Systematic investigation of pipe flows and installation effects using laser Doppler anemometry—Part I. Profile measurements downstream of several pipe configurations and flow conditioners

The first part of this paper reports on an automated facility designed to investigate the influence of disturbed flows in pipes on the shift of the error curves of gas flowmeters in situ. This facility can be equipped with several pipe configurations (single and double bends, convergent and divergent sections, straight pipes up to 40 diameters in length etc.) as well as with various types of flow conditioners. It works with atmospheric air at flowrates of up to 5500 m³/h. A two-component semiconductor Laser Doppler Anemometer (LDA) is used to measure the spatial velocity and turbulence fields of the flow along the entire cross section in front of the flowmeter to be investigated. More than 150 velocity distributions have been determined for different pipe configurations at several flowrates and data have been collected to describe the corresponding flowmeter's behaviour. Some typical velocity profiles for the most usual pipe elements and flow conditioners are shown.

The second part of the publication, which will be presented in the following, compares the changes in the meter behaviour with the specific development of the flow characteristics downstream of the pipe configurations investigated. The model found to explain these metering effects will be described, evaluated and verified.     

Evaluation of hot-film,dual optical and Pitot tube probes for liquid–liquid two-phase flow measurements

An experimental study of kerosene–water upward two-phase flow in a vertical pipe was carried out using hot-film, dual optical and Pitot tube probes to measure the water, kerosene drops and mixture velocities. Experiments were conducted in a vertical pipe of 77.8 mm inner diameter at 4.2 m from the inlet (L/D=54). The tests were carried out for constant superficial water velocities of 0.29, 0.59 and 0.77 m/s (flow rates = 83, 167 and 220 l/min) and volume fractions of 4.2%, 9.2%, 18.6% and 28.2%. The Fluent 6.3.26 was used to model the single and two-phase flow and to reproduce the results for the experimental study. Two methods were used to evaluate the accuracy of the probes for the measurement of the velocities of water, drops and mixture for two-phase flow: (i) comparison of measured local velocities with predictions from the CFD simulation; (ii) comparison between the area-averaged velocities calculated from the integration of the local measurements of water, drops and mixture velocities and velocities calculated from flow meters’ measurements.The results for single phase flow measured using Pitot tube and hot-film probe agree well with CFD predictions. In the case of two-phase flow, the water and drops velocities were measured by hot-film and dual optical probes respectively. The latter was also used to measure the volume fraction. These three measured parameters were used to calculate the mixture velocity. The Pitot tube was also used to measure the mixture velocity by applying the same principle used for single phase flow velocity. Overall the mixture local velocity measured by Pitot tube and that calculated from hot-film and dual optical probe measurements agreed well with Fluent predictions. The discrepancy between the mixture area-averaged velocity and velocity calculated from flow meters was less than 10% except for one test case. It is concluded that the combined hot-film and optical approach can be used for water and drop velocity measurements with good accuracy for the flow conditions considered in this study. The Pitot tube can also be used for the measurement of mixture velocities for conditions of mixture velocities greater than 0.4 m/s. The small discrepancy between the predictions and experimental data from the present study and literature demonstrated that both instrumentation and CFD simulations have the potential for two-phase flow investigation and industrial applications.     

Mass flow measurement of pneumatically conveyed solids using radiometric sensors

This paper describes the design and experimental evaluation of a novel radiometric instrumentation system that has been recently developed for the measurement of velocity, concentration and mass flow rate of solids in a pneumatic pipeline. The system employs a novel multi-beam, micro-sensing field interrogating the entire pipe cross-section to accommodate the irregularity of the solids velocity profile and inhomogeneity of the solids distribution within the pipeline. Experimental results obtained on a pneumatic conveyer circulating ilmenite powder demonstrate that the system is capable of detecting various velocity profiles and solids distributions across the pipe section and providing an absolute mass flow rate of solids within a good agreement to the reference reading from load cells.     

