Closed loop pulsating heat pipes Part A parametric experimental investigations Piyanun Charoensawan  Sameer Khandekar b  Manfred Groll Pradit Terdtoon Department of Mechanical Engineering Chiang Mai

Closed loop pulsating heat pipes Part A parametric experimental investigations Piyanun Charoensawan Sameer Khandekar b Manfred Groll Pradit Terdtoon Department of Mechanical Engineering Chiang Mai - Description

hydrodynamic coupling governing the thermal performance 0n this paper a wide range of pulsating heat pipes is experimentally studied thereby providing vital information on the parameter dependency of their thermal performance The in64258uence charac ID: 28983 Download Pdf

158K - views

Closed loop pulsating heat pipes Part A parametric experimental investigations Piyanun Charoensawan Sameer Khandekar b Manfred Groll Pradit Terdtoon Department of Mechanical Engineering Chiang Mai

hydrodynamic coupling governing the thermal performance 0n this paper a wide range of pulsating heat pipes is experimentally studied thereby providing vital information on the parameter dependency of their thermal performance The in64258uence charac

Similar presentations


Download Pdf

Closed loop pulsating heat pipes Part A parametric experimental investigations Piyanun Charoensawan Sameer Khandekar b Manfred Groll Pradit Terdtoon Department of Mechanical Engineering Chiang Mai




Download Pdf - The PPT/PDF document "Closed loop pulsating heat pipes Part A ..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.



Presentation on theme: "Closed loop pulsating heat pipes Part A parametric experimental investigations Piyanun Charoensawan Sameer Khandekar b Manfred Groll Pradit Terdtoon Department of Mechanical Engineering Chiang Mai"— Presentation transcript:


Page 1
Closed loop pulsating heat pipes Part A: parametric experimental investigations Piyanun Charoensawan , Sameer Khandekar b, , Manfred Groll Pradit Terdtoon Department of Mechanical Engineering, Chiang Mai University, 50200 Chiang Mai, Thailand Institut f r Kernenergetikund Energiesysteme (IKE), Universit t Stuttgart, Pfaffenwaldring 3(, )0569 Stuttgart, ,ermany Received $ April %003( accepted ) May %003 Abstract Closed loop pulsating heat pipes *C+P,Ps- are complex heat transfer devices having a strong thermo. hydrodynamic coupling governing the thermal performance/ 0n

this paper, a wide range of pulsating heat pipes is experimentally studied thereby providing vital information on the parameter dependency of their thermal performance/ The influence characterization has been done for the variation of internal diameter, number ofturns, working fluid and inclination angle *from vertical bottom heat mode to horizontal ori. entation mode- ofthe device/ C+P,Ps are made ofcopper tubes ofinternal diameters %/0 and )/0 mm, heated by constant temperature water bath and cooled by constant temperature water–ethylene glycol mixture *405 each by volume-/ The

number ofturns in the evaporator is varied from 4 to %3/ The working fluids employed are water, ethanol and R.)%3/ The results indicate a strong influence ofgravity and number ofturns on the performance/ The thermophysical properties ofworking fluids affect the performance which also strongly depends on the boundary conditions ofP,P operation/ Part 7 ofthis paper, which deals with development ofsemi.empirical correlations to fit the data reported here coupled with some critical visu. alization results, will appear separately/ %003 9lsevier +td/ All rights

reserved/ Keywords- Pulsating heat pipe( Parametric experimental study Corresponding author: Tel/: :$;.7)).684.%)$%( fax: :$;.7)).684.%0)0/ E.mail address- khandekar@ike/uni.stuttgart/de *S/ Khandekar-/ )34;.$3))@A . see front matter %003 9lsevier +td/ All rights reserved/ doi:)0/)0)6@S)34;.$3))*03-00)4;.4 Applied Thermal 9ngineering %3 *%003- %00;–%0%0 www/elsevier/com@locate@apthermeng
Page 2
1. Introduction Bscillating, loop type or pulsating heat pipes *P,Ps- are a relatively new type ofheat transfer devices, which may be classified in a special category ofheat pipes/ They

have been introduced in the mid.);;0s/ The first predecessor ofthe family ofP,Ps appeared in the );;0s C)–$D, a few examples ofwhich are shown in Eig/ )/ The basic structure ofa typical pulsating heat pipe consists ofmeandering capillary tubes having no internal wick structure/ 0t can be designed in at least three ways: *i- open loop system, *ii- closed loop system and *iii- closed loop pulsating heat pipe *C+P,P- with additional flow control check valves, as shown in Eig/ %/ The closed passive system thus formed is evacuated and subseFuently filled up partially with a pure

