Session T09: Experimental Techniques: Quantitative Flow Visualization. PIV, PTV, PLIF (8:00am - 8:45am CST)

Improvement of particle detection accuracy in digital holographic microscopy by phase retrieval method.

Digital Holographic Microscopy (DHM) is a powerful tool for the measurements of particle motions and fluid flows in three-dimensional (3D) space at microscale. However, the DHM experiences the long-standing virtual image problem, which creates noises in the digital reconstructions and limits the hologram plane to be placed outside of the sample volume. Here, we implement the phase retrieval method in order to solve the virtual image problem. This new method is based on recording two holograms whose planes are displaced along the optical axis, and then reconstructing the complete optical waves estimated by the iterative phase retrieval algorithm. We demonstrate this new method by both numerical simulations and experimental measurements with dense particle suspensions exceeding 2000 particles/mm$^{\mathrm{3}}$. Results show that this new method totally eliminates virtual images, and recovers the original particle distributions even when the hologram planes are within the particle suspensions. Moreover, this phase retrieval method has a lower false particle detection rate and a higher particle localization accuracy, compared to the traditional method. In this presentation, we will introduce the optical setup and data analysis procedures of the phase retrieval method, as well as its application in turbulent boundary layers for velocity measurements.   

Quantifying biogenic turbulence through 3D scanning particle image velocimetry

Biogenic turbulence from aggregates of vertically migrating swimmers remains a poorly understood, and potentially underrepresented, source of scalar transport and ocean mixing. Though turbulent scales created by a single swimmer might be limited to those of the individual organism, the larger motions associated with the swimmer aggregates can introduce mixing scales relevant to the surrounding water column. Quantifying this process requires a volumetric, three-component velocimetry technique (3D-3C) capable of resolving the full range of flow scales in and around the swimmer aggregate. Here, we present a scanning particle image velocimetry system for reconstructing the three-dimensional configurations of vertically migrating swimmers and quantifying their volumetric, three-component velocity fields. The approach relies on a laser sheet that rapidly scans through the volume of interest, selectively illuminating slices of seed particles and swimmers. These images slices are captured by a single high-speed camera, encoding information about the third spatial dimension within the image time-series. The capabilities of the technique are evaluated on an induced vertical migration of brine shrimp \textit{Artemia salina}.   

An experimental study of the effect of water-soluble fluorescent surfactant on the interfacial wave characteristics of annular flows

We study experimentally the effects of surfactants on gas-liquid annular flows. We present a novel method to prepare water-soluble fluorescent surfactant solution and its detailed characterisation. In the presence and absence of this fluorescent surfactant, we conduct a detailed study on air-water annular pipe flows. The liquid and gas Reynolds numbers range between 500 to 1375 and 0 to 40000, respectively. We perform structured planar laser-induced fluorescence (S-PLIF) and capacitance probe measurements to accurately obtain film-thickness measurements and reveal the temporal characteristics of the waves. We further explore the differences on the gas entrainment rates, entrainment depths, and size of the bubbles in the liquid films for both cases. Diagnostic methods to track surfactant concentration are also currently being developed.   

Meta Uncertainty for Particle Image Velocimetry

There have been widespread efforts in the field of Particle Image Velocimetry (PIV) over the past decade to develop a-posteriori uncertainty quantification methodologies to report local, instantaneous uncertainties for each displacement measurement. However, multiple collaborative assessments show that none of the displacement uncertainty quantification methods perform well under all situations, and the methods can predict very different uncertainties for the same flow field as they operate under different assumptions and feature different calculation procedures. To address this issue, we propose an uncertainty combination framework to combine the uncertainty estimates from different schemes to provide a combined estimate, that is on average better than the individual schemes. We introduce ideas from the consensus forecasting literature, and combine estimates from different models based on a meta-analysis of the individual models/methods. We use a particle resampling approach to estimate the ‘response function’ of a given uncertainty estimation scheme to a given perturbation, and use this to assign weights for the individual uncertainty schemes. The methodology is assessed with synthetic and experimental images for planar and stereo PIV.   

