基于几何光学的图像矫正法在圆管PIV实验中的应用研究

粒子图像速度技术被广泛用于流体流动测量,介质折射率差异使光在圆管壁面发生偏折,导致图像失真,直接影响速度测量精度。本文建立了光学折射的物理模型,得到圆形管道中物点和图像点之间的函数关系进而得到矫正后图像的像素坐标,使用双线性插值算法得到像素灰度值重建出矫正后的粒子图像,最后根据多重网格迭代算法计算管内速度场。分别对流体进行管内静态流体与管内层流速度场测量实验,对比了光学矫正箱法、线性矫正以及基于光学模型的畸变矫正方法误差。结果表明,本文提出的基于几何光学的图像矫正方法精度优于光学矫正箱法和线性矫正方法,并通过静态与流动实验充分验证了所建立几何光学模型的准确性和有效性。     

An experimental device for measuring radial flow velocity profiles of yield stress fluids

Measuring the radial flow velocity field of yield stress fluids (YSFs) between two parallel disks provides crucial data to understand the underlying flow phenomena. However, direct velocimetry of YSFs in the radial flow configuration remains a challenge, due to the complex fluid rheology and geometry constraints. In this paper, we present an experimental device for measuring YSF radial flow velocity profiles. Ultrasound Velocity Profiling (UVP) is used to non-intrusively measure the velocity profiles. The Tikhonov regularization method is implemented to obtain smooth velocity profiles, which are used to calculate the plug-flow region. Compared to our previous work on radial flow, the current contributions include: (i) additional structural frame members to maintain a constant aperture, (ii) wall slip reduction, and (iii) an improved velocity profile plug-detection algorithm. The results show that the experimental device and the measurement method are effective for further studying radial flow behavior of YSFs for industrial applications.     

Estimation of spray flow characteristics using ensemble Kalman filter

Spray flows are widely used in several industrial applications, such as combustion engines. Accurate measurement of spray flow characteristics requires sophisticated equipment and techniques. In recent years, the discrete droplet method (DDM), which analyses droplet scattering, has become a mature technique and has been applied to various analyses. We propose an estimation system based on particle image velocimetry (PIV) measurements and an ensemble Kalman filter, together with DDM, to efficiently investigate spray flow characteristics. The proposed method performs data assimilation on the velocity distribution in a two-dimensional cross-section obtained by PIV to estimate the characteristics of the spray flow in three dimensions. In this study, the system was constructed so that droplet particle is ensembled during data assimilation to estimate the droplet diameter distribution indirectly. The proposed method can be used to estimate the spray velocity and droplet size distribution. The numerical solution obtained using DDM was used as a criterion for assimilation and validated by conducting twin experiments. The results showed that, in terms of spray velocity, the estimation error for the velocity component parallel to the main flow was 2% and that for the velocity component perpendicular to the main flow was around 10%. Finally, the velocity and particle size distributions of the spray stream and the three-dimensional droplet distribution were estimated by assimilating the velocity distributions measured by PIV. This technique predicts the spray angle and droplet size distribution from the two-dimensional velocity field of the PIV only and is expected to contribute to the development of injectors and atomizers.     

基于维诺图匹配的粒子跟踪测速法

针对密集颗粒流速度场分布的测量问题,提出了基于维诺图匹配的粒子跟踪测速法。首先,通过对图像粒子进行维诺图构建,给出面积相似度筛选匹配粒子的条件;其次,引入Delaunay三角网搜索结构,通过计算维诺多边形的形状相似度来匹配粒子;再次,研究了去除错误匹配粒子矢量的方法和匹配算法中的关键参数;最后,通过模拟二维旋转流场运动以及二维转盘中的颗粒流实验对算法进行了测试。结果表明:维诺图匹配的匹配准确率高于DT-PTV并且在处理密集粒子匹配效果上更好;维诺图匹配算法适用于测量密集颗粒流速度场分布,颗粒匹配准确率高达99%,并由得到的颗粒流速度场分布验证了算法的有效性。     

Review of certain key issues in indirect measurements of the mass flow rate of solids in pneumatic conveying pipelines

On-line mass flow measurement of particulate solids in pneumatic conveying pipeline is a technically challenging area, where mass flow measurement presents a range of problems. These problems are not normally relevant to a single phase flow, but are always involved in gas–solids two-phase flows, like inhomogeneous distribution of solids over the pipe cross section, irregular velocity profiles, variations in particle size, moisture content, and deposition of fine particles on the inner wall of the pipeline. These variables may affect the response of a solids flow meter in ill-defined ways. All of these make the design and the calibration of solids’ mass flowmeter more difficult. Based on a review of non-invasive mass flow measurements of particulate solids, this paper summarizes and highlights several key issues, which often rely on structures of sensors or measurement methods, in indirect mass flow metering of pneumatically conveyed solids. They are: (i) spatial filtering effect; (ii) averaging effect; (iii) measurement resolution and sensitivity of array structures in tomography sensors.     

