Scribd
Upload
75 views14 pages

Effect of Pulsation Intensity On The Flow

Uploaded by

Kuldhir
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
75 views14 pages

Effect of Pulsation Intensity On The Flow

Uploaded by

Kuldhir
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 14
International Jourel of Mechanical Selene 195 2021) 105162 Content lists available at ScienceDirect _ Msc International Journal of Mechanical Sciences a ELSEVIER journal homepage: www elsevierconocteimecsel _a Effects of pulsation intensity on the flow and dispersion of pulsed dual ® plane jets enger Sanjay Kumar®, Rong Fung Huang**, Ching Min Hsu * Dear of Mehanl Bering Neto! Ta Uninet of cee and Tey, Tagl Tara 10507, Repub of Cine “peeramnt hanel Dos Eine nal Porm erty Ya, Tao 207, Rea of China Ronse The ees of jet publion Invasiy on de Now and alspesion characteristics f dal parallel plane jes ware a plane es ‘experimenaly studied. The Jt pulsations wore generate by acoustic excitations. The astightshest assed aeton esty smoke flow visualization method was used o study the Now evolution proceses. The ltera preaing ch tet of the jets were measured using the binary edge detection method. A hotwie anemometer was used to mearure the velocity pusstions, mean velocities, turbulence intensities, and Lagrangian turbulent length and time scales. The dispersion charactensts of te jt Huis were measured by the Wace gas concentration dete tion method, The et Reynolds mimber and excitation Strouhal number were Gea 60 an 0,106, respectively “Te et polston intensities were varied from Oto 1,6. As the jets were pase, the trblence intensity, literal spread wide, and jt Sid dispersion capability marke increased compared with those ofthe non puedes The lange the pulsation intensity was, the higher the turbulence incensiy, lateral spread width, and jet Ould 2, ‘Anderson and Spall [15] compared the experimental and numerical results of two parallel plane jes, including the mean velocities, stream. ‘wise and transverse notmal stresses, and locations of merging and com bined points. Anderson etal (19) and Anderson (20) extended their previous work and determined the instability characteristics for low jet spacing to jet width ratios, The instability behavior in the spectra was Indiated by sharp peaks. They concluded that the vortex shedding phe ‘nomenon, similar to flows over bluff bodies, eppeered. Furthermore, ‘Anderson observed thatthe periodic behavior inthe jets was indepen ‘dent of the Reynolds number within the nozzle spacing ratio range of S/d = 05-1,7. At higher nozzle specing ratios, the instabilities were ~ Garesponding author. Department of Mechanical Enginsering, National Taiwan University of Science and Technology, No, #9, Section 4, Keshing Road, Tips 607, Taiwan, RO mal ede: sang ns sie (RF. Hang. napa org/10.1016/}.ncese 2020105182 Received 11 August 2020; Recelved in revised form 1 Octobe 2020; Accepted 30 October 2020 ‘Availabe online 4 Novernber 2020 (0020-7403/ 2020 Elsevier Led. Al rights reserved. Nomenclature © toca eoncenteaton of carbon dowide €, concentration of carbon donde a jet ext aids oft et, 4m 4% voltage of signal output Som function generator £, voltage of signal ourpt fom power amplifier foe excitation feequeney TE” jeepalsation intensity (=1'/4,) Tenge of jet exit 145 mm 1, eng of nial jet cotamn Lagrangian integral length sale of turbulence ees 1m mass low rte measured by mass flow meter N numberof datapoints employed for calculation of @ ——Wolumerie Now rate at STP converted fom measired sass low Rey Reynolds umber of et © ted /) S-' _ separation distance between exits of parallel plane jet Stage ekitation Strouhal number based on jt ext velocity fd width feed /u) Tage peo of excitation signal ( 1 / fa) 2 Gvalvng time © ondimensional time (= Tac) ‘At phase agin time te time-averaged anil velocities tg mean velocities at je eit (= @/ 210) 4 rootmeansquare vale of local flow velocity uta sions = Ee, x lateral coordinate originated from center on ext plane of dua pralel plane ets ¥—axll coordina originated from center on exit plane of ual parle plane jets transverse coordinate originated from center on exit plane of dual parallel plane jets #——_exito inlet area ratio of ASME nozzle + convolutional time scale of eutoorrelaion ,—_Tageangianinegral ime scale of turbulence ees © power spectrum density funeion 1» centrine concentration reduction index 1 Klnematie viscosity of air not detectable. Spall (21 attributed the instability characteristics tothe equal momentum fx of two parallel plane ets of different slot witb The instability behavior of the jets was characterized by “flapping” of the jet columns. Thu, it was difficult to identify the merging ané com- bined points because ofthe instability ofthe jet columns when these jets mixed strongly with ambient fluids at their matched momentum. Mon- dale al (22) studied the effet of separation cistancesofrwo plane tur- bulent parallel jets onthe periodie vortex shedding phenomenon. There- sults inicated that the periodic vortex shedding phenomenon observed near the nozale plate was behaved similarly to obtained in the near wake region of the two-dimensional bil? body at separation distance range 08 5 S/d < 1.4 However, at $/d = 05 and 1.5, two counterrotating stables seady vortices appeared inthe near flow field region. The veloc- fay signals csplayed pure sinusoidal orilation at the same separation distance range. The fart Fourier transform of velocity signals exhibited 2 frequency peaks at the same value of Strouhal number and gradu ally reduced with the increase in separation distances between two ets “Moreover, turbulence characteristics such as mean velocity profiles, tur- boulent kinetic energy distributions indicated greater interaction of rin jets ithe shear layers along the central path as separation distance de- mrt oma of Mehl cee 99 (2021 108182 ‘creased. Spectral studies revealed two distinct peak frequencies were found with formation of vortex associated with inner and outer shear layers (23), ang et al. (24) investigated the turbulent mixing of parallel jets for the design of cooling systems in nuclear reactors using Laser Doppler measurements, The mean velocities, turbulence intensities, and Reynolds stress were computed at equal and non-equal exit velocities The merging point wes appeared at anaxil distance between y/d'= 1.72 03.45 anda combined point obtained at y/d = 18.52. The higher turbu lence level observed atthe outer edges of two jets and the outer bound! ary of the combined jets. The maximum Reynolds stress appeared inthe ‘merging region indicated the stronger mixing inthe region. Wang tal 25] extended hie work and attributed the flow field in the converging region of two parallel plane jets by using particle image velocimetry measurement, The mean velocities revealed thatthe flow moved in the ‘opposite direction generating a significant amount of recirculation. The velocity profiles obtained from PIV measurements were agreed well to LDA results, The turbulence characteristics depicted that strong momen: tum transfer occurred in the merging region. The instantaneous vector ‘eld was evaluated to identify the presence of vortices, is growth, and ieraction patterns. Wang and Tan (26] experimentally investigated the interaction between dual-et flow by wall jet and a parallel off jet. ‘The dual jet flow was characierized by the Karman-like vortex stret in the inner shear layer regions. These vortices roll up and shedding pe Fiodically revealed interaction between two jes. Besides, these vortices sigalficantly affect the outer shear