Images of fluid flow: Art and physics by students

In Spring 2003, a new experimental course on flow visualization was offered to a mixed class of Fine Arts Photography and Engineering students. Course content included fluid flow physics, history of photography with respect to the relationship of science and art, as well as flow visualization and photography techniques. Issues such as “What makes an image art? What makes an image scientific?” were addressed. The class focused on studio/laboratory experiences for mixed teams of students. In Spring 2004 these concepts were distilled into an engineering outreach experience for middle school girls. The spectacular images resulting from these experiments show that flow visualization can be both performed and appreciated by a broad spectrum of people. Thus flow visualization may represent a new bridge between scientists and non-scientists.     

Ring vortex scenario in engraved champagne glasses

The simple idea this study rests on is that one cannot be concerned by the bubbling and aromatic exhalation events in champagne tasting without being interested in the study of the flow mixing mechanisms inside the glass. Indeed, a key assumption is that a strong link of causality may exist between inherent liquid-phase flow structures due to bubble motion and the flavors exhalation process. This is the reason why, to underscore the impact of glass-shape and glass-engravement conditions on mixing flow phenomena, classical flow visualization techniques were used to capture fluid motion in traditional flutes and coupes poured with champagne. Laser tomography combined with fluorescent dyes and solid tracers have been used to give the quasi-instantaneous velocity field from which streamline patterns are deduced as well as the vorticity convection.     

Flow visualization and scanning PIV measurement of three-dimensional structure in near field of strongly buoyant jet

Abstract  

The near-field structure of strongly buoyant jet issuing from a square nozzle at low Froude and Reynolds numbers is studied by using LIF flow visualization and time-resolved scanning PIV. These experimental techniques allow the visualization of unsteady three-dimensional flow phenomenon occurring in the near-field of strongly buoyant jet. It is found that the buoyant jet is unstable to the positive buoyancy forces, which promote the inflow motion near the nozzle exit. The surrounding low temperature fluid moves into the nozzle inside along the nozzle corner and mixes with the high temperature fluid deep into the nozzle. Then, the flow pattern inside the nozzle becomes highly complex to promote the laminar to turbulent transition of the jet. The statistical flow characteristics of the strongly buoyant jet are evaluated from the scanning PIV measurement, and the result indicates the presence of axisymmetric distributions of mean flow and velocity fluctuations in the circle of diameter equal to the square side of the nozzle.     

Application of k- and q-space encoding NMR techniques on granular media in a 3D model fluidized bed reactor

A combination of PFG-NMR imaging and velocity encoding methods was applied to investigate the dynamic behavior of a bed of poppy seeds subjected to air flow, representing a model setup for fluidized bed reactors. The particle motion is described both from a statistical point of view, by determining propagators and dispersion coefficients representing an average over the whole bed volume, as well as combined with spatial resolution by generating velocity maps. Velocity images of different horizontal slices in the bed confirm the notion of a toroidal particle flow pattern inside the shallow granular bed. Despite the need of considerable averaging due to the random motion of the relatively few particles in the bed, quantitative velocity images and statistical information about the random particle motion can be obtained from monitoring the fluid component in the seeds by conventional spin-echo techniques.     

Visualization of scientific arts and some examples of applications

In this paper, implementation and visualization of scientific arts are described using some examples of application in subject research areas, such as sculpture, archeology, fine arts and information aesthetics, which have been discussed through the Scientific Art Session at FLUCOME9, Tallahassee, Florida, 2007–9. In the application to sculpture, stereo visualization techniques, such as anaglyph stereo visualization and integral imaging technique, are introduced to realize the three-dimensional geometry of sculpture to enhance visual impact on the art. The second application is the flow visualization technique for archeology, where the vortices behind the river stones are studied to understand the origin of patterns on Jomon pottery. Interestingly, such vortex patterns also appear in the paintings of fine arts. The third example is the visualization of information aesthetics, where the Web information, such as public media and stock market, are visualized through scientific techniques. These examples of visualization of scientific arts provide the present state of the art in interdisciplinary visualization.     

