Imaging laser Doppler velocimetry

Imaging laser Doppler velocimetry (ILDV) is a novel flow measurement technique, which enables the measurement of the velocity in an imaging plane. It is an evolution of heterodyne Doppler global velocimetry (HDGV) and may be regarded as the planar extension of the classical dual-beam laser Doppler velocimetry (LDV) by crossing light sheets in the flow instead of focused laser beams. Seeding particles within the flow are illuminated from two different directions, and the light scattered from the moving particles exhibits a frequency shift due to the Doppler effect. The frequency shift depends on the direction of the illumination and the velocity of the particle. The superposition of the two different frequency-shifted signals on the detector creates interference and leads to an amplitude modulated signal wherein the modulation frequency depends on the velocity of the particle. This signal is detected using either a high-speed camera or alternatively a smart pixel imaging array. This detector array performs a quadrature detection on each pixel with a maximum demodulation frequency of 250 kHz. To demonstrate the feasibility of the technique, two experiments are presented: The first experiment compares the measured velocity distribution of a free jet using ILDV performed with the smart pixel detector array and a high-speed camera with a reference measurement using PIV. The second experiment shows an advanced setup using two smart pixel detector arrays to measure the velocity distribution on a rotating disk, demonstrating the potential of the technique for high-velocity flow measurements.     

Application of digital speckle pattern interferometry for fluid velocimetry in wind tunnel flows

This paper shows the feasibility of using digital speckle pattern interferometry (DSPI) as a fluid velocimetry technique in high speed gaseous flows. The light scattered from an illuminated plane was recorded with a CCD camera at the same time as a uniform reference beam. A fibre optic was used to bring this reference beam from the laser cavity to the CCD camera. The comparison of two subsequent frames gives information about the velocity field. DSPI was applied to a Von Karman street flow set up in a wind tunnel. Particle image velocimetry (PIV) measurements were also obtained for comparison with the information provided by DSPI. A system for increasing the measurement region when using short coherence length lasers is proposed. Received: 15 June 2000/Accepted: 8 September 2000     

Turbulence spectra from individual realization laser velocimetry data

Two techniques — a modified correlation method and a direct transform method — have been evaluated for use in making spectral estimates from randomly-sampled random data (such as turbulence data obtained with an individual realization laser Doppler velocimeter (LDV)). The effect of bad points (which usually appear randomly in actual LDV data sets) on the spectral estimates has been studied. Necessary modifications and extensions to the techniques have been determined based on studies using simulated data with known spectral characteristics. The direct transform method is found to have certain advantages over the correlation method.From the LDV measurements taken in a simulated coal-fired magnetohydrodynamic flow field, major differences in turbulence spectra obtained from the two methods are observed in the region immediately downstream of the combustor where the flow was evidently dominated by (uncorrelated) combustion instabilities. Spectra are reported for three axial positions, and comparisons with classic turbulent pipe flow data are presented.     

Taylor length scale measurement by laser Doppler velocimetry

 A new method for making direct measurements of the spatial velocity correlation coefficient, based on two-point laser Doppler velocimetry (LDV), has been developed. In this paper, the effects of control parameters on the correlation coefficient are being investigated. The main sources of experimental error have been identified and analysed. It appears that the probe volume length has a key effect on the accuracy of Taylor micro-scale measurement. A data processing procedure has been established and validated for the determination of this scale. The procedure shows that the portion of the correlation curve used to determine Taylor scale is a function of the integral scale to Taylor micro-scale ratio. Received: 7 June 1995/Accepted: 8 December 1997     

