Digital Particle Image Velocimetry Research Articles

Failure mechanics of snow layers through image analysis

This paper is devoted to analyze the micro-mechanical behavior of snow layers subjected to shear force. A number of cold laboratory tests were performed on reconstructed snow samples on a specially designed direct shear apparatus and the loading process leading to the failure was recorded by a high-speed camera. The specimens were prepared under isothermal conditions and immediately tested in order to avoid sintering between the snow grains. Image analysis techniques (digital image correlation and particle image velocimetry) were applied on the camera records in order to obtain the snow strain history (during the loading phase) and the velocity field after the failure. Sequences of jamming and unjamming phases with the formation of vortices were observed. These observations confirm that snow layer failure process is similar to the one observed in other natural materials (e.g., sands). These results give an insight into triggering and failure propagation in a weak snow layer.

Efficiency of a Digital Particle Image Velocimetry (DPIV) Method for Monitoring the Surface Velocity of Hyper-Concentrated Flows

Digital particle image velocimetry records high resolution images and allows the identification of the position of points in different time instants. This paper explores the efficiency of the digital image-technique for remote monitoring of surface velocity and discharge measurement in hyper-concentrated flow by the way of laboratory experiment. One of the challenges in the application of the image-technique is the evaluation of the error in estimating surface velocity. The error quantification is complex because it depends on many factors characterizing either the experimental conditions or/and the processing algorithm. In the present work, attention is devoted to the estimation error due either to the acquisition time or to the size of the sub-images (interrogation areas) to be correlated. The analysis is conducted with the aid of data collected in a scale laboratory flume constructed at the Hydraulic laboratory of the Department of Civil, Environmental, Aerospace and of Materials Engineering (DICAM)—University of Palermo (Italy) and the image processing is carried out by the help of the PivLab algorithm in Matlab. The obtained results confirm that the number of frames used in processing procedure strongly affects the values of surface velocity; the estimation error decreases as the number of frames increases. The size of the interrogation area also exerts an important role in the flow velocity estimation. For the examined case, a reduction of the size of the interrogation area of one half compared to its original size has allowed us to obtain low values of the velocity estimation error. Results also demonstrate the ability of the digital image-technique to estimate the discharge at given cross-sections. The values of the discharge estimated by applying the digital image-technique downstream of the inflow sections by using the aforementioned size of the interrogation area compares well with those measured.

An analytical model of final average droplet size prediction of wave spray cloud

This study introduces an analytical model to predict the final average droplet size within a spray cloud that is formed due to wave interactions with marine objects. The process of spray cloud formation from wave impact involves several complicated processes, and each of these phenomena should be taken into account when considering an analytical approach. Following this process-based approach, the model first considers the wave characteristics and relates wave properties to the maximum impact pressure also considering the effect of air entrapment occurrence. Secondly, the maximum impact pressure is correlated empirically to the maximum wave run-up velocity. Finally, wave run-up velocity is linked with water sheet breakup models, which leads to the prediction of the final average diameter of ligaments and droplets. The prediction model is compared with experimental measurements from experiments conducted in a tow tank to form a spray cloud due to wave impact with two lab-scale models. Several measuring techniques such as image processing, bubble image velocimetry (BIV), and digital particle image velocimetry (DPIV) methods were implemented to measure the physical quantity of each phenomenon during the event of wave interaction. The effects of initial wave characteristics, the geometrical parameters of the impact condition and water sheet thickness at the moment of impact, on the final average droplet diameter were investigated. A sensitivity analysis of most of the principal parameters was performed to show the high veracity of the prediction model.

Feasibility Study of Anchored Fiber-Optic Strain-Sensing Arrays for Monitoring Soil Deformation beneath Model Foundation

High-accuracy deformation measurement is an essential part of geotechnical model testing, in which fiber-optic sensors have great potentialities. In this study, a 1-g plane-strain model test was performed to investigate the feasibility of using fiber Bragg grating (FBG) strain-sensing arrays for monitoring soil deformation beneath a strip footing. The FBG array with specially made anchors was designed to enhance the fiber-soil interfacial bond and, hence, to improve the measurement quality. Three arrays were embedded horizontally within the soil to monitor internal linear strains, while digital photography-based particle image velocimetry (PIV) was employed to obtain superficial displacement and strain fields. Test results show that the strains captured by FBGs were comparable with equivalent strains determined via PIV analyses in terms of strain development, which reflected the evolution of soil deformation under incremental loads. Finally, the benefits and drawbacks of anchored FBG arrays for monitoring laboratory-scale models were discussed, with the conclusion being that they are capable of capturing internal strains or a strain profile of soil with low noise and high resolution, but there has to be a trade-off between robustness and sensitivity.

