Search

In the past few decades various particle image–based volumetric flow measurement techniques have been developed that have demonstrated their potential in accessing unsteady flow properties quantitatively in various experimental applications in fluid mechanics. In this review, we focus on physical properties and circumstances of 3D particle–based measurements and what knowledge can be used for advancing reconstruction accuracy and spatial and temporal resolution, as well as completeness. The natural candidate for our focus is 3D Lagrangian particle tracking (LPT), which allows for position, velocity, and acceleration to be determined alongside a large number of individual particle tracks in the investigated volume. The advent of the dense 3D LPT technique Shake-The-Box in the past decade has opened further possibilities for characterizing unsteady flows by delivering input data for powerful data assimilation techniques that use Navier–Stokes constraints. As a result, high-resolution Lagrangian and Eulerian data can be obtained, including long particle trajectories embedded in time-resolved 3D velocity and pressure fields.

Impinging jets are widely used for heat and mass transfer because they are applicable to any type of body and can be easily implemented. They are also used in various industrial fields and design techniques. Previous research investigated e.g. a jet impinging onto a flat surface. However, since most of the mechanical parts have curvature on their surface, it is necessary to study more detailed properties of the jet impinging on a curved surface using advanced measurements. Therefore, in this study, three-dimensional flow structures of a round jet impinging on a convex cylinder surface were measured using volumetric Lagrangian particle tracking (LPT).

In recent years, Lagrangian Particle Tracking (LPT) has become more and more important in the field of 3D flow measurements, largely due to the introduction of the methods of Iterative Particle Reconstruction (IPR, Wieneke 2013) and Shake-The-Box (STB, Schanz et al. 2016). Applying the STB algorithm, particles can be tracked in large numbers, with a nearly complete suppression of ghost particles. The introduced extension to the STB method extends its working range to flows with high vleocity range, i.e. large areas within the measurement region that exhibit very slow flow velocities compared to the fastest ones, which determine the applied frame-rate

Three-dimensional Lagrangian Particle Tracking (LPT) was performed using the Shake-The-Box (STB) algorithm on a subsonic jet flow at Mach 0.845. The STB technique for multi-pulse data was employed to reconstruct particle tracks along the four-pulse sequences providing highly resolved 3D flow velocity and material acceleration data. A description of the experimental methodology is given followed by instantaneous accelerations and averaged flow statistics.

We present spatially and temporally resolved velocity and acceleration measurements of turbulent Rayleigh-Bénard convection covering the complete volume of a cylindrical sample with aspect ratio one. Using the Shake-The-Box Lagrangian particle tracking algorithm, we are able to instantaneously track up to ~ 600,000 particles in the complete sample volume (~ 1 m³). The data assimilation scheme FlowFit with continuity and Navier-Stokes-constraints is used to interpolate the scattered velocity and acceleration data by continuous 3D B-Splines in a cubic grid, enabling to recover the smallest flow scales. The frame rate was chosen to be at least 7 times faster than the Kolmogorov time scale. In the presentation, we show Lagrangian and Eulerian visualizations of the large scale circulation (LSC) as well as small scale structures, such as thermal plumes and turbulent background fluctuations and unveil the dynamics of their complex interplay. Rare dynamic events, such as cessations and reo rientations of the LSC will be presented in addition to combined and conditioned Lagrangian and Eulerian statistics of turbulent quantities.

We present measurements of the full velocity gradient and dissipation rate tensor based on dense fields of fluid particle trajectories in homogeneous turbulence at Rel~270 and ~350 in a von Kármán flow between two counter-rotating propellers. Applying the Shake-The-Box (STB) particle tracking algorithm, we are able to instantaneously track up to ~100.000 particles in a measurement volume of 50 x 50 x 15 mm³. The mean inter-particle distance is lower than 7 Kolmogorov lengths. The data assimilation scheme FlowFit with continuity and Navier-Stokes- constraints is used to interpolate the scattered velocity and acceleration data by continuous 3D B-Splines in a cubic grid, enabling to recover (locally) the smallest flow scales. In the presentation, we show Lagrangian velocity, acceleration and jolt statistics, as well as the Eulerian counterparts on velocity and pressure gradients with respective Q-R-diagrams, compute the energy dissipation rate directly by using local velocity gradient information gained by FlowFit at midpoints of particle tetrahedra in close proximity of a few Kolmogorov lengths and compare it to known indirect approaches.