Two-color particle-imaging velocimetry using a single argon-ion laser

A swept-beam, two-color particle-imaging velocimetry (PIV) technique has been developed which utilizes a single argon-ion laser for illuminating the seed particles in a flowfield. In previous two-color PIV techniques two pulsed lasers were employed as the different-color light sources. In the present experiment the particles in a two-dimensional shear-layer flow were illuminated using arotating mirror to sweep the 488.0-nm (blue) and 514·5-nm (green) lines of the argon-ion laser through a test section. The blue- and greenparticle positions were recorded on color film with a 35-mm camera. The unique color coding eliminates the directional ambiguities associated with single-color techniques because the order in which the particle images are produced is known. Analysis of these two-color PIV images involved digitizing the exposed film to obtain the blue and green-particle image fields and processing the digitized images with velocity-displacement software. Argon-ion lasers are available in many laboratories; with the addition of a rotating mirror and a few optical components, it is possible to conduct flow-visualization experiments and make quantitative velocity measurements in many flow facilities.List of symbols d length of displacement vector - d _m distance between rotating mirror and concave mirror - n _f number of facets on rotating mirror - R seed-particle radius - v velocity in x, y plane - v _s sweep velocity of laser beams, assumed to be in y direction from top to bottom of field of view - v _x, v _y, v _z x, y, and z components of velocity - x ₁, y ₁ color-1 particle coordinates - x ₂, y ₂ color-2 particle coordinates - y _max y dimension of field of view, assumed to be the long dimension - s spatial separation of beams as they approach rotating mirror - t time separation of laser sheets or of swept beams passing fixed point - t _b time between successive sweeps through test section by same beam - t _s time required for both beams to sweep through test section - angular separation of beams reflecting from rotating mirror - fluid viscosity - v angular velocity of rotating mirror in cycles per second - seed-particle density - seed-particle response time - _v, _d, _t standard deviation of velocity, displacement, and time - vorticity This work was supported, in part, by the Aero Propulsion and Power Directorate of Wright Laboratory under Contract No. F33615-90-C-2033.     

Acoustoelastic determination of residual stress with laser doppler velocimetry

The main object of this study is to develop a new technique for stress nondestructive measurement. A noncontact measurement technique of ultrasonic wave velocity is proposed. In the measurement system, a laser Doppler velocimeter, which is noncontact, is used to detect wave motions due to Rayleigh waves instead of a piezoelectric transducer. The noncontact measurement technique is applied to determine the stress-acoustic coefficient of Rayleigh waves for aluminum 5052 and a structural steel, and the results are in good agreement with those obtained using knife-edge piezoelectric transducers. The technique is also used to evaluate residual stress existing in an H-section rolled beam of the structural steel. The distribution of residual stress is reasonable.     

Analysis of helicopter blade vortex structure by laser velocimetry

In descent flight, helicopter external noise is mainly generated by the Blade Vortex Interaction (BVI). To under-stand the dynamics of this phenomenon, the vortex must be characterized before its interaction with the blade, which means that its viscous core radius, its strength and its distance to the blade have to be determined by non-intrusive measurement techniques. As part of the HART program (Higher Harmonic Control Aeroacoustic Rotor Test, jointly conducted by US Army, NASA, DLR, DNW and ONERA), a series of tests have been made in the German Dutch Wind Tunnel (DNW) on a helicopter rotor with 2 m long blades, rotating at 1040 rpm; several flight configurations, with an advance ratio of 0.15 and a shaft angle of 5.3°, have been studied with different higher harmonic blade pitch angles superposed on the conventional one (corresponding to the baseline case). The flow on the retreating side has been analyzed with an especially designed 3D laser velocimeter, and, simultaneously, the blade tip attitude has been determined in order to get the blade-vortex miss distance, which is a crucial parameter in the noise reduction.A 3D laser velocimeter, in backscatter mode with a working distance of 5 m, was installed on a platform 9 m high, and flow seeding with submicron incense smoke was achieved in the settling chamber using a remotely controlled displacement device. Acquisition of instantaneous velocity vectors by an IFA 750 yielded mean velocity and turbulence maps across the vortex as well as the vortex position, intensity and viscous radius. The blade tip attitude (altitude, jitter, angle of incidence) was recorded by the TART method (Target Attitude in Real Time) which makes use of a CCD camera on which is formed the image of two retroreflecting targets attached to the blade tip and lighted by a flash lamp. In addition to the mean values of the aforementioned quantities, spectra of their fluctuations have been established up to 8 Hz.This work has been achieved with the financial support and the encouragements of contracts from Ministry of Defense (DGA-DRET/6 and STPA/HE). The recognition of the authors goes also to their ONERA colleagues (G. Losfeld, D. Soulevant and R. Soares) who very actively participated in the tests at DNW.     