working fluid/ The optimum Fuantity ofworking fluid needed depends on various parameters and is still an area of research C4,6D/ The entire essence ofthermo.mechanical physics lies in the closed *constant vol. ume-, two.phase, bubble–liFuid slug system formed inside the tube.bundle due to the dominance of surface tension forces/ This tube.bundle receives heat at one end and is cooled at the other/ Temperature gradients give rise to temporal and spatial pressure disturbances in the wake of Nomenclature tube diameter, m acceleration due to gravity, m@s length, m number ofturns heat

transfer rate, inclination angle from horizontal axis Su/scripts a adiabatic section c condenser section crit critical value e evaporator section i inner max maximum Eig/ )/ Some practical designs ofpulsating heat pipes *adapted from C)–$D-/ %0)0 P0 Charoensawan et al0 1 2pplied Thermal Engineering 23 (2003) 2009–2020
Page 3
phase change phenomena *bubble generation and growth in the evaporator and simultaneous collapse in the condenser-/ The generating and collapsing bubbles act as pumping elements transporting the entrapped liFuid slugs in a complex

oscillating–translating–vibratory fashion( a direct conseFuence ofthermo.hydrodynamic coupling ofpressure@temperature fluctuations with the void fraction *mal.- distribution/ This causes heat transfer, essentially as a combination of sensible and latent heat portions/ The relative magnitude ofthese portions is also ofprofound interest and decisive to the overall thermal performance of the structure/ 0t has been indicated earlier that the sensible heat transfer is the maGor contributor in the overall heat exchange C6–8D/ Eurther studies have indicated that after a certain input heat

flux, the bubble–liFuid slug flow may break down into annular flow regime C;D/ The relative magnitude ofsensible and latent portions thus changes and is dependent on the flow pattern existing inside the tubes/ This aspect is another critical area that reFuires further investigations/ 0t has been shown by previous studies that a closed loop pulsating heat pipe is thermally more advantageous than an open loop device because ofthe possibility offluid circulation/ Although a certain number ofcheck valves have shown to improve the performance, miniaturization ofthe

device makes it difficult and expensive to install such valve*s- C3,)0D/ Therefore, a closed loop device without any check valve*s- is most favorable from many practical aspects/ Studies *mostly Fualitative- have already identified various design parameters affecting the performance of C+P,Ps C6,))D/ This paper presents results ofa large experimental matrix aimed at better un. derstanding the Fuantitative parameter dependency ofC+P,Ps/ 2. Experimental setups and procedure 0n conventional heat pipes, the adiabatic vapor temperature gives a very convenient way of standardizing

an experimental procedure/ 0n contrast, there is no well.defined adiabatic temper. ature in the case ofC+P,Ps/ Thus, in general, performance testing ofC+P,Ps may be conducted in two ways: *i- controlling the input heat flux and the condenser temperature, in which case the Eig/ %/ Schematic ofa pulsating heat pipe and its design variations/ P0 Charoensawan et al0 1 2pplied Thermal Engineering 23 (2003) 2009–2020 %0))
Page 4
evaporator temperature is a dependent variable and, *ii- controlling the evaporator and condenser temperature to give dependent heat throughput/ 0n the

present experimental plan, the latter strategy was adopted/ 9ssentially three parameters were fixed at the outset: *a- the average evap. orator temperature was always maintained at 80 C with the help ofa large water cooling bath *,AAK9, 8I3.7, and 0/04 C accuracy-/ The imposed mass flow always insured near iso. thermal conditions within 0/4 C, *b- in the condenser, an aFueous solution ofethylene glycol *405 by volume- with inlet temperature always maintained at %0 C circulated from a cold bath *,AAK9, I6.C$), and 0/04 C accuracy- and *c- the filling ratio *working

fluid volume inside the device@total internal volume ofthe device- was always maintained at 405 in all experimental set.ups/ The experimental set.up is shown in Eig/ 3/ 0t consisted ofthe tested C+P,Ps, the heating and cooling baths, a temperature data logger *Comark, C84)0, )0 channels, overall accuracy 0/4 C- and a flow meter *Platon, PG7$)) with accuracy of0/) l@min- to measure the flow rate ofthe coolant solution/ Eour chrom–alumel thermocouples *BM9GA.Type - were used to measure the temperature ofthe cooling solution, two each at the inlet and outlet sections ofthe