两端固定输流管道的参数共振实验

将非接触式测振方法引入到输流管道的参数共振问题的实验研究。首先,通过激光测振技术获取了实验管道中部在脉动流激励下的振动信息;其次,通过确定管道振动频谱图中1/2倍频出现和消失时对应的激励频率,并在多组测试结果的基础上拟合出了管道第一振型1/2次谐波参数共振区域,且实验结果与利用平均法计算得到的理论结果定性一致;最后,对实验结果的误差成因进行了较深入地分析。得到如下结论：在一定平均流速下,两端固定管道在适当脉动幅值和脉动频率下可以发生参数共振现象,且参数共振区域的位置与流速关系较大;管道实际发生参数共振的范围要大于理论计算结果,这可能与选用平均法进行理论计算有关。     

Flow rate measurement in flows with asymmetric velocity profiles by means of distributed thermal anemometry

Flow rate in closed conduits is one of the most frequently measured parameters in industrial processes and in gas and water supply. For an accurate measurement, flow meters typically require a fully developed symmetric flow profile with preferably no radial or tangential velocity components. This is commonly secured by mounting flow meters in a pipe at a sufficiently long distance downstream any change in cross-section or pipe direction. In this paper, we introduce a new approach for flow rate measurement of gases or liquids that employs a novel spatially resolving fluid velocity sensor basing on thermal anemometry. The new principle allows accurate flow rate measurements for non-axisymmetric velocity profiles, even directly after pipe bends, T-junctions or other alterations in the pipe geometry. This is exemplified for air flow in three different pipe bend configurations.     

Assessment of combined V/Z clamp-on ultrasonic flow metering

Clamp-on ultrasonic flow metering can provide a non-invasive and portable means for flow measurement. However, it indicates flow rates with low measurement accuracy at low flow velocity in pipe flows. Typical accuracy of the clamp-on ultrasonic flow metering amounts as low as ±1% if the flow velocity in a pipe is greater than 0.5 m/s. The accuracy can be increased greater than ±2% if the flow velocity is lowered smaller than 0.5 m/s. Inner pipe diameter can be also an influential factor in flow metering when the exact value of the inner diameter is not known. The inner pipe diameter cannot be found if the pipe is too large to measure or if there are erosions or adhesions on the inner pipe surface due to small particles in the flow. These shortcomings of the clamp-on ultrasonic flow metering can be overcome by combining two transit times along a Z-shaped and a V-shaped ultrasonic path. This technique is termed combined V/Z clamp-on ultrasonic flow metering. With the water flow standard system in KRISS, this combined technique exhibited intermediate performance between the two flow metering techniques along the Z-shaped and the V-shaped ultrasonic paths. Notably, the combined technique showed better performance (expanded uncertainty less than 0.76%, k = 2) than the two flow metering techniques (1.61% and 1.17%, k = 2) in the flow range of (100–400) m³/h with pipe diameter of 250 mm.     

Dynamic analysis of Coriolis flow meter using Timoshenko beam element

Coriolis mass flow meter (CFM) is used to measure the rate of mass flow through a pipe conveying fluid. In the present work, the Coriolis effect produced in the pipe due to a lateral excitation is modeled using the finite element (FE) method in MATLAB©. The coupled equation of motion for the fluid and pipe is converted to FE equations by applying Galerkin technique. The pipe conveying fluid is excited at its fundamental natural frequency. The time lag observed between symmetrically located measurement points which are equidistant from the point of excitation, is utilized to predict the mass flow rate. The results predicted by the present code is validated using the experimental, and numerical results published in the literature. The main contribution is the development of a FE model, using three node Timoshenko beam element to analyse the dynamics of fluid conveying pipes subjected to external excitation. The direction of the Coriolis force is perpendicular to the plane containing the velocity of flow vector and angular velocity vector of the pipe. Hence a three dimensional FE model is essential. This model can include curved geometry, damping, velocity and gyroscopic effects for three dimensional flexible tubes. The reduced integration used for overcoming shear locking in two node elements, will result in the formation of spurious modes leading to an incorrect prediction of natural frequencies and velocity. These modes will not occur while using three node elements. Influence of spatial as well as temporal discretisation on the time lag and frequency are also discussed. The sensitivity analysis shows that the time lag varies linearly with the mass flow rate.     

A system to estimate coarse particle velocities at the pipe wall in settling slurry flow

Understanding and modelling of coarse particle (settling) slurries in laminar flow in pipes is still not complete. Better analytical procedures will enable more efficient pipeline operation which could result in significant power savings in the transport of coarse materials. The depth and velocity of the settled bed significantly affect the pressure gradients required to pump these slurries and knowledge of these parameters is of importance. Existing measurement techniques that can be used for monitoring of these complex fluid systems are either very expensive, not easy to implement or mostly used under controlled laboratory conditions. A new measurement system was developed that is capable of detecting the deposition of solids, the depth of a settled bed and the velocity of coarse particles at the pipe wall, around the pipe circumference. Particle velocities were determined by cross-correlating modulated signals from pairs of electrodes mounted flush with the pipe wall, in contact with the slurry. Tests were conducted using a mixture of acetal beads in water, at bulk velocities between 0.5 and 4 m/s. Estimated particle velocities from the cross-correlation analyses were compared with those obtained using a 30 fps video camera combined with visual inspection, and found to be within ±6%, thus validating the viability of the system. Particle velocity resolution using the cross-correlation technique is limited by block size, sample rate and the measurement distance between electrode pairs. Further test work with a range of real slurries (different particle sizes, solids concentrations and rheologies) needs to be conducted, along with more extensive verification of the results, to establish the limits of the system. Initial testing and evaluation of the system, which is capable of coarse particle flow monitoring in real time, showed significant potential for development of a new commercial sliding bed detector that can be used in a wide range of industrial applications in which particles are transported hydraulically.     