Vortical structures and behaviour of an elliptic jet impinging upon a convex cylinder

A study on a Reh = 2100, AR = 3 elliptic jet impinging upon different convex cylinders at a jet-to-cylinder separation distance of H/dh = 4 has been conducted. Laser induced fluorescence (LIF) and digital particle image velocimetry (DPIV) techniques were utilized to investigate the effects of cylinder-to-jet diameter-ratio (i.e. D/dh = 1.15, 2.3 and 4.6) and jet orientation upon the vortical structures and their behaviour. Results show that comparable flow developments occur along both convex surfaces and straight-edges when the elliptic jet minor-plane is aligned with the cylindrical axis (i.e. EJ1 configuration), while more non-uniform flow behaviour occurs when the elliptic jet major-plane is aligned with cylindrical axis (i.e. EJ2 configuration). Additionally, significant vortex engulfment behaviour between adjacent ring-vortices upon impingement is observed along the elliptic jet minor-plane regardless of exact impingement configuration, which subsequently leads to different flow modes. Braid vortices play a surprisingly interesting role, where they lead to cross-stream and upstream vortical motions for D/dh = 1.15 and 2.3 cylinders under EJ1 configuration. In contrast, they interact with adjacent jet ring-vortices and rib structures for D/dh = 4.6 cylinder. Proper orthogonal decomposition (POD) analyses provided additional information on the unique vortical behaviour identified in the flow fields, while momentum thickness profiles characterize the mixing layers in relation to the different configurations. Lastly, wall shear stress distributions under the above-mentioned flow conditions have also been determined and related to the vortical structures and behaviour.

The effects of wing twist in slow-speed flapping flight of birds: trading brute force against efficiency

In aircraft propellers that are used to propel aircraft forward at some speed, propeller blade twist is important to make the individual propeller ‘wings’ operate at a relatively constant effective angle of attack over the full span. Wing twist is sometimes also assumed to be essential in flapping flight, especially in bird flight. For small insects, it has however been shown that wing twist has little effect on the forces generated by a flapping wing. The unimportance of twist was attributed to the prominent role of unsteady aerodynamic mechanisms. These were recently also shown to be important in bird flight. It has therefore become necessary to verify the role of wing twist in the flapping flight of birds.The aim of the study is to compare the efficiency and the aerodynamic forces of twisted and non-twisted wings that mimic the slow-speed flapping flight of birds. The analyses were performed by using physical models with different amounts of spanwise twist (0°, 10°, 40°). The flow was mapped in three-dimensions using digital particle image velocimetry. The spanwise circulation, the induced drag, the lift-to-drag ratio and the span efficiency were determined.Twist and Strouhal number (St) both determine the local effective angles of attack of the flapping wing. Wings with low average effective angles of attack (resulting from high twist and/or low St) are more efficient, but generate significantly lower aerodynamic forces. High average effective angles of attack result in lower efficiency and high aerodynamic forces. Efficiency and the magnitude of aerodynamic forces are competing parameters. Wing twist is beneficial only in the cases where efficiency is most important—e.g. in cruising flight. Take-off, landing and maneuvering, however, require large and robust aerodynamic forces to be generated. The additional force comes at the cost of efficiency, but it enables birds to perform extreme manoeuvres, increasing their overall fitness.