layers of offset jet and wall jet which resulted in concentrated Karman-like vorticity. Zhao et al. (27) exper mentally studied the interaction between two parallel planar Jes in a ‘rosflow. A prominent recirculation zone was formed which was larger in size atthe windward side and salle in size at the leeward side, The windward side shear layer was more turbulent than the leeward side shear layer. The formation of coherent structures and rollup of large ‘eddies, its growth, and fracture into smaller eddy sizes were occurred thickness) ofthe nozzle plate ‘were 360 mim x 160 mm x 6.4 mm. The nozzle plate contained two ‘denies prallel ASME low.p series of converging nozzles (37]-In some studies, for instance, Raman and Taghvi [38] the term “twin jets” de. notes two supersonic jets issued from two inline-arranged rectangular nozzles (usually having an aspect ratio less than or equal to 5). Under this arrangement, weak and strong interaction between the coherent low structures and the shock waves may produce broadband and dis ‘rete tone (screech). The tone propagates upstream tothe jet exits and couples with the hydrodynamic disturbances. The hydrodynamic di turbances may induce structural fatigue. Ths phenomenon was usually called the “twin jets coupling", which may include two modes: symme ‘uy and anti-symmety. Inthe present study, low-speed jets were issued from two slots with an aspect ratio of 36.25 arranged to be parallel to ‘each atheras shown in Fig. 1, No shock waves were created and hence no Interactions between the coherent flow structures and the shock waves ‘occurred, therefore no coupling types were identified. The exits ofthe ‘oeo parallel nozzles had a length (2) and width () of 145 mm and 4 mm, respectively. The aspect ratio ofthe slots (nozzle ents) was about 36.25, ‘whic was larger than the ratio of 30 proposed by Deo et al. (39) for forming two-dimensional jets. The distance between the centers ofthe ‘hwo slots was $ = 20 mm, so the dimensionless separation ratio (the nozale spacing ratio) was S/d = 5. Transparent Plexiglas plates were in stalled at the two ends ofthe jet-generation chamber vertically from the Tevely = -450 mim to 1000 mm ‘A loudspeaker was installed in the center of ane ofthe sidewalls of, the jet-generation chamber to impose pulsations on the jets. he lou: speaker had a face diameter and a voice-ol electrical resistance of 20, ‘em and 5.5 Q, respectively. The weight of the loudspeaker, including the voice col, dust cap, surround, spider, diaphragm, magnet, and outer frame, was 1.60 kg. A function generator was used to provide the control signals of = 25 V to a power amplifier, which subsequently output a larger cleric current and voltage to the loudspeaker col. Prelim nary tests shoved thatthe flows subjected to the acoustic excitations us ing the triangular, sinusoidal, and square waveforms from the function ‘generator exhibited similar behaviors, but the square waves induced larger jet pulsations than the other two waveforms. This phenomenon ‘coincided withthe observations of previous investigations by MCloskey etl. (40) and Shapiro et al. [17], Square waves were employed inthe present stad to fore the ets. The Frequencies ofthe square waves pro vided by the function generator were varied from 20 to 500 Hz. The ‘maximum attainable excitation voltae at the output terminals ofthe power amplifier was E, = 15 V. ‘he motions of the loudspeaker diaphragm induced convective ds turbances on the flow inthe jet generation chamber andthe propaga ing sound waves. The propagating sound waves exhibited asthe sound pressure exerting on the flow. The sound waves, gravitational waves, ‘and random waves had insignificant viscous losses or dissipations and therefore are not classified a turbulence, but, they are dispersive (42 Both the convective disturbances induced by the loudspeaker diaphragm ‘motions and the sound pressure could affect the flow id, At large How pulsation intensities, the contribution of inducing flow pulsations bythe sound pressure may not be as large as that by the convective distur bbances induced by the loudspeaker diaphragm motion. In the present study, the combined effects ofthe low pulsations induced by the con- vectve disturbances and the sound pressure were lumped into the inde Pendent parameter measured a the jet exits, Le, the pulsation intens!- 2.2. Velocity pulsation measurements To detect the pulsating velocities near the eit ofthe dval parallel plane jets, two homemade one-component hot-wire anemometers were placed at the centers of both nozzle exits at (e/d y/e 2/d) = (22.5, 15, 0). To measure the instantaneous flow velocity variations at the other locations of the jets, a one-component hot-wire anemometer was, used. The hot-wire probe had a platinum wire with a length and a di- ameter of 1.5 mm and 5 ym, respectively. The response frequency of the platinum wire corresponding tothe electronic square-wave test was, adjusted to 20 k#x, The postion of the hot-wire probe was controlled ‘sing a stepping-motor controlled three-dimensional traversing mecha nism. The hoc wire anemometer was calibrated in a wind tunnel using 4 pltot tube associated sith a high-precision electronie pressure tans= Acer, The dynamic behavior ofthe hotwire anemometer output signal was analyzed by feeding these signals toa high-speed PC-based data- acquisition system. The sampling rate and the elapsed time ofthe data- acquisition system were set to 30,000 samples/s and 3s, respectively. 2.3, Flow visualization Before air was supplied to the jet-generation chamber, it was fist passed through a smoke generator io entrain smoke. The Mic seatering technique [3] was applied to the smoke-seeded flow feld to visualize ‘he flow patterns by shinning a laser light sheet across the smoked flow. The smoke was generated from a homemade smoke generator that con- tained a nicke-chzomium ally spiral col and kerosene ol. The metalic coil was heated by @ DC generator and regulated at an optimum output to generate a large amount of oil mist. The diameter ofthe kerosene-cil, mist, which was about 4.7 4 0.2 ym, was determined by the Malvern 2600 C particle analyzer. The density of the kerosene-oll mist was 5.31 keg. The Stokes numberof the smoke partiles was on the order of 10°S, which was significantly smaller than unity, indicating that the smoke particles could follow the flow motion reasonably well (44). The light source use to scatter the traces ofthe smoke particles was a con- tinuous greenlight laser with a pover and a wavelength of 3 W and '532:nm, respectively. By installing a cylindrical opie at the ext ofthe Taser head, a thin planer laser ight-sheet of 0.5 mm in thickness was obtained. Ths laser lghtsheet was aligned vertically across the x = 0 plane for flow visualization Instantaneous flow images were captured with a highspeed dig- tal camera (IDT Motion Pro 4). The camera had a complementary ‘metal-oxide semiconductor (CMOS) active monochrome sensor array of 1024 x 1024 pixels. The images were recorded ata frame rate of 600 frames per second (ps) with an exposure time of 0.6 ms. The frame rate ‘of 600 fps was cen times the pulsation frequency (60 Ha) so hat ten consecutive images were captured within one pulsation cycle. A Nikkon still eamera (Model D200) was used to record the Tong-exposure smoke Flow images. The camera had a CMOS sensor of 6000 x 4000 pixels, The exposure time of the long-exposure images ws 4.0 s, All images were captured using the same iismination intensity ofthe laser Hightsheet, exposure time, and image depth of field. These long-xposure images were used