Table of contents

Particle Image Velocimetry (PIV) has gone through a period of rapid development in recent years, in part due to improved laser sources and cameras, but also due to increasing experience with specific applications and with the processing of data into meaningful quantities for fluid mechanics. The result is evident in the temporal and spatial resolution now achieved with PIV and the corresponding widening of application horizons.

Measurement Science and Technology (MST) has always considered itself a home to authors of PIV developments and applications. In the December 1997 issue (vol. 8, no. 12) Dr J Kompenhans of the DLR, Göttingen, assisted in assembling a special issue of MST dedicated to state-of-the-art developments in Particle Image Velocimetry. From 29-31 May 2000 Dr M Trinité and Dr B Lecordier of CORIA, Rouen, hosted the Euromech 411 on Development of 3C Stereoscopic and Holographic Techniques - Application of PIV to Turbulence Measurements and subsequently acted as guest editors to assemble a feature issue of MST comprising selected papers from that meeting (vol. 12, no. 9, September 2001). Therefore, when plans were announced to hold the 4th International Symposium on Particle Image Velocimetry at DLR, Göttingen, on 17-19 September 2001, it was quite natural to approach selected authors with an invitation to submit their manuscripts for another special issue of MST. The present issue is the result of those invitations.

Perhaps it is worthwhile to explain that all the contributors to this issue were invited to submit their conference manuscripts; however, these were nevertheless subjected to the normal refereeing procedure used at MST - at least two anonymous referees. We feel that this is the only way in which the high standards that our readers are accustomed to can be maintained, while still associating ourselves with conference events as a source of manuscripts. The benefit we see is a concise but broad coverage of the subject area in a timely manner.

Dr Kompenhans again assisted in organizing this special issue and our thanks are extended to him for his excellent work. Furthermore, it is never tiresome to draw attention to the fact that a special issue demands exceptional discipline from authors, referees and the publishing staff, simply because everyone has to finish their job at the same time. Therefore a special thanks to all individuals involved is due for this. Hopefully, the compensation in the form of this excellent collection of papers on Particle Image Velocimetry will be considered reward enough!

Holography is truly the key to three dimensions in particle image velocimetry, i.e. the measurement of all spatial components of the velocity vector - and this over a deep measuring field. Sophisticated instruments have been designed that successfully tackle practical problems such as the low scattering efficiency of particles, the inferior depth resolution or the aberrations and distortions in the reconstruction. Furthermore, efficient strategies are introduced to interrogate the holographic storage and process the huge amount of data towards a final flow field representation. Recently, phase-sensitive metrology, familiar in many fields of experimental mechanics, has been examined for use in particle velocimetry. Suitable methods are holographic and speckle interferometry or the optical processing of data for three-dimensional correlation. While in these techniques the power of optics is unrivalled, the practical advantage of video and digital techniques over photographic recording is obvious. The electronic version of speckle interferometry (ESPI/DSPI) is a well-established method used in laser metrology and has received further exploitation for applications in flow analysis recently. Finally, the state-of-the-art of digital particle holography is reviewed to allow estimates of its future in experimental flow analysis.

We present a procedure for analysing images obtained in particle image velocimetry. This consists of weighting, re-shaping and re-orientating the interrogation windows to improve the measurements in flow regions exhibiting velocity differences in space as in shear flows. The procedure is applied recursively together with that used to determine the optimal window offset and does not require any window deformation. Tests are performed on synthetic images in different conditions (on a two-dimensional wall shear flow provided in the web by the Visualisation Society of Japan) and on images acquired in the near field of a turbulent axisymmetric water jet (at a Reynolds number equal to 8000). Under these conditions, the new algorithm substantially improves the results obtained with the standard procedure. The use of an offset procedure with a variable iteration scheme (not simple halving) is also considered; although it introduces some degree of arbitrariness, sometimes it can also improve the quality of the results in comparison to the standard procedures.

Multigrid particle image velocimetry (PIV) is an open path in the search for high-resolution PIV methods. It is based on an iterative scheme that uses the information of initial processing to adapt the method parameters in order to improve the measurements. This is mainly performed by reducing the size of the interrogation windows and shifting them. In multigrid PIV, two sources of error can significantly affect the final measurement quality: (1) the error coming from the amplitude response of the initial large interrogation windows to spatial frequencies; (2) the error originating from the truncation of particles at the borders of the final small interrogation windows. By applying weighting functions and using symmetric direct correlation both errors can be reduced, respectively. These techniques have been separately tested in the past, but a joint implementation has not yet been analysed. This task is fulfilled and both sources of error are further clarified. For this purpose, a one-dimensional single wavelength displacement field is used. This gives us the opportunity to analyse the non-linear behaviour of PIV, together with the influence of basic parameters on it. In addition to this, the multigrid method, so far described, is enhanced by compensation of the particle pattern deformation. The metrological performance of this advanced method is tested using synthetic images and the results are compared with those delivered by established PIV methods. Coherence between these results and those obtained in a real image is also detailed.