condenser/ The heat throughput was thus measured by calorimetric method applied to the condenser.cooling Gacket/ 0n addition, two thermocouples on the evaporator tube sections, four thermocouples on the adiabatic tube sections and two thermocouples on the condenser section completed the in. strumentation/ The tested C+P,Ps were made ofcopper capillary tube/ 7oth ends ofthe tube were connected together to form a closed loop structure which was located in the condenser in all the experiments/ The adiabatic section was well insulated with foam insulation *Armaflex-/ Eirst, the C+P,P was

evacuated *)0 Pa- and then filled with 405 ofthe total volume with the working fluid/ The inlet temperature ofthe hot and cold baths were set at the fixed values and the hot and cold fluids were supplied to the Gackets ofboth the evaporator and condenser sections/ After a Fuasi.steady.state was reached, the temperatures and flow rate were recorded/ Thus for a given configuration the heat throughput could be evaluated/ Then the influence parameters were varied according to the reFuired conditions/ The value ofcalculated was subGect to experimental Eig/

3/ Ketails ofthe experimental set.up/ %0)% P0 Charoensawan et al0 1 2pplied Thermal Engineering 23 (2003) 2009–2020
Page 5
uncertainties and errors that were later evaluated/ The complete experimental matrix is as shown in Table )/ 3. Results and discussion 30(0 Data accuracy Since the heat transfer rate of the C+P,P is calculated by measuring the volume flow rate and the inlet and outlet temperatures ofthe coolant flowing through the condenser section, the ac. curacy ofeach recorded data is inspected/ After carrying out a detailed error analysis with re. spective

accuracy ofindividual measurements and thermal losses, the data on maximum thermal performance reported here, as a whole, is within 305 accurate/ Lhen the device performance is higher, a measurable temperature gradient exists in the condenser cooling fluid inlet and outlet resulting in an error ofless than )05/ Lhile it may be argued that accuracy ofthe present reported data is not ofexcellent category, the essence ofthermofluidic characteristics and influ. ence parameter trends could be clearly demonstrated after data reduction/ 3020 Effect of operating orientation Bne

of the aims of good C+P,P design is to make the thermal performance, as far as possible, independent ofthe operating orientation/ At a first glance, two physical phenomena affect the C+P,P performance with respect to orientation/ The first is ofcourse, the effect ofgravity on slug flow and the second is the effect oftotal number ofmeandering turns on the level ofinternal temporal and spatial dynamic pressure perturbations/ 0n addition to these two, the input heat flux is also a strong parameter, which affects dynamic instability C)%,)3D,

especially in density wave oscillations, and is therefore believed to affect the thermal performance of C+P,Ps with respect to orientation/ This aspect remains to be further explored and will not be highlighted in this paper/ 0t is to be noted that for performance in vertical orientation, the effect of input heat flux has already been experimentally demonstrated C)$,)4D/ Table ) Complete experimental matrix Lorking fluids *mm- total *m- *m- *number ofturns- Later–ethanol–R.)%3 %/0 4 0/)4 4 %/0, )/0 4 0/)0 7 )0 0/)0 )6 0/)4 )) )4 0/)0 %3 0/)4 )6 Eill ratio always

maintained at 405 in all configurations All configurations tested at inclinations of0 *horizontal- to :;0 *vertical, evaporator down- P0 Charoensawan et al0 1 2pplied Thermal Engineering 23 (2003) 2009–2020 %0)3
Page 6
Classical experiments on the rise velocity ofa single bubble in a cylindrical tube have shown that as the 7ond number approaches a critical value approximately eFual or less than %, surface tension forces start predominating over gravity forces C)6,)7D/ There exists a discrepancy in the agreement ofthis critical limit with some sources Fuoting slightly

different values, e/g/ )/8$ C)8D/ This discrepancy is generally attributed to the tube material@working fluid contact angle char. acteristics, especially the hysteresis phenomenon/ 0fit is assumed that surface tension is indeed dominating in a particular experimental set.up that satisfies Bo %, then the shape ofa typical slug.bubble element should not change in vertical or horizontal orientation, especially regarding the symmetry ofliFuid film thickness around the bubble/ Eig/ $ shows the photograph ofstatic ethanol vapor bubbles suspended in liFuid ethanol inside 0 and

