Lingzhen Kong,Jiaqing Chen,*,Tian Lan,Huan Sun,Kuisheng Wang

1 School of Mechanical Engineering,Beijing Institute of Petrochemical Technology,Beijing 102617,China

2 Beijing Key Laboratory of Pipeline Critical Technology and Equipment for Deep Water Oil&Gas Development,Beijing 102617,China

3 College of Mechanical and Electrical Engineering,Beijing University of Chemical Technology,Beijing 100029,China

Keywords: Atomization mixing Liquid jet Primary breakup Droplet breakup Droplet size

ABSTRACT For the design and optimization of a tubular gas–liquid atomization mixer,the atomization and mixing characteristics of liquid jet breakup in the limited tube space is a key problem.In this study,the primary breakup process of liquid jet column was analyzed by high-speed camera,then the droplet size and velocity distribution of atomized droplets were measured by Phase-Doppler anemometry(PDA).The hydrodynamic characteristics of gas flow in tubular gas–liquid atomization mixer were analyzed by computational fluid dynamics(CFD)numerical simulation.The results indicate that the liquid flow rate has little effect on the atomization droplet size and atomization pressure drop,and the gas flow rate is the main influence parameter.Under all experimental gas flow conditions,the liquid jet column undergoes a primary breakup process,forming larger liquid blocks and droplets.When the gas flow rate(Qg)is less than 127 m3·h?1,the secondary breakup of large liquid blocks and droplets does not occur in venturi throat region.The Sauter mean diameter(SMD)of droplets measured at the outlet is more than 140 μm,and the distribution is uneven.When Qg>127 m3·h?1,the large liquid blocks and droplets have secondary breakup process at the throat region.The SMD of droplets measured at the outlet is less than 140 μm,and the distribution is uniform.When 127

1.Introduction

Atomizing the liquid solvent into micron-size droplets can significantly increase the specific contact surface area between gas and liquid.The excellent followability of micron droplets in the air flow is also conducive to the uniform dispersion and mixing between gas and liquid,which can intensify the gas–liquid mass transfer process in the energy and chemical industry.The droplet distribution and gas–liquid mass transfer characteristics of Venturi scrubbers [1–3]and spray towers[4–6]have been widely studied,in which enhance gas–liquid mass transfer by atomizing absorbents.Based on the advantages of gas–liquid atomization mixing enhancing mass transfer,relevant scholars have developed tubular gas–liquid contactors,such as tubular gas–liquid contact absorbers for natural gas dehydration and deacidifi-cation[7,8],and tubular gas–liquid contact absorbers for volatile organic chemicals(VOCs)purification[9,10].This technology has a broad application prospect in energy and chemical industries.The common feature of these compact and efficient tubular gas–liquid contactors is that the liquid is injected into a high-speed gas folw through uniformly distributed jet holes in the pipeline space,and the liquid jet is sheared and broken into micron droplets with small diameter and uniform size distribution.Those micron droplets are dispersed and contacted with gas in the strong turbulent folw,thus making gas–liquid mixing and mass transfer more effciient.The key for gas–liquid atomization mixing is to effectively atomize the liquid phase,and the mixing effect is also related to the droplet dispersion space and the gas phase folw conditions in the tube.As shown in Fig.1,a special folw channel structure is designed to form a high-speed air folw zone in the tube,so that the large droplets or liquid blocks generated by the primary breakup can have secondary breakup to form tiny droplets.The primary breakup and secondary breakup mode of the liquid jet is related to the droplet number,droplet size and spatial distribution of the atomized droplet group,and directly affects the design of the gas–liquid atomization mixer.

Fig.1.Schematic diagram of liquid jet atomization process in a tubular gas–liquid mixer.