Filming the invisible – time-resolved visualization of compressible flows

Essentially all processes in gasdynamics are invisible to the naked eye as they occur in a transparent medium. The task to observe them is further complicated by the fact that most of these processes are also transient, often with characteristic times that are considerably below the threshold of human perception. Both difficulties can be overcome by combining visualization methods that reveal changes in the transparent medium, and high-speed photography techniques that “stop” the motion of the flow. The traditional approach is to reconstruct a transient process from a series of single images, each taken in a different experiment at a different instant. This approach, which is still widely used today, can only be expected to give reliable results when the process is reproducible. Truly time-resolved visualization, which yields a sequence of flow images in a single experiment, has been attempted for more than a century, but many of the developed camera systems were characterized by a high level of complexity and limited quality of the results. Recent advances in digital high-speed photography have changed this situation and have provided the tools to investigate, with relative ease and in sufficient detail, the true development of a transient flow with characteristic time scales down to one microsecond. This paper discusses the potential and the limitations one encounters when using density-sensitive visualization techniques in time-resolved mode. Several examples illustrate how this approach can reveal and explain a number of previously undetected phenomena in a variety of highly transient compressible flows. It is demonstrated that time-resolved visualization offers numerous advantages which normally outweigh its shortcomings, mainly the often-encountered loss in resolution. Apart from the capability to track the location and/or shape of flow features in space and time, adequate time-resolved visualization allows one to observe the development of deliberately introduced near-isentropic perturbation wavelets. This new diagnostic tool can be used to qualitatively and quantitatively determine otherwise inaccessible thermodynamic properties of a compressible flow.     

Three-dimensional particle cluster tracking algorithm using affine transformation

In Particle Image Velocimetry (PIV), the cross correlation tracking technique is widely used to analyze the particle images. The technique assumes that the fluid motion, within small regions of the flow field, is given by parallel movements over short time intervals. However, actual flow fields may have some distorted motion, such as rotation and shear. Therefore, if the distortion of the flow field is not negligible, the fluid motion can not be tracked well using the cross correlation technique. The author proposed a new particle tracking technique, based on the particle cluster matching using linear Affine Transformation. The algorithm can be applied to flow fields which exhibit characteristics such as rotation and shear. The deformation of the cluster pattern is expressed by the linear Affine Transformation. The parameter of the transformation can be determined using the least square technique from the particle positions. The effectiveness of the tracking techniques, including 3D cross correlation, Spring Model and Affine Transformation, were evaluated with synthetic data of three-dimensional flow field. The cross correlation technique could be applicable to the small deformation cases. When the deformation of particle pattern between two images are very large, the pattern deformation could not be expressed by the Affine Transformation, i.e., linear transformation, resulting in mis-tracking. However, the Spring Model technique was found to be more effective even in the larger deformation condition, because the Spring Model does not assume the linear transformation.     

Visualization of flow fields in a Bubble Eliminator

Bubbles and dissolved gases in liquids greatly influence the performance of fluid power systems, coating solutions, plants in the food industry and so on. To eliminate bubbles from working fluids and to prevent degradation of liquids as well as to avoid possible damage of fluid components is an important engineering issue. Recently one of the authors, Ryushi Suzuki, has developed a new device using swirling flow with the capability of eliminating bubbles and of decreasing dissolved gases in fluids. This device is called “Bubble Eliminator.” The swirling flow pattern and pressure distributions in the bubble eliminator greatly influence the effective performance of the bubble removal. In this paper the swirl flow pattern in a transparent bubble eliminator is experimentally visualized and processed as digital images by a high-speed video camera system. Velocity profiles and pressure distributions in the bubble eliminator are calculated and graphically visualized by a three-dimensional numerical simulation. The results of the flow visualization are compared with the numerical simulation. The performance evaluation of the bubble removal effectiveness is numerically and experimentally verified. It is also proposed to augment understanding of 3D flow fields for the swirling flow in the bubble eliminator with scientific flow visualization methods, which combine graphics or real images with haptic displays.     

A PIV and LSV study in the wake of an aircraft model

An experimental investigation was undertaken of the wake aerodynamics of a 1/12.81 model of a commercial passenger aircraft. The tests were undertaken in the 3.05 m by 3.66 m working section of a closed-return wind-tunnel. The program made use of laser sheet visualization (LSV) and particle image velocimetry (PIV) in planes, normal to the mean flow, to obtain images of flow-following, seed particles. The images were processed to obtain raw velocity vectors, flow divergence, vorticity, crossflow energy and high quality visualizations. Image sequences obtained under identical incident flow conditions were used to conceptualize the variability of principal wake features, such as regions of fluid of high vorticity.     

The design secret of Kyokusui-no-En’s meandering channel

The purpose of this study is to investigate the characteristics of the flow through the Jonangu channel which is used for ceremonial game called as ‘Kyokusui-no-En’ in Japanese. The geometry of the channel is measured, a visualization technique is used to measure the actual flow characteristics, and then a numerical flow model is used to represent the flow including unsteady flow characteristics. Numerical model of drifting cup is introduced to investigate an interaction between flow and motion of the cup. Finally, the intention of the channel design is interpreted from the viewpoint of fluid mechanics using observed and calculated results.     