)/0 mm glass tubes, re. spectively, at room temperature *taken by I0KBI Coolpix 4700 Kigital Camera-/ Although the boundary conditions meet the critical 7ond number criterion, the effect ofgravity is clearly seen by the unsymmetrical shape ofthe bubble in the side view *Miew 7-/ R.)%3 bubbles are more unsymmetrical as surface tension is still lower/ 0t is also clear that in a non.operating, isothermal, partially filled C+P,P, the static pressure distribution traversing across the tube through the liFuid slugs and vapor bubbles is drastically different in vertical and

horizontal orientations C;D/ Thus, gravity does affect C+P,P dynamics even though the boundary conditions satisfy the critical 7ond number criterion/ This has indeed been demonstrated by the experimental results that follow/ The second design aspect is related to the total number ofmeandering turns in a given C+P,P/ Eigs/ 4 and 6 summarize the thermal performance for the entire experimental matrix, with respect to the inclination angles/ Eor a given case, the performance is scaled by the maximum perfor. mance achieved for that case during operation in the full range of inclination

angles/ 0t can be clearly seen that the performance independence with orientation is affected by the number of Eig/ $/ 0mages ofethanol slugs and bubbles in glass tube under static isothermal conditions/ %0)$ P0 Charoensawan et al0 1 2pplied Thermal Engineering 23 (2003) 2009–2020
Page 7
turns/ Eor 0 mm devices, the effect could be clearly separated into two cases by using a certain critical value ofnumber ofturns * crit -/ 0n this case, the critical number ofturns was approximately )6 turns *with the exception of )4 cm, )6 turns and ethanol as working fluid-/ 0n

case of 0 mm devices too, similar trends are seen as depicted in Eig/ 6*a- and *b-/ Eor this case, the critical value ofnumber ofturns tends to be higher than for % mm tubes/ 0n addition, for ) mm tubes, measurable heat transfer was not possible with water filled devices in the entire range ofoperating orientation/ Lhen is less than a certain crit , the C+P,P cannot satisfactorily operate in the horizontal orientation and vice versa/ Eor crit , the highest thermal performance normally occurs at vertical bottom heating mode decreasing continuously as the device is turned towards

horizontal/ Eig/ 4/ Thermal performance for % mm: *a- crit and *b- crit P0 Charoensawan et al0 1 2pplied Thermal Engineering 23 (2003) 2009–2020 %0)4
Page 8
,owever, when the number ofturns is higher than the critical value, although the performance improves with increasing the inclination angle from horizontal orientation, the performance re. mains nearly comparable from vertical position to about 60 / The critical number ofturns de. pends on the working fluid and size ofthe used capillary tube *and may also depend on the input heat flux, as previously mentioned-/ 3030

Effect of construction on performance Erom the previous section it is clear that gravity is certainly affecting the thermal performance/ The C+P,Ps tested in the present work are inline designs in which all the tubes are in one plane *as in Eig/ %-/ Thus when such a design is made horizontal, the gravity vector is non.existent on all the tubes simultaneously/ Eig/ 7*a- and *b- suggests two design variations which, only by virtue of the construction and physical arrangement oftubes, should have a favorable effect on thermal Eig/ 6/ Thermal performance for ) mm: *a- crit and

*b- crit %0)6 P0 Charoensawan et al0 1 2pplied Thermal Engineering 23 (2003) 2009–2020
Page 9
performance with respect to orientation/ Kesign variation has partially bent tubes turns and so ifthis structure is operated horizontally, the gravity vector will still be partly functional/ 0n variation the tubes are bent in a three dimensional manner and so the gravity will affect the flow irrespective ofany global orientation ofthe heat pipe/ These constructional variations are cer. tainly believed to enhance the performance/ 3040 Effect of tu/e inner diameter The

internal diameter is a parameter which necessarily affects the very definition ofa pulsating heat pipe/ 7eyond a particular limit, all the working fluid will tend to settle down by gravity and the device will stop functioning as a pulsating heat pipe/ 0t will rather behave like an intercon. nected array ofclosed two.phase thermosyphons C6D/ Eig/ 8*a-–*c- shows the effect oftube inner diameter for vertically operating devices having )0 cm/ Eor a given number ofturns, the performance improved with internal diameter/ This is realized since there is more mass inventory

