A modified optical flow algorithm based on bilateral-filter and multi-resolution analysis for PIV image processing

Particle Image Velocimetry (PIV) technology is an efficient and powerful testing method to investigate the characteristics of flow field. The topic of PIV post-processing techniques has roused researchers׳ wide concern for its great influence on the success of flow field measurement. The traditional correlation algorithms have their innate defects. In the present study, a modified optical flow algorithm is proposed to overcome these deficiencies based on bilateral-filter and multi-resolution analysis of PIV image processing. The algorithm is designed based on the principle of multilayer segments, in which the isotropic diffusion method is employed to calculate the low-resolution layer of the image and the nonlinear filtering method is used to process the high-resolution layer. This new algorithm can reduce image noise effectively and maintain the details of the image boundary. In addition, the design of nonlinear filter makes the optical flow equation simpler, and the optimal velocity mapping factor method needs less iteration and reduces the computational load. The algorithm is first tested on synthetic time-resolved channel flow images, and the computational results from the simulated particle images are found to be in reasonable agreement with the given simulated data. The algorithm is then applied to images of actual up-channel flow, and the results also confirmed that the algorithm proposed in the present study has good performance and reliability for post-processing PIV images.     

The development of an automated PIV image processing software—SmartPIV

Since the popularity of digital particle image velocimetry technique (DPIV), many PIV image processing algorithms have been proposed. Amongst them, fast Fourier transform (FFT) Cross Correlation, Discrete Window Offset Cross Correlation, Iterative Multigrid Cross Correlation, Iterative Image Deformation Cross Correlation and cross correlation based particle tracking methods are widely used algorithms and have been extensively studied by researchers. All of these algorithms have their advantages and disadvantages in terms of computational load and measurement accuracy. To choose a suitable algorithm, researchers not only need to understand the complex principles of these algorithms, but also need to find out their applicable flow conditions. This could greatly increase work load for PIV users who focus more on flow structure itself instead of PIV algorithms. It is therefore necessary to develop a method which can choose PIV algorithms wisely according to the input PIV images. This paper firstly reviews the development of PIV algorithm with mainly focus on analysing advantages and disadvantages of six widely used algorithms. By using both synthetic and real PIV images, comparative studies are then carried out among these algorithms. The tests give a rate for the performance of the algorithms and provide a parameter to automatically separate pattern match and particle tracking algorithms. Based on qualitative and quantitative analysis, an automated PIV image processing method—SmartPIV is proposed and tested by both synthetic and real PIV images. For all the three test cases, the SmartPIV successfully picks the most suitable algorithm and gives very promising results.     

粒子图像测速（PIV）技术在烟丝流量检测中的研究

粒子图像测速技术（PIV）是一种全新的非接触式的,瞬时的,全场流速测量方法,广泛应用于流体力学中。本研究将PIV技术引入烟丝运动研究是一个尝试,旨在为烟丝流量的检测提供一个新的测量手段。通过分析烟丝在风送管道内的运动,利用PIV技术并结合PTV技术,采用图像处理得到烟丝的运动速度。粒子图像测速技术具有一定的优势,可以克服传统测量手段的不足。     

Measurement of particle/bubble motion and turbulence around it by hybrid PIV

Characteristics of bubble flow are influenced by bubble motion, liquid flow and interactions between bubbles, and between a bubble and liquid phase. Thus because behavior of a single bubble and liquid around it is regarded as one of the basic elements characterizing bubble flow, the single bubble motion in stagnant water was investigated experimentally by using flow visualization and image processing methods. The bubble motion is influenced by several factors, that is, bubble size, density difference between gas and liquid, bubble shape and deformation in motion. In order to separate the effect of each factor, some solid particles with different size, shape and/or density were also measured and the characteristic of each factor was discussed. Two-dimensional water velocity field and the motion of a rising particle/bubble in the water were simultaneously measured by PIV (Particle Image Velocimetry) and PTV (Particle Tracking Velocimetry), respectively (Hybrid PIV). The experimental results showed that the large density difference between a particle and water caused high relative velocity and induced zigzag motion of the particle. Furthermore, the turbulence intensity of a bubble was about twice in the case of the spherical solid particle of similar diameter.     