The effect of blade pitch on the flow dynamics inside the rotor of a three-straight-bladed cross-flow turbine

This study analyses the variations in flow dynamics inside the rotor of a three-straight-bladed cross-flow turbine for different blade pitches. A water towing tank facility has been used with a turbine model based on symmetric NACA-0015 profiles with a chord-to-diameter ratio of 0.16. The set-up was especially designed to measure the flow field inside and around the rotor, using planar digital particle image velocimetry. The experiments were made at a constant turbine diameter Reynolds number, [Formula: see text], of 6.1 × 104. The forced rotation of the cross-flow turbine using a direct current motor allowed changes in the ratio of the blade tangential speed to the freestream velocity (tip speed ratio), covering the range of 0.7–2.3. Toe-in and excessive toe-out angles have been associated in the past to low performances in this type of turbines. Pitch angle cases with expected low performances have been chosen, namely, [Formula: see text] of 8° toe-in and 16° toe-out, as well as a case with a high expected efficiency of 8° toe-out and another with [Formula: see text] as the reference case. The the goal is to better understand the flow dynamics and the blade–wake interactions inside the rotor of this type of machines, by studying cases with very different expected performances.

Studies on Liquid Flow through the Model of the Crystallizer Segment

In the sugar industry, in the process of crystallizing the suspension of the afterproduct (crystals III), among others cooling crystallizers with vertically moving tubular cooling elements are utilized. According to the designers and users, in this crystallizer type there is potential for heat transfer intensification leading to a significant reduction of the apparatus dimensions. Liquid flow through the model of a single tubular segment (part of the moving cooling element) of the cooling crystallizer was studied. An experimental model of the crystallizer segment was built. The model was utilized for the experimental research on liquid flow, conducted in parallel with the simulation of the segment operation. The method of digital particle image velocimetry was applied to analyze the experimental flow. The experimental results enabled verification of the assumptions of the numerical model based on CFD code Fluent. The finite volume method was applied in the numerical simulation. The results of experimental research and of numerical simulations were presented to improve understanding of physical conditions that promote or disturb homogenous velocity distribution of the liquid flowing around the segment of the crystallizer model.

Experimental investigation of aerodynamic forces and flow structures of bionic cylinders based on harbor seal vibrissa

In the present study, we investigated a new type of non-uniform surface cylinder based on a harbor seal vibrissa to reduce aerodynamic forces and suppress vortex shedding. Three vibrissa-based bionic cylinder models (i.e., models #1, #2, and #3) with different undulation wavelengths were manufactured and tested in a wind tunnel. The wind tunnel experiments were conducted at a Reynolds number of Re ≈ 50,000 based on the hydrodynamic diameter (Dh) and incoming airflow speed. The control effects on the aerodynamic forces and flow structures around the bionic cylinders were studied and compared with those of a baseline circular cylinder. In addition to measuring the surface pressure distributions using an array of digital pressure transducers, a high frequency force balance was used to determine the aerodynamic forces acting on the test models. Moreover, a digital particle image velocimetry (PIV) system was utilized to quantify the stream-wise flow structure and span-wise flow field to assess the effectiveness of the bionic cylinder models. The results revealed that a bionic non-uniform surface distribution could reduce the mean drag and suppress the fluctuating lift to a certain extent. Bionic cylinder model #2, which had a wavelength of two minor axes, worked best among the three models, and it was found to achieve a drag reduction of 15% and a maximum fluctuating lift suppression of 58% when the angle of the incoming airflow was 0°. The PIV results indicated that the non-uniform surface structure would disrupt the consistency of span-wise flow and decrease the span-wise correlation in the near wake of the test models, which could decrease the turbulent kinetic energy, mean drag, and fluctuating lift of the test models.

Investigation on Drag Reduction Characteristics and Shell‐Side Hydrodynamics in Hollow Fiber Membrane Modules

AbstractA novel hollow fiber membrane module shell‐side model is fabricated, simulated, tested, and analyzed. The mathematical model consists of a six spaced bar‐shaped fiber and a vortex‐control tangential entrance with protuberance radians. The vortex shedding behavior is captured by the digital particle image velocimetry technique. The velocity vectors and vorticity fields are analyzed experimentally. Meanwhile, the turbulent model is established to identify the flow stagnate point/region transformation. Investigation of wall shear stress and surface pressure distributions around circular cylindrical fibers illustrate the mechanism of drag reduction tentatively. Afterward, the novel optimization module produces higher recovery rates of up to 1.56 times as compared to that of a conventional module in equivalent filtration condition. It indicates that by means of a hydrodynamics flow control strategy, the fabricated membrane module structure could potentially improve energy utilization efficiency in filtration processes.