The primary breakup of liquid jet is one of the basic process of spray.The current research is mainly focused on the characteristics of liquid jet breakup mode,penetration depth and so on.The liquid jet column bends under the action of aerodynamic force in the tube and produces surface wave[11].The shape change of jet column affects the force acting on its surface,and the force and shape change frequently.When the force acting on the liquid jet column is instantaneously unbalanced at a certain time,the liquid jet column is randomly broken into different sizes of droplets.Wu et al.[12]and Mazallon et al.[13]carried out the experimental study on the breakup process of liquid jet in subsonic gas crossflow.The breakup mechanism of liquid jet was observed and analyzed by pulse shadow technique.The liquid jet breakup mode can be divided into enhanced capillary breakup,bag breakup,multimode breakup and shear breakup.Wu also discussed the wavelength of the surface wave along the liquid column,the deformation of the liquid column happened before the liquid surface began to break,and the conditions under which the liquid droplets began to form on the liquid surface,and analyzed the influence of the liquid jet velocity on the penetration depth of the liquid jet.By the analysis of the aerodynamic force imposed on the cylindrical liquid jet column,the liquid jet trajectory correlation in terms of the liquid–gas momentum ratio is derived,and the resistance coefficient of the liquid jet is estimated by the jet penetration measurement.Following the work of Wu and Mazallon,Sallam et al.[14]studied the primary breakup process of turbulent liquid jet in subsonic flow,and analyzed the velocity,trajectory,time and length of jet breakup.At the same time,the secondary breakup regimes of liquid jet can be divided into:vibrational breakup,bag breakup,multimode breakup,shear breakup and catastrophic breakup.Inamura[11,15]and Oda et al.[16]have studied jet penetration,jet width and spray characteristics in subsonic crossflow.Miller [17]studied the spray characteristics of liquid jet in subsonic gas crossflow near the downstream field of the nozzle by microscale holographic measurement technique,and the reconstruction hologram of liquid jet atomization was obtained.Olinger et al.[18]holographically measured the atomization effect of liquid jet in subsonic air crossflow and reconstructed the spray field in three dimensions.Numerous regime maps for predicting the primary breakup modes of a non-turbulent liquid jet injected into subsonic air crossflow have been proposed in the literature.By analyzing these maps,it was found that the gas Weber number(Weg)seems to be the most effective dimensionless parameter for defining the primary breakup modes.However,there are other different types of maps in the literature,such as Weg?q [12,19–21],Weg?Wej[22],Oh ?Weg[13,14],Weg?(λc/dj) [13,23],Weg?(λs/dj)[13,14].Mashayek et al.[24]summarized the atomization model of droplets for liquid jet breakup process.For most of the modeling methods,the liquid jet column is treated as a group of discrete large droplets.Some models are to make the atomization droplets follow a certain trajectory.By modifying the resistance relationship of the droplets,Some models ensure that the trajectories of the movement for the droplets is similar to the trajectories of the transverse liquid jet and the droplets can deflect and break up quickly under the strong action of air crossflow.

The formation mechanism of liquid droplets in the near-field and far-field regions of the liquid jet in subsonic crossflow is different[25].In the breakup process of liquid jet,liquid blocks and droplets are dominated by the breakup of liquid column and surface wave,and small droplets are mainly generated by surface wave breakup of the liquid jet column[26].Liquid blocks and large droplets generated by the primary breakup are broken up into small droplets in secondary breakup,thus forming a spray in the far-field regions.During the primary breakup process,the breakup characteristics of droplets are weakly affected by the cross-flow velocity,and are mainly affected by the exit conditions of the nozzle(such as turbulence,internal geometry of the nozzle).Under non-turbulent and turbulent conditions,the size of the ligaments and droplets formed by the primary breakup of liquid jet increases along the flow direction of liquid jet,and the velocity of droplet has nothing to do with the droplet size.However,the characteristics of the droplets in the secondary breakup strongly depend on the velocity of the air crossflow,and the size of the droplets is obviously related to the velocity of the transverse airflow.From the droplet size distribution of the downstream liquid jet column measured in the experiment,it is found that the droplet size distribution is related to the transverse airflow velocity ug[19,27–29].Some studies have also analyzed the relationship between the droplet distribution and the downstream position of the liquid jet.The results show that at low Weg(that is,the main primary breakup mode is column breakup),the droplet size reaches the maximum at the outer edge of the spray plume,where the droplet is generally large[11,30,31].This is because large droplets have higher momentum and longer velocity relaxation time,so that they can penetrate more distant transverse airflow.However,at high Weg(where surface breakup is dominant),the droplet size reaches its maximum in the core area,because larger droplets cannot penetrate farther due to the greater aerodynamic force of the transverse airflow.In essence,the surface breakup produces smaller droplets in the primary breakup process,while the droplets produced in the column breakup process can easily follow the air crossflow.Nevertheless,under the condition of low Weg(~33),Lubarsky et al.[31]also obtained the bimodal droplet size distribution in the downstream region of liquid jet,which indicates that there is a multi-mode breakup process of the liquid jet under the condition.Becker and Hassa[19]reported that the effect of air crossflow velocity under high pressure is much weaker than that under atmospheric test conditions,which means that there is no strong correlation between droplet size and Wegunder high pressure test conditions.Farvardin et al.[32]studied the effect of liquid viscosity on the downstream droplet size distribution of the nozzle with a diameter of 0.5 mm,and reported that the droplet size distribution is similar at 50 mm.This is because the droplet becomes very small at this flow position,so secondary breakup will not occur again.Song et al.[33]established a relationship predicting the average droplet size in terms of Weg,Rej/Reg(or q),liquid–gas density ratio and viscosity ratio,which considered the airflow pressure and liquid characteristics.Poozesh et al.[34,35]used a time-resolved particle image velocimeter to study the behavior of near-field droplets at the tip of the two-fluid nozzle.The results showed that the non-dimensional break-up length was purely dependent on the momentum flux ratio per unit volume,M.On the other hand,the spray angle was a function of both M and We numbers.By increasing We number it was possible to shift the droplet velocity distributions to produce wider distributions in droplet sizes.