Characterization of flow homogeneity downstream of a slotted orifice plate in a two-phase flow using electrical resistance tomography

Two-phase flows are complex and unpredictable in nature, commonly encountered in a majority of fluid transport systems. The accurate measurement of two-phase flow is critical for a wide range of applications from wet stream to multiphase flows. There are different methods to meter two-phase flow in various industries. One approach is to produce a flow meter that does not require the individual flow components to be separated and measured separately. This goal can be met if a homogenized mixture is produced which can be measured by a standard single phase flow meter. The slotted orifice plate was invented as a flow meter for single phase flows, it is independent upon upstream flow conditions. Slotted orifice plate flow meter's utilization in two-phase flow revealed that it is highly capable of working as a flow conditioner transforming most of the multiphase flow regimes into a fairly uniform mixture. This study measures how the relative homogeneity of an air/water mixture varies downstream of the slotted plate in a horizontal pipe for various upstream conditions including elongated bubble and slug flow regimes using electrical resistance tomography (ERT). According to this study, the optimal location with a maximum homogeneity was determined to be between 1.5 and 2.5 pipe diameters downstream of the slotted orifice plate. This indicates that placing a slotted orifice plate at the obtained distance upstream of another flow meter such as a venturi coupled with a density measuring device like a radiation based densitometer or an electrical impedance device will help in obtaining accurate multiphase flow rate measurement.     

软性磨粒流磨粒入射壁面过程及其加工特性研究

针对利用两相流模型无法计算高浓度固—液两相流固相颗粒撞击壁面时颗粒速度的问题,提出一种边界层内颗粒运动轨迹计算模型,基于Mixture两相流模型和Realizable k-ε湍流模型仿真计算结果,通过分析提取颗粒入射前速度、计算边界层厚度、建立边界层内速度场和颗粒运动分析可以得到颗粒撞击壁面时的速度和入射角度。分析加工表面动压力分布和磨粒体积分数分布,结合两种结构约束流道验证仿真结果与加工效果的对应关系。通过对试验结果的分析,为约束模块的设计提供依据。研究结果表明:磨粒入射速度、磨粒体积分数和加工表面所受动压力大小直接影响工件加工效果,并与材料去除量成正相关关系;在本次试验中选择的工艺参数导致加工材料去除量小,适合初始粗糙度低的工件表面加工,对于此次试验的初始粗糙度应在0.2μm以下;约束模块的设计除了要考虑磨粒流流场特性之外,还要对加工表面的原始加工痕迹作详细了解,为约束模块的设计及加工工艺参数提供参考。     

An instrumentation system for multi-parameter measurements of gas-solid two-phase flow based on Capacitance-Electrostatic sensor

In this paper, an instrumentation system for the measurements of local solid volumetric concentration, local solid velocity, local solid mass flowrate and solid mass flowrate in gas-solid two-phase flow system is developed. It is based on a new type of a Capacitance-Electrostatic sensor (CES). The CES sensor is mainly composed of a capacitance electrode array and two electrostatic electrode arrays. The optimum design of the sensor is achieved by finite element method. The capacitance electrode array is employed to detect the solid distribution over the cross-section of the pipe, and the local solid volumetric concentration measurement is further derived. The electrostatic electrode arrays are used to measure the local solid velocities in conjunction with cross-correlation method. From the local solid velocity and local volumetric concentration, the solid mass flowrate and the local solid mass flowrate can be achieved. The developed system for the local solid volumetric concentration measurement is verified through analogue simulation experiments and static experiments. Finally, the system is employed to measure the local solid volumetric concentration, local solid velocity, local solid mass flowrate and solid mass flowrate on a belt conveyor. The experimental results show that the measurement error of the local solid volumetric concentration measurement results are less than 10.43% for solid local volumetric concentration ranging from 0.02 to 0.56, the standard deviations of the local solid velocity measurement results are less than 0.42 for solid velocity ranging from 3.5 m/s to 15.0 m/s, and the relative error of the solid mass flowrate is within −19.6% to +14.9% for solid mass flowrate ranging from 0.006 kg/s to 0.103 kg/s, indicating that the system is capable of achieving multi-parameters measurement in gas-solid two-phase flow system.     