The recent introduction of the Multi-Pulse Shake-The-Box technique (MP-STB [3]) extended the capabilities of the Shake-The-Box (STB [4]) 3D Lagrangian particle tracking (LPT) algorithm in terms of applicable flow velocity. In particular, the MP-STB allows for 3D LPT investigations at velocities relevant for industrial and aerodynamic applications in relatively large volumes where the adoption of a time-resolved recording strategy would result in significant compromises in terms of spatial range or resolution (in order to limit the particle displacement between subsequent recordings). Multi-pulse sequences are obtained by synchronizing multiple illumination systems to generate bursts of laser pulses (typically four) where the time separation can be freely adjusted down to less than a microsecond (e.g. for supersonic applications). A 3D imaging system consisting of dual-frame cameras is adopted, where each frame is double-exposed to record the four pulses [2]; the MP-STB technique makes use of an iterative approach to overcome the limitations posed by the short observation time offered by the multi-pulse recording sequence, and allows for the reconstruction of individual four-pulse particle tracks in 3D [3]. The acquisition of long sequences of four-pulse recordings at low frequency (10-15Hz) enables the evaluation of flow statistics. In the present study, the MP-STB technique is employed for the investigation of turbulent boundary layer (TBL) flows at relatively high Reynolds numbers. A TBL at 15 m/s (Reθ ≈ 10; 000) is investigated in the SWG facility at DLR Göttingen; approximately 95; 000 individual tracks are reconstructed within the 80×120×10 mm^3 domain for each four-pulse recording. Spatial gradients, as well as the instantaneous 3D pressure field, are evaluated by means of the FlowFit data assimilation algorithm [1] (fig.1-left). A TBL with adverse pressure gradient (APG) at 29 m/s (Reθ ≈ 30; 000) is investigated in the atmospheric wind tunnel of the University of Armed Forces in Munich; both planar PIV and MP-STB in a 80 × 90 × 7 mm^3 domain are employed; the direct comparison of the results in terms of turbulent fluctuations (fig.1-right) highlights the higher spatial resolution of the LPT approach which, unlike PIV, does not suffer from the signal modulation caused by the finite size of the cross-correlation window. In the final presentation the MP-STB processing technique will be introduced together with the main parameters affecting its performances. Moreover, results regarding the TBL structure and flow statistics (e.g. TBL profiles, two-point correlations) will be presented.

We report on an experimental undertaking with the goal of detecting so-called turbulent superstructures within a turbulent boundary layer and describing their temporal development while propagating from a region of zero pressure gradient (ZPG) into a region of adverse pressure gradient (APG). The flow was characterized by performing Lagrangian Particle Tracking (LPT) of Helium-filled soap bubbles (HFSBs) using the Shake-The-Box algorithm.