The flow wake region and droplet size caused by the liquid jet breakup are the two main factors affecting the downstream droplet velocity distribution of the liquid jet in the air crossflow.The interaction between these two factors determines the droplet velocity distribution [24].There is a lower-velocity zone in the spray wake behind the liquid jet(that is,the flow wake zone),but the peak velocity is on the outer and inner peripheries of the jet [11,30,32].Therefore,the wake region is used to indicate the level of momentum exchange between the liquid jet column and the transverse airflow [30].Mashayek et al.[24]asserted that due to the breakup of liquid column,the influence of liquid column on the flow decreases with the increase of the liquid jet streamwise distance,z.Therefore,this leads to a more direct momentum exchange between the droplets and the transverse airflow,resulting in the acceleration of the droplets.The larger the droplets are,the more slowly they move downstream.Moon et al.[36]use the ultrafast Xray phase-contrast imaging(XPCI)technique to characterize the velocity and turbulence intensity fields of the high-speed diesel sprays in the near-field.The effects of the orifice inlet geometry and injection pressure on the near-field dynamics of the diesel sprays are investigated.The spray shows stronger momentum exchange with surrounding gas from around 35 mm location.Inamura et al.[11]believe that the downstream droplet velocity of the liquid jet in a cross airflow is a function of the gas average velocity ug(or Weg),while Elshamy et al.[37]attributed the variation in the droplet velocity distribution to the change of the wake region size and the Weg.With the increase of Weg,the breakup mode of liquid jet changes from column breakup to surface breakup,resulting in corresponding changes of droplet size and velocity distribution.By studying the droplet velocity components under different Weg,Lubarsky et al.[31]found that there is a considerable lag between the droplet velocity in the cross airflow direction and the average velocity of droplets entering the air flow,especially in the core of spray.

The existing research work shows that the spray characteristics of a liquid jet in air crossflow are mainly affected by gas–liquid velocity ratio,Weber number and specific parameters of airflow.However,the liquid jet breakup mode map,jet trajectory empirical formula and droplet size prediction formula are only limited to specific experimental conditions,so the results do not have universality.The above studies,in addition,mainly focus on the liquid jet breakup process in a uniform airflow,but there are limited studies on the liquid jet breakup mechanism in a non-uniform airflow.The atomization mixing characteristics of a liquid jet in the complex airflow,caused by the Venturi and cone structure in the tubular gas–liquid atomization mixer,have not been fully studied.It's easy to atomize liquid and conduct gas–liquid mixing in a large airflow space.Generally,the liquid medium is sprayed into the gas phase medium by an atomization nozzle.Losing a certain pressure energy,the liquid medium has a certain kinetic energy and interacts with the gas phase medium to form droplets.The atomized droplets have a certain residual kinetic energy and diffuse in a specified area of the gas space,thus achieving the mixing between the gas and liquid media.The atomizing and mixing effect of liquid depends on the space of gas medium.However,the diameter of tubular gas–liquid mixing contact equipment is generally in the range of 50 to 300 mm.Although the pipe length is little limited,but the limited pipe diameter restricts the atomization space,mixing region and mass transfer space in the radial direction.In order to achieve efficient gas–liquid mixing in the narrow tube space,improvements such as liquid jet and variable cross-section structure are needed[38].For ensuring simple structure design,efficient liquid atomization and excellent mixing effect,the difficulty of realizing large flow rate gas–liquid atomization mixing in a small pipeline space should be solved.

Based on the gas–liquid atomization theory,the basic structure of tubular gas–liquid atomization mixer is proposed,and the testing experimental platform is designed.The atomization mechanism and mixing characteristics of liquid jet in the flow channel of preliminarily designed tubular atomizer are tested,and the basic atomization conditions of liquid jet are analyzed.Computational fluid dynamics(CFD)numerical simulation,high-speed camera and Phase-Doppler anemometry(PDA)are used to analyze the primary breakup form of the liquid column,the size and velocity distribution of droplets at the outlet,and the velocity and turbulence intensity distribution of airflow in the experimental section.The study results can be utilized as the basis for designing the tubular atomization mixer,and provide theoretical guidance and data support for the optimization of the tubular atomization mixer.

2.Experimental Apparatus and CFD Modeling Methods

2.1.Experimental apparatus and test methods

The atomization characteristics of tubular atomization mixer are tested by the self-designed experimental platform,as shown in Fig.2.The experiment is carried out at an ambient room temperature and atmospheric pressure.Compressed air is provided by a variable frequency screw air compressor,which can adjust the rotate speed according to the gas flow rate,and the gas buffer tank increase the stability of the gas supply.Gas pressure for the inlet of the experimental pipeline is adjusted by the pressure reducing valve.The gas flow rate is controlled by the regulating valve and measured by the vortex flowmeter with temperature and pressure compensation.To ensure the stability and uniformity of the gas flow in the inlet of experimental section,a orifice type gas distributor is installed in the upstream of experimental section.The outlet pressure of the experimental tube section is atmospheric pressure,so it can be considered that the pressure drop of the atomization element is equal to the measured inlet pressure.A pressure sensor(0–10 kPa with an accuracy of 0.5%FS)is installed at the inlet of the test section to measure the pressure drop when the gas pass through the atomization element.