to quantify the jet spread width (W) at various flow coadi- ‘ons to identity the lateral jet expansion characteristics. The binary edge detection method proposed by Shapiro and Stockman (45) and used by Huang and Hsu [46] and Murugan etal. [47] was employed to deter- mine the boundaries of the smoked jets. The jet spread width at an axial level was obtained by estimating the distance between the left and right boundaries ofthe jets. murat! oma of Mehl Sees 99 (021 106182 2.4. Trocer-ga concentration detection “The jet dispersion characterises were examined using the tracer ‘gas concentration detection method. The tracer gos used was carbon 10-8 m/s, re spectively, which were close to the correspondingly values ofthe air (of 28.97, 1.205 kg/m’, 1.801 x 107° kg/m, and 1.981 x 10° m/s "Therefore, the jet dispersion properties were approximately simulated, ‘The probe used to sample the gas was a small-diameter, L-shaped stainlese stel tube, The inner and outer diameters ofthe tube were 2.6 mm end 3 mm, respectively. The tip of the sampling probe was aimed toward the upstream direction so thatthe L-shaped part ofthe sampling ‘probe stayed downstream, The mixture was continuously drawn into the lip of the sampling probe by a suction pump. A dehumidiying system was installed inthe piping system to remove redundant moisture. The debumidified mixture was Finally directed into a nondspersive infrared sas analyzer to measure the time-averaged carbon dioxide concentra tons. The suction flow rate was preliminarily tested before the formal ‘measurements were performed. A suction flow rate of 800 cm /min was selected because it caused almost negligible interference on the fow Pattees upstream of the sample tp within the functioning rage ofthe ‘ga analyzer. Preliminary tests also showed that a sampling time longer than 2 min was sufficient enough to obtain sablizd time-averaged con ‘entration dats. sampling time of about 5 min was employed forthe present study, 2.5. Uncertainty estimation Using the method proposed by Steele ea [8], the uncertainties of the measurements were estimated using the following equation: u,= (apse? en ‘where Us isthe uncertainty, B isthe bas limi ofthe estimation of Fixed ‘values and it should be 9586 ofthe experimente’s confidence, ad Ps the precision limit of the measurement points or lack of repeatability ‘aused by the random ercors. The measurements of present experimental study were repeated ‘more than 5 times eareflly to ensure the accuracy of the measured re. sults. The leakage tests were firstly performed on the piping system and ‘the jet generation chamber. After that, the rotameters used to measure ‘the flov rates of air and CO,/N, gas mixture were separately calibrated via a mass flovimeter. The hot-wire anemometer used to measure fow velocities was calibrated using a Pitot tube with a high-precision elec teoniepressue transducer. The hor-wire anemometer was calibrated ev ‘ery 3.04 hours o prevent from the aging eect of hot-wire probe. The Dinary edge detection technique was used to Identify the Jet contours ‘and to obtain the jt spread widths from the long-exposure smoke flow pictures using the iterative program executed in Matlab, The data were ‘obtained by averaging over 40 to 60 images of 4 sec exposure time. The ‘measurement uncertainties ofthe present stody are listed in Table 1 3 Results and diseussion 3.1. Jee velocity pulsations near the jet exits “The non-dimensional jet velocities u/u, near the jet exits subjected to various jet pulsation intensities J at an excitation Strouhal aumber ‘tg = 0106 are shown in Figs. 2(a<. The corresponding output signals ‘of the function generator and power amplifier (denoted by Fy and F, respectively) are also marked inthe figures for comparison, The function ‘Benerator Output FE; was square waves of =8 V. The power amplifier Rar, RF Mag ond Ca, 8 ‘Table Unceraites of mesmurements ‘Mass aw ate mesure by mass fowineter ow velocities measured by bre anememeter Jet spead with by binary e€ge detection method ©, eonentaton neta nT & & {0 2 2 eae een “ th DANTE 0] a. ~ me 1) — tte ad 7 20 4 ” ages Onnt8” ale 7 to, ‘ Ee he Vd? 730 £, i one Oy hp-16 ite “log fate) output signals E, which were used to drive the loudspeaker to pulse the jets, were distorted bythe inductance of the loudspeaker voice coll “The normalized instantaneous jt exit velocities uu, exhibited periogie waveforms with a phase lag Ac when compared with B; and E., The non-dimensional phase lags A1/P between the jet exit velocities and the ower amplified sigoals F, were about 0.27, 0.21, 0.1, and 0.10 atthe jet pulsation intensities 0.4, 0.8, 1.2, and 1.6, respectively. The non- dimensional phase lags ACT were larger at low et pulsation intensities and smaller at high pulsation intensities, These phare lags were induced, by the butfer efact of the jet generation chamber. ‘The amplitudes of the periodic u/u, wave forms increased as the jet pulsation intensity J increased, For example, the amplitude of u/, signals was about 1.43 at [p = 0.4 (Fg. (8) and about 2.38 at I, = 0.8, (ig. 200). A€ large pulsation intensities, 1p = 1.2 (ig. 2(2)) and 1.6 (Wig. 2€), the lower parts ofthe u/u, wave forms exhibited hump-ike signals above the level of w/t, = 0 atthe end of each pulsation. The aa murano of Mecha Sens 195 (2021) 106182 Primed mania cern OO 5 O2k ofreaing 2 kof reading Fowe of ening Fst ofeading 24 Fig. 2, Jet plations at exe of nals memsured by hotovire ‘memometer. (Gd) instantaneous velocities and output signals 6 fof fnetion generator and power amplifier, (@2) power spec: ‘ram density futons. Pane jet Reynolds number Rey ~ 600, log@ — scrochal number ig, ~ 0.106.Measurement locations: ey, 8 (ails) 2/6) = 225,05,03 ° 24 6 logo 8 (als) ° 4 6 logo 8 (als) ° 24 6 logo 8 (ails) ° time-evolving velocity signals were compared with the corresponding flow visualization video clips, and reverse flows (suction back) were ‘observed near the jer exis in the flow visualization movies within the time periods ofthe hump-lke signals. ‘We Inferzed that these hump-lie wave forms ae induced by the “- rection ambiguity” of the one-component hot-wire anemometers used for measurement. We added dotted piecewise curves to mieror-reflect, the hump-ike signals within those periods in Figs. 2(e and (dl) to present the suction back phenomenon. These negative velocities near the jet ex its were induced by the large jet pulsation intensities. Atlarge pusation intensities, fy~1.2 (Pig. 2(6) and 1.6 (Fig, 26), the amplitudes of u/, ‘wore about 3.97 and 6.04, respectively, which were large than those at Tow jet pulsation intensities without suction back. “The digital fast Fourier transform (DEFT) method [49] was applied to the time-evolving signals, and the power spectrum density func tions © corresponding tothe instantaneous jet ext velocities shown in 420246 afd igs 2@)40) are show in igs. 2(€)-). In all Agures, the power spec- trum density function exhibited many peaks. The first peaks in Figs 2(6)- () correspond to the primary frequency ofthe signals, and the other peaks in the high frequency domain with values smaller than those ofthe first peaks were induced by the harmonics when calculating the power spectrum density functions. The frst peak values in Figs. 2(e}h are all atthe frequency f= 60 Fz, which corresponds tothe excitation fre- quency fig Used i the work. The etext velocities faithfully responded. to the acoustic excitations but with a time lag from the excitation sig nals, 43.2, Instantaneous smoke flow pater Fig 9 shows the instantaneous smoke flow patter in the z= 0 plane ofthe