Unsteady flow in rotating drums using laser Doppler velocimetry

Non-destructive measurements by laser Doppler velocimetry is employed to study unsteady flow in a hollow drum filled with liquid. The drum is suddenly accelerated from rest or is suddenly decelerated from a steady rotation to rest. Pure water and glycerin-water mixtures are used as the test liquid in which polyethylenelatex particles are mixed as the light scattering tracer. The boundary layer formation, the time history of velocity, momentum and kinetic energy of the liquid, the wall-to-fluid force transfer, and the transient response time are determined. Also determined are the effects of side walls and fluid viscosity on the transient flow response. Of importance is the disclosure of Ekman layer instability near the inner radial wall of the test drum. It is actuated by the centripetal acceleration-induced buoyancy force.List of symbols A wetted surface area of test drum, cm² - a reciprocal of characteristic velocity, = t _sH, s/cm - B width of test drum, cm - b axial coordinate of test drum, cm - D diameter of test drum, cm; D ₁, inner diameter; D ₂, outer diameter - d diameter of laser beam, mm - d _p particle diameter, mm - E kinetic energy of liquid, kg · cm²/s²; E _s, steady value - F force transferred from drum walls to liquid, N - f focal length of lens, mm - G one-half of spacing between two parallel split beams, Fig. 1 - H characteristic length of test drum, cm; = V/A - M momentum of liquid, kg·cm/s; M _s, steady value - m mass of control volume, kg - r radial coordinate of test drum, cm - S fringe spacing, mm - t time, s - t _p time for particle to travel through fringe spacing, s - t _s transient time, s - u liquid velocity, cm/s - V liquid volume in test drum, cm³ - V _s effective volume of sample volume, mm³ - v velocity of tracer particle, cm/s; = S/t - W waist diameter of parabola in Fig. 2, mm - (x, y, z) coordinates for paraboloid in Fig. 2, mm - crossing angle of splitting beams, degrees - wavelength of laser length, cm - v kinematic viscosity, cm²/s - liquid density, kg/cm³ - Doppler frequency, l/s - s at steady state - 1 outer - 2 inner On leave from the Dept. of Mechanical Engineering, Musashi Institute of Technology, Tokyo, Japan     

Assessment of turbulent spectral bias in laser Doppler velocimetry

A critical evaluation is made of the spectral bias which occurs in the use of a laser Doppler velocimeter (LDV). In order to accommodate the randomly sampled LDV data, statistical treatments of particle arrival times are needed. This is modeled as a doubly stochastic Poisson process which includes the intensity function of the velocity field. Three processing algorithms are considered for spectral estimates: the sample and hold method (SH), the modified Shannon sampling technique (SR), and the direct transform (RG). Assessment is made of these for varying data densities (0.05 ≤ d.d ≤ 5) and turbulence levels (t.i.=30%, 100%). The effects of the values of the Reynolds stress coefficients and the transversal standard deviation on the spectral contents were examined. As an improved version of the spectral estimator, the utility of POCS (the projection onto convex sets) has been tested in the present study. This algorithm is found useful to be in the region when d.d. ? 3.     

High image-density particle image velocimetry using laser scanning techniques

Laser scanning, corresponding to time-dependent deflections of laser beam across a field of interest, can provide relatively high illumination intensity of small particles, thereby allowing implementation of high image-density particle image velocimetry (PIV). Scanning techniques employing a rotating (multi-faceted) mirror, an oscillating mirror, and an acousto-optic deflector are addressed. Issues of illumination intensity and exposure, rate of scan of the laser beam, and retrace time of the scanning beam are assessed. Representative classes of unsteady separated flows investigated with laser-scanning PIV are described.     