复杂颗粒流速度场分布的空间滤波测量

空间滤波测速法具有结构简单、稳定性好和适用性强等优点,但传统线阵CCD空间滤波测速法要求CCD阵列方向与待测物体的运动方向一致,因此不适合测量复杂的流场。针对这一问题,本文提出了基于面阵CCD相机的空间滤波测速方法。对采集到测量区域的一系列面阵CCD输出图像进行处理,在图像水平方向和垂直方向分别进行隔行采样,模拟多狭缝空间滤波特性,实现了对障碍流颗粒速度的光学非接触式测量。针对复杂流场功率谱密度的特点,利用能量重心校正频谱提高了系统测量精度。通过调节传送带速度对本系统进行标定,实现了不同速度下的流速测量,平均误差小于4%。利用本系统还测量了由玻璃砂模拟的泥石流速度场分布,最后讨论了空间周期和测量时间对测量结果的影响。结果表明,采样时间大于0.5s时,测量结果趋于稳定,空间分辨率最小可达1.28mm。     

An experimental study for the flow rates of automatic pressure smoke dampers and their applications

Smoke spread from a fire is a major cause of mass casualties. To suppress smoke spread, escape routes are pressurized and smoke dampers are used. In Korea, the automatic pressure smoke damper (APSD) was recently adopted in escape routes, and the number of APSDs is increasing concomitantly with the increase in the number of high-rise buildings. In this study, a more efficient method for the assessment of the flow rates of APSD was experimentally studied for multi-blade types of which the blade angle is controlled. A modified wind tunnel was used to measure the relationship between the pressure difference and the flow rate with a blade angle variation for three kinds of APSD. As a result, the performance of the APSDs could be presented as a function of the pressure difference and blade angle. Using this performance function, a volume of information regarding the characteristics of the APSDs could be easily extracted. Furthermore, some experimental methods, based on the performance of the APSD, were suggested to assess the smoke suppression velocity and the effective leakage area in practical buildings.     

Double S-type Pitot Tube for velocity field study of fire whirls

In this paper we present a pressure measurement instrument based on the S-type Pitot tube modified to measure the two velocity components of a high temperature flow assuming that the flow is locally two-dimensional. The development of this new device, which we designate as the Double S-type Pitot Tube, is related to the difficulty and the lack of precision of measurements with a standard S-type Pitot tube in flows with unknown directions like the case of fire whirls in laboratory experiments. The design construction and calibration method of the device are described. The pitch angle of the flow and the velocity coefficients of the probe are analysed experimentally in a wind tunnel calibration as well as the associated errors. The use of this sensor in a fire whirl laboratory test is shown and the results are compared with those of simple S-type Pitot tubes in the same test. The obtained results show the applicability and better performance of the novel device.     

Measurement of velocity profiles in multiphase flow using a multi-electrode electromagnetic flow meter

This paper describes an electromagnetic flow meter for velocity profile measurement in single phase and multiphase flows with non-uniform axial velocity profiles. A Helmholtz coil is used to produce a near-uniform magnetic field orthogonal to both the flow direction and the plane of an electrode array mounted on the internal surface of a non-conducting pipe wall. Induced voltages acquired from the electrode array are related to the flow velocity distribution via variables known as ‘weight values’ which are calculated using finite element software. Matrix inversion is used to calculate the velocity distribution in the flow cross section from the induced voltages measured at the electrode array. This paper presents simulations and experimental results including, firstly the effects of the velocity profile on the electrical potential distribution, secondly the induced voltage distribution at the electrode pair locations, and thirdly the reconstructed velocity profile calculated using the weight values and the matrix inversion method mentioned above. The flow pipe cross-section is divided into a number of pixels and, in the simulations, the mean flow velocity in each of the pixels in single phase flow is calculated from the measured induced voltages. Reference velocity profiles that have been investigated in the simulations include a uniform velocity profile and a linear velocity profile. The results show good agreement between the reconstructed and reference velocity profiles. Experimental results are also presented for the reconstructed velocity profile of the continuous water phase in an inclined solids-in-water multiphase flow for which the axial water velocity distribution is highly non-uniform. The results presented in this paper are most relevant to flows in which variations in the axial flow velocity occur principally in a single direction.     