We would like to report on the techniques used around 1930 for the creation of a film consisting of sequences visualizing various forms of flow separation produced by L. Prandtl and his colleagues O. Tietjens and W. Müller in open-surface water channels. After first experiments with cinematography in the 1910s, Prandtl began to produce flow visualisation films in the 1920s, in order to better capture the dynamics of the unsteady flow phenomena. However, due to the short exposure times provided by the film camera the particles appeared as dots rather than the anticipated streaks, which were desired as representations of the streamlines. Therefore Prandtl and his colleagues first decided to work with a modified camera to provide longer exposure time and consequently streaked images. At this point, the real potential of animated film, for instance for the demonstration of flow separation, was realized and resulted in a flow visualization film, originally named "The Production of Vortices by Bodies Travelling in Water", that Prandtl showed 1927 in London and while travelling the globe from 1929 to 1930. The sequences of this film were found to be highly instructive from an educational point of view such that the film, now named C1, was made available in 1936 through the Reich Office for Teaching Films and in the 1950s by the Institute of Scientific Film. During the preparation of this paper, new movies recorded by Prandtl and his colleagues have been found in the archives. They are discussed in this paper in order to demonstrate the feasibility of such flow visualization movies for evaluation with modern PIV correlation and tracking analysis methods in a broader frame. While the film sequences were mainly intended to illustrate the temporal evolution leading to boundary layer separation and separated flow, numerous references confirm that Prandtl and his colleagues were fully aware of the quantitative nature of the acquired imagery. With much of the material preserved over the decades, it nowadays is possible to use advanced particle image velocimetry techniques such as PIV and Lagrangian particle tracking to retrieve this quantitative information and make a clear statement on the importance of proper data archival.

The flow around a surface mounted cube with incoming turbulent boundary layer has been topic of many experimental and numerical investigations in the past decades. Despite its simple geometry the flow generates a set of complex vortical structures, includes flow separation at the three front plane edges with corresponding subsequent shear layer dynamics and recirculation zones and organizes a large unsteady flow separation region downstream of the cube which is associated with typical quasi-periodic bluff-body wake dynamics. Because of its generic role for geometry induced flow separation and ist complex unsteady vortical properties the flow around a surface mounted cube is often used for benchmarking purposes to validate turbulence models and advanced numerical methods, while an increasing interest of DNS investigations in the recent years including the laminar boundary layer flow case can be registered. Furthermore, the flow configuration is well suited to enhance the understanding of similar unsteady and separated flow phenomena in many aerodynamic and engineering applications. In the present experimental investigation we aim at resolving a large spectrum of spatial and temporal scales in the flow around a cube with incoming laminar and turbulent boundary layers by using the most recent developments of 3D Lagrangian Particle Tracking (LPT) and high resolution TR-PIV. The results are partly documented in the present paper and consist of long time-series of highly resolved 3D and 2D velocity fields as well as unsteady 3D pressure volumes allowing for the analysis of coherent structure dynamics and of corresponding statistical Lagrangian and Eulerian flow properties.

We report on an experimental undertaking with the goal of detecting so-called turbulent superstructures within a turbulent boundary layer and describing their temporal development while propagating from a region of zero pressure gradient (ZPG) into a region of adverse pressure gradient (APG). The flow was characterized by performing Lagrangian Particle Tracking (LPT) of Helium-filled soap bubbles (HFSBs) using the Shake-The-Box (STB, Schanz et al. 2016) algorithm. This work will present first results, but will concentrate on the description of the experimental efforts and the evaluation approaches that were necessary to successfully conduct this investigation. The tracking of a large number of tracer particles (several 100,000) within an extensive volume (nearly three meters in length) was made possible by the use of a system consisting of 12 cameras, forming three successive volumetric systems. Calibration and evaluation methods for fusing the different systems into one large camera system are described. Challenges in calibrating camera systems with a large field-of-view due to small-scale distortion effects introduced by the observation windows are discussed and correction methods are introduced. Methods for achieving a perfect transition between overlapping camera systems are introduced. For higher flow velocities, a double illumination approach was created in order to limit the pixel displacement between frames. To the knowledge of the authors, this is the first time such an approach was realized for time-resolved image sequences.

The flow on the suction side of a half-wing model is investigated by means of 3D Lagrangian particle tracking (LPT) with Multi-Pulse Shake-The-Box and stereoscopic particle image velocimetry. The experiment is carried out in the Low Speed Wind-Tunnel facility (LSWT) at the Airbus UK site (Filton, England). Illumination is provided by a multi-pulse system consisting of two dual-pulse lasers. A 3D imaging system is composed of five double-frame cameras; a two-camera subset can be employed in stereoscopic fashion. Due to the constrains posed by the limited optical access to the measurement region, three cameras are installed inside the wind-tunnel test section. As a consequence, significant vibrations occur during the operation of the wind-tunnel, leading to large unsteady decalibration of the imaging system (up to 120 px). A tailored single-snapshot self-calibration approach is proposed here which makes use of standard volume self-calibration and particle triangulation from the 3D imaging system to detect and correct the camera decalibration for each instantaneous recording. Results shows that the disparity error can be reduced to less than 0:1 px, a suitable value for both accurate 3D particle triangulation using the full imaging system and for stereoscopic evaluation of the out-of-plane velocity component with PIV making use of a two-camera subset.