Analysis of two-phase flow regimes and parameter distribution in 5 x 5 rod bundle using image processing techniques

With the recent developments in image processing and analysis, this paper presents bubble characteristics distribution in adiabatic air-water two-phase flow through a 5 × 5 rod bundle. The experiment covered water superficial velocities (J_l = 0.012 m/s – 0.421 m/s) and air superficial velocities (J_g = 0.042 m/s – 0.987 m/s) in which three distinct flow regimes were identified. The flow regime map was compared with existing flow regime transition criteria for vertical rod bundles. Distinct features from the two-phase flow images were extracted to train a classifier model to distinguish between regimes from a separate experiment. The model distinguishes between the bubbly flow regime and others accurately. The void fraction and velocity distributions were also extracted from the R–CCN masked images. Bubble-induced turbulence that was dominant in the subchannel at (J_l = 0.28 m/s) shifted to the outer subchannels and gaps when the flow rate increased (J_l = 0.42 m/s). These methods over-predicted the void faction around the surfaces of the inner rods.     

Evaluation of recursive PIV algorithm with correlation based correction method using various flow images

The hierarchical recursive local-correlation PIV algorithm with CBC (correlation based correction) method was employed to increase the spatial resolution of PIV results and to reduce error vectors. The performance of this new PIV algorithm was tested using synthetic images, PIV standard images of Visualization Society of Japan, real flows including ventilation flow inside a vehicle passenger compartment and wake behind a circular cylinder with riblet surface. As a result, most spurious vectors were suppressed by employing the CBC method, the hierarchical recursive correlation algorithm improved the sub-pixel accuracy of PIV results by decreasing the interrogation window size and increased spatial resolution significantly. However, with recursively decreasing of interrogation window size, the SNR (signal-to-noise ratio) in the correlation plane was decreased and number of spurious vectors was increased. Therefore, compromised determination of optimal interrogation window size is required for given flow images, the performance of recursive algorithm is also discussed from a viewpoint of recovery ratio and error ratio in the paper.     

Liqnet: A real-time monitoring network for two-phase flow patterns

Real-time identification of gas-liquid two-phase flow can help fluid systems maintain safe operating conditions. A flow pattern identification method based on a convolutional neural network (CNN) algorithm (after this referred to as liqnet) is proposed in this paper to realize automatic detection and real-time identification of two-phase flow patterns. This paper mainly focuses on solving two problems of CNN algorithm flow pattern identification (1): the experimental samples for two-phase flow classification are few, and (2): the existing methods do not fully consider the real-time nature of two-phase flow identification. Therefore, this paper constructs a two-phase flow database containing 6242 images using data enhancement, proposes a lightweight network liqnet, and compares it with six mainstream CNN models. The results show that liqnet can achieve the highest accuracy (98.65%), has the least amount of parameters (1.3708 M), and can achieve the purpose of real-time prediction (32.11FPS).     

Form drag coefficient quantification on rising bubbles using particle image velocimetry

Two-phase flow transport heavily depends on the generalized interfacial drag force term in the two-fluid model. The impact of accurate design and prediction associated with thermal energy systems is highly sensitive to multi-phase heat transfer characteristics. Because of this, the interfacial drag force has been studied with rigor for some time. The steady state drag force component in particular has been well characterized for rising single bubbles but has not been previously experimentally separated into its skin and form drag components. Historically, experimental studies were unable to measure the pressure distribution around a bubble to determine the form drag force along the bubble interface. This paper presents the outcomes of an experimental study wherein a new experimental method was developed which, for the first time, separates the form and skin drag coefficients on rising bubbles. Eleven air bubbles sizes representing spheroidal, ellipsoidal, and transition to spherical cap regimes (10²<Re<10⁴) were studied in a water test loop with velocity fields measured via particle image velocimetry; pressure fields were then synthesized from these velocity fields through the Queen2 algorithm. The skin and form drag coefficients were separated for single bubbles which showed a nominal trend of increasing form drag contribution with increasing Reynolds number. This work presents a new method and new outcomes for rising bubbles over several bubble regimes and includes a comprehensive uncertainty characterization of the resulting data.     