Comparison of the radial profiles of particles velocity between invasive and non-invasive measurement techniques

In this work, the radial profiles of the particles velocity in gas-solid fluidized beds were measured using two techniques: (1) an advanced optical fiber probe, an invasive technique, and (2) radioactive particle tracking (RPT), a noninvasive technique. Two gas-solid fluidized bed columns of 0.14 and 0.44 m inside diameters were used to estimate the radial profile of the particles velocity. The particle velocity was measured at different operating conditions and axial heights. It was found that the implementation of the noninvasive radioisotope particle tracking (RPT) technique was able to successfully guide the measurement of the radial profile of the particle velocity using the optical fiber probe technique. The average inversion point of the radial particle velocity, from positive to negative, was identified from the measurements using both techniques for different operating conditions and axial heights.     

In-line measurement of pneumatically conveyed particles by a light transmission fluctuation method

An optical method for particle sizing in two-phase flow, the so-called light transmission fluctuation method, is presented in this paper. On the basis of this method, a novel optical instrument capable of in-line measurement of pneumatically conveyed particles is introduced. Quite a few medium-term tests validate the reliability and the applicability of this technique. Some measurement results are shown and discussed, concerning real-time in-line continuous monitoring of pulverized coal flow at coal-fired power plants and relevant laboratory experiments.     

基于降质图像-光流场复原的智能轮椅测速

提出一种运动图像去模糊复原和基于仿射运动模型的光流场去抖动方法,以提高智能轮椅中光流里程计测速方法的精度。当轮椅线速度或角速度较大时,导致机载相机成像产生显著的运动模糊;且轮椅机器人的机械抖动也易产生光流场的偏差,进而影响速度估计的精度。针对于此,首先利用一种基于自适应模糊核的运动图像去模糊方法实现图像复原,以改善视频帧质量;其次,针对智能轮椅在行进中的机械抖动,利用随机抽样一致(RANSAC)排异后的光流场,在卡尔曼滤波框架下估计同名像点的仿射运动模型参数,进而实现光流补偿。实验结果表明所提方法能够提升基于光流场的智能轮椅视觉测速精度。     

Estimation of uncertainty of mean velocity of flue gas in a conduit as determined by the traverse method within the gravimetric measurement of particulate concentration

In gravimetric measurements of dust emissions from industrial technological plants, the required mean gas velocity in a conduit is often determined by Pitot traverse method. It is commonly seen as a method giving good approximate values of mean gas velocity, although the actual rate of this approximation is not considered in the analysis of measurement results. It was seen that there was a need to establish what magnitude of error might occur in practice due to the small number of measurement points and typical non-uniformity of the gas velocity profiles in conduits of rectangular cross-section. The calculations were based on the concept of treating a measurement plane as one consisting of a set of elementary planes. The elementary gas velocity profiles in these elementary planes were simulated, the mean velocity for these profiles were calculated based on point velocity values, and the measurement uncertainty of this mean velocity determined. This uncertainty results in the uncertainty of the mean velocity across the entire measurement plane. It appears that, depending on the number of measurement points and gas velocity profile non-uniformity, the value is not small and is of the order of several percent, and hence needs to be taken into account in the budget of the combined uncertainty of mean velocity, which in turn contributes to the uncertainty of gas volumetric flow rate and dust pollutant mass flow rate.     

Particle image velocimetry of a flow at a vaulted wall

The assessment of flow along a vaulted wall (with two main finite radii of curvature) is of general interest; in biofluid mechanics, it is of special interest. Unlike the geometry of flows in engineering, flow geometry in nature is often determined by vaulted walls. Specifically the flow adjacent to the wall of blood vessels is particularly interesting since this is where either thrombi are formed or atherosclerosis develops. Current measurement methods have problems assessing the flow along vaulted walls. In contrast with conventional particle image velocimetry (PIV), this new method, called wall PIV, allows the investigation of a flow adjacent to transparent flexible surfaces with two finite radii of curvature. Using an optical method which allows the observation of particles up to a predefined depth enables the visualization solely of the boundary layer flow. This is accomplished by adding a specific dye to the fluid which absorbs the monochromatic light used to illuminate the region of observation. The obtained images can be analysed with the methods of conventional PIV and result in a vector field of the velocities along the wall. With wall PIV, the steady flow adjacent to the vaulted wall of a blood pump was investigated and the resulting velocity field as well as the velocity fluctuations were assessed.     

Simultaneous measurement of continuous and dispersed phase in a two phase jet flow

In the present paper, a method is proposed and applied in order to increase the size range sensitivity of a Phase–Doppler–Anemometry (PDA) system, allowing simultaneous measurement of size and velocity distributions of continuous and dispersed phases in gaseous, dispersed two phase flow. Discrimination between the two phases is directly achieved here from the detected phase difference. Design considerations are presented to increase the size range sensitivity of the PDA system to 1:200. Therefore, the optical properties for light scattering of the particles are properly adjusted to the measurement problem by colouring the particles homogeneously. The application of the system is described by measurements of velocity profiles in a zone close to the nozzle of a two phase jet with a tracer to particle diameter range of 1–160 μm.     