Particle tracking experiments to capture droplet velocities in human exhalations.

Infection control guidelines suggest a spatial separation of 1 - 2 m as the safe distance between a health worker and an infected patient. This is based on assumptions of the risk of droplets spread from various respiratory exhalations. Most flow visualisation studies to date, provide only qualitative data, and do not provide sufficient details to accurately estimate the flow velocity of respiratory droplets. Here, we present a method to visualize droplets expelled during various exhalations and a framework to understand their dynamics. This method is tested to resolve the flow velocity of droplets expelled during exhalations, towards understanding their motion and dispersion. Preliminary results are presented by applying the methodology over various respiratory exhalations. Data from this work will be useful in understanding the transmission of infections and to inform infection control guidelines.   

High speed two color VLIF

A novel two-color volumetric laser induced fluorescence (VLIF) imaging system for fluid dynamics is presented. The lasing scanning system allows for a flexible trade-off between speed and resolution with a throughput of over 15 gigavoxels per second (e.g. at 512 x 512 x 512 we can record at up to 133 volumes per second). The portable scanning and imaging system is capable of both one-color and two-color VLIF on demand. Single shot (3D) measurements are demonstrated to illustrate the vast range of capabilities of the VLIF imaging system. Example data from our scanning VLIF technique, paired with a scanning particle tracking velocimetry (SPTV) technique, is presented for characterization of the flow past a sphere for a range of Reynolds number.   

Improved 3D reconstruction of velocity and density fields from SPIV and PLIF images by relaxation of Taylor's hypothesis

Given the high cost and complexity of setting up simultaneous tomographic PIV and LIF for the analysis of variable-density turbulent flows, we developed a novel algorithm that reconstructs from stereoscopic-PIV and planar LIF accurate 3D velocity and density fields with spatial resolution and size comparable to those achievable with direct volumetric measurements. This new algorithm is based on the use of the local instantaneous velocity for the data convection and the relaxation of the Taylor's hypothesis by the iterative enforcement of the incompressibility constraint on the velocity field. With application to numerical and experimental data, we demonstrate that this new method provides reconstructed 3D fields of variable-density flows with strong shear layers that are more accurate and that better satisfy the conservation of mass and the vorticity transport equations than those provided by the traditional method based on the convective mean field.   

Approximate Bayesian approach for volumetric reconstruction in a 3D PIV measurement

Volumetric Particle Image Velocimetry (PIV) is a non-invasive flow measurement technique which resolves the 3D flow field by recording multi-camera projections of the tracer particle motion. A key step in the measurement process is the volumetric reconstruction, which solves the inverse problem of estimating the 3D intensity field from the 2D particle image projections. This inverse problem is underdetermined and often leads to a high number of false reconstructions, especially for higher particle concentrations. The MART algorithm introduced by Elsinga et al. (2006) is the most widely accepted tomographic reconstruction method. However, the accuracy in such a reconstruction decreases with increasing seeding densities (\textgreater 0.05 ppp). The process is also computationally intensive. Here, we develop a Bayesian formulation to solve the inverse problem in a probabilistic sense. A maximum a posteriori (MAP) estimate is formulated using both uniform and Dirichlet process prior distributions for the 3D particle locations. The posterior is calculated using a likelihood function incorporating the camera calibration function and a Gaussian image noise. The MAP problem is recast as a stochastic optimization problem and it is solved using a stochastic gradient ascent algorithm which, in general, finds a better local maximum than a classical gradient based optimization. The cost function is iteratively solved using Tensorflow. This framework also provides an uncertainty bound on the estimate. The model is validated using a synthetic vortex ring data and an experimental pipeflow case.   