Quantitative analysis of spatial irregularities in RBCs flows

The spatial irregularities of red blood cells (RBCs) in continuous pulsing flow condition in micro-channels were investigated by analyzing the time variability of optical signatures obtained by recording transmitted light variability at specific location in micro-flow channels. Different flow conditions were filmed and analyzed by the digital particle image velocimetry (DPIV) to characterize local flow velocity across the whole micro-channel and in four sub-areas selected to study the particles behaviors close to the walls and in the micro-channel bulk. Starting from a behavioral classification based on the three flow patterns identified as {Weak Activity, Vorticity, Alignment}, an analysis to detect the spatial irregularity in the flow distribution was carried out. The velocity gradients and four nonlinear parameters (shear rate, strain rate, vorticity, divergence) were computed from the time-varying velocity maps obtained by DPIV and used to provide a quantitative characterization of the flow features. The comparison of the results obtained in the four experiments has made possible an overall understanding of the RBC movements in different conditions and, as well, the establishment of a analysis procedure for flows spatial irregularity detection.

Quantitative flow visualization in the distributor of a plate-fin heat exchanger

The flow distribution and head loss in the distributor of a plate-fin heat exchanger were experimentally investigated using quantitative and qualitative flow visualization methods. The flow distribution and head loss are important parameters related to the thermal performance and economic design of the heat exchangers. The effect of the different distributor angles (30°, 45°, and 60°) and the alignment configuration at the junction in the distributor were studied. The Reynolds number based on the hydraulic diameter was varied from 100 to 1200. Flow visualization using the dye injection and hydrogen bubble methods was applied to capture the qualitative flow structure inside and around the distributor. Digital particle image velocimetry was used to measure the instantaneous velocity fields, and the flow maldistribution was quantified by using the averaged velocity information. The qualitative flow visualization showed the variation of streamwise velocity in the end of the distributor and the existence of the secondary flow at the junction. From the velocity and pressure measurement, we found that the 30° model has the worst flow distribution and the smallest head loss among other models and the 60° model with misaligned configuration has the best flow distribution and the largest head loss.

Effects of superhydrophobic surfaces on the flow around an NACA0012 hydrofoil at low Reynolds numbers

In the present study, the effects of superhydrophobic surface on the flow around an NACA0012 hydrofoil are experimentally investigated at low Reynolds number range of 0.2–1.0 $$\times 10^{4}$$ . The velocity fields were measured using two-dimensional digital particle image velocimetry in a water tunnel while varying the angle of attack from $$0^\circ$$ to $$20^\circ$$ . The spray-coating of hydrophobic nanoparticles was used to create superhydrophobic surfaces. Depending on the Reynolds number and angle of attack, we found that the effects of superhydrophobic surface show up differently, which is determined by the relative strength of turbulence caused by both the surface slip (influence of trapped air pockets) and roughness, compared to that of background (i.e., uncontrolled) flow. In general, the superhydrophobic surface imposes a little influence on the wake behind a hydrofoil when the angle of attack is very low (attached flow) or high (fully separated flow). At intermediate angles of attack (flow separates between the leading and trailing edges), however, it is measured that the flow over superhydrophobic surface has a stronger turbulence, and thus, the enhanced shear-layer instability forces the early vortex rollup in the wake and reduction of vortex formation length. Interestingly, there is a transitional range of angle of attack, in which this effect is reversed, and thus, the vortex rollup is slightly delayed. This trend can be explained based on the changes in the uncontrolled flow structures, and finally, we classify the effects of superhydrophobic surface in terms of Reynolds number and angle of attack, in the considered ranges of both.

Interaction of a rigid beam resting on a strong granular layer overlying weak granular soil: Multi-methodological investigations

In the geotechnical and terramechanical engineering applications, precise understandings are yet to be established on the off-road structures interacting with complex soil profiles. Several theoretical and experimental approaches have been used to measure the ultimate bearing capacity of the layered soil, but with a significant level of differences depending on the failure mechanisms assumed. Furthermore, local displacement fields in layered soils are not yet studied well. Here, the bearing capacity of a dense sand layer overlying loose sand beneath a rigid beam is studied under the plain-strain condition. The study employs using digital particle image velocimetry (DPIV) and finite element method (FEM) simulations. In the FEM, an experimentally characterised constitutive relation of the sand grains is fed as an input. The results of the displacement fields of the layered soil based DPIV and FEM simulations agreed well. From the DPIV experiments, a correlation between the slip surface angle and the thickness of the dense sand layer has been determined. Using this, a new and simple approach is proposed to predict theoretically the ultimate bearing capacity of the layered sand. The approach presented here could be extended more easily for analysing other complex soil profiles in the ground-structure interactions in future.