When the angular displacement method is used in stereoscopic particle image velocimetry, a projection algorithm is required to map the image coordinates onto physical coordinates. One difficulty is to find an accurate mapping function which describes the projection. A new approach for determining a mapping function from calibration-grid images is presented. The image of a Cartesian grid is processed using a windowing technique. The image plane is covered by small overlapping windows. The Hough transformation is used to detect straight line segments in each window. The straight line segments correspond locally to the possibly curved grid lines in the image. Then the data of these line segments are used to compute the coefficients of the mapping function. The algorithm is robust and requires only a little interactive control. Thus, an almost automatic evaluation of the calibration-grid images is possible.

A novel hierarchical processing scheme is proposed to efficiently increase the spatial resolution and dynamic range of detecting particle image displacements in PIV images. The technique takes full advantage of the multi-resolution characteristic of the discrete correlation function by starting the processing at the smallest scale and, if necessary, gradually building correlation planes into larger interrogation areas based on the result of inter-level correlation correction and validation. It is shown that the algorithm can be implemented in both direct and FFT based correlation algorithms with greatly reduced computational complexity. The technique opens new perspectives for locally adaptive super-resolution processing taking flow field, seeding, and imaging anomalies into account. Processing at the lowest scale (e.g. pixel or particle image size) allows the combination of correlation planes on any shape. Hence the proposed reverse hierarchical processing represents interrogation area optimization both in size and shape in order to maximize the correlation plane signal-to-noise ratio. The method is successfully demonstrated on experimentally obtained images.

The displacement (velocity) precision achieved with digital particle image velocimetry (PIV) was measured. The purpose of this work was to determine the precision and sensitivity of digital PIV using real rather than theoretical images at 1 and 2 mm spatial resolution. The displacement measurement precision was determined by measuring the RMS noise from 60 identical displacement distributions. This work is unique in that it uses electro-optical image shifting to create a repeatable image displacement distribution of random particle fields. The displacement variance between images is caused by the shot-to-shot variation in: (1) the particle-image fields, (2) the camera noise and (3) the variance in the correlation peak detection. In addition to magnification variations, the particle-number density, imaging-lens f-stop and image-plane position errors were varied to determine the best configuration. The results indicate that both the ensemble-mean and the RMS fluctuations of the image displacements are affected by these parameters and comparisons with results found in the literature are presented. The extents of these variations are quantified. This variance does not, of course, include errors due to random gradients and out-of-plane pairing losses, which exist in real turbulent flows.

For two decades, there has been active research to enhance the performance of particle image velocimetry (PIV) systems. However, the resulting systems are rather costly, cumbersome and delicate. In this paper, we address the design and some first experimental results of a PIV system belonging to the opposite paradigm. The miniature PIV or MPIV system features relatively modest performance. For example, our MPIV prototype has a field of view of 6 mm×5 mm and can measure about 50 velocity vectors for liquid flows slower than 1 m s^-1. However, it is considerably smaller (our MPIV could be held in ⌀40 mm×120 mm), cheaper (our MPIV total cost is less than $500), easy to handle and less hazardous (accidental exposure of the eyes to the strobe light). Potential applications include industrial velocity sensors.

The proposed MPIV system uses a one-chip-only CMOS camera with digital output. Only two other chips are needed, one for a buffer memory and one for an interfacing logic that controls the system. Images are transferred to a personal computer via its standard parallel port. No extra hardware is required (in particular, no frame grabber board is needed). In the MPIV prototype presented in this paper, a ring of 12 super-bright LEDs in line with the optical axis is used as a strobe light in a forward-scatter configuration. Experimental results are presented and discussed.

A novel non-cross-correlation based diffraction method is proposed to measure fractional particle displacement. This method directly processes the complete set of spot intensity distribution information in the image plane without iteration. After subtracting two consecutive particle image velocimetry images, image pairs are identified and each peak height reveals local displacement. This method possesses the potential of being able to calculate the accurate sub-pixel displacement at a significant reduction in computational cost and reveals sub-pixel displacement variations across finite interrogation regions.