A high-resolution velocimetry technique based on decaying streaks from individual phosphor particles

A new high-resolution two-dimensional velocimetry technique is presented which is based on decaying streaks formed by individual phosphor particles following pulsed excitation. Tin-doped micron-sized phosphor particles are dispersed into flows and excited by a pulsed UV laser light sheet. Emission streaks are recorded as a result of the particle motion during the persistence of the luminescence ($\sim30$ $\mu$s). The two components of the flow velocity are derived from the streaks without directional ambiguity by applying to each streak a two-dimensional fit describing a linearly moving point source with a mono-exponential decaying emission. In addition, the frequency shifted luminescence allows rejection of reflected laser light, e.g., very near walls. The approach is first validated against particle tracking velocimetry and PIV in turbulent and laminar jets, where uncertainties were below 5\% in the 0.5 to 8 m/s range. Finally macroscopic velocity imaging measurements are presented in a boundary layer as close as 60 microns from the wall. This technique is particularly well suited for near-wall turbulent flow velocity measurements and, given the temperature dependence of the phosphor particles’ emission spectrum, it has the potential for simultaneous temperature measurements.   

Investigation of the shear-layer instabilities in supersonic impinging jets using double-PIV measurements.

The fundamental study of the phase-locked flow in supersonic impinging jets, generating strong resonance tones due to the presence of aero-acoustic feedback loop, is important for the aerospace propulsion and other industrial applications. While the shear-layer characteristics in such flows has been experimentally explored in various research studies using time-unresolved particle image velocimetry (PIV) technique, the understanding is limited due to the absence of temporal information. Due to the small time-scales associated with supersonic flows, the time resolved PIV measurements require a large bandwidth, which is challenging for the current state of the technology. An alternate approach using time unresolved double-PIV measurements is presented in the current study, which can generate multiple samples of dual-time data. The application of techniques like dynamic mode decomposition (DMD) on such data is shown to provide valuable information about the frequencies and the associated flow structures involved in the aero-acoustic feedback loop.   

Performance estimates of ratiometric quantum dot thermometry for the NASA ZBOT experiment

A whole-field planar optical technique for thermometry in the NASA Zero Boil-off Tank (ZBOT) experiment has been previously presented. The technique is based on ratiometric Laser Induced Fluorescence (LIF) using nanocrystal quantum dots (QD) that are modified to dissolve into the working fluid (perfluoropentane) of the ZBOT experiment. In this talk, we will discuss performance modeling of this two-color ratiometric LIF thermometry approach using two different sensing strategies: dual monochrome cameras with an optimized color filter set and a single camera with a standard Bayer color filter. The developed performance model is compared to measurements of both detection strategies in the ZBOT Breadboard setup.   

Deep optical flow for experimental fluid dynamics: sensitivity to network training

New advances in non-invasive medical imaging are emerging which hold promise to revolutionize patient screening and early disease detection. Optical flow through deep convolutional neural networks have shown significant promise. We utilize the LiteFlowNet[1], which is state-of-the-art architecture. LiteFlowNet with pre-trained weights shows a volatile performance on particle flows when compared to standard techniques in PIV. We first investigate how this depends on the specific examples seen during training and on the training modality (staged or end-to-end). We then train using a new dataset generated based on a Rankine vortex flow configuration, solutions to Stokes first and second problems, and CFD dataset from the literature[2]. Finally, we discuss the integration of prediction uncertainty in the LiteFlowNet architecture. \newline [1] Hui T-W., Tang X., Change Loy C., Liteflownet: A lightweight convolutional neural network for optical flow estimation, Proceedings of the IEEE conference on computer vision and pattern recognition, pp.8981-8989, June 18-22 2018, Salt Lake City, UT. \newline [2] Cai S., Zhou S., Xu C., Gao Q., Dense motion estimation of particle images via a convolutional neural network, Experiments in Fluids, 60(4), pp. 73, 2019.   

High PM concentration measurement with deep-learning based holographic speckle patterns.

A novel measurement technique of high particulate matter (PM) concentration was developed by adopting holographic speckle pattern analysis with deep learning. Conventional air-quality monitoring methods are usually cumbersome to handle and require expensive equipment for precise measurement and high throughput. The proposed technique, called Holo-SpeckleNet, can predict high PM concentration from holographic speckle patterns of PM particles. The speckle patterns of PMs were acquired for a wide range of PM concentrations, using a digital holographic microscopy (DHM) setup. Deep autoencoder (DAE) and regression algorithms were used to train the captured speckle patterns, and their concentrations were measured with a particle counter. Hyperparameter optimization and comparison with a typical convolutional neural network (CNN) algorithm were conducted to enhance measurement accuracy. The proposed measurement technique was found to exhibit high accuracy and speedy measurement under highly concentrated PM conditions unhealthy for human exposure. It would be applied to the design of a rapid, reliable, and accurate air-quality monitoring device. This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2017R1A2B3005415).   