Flow-induced vibration control of a circular cylinder using rotational oscillation feedback

The effect of imposed rotation on a slender elastically mounted circular cylinder free to oscillate transversely to the incident flow has been studied experimentally in a free-surface water channel. Rotation rate and direction are imposed to be proportional to either the cylinder’s transverse displacement or the cylinder’s transverse velocity to determine the effectiveness of these rotation laws to control the dynamics of the cylinder, either to reduce or to enhance oscillations. The former can be of interest for energy harvesting purposes whereas the latter can be useful to avoid unwanted oscillations. In all cases, non-dimensional mass and damping are fixed ($m^{\ast }=11.7$,$\unicode[STIX]{x1D701}=0.0043$) so the analysis is focused on the role of the rotation law and the reduced velocity. The Reynolds number based on the diameter of the cylinder ranges from 1500 to 10 000. Results are presented in terms of steady-state oscillation characterization (say, amplitude and frequency) and wake-pattern topology, which was obtained through digital particle image velocimetry. Both laws are able to either reduce or enhance oscillations, but they do it in a different way. A rotation law proportional to the cylinder’s displacement is more effective to enhance oscillations. For high enough actuation, a galloping-type response has been found, with a persistent growth of the amplitude of oscillations with the reduced velocity that shows a new desynchronized mode of vortex shedding. On the other hand, a rotation law proportional to the cylinder’s transverse velocity is more efficient to reduce oscillations. In this case only vortex-induced-type responses have been found. A quasi-steady theoretical model has been developed, which helps to explain why a galloping-type response may appear when rotation is proportional to cylinder displacement and is able to predict reasonably the amplitude of oscillations in those cases. The model also explains why a galloping-type response is not expected to occur when rotation is proportional to the cylinder’s velocity.

Experimental Measurement of Dolphin Thrust Generated during a Tail Stand Using DPIV

Estimation of force generated by dolphins has long been debated. The problem was that indirect estimates of force production for dolphins resulted in low values that could not be validated. Bubble digital particle image velocimetry (DPIV) measured hydrodynamic force production for swimming dolphins and demonstrated high force production. To validate the bubble DPIV and reconcile force production measurements, two bottlenose dolphins (Tursiops truncatus) performing tail stands were measured with bubble DPIV. Microbubbles were generated from a finely porous hose and compressed air source. Displacement of the bubbles by the propulsive motions of the dolphin was tracked with a high-speed video camera. Oscillations of the dolphin flukes generated strong vortices and a downward directed jet flow into the wake. Application of the Kutta–Joukowski theorem measuring vortex circulations yielded forces up to 997.3 N. Another video camera recorded body height above the water surface to determine the mass-force of the dolphin above the water surface. For the dolphin to hold its position above the water surface, the mass-force approximately balanced the vertical hydrodynamic force from the flukes. The results demonstrated the fluke motions generate high sustained forces roughly equal to the dolphin’s weight out of the water. Bubble DPIV validated high forces measured previously for thrust generated in swimming by animals and demonstrated a more accurate technique compared to standard aerodynamic analysis.

Single-camera three-dimensional tracking of natural particulate and zooplankton

We develop and characterize an image processing algorithm to adapt single-camera defocusing digital particle image velocimetry (DDPIV) for three-dimensional (3D) particle tracking velocimetry (PTV) of natural particulates, such as those present in the ocean. The conventional DDPIV technique is extended to facilitate tracking of non-uniform, non-spherical particles within a volume depth an order of magnitude larger than current single-camera applications (i.e. 10 cm × 10 cm × 24 cm depth) by a dynamic template matching method. This 2D cross-correlation method does not rely on precise determination of the centroid of the tracked objects. To accommodate the broad range of particle number densities found in natural marine environments, the performance of the measurement technique at higher particle densities has been improved by utilizing the time-history of tracked objects to inform 3D reconstruction. The developed processing algorithms were analyzed using synthetically generated images of flow induced by Hill’s spherical vortex, and the capabilities of the measurement technique were demonstrated empirically through volumetric reconstructions of the 3D trajectories of particles and highly non-spherical, 5 mm zooplankton.