We present an experimental adverse pressure gradient turbulent boundary layer (TBL) flow investigation at high Reynolds numbers (approx. 10.000 < Re < 40.000) using large field Multi camera 2D PIV and three different particle tracking methods based on the Shake The Box (STB) technique, namely time resolved 2D and 3D STB and Multi Pulse (MP) STB. The experiments were performed within the frame of the DLR project Victoria and conducted in the Eiffel type atmospheric wind tunnel of the University of Armed Forces in Munich (AWM), which has a 22m long test section with a rectangular cross section of 1.8 x1.8 m^2. After a ramp the TBL develops along a flat plate with nearly zero pressure gradients (ZPG) to an equilibrium state before it enters into a 2D diffusor geometry following a smooth and moderate curvature into a flat plate at ~18° inclination angle while undergoing a significant adverse pressure gradient (APG) leading to flow separation. The measurements have been performed with an embedded field of view (FOV) strategy. A statistical significant number of samples for four different free stream velocities have been acquired at several positions in all pressure gradient regimes down to the region of (intermittent) flow separation. A large number of instantaneous velocity vector fields and time resolved particle tracks of the boundary layer flow have been achieved in order to gain high resolution wall normal profiles of the mean velocities and related Reynolds stresses in a first step.

Proceedings 18th International Symposium on Flow Visualization

Lagrangian particle tracking (LPT) enables the accurate measurement of position, velocity and acceleration of single seeding tracers imaged by a multiple camera system. The Shake-The-Box (STB) technique extended the capabilities of 3D particle tracking methods to higher seeding densities, thus providing access to the full velocity gradient tensor. The STB method, initially proposed for the reconstruction of time-resolved sequences of recordings, has been extended to the case of short recording sequences acquired by multi-pulse systems for the investigation of high-speed flows. The Multi-Pulse Shake-The-Box technique (MP-STB) exploits an iterative strategy to make up for the lack of time-resolved recordings and progressively simplify the reconstruction and tracking problem, therefore increasing the number of successfully identified particle tracks. Individual track fields from MP-STB can be exploited to describe the instantaneous flow structures by means of interpolation to a regular grid or collected into bins with an ensemble-average approach to produce highly spatially resolved statistics. Several applications of the MP-STB technique to turbulent boundary layer flows and to a high speed free jet in air are presented here. The experimental conditions are described together with the main processing parameters adopted for the iterative STB strategy. The multi-pulse recording sequences are obtained by means of two pulse separation techniques (polarization- and timing-based) and by adopting Multi-Exposed recordings. Results are shown in terms of instantaneous fields and flow statistics.

Proceedings 18th International Symposium on Flow Visualization

The aim of this work is to capture the spatiotemporal dynamics of large-scale flow structures in a turbulent boundary layer flow in a volumetric and time-resolved way. The applied means are dense Lagrangian Particle Tracking of Helium-Filled Soap Bubbles (HFSBs) using the Shake-The-Box (STB) algorithm and a subsequent interpolation to a regular grid using the physically regularized FlowFit method. The presented experiment was performed as a pre-test to a full campaign to be held in August 2018 at the Atmospheric wind tunnel at the University of the armed forces in Munich, Germany. This test demonstrates the principle applicability of the measurement techniques to a low-speed wind tunnel environment and shows that high spatial accuracy and resolution can be reached despite of a large measurement volume. Problems were encountered achieving a homogeneous bubble distribution and in deploying large amounts of bubbles, while not disturbing the flow.