Direct measurement of vorticity using tracer particles with internal markers

We demonstrate an optical imaging technique to obtain a direct measurement of 3D vorticity in a flow field, based on the measurement of instantaneous rotational rate of microscale tracer particles. The tracer particles of $\sim$ 50 $\mu$m with internal markers ($\sim$ 2 $\mu$m) are fabricated using a flow-focusing microfluidic device. Digital inline holography (DIH), which consists of a collimated coherent light beam and a digital camera that capture the diffraction signals (holograms) from the objects within the beam path, is employed to image several tracer particle within a field of view of centimeter scale. The holograms are then processed using an inverse reconstruction approach to obtain the 3D positions of each internal marker within a tracer particle. The translation and rotation of the particles are then derived from the time-resolved 3D positions of internal markers. The proposed approach is calibrated using a solid-body-rotational flow system and would be applied to probe the small-scale dynamics in different types of turbulent flow.   

Aerosol and splatter in a dental procedure; an experimental approach using PIV/PTV

With the novel coronavirus, SARS-CoV-2, outbreak there has been a directive to improve medical working conditions. In the dentistry field, airborne particles are produced through aerosolization during oral procedures, which utilize high-speed dental instruments. These droplets and particles could spread diseases such as influenza, Legionnaire's disease, or severe acute respiratory syndrome (SARS). This study explores the reflected spray from an ultrasonic scalar. Particle image velocimetry (PIV) and particle tracking velocimetry (PTV) were used to study the velocity and trajectories of splatter produced during the scalar process. It was observed that the maximum droplets velocity occurs directly below the tip of the scalar. Shadowgraphy was then used to measure the size and the velocity of individual droplets of the splatter produced by the high-speed scalar process after the device starts up. The droplets diameters were found to vary between 5$\mu $m to 600$\mu $m. Understanding the velocity and particle size distribution of droplets will help develop safety precautions to protect dentistry professionals from possible Coronavirus exposure.   

Lagrangian Strain Rate Tensor Evaluation based on Multi-pulse Particle Tracking Velocimetry and Radial Basis Functions

Physical conservation laws are inherently Lagrangian. However, analysis in fluid mechanics using the Lagrangian framework are often forgone in favor of those using the Eulerian framework. This is perhaps due to a lack of experimental techniques with high temporal and spatial resolution that track the movement of every fluid parcel in a flow domain. Development of time-resolved Particle Tracking Velocimetry/Accelerometry (TR-PTV/A) that measures flows with high seeding density has made the use of the Lagrangian framework more accessible. A challenge facing PTV/A is the need for robust mesh-free numerical schemes that handle random particle locations. Such a scheme can be created with high-order accuracy using Radial Basis Functions (RBFs). RBFs allow direct evaluation of derivatives of vector and scalar fields at random locations with infinite-high order of smoothness. The current work uses RBF-based differential schemes to develop a post-processing tool for PTV/A data, which can accurately evaluate spatial derivatives directly from Lagrangian particle tracks. This RBF-based strain rate tensor evaluation tool is validated with two and three-dimensional flows from analytical solutions and is then tested with experimental data measured by a multi-pulse PTA system.   

3D localization of the tracer particles in digital inline holographic microscopy PIV/PTV: do the bright regions in the intensity reconstruction volume really correspond to the tracer particles?