Measuring the wind-induced water surface flow by laser Doppler velocimetry

The way to adjust the laser Doppler velocimeter for measuring the water motions near a moving air-water interface is described by specifying the experimental arrangement, the choice of the scattering configuration, the seeding problems and the preliminary tests which were performed. This technique is then used to explore the motions induced by the wind near the water surface. Illustrative results on the instantaneous longitudinal and vertical velocities, the mean profiles and the respective spectra are presented and discussed. It appears that the Eulerian measuring method when applied to such flows and the presence of dropouts on the signals may require significant corrections of the original data before any physical interpretation.     

Power spectra of fluid velocities measured by laser Doppler velocimetry

The power spectrum and the correlation of the laser Doppler velocimeter velocity signal obtained by sampling and holding the velocity at each new Doppler burst are studied. Theory valid for low fluctuation intensity flows shows that the measured spectrum is filtered at the mean sample rate and that it contains a filtered white noise spectrum caused by the steps in the sample and hold signal. In the limit of high data density, the step noise vanishes and the sample and hold signal is statistically unbiased for any turbulence intensity.List of symbols A cross-section of the LDV measurement volume, m² - A empirical constant - B bandwidth of velocity spectrum, Hz - C concentration of particles that produce valid signals, number/m³ - d _m diameter of LDV measurement volume, m - f(₁, ₂ | u) probability density of t _i; and t _j given (t) for all t, Hz² - probability density for t _j-t_i, Hz - n (t, t) number of valid bursts in (t, t) = N + n - N (t, t) mean number of valid bursts in (t, t) - N _e mean number of particles in LDV measurement volume - valid signal arrival rate, Hz - mean valid signal arrival rate, Hz - R _uu time delayed autocorrelation of velocity, m²/s² - S _u power spectrum of velocity, m²/s²/Hz - t ₁, t ₂ times at which velocity is correlated, s - t _i, t _j arrival times of the bursts that immediately precede t ₁ and t ₂, respectively, s - t _ij t _j–t _i s - T averaging time for spectral estimator, s - T _u integral time scale of u (t), s - T Taylor's microscale for u (t), s - u velocity vector = U + u, m/s - u fluctuating component of velocity, m/s - U mean velocity, m/s - u _m sampled and held signal, m/s Greek symbols (t) noise signal, m/s - _m (t) sampled and held noise signal, m/s - bandwidth of spectral estimator window, radians/s - time between arrivals in pdf, s - Taylor's microscale of length = UT m - kinematic viscosity - ₁, ₂ arrival times in pdf, s - root mean square of noise signal, m/s - _u root mean square of u, m/s - delay time = t ₂ - t ₁ s - B duration of a Doppler burst, s - circular frequency, radians/s - _c low pass frequency of signal spectrum radians/s Other symbols ensemble average - conditional average - ^ estimate     

Signal-processing techniques for low signal-to-noise ratio laser Doppler velocimetry signals

A variety of methods have been developed to obtain acurate frequency estimates from laser Doppler velocimetry (LDV) signals. Rapid scanning and fiber optic LDV systems require robust methods for extracting accurate frequency estimates with computational efficiency from data with poor signal-to-noise ratios. These methods typically fall into two general categories, time domain parametric techniques and frequency domain techniques. The frequency domain approach is initiated by transforming the Doppler bursts into the frequency domain using the fast Fourier transform (FFT). From this basic transformation a variety of interpolation procedures (parabolic, Gaussian, and centroid fits) have been developed to optimize the frequency estimation accuracy. The time domain approaches are derived from the parametric form of a sinusoid. The estimation of constants in this relationship is performed to satisfy specific constraints, typically a minimization of a variance expression. A comparison of these techniques is presented using simulated signals and additive Gaussian and Poisson white noise. The statistical bias and random errors for each method are presented from 200 signal simulations at each condition. Frequency estimation via the FFT with zero-padding and a Gaussian interpolation scheme was found to produce the lowest bias and random errors.List of symbols A(z) eigenfilter or characteristic polynomial - a _{m + 1} eigenvector - f frequency, Hz - f normalized frequency f = f/f _s - _d Doppler frequency estimate - f _i frequency of FFT spectral bin - f ^s sampling frequency - N number of sample points in data set - P _i ith power spectral line from PSD - r _xx (i) autocorrelation coefficient for time lag i - R_{Mn + 1} autocorrelation matrix of order M+1 - T sampling period - f spectral resolution for FFT, f = 1/N t - t sampling interval     