The trajectories of heavy particles (ρ_particle/ρ_fluid>>1) are simulated in a two-dimensional free vortex flow. The results show that heavy particles, even with small diameters, cannot properly trace the fluid and develop a centrifugal motion. This behaviour leads to a rapid depletion of particles in the vortex core, which, in seed-dependent measurements such as particle image velocimetry (PIV) or laser Doppler velocimetry, produces a marked increase in measurement errors. Some examples are given and the evolution of the local concentration of particles is simulated. Synthetic images have been generated using the information about its motion in a two-dimensional free vortex flow. A wing-tip vortex is simulated, showing good agreement with experimental results present in the literature. Standard PIV measurements have been performed over the synthetic images, showing the effect of core depletion of particles on the incidence of erroneous measurements.

When making particle-image velocimetry measurements through the quartz cylinder of a reciprocating engine, the particle images are aberrated. This work quantifies the practical field-of-view and the errors in the velocity measurements caused by those aberrations. Electro-optical image shifting was used to create a repeatable particle-image displacement distribution for 60 images. Ensemble averaging of these images is used to quantify the rms errors due to the shot-to-shot variation in (1) the particle-image fields, (2) the camera noise, (3) the variance in the correlation-peak detection and (4) the particle-image aberrations. These results demonstrate that the field-of-view is restricted to the centre 66 mm of the 86 mm inside-diameter cylinder due to decreased accuracy, decreased image-to-image precision and decreased displacement-peak detectability of the image-displacement correlation. The correlation-peak detectability was degraded by both particle-image aberrations and decreased transmission of the scattered light.

Experimental data are presented for the turbulent velocity field generated during flame/solid wall interactions in explosions. The presence of turbulence in a flammable gas mixture can wrinkle a flame front, increasing the flame surface area and enhancing the burning rate. In congested process plant, any flame propagating through an accidental release of flammable mixture will encounter obstructions in the form of walls, pipe-work or storage vessels. The interaction between the gas movement and the obstacle creates turbulence by vortex shedding and local wake/recirculation, whereby the flame can be wrapped in on itself, increasing the surface area available for combustion. Particle image velocimetry (PIV) was used to characterize the turbulent flow field in the wake of the obstacles placed in the path of propagating flames. This allowed the quantification of the interaction of the propagating flame and the generated turbulent flow field. Due to the accelerating nature of the explosion flow field, the wake flows develop `transient' turbulent fields and PIV provided data to define the spatial and temporal variation of the velocity field ahead of the propagating flame, providing an understanding of the direct interaction between flow and flame.

Digital particle image velocimetry was used to study the pulsatile flow through a new mechanical heart valve prosthesis. The design of the new prosthesis with three leaflets in symmetric arrangement is similar to the human aortic heart valve. A fully transparent 1:1 model of the valve was used for laser-based flow visualization and velocity measurements to optimize the valve design. Planar velocity field measurements of the time-dependent flow field and high-speed recordings of the occluder motion were carried out in physiological pulsatile flow. The new valve shows a well-defined core flow with small wakes behind the leaflets. The specific three-leaflet design yields a desired wash-out effect of the stagnant regions in the aortic sinuses during valve opening and closure. The leaflets begin to close in an early stage of the end of the flow cycle which is promoted by induced backflow near the walls and the sinuses. This leads to a relatively smooth valve closure and a reduced impact velocity.

The gun-type burner is an oil burner frequently used for industrial and domestic applications. The oil droplets are mixed with air and the flow generated is extremely complex and requires the combination of different techniques for its complete description. Here, for the first time, the full three-dimensional velocity vector map has been measured for different droplet sizes in non-reacting conditions using stereoscopic PIV together with a technique called multi-intensity layer, that allows size discrimination by accounting for differences in the scattered light intensity. Different spatial structures are detected depending on the droplet size: droplets under 30 µm in diameter follow the incoming airflow and entrain the recirculation zones while droplets bigger than 50 µm penetrate due to their large momentum. Differences are also found for the swirl component near the baffle plate, measuring stronger swirl for droplets smaller than 30 µm.