Digital Inline Holographic Microscopy based PIV/PTV (DIHM-PIV/PTV) techniques are becoming increasingly popular because of their ability to provide 3C-3D flow measurement with high spatio-temporal resolution, minimal optical setup and easy calibration compared to other 3D flow measurement techniques. One of the major challenges of DIHM-PIV/PTV technique is to set the criteria for the localization of the tracer particles in the reconstruction volume. Conventionally this is performed by finding the bright regions with peak local intensity. However, these bright regions may not always correspond to the actual tracer particles, especially in case where micrometer-sized dielectric spheres are used as tracer particles. In conventional DIHM setup these microspheres generate highly localized light, called photonic nanojets, further away from their centroid in the shadow side of the particles. Therefore, the peak intensity of the nanojet region is much higher than that of the actual particle and if not taken into account could be misjudged as the tracer particle. In this study, we explore the effect of the size and optical properties of some common tracer particles and propose a novel algorithm based on the reconstruction phase for accurate 3D localization of the tracer particles.   

Estimating turbine rotor thrust using near-wake stereo-PIV data

This study presents a method to estimate turbine rotor thrust using near-wake velocity data. Traditionally, wake surveys for force estimation have been conducted far downstream where the gauge pressure tends to zero. Herein, the need for pressure information is avoided by using vortical impulse theory, permitting the use of a stream-normal survey plane just downstream of the rotor. The derived equations depend only on the three components of velocity and their radial and azimuthal gradients, such that stereo-PIV data on this plane is sufficient to facilitate force estimation. Two force formulations are presented, both of which depend on the assumptions of steady rotation, a steady freestream velocity, and a rigid wake. The simpler of the two formulations depends on the additional assumption that the trailing vortex sheet is thin. The estimated forces using stereo-PIV data will be compared with direct measurements using a force transducer to determine the accuracy of the proposed methodology.   

Numerical Study of the Accuracy of the Particle Image Velocimetry Technique in High-Speed Turbulent Flows

In this study, we investigate the accuracy of the Particle Image Velocimetry (PIV) technique for the characterization of high-speed turbulence, which is representative of the flows typically found in modern practical and experimental combustion systems. We carry out direct numerical simulations and subsequently perform synthetic PIV reconstruction of the resulting three-dimensional turbulent flow field at Reynolds numbers $\approx 5000$ and Karlovitz numbers $\approx 200$ (premixed CH$_4$-air under atmospheric conditions). The flow field is uniformly seeded with 0.3 $\mathrm{\mu}$m monodispersed Al$_2$O$_3$ particles that represent typical PIV particles used in experiments, along with initially co-located massless Lagrangian tracer particles to recover the actual flow pathlines. We address the following questions: 1) How do PIV particles affect the carrier flow field? 2) How well do PIV particles sample the flow field of interest? 3) How closely do PIV particles follow the flow pathlines? 4) What is the accuracy of the PIV reconstructed flow field when compared to the true flow field? Finally, we conclude by discussing the implications of using PIV as a diagnostic tool for high-speed reacting and non-reacting flows.   

3D Unsteady Fluid-Structure Interactions Diagnostics with a Single Plenoptic Camera

Rising interests in fluids-structure interactions (FSI) stem from wide-ranging applications including NASA's supersonic parachute deployment, arterial flow, biomimicry and wind/wave power turbines. Conventional experimental studies of FSI nominally treat the flow and structural measurements separately, resulting in a loss of true FSI physics, especially for aperiodic flows. More recently, simultaneous six-camera approaches combining tomographic-particle image velocimetry (PIV) and stereographic-digital image correlation have been explored. However, these approaches are not economically viable and often challenging to implement especially in confined test facilities. Here, we demonstrate an alternative method of simultaneous FSI measurement using a single plenoptic camera. Equipped with a microlens array, plenoptic cameras are specialized imagers that preserve the 4D light-field of a measured volume within a single image. This enables single-camera 3D measurements to be achieved, typically with a higher degree of robustness against particle-particle occlusion than four-camera tomo-PIV. We utilized this principle and employed a newly developed \textit{kHz}-rate plenoptic camera to simultaneously measure the surface morphology of a rectangular blade and the unsteady flow around it. The 3.8\textit{cm}-chord blade was immersed in a water-tunnel, held at known static angles-of-attack (0-45$^{o})$, oscillated at known frequencies (0.5-1.0\textit{Hz}) and alternately switched between rigid and flexible constructions, thus providing a large dataset to benchmark the plenoptic-FSI technique's viability and accuracy.   