non-pulsed dual parallel plane jet at Rey = 600. The jets emerged from the parallel nozzles 1 and 2, and then traveled downstream. The jet columns remained clearly separated aty/d < 5. Within this xia dis- tance, the two jets delected very slightly toward the centerline, and very litle smake was observed between the jets. The jt surfaces exhibited instability waves induced by the shear effect ofthe ambient ar. At 5 < 1y/d-<10, the instability waves onthe et surfaces rolled up and formed Coherent structures, The coherent structures grew in size with inereases| in the axal distance and interacted with each other s0 thatthe smoke filled the region within 5 < y/d < 10. The two jet columns deflected toward che centerline significantly within 8 < y/d < 10 to form a neck ing region. At y/d ~ 10, the inner coherent structures of the two jet ‘merged and mited together to form a single jet. Aty/d > 10, the outer coherent structures ofthe two jets formed a vortex pair onthe surfaces of the merged single jet and inereased in size. The vortices travelled downstream, constricted inthe center part, and underweat a sma at eral expansion at y/d = 14, At14 < y/d < 17, the traces of the large vortices were stil distinguishable. Aty/d > 17, the coherent structures broke up and disintegrated into turbulent eddies. ig. # hows the instantaneous smoke low pattems in the 2 =O plane ofthe pulsed dual parallel plane jets at Rey = 600 at various pulsation intensities at an instant of = 0.6 from the beginning of 2 pulsation cycle Ata pulsation intensity fp = 0.4, as shown in Fig. (2), an intial ‘mushtoom-shaped flow structure appeared near each jet exit due to the pulsation ofthe jet. The inner parts ofthe votial structures interacted ‘wth each other so that thick smoke filed the space between the jets at mrt oa of Mehl Sees 99 (2021 106182 4 4 H 0 0 xd x/d ig. 4 Insantancous smoke flow pattems of pulsed dal plane Jets at vari ‘ous pulstion intensity 1 at a xed Strouhal number Sta = 0.106. Plane jet ‘Reynolds number Ke, = 60, Framing rate = 600 fps. Exposure time = 06 1. ‘y/d < 4, The jets underwent a constriction at y/d = 5, Two mushroom: ‘Shaped structures that evolved from the previous cycle of pulsation were loctted downstream ofthe constriction ares. Around the center axis be ‘ween the two evolved downstream misshroom-thaped structures, there ‘was a narrow jet eolumn that evolved from the constriction area tothe 16, the coherent seetures broke up and disintegrated Into Fine turbulent ees. ‘Ata pulsation intensity fp = 0.8, shown in Fig. (), the Mow pre ‘sented a pair of intial mushroom shaped structures near the etelts due tothe pulsation, which was similar to that shown in ig, (a) fo ly = 04. ‘The jets constricted at y/d 6, and two large-sized mushroom-shaped structures that evolved from previous cycle of pulsation were located ‘downstream ofthe constriction area. These two downstream mushroon shaped structures appeared within & < y/d < 16. Because of the in ‘crease in J turbulent Auctuations were imposed on the dovnstream mushroomshaped structures so that che ewo downstream mushroom shaped structures were not as coherent a the images but were larger in size compared with those shown in Fig. (a) for Ip = 04. The jets derwent a lateral expansion at about y/d = 9, an axial level lower than. that of lateral expansion for I, = 0.4 shown in Tig. (a). At y/d > 26 the turbulent downstream mushroom-shaped structures broie up and disintegrated into turbulent eddies. 'At @ pulsation intensity fp = 1.2, shown in Fig. 4(@), the inal rmushroom-shaped strutures heat the jet exits became more turbulent than their counterparts at fy = 0.4 and 0.8 shown in Figs. (a) and @), respectively. The jets constricted around y/d = 8. A thin turbulent jt column evolved dovinatream ofthe constriction ares. No clear images of the downstream mushroom-shaped structures evolved fom the constric- tion aea like those observed in igs. (a) and (b). The jets exhibited an abrupt lateral expansion with fine curbulent eddies downstream of the Constriction area. At pulsation intensity ly = 1.6, shown in Vig. 4), the mushroomshaped structures near the et exits were strongly im- posed by the large turbulence due to the stong pulsation intensity, fausing the images of the structures tobe blurred The jets constricted around y/d = 9. A narrow turbulent jet column evolved downstream fom the constriction area, The jets exhibited an abrupt lateral expan- sion with strong turbulence of fine eddies downstream ofthe constric- tion area. The range ofthe lateral expansion was wide. Similar to that observed at , = 1.2 in Fg. 4(2, no traces of downstream mushroom- shaped structures that evolved from the jet consrition area were ob- served. In summary, an intial pair of mushroom-shaped structures formed near the jet exits shorly after the pulsation cycle began. The subsequent evolution behaviors ofthe jets were diferent at low and high pulsation intensities, Palsing the jets at low pulsation intensities of about fy < 1.0 (a Ip = 04 and 08 in Figs. (8) and (), respectively) induced rel. atively’ laminar initial mushroomshaped structure, while pulsing the jets at high pulsation intensities of about Jy > 1.0 (fp = 1-2 and 16 in Figs, 4(€) and (@, respectively) caused relatively turbulent ini= ‘il mushroom shaped structures to form. After formation ofthe intial ‘mushroom-shaped structures, the jets constricted toward the center. At low pulsation intensities, a pair of mushroom-shaped structures evolved into the downstream region from the jet constriction area and under- went a lateral expansion ‘At high pulsation intensities, there was no second par of mushroom shaped structures dovmstream ofthe initial pair of mushroom-shaped structures because as shown in the movies, the initial mushroom-shaped structures evolved downstream and quickly broke up into turbulent ed dies, This phenomenon was eaused by the strong vortex stretching effect [42) induced by the intensive acceleration and deceleration of jet pul- sations causing cascade of vortex stretching in three-dimensions, and this caused the large-scale coherent structures to break up quickly into fine turbulent eddies and decrease in axial momentum within a short axial distance. Thus, the jets spread abruptly and widely inthe lateral Aiection, 3.3. Long-xposure smoke flow patterns and et spread width Fig. 5 shows the long-exposure smoke flow image inthe x= plane ofthe non-pulsed dual parallel plane jets at Reg = 600. Near the jt exits, the two Jets that ised from the nozzles 1 and 2 were delineated by two bright, curved-inward narrow bands. Smoke Milled the region between the jets, The lateral expansion ofthe smoke followed the necking area at y/d 8 The outer boundaries of the smoke near the jet exits were relatively sharper than those downstream of the necking area because the jet flow ater lateral expansion became turbulent Fig. 6 shows the long-exposure smoke flow patterns in the =O plane ofthe pulsed dua parallel plane jets at Rey = 600 at various pulsation intensities. all images shown in Fig. 6 exhibit a shore upright columa of smoke near the jet exits and followed by an abrupt lateral expansion The larger the et pulsation was, the larger the laeral expansion that was seen. The lateral expansions ofthe pulsed dual parallel plane jets shown mural mal of Mech Scenes 99 (021 108182 é x/d Fig. 5. Long-xposure smoke fw pater of nampa plane ets shwtng timeaveraged flow behavorand qualitative e spread width, Plane je Reynos In Pig. 6 are significantly more obvious than that shown in Fig. 5 forthe ‘non-pulsed jets. "The 4s exporure