A study of mass flow rate measurement based on the vortex shedding principle

Mass flow rate measurement is very important in the majority of industry processes because the mass of fluid is not affected by ambient temperature and pressure as the volume will be. Conventional mass flow rate is normally derived from the volumetric flow rate multiplied by fluid density. The density can be obtained by a densitometer or calculated according to the temperature and pressure measured by a thermometer and pressure gauge respectively. However the measurement accuracy is not always satisfactory. Flowmeters directly measuring mass flow rate have been studied and developed recently, such as Coriolis and thermal flowmeters. Unfortunately they still have some limits in practical applications. A new method in which mass flow rate can be directly measured based on the vortex shedding principle is presented in this paper. As a vortex flowmeter, von Kàrmàn vortex shedding is generated by a bluff body (vortex shedder), leading to a pressure drop and pressure fluctuation. A single differential pressure sensor is employed to detect the pressure difference between upstream and downstream sides of the vortex shedder. Both vortex shedding frequency and pressure drop are contained from the output signal of the differential pressure sensor, so that the mass flow rate can be obtained from the pressure signal. Numerical simulation has been done to analyze the characteristics of the fluid field and design the measurement device. The Computational Fluid Dynamics (CFD) codes Fluent were used in the numerical simulation. Experiments were carried out with water and gas, and the results show that this method is feasible and effective to measure the mass flow rate. This method has also robustness to disturbances such as pipe vibration and fluid turbulence.     

Numerical study of gas-solid reactive flow in a cylinder

A theoretical model is described for the numerical simulation of unsteady gas-solid reactive flow in packed beds of granular propellants in a tubular geometry. The approach couples continuity, momentum and energy equations in each phase along with porosity and granular stresses. Propellant bed combustion process is neglected, but the mass flow rate due to the combustion is derived using a pressure-based burning rate correlation. As the burning of solid propellant begins, pressure, temperature, density and other gas parameters begin to change rapidly. To catch these changes, a CFD approach with explicit McCormack method is used to solve the coupled system of equations. A moving mesh is used to consider the moving boundary. Pressure history, moving boundary velocity and other parameters of the mixture are obtained and compared with other numerical data.     

The fractal flow conditioner for orifice plate flow meters

The sensitivity of orifice plate flow meters to the quality of the approaching flow continues to be a cause for concern in flow metering. The distortions caused by pipe fittings such as valves, bends, compressors and other devices located upstream of the orifice plate can lead to non-standard velocity profiles and give errors in measurement. The design of orifice plate meters that are independent of the initial flow conditions of the upstream is a major goal in flow metering. Either using a long straight pipe, or a flow conditioner upstream of an orifice plate, usually achieves this goal.The effect of a fractal flow conditioner for both standard and non-standard flow conditions was obtained in experimental work and also using simulations. The measurement of mass flow rate under different conditions and different Reynolds numbers was used to establish a change in discharge coefficient relative to a standard one. The experimental results using the fractal flow conditioner show that the combination of an orifice plate and a fractal flow conditioner is broadly insensitive to upstream disturbances.The simulation results also show that the device can be used as a part of a flow metering package that will considerably reduce installation lengths. Previous work with orifice plates has shown that a combination of flow conditioner and orifice plate was promising. The results of using a combination of the fractal flow conditioner and orifice plate for non-standard flow conditions including swirling flow and asymmetric flow show that this package can preserve the accuracy of metering up to the level required in the Standards.     

Erosion in a curved pipe

The erosion in a curved pipe carrying a gas-particle mixture has been investigated. The fluid mechanics of such a system were solved under several idealized assumptions to obtain information about the impact velocities, the impingement angle and the mass of particles impinging per unit time per unit area. The results are presented in terms of the maximum relative erosion rate at the central plane of the curved pipe and it has been found that under the assumptions used the maximum relative erosion rate E_max can be expressed as being proportional to W³Z_L for large W and to W^3.93Z_L for small W, where W is the initial flow velocity of the mixture and Z_L is the loading ratio (mass of particles/mass of gas).     