A 50 mm×50 mm plexiglass square tube,in which atomization elements are installed,is used in the experimental section.Since the flow rate of atomized liquid is small,the conventional liquid supply pump is difficult to ensure the stable supply of liquid.The compressed air in the high pressure gas cylinder is injected into the pressure tank after decompression through the pressure reducing valve.The gas pressure,driving the liquid flow,in the supply tank can be adjusted by the pressure reducing valve,so as to achieve the stable supply of a micro amount of liquid.The morphological change during liquid jet primary breakup in the air crossflow is captured by a high-speed camera,and the injecting,bending and breaking process of the liquid column are analyzed.The light is illuminated by an adjustable LED matrix,and the clarity of images is ensured by adjusting the incident angle and brightness of the LED matrix.The MotionPro Y3 high-speed camera,made by IDT Company of the United States,is adopted,and the highest frame rate is 6000 fps at its maximum resolution(1280×1024 pixel).The droplet size and velocity are measured by PDA,and the droplets in the gas–liquid two-phase flow are collected through an L-shaped pipe whose diameter is larger than that of the test section.

The detailed structure of the test pipe section is shown in Fig.3.The experimental tube section is made of transparent plexiglass,and the atomization element is composed of Venturi structure and inverted triangle structure.The liquid is injected into the high-speed airflow through the small holes in the triangular cone wall.The diameter of the nozzle hole is 0.5 mm,and the ratio of length to diameter(L/dj)is 9.In order to measure droplet size and droplet velocity by PDA,7 measuring points are set at 5 mm form outside the outlet of the test section.The distance between the outermost measuring point and the wall is 1 mm,and the uniform distribution distance of each measuring point is 8 mm,as shown in Fig.3(c).Each measurement point needs to collect effective data of 2000 droplets,or the measurement time is up to 6 min.For the PDA system,the measurement accuracy of the droplet size is less than 1%Dmax+1%Dmeasurement,and repeatability is 0.5%.The Sauter mean diameter of droplets were based on 2000 samples,which results in statistical uncertainties around 3% for the droplet size value.With the statistical uncertainties,systematic errors and random errors taken into consideration,the overall uncertainties of the droplet Sauter mean diameter can be estimated to be±5%.

The influence of gas flow rate on atomization characteristics is analyzed when the liquid–gas volume ratio(L/G)is about 2.1×10?5.The experimental parameters are shown in Table 1.When the gas flow rate is 127 m3·h?1,162 m3·h?1and 199 m3·h?1,the influence of liquid flow rate on atomization characteristics is analyzed respectively.The experimental parameters for different liquid–gas volume ratios(L/G)are shown in Table 2.When the gas flow rate is 127 m3·h?1,three liquid flow rates are considered,and four liquid flow rates are considered under the other two gas flow rate conditions.The range of L/G is between 1.5×10?5and 2.3×10?5.

Table 1Gas–liquid flow parameters for the experiments with different gas flow rates

Table 2Gas–liquid flow parameters for the experiments with different liquid–gas volume ratios

Fig.2.Schematic diagram of the experimental platform.(1-Air compressor,2-Gas buffer tank,3-Pressure reducing valve,4-Regulating valve,5-Gas flowmeter,6-Gas distributor,7-Pressure sensor,8-High speed camera,9-Test section,10-PDA,11-Collector tube,12-Liquid collecting tank,13-LED matrix,14-Needle valve,15-Liquid flowmeter,16-Liquid supply tank,17-High pressure air cylinder).

Fig.3.Schematic diagram of the experimental section:the structure dimension(a),actual experimental structure(b),and the layout of measurement points for PDA at the outlet(c).

2.2.CFD modeling method of gas phase flow field

The primary and secondary breakup processes of liquid jet in a limited pipeline space are influenced and dominated by the high-speed air flow.The flow field distribution of air flow in the atomizer plays an important role in analyzing the characteristics of liquid atomization and gas–liquid mixing.For numerical simulation method,if the multiphase flow model considering gas–liquid interface tracking is adopted,the calculation process will be more complex and has enormous computational cost.In this work,the flow rate of liquid injection is small,so the liquid column and droplets have little influence on the gas flow field.In order to simplify the problem,the single-phase air flow field in the experimental section issimulated by FLUENT software.According to the structural parameters of experimental pipe section,three-dimensional geometry of the gas channel is established and meshed,as shown in Fig.4.A hexahedral grid,containing 1.92 million elements,was exported into the solver software,Ansys Fluent v.15(Ansys Inc.,Canonburg,PA).The mesh is refnied near the wall of the tube and atomizing component where large velocity gradient exists.Mesh dependence check shows that this meshing scheme can meet the requirements of numerical simulation.