Time-resolved PIV measurements of an axis-symmetric forward-facing cavity.

The flow field of an axis-symmetric forward-facing cavity was experimentally investigated at Reynolds Number of 5,000 and 20,000 for three different aspect ratios (length/diameter) 1, 1.5 and 2 in the water tunnel using time-resolved particle image velocimetry. The location of the primary singular point was observed to depend upon aspect ratio for a given Re. Spatial oscillations of the primary singular point were observed that influence the formation of the separation bubble at the lip of the cavity and its convection downstream. The dividing streamlines shifted towards the outer wall for aspect ratio 2 and Re 20,000, two counter structures were observed inside the cavity. These vortices started to merge at aspect ratio 1.5 and transformed into one large structure for aspect ratio 1. The POD analysis of the PIV data was performed to understand the role of eddies emanating from the face of the cavity. The wake power spectra also contain the frequency of oscillations indicating close coupling between the two flows (external and internal).   

Development of a Low Cost Field PIV System for Shallow Environmental Flows

Field measurements of natural environmental flows can provide new insight into hydrodynamic processes. In contrast to the traditional technique of acoustic Doppler velocimetry, particle image velocimetry (PIV) offers a number of advantages such as the ability to directly calculate vorticity, observe coherent structures, and measure near boundaries. However, the application of PIV to the field has been limited by its expense and complexity. Here we present a novel, low cost PIV system suitable for measuring shallow environmental flows. The system uses multiple continuous wave green waterproof lasers equipped with Powell lenses in 3D-printed mounts to illuminate natural seeding at nighttime within an area up to approximately 1 m x 0.5 m. Two cameras, a Nikon D810 DSLR in an underwater housing and a GoPro 7 Hero, both filming at 60 Hz are tested, thereby providing time-resolved flow field for tens of minutes. All components are mounted on an easily transported, lightweight, aluminum frame. The system is tested in the optically clear waters of the run downstream of Lithia Springs, a second order magnitude freshwater spring in Hillsborough County, FL and is used to measure vertical profiles of instantaneous and mean streamwise velocity extending from the bed to the water surface. -/abstract- Authors: Azher Hamid, Stefano Mahairas, Sanjib Gurung, Tristen Mee, Da   

Tomographic Background Oriented Schlieren using Plenoptic Cameras

Tomographic background oriented schlieren (BOS) is a novel technique used to reconstruct the three-dimensional (3D) density or refractive index field in a compressible flow. A four-camera plenoptic BOS experiment was designed to perform a systematic study of varying length scale features in a 3D flow and their influence on the final tomographic reconstruction of a volumetric refractive index field. Solid transparent cylinders were submerged in a nearly refractive index matched solution to act as features within a static flow. This well-controlled experiment varied the cylinder size, separation distance between cylinders, and position with respect to the four-camera configuration. Final reconstructions aim to: (1) determine the separation distance limit between the two features before they can no longer be individually resolved, (2) observe how the separation distance changes as a function of cylinder size and position, and (3) compare the performance of the reconstruction as a function of how many viewing angles were used. Such results will provide both the schlieren and compressible flow communities with a better understanding of what limitations might be present in a final reconstruction with respect to the interference of features across a wide range of length scales.   

Development of 3D Plenoptic PIV to Study Flow over Rotating Wings

While the leading-edge vortex (LEV) is of great interest in rotor aerodynamics, research in this phenomenon has been hampered by difficulties in indefinitely following the 3D flow-field evolution, thereby necessitating measurements in the rotating frame of reference. Rotating 3D Velocimetry (R3DV) is a technique designed to fill this void using a plenoptic camera. The primary distinction between a conventional and plenoptic camera is an additional microlens array in front of the image sensor of the latter that splits incoming light rays based on incidence angle, allowing for reconstruction of 3D volumes from a single camera. Here, a volume of flow over a rotating wing is imaged with a stationary plenoptic camera via a hub-mounted rotating mirror locked to the rotor’s view. Quiescent flow measurements were also recorded in the absence of the wing to account for calculated velocity vectors being skewed by the non-inertial frame of reference. Implementation of a previously established plenoptic calibration method for stationary volumes has been adapted to incorporate systematic volume rotation. An overview of the methodology will be presented along with resulting visualization of the leading-edge vortex development over a rotating wing.   