Application of dual-plane ERT system and cross-correlation technique to measure gas–liquid flows in vertical upward pipe

The progress of process tomography provides a new method for two-phase flow measurement. The dual-plane electrical resistance tomography (ERT) is combined with the correlation measurement technique to carry out the two-phase flow measurement in which the continuous phase is conductive liquid. The method of the estimation of void fraction and the disperse phase velocity by extracting the eigenvalue of the dual-plane ERT boundary measured data is presented. This method is applied to the transient flow-rate measurement of the air–water two-phase flow in vertical pipe. The information of disperse phase void fraction and distribution variation with time change can be considered adequately, and the estimated value of disperse phase void fraction and velocity can be gained fairly accurately in this method, which provides the data for the calculation of the transient flow-rate. The experiment results indicate that this kind of measurement method is valid when the distance between the upstream and downstream measured cross section is short enough.     

Comparison of Particle Image Velocimetry and Laser Doppler Anemometry measurement methods applied to the oil–water flow in horizontal pipe

In this work, a comparison of Particle Image Velocimetry (PIV) and Laser Doppler Anemometry (LDA) measurement methods was made applied to oil–water two-phase flow in a horizontal pipe. The experiments were conducted in a 15 m long, 56 mm diameter stainless steel pipe using Exxsol D60 oil (density 790 kg/m³ and viscosity 1.64 mPa s) and water (density 996 kg/m³ and viscosity 1.0 mPa s) as test fluids. The experiments were performed at different mixture velocities and water cuts. Mixture velocity and water cut vary up to 1.06 m/s and 0.75, respectively. The instantaneous local velocities were measured using PIV and LDA, and based on the instantaneous local velocities mean velocities and turbulence profiles are estimated. The measurements are performed in the vertical plane through the pipe center. A double-pulsed Nd:yttrium aluminium garnet (YAG) laser and a high-speed camera with 1260×1024 px resolution (1.3 Mpx) were used for the PIV measurements. The LDA set-up is a two-colour backscatter system with 3 W Argon-Ion Laser. The time averaged cross-sectional distributions of oil and water phases were measured with a traversable gamma densitometer. The measured mean axial velocity and turbulence profiles using PIV were observed to compare favourably well with LDA measurements. Nevertheless, the PIV measurements are more sensitive for optical disturbances in the dispersed region close to the oil–water interface. Hence, this region cannot be confidently analyzed using PIV, whereas LDA offers full-field measurements even at higher mixture velocities.     

Systematic investigation of pipe flows and installation effects using laser Doppler anemometry—Part II. The effect of disturbed flow profiles on turbine gas meters—a describing empirical model

For systematic investigations of installation effects and for finding efficient ways to minimise these effects, a research project was initiated at the PTB. It covers the design of an automated test facility using a laser Doppler anemometer, the measurement of velocity profiles downstream of several pipe configurations and flow conditioners, as well as the measurement of the change in the gas meter behaviour, namely the shift of the error curve due to the disturbed velocity profiles.

Part I of this paper (presented in this issue) describes the test facility for the investigation of installation effects and shows the relation between pipe configuration and disturbed flow profile for a wide variety of pipe configurations and flow conditioners.

The second part compares the error shift of turbine meters with the characteristic of disturbed flow profiles. For this, three flow field parameters are used to quantify the disturbances of the velocity profiles such as the swirl intensity, flatness and asymmetry of the profile. Considering this, an empirical model is presented to explain the error shift of a turbine meter as a function of these three flow field parameters. The model will be verified for three types of turbine meters and the results will be discussed.     

On-line measurement of oil/gas/water mixtures,using a capacitance sensor

This paper describes the use of a non-intrusive capacitance transducer for the simultaneous on-line measurement of water and undissolved gas in crude oil. The water concentration of the flow is determined from the mean capacitance of the flowing mixture. The same transducer can be used simultaneously to determine the void fraction of the flow by measuring the instantaneous variation in the permittivity of the mixture created by fluctuations of the gas component. There is an interaction between these two measurements but this can be decoupled. This measurement principle was investigated experimentally, using process oil/water/air mixtures. Results are reported for mixtures with water contents up to 40% v/v and void fractions up to 15% v/v.A three-component volumetric flowrate measurement system, based on the capacitance transducer discussed in this paper was proposed. The work reported and the conclusions drawn relate to tests using mixtures of lubricating oil/fresh clean water/air. No work has so far been carried out in mixtures of crude oil/saline waters/hydrocarbon gases typical of the intended applications, nor have estimates been made of the effects that the variable nature of these actual components might have on the performance of the technique. Also, the tests were made on a Perspex section of pipe at nominal pressure whereas the real-life application would require measurements on a high-pressure/high-integrity stell pipeline; the practical problems of adapting the technique to the latter situation have not been examined though they are considered to be resolvable.