images shown in Figs, Sand 6 were processed to devermine the lateral jt spread width W using the binary edge dere tion method of Shapizo and Stockman [45]. This method allocates the left and right boundaries of the long-xposure smoke images ofthe jets bby searching forthe local maximum grey-level gradients across an axial level. The local maximum grey-level gradients would be found around the jet boundaries. The distance between the locations ofthe left and right maximum grey-level gradients aross the target axial level i taken as the Jet spread width across that axial level. The caleulated results of the lateral et spread width using che 4 s expesure smoke Images were scattered because the 4s exposir time was not long enogh [45,47] ‘We therefore performed a preliminary test of averaging over multiple 4 5 exposure images. The results ofthe preliminary test indicated that averaging over more than 15 smoke images of 4 s exposure could givea ‘converged steady value of jet widths, Pgs. ? shoves the non-dimensional {et spread width Wad of che daca averaged over 20 images of 4s expo: ig. 762) shows the nos-dimensional jet spread width W/d of the non-pulsed an pulse jets with varying of the non-dimensional axial distance y/d. The non-pulsed jets presented the smallest values among the data, Aty/d < 8, the W/d value of the non-pulse Jets decreased slightly with inereasing yd with an approximate average valle of about 6, which accords with the observation shown in Fg, 5. Aty/d> 8, W/dl {increased gradually with increasing y/d, with an increase rave of about 5496, When the jets were pulsed, the initial jet columns at al jt psa tion intensities exbibited a larger spread width than that with no pul sation. At low jet pulsation intensities, ly = 0.4 and 0.8, the values of W/d decreased slighty or almost remained constant with increasing y/d However, at high jer pulsation intensities, Jp = 1.2 end 1.6, the vslues ‘of W/d increased slighty with increasing y/a. The Jet spread widths of the pulse jets all exhibited an abrupt increase as y/d increased beyond ‘short distance from the jet exis. The distance from the jet exits where the jets expanded abruptly became shorter asthe jet pulsation intensity Jncreased. The W/ad values of the pulsed jets were markedly larger than those of the non-pulsed jess inthe jet expansion region. For example the value of W/d of the non-pulsed jets at y/d = 32 was only about 22, while aty/d = 32, the values of Wi at J, = 04 and 1.6 were 48 and 67, respectively. The increase rate of the jet spread width of the pulsed jets ‘was significantly larger than that ofthe non-puleed jets, For example, the increase rates ofthe pulsed jets aty/d > 20 were almost all between the values of 110% and 125%, which were significantly higher than the ‘value of 4586 for the non-pulsed jets. (@) p= 0.4 (b) p= 0.8 murat! oma of Mehl Sees 99 (021 106182 eb ae abba oe x/d 0 2 @ % 60S 9 Lanne ON pute - 13) putsea : E oh | 46k wow N oo 1 J [ avi = dg decreasing ——— oly — [ . fF 030s us a2 se yd Ip Lr — 8 of len ty ot ema eg ef ll oo tin Lyd ae yd decree ds erening Deseaing ras In snwiat (4 Sudden Sein noc hha rhs 90 ee ae 3 expansion Der Sige = 0.106 colitan Ye 4 ERE 3 © oo | d ow The values of wtih eel column ag oe use wok. f—jetaidnorvn stay wi changr thet pulses cst. rn ‘ee mouty seta eno the da jr eet ot ¥%g.7. Non-dimenslonsl jet spread width W/dobalned by applying binary edge detection method to meaveraged stoke ow patterns (2 vacations of W/d With normalized aul distance y/, (D) variations of Wd wil jet pulsation lates ey = 600, Se = 0-108, Fig. 7(0) shows the effect of pulsation intensity on the non- dimensional jet width of the non-pulsed and pulsed jets. The non- dimensional Isteral jet spread wicths of the non-pulsed jets = 0) at the axial Tevels corresponding to those of the pulsed jets were relatively the inital je column ofthe dua jets is reiced because of the decrease ‘ofthe axial length ofthe recirculation zone [36 ig. 6 presents results for the non-dimensional length Ly/d of the intial jt column. The value of Ly/d of the non-pulsed jets was large, about 8.05. Ar the smal jet pulsation intensity off, = 0.4, the value of 1by/d decreased a litle compared with that ofthe non-pulsed jets. As the {et pulsation intensity increased, the value of ly/d decreased quickly. ‘Compared with the Ly/d of the non-pulsed jets, the non-dimensional Inia jet column lengths at fy = 0.4, 08, 1.2, and 1.6 decreased by about 3.1, 9:3, 21,7, and 51.6%, respectively. Rar, Magan CM. Hs murano of Meche Secs 195 (2021) 105182 1 2h Pig 9. Jet pete characteitic nase i he eye, Ne @h=04 ©@hn04 surement lestions (ee) noted in gre. a) atetancos wl conz ‘clot, (et) power sper deny neon. Plane jo Reyes O8- (x/d, vid) Pa ue ‘number Re, ~ 600, trutal number St. 0.108. uw |, lone uy 94 (a's) 0 | 04 .8 2.0) 24 T= 08 Oe (80 He 16 i log? u log We 8 aris) 0 ° “10 8 sg oa 4 Op 12 7 sour | (xia, yid)= (4.1, 5.4) Pa dis 20 : u loge & AAA vain ° 4 ° 20 -8 601 @ip= 1.6 O=16" Go |°* 4.0 (x/d, pid) = (4.1, 6.04 “ 16 uw log? th 29 8 aris ° o 2.0 o 1 2 3 40 2 4 6 log f(Hz) BA, Velocity characteristics Figs. {@)(d) show the non-dimensional instantaneous velocities tu. measured in the shear layers of the initial jet eolumas of the pilse jets, All of the velocity histories exhib ‘ause the mushroom-shaped structures swept over the measurement lo- cation. Fach periodic eyele of the velocity signals occupied an acoustic pulsation period, meaning that one pulsation quickly caused the et low to exit the nozzle, form a mushroom-shaped structure near the Jet ext, and then travel downstream, The amplitudes of the periodic velocities increased with increases inthe jet pulsation intensity. Therefore, the velocities ofthe initial mushroom-shaped structure near the jt exit re- sion increased a J, increased. The imposed turbulent fluctuations also increased in magnitude as the et pulsation intensity increased, confirm- ing the flow visualization images shown in Fig. 4 "The power spectrum density functions @ in the frequency domain ‘shown in Figs 9{e)(3) correspond othe instantaneous velocities shown ln Figs. 9(@)@), respectively. All ofthe figures forthe power specrumm density functions exhibited a primary peak at 60 Hz, which was the excitation frequency (corresponding 0 an exclation Steouhal number of 0.106). The value of the primary peak of the power spectrum den- sity function @ increased a the jet pulsation intensity fy increased be- ‘cause the amplitude ofthe periodic velocity signals increased with in- ‘ereases inf. The ober peaks that appeared at frequencies higher than the frequency corresponding tothe primary peak and had smaller mag. nitudes of © were subharmonics induced curing the calculation process sing the DFFT method. Hg. 9 shows that pulsating the jts induced the ‘mushroom-shaped structures (a shown in Fig. 4), which dominated the ‘dynamic behavior ofthe pulsed jets and therefore the time-averaged appearances of te jets. The spectrum analysis reveals that the inst Diy in the shear layer is due tothe faster and stronger development ‘of the fundamental vortex appearing in the outer and inner shear lay ers as examined by Lai and Nase [36]. The instability of shear layer ‘becomes stronger asthe jet pulsation intensity increases. The high fe ‘quency peaks in the spectra were also observed in the shear layers of the excited plan jet by Thomas (50 and in the excited circular jet by Zaman and Hussain (51) "ig. 10 shows the lateral distributions of the normalized mean axial velocities tyq/p at various aia levels. Aty/d=6 (Fig. 10(@), the nor- ‘malized