Fig.4.3D geometric model(a)and mesh(b)of the experimental section.

The airdensity andviscosityin thenumericalsimulationaresetto 1.225 kg·m?3and 1.7894×10?5kg·m?1·s?1respectively.Theinlet boundary condition is set as the velocity inlet,and the inlet gas velocity is calculated according to the gas flow rate set by the experiment.Pressure outlet is selected as the outlet boundary condition of the test section,and the pressure is set at 0 Pa.Gas turbulence flow was simulated by using the Reynolds Stress model with enhanced wall functions.Pressure–velocity coupling is achieved by employing the SIMPLEC algorithm.The second-order upwind schemes are selected for the discrete method,which can avoid the false diffusion problem of the first-order upwind schemes when the Peclet number (Pe) is large.After the above settings are completed,the gas phase flow field in the experimental section for different gas flow rates is numerically simulated.

3.Results and Discussion

3.1.Morphology of liquid jet breakup

The primary breakup characteristic of the liquid jet in a high-speed airflow region of experimental tube section is very important to improve the performance of tubular atomization mixer.Fig.5 shows the breakup images of liquid jet column ejected from the small holes in the triangular cone under different gas flow rates.There is a large relative velocity between the liquid jet and the air crossflow in the tube.Under the action of external gas drag and internal turbulence,the surface of the liquid jet column becomes unstable and bends along the airflow direction.The breakup of the liquid jet column appears in the tail region of the triangular cone,resulting in liquid blocks,liquid lines and droplets with different shapes and sizes in a very short time and space.The liquid blocks,liquid lines and large droplets,produced by the primary breakup process,move with the airflow to the throat region of the Venturi structure,in which liquid still has relative movement to the surrounding airflow.Under the combined action of gas shear force and surface tension,the deformation of liquid blocks and large droplets will continue,and the amplitude of the disturbance wave on the liquid surface will continue to increase.When the gas flow rate is in the range of 72.64 to 106.97 m3·h?1,the secondary breakup of droplets does not appear in the throat region.When the gas flow rate is in the range of 127 to 216 m3·h?1,for the droplets passing through the throat region,the secondary breakup will happen.The secondary breakup mode of droplets changes with the increase of gas flow rate.For the gas flow rate is in the range of 127 to 162 m3·h?1,the secondary breakup mode of droplets is bag breakup.For the gas flow rate is in the range of 181 to 216 m3·h?1,the secondary breakup mode of droplets is shear breakup or catastrophic breakup.

In order to analyze the influence of liquid flow rate on the breakup morphology of liquid jet column,when the gas flow rate is 127 m3·h?1,the breakup morphologies of liquid column with different liquid flow rates are photographed by a high-speed camera respectively.It can be observed from Fig.6 that the liquid column ejected from the orifice bends and then breaks up into droplets under the action of air flow.For low liquid flow rate,the liquid jet column is close to the side wall of triangular cone.The distance from the liquid column to the side wall of triangular cone increases with the increase of liquid flow rate.Liquid jet column swings and breaks up into large droplets and liquid blocks in the air flow.When flow through the Venturi throat,these large droplets and liquid blocks break up into smaller droplets.The variation of liquid flow rate does not affect the secondary breakup mode of droplets in the throat region,and the secondary breakup mode of droplets is bag breakup when the gas flow rate is 127 m3·h?1.

3.2.Gas flow characteristics in the test section

Fig.5.Images of liquid jet breakup under different gas flow rate conditions for L/G≈2.1×10?5.(a) Qg=72.64 m3·h?1,(b) Qg=92.84 m3·h?1,(c) Qg=106.97 m3·h?1,(d) Qg=127.59 m3·h?1,(e) Qg=144.11 m3·h?1,(f) Qg=162.14 m3·h?1,(g) Qg=181.87 m3·h?1,(h) Qg=199.60 m3·h?1,(i) Qg=216.52 m3·h?1.

Fig.6.Images of liquid jet breakup under different liquid flow rate conditions.(a)Qg=127.89 m3·h?1,QL=0.039 L·min?1;(b)Qg=127.59 m3·h?1,QL=0.045 L·min?1;(c)Qg=127.50 m3·h?1,QL=0.050 L·min?1.

Fig.7.Velocity distribution of the central section for Qg=162 m3·h?1.

Fig.8.Streamline in the experimental tube for Qg=162 m3·h?1.

Fig.9.Turbulence intensity distribution of the central section for Qg=162 m3·h?1.

From the above analysis,it can be seen that the flow field distribution of air flow in the experimental tube is the key factor for the atomization process of a liquid jet.Fig.7 shows the velocity field distribution on the central section of the experimental tube for Qg=162 m3·h?1.It is observed that the velocity distribution in the inlet straight tube region shows the fully developed pipe flow characteristics,and the air flow velocity in the center region of the tube is about 16 m·s?1.An oval low-velocity airflow region appears at the upstream end of the central triangular cone.In the flow region between the triangular cone and the Venturi pipe wall,the gas velocity increases with the decrease of the flow channel area.The maximum gas velocity appears near the throat,and the value is 56.8 m·s?1.In the downstream area of the triangle cone,there is a low air velocity region in the center of the flow channel and a high gas velocity region near the pipe wall.In the outlet straight pipe region near the expansion section,the velocity is larger near the wall and smaller in the center.