Improved numerical integration method for flowrate of ultrasonic flowmeter based on Gauss quadrature for non-ideal flow fields

Multipath ultrasonic flowmeters with large diameter are widely used in industry. And their measurement performances are sensitive to velocity profiles in conduits. Gauss–Jacobi and Optimized Weighted Integration for Circular Sections (OWICS) method are commonly applied in flow measurement of multipath ultrasonic flowmeters, both of which assume ideal flow in pipes. They are not proper for non-ideal flow measurement. Therefore, an improved numerical integration method for flowrate based on Gauss quadrature is proposed. With this method, optimum relative path heights and corresponding weights are determined according to specific disturbed flows. By comparison Gauss–Jacobi, OWICS with the improved method, the validity of the proposed method is verified for typical disturbed flows based on both theoretical analysis and experiments, and measurement performances of ultrasonic flowmeters are improved significantly.     

Performance measurement of a smoke extraction system for buildings in full-scale hot smoke test

The current inspection method for a smoke extraction system in Taiwan measures whether the air flow of the smoke vent conforms to the design value. However, this method is inapplicable to natural smoke extraction systems. As the plugholing effect is disregarded, adequate air flow cannot ensure that the system will exhaust smoke effectively during a fire. Hence, a full-scale hot smoke test for smoke extraction systems is necessary. Some international test criteria are formulated using visible smoke and alcoholic fuel which is pollution-free after combustion. The effectiveness is judged only by visually observing the smoke’s flow direction, which is indeed unscientific. This study used a string of vertical smoke layer measuring instruments composed of several approved photoelectric smoke detectors, as well as a light attenuation measuring device composed of illuminance meters to conduct tests on the effect of makeup air. The results proved that the former one uses different height induction times to judge the position of the smoke layer effectively, while the latter one uses the light attenuation rate to distinguish the smoke density instantly and accurately. The obtained experimental data were consistent with the onsite smoke distribution. The two sets of equipment designed in this study can be used for full-scale hot smoke tests to obtain performance data for a smoke extraction system.     

Deep particle image velocimetry supervised learning under light conditions

Particle image velocimetry (PIV) is an important fluid visualization technology which extracts the velocity field from two successive particle images. Recently, some researchers have begun to use convolutional neural network (CNN) to tackle the PIV problem successfully. Some supervised learning methods make use of the PIV dataset with ground truth for network training. However, the existing dataset is composed of pairs of particle images under ideal light conditions and does not take into account the changes in actual experimental conditions. In this paper, we firstly generated a new and more challenging dataset called Light-PIV which fully simulates the change of the brightness of particle images in the real PIV experiment. Secondly, we present here a novel approach for fluid motion estimation which is based on an optical flow network LiteFlowNet. The proposed approach is verified by the application to a diversity of synthetic and experimental PIV images. We not only improve the structure, but also combine the traditional prior assumptions knowledge with the loss function to better guide the network training. The proposed approach is verified by the application to a diversity of synthetic and experimental PIV images. The experimental results show that our proposed method has advantages of high accuracy, obtaining detailed information and strong robustness in our PIV dataset compared with classical PIV methods such as HS optical flow and WIDIM, and even outperforms these existing approaches in some flow cases.     

Point-velocity methods for flow-rate measurements in asymmetric pipe flow

Laser Doppler velocimetry (LDV) is characterized by its ability to determine local fluid velocities with high accuracy. Therefore, LDV may also be used for precise flow-rate measurements of turbulent flow in circular ducts. The uncertainty of the measurement depends mainly on the asymmetry of the axial velocity distribution and on the point-velocity method chosen to estimate the flow-rate. LDV-measurements in conjunction with velocity-area methods have been performed under different asymmetric flow conditions yielding errors in the range of one per cent. The experimental data have been transformed into analytical flow profiles, in order to investigate combinations of single-point measurements. As a result, a new multi-point method with variable centre-point factor is introduced, that reduces both the effort and uncertainty in the measurement.     

立体网状静电传感器空间灵敏特性研究

提出一种全新的立体网状静电传感器,以克服现有静电传感器对流场内速度分布无法测量的不足。通过建立立体网状静电传感器的三维仿真模型,利用有限元分析法获得了立体网状静电传感器的灵敏度空间分布规律,并对其动态灵敏度进行了分析。结果表明:立体网状静电传感器具有局部敏感特性,对于管内颗粒分布比较敏感。而且由于其极棒间相互交错的特性,可以探测到颗粒的运动方向;传感器的动态灵敏度可以反映带电颗粒在各个极棒间的空间位置及其速度,即带电颗粒距离极棒交织点越近,速度越快,动态灵敏度越大。最后通过重力输送颗粒实验,证明理论分析的结果与实验结果一致,从而为立体网状静电传感器实现颗粒流动参数测量提供了理论依据。