mean axial velocity distributions exhibited self-similar profes for the non-pulsed and pulsed jets. The lateral distributions of tye/ty ‘exhibited dual peaks at x/d ~ 42.5 in the paths ofthe jets and had deep valleys around the centerline x/d = 0, The valleys indicated that the ax fal level y/d = 6 was within the merging region, which is between the ‘merging and combined points [9,10,17,36,52-55]. The uyy/ty values ‘of the non-pulsed jets were lower than those of the pulsed jets. With Rr RF Mag and Ca, 9 muerte of Mecha Sens 195 (2021) 105182 20 ™ T (wid 12 T yao 1s Fig 10, tater! distributions of non-dimensional mesn velocities Shoes of ron-palsed and pulsed deal plane jets a various 3. Jal distances 9/4 Plane jet Reynolds number Rey ~ 600, Strohl | numbers, 008. pulsations the values of taye/ty increased with inreases inthe jet pul- ‘sation intensity, The peak values of ug/g ofthe pulsed jets ap = 1.6 reached about 1.5, which is about 25 times of that (= 0.6) ofthe non- pulsed jets, At y/d = 12 (Pig. 1000), the tye/uy values were smaller than those at y/d = 6 (Fig. 10(@)) because the axial momentum ofthe Jets decreased with increases in the axial distance. The differences be- ‘ween the values ofthe peaks and valleys aty/d = 12 were not as large those aty/d = 6 because the profiles became a litle fate. The shal- low valleys implied that the axial level y/d = 12 was still within the merging region but was closer to approaching the combined point than those in Fig. 10(). [Aty/d= 18 and 24, a8 shown in Figs. 10(¢) and (), respectively, the ‘dual-peak patterns disappeared, The non-pled ets had a Gaussian ike profile resembling tha ofa single jet, indicating thatthe two jets com- bined and conserved momentum (9,10,17,36,52-85]- The axal momen tum ofthe combined non-pulsed jets was concentrated around the enter core, The pulsed jets exhibited relatively flattened lateral velocity die- toibutions compared with those at y/d = 6 and 12 and with those ofthe non-pulsed jes. This phenomenon implies that the axial momentums ofthe pulsed jets spread laterally faster than those of the non-pulsed jets It also implies thatthe axial levels ofthe combined points of the Pld jets were located more upstream than those ofthe non-pulsed jets. In general, the larger axial-momentum loss due to the higher jet pulsation intensity ultimately results in enhancing the mixing process by allowing jets to expand in lateral directions, Similarly, the effet of jet pulsation intensities on the axial velocity is observed in the swicing Aouble-concentric jets by Jular et al. (21), ig. 11 shows the lateral distributions ofthe axial turbulence inten- sites wu, at various axial levels. Aty/d = 6 (Fig. 12(@), the axa tur. boulence intensities exhibited dua-peak profiles for both the non-pulsed and pulsed jets. The peak values of the non-pulse jets were about 12% ‘The axial turbulence intensities of the pulsed jets were significantly larger than those of the non-pulsed jets. The larger the jet pulsation Intensity was the higher the axel turbulence intensity. The peak values ‘of u/s ofthe pulsed jets at = 1.6 were markedly high—about 110%, ‘whieh was about 9.2 times that (12%) of the non-pulsed jes ‘he dual-peak patterns beeame inereasingly less appazent a the ax {al distance increased, The axial turbulence intensities of the pulsed jets decreased quickly 25 the axial distance inereased. However, fr in the ‘non-pulsed jets, the axial turbulence intensities decrease slowly a5 the axial distance increased. For example, at y/d = 24, the values of w/t, ‘decreased to about 25% at fy = 1.6. With no pulsation, the values of Lg were lower than 10% because the axial momentum of the pulsed {ets spread laterally significantly faster than forthe non pulsed jets. The ‘turbulence intensities remain hgh in the outer shear layer in the pulsed ‘ease are associated with the local development of mushroom-shaped structures, which has been attributed by Thomas (50) for the exited planar je. 35. Time and length scales of turbulence inthe centerline Fig. 12 shouts the typical autocorrelation coefficients Ry (42) of ‘the aval velocity fluctuations in the centerlnes of the non-pulsed and pulsed jets measured at (x/d, y/a) = (0, 6). The autocorrelation coef cient rfiects he extent zat the past history is related to a given event. ‘The symbol r denotes the shift ine in performing convotution calcu Lions for Ry The autocorrelation coefficient of the axial velocity ue twations ofthe non-pulsed jets shown in Fig. 12(a) decreased fom unity at r= 0 tozeroat ¢ = 65.7 ms, followed by small uctuations at ¢ > 65.7 sms. The sift time + = 65.7 ms corresponding to the fst zero crossing ‘ofthe autocorrelation coefficient denotes the statsticl turbulence time scale. The statistical turbulence time scale is usually presented using ‘the Lagrangian integral time scale +, 42], which isan artificial quan tity sed to roughly estimate the statcical time interval of the macro scale turbulent eddies, The value ofthe Lagrangian integral time scale In Fig, 12) forthe non-pulsed jets measured at (e/a y/d) = (0, 6) was 5, = 32.8 ms Rar, Magan CM. Hs mural Jur of Mecha Sees 195 (2021) 105182 150 T (oid Fig 11, Lateral dstebatons of turbulence intense w/a st ae fous normalized axial distances y/d, Plane jet Reynolds number eq ~ 600, Sohal number Sg ~ 0.10. i (%) 20 ‘The avtocorrelation coeficient ofthe axial velocity uctuations of the pulsed jets shown in Fig. 12(b) decreased sharply fom unity at to null @t r= 3.96 ms, followed by periodic wave forms at ¢ > 3.96 rms. The Lagrangian integral time scale in Fig. 1200) forthe pulsed jets measured at (x/d y/a) = (0, 6) was r= 198 ms, which was markedly ‘smaller than that othe corresponding non-pulsed jets. The jet pulsations induced the fst breakup ofthe coherent structures into fine turbulent ‘eddies because of the vortex-stretching effect, so the time scale ofthe pulsed jets was significantly smaller thax that ofthe non-pulsed ets. The Periodic wave forms tha appeared inthe tal parts ofthe autocorrelation ‘coefficient were induced by the et pulsations. The period ofthe periodic wave forms was 16.7 ms, which with inversion equaled jet pulsation frequency of 60 Ha. “The normalized Lagrangian integral time Scales =/Tixe with Vary lng jet pulsation incensities fate shown in Fig. 13(@. The normalized Lagrangian integral time cals Fj /Tox of the axial velocity Mutuations measured at (x/d, y/€) = (0, 6) remained at small and almost unchanged values of about 0.13 when the jet pulsation intensity was increased. For the aval levels downstream of y/d = 6, the values of ¢:/Taye became larger than that at (xd, y/a) = (0, 6). The higher the axial level was, the larger the normalized Lagrangian integral timescale. Moreover, at the same axial level, ¢,/Toe decreased quickly as che jet pulsation inten- sity increesed. At he high pulsation intensity of, = 1.6, the normalized [Lagrangian integral time scales at al axial levels decreased to small val- ues of almost the same magnitude. The jet pulsations hada strong effec ‘on reducing the time scales of the turbulent eds. The larger the jt pulsation intensity was, the smaller the Lagrangian integral time scale of turbulent eddy, Because the inital mushroom.shaped structures broke ‘up into fine turbulent eddies quickly after the constrition are, the time scales of the turbulent eddies decreased asthe jt pulsation intensity in- creased ‘The Lagrangian integral length scales L, of the macro-scale rurbu- lence eddies (56) can be estimated by using the Lagrangian time scales rand Taylors frozen flow field hypothesis (51). Taylor’ frozen flow fold hypothesis assumes thatthe time derivatives of flow properties can ‘be obtained approximately by multiplying the local time-averaged flow velocity by the spatial derivative ofthe properties ithe turbulent eddies passing the observation point are fast enough thatthe properties do not hange substantially ding the passing time. Figure 136) shows the ‘normalized Lagrangian integral length scales I./d eaeutated using #, and Taylors frozen flow field hypothesis. The value of ./d ata fixed axial level decreased asthe jt pulsation intensity increased because the effect of vortex stretching increased as the jet pulsation intensity in creased. Ata fixed jet pulsation intensity, the value ofl increased 1s the axial level increased because the vortex stretching was more ef fective in the near feld so the turbulent eddies were smaller because ‘of the cascade of turbulence kinetic energy. These results coincide with ‘the flow visualizations shown in Fig, 4 3.6. Dispersions of jt fluids ig. 14 shows the lateral distributions of the carbon dioxide con ‘centrations measured by the tracer gas detection method. At y/d = 6 (ig. (a), dual-peak distribution profiles were observed forthe non pulsed jets A peak value of about 4% appeared in the path of each indi vidual jt. The eoncenteaton profiles ofthe pulsed jess exhibited humps around the central area because the jets merged earlier (Cig. 4) and the Initial jet column became shorter (Fig. 6) compared with those of the ‘non-pulsed jets (Pigs. 8 and 5). When the acoustic excitations were ap plied tothe jets, the carbon dioxide concentrations became lower than those ofthe non-pulsd jets and decreased as the jet pulsation intensity ‘increased. At low pulsation intensities, the carbon dioxide concentra tons were appreciably smaller than those of the non-pulsd jets. For ex ample, the peak values of the carbon dioxide concentrations were about 2.5% at J, = 0.4. At high pulsation intensities, the carbon dioxide con ‘centrations were markedly lower than those ofthe non-pulsed ets. The peak carbon dioxide concentration values decreased to about 0.6% at 1, = 1.6. Aty/d= 12 (Pig. 14(D), the earbon-dioxide concentrations of Rar, Marg ond CM. 10 eee 20 ye sea) Fig. 12. Typical autocorrelation coefcens of axle veloc Mustustons at (x 9/2) =(0, 8) (a) noa-pulsed es, () pulsed Jets at y= 1.2, Se = 0.1087. Re, = 600, ‘he non-pulsed jets exhibited a large central peak of about 4.5% because the axial fevel was inthe merging region between the merging and com bined points (8,10,17,36,52-55]. The carbon dioxide concentrations of the pulsed jets were smaller than those at y/d» 6 murano of Mecha Sens 195 (2021) 106182 ; a va = a 7 we 52 a Tee. 1 oL 8 o ’ 4 he ; ole Ip Fig. 15. ec of pulsation ftensity J on sae af turbulent edie in cet ans at several ail sages y/d (3) Lagrangian itepral tie sale, /Tee () Lagrangian feral length Seale /a Pane jet Reypols number ey 600, ‘Soubal umber Sige = 0.106 ‘Aty/d'= 18 and 24, as shown in Figs. 14() and (4), respectively, the ‘concentration profiles of the non-pulsed jets exhibited large values for ‘the humps around che center part because the aval levels were within ‘the combined region [9,10,17,36,52-55]. The values of the pulsed jess ‘were significantly smaller than those ofthe non-pulsed jes. Aty/d'= 24, the carbon dioxide concentration profiles exhibited humps in the center 50 T (id= 12 Fig, 14. Lateral distributions of carbor-dionde concentrasons C af onepled snd pled jar at vere aleve 9/d ne et a all -] ey uty = 66, Sha mbes Sne = 0308 30 4b 4 c » dL 4 us 4 3 so —1 o [02 ertat of | SZ) _ pulsed (O84 30 4 &) (%) ok J Lo 4 100 1s so vid . 6. 2s Sh Ss =u o ots 08 12 16 20 Ip Fig. 15, Bee of pusntion incest J, on centerin concentration reduction index Pane jet Reynolds amber Rey ~ 600, Stouhel namber Stax ~ 0.206. cone regions. At fp = 1.6, the values of the central hump ofthe pulsed Jets decreased to ‘only about 0.5%, which was deastieally lower than {hat (4.684) ofthe non-pulsed jets. The lateral dispersions ofthe jet Mu ids ofthe pulsed jets were clearly significantly better than those ofthe non-pulsed ets. Aydemir eta (57) discussed Garib et als (58) results for the strongly forced round jet. They concluded that increase in the pulsation intensity could increase the circulation of vortex rings, and therefore results in an increase inthe entrainment and dispersion rate The results of Aydemir etal. (57) agree with the results of Huang eta. (32) and the present study The mixing capabilites of the non-pulsed and pulsed ets ean be com- pared quantitatively using a centerline concentration reduction index defined in Fo (1) as follows (Cxseret ~ Grin ss 1 “ (Conmpaise) jane Where Cups 84 Cyuey denote the carbon dioxide concentration at the centerline x/d = 0'f the non-pulsed and puleed jets, respectively. igure 15 shows the effect ofthe jet pulsation intensity onthe center- line concentration reduction index 7 a several axial level. At all axial levels, the centerline concentration reduction index m increased as the jet pulsation intensity increased. [At the lw et pulsation intensity of fy = 0.4, the values of m were in the range of 25.37% at y/d = 6-24, ALthe high jt pulsation intensity Of fy = 1.6, the values of y were increased to the ange of 74.82% at ‘yd 6-24, Massive reduction in the centerline carbon dioxide concen- trations were obtained when the jets were excited, particularly at the high jet pulsation intensities. In other words, comparing with the non- pulsed jets, the mixing capability was significantly improved a high jet pulsation intensities. The mixing index results are in good agreement With that reported by Jufar et al. [31] 4. Conclusions ‘The effects of pulsation intensity on the flow and dispersion charac- teristes of pulsed dual parallel plane jets were studied experimentally The following conclusions were drawn based onthe results of study 1 With mo pulsation, instability waves induced by the shear effect of the ambient air appeared on the surfaces of the dual parallel plane jets near the jet exits. The instability waves onthe jet sucfaces evolved into rollups and formed colerent structures as the axial dis- tance increased, The coherent structures grew in size os the axial stance increased, interacted with eachother, deflected toward the centerline, merged into a single jet because of the merging ofthe in- ner vortices, and then constricted in the center part and underwent ‘smal lateral expansion beeause of the breakup of the coherent structures mre al of Mecha cee 99 (2021 106182 2 When the jets were pulsed, 2 par of intial mushroom shaped struc tures was formed near the jet exits shortly after the pulsation cycle began. Relatively laminar and turbulent inital mushroom-shaped structures were Induced at low and high jet pulsation intensities, respectively, At low pulsation intensities, the jets constricted to ward the center at the top of the inital mushcoom-shaped struc tures, evolved into a downstream pair of mushroom-shaped struc tures, and then broke up into turbulent eddies with a notably wide lateral expansion because of the loss of axial momentum, At high pulsation intensities, because of the strong vortex stretching effect induced by the intensive acceleration and deceleration of et pulse tions, the jets broke up into fine turbulent eddies and underwent a wide and abrupt lateral expansion downstream of the initial rurb- Jent mushroom shaped structures. ‘The lateral spread widths of the pulsed jets were drastically larger than those of the non-pulsed jets. The larger the pulsation intensity was, the wider the jets expanded laterally downstream ofthe in {isl musheoomeshaped strictures. The axial Tength of the inal jet columa located upsizeam the lateral expansion decreased asthe jet pulsation intensity increased. 