Imaging of a convective field in a rectangular cavity using interferometry, schlieren and shadowgraph

The present study is concerned with the quantitative imaging of buoyancy-driven convection in a fluid medium that is confined in a horizontal differentially heated rectangular cavity. The horizontal surfaces of the cavity provide a temperature difference, for initiating convection in the fluid. The vertical side walls are thermally insulated. Three imaging techniques, namely laser interferometry, schlieren, and shadowgraph have been utilized. Experiments have been conducted in a cavity of length 447 mm and 32 mm vertical height. The cavity is square in cross-section, and the imaging direction is parallel to its longer side. Convection in air and water have been investigated. Temperature differences in the range of 5–50 K for air and 3–10 K for water have been employed in the experiments. Quantities of interest are the temperature profiles in unsteadiness in the thermal field. At lower temperature differences across the fluid region, temperatures as recorded by interferometry and schlieren are in good agreement with each other. Further, they match the numerical predictions, as well as correlations available in the literature. Imaging based on shadowgraph is not as satisfactory at lower temperature differences. At larger cavity temperature differences, the shadowgraph images become clear enough for quantitative analysis, but the flow becomes time-dependent. The three techniques reveal similar trends in terms of the spatial distribution of temperature gradients and the time scales of unsteadiness. The schlieren and shadowgraph are more suitable for high gradients and interferometry is suitable for low gradients and all these three techniques are not flow visualization tools alone but are appropriate for quantitative imaging of thermal field.     

Passive imaging of wind surface flow using an infrared camera

We present a method for passive imaging of wind motion against surfaces in a scene using an infrared video camera. Because the method does not require the introduction of contrast agents for visualization, it is possible to obtain real-time surface flow measurements across large areas and in natural outdoor conditions, without prior preparation of surfaces. We show that this method can be used not just for obtaining single snapshot images but also for real-time flow video, and demonstrate that it is possible to measure under a wide range of conditions.     

Three-dimensional bubbly flow measurement using PIV

The experimental flow visualization tool, Particle Image Velocimetry (PIV), is being extended to determine the velocity fields in three-dimensional, two-phase fluid flows. In the past few years, the technique has attracted quite a lot of interest. PIV enables fluid velocities across a region of a flow to be measured at a single instant in time in the whole volume (global) of interest. This instantaneous velocity profile of a given flow field is determined by digitally recording particle (microspheres or bubbles) images within the flow over multiple successive video frames and then conducting flow pattern identification and analysis of the data. This paper presents instantaneous velocity measurements in various three-dimensional, bubbly two-phase flow situations. This information is useful for developing or improving existing computer constitutive models that simulate this type of flow field. It is also useful for understanding the detailed structure of two-phase flows.     

Phase contrast MRI with improved temporal resolution by view sharing: k-space related velocity mapping properties.

Phase contrast techniques in combination with k-space segmented CINE imaging are widely used for the quantitative assessment of blood flow or tissue motion. The temporal resolution of the corresponding pulse sequences plays an important role concerning the potential of the method to fully detect time resolved flow or motion patterns. A further improvement of temporal or spatial resolution in phase contrast CINE MRI can be achieved by the application of view sharing. Based on simulations with point-spread-functions resulting from different cyclic flow or motion patterns an analysis of view sharing techniques in combination with phase contrast MRI is presented. Velocity mapping properties and the role of different k-space regions concerning the resulting values in the phase images and thus encoded velocities were investigated. It could be shown that the velocity induced phase shifts in phase contrast techniques are mainly encoded in the central sections of k-space which makes view sharing also suitable for velocity mapping. As a result the use of appropriate sampling and data acquisition schemes permits the assessment of flow or motion patterns with significantly improved temporal resolution without loss of functional information. In addition phantom measurements with an oscillation phantom were performed in order to validate the simulation results and to demonstrate the potential of view sharing techniques to accelerate phase contrast imaging and improve the detection of the underlying flow or motion dynamics.     