Model analysis for differential pressure two-phase flow rate meter in intermittent flow

Multiphase flow rate metering is a challenging problem, specially for flow patterns other than wet-gas. This paper brings forward a new comparative analysis of three differential pressure calibration models suited for liquid dominated two-phase flows, in a total of seven model configurations. First, the models are compared theoretically and classified in terms of the type of input data required. Then, experimental data of over 300 horizontal air–water experiments, for $1$” and $2$” pipe diameters, supports quantitative analyses of the prediction accuracies and sensitivity of the superficial velocities of gas and liquid to measurement errors in the model input variables. Finally, a method for assessing the decoupled measurement errors for the void fraction and gas velocity is shown, as these variables are typically subject to higher uncertainties. It results that, though the void fraction is shown to be systematically under evaluated in more than 10%, the total mass flow rate is estimated through the Paz et al. (2010) model with an overall root mean squared deviation (RMSD) of 5.75% for the $2$” data. Also, the use of gas velocity measurements, even if subject to considerable errors, decreased the RMSD for the gas superficial velocity by more than half for the $1$” data.     

Velocity distribution of liquid phase at gas-liquid two-phase stratified flow based on particle image velocimetry

Horizontal gas-liquid flows are commonly encountered in the production section of the oil and gas industry. To further understand all parameters of the pipe cross-section, this paper use particle image velocimetry to study the circular pipe cross-section liquid velocity distribution rule. Firstly the focus is on the software and hardware combination of image correction system, to solve the influence of different refractive indexes of medium and pipeline curvature caused by image distortion. Secondly, the velocity distribution law of the corrected stratified flow (the range of liquid flow of 0.09-0.18 m³/h, and gas flow range of 0.3-0.7 m³/h) cross-section at different flow points of the pipeline cross-section at x=0 and in the Y direction at the maximum liquid velocity is studied. It is found that these distribution laws are caused by the influence of the interphase force of the gas-liquid interface and the resistance of the pipe wall. The current measurements also produce a valuable data set that can be used to further improve the stratified flow model for gas-liquid flow.     

Coriolis flowmeter damping for two-phase flow due to decoupling

Coriolis flowmeters experience measurement errors due to both single- and two-phase flow. For two-phase flow, severe damping may occur, which leads to a (temporary) inability of the flowmeter to operate. The dominating part of the damping is caused by decoupling of the continuous and the dispersed phase. This paper presents the theory of damping due to decoupling in two-phase flow. Using a simple structural model, we provide examples of mixtures with water as the continuous phase. The dispersed phase is either air, or oil or sand.     

Image processing techniques for high-speed videometry in horizontal two-phase slug flows

Two-phase flow measurements are very common in industrial applications especially in oil and gas areas. Although some works in image segmentation have analyzed gas–liquid slug flow along vertical pipes, few approaches have focused on horizontal experiments. In such conditions, the detection of the Taylor bubble is challenging due the great amount of small bubbles in the slug area and, thus, requires a special treatment in order to separate gas from liquid phases. This article describes a new technique that automatically estimates bubble parameters (e.g. frequency, dimension and velocity) through video analysis of high-speed camera measurements in horizontal pipes. Experimental data were obtained from a flow test section where slug flows were generated under controlled conditions. Image processing techniques such as watershed segmentation, top-hat filtering and H-minima transform were applied to detect and estimate bubble contour and velocities from the observed images. Finally, the estimated parameters were compared to theoretical predictions, showing good agreement and indicating that the proposed technique is a powerful tool in the investigation of two-phase flow.     