ofworking fluid coupled with reduced pressure drop/ 0n general, the entire experimental matrix exhibited this trend/ Eurther, doubling the diameter did not double the performance/ At the same internal diameter and evaporator length, the performance is higher with increasing the number of turns/ Thus, for a specified temperature gradient between evaporator and condenser, the per. formance can be increased by increasing the tube inner diameter and@or the number of mean. dering turns/ 3050 Effect of working 5uid The thermophysical properties ofthe working fluid coupled

with the geometry ofthe device have profound implications on thermal performance of the device/ This affects the following: Eig/ 7/ Kesign variations for improving performance with respect to orientation/ P0 Charoensawan et al0 1 2pplied Thermal Engineering 23 (2003) 2009–2020 %0)7
Page 10
the relative share oflatent and sensible heat in the overall heat throughput( the possibility ofhaving different flow patterns in the device, e/g/ slug–annular flow regime inter.transition( the average flow velocity and overall pressure drop *including effect

ofgravity-( bubble nucleation, collapse, shapes, agglomeration and breakage( bubble pumping action, etc/ 0n vertical orientation for the %/0 mm devices, water filled devices showed higher performance as compared to R.)%3 and ethanol/ 0n contrast R.)%3 and ethanol showed comparable perfor. mance in case of)/0 mm devices with water showing very poor results/ This is seen in Eig/ ; and Eig/ 8/ 9ffect ofdiameter on the heat throughput * )00 mm, vertical orientation-/ Eig/ ;/ 9ffect ofworking fluid on the thermal performance/ %0)8 P0 Charoensawan et al0 1 2pplied Thermal

Engineering 23 (2003) 2009–2020
Page 11
also in Eig/ 8/ Later has a very high surface tension, a very low sat , a very high latent and specific heat and reasonably higher dynamic viscosity as compared to R.)%3/ Since the thermal performance is a complex combination of the above noted it is certainly difficult to prescribe or proscribe a certain working fluid unless all the boundary conditions are exactly known and in. dividual effects have been explicitly isolated and Fuantified/ 4. Summary and conclusions A range ofclosed loop pulsating heat pipes has

been experimentally investigated to study the effects ofvarious influence parameters/ The effect ofinternal diameter, operating inclination angle *gravity-, working fluid and number ofturns on the thermal performance has been demonstrated/ The following main conclusions can be drawn from the study: Gravity certainly affects the heat throughput/ Although the internal diameter ofthe tubes tested in the present study, as governed by the critical 7ond number, is well within the specified limit, bubble shapes are affected by the buoyancy forces/ A certain

critical number ofturns is reFuired to make horizontal operation possible and also to bridge the performance gap between vertical and horizontal operation/ This is attributed to the increase in the level ofinternal perturbations/ Kifferent fluids are beneficial under different operating conditions/ An optimum tradeoff ofvar. ious thermophysical properties has to be achieved depending on the imposed thermo.mechan. ical boundary conditions/ Eor a given temperature differential, performance improves with increase in internal diameter/ The internal diameter is a

parameter which necessarily affects the very definition ofa pulsating heat pipe/ 0t may also be safely concluded that thermo.mechanical interactions and instabilities in a pulsating heat pipe in particular, and in capillary sized tubes *mini.micro channels- in general, is Fuite complex and further experiments are indeed needed/ The fact that pulsating heat pipes are closed systems in which the velocity scale is dependent on the imposed thermal boundary con. ditions *and is not known a priori- makes it all the more difficult for analysis/ This aspect, in. cluding semi.empirical

modeling approach coupled with critical visualization results is addressed in Part 7 ofthis paper C);D/ Acknowledgements This research work was done Gointly by Eaculty of9ngineering, Chiang Mai Nniversity, Thailand and 0nstitut f ur Kernenergetik und 9nergiesysteme *0K9-, Nniversit at Stuttgart, Ger. many under the auspices ofRoyal Golden Oubilee Scholarship ofthe Thailand Research Eund *under Contract Io/)/M/CM@$3@A/)- and Keutscher Akademischer Austauschdienst *KAAK-/ The work was also partly supported by Keutsche Eorschungsgemeinschaft *KEG- under Grant GR.$)%@33.)/ P0 Charoensawan et al0 1