Visualization of respiratory flows from 3D reconstructed alveolar airspaces using X-ray tomographic microscopy

Abstract  

A deeper knowledge of the three-dimensional (3D) structure of the pulmonary acinus has direct applications in studies on acinar fluid dynamics and aerosol kinematics. To date, however, acinar flow simulations have been often based on geometrical models inspired by morphometrical studies; limitations in the spatial resolution of lung imaging techniques have prevented the simulation of acinar flows using 3D reconstructions of such small structures. In the present study, we use high-resolution, synchrotron radiation-based X-ray tomographic microscopy (SRXTM) images of the pulmonary acinus of a mouse to reconstruct 3D alveolar airspaces and conduct computational fluid dynamic (CFD) simulations mimicking rhythmic breathing motion. Respiratory airflows and Lagrangian (massless) particle tracking are visualized in two examples of acinar geometries with varying size and complexity, representative of terminal sacculi including their alveoli. The present CFD simulations open the path towards future acinar flow and aerosol deposition studies in complete and anatomically realistic multi-generation acinar trees using reconstructed 3D SRXTM geometries.     

MRI of pulsatile CSF motion within arachnoid cysts

Flow void due to pulsatile motion of cerebrospinal fluid (CSF) has recently been demonstrated by a variety of magnetic resonance techniques with sensitivity to slow flow. It has been suggested that within fluid collections not communicating with the physiologic CSF space, there is less signal loss than with the physiologic CSF spaces. Utilizing the SSFP MR technique, which is sensitive to flow as slow as 1 mm/sec, we evaluated three patients with isolated arachnoid cysts. Irregular signal loss consistent with fluid motion was noted within all of the cysts, as well as within the physiologic CSF spaces. Definitive anatomic evaluation of these lesions, though, required ventriculography, an invasive technique.     

Visualization of multiple scalar and velocity fields in a lifted jet flame

The stabilization of lifted jet diffusion flames has long been a topic of interest to combustion researchers. The flame and flow morphology, the role of partial premixing, and the effects of large scale structures on the flame can be visualized through advanced optical imaging techniques. Many of the current explanations for flame stabilization can benefit from the flow and flame information provided by laser diagnostics. Additionally, the images acquired from laser diagnostic experiments reveal features invisible to the eye and line-of-sight techniques, thereby allowing a deeper insight into flame stabilization. This paper reports visualizations of flame and flow structures from Particle Image Velocimetry (PIV), Planar Laser-Induced Fluorescence (PLIF) and Rayleigh scattering. The techniques are surveyed and the success of visualization techniques in clarifying and furthering the understanding of lifted-jet flame stabilization is discussed.     

Evaluation of shear and recirculation in centrifugal artificial heart by flow visualization

A centrifugal blood pump for artificial heart has to have low hemolysis and no thrombus formation. For finding and removing the fluid dynamic causes of the above mentioned phenomena, flow visualization was utilized, as flow visualization can identify the high shear and stagnation locations which relate to hemolysis and thrombus formation respectively. In the present paper, general fluid dynamic characteristics of centrifugal blood pumps will be extracted, through analyzing four typical centrifugal pumps with flow visualization technique. Finding shear in the volute and diffuser regions, confirming vortex formation at washout holes, and quantitative analysis of the flow in the inlet regions are the shown examples. The obtained results correspond well to hemolysis tests, which proves that flow visualization is a useful tool in developing and analyzing blood pumps for artificial hearts.     

An improved, free surface, topographic technique

Current techniques of water wave visualization such as shadowgraphy and stereo photography are widely used but are deficient in many aspects. Refraction based visualization observes the bending of light as it traverses across a liquid-air interface. This work describes the continued development of techniques to measure the surface height of a liquid free surface. The method, Reference Image Topography, utilizes refraction of light at the free surface as a function of the local angle of that surface. Particle Image Velocimetry (PIV) software is used to evaluate apparent dislocations of the target image viewed through the free surface, which are approximately proportional to the surface angle. High-resolution images are presented of the dynamic surface topography for a point source and the shallow water flow around a vertical cylinder.     

Segmentation of CT brain images using unsupervised clusterings

In this paper, we present non-identical unsupervised clustering techniques for the segmentation of CT brain images. Prior to segmentation, we enhance the visualization of the original image. Generally, for the presence of abnormal regions in the brain images, we partition them into 3 segments, which are the abnormal regions itself, the cerebrospinal fluid (CSF) and the brain matter. However, for the absence of abnormal regions in the brain images, the final segmented regions will consist of CSF and brain matter only. Therefore, our system is divided into two stages of clustering. The initial clustering technique is for the detection of the abnormal regions. The later clustering technique is for the segmentation of the CSF and brain matter. The system has been tested with a number of real CT head images and has achieved satisfactory results.