4 The larger the pulsation intensity was, the higher the turbulence in tensity. The time and length scales ofthe turbulent eddies decreased as the jet pulsation intensity inereased beeause the mushroom: shaped structures generated by the jet pulsations broke up into smaller turbulent eddies at high jet pulsation intensities because of the strong vortex-stretching effet. ‘Tae dispersion capability ofthe jt fuids was significantly enhanced by increasing the jet pulsation intensity. The marked improvement {nthe dispersion of the et fluids was Induced by the character fstes of the flow field (eg, the formation and breakup of the ‘mushroom-shaped structures, te enbanced turbulence intensities, and the abropt expansion ofthe jets) caused by the pulsations ap pliod to the jets. 5. Future perspectives “The effect of spacing between the dual parallel plane jets should play ‘an important role on the flow and dispersion. Understanding the spacing ‘lfect when the jets subject tothe excitation may be considered inthe future. The particle image velocimetry (PIV) can be employed forthe in-depth physics ofthe coherent structure development, instability, and vorticity, which play significant roles inthe dynamics of entrainment and mixing processes. The effec of erossfiow onthe excited dual parallel plane jets an also be investigate inthe future Declaration of Competing Interest The authors declare that they have no known competing financial Interests oF personal relationships that could have appeared to influence the work reported in this paper. (CRediT authorship contribution statement Sanjay Kumar: Data curation, Weiting- original drat. Rong Fung ‘Huang: Conceptualization, Methodology, Supervision, Writing -review ‘editing, Ching Min Hsu: Visualization, Investigation, Writing review eciting References cain ego fe eri of a dines ed (linseed bee meena ma ei a 3} Miller DR, Comings EW. Force-momentum feids in a dul-jet low. Iovorrzren hye inc toy ssa sas 1 tr Gon fen, els A TPB Mera teisesenn Rar, Magan CM. Hs 1 ® v1 ® 00) on al 3) oa) ost be 7 oo) st eo eu (221 a) wa 2s) (2s) a ps1 2s ¥, Nga Sb, We FD glne ae ean es on. town Bataene Py fev He SORSBUEO i100 {Ei And Nhe YA. Compu sly of tbls ow Son ntwace erin tangas Janets Msn S16 920097, pe /otog10 lot Sstnescoter 20171208 UL Aeand Attu Ya, Nowy Ts Large ey maton of het Sit ows iin pra jo tJ Het Nae" 200191200. pe /tog0 Toto yesmasrener 29181008 aA a 3 Ma venga ping Rens Hye lon re pleadon raga, Feaka"E Te.lenineae of trodes pall jr at Re pore experimen nal je "buleun ot Se 1a1072-90 Kip Aano tzoy aoa ara, “aka ET Inter of orion pl jt 2a Rep, periments the comin fw Gn Rol IME Toa Spey ier patos 90. Na Ata XS, Comprion of fw curacy in he oa flo {wo pr pln fran nf plane je Ppe Fh 291% Iie ki ees ss Tome Song ag Mich Damar veigtin of ow ace a ir ‘son gato toed pe tie wpe i 1900970. fepe/adbhrarc/nk tee 967A 102910875010 1087 {aman Davin MR Crow hrc oiling losin hd twa mol deg 90 Sao eng OS on, {ane Ni bacon BR. Stained lien sboege se tara’ rcungur ewig Thos and "Saar Sn SSR Ripsvntng toga 27 ‘Span cing tei ty on fy ibis a Tashio A, Kine Kane Sunes sy on i Sep of ecling ques jh ow eye ner bse ‘Se od bet a cpm mii’ Rucor fp Dag 027-95 Sie esas seta [etait ud adit Soe ona ‘Siete ta Spl Re Exped sed none liven of. ‘Scns pu jo Ps ng easton: Ost85 or ‘ilcnoe a yder0, Gree epee of pre Shel ro aes ewe ig Adc tie yf ye ro ha Ap ISS hal nies olen Kale a a 8 208 Moni, ar ae Pavone ding prec wi op Stn cee Stns to pan tent paral jit Ts Dieposneae hig tari Oy Seeinacses 20960885 eg, New Ti the wakeie eclsrangenent sn bb a Siew tn jet ce proxy fap Therm Fd S289 12720 inp dotong10 1016 expen 2015 0808, Viane tat 5 Then YA. Rages AE LavesDopple mesa Dens ot thal of. etal mee np Ifo a satay "pote Heat” a 20969020027 Ip And org10 106 picts 018.0088 vag He Haan YAP age ciety merc fhe ow Tne comer repon of pra, Nar Eg Design 216905897 tps tadacorg i ohrmenghes 98.09, Vian Xs TSK Experiment Twente terion ewes pho wall Joana panel eS" a Sasa haps ritta br posekabraass> he KO easy Rome 2 PV erent on he inca {ere ween mo ure fae row A Aan 27790008 Repevrorgno ae sot Ser ch Dnieper ier fe al ine [rw vraton tne del cone Se Om tres 2010008 ieee 278 ND ‘Gon yun 8, Toe Mow Chun oe. ne i te Spec ange Vannes" Tce) ~ Une et, Tne Ste ata Sc of Maca tains sone) 090 Fpl sto nao 37 murano of Mecha Sees 195 (2021) 106182 (20) Kee KD, ty SY. Experiment evetiain of To Parle Pe Jet eae Sito ofthe Koran Soca of eshte Eagles 20082003212 ‘pe ncong/10.5735 Rod 8.200529 2214 (ot Jafar Si Hoang 27, tha CAL ee of Puen teemty on Faw {he Ming of Siting DauhleConsotic Je AIAN 3 2013z1@}I922-4, ‘py ens io 2st 252, (22) Iieg RE, Jar SK ew CM Tow and mining charac of swing Aoabeconcatie je abject seme exsatan Esp Hide 201554150 ‘tp ano 100? ss4s0123 4212 (55) amar OA, Hoang RE Hb CML Efe of seat ean oe Omer of ‘hing dash ou concent es ep Therm Fd St 701910578, Nip ator 1016 jesptberase 201608018, (se) 1st tus fect acute extn a reste Sou names raters an leat trys ot xp Ther Fad e201 951370-8, Into 10.2086 exper 20110520, (35) Hung, Kia AK au CM. Elec of jt plo intensity on # wake Stile snpremizel ene a cosow xp Tou Thi Se! 201578 153— erin! dude 10 sole fear: deem, (36) Mr 14S Twe pa! plane et ta ow an ls aes eacaton, tap Ps 1997 22:28-40 s/c 910 087 DSHS 19ND [57] Bowler undnenalvofTespetance, Pes, ow Mesut New “ots abn Wey 8 Sos 1968, (38) aman, Tg Re Caupng of lo rectangular Speen J Fd ech Toe... hp on/ 1017/0022 97a 99) Beo RG, M3} Nahant ence of noma spect ato on plane Jt ep ‘Them ult 4007-78-98 ao Votes oxpnefs I 8 (Wo) Weloney RE Ring IM Gates! 1 Ranglin AR The a ‘vey cong “Jet inertia. J “Had ech” 200245205. ‘pes ong/10. 1047/0024 a0 0065. [at Shap ing, Psy RM Kragosan A Opsaton of etal Jin croton: AMA 20064(61392-8 pe ooo 1oaeT4/1 ur aa] Teper aly JAP Course in Trelence. Cambridge, MAH Pre [a] Mei Roloc Sdeiy of flow ace pais, tap. Hae 1996222, ‘pene 10.107 8300. We) Hagen Re. Sela. Pandata of Air Pllton Hagler. Englewood lf New Jou: Prete Hal 988 Ws) Shon, Seeman GC Camper Vien. Upper Sia iver, Now Jee: Pre Seine 201 [oan Ho CM ow ad aig charac ofa etd pig rae ‘ee et Pa ade 201224 0151041-21 hig dono WO rts a Maegan, fang, le Be faa a plone a ng ics of al cn os tow et ot ce td [aa] Ste We, ayn D, Buel RE, Clean IM, Us previo experince ‘st precisbe weeny ofall spe epeinen IAA J 1951-18818, Intpe lancer 102s14/308" 149] Bigham 0. The Fat Furr Teor Aplin. Enlenond i, New 150) Thomas 60. An experimental investigation ito the role of simultaneous amplitde ‘nd pase modsation i the wnston a lane ja Ps Hs 109025893, Rupe nog 101085/1 657758 [51] Zaman RBM, Hsin A. Tyor ype and ages coherent sts {Fd ne 9811120279, 57996. esas faved heey (s2] Mora Ky Tapa MG Agave An experi study encom uwace oon je." Bal ASME "19753998066, igs eo onsn0 oar 7a. 6368415, 155) owes Don plane pce alent js PD. hess) The Ulesy of Wot hos ar7 [54] tin Yr, Show Me Inveation of opine paral vende Ep Fe ose 722 sao T0007 Re0187867, (ss) in YE, Shew MU. Incracion of paral urea pine Jes AIAA J wor. a919.19723. hes /ansoy 102818/3107°9. (36) op. Tres! ows Bg UK: Cambridge niesy pes, 200, (57) Aydor Wor NA, Dawson JR The fomaton of veneers eet stongy free! round jeu fap. Ms 0125209).729-42, pe dour 1007 n008-01 10, (38) Gaskin Shan XA niet in cle fr vrei rman J oid ech Tops, poor io 1017/ Suen 057 00ee

You might also like