Near-wake characteristics of rigid and membrane wings in ground effect

Wind tunnel measurements are conducted at a Reynolds numbers of Re=56,000 to examine the characteristics of near-wake vortices of rigid flat-plates and membrane wings from free-flight into ground-effect conditions. Synchronised high-speed load cell measurements, digital image correlation and particle image velocimetry are performed to resolve lift, drag and pitch oscillations simultaneously with membrane deformation and flow dynamics. Flow measurements are acquired in a crossflow-plane, one chord downstream of the trailing-edge, allowing the examination of time evolution of the wake and its relationship to the forces and membrane deformation. Membrane wings are found to delay ground-effect or high angles-of-attack induced tip-vortex break-down and result in larger tip-vortex push-out (beyond the wing span) compared to rigid flat-plate wings. The leading-edge vortex appears to shed with streamwise vorticity close to the root of the wing and this frequency of this shedding is found to match with the dominant frequency observed in membrane fluctuations. In specific resonance conditions, membrane and flow fluctuations are found to correlate well to the fluctuations in loads and moments.

Patterns for efficient propulsion during the energy evolution of vortex rings

Compared with steady-jet propulsion, pulsed-jet propulsion can exhibit more efficient by manipulating the unsteady formation of vortex rings in the near wake. That is, energy is much better transferred and used in the form of vortices to generate propulsive force rather than a steady jet. This study is aimed at analyzing the energy evolution of vortex rings and revealing the patterns for efficient propulsion during the energy evolution. Of particular importance to the energy evolution of vortex rings is the pinch-off mechanism, which is owing to the limiting effects of energy and causes that vortex rings cannot grow indefinitely. Canonical vortex rings are generated by using a piston–cylinder apparatus and their time-dependent flow fields are measured using digital particle image velocimetry. During the energy evolution of vortex rings, overpressure and fluid flux at the exit plane are found to significantly contribute to their energy. Moreover, overpressure is the dominant pattern contributing to the energy of vortex rings and has a greater contribution than fluid flux at the early evolution stage, whereas the contribution of overpressure nearly disappears once vortex rings pinch off. By contrast, fluid flux continuously contributes to the energy of vortex rings until they physically separate from the trailing vortices. Based on the hyperbolic LCSs, distinguishable flow patterns are detected to correspond to the energy evolution of vortex ring. The appearance of a newly disconnected LCS in the trailing jet indicates the initial roll-up of a trailing vortex and the onset of pinch-off, which coincides with the end of overpressure contribution to the energy of vortex rings. A LCS barrier behind the vortex rings appears and prevents any fluid flux from entering the vortex rings, thereby indicating the end of pinch-off and corresponding to the end of fluid-flux contribution. Moreover, it is suggested that transferring or using energy by pressure is a more efficient pattern than fluid flux to generate hydrodynamic force, and enlarging the energy contribution to vortices can further enhance propulsive performance.

Flow patterns through vascular graft models with and without cuffs.

The shape of a bypass graft plays an important role on its efficacy. Here, we investigated flow through two vascular graft designs–with and without cuff at the anastomosis. We conducted Digital Particle Image Velocimetry (DPIV) measurements to obtain the flow field information through these vascular grafts. Two pulsatile flow waveforms corresponding to cardiac cycles during the rest and the excitation states, with 10% and without retrograde flow out the proximal end of the native artery were examined. In the absence of retrograde flow, the straight end-to-side graft showed recirculation and stagnation regions that lasted throughout the full cardiac cycle with the stagnation region more pronounced in the excitation state. The contoured end-to-side graft had stagnation region that lasted only for a portion of the cardiac cycle and was less pronounced. With 10% retrograde flow, extended stagnation regions under both rest and excitation states for both bypass grafts were eliminated. Our results show that bypass graft designers need to consider both the type of flow waveform and presence of retrograde flow when sculpting an optimal bypass graft geometry.