2pplied Thermal Engineering 23 (2003) 2009–2020 %0);
Page 12
References C)D ,/ Akachi, NS Patent, Patent Iumber $;%)0$), );;0/ C%D ,/ Akachi, NS Patent, Patent Iumber 4%);0%0, );;3/ C3D ,/ Akachi, NS Patent, Patent Iumber 4$;0448, );;6/ C$D ,/ Akachi, E/ Pol sek, P/ Stulc, Pulsating heat pipes, in: Proceedings ofthe 4th 0nternational ,eat Pipe Symposium, Melbourne, Australia, );;6, pp/ %08–%)7/ C4D S/ Khandekar, I/ Kollinger, M/ Groll, Nnderstanding operational regimes ofpulsating heat pipes: an experimental study, Appl/ Therm/ 9ng/ %3 *6- *%003- 707–7); *0SSI )34;.$3))-/ C6D M/

Groll, S/ Khandekar, Pulsating heat pipes: a challenge and still unsolved problem in heat pipe science, Arch/ Thermodyn/ %3 *$- *%00%- )7–%8 *0SSI )%3).0;46-/ C7D S/ Iishio, S/ Iagata, S/ 7aba, R/ Shirakashi, Thermal performance ofS9MBS heat pipes, in: Proceedings of)%th 0nternational ,eat Transfer Conference, Grenoble, Erance, vol/ $, %00%, pp/ $77–$8% *0S7I %.8$%;;.307.)-/ C8D M/ Shafii, A/ Eaghri, P/ Zhang, Thermal modeling ofunlooped and looped pulsating heat pipes, ASM9 O/ ,eat Transfer )%3 *%00)- ))4;–))7%/ C;D S/ Khandekar, M/ Groll, An insight into thermo.hydrodynamic coupling in

pulsating heat pipes, 0nt/ O/ Therm/ Sci/ *Rev/ Gn/ Therm/- in press *0SSI: )%;0.07%;-/ C)0D S/ Kuminy, 9xperimental investigation ofpulsating heat pipes, Kiploma Thesis, 0nstitute ofIuclear 9ngineering and 9nergy Systems *0K9-, Nniversitt Stuttgart, Germany, );;8/ C))D S/ Maezawa, K/ Minamisawa, A/ Gi, ,/ Akachi, Thermal performance of capillary tube thermosyphon, in: Proceedings ofthe ;th 0nternational ,eat Pipe Conference, AlbuFuerFue, Iew Mexico, );;4/ C)%D O/ 7oure, A/ 7ergles, +/ Tong, Review oftwo.phase flow instability, Iucl/ 9ng/ Kesign %4 *);73- )64–);%/ C)3D P/ Saha, M/ 0shii,

I/ Zuber, An experimental investigation ofthermally induced flow oscillations in two.phase systems, ASM9 O/ ,eat Transfer ;8 *);76- 6)6–6%%/ C)$D 7/ Tong, T/ Long, K/ Boi, Closed.loop pulsating heat pipe, Appl/ Therm/ 9ng/ %) *)8- *%00)- )8$4–)86% *0SSI )34;.$3))-/ C)4D S/ Khandekar, M/ Groll, P/ Charoensawan, P/ Terdtoon, Pulsating heat pipes: thermo.fluidic characteristics and comparative study with single phase thermosyphon, in: Proceedings ofthe )%th 0nternational ,eat Transfer Conference, Grenoble, Erance, vol/ $, %00%, pp/ $4;–$6$ *0S7I %.8$%;;.307.)-/ C)6D T/ ,armathy,

Melocity oflarge bubbles in media ofinfinite or restricted extent, A0Ch9 O/ 6 *%- *);60- %8)–%88/ C)7D 9/ Lhite, R/ 7eardmore, The velocity ofrise ofsingle cylindrical air bubbles through liFuids contained in vertical tubes, Chem/ 9ng/ Sci/ )7 *);6%- 34)–36)/ C)8D E/ 7retherton, The motion oflong bubbles in tubes, O/ Eluid Mech/ *)0- *);6)- )67–)88/ C);D S/ Khandekar, P/ Charoensawan, M/ Groll, P/ Terdtoon, Closed loop pulsating heat pipes––part 7: visualization and semi.empirical modeling, Appl/ Therm/ 9ng/ *under review- *0SSI )34;.$3))-/ %0%0 P0 Charoensawan et al0 1 2pplied Thermal

Engineering 23 (2003) 2009–2020