Fig.8 shows the streamline in the experimental section of the atomizer for Qg=162 m3·h?1.It can be observed that the high-speed air flow between the expansion section and the downstream straight pipe section blows to the wall and make the flow direction parallel to the straight pipe wall.In the expansion section,the flow direction of the high-speed air flow points to the straight pipe wall.There is a symmetrical vortex region for the down stream of the triangular cone,which has symmetrically small vortex between the throat and the expanding section,and has symmetrically large vortex between the expanding section and the straight section.

Fig.9 shows the distribution of the turbulence intensity on the central symmetric plane of the experimental tube.The turbulence intensity in the expansion section and downstream straight pipe section increases significantly,and the region with the maximum turbulence intensity is located in the straight pipe section near the expansion section.The secondary breakup of droplets occurs in the throat and expansion section,and the micron droplets produced by secondary breakup can be better dispersed and mixed in the strong turbulent region of air flow.In the downstream region where the distance from the expansion section is three times the diameter of the pipe,the turbulence intensity gradually decreases and tends to be more uniform.

Droplet breakup mainly occurs in the area of high-speed air flow at the throat,and the value of gas velocity at the throat affects the droplet size and uniformity of the final atomized droplets.The velocity value at the dotted line in Fig.7 is extracted,and the velocity distribution curves of the throat under different gas flow rates are obtained,as shown in Fig.10.The velocity distribution at the throat shows the tendence that the value of velocity is relatively low in the middle and relatively high on the both sides.In the tail region of the triangular cone,for the coordinate from-7 to 7 mm,the gas velocity is in the range of 1.1–6.5 m·s?1.There is a high-speed zone from the coordinate point of 11 mm to the side wall of the throat.The gas velocity is about 29.5 m·s?1for Qg=72 m3·h?1and about 88.3 m·s?1for Qg=216 m3·h?1.Combined with the analysis in Fig.5,when the gas flow rate is within the range of 72.64–106.97 m3·h?1,the droplets will not break up when they pass through the throat,where gas velocity is between 29.5 to 44.2 m·s?1.When the gas flow rate is in the range of 107–162 m3·h?1,the secondary breakup mode of droplets is bag breakup,and the gas velocity at the throat is between 44.2 to 66.3 m·s?1.When the gas flow rate is in the range of 162–216 m3·h?1,the secondary breakup mode of droplets is shear breakup or catastrophic breakup,and the gas velocity at the throat is between 66.3 to 88.3 m·s?1.

Fig.10.Gas velocity distribution at throat under different gas flow rates.

Fig.11.Droplet velocity distribution at the outlet of the experimental tube section under different gas flow conditions.

3.3.Droplet velocity distribution

Fig.11 shows the droplet velocity distribution measured at the outlet of the experimental tube under different gas flow conditions.The distribution trend of droplet velocity is remains basically unchanged under different gas flow rates.With the same gas flow rate,the droplet velocity near the wall surface is becomes low,and the droplet velocity in the middle position becomes high,and the droplet velocity measured at the five measurement points in the middle of the pipeline tend to be the same.As the gas flow rate increases,the droplet velocity also gradually increases.Fig.12 displays a comparison between the gas velocity obtained by numerical simulation at the outlet of the test tube and the droplet velocity measured by experiment.Under the conditions of gas flow rates with 72 m3·h?1,162 m3·h?1and 199 m3·h?1respectively,the droplet velocity measured by PDA and the gas velocity obtained by numerical simulation are consistent with the velocity distribution of the fully developed turbulent in a tube.The droplet velocity value is slightly smaller than the gas velocity value.The micron droplets formed by atomization have a strong followability in the airflow.The liquid injected from the small holes on the cone surface is broken and atomized into small droplets under the action of high-speed airflow.There is a speed difference between the generated small droplets and the airflow.The droplets continue to accelerate under the action of the airflow.At the outlet of the test tube section,the acceleration force of the airflow on the droplets basically reaches a state of equilibrium,and the velocity of the droplets and the gas flow is essentially the same.

Fig.12.Comparison of the distribution of droplet velocity and gas velocity at the outlet of the test section.