Performance of kHz-rate plenoptic-PIV versus tomo-PIV on a 10mm-scale pipe flow

Plenoptic-Particle Image Velocimetry (PIV) is a 3D velocimetry technique performed via custom cameras imbued with microlens arrays (MLA). The MLA serves to encode both spatial and parallax information of a subject into a single recorded image, allowing depth inference using as few as one plenoptic camera. Thus, plenoptic-PIV represents a viable solution for 3D velocimetry in facilities with limited optical access, while also reducing alignment complexities and hardware costs relative to multi-camera tomographic-PIV. Here, we show the performance comparison of a single-camera high-speed plenoptic-PIV against the 4-camera tomographic-PIV, to characterize 3D flow in a 10mm scale pipe. The flow was varied from steady laminar (Re$_{\mathrm{d}}=$621-1167) to turbulent (Re$_{\mathrm{d}}=$3283) to pulsatile ($\omega_{\mathrm{0}}=$9.5) to produce different velocity profiles. Results, after 3D reconstructions and cross-correlations, were benchmarked using metrics such as particle counts, mutual information between correlating frames, and reconstructed particle shape. The velocity-field probability density function and the average velocity profiles were compared relative to each other and to the expected profile from pipe-flow theory. ~   

Mapping an outdoor odor plume using a mobile chemical sensor

How insects follow turbulent odor plumes over long distances under shifting wind conditions is an active area of investigation. Small scale wind tunnel experiments and simulations have hypothesized that they may use odor plume encounter intermittency as an indicator of the distance to the odor source. To determine if this strategy might work under natural outdoor wind conditions at large spatial scales on the order of 10 meters, we set out to map a real world odor plume using a mobile chemical sensor. The experiment was set up in an open space with no high-rise buildings, land surfaces, or trees within a mile radius. The setup included a propylene odor plume source, which was surrounded by three wind sensing stations coupled with GPS units. A mobile sensor hub was assembled which was mounted with a high precision GPS, wind and odor sensor, an inertial measurement unit, and a camera for visual odometry. The data we collected will be analyzed to estimate outdoor odor plume dynamics on a scale of 10 meters to test the hypothesis that plume encounter intermittency is correlated with distance to the source. Furthermore, we will explore the possibility of using our data to build an efficient virtual odor plume simulator to test algorithms for tracking turbulent odor plumes.   

Focused tracer detection algorithm for particle shadow velocimetry.

Particle Shadow Velocimetry (PSV) is an optical measurement technique which shares much with Particle Image Velocimetry (PIV), with some key differences. The light source shares a common optical axis with the camera and illuminates a volume, rather than the perpendicular light sheet used in PIV. To isolate tracers to a plane, PSV relies on the optical depth of correlation (DoC) rather than the thickness of the light sheet. PSV tracers thus appear as dark spots in a bright background rather than the converse, with out-of-focus particles much more visible than in comparable PIV images. Common practice is to invert the images, then use PIV algorithms. However, typical PIV image prefilters retain significant noise from the out-of-focus tracers. We present a new approach which detects focused tracers in PSV images with nontrivial DoC and lighting intensity variation, with an estimated error of 0.5{\%}. We then construct new images based on the detected positions of in-focus tracers, completely eradicating the out-of-focus noise. Tracer positions are detected via three primary operations: contrast stretching, edge detection, and thresholding on the image diameter, circularity, and darkness. We also show the application of our algorithm to the flow generated by the appendages of a gelatinous marine zooplankton. By presenting details of our focused tracer detection algorithm, we intend to help other PSV users improve their own data quality, and provide techniques that may also be useful to PIV users with noisy images.