This paper concerns the computation of derivatives from particle image velocimetry (PIV) velocity fields with the goal of obtaining the vorticity component normal to the plane. A variety of derivative schemes are characterized by their transfer function, taking into account the truncation and noise amplification. The PIV measurement noise is supposed to be a white one in the Fourier space. A spectral approach is used in order to choose the best filter for turbulent flows. The derivative spectra are discussed. An application is presented on a real turbulent flow with two interrogation window sizes and different derivative schemes. The most significant schemes are also applied to a velocity field containing a single vortex. A comparison of the maximum of vorticity obtained with each scheme and through a least-square fit with an Oseen vortex, allows us to quantify the effect of the band pass filter and to select the best scheme.

The current work demonstrates an optical technique that utilizes micron-resolution particle image velocimetry (µ-PIV) for temperature measurement. The technique is based on the premise that Brownian motion will cause width-wise broadening of the cross-correlation peak. A correlation-based PIV algorithm detects the magnitude of Brownian particle motion and can be used to determine the temperature of the fluid. Results were obtained using fluorescing (rhodamine 542/612) spherical 700 nm diameter polystyrene-latex particles in water. Temperature changes up to 25 °C were determined with an experimental accuracy of ±3 °C.

Methods of analysing and interpreting two-dimensional velocity field data in order to understand the scales of turbulence of bubbly two-phase flow were developed. Reynolds decomposition and large eddy simulation (LES) decompositions (low-pass filtering), in conjunction with proper orthogonal decomposition (POD) energy spectra analysis, as well as adjusted convective decomposition (constant convection velocity is gas bubble velocity) were applied to analyse the structure of turbulence. Particle image velocimetry was applied for velocity measurements. Decomposition analysis was performed for the local velocity around gas bubbles in a chosen region. Various decomposition methodologies were applied for interpretation of the results and it was found that the number of eddies revealed and vorticity magnitude varied with the decomposition method used. In particular, LES decomposition was found to perform better at showing smaller eddies. POD indicated the energy changes quantitatively through the spatial energy spectra, while the comparison of single-phase flow with bubbly two-phase flow offered an efficient way of decomposing the total velocity.

A combination of convective decomposition and POD was applied to the energy spectra level in order to obtain a novel view of the turbulence energy introduced by the gas bubbles.

In this communication, we describe a fast thickness profile measurement method for a transparent film, thinner than the white-light coherence length of 3-4 µm that is deposited on pattern structures. A visible acousto-optic tunable filter is employed for real-time wavelength scanning and the three-dimensional volumetric thin-film thickness profile information is obtained using a simple peak detection method in the spectral domain. The key idea is to divide the measurement into two states using a beam blocking mechanism to separately obtain the two unknowns of thickness and surface profile. Such separate measurements are required to compensate for the phase change effect caused by the multi-reflected beams from the thin film. The final thin-film surface profile information is measured by obtaining the number of peaks and phase deviations from the two separately scanned spectral intensity values.

This paper describes the design of a high resolution low cost imaging system for the analysis of high speed particle projection. This system, based on a camera and a set of flashes, is used to characterize the centrifugal spreading of fertilizer particles ejected at speeds of ≈30 m s^-1. Multiexposure images collected with the camera installed perpendicular to the output flow of granules are analysed to estimate the trajectories of the fertilizer granules. Very good results are obtained with the Markov random fields method, in comparison with others.

We present a novel method to study the behaviour of the optical properties of photopolymer materials with temperature. The photopolymer is deposited on the tip of optical fibres by dip coating to fabricate low-finesse Fabry-Perot microcavities. The signal processing technique utilized to interrogate the cavity is based on the generation of two quadrature phase-shifted interferometric signals using two Bragg fibre gratings. This technique enables the determination of the values of the thermo-optical coefficient and the linear coefficient of thermal expansion of the photopolymer. The effectiveness of the processing technique is also exploited in the study of the dependence of the temperature sensitivity on the cavity thickness.

The calibration of five-hole probes has been limited in the range of accepted angles of calibration. Potential flow theory is used to model a traditional method of calibration, and the theoretical limitations and consequences are examined. A theoretical basis for extending the range of angles included in calibration is subsequently given.

The high-frequency components of engine block vibration signals during misfire and normal combustion show clear differences of duration and scale by wavelet analysis - continuous wavelet transform (CWT). A new factor, combustion noise intensity (`CNI_CWT'), has been defined, which has been derived from the CWT scalogram. It has been shown that CNI_CWT can detect misfire more accurately than fast Fourier transform power spectral density. Similarly, it has been found that CNI_CWT can also be applied to knock detection. The durations of both knock and misfire detection overlapped each other at the same wavelet scale 8; misfire, normal and knock cycles can be indicated on the same histogram by CNI_CWT. Moreover, knock and misfire detection can be combined to one CNI_CWT computation.