3.4.Pressure drop characteristics

The pressure drop characteristic of gas flow is also an important indicator to evaluate the performance of tube atomizer.The pressure drop of the gas flowing through the atomizer is partly due to the energy loss during the gas acceleration and deceleration process in the contraction and expansion sections,and partly caused by the energy transfer between the air flow and the liquid jet in the atomization process.The inlet gas velocity of the experimental pipe section is converted according to the gas flow rate,and the relationship curve between the atomizer pressure drop and the square of gas velocity is obtained by squaring the gas velocity,as shown in Fig.13.With the increase of the gas flow rate,the pressure drop of the atomizer is increasing,and the resistance loss of gas is linearly related to the square of the gas velocity in the pipeline.When the atomized liquid is injected into the atomizer,withL/G≈2.11×10?5in the different gas flow rates,the pressure drop also increases linearly with the square of the gas velocity.The pressure drop is slightly increased compared to the single-phase gas flow.In order to further analyze the pressure drop characteristics of the atomizer,the resistance coefficient is also used to evaluate the resistance characteristics of the atomizer.The resistance coefficient is calculated by Eq.(1):

where,ΔPis the pressure drop,ΔP=Pin-Pout;PinandPoutare inlet pressure and outlet pressure of the experimental pipe section,respectively;ugis the gas velocity at the inlet of the atomizer;and ρgis the gas density.The variation curve of resistance coefficient with the gas flow rate obtained by processing the pressure drop data of the experiment is shown in Fig.13.The resistance coefficient is in a horizontal straight line with the increase of the gas flow rate,and the resistance coefficient of the single-phase gas flow is slightly smaller than that of the gas–liquid atomization process.The resistance coefficient under single-phase gas flow is about 2.55,and the resistance coefficient under gas–liquid atomization is about 2.73.

Fig.13.Variation of pressure drop and resistance coefficient with gas flow rate.

Fig.14.Effect of liquid–gas volume ratio on pressure drop of atomizer.

Fig.14 shows the effect of different liquid gas volume ratios on the pressure drop of the atomizing mixer.At the same gas flow rate,the pressure drop remains basically unchanged as the L/G increases.The main reason is that the liquid flow rate is much smaller than the gas flow rate,and the gas energy consumption in the liquid atomization process is far less than the total gas pressure energy.Therefore,when the L/G increases from 1.55×10?5to 2.29×10?5with the same gas flow rate,the increase of liquid flow rate causes a very small increase for the gas pressure drop,and the pressure value measured by the sensor dose not change obviously.

3.5.Droplet size distribution

Fig.15 shows the distribution of droplet size at the outlet of the test section under different gas flow rate conditions.It can be seen that under the same gas flow rate,the droplet sizes at different measurement points are basically the same,and the droplet size decreases gradually with the increase of gas velocity.With the increase of gas velocity,the droplet size distribution of the measurement points is more similar to a straight line,and the uniformity of atomized droplet distribution in the tube is better.With the decrease of gas velocity,the droplet size fluctuation between measurement points increases.This is mainly because large droplets are produced at low gas velocity and the followability of large droplets in the gas flow is relatively poor,resulting in a less uniform distribution of droplets in the tube.However,the size of atomized droplets becomes smaller with the increase of gas velocity,which enhances the followability of the droplets in the high-speed gas flow and finally makes the droplet distribution at the outlet of atomization mixing section more uniform.With the value of the gas velocity 72 m3·h?1,the sauter mean diameter of droplets is about 200 μm,and with the value of the gas velocity 216 m3·h?1,the sauter mean diameter of droplets is about 50 μm.

Fig.15.Droplet size distribution along the y direction at the outlet of the experimental section under different gas flow rates.

With the value of the gas flow rate 127 m3·h?1,162 m3·h?1and 199 m3·h?1,the distribution curves of droplet size at the outlet of the test section for different liquid injection flow rate are shown in Fig.16.The droplet size decreases with the increase of gas flow rate.When the gas flow rates are the same,the droplet size increases gradually with the increase of liquid injection flow rate.The droplet size distribution at different measurement points shows a W-shaped distribution with the gas flow rate 127 m3·h?1.When the droplets have secondary breakup in the throat region,if it's a bag breakup,the liquid film in the middle part of the droplet is broken into a plentiful of small droplets,and the liquid band in the edge part is broken into the large droplets.The velocity component of the large droplets,which is perpendicular to the direction of the gas flow,need a long time to be eliminated by the momentum exchange between the gas and liquid,leading to most of the large droplet being distributed near the central axis and the wall.The followability of small droplets is excellent,so the small droplets can keep the same velocity with the air flow through momentum exchange in a short time,and it is eventually distributed evenly at the outlet of the experimental tube section.The droplet size is larger in the area near the wall and central axis than somewhere else,and the size distribution of droplets has a W-shaped trend in the tube.Under the condition of gas flow rate of 162 m3·h?1and 199 m3·h?1,the droplet size distribution at different measured points tends to be a straight line.The droplets in the Venturi throat have pouch or catastrophic breakup that generates small droplets.Such small droplets have good followability in the air flow,and can be uniformly dispersed and mixed with the air flow in the dispersion mixing section.Under the condition of high gas velocity,the droplet distribution at the outlet of the test section can be more uniform.