Theoretical and experimental investigation for developing a gas-liquid two-phase flow meter

Despite the intricacy, inline metering of two-phase flow has a significant impact in multitudinous applications including fusion reactors, oil, nuclear, and other cryogenic systems. Since measurement of individual flow rate is prominent in various systems, it warrants the establishment of a flow meter system that can monitor the mass flow rates of liquid. In this regard, an approach was taken towards the development of a two-phase flow meter system in the present study. The concept involves two-phase flow through narrow parallel rectangular channels resulting in laminar, stratified flow with a slope at the liquid-vapor interface. The height of the liquid column at specific channel locations is measured for determining the flow rate. However, the geometric configurations of the channels and fluid properties are pivotal in ensuring accurate measurement. Consequently, theoretical and experimental studies are performed to investigate the correspondence between flow rate and change in liquid height. Based on the governing equations, a theoretical model is established using MATLAB®. The model investigated the intricate influence of various flow and fluid properties in the estimation of the mass flow rate. The experimental investigation was done with various conditions under different liquid and vapor volume flow rates for validating the proposed supposition and the theoretical model. Both the theoretical and experimental analyses showed fair correspondence. The proposed system estimated the mass flow rate within a tolerance of ±10% and showed potential towards the development of the cryogenic two-phase flow meter.     

Quality measurement of refrigerant two-phase flow in refrigeration cycles

In refrigeration cycles, quality measurement of two-phase refrigerant flow is required to monitor the cycle operation. Although sectional void fraction of the two-phase flow can be detected in several ways, the quality of the two-phase flow is hardly obtained from the sectional void fraction since velocities of liquid- and gas-phase in the pipe are different from each other. In this study, a new quality measuring method was developed by installing multiple narrow tubes in a test section. By installing a gas bypass tube with the multiple narrow tubes, the quality measurement having an accuracy within 0.03 was achieved in the quality range from 0.05 to 0.8. In addition, the influence of oil contamination in the refrigerant flow on the flow pattern in the narrow tube was examined. It was found that the flow pattern in the narrow tube became bubble flow by the mixing of oil.     

A hybrid phase boundary detection technique for two-phase-flow PIV measurements

Particle Image Velocimetry (PIV) measurement accuracy is lower along the phase boundaries of two-phase-flows, because the interrogation windows contain information from both phases. Different seeding density, background intensity, velocity magnitude and flow direction conditions often exist across the boundary, and the cross-correlation-based PIV algorithm selects only the highest correlation peak. The highest correlation peak is either influenced by the wrong phase (across the boundary), or the correctly calculated displacement is erroneously detected as an outlier at a later stage and is subsequently replaced. Phase-separated PIV measurements minimize this problem, and increase accuracy along the boundary by treating each phase separately. This type of measurement requires for each time step; (i) the accurate detection of the phase boundary in consecutive frames, (ii) generation of dynamic phase masks, (iii) an accurate PIV evaluation of each phase and (iv) recombination of the flow fields. In this article, we focus on the first step and test a hybrid phase boundary detection (PBD) technique in three different two-phase-flow configurations which manifest different challenges: The first configuration is the mixing of two liquids in a magnetic micromixer, the second is a combustion experiment where a turbulent, pre-mixed, low-swirl, lifted flame is investigated, and the third is a bubble column reactor where air bubbles are rising in a water tank. The PBD implementation uses a three-step procedure: approximate global thresholding, local Otsu thresholding, and discrimination of image gradients. Comparison of results with and without the use of PBD and phase separation indicate that there are significant measurement accuracy improvements along the boundary.     

