higher frequencies if the field of view is greatly reduced and lower laser pulse energy is accepted. Current technology allows image acquisition at 400 kHz for sequences exceeding 4,000 frames, but for a small array of only 128 × 120 pixels, giving rise to the moniker of “postage-stamp PIV.” Despite the limited spatial extent, this approach is well-suited to measuring turbulent velocity spectra in high-speed flows because it is not subject to frozen turbulence assumptions and it can employ advanced algorithms using the closely-spaced laser pulses and local spatial information. The resulting spectra reveal two regions exhibiting power-law dependence describing the turbulent decay. One is the well-known inertial subrange with a slope of -5/3 at high frequencies. The other displays a -1 power-law dependence for a decade of mid-range frequencies corresponding to the energetic eddies measured by PIV, which appears to have been previously unrecognized for high-speed free shear flows. [Preview Abstract] |
Sunday, November 20, 2016 9:18AM - 9:31AM |
A31.00007: Validation of Multi-plane Particle Shadow Velocimetry to Quantify Turbulence Scales Christine Truong, Steven Hinkle, Kyle Sinding, Jeff Harris, Michael Krane, Rhett Jefferies Estimates of radial integral length scales using multi-plane Particle Shadow Velocimetry (PSV) are presented using measurements from the 11.2-inch diameter glycerin tunnel in the Applied Research Lab Garfield Thomas Water Tunnel. Particle shadow velocimetry (PSV) enables illumination of a volume and is an efficient means of obtaining multi-plane illumination. The combination of two colors in the LED backlight and a dichroic mirror makes possible the imaging of two planes in space. Thus, velocity fields in two imaging planes separated radially along the optical axis can be simultaneously measured. These multi-plane velocity fields are correlated over a range of separations to obtain integral length scales. Integral time scales are also calculated and converted into a streamwise length scale using Taylor's hypothesis for further confirmation. The inter-plane radial length scales, the in-plane length scales, the converted time scale in the inter-plane radial direction, and multi-plane turbulent statistics are compared with published studies, which used proven measurement methods. An additional, independent check is provided from PSV measurements in a single radial-axial plane. [Preview Abstract] |
Sunday, November 20, 2016 9:31AM - 9:44AM |
A31.00008: Application of Multi-Plane Particle Shadow Velocimetry to Obtain Velocity Fields Through an Optically Clear Object . Steven Hinkle, Christine Truong, Kyle Sinding, Rhett Jefferies, Jeff Harris, Michael Krane Particle Shadow Velocimetry (PSV) is performed using an LED array to illuminate a volume of fluid rather than individual two-dimensional laser sheets as is done in Particle Image Velocimetry (PIV). Multi-plane PSV is a technique that is able to take advantage of the volumetric illumination of PSV to simultaneously take velocity field measurements in two different planes along the same optical axis within the fluid flow. This technique can be further extended to resolve flow fields around and through clear objects to obtain measurements along the optical axis both in front of and behind the object. A proof of concept application of taking images both in front of and behind cylindrical rods is presented. The rods, one made of clear acrylic and the other borosilicate glass, were chosen to have an index of refraction close to that of the surrounding fluid. Two different calibration targets are arranged on either side of the cylinder and simultaneous images are taken to show that velocity measurements in front of and behind the rod are possible to obtain utilizing multi-plane PSV. This methodology will be implemented in future measurements to obtain velocity fields for an airfoil on both the suction and pressure sides simultaneously in fully developed turbulent flow. [Preview Abstract] |
Sunday, November 20, 2016 9:44AM - 9:57AM |
A31.00009: Recovering mean flow quantities from limited time-resolved PIV measurements through a data assimilation framework Sean Symon, Peter Schmid, Denis Sipp, Beverley McKeon Data assimilation combines experimental and numerical realizations of the same flow to produce hybrid flow fields. These have the advantages of less noise contamination and higher resolution while simultaneously reproducing the main physical features of the measured flow. This study investigates data assimilation of the mean flow around an idealized airfoil (Re = 13500) obtained from time-averaged PIV data. The experimental data, which represents a low-dimensional representation of the full flow field due to resolution and field of view limitations, is incorporated into a simulation governed by the incompressible RANS equations with an unknown momentum forcing. This forcing, which corresponds to the divergence of the Reynolds stress tensor, is calculated from a direct-adjoint optimization procedure to match the experimental and numerical mean velocity fields. The simulation is projected to the low-dimensional subspace of the experiment to calculate the discrepancy and a smoothing procedure is used to recover adjoint solutions on the higher-dimensional subspace of the simulation. The study quantifies how well data assimilation can reconstruct the mean flow and the minimum experimental measurements needed by altering the resolution and domain size of the time-averaged PIV. [Preview Abstract] |
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