Measurement of velocities in gas-liquid two-phase flow using laser Doppler velocimetry

Measurements of bubble and liquid velocities in two-phase flow have been made using a new forward/backward scattering Laser Doppler Velocimetry (LDV) technique. A standard LDV fiber optic probe was used to measure the bubble velocity using direct backscattered light. A novel retro-reflector and lens assembly permitted the same probe to measure the liquid velocity with direct forward-scattered light. Preliminary results show the usefulness of the technique in a duct of narrow thickness dimension.     

Full-field bubbly flow velocity measurements by digital image pulsed laser velocimetry

Digital Pulsed Laser Velocimetry (DPLV) is a full-field, two dimensional, noninvasive, quantitative flow visualization technique. The technique described here includes the novel use of direct digitization of two-phase bubbly flow images using a high resolution imaging system. The image data is stored for further analysis by new image processing and analysis software developed for flow experiments.In the technique, ten consecutive frames of data separated by a time increment of 150 ms, are recorded. Each of these ten frames contains the images of bubbles at that one instant of time. A program smooths the instantaneous image and calculates bubble parameters. Another program matches the bubbles from each of the frames into tracks of bubbles through time. This program uses a statistical technique to determine the best possible path of the bubbles.The ability of pulsed laser velocimetry to capture simultaneous and quantitative rather than qualitative information along with these image processing techniques gives the experimentalist a powerful tool to perform flow visualization and analysis.     

Application of nonlinear spectral analysis and signal reconstruction to laser Doppler velocimetry

Spectral analysis is a useful tool in engineering applications but many practical problems are encountered in computing power spectral densities (PSD), particularly in laser Doppler velocimetry where data are irregularly sampled. Methods to compute PSD are usely based on discrete Fourier transforms such as FFT. A non linear spectral analysis approach is tested and found to give better results, particularly with short data records. Results are also quite promising when non linear spectral analysis is combined with signal reconstruction to process irregularly sampled data.A preliminary version of this paper has been presented at LDA specialist meeting on the use of computers, May 1987, Institut Saint Louis, France     

Irregular particle sizing using speckle pattern for continuous wave laser applications

We present velocity power spectra computed by the so-called direct method from burst-type laser Doppler anemometer (LDA) data, both measured in a turbulent round jet and generated in a computer. Using today’s powerful computers, we have been able to study more properties of the computed spectra than was previously possible, and we noted some unexpected features of the spectra that we now attribute to the unavoidable influence of a finite measurement volume (MV). The most prominent effect, which initially triggered these studies, was the appearance of damped oscillations in the higher frequency range, starting around the cutoff frequency due to the finite size of the MV. Using computer-generated data mimicking the LDA data, these effects have previously been shown to appear due to the effect of dead time, i.e., the finite time during which the system is not able to acquire new measurements. These dead times can be traced back to the fact that the burst-mode LDA cannot measure more than one signal burst at a time. Since the dead time is approximately equal to the residence time for a particle traversing a measurement volume, we are dealing with widely varying dead times, which, however, are assumed to be measured for each data point. In addition, the detector and processor used in the current study introduce a certain amount of fixed processing and data transfer times, which further contribute to the distortion of the computed spectrum. However, we show an excellent agreement between a measured spectrum and our modeled LDA data, thereby confirming the validity of our model for the LDA burst processor.