Branching of two-phase flow from a vertical header to horizontal parallel channels

The objective of the present experimental work is to investigate the two-phase flow distribution from a vertical main to parallel horizontal branches. Both the main and the branches have rectangular cross-sections simulating the header and the channels of the compact heat exchangers for air-conditioning systems. The cross section of the main is 8 mm × 8 mm while that of the parallel branches is 8 mm × 1 mm. Here, the second (downstream) junction was taken as the reference. The effect of the distance between the branches was mainly examined by changing it from 9 mm to 49 mm for the given flow conditions at the inlet of the downstream junction. Air and water were used as the test fluids. The superficial velocity ranges of air and water at the test section inlet were 13.2–21.4 m/s and 0.08–0.28 m/s, respectively. When the branch spacing becomes smaller, the fraction of liquid separation through the downstream branch decreases. The trend remains the same over the entire range of the present experiment, i.e., for different values of quality and the mass flow rate at the inlet of the downstream junction. Based on the correlation for single T-junctions, a modified correlation was proposed to take into account the effect of the branch distance in predicting the fraction of liquid separation. The correlation represents the experimental results within the accuracy of ±15 %. This paper was recommended for publication in revised form by Associate Editor Kyung-Soo Yang Jun Kyoung Lee received his B.S. degree in Mechanical Engineering from Busan National University in 1999. He then received his M.S. and Ph.D. degrees from KAIST in 2001 and 2005, respectively. Dr. Lee is currently a Professor at the School of Mechanical Engineering and Automation at Kyungnam University in Masan, Korea. Dr. Lee’s research interests are in the area of two-phase flow and heat transfer, micro-fluidics, cryogenic devices for superconductivity, thermal management systems for automobiles.     

Pipe two-phase flow non-invasive imaging using Ultrasound Computed Tomography: A two-dimensional numerical and experimental performance assessment

Pipe two-phase flow non-invasive imaging is of great interest in the field of industry. In particular, small bubble flow imaging through opaque pipes is challenging. Ultrasound computed tomography can be a relevant technique for this purpose. However, perturbation phenomena that are inherent to the configuration (acoustic impedance mismatching, circumferential propagation, reverberation) limit two aspects: the performance of the technique and the use of conventional inversion algorithms. The objectives of the presented work are: (i) to predict the effects of the pipe wall on ultrasonic waves for both metallic and plastic pipe, (ii) to define a consistent inversion algorithm taking into account those effects, (iii) to validate and to assess the limitations of the designed imaging numerical tool using an experimental setup. The benchmark configuration consists of 150 mm diameter 3 mm thick pipes containing 6 mm diameter rods acting as reference scatterers. Two materials of very different acoustical properties were tested: aluminum and PMMA. The results highlighted that the quality of the reconstructed image is very dependent on the pipe material. The results showed that, using an adapted inversion model, consistent target reconstruction is obtained. Based on numerical predictions, performance limitations are reached for metallic pipes.     

Analytical investigation of an inductive flow sensor with arc-shaped electrodes for water velocity measurement in two-phase flows

An inductive flow sensor with spot-shaped electrodes (IFS-SE) is sensitive to the shape of the flow profile and is restricted to be used to measure the flow rate of axisymmetric single-phase flows in a circular pipe. In many cases of application, it is not possible to provide a fully developed flow profile. Therefore, the inductive flow sensor has to cope with flow profiles that are not fully developed. To improve the accuracy, an inductive flow sensor with a pair of arc-shaped electrodes flush-mounted on the internal surface of an insulating section of a pipe is proposed in this article to investigate the characteristics of vertical gas-water two-phase flows. The effect of the flow profile on the inductive flow sensor is analyzed. A key contribution of the present work is to estimate the relationship between the induced voltage and the velocity of the conductive phase in two-phase flows. The estimation is achieved by the analytical calculation of magnetic-inductive equations through the method of variables separation. The analytical solution is compared with the results from an ideal model and from numerical simulation. Experiments are conducted to calibrate the inductive flow sensor with arc-shaped electrodes (IFS-AE). It is noted that the proposed IFS-AE can be adopted to obtain the velocity of the conductive phase in two-phase flows by measuring the voltage induced on the arc-shaped electrodes.