Thermal manikins are necessary instruments for measuring the thermal insulation and moisture vapour resistance of clothing systems, which are important parameters relevant to clothing thermal comfort. Although many thermal manikins have been developed since the first example in the 1940s, simulation of human perspiration in thermal manikins remains a challenge. This paper reports on a novel perspiring fabric thermal manikin, which simulates gaseous perspiration by moisture transfer through a `skin' made of a breathable fabric. The manikin has been used to measure the thermal insulation and moisture vapour resistances of clothing ensembles, and demonstrated high accuracy and reproducibility.

A simple analytical model for the calibration of a flux-gate magnetometer using relative sensor motion in a constant magnitude magnetic field (B) is presented.

Sensor motion is parametrized in terms of elementary rotations about one axis. The number of elementary rotations constitutes the number of degrees of freedom of the motion.

A generalization is performed by investigating cases with known/unknown B, one/several different values of B, one/several degrees of freedom. The maximum number of calibration parameters, which can be determined in each case, is established.

The conclusion is that the determination of all calibration parameters, i.e. an absolute calibration of the instrument, is already possible if the relative motion has at least two degrees of freedom at a known, constant B value.

Two experimental applications of the model are described briefly.

The frequency and temperature dependence of the surface resistance of metallic films was measured by a microwave micro-strip method under a T-junction structure. Numerical analysis of micro-strips made of silver-tin (Ag-Sn) alloy, or good conducting niobium (Nb) films reveals the surface resistances behaving as nearly a one-half power law dependence on the frequency, which is in congruence with the results derived from the free-electron model in simple metals. In addition, we have specifically investigated the electron transport with a strong localization effect on the DC temperature-dependent resistivity in abnormal and normal Nb films. The results indicate a deviation from a one-half power law may occur in the abnormal film. This work can be further exploited to measure the conductivity and penetration depth of metals in multilayered structure or of superconducting films.

A high-accuracy gas flow dilutor has been developed using mass flow controllers operating at flow rates ordered in terms of factors of two. It uses a novel self-referencing calibration procedure to generate variable dilutions with a typical accuracy of ±0.4%. There is no requirement for calibration to any external standard of flow. A detailed analysis of the uncertainty of configurations giving constant total gas and constant complementary gas flow is presented. Applications of the dilutor to the analysis of standard gas mixtures and to testing the linearity of gas analysers are demonstrated.

The study presented concerns the behaviour of an electric high current and low voltage breaking arc. The arc dynamics in a circuit breaker is analysed according to two diagnostic methods. A magnetic diagnostic based on precise measurements of induction restores the average line of current that magnetically represents the arc. This method is compared to an optical diagnostic using a high-speed mechanical camera and a CCD camera. These methods allow us to follow the movement and shape of the arc and to study the influence of the quenching chamber on the break quality. This quenching chamber is composed of splitter plates whose material can been varied.

A two-component Doppler global velocimeter (DGV) system has been improved through the use of vapour-limited iodine cells that have temperature-independent responses, along with nonpolarizing beam splitters and lower f-number lenses. Two-component DGV velocity measurements have been obtained for a 1 inch diameter uniform circular jet flow at a nominal exit velocity of 60 m s^-1, as well as for an annular jet and a swirling jet. These data generally agree with earlier point Doppler velocimeter and hot wire anemometer results to within about 2-4 m s^-1, and display a total variability from a smooth curve of ±2-3 m s^-1. This level of accuracy has been obtained for a system that uses a cw argon ion laser and eight-bit CCD cameras and digitizers. Exceptions to this level of accuracy are noted in regions of significant secondary scattering, due to scattered laser light that is reflected off the lip of the jet nozzle, as well as in regions of low smoke seeding levels, resulting in low signal-to-noise ratios. A significant amount of the variability of the data from a smooth curve is due to the flat field correction.

In this design note, an air-gapped ring core type Faraday-effect optical current sensor using a new sensing element design is described. The sensing element is a finely polished glass prism with highly reflective coatings on the two end surfaces, which is designed in such a way that the light beam propagating through it undergoes 20 critical-angle reflections at the glass-air interfaces. The current sensitivity of this new design is almost three times higher than that of a previous design based on a similar principle.

A pump-probe system using infrared-visible light is proven to have sensitivity to measure surface deformations produced by the infrared light through changes in the transmittance to visible light.