Averaging the droplet sizes obtained from each measurement point,the sauter mean diameter(SMD)at the outlet under different gas–liquid conditions are shown in Table 3.When the gas flow rates are the similar,the SMD at the outlet increases with the increase of liquid flow rate.When the gas flow rate is 127 m3·h?1,with the increase of L/G from 1.83×10?5to 2.35×10?5,the average SMD at different measuring points increases from 123.16 to 133.87 μm,with an increase of 10.71 μm.When the gas flow rate is 162 m3·h?1,with the increase of L/G from 1.55×10?5to 2.29×10?5,the average SMD at different measuring points increases from 70.47 to 83.15 μm,with an increase of 12.68 μm.When the gas flow rate is 199 m3·h?1,with the increase of L/G from 1.50×10?5to 2.28×10?5,the average SMD at different measuring points increases from 54.57 to 62.48 μm,with an increase of 7.91 μm.With the increase of gas flow rate,the effect of liquid–gas volume ratio on the average SMD at the outlet decreases.

Table 3Statistics of average SMD at different measurement points under different liquid–gas volume ratios

Fig.16.Variation of droplet size with liquid jet flow rate.

Through the above analysis,it can be seen that the liquid flow rate has little influence on the droplet size and atomization pressure drop,and the gas flow rate is the main factor affecting the atomization droplet size and atomization pressure drop.When the liquid jet column is effectively broken up and atomized in the air flow,the liquid surface tension and viscosity force are the internal forces that maintain the original shape of the liquid column,and the aerodynamic force acting on the liquid column and the liquid block is the external force that makes the droplets unstable and break up.Because the change of the internal force of the liquid is limited by many conditions,the reasonable design of the gas channel structure of the tubular atomizer mixer can increase the gas–liquid velocity difference and the turbulence of airflow,which can produce a good atomization effect in the pipeline space.In the limited pipeline space,the setting of venturi and triangular cone structures can accelerate and decelerate the flow,and create the high-speed flow region in the liquid jet area and the throat area.Which not only increases the gas–liquid velocity difference but also increases the turbulence degree of the gas flow.After the liquid jet leaves the nozzle,the high-speed gas flow provides appropriate and sufficient aerodynamic force for the liquid jet to overcome the internal force maintained the present shape,and make the surface of the liquid jet change dramatically and become unstable in a very small distance.Thus,the continuous liquid jet is transformed into a group of dispersed droplets to achieve good atomization.Under all the gas flow conditions,the liquid jet has the primary breakup process to produce large liquid blocks and droplets.When the gas folw rate is less than 127 m3·h?1,the large liquid blocks and droplets do not have secondary breakup at the throat region,and the average SMD measured at the outlet is about 140 μm and the distribution of droplets is uneven.When the gas folw rate is greater than 127 m3·h?1,the large liquid blocks and droplets at the throat have secondary breakup and the distribution of droplets iseven.When the gas flow rate is 127 m3·h?1,the throat gas velocity obtained by numerical simulation is 51 m·s?1.In order to achieve excellent atomizing mixing effect,the designed tubular atomizing mixer shall ensure that the throat gas velocity is more than 51 m·s?1under the lowest operating gas flow rate.

4.Conclusions

Considering the variation of gas flow rate and liquid gas volume ratio,the liquid jet breakup morphology,droplet size and droplet velocity in tubular atomization mixer are studied by a specially designed experimental platform.Under different gas flow conditions,the CFD numerical simulation method is utilized to analyze the gas flow field characteristics of the experimental tube section.When the gas flow rate is in the range of 107 to 162 m3·h?1,the droplet secondary breakup mode is bag breakup,and the gas velocity in the throat is between 44.2 and 66.3 m·s?1.When the gas flow rate is in the range of 162 to 216 m3·h?1,the droplet secondary breakup mode is shear or catastrophic breakup,and the gas velocity in the throat is between 66.3 and 88.3 m·s?1.In the test tube section,the droplets accelerate and disperse continuously under the action of aerodynamic force,and the distribution of droplet velocity and gas velocity at the outlet are basically the similar.Regardless of the single-phase gas flow or the gas–liquid atomization mixed flow condition,the pressure drop of the experimental tube is linearly related to the square of the gas velocity.However,when the liquid to be atomized is injected,the pressure drop of the experimental tube increases slightly.With the increase of the gas flow rate,the droplet size decreases gradually.Under the same gas flow rate,the droplet size increases slightly with the increase of liquid flow rate.In order to achieve excellent atomizing mixing effect,the designed tubular atomizing mixer shall ensure that the throat gas velocity is more than 51 m·s?1under the lowest operating gas flow rate.

Declaration of Competing Interest

The authors declared that they have no commercial or associative interest that represents a conflict of interest in connection with the work submitted.

Acknowledgements

The financial supports from the National Natural Science Foundation of China(21808015)and the Project of Construction of Innovative Teams and Teacher Career Development for Universities and Colleges under Beijing Municipality(IDHT20170507)are gratefully acknowledged.