Review of Reduced Gravity Boiling Heat Transfer: European Research

J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 (252–263)

Special Issue: Thermal Engineering in Microgravity Part 1: Boiling (Review)

Paolo Di MARCO

Abstract The ˆrst section of this paper (overview on boiling heat transfer characteristics) is directed to those who are not specialists in the ˆeld of boiling heat transfer, in order to make them aware of the fundamental physical mechanisms, the advantages and the main problems of this technique. Next, some of the basic knowledge achieved on terrestrial boiling heat transfer is outlined, along with the beneˆts and the problems raised from its application in space. Finally, the past and forthcoming research activities of the European teams in this ˆeld are reported, in addition to the main results achieved and a short overview on the available test facilities.

the rejected heat from the utilities to the external radi- 1. Introduction ators where power is radiated to space. Thermal con- The safe operation of spacecrafts relies on the trol systems of individual modules are connected to it. e‹ciency of their heat removal systems; besides, sever- Thermal bus concept must provide stable thermal re- al sophisticated scientiˆc apparatuses require very pre- gime at any number of attached module and at any cise temperature control. The ultimate heat sink is con- variation of thermal load. At present, thermal buses stituted by the cold outer space, to which, being in are generally mechanically-pumped ones, in which a vacuum, heat can be transferred by radiation unique- single-phase ‰uid is operated, hence their heat removal ly. As the outer surface of the spacecraft has in most capability is based on the so-called sensible heat of the cases poor heat exchange characteristics, in the ab- ‰uid, i.e. in its capacity of absorbing energy by rising sence of dedicated systems the internal temperature its own temperature. However, as well known, a ‰uid would rise to unacceptably high values for both system can exchange energy in a diŠerent way as latent heat, operation and human survival. Furthermore, systems i.e. changing its aggregation state from liquid to vapor are needed to route heat from internal utilities, like and vice versa (boiling and condensation, respective- electronic equipment, payloads, air conditioning sys- ly); the Gibbs' rule states that, as long pressure is kept tems and so on, to outer heat sink. constant and the substance is pure, this process is To date, spacecraft active thermal control has been isothermal. The main advantages of boiling systems accomplished using pumped single-phase liquid loops. are that they are nearly isothermal, require smaller Systems of this kind (termed Active Thermal Control heat exchange surface, have high power density Systems, ATCSs) have been used on Mercury, Gemini, removal and consequently require lower pumping Apollo, MIR space station and are currently used on power. As a result, boiling is recognized as a very eŠec- US Space Shuttle, and Russian Soyuz spacecraft. tive technique to exchange high heat ‰uxes from heat- However for larger spacecrafts, like the International ed bodies and is widely applied in on-earth technology Space Station (ISS), some basic diŠerences with in component heating and cooling. respect to the actual systems are present: As detailed in the following, boiling systems are dis- the total heat load will pass from a few kilowatts to tinguished in forced-convection ones, in which the tens of kilowatts; ‰uid is driven by an external device like a pump, and the heat transport distance will be longer (tens of natural convection or pool boiling systems, in which meters); the motion of the ‰uid is originated by the thermal a greater ‰exibility is required to accommodate the gradients within it: these in turn induces variations of needs (heat load, sink temperature) of diŠerent other properties like density, electrical or magnetical modules as they are put into operation. permittivity etc., on which an external force ˆeld can Heat removal in a spacecraft is generally pursued via act to induce motion. Pool boiling systems require no a thermal bus, i.e. a loop in which a ‰uid transports external pumping power, henceforth their interest is

LOTHAR, Department of Energetics ``L. Poggi'', University of Pisa, Italy via Diotisalvi 2, I–56126 Pisa, Italy (E-mail lothar＠ing.unipi.it)

2 ― 252 ― Review of Reduced Gravity Boiling Heat Transfer: European Research even greater for low-power demanding environments. designers. These correlations are mostly empirically Summarizing, the adoption of boiling systems may based and, although they have proven very eŠective in yield substantial saving of weight, space and power developing terrestrial equipment, their validity aboard spacecrafts in order to work out the crucial diminishes very rapidly outside their range of validity, problem of heat rejection. However, as detailed in the that is they might poorly represent the situation at following, a number of problems have still to be solved diŠerent gravity levels, and in particular in microgravi- to adapt these systems to microgravity environment. ty conditions. The former arguments, recently surveyed by Delil1) As a consequence, an extensive and careful ex- stress the need and the importance of research in this perimental activity under conditions as close as possi- êld. ble to the actual ones is required to improve design of boiling equipment in space. The physical insight 2. Overview on Boiling Heat Transfer Characteris- gained in this way in conditions diŠerent from the tics ground ones may also lead to a substantial improve- In any heat transfer process between a solid surface ment of the comprehension of physical mechanisms and a ‰uid, the heat rate Q is commonly expressed by governing boiling phenomena and their interaction, the convection law (usually referred to as Newton's) which, as shown later on, is still an open question. In the following, a short review of the main aspects Q＝aA(T －T )(1) w ref of pool boiling heat transfer will be carried out, mainly where a is the so-called heat transfer coe‹cient, A is devoted to those who are not specialists in the êld. the heater area and (Tw–Tref ) is the diŠerence between The process of boiling is intrinsically a non-stationary, the temperature of the surface and a convenient refer- non-equilibrium one, although quasi-cyclic repetitions ence one taken in the ‰uid; in boiling phenomena, the are typical. Its study has to do with the most complex saturation temperature Tsat is generally adopted for transport phenomena encountered in engineering, and Tref. Both the heat transfer area and the temperature several complex factors are involved, like interaction diŠerence should be kept as small as possible, the between the solid surface of the heater and the ‰uid, former to minimize weight and investment costs, and interaction between liquid and vapor phases, and the latter to minimize entropy generation and avoid phase transport mechanisms. As a consequence, a surface overheating, which in turn may lead to equip- satisfactory description of the transport phenomena in ment failure. Consequently, the heat transfer coe‹ci- the ‰uid in terms of the diŠerential equations of con- ent has to be as high as possible to accommodate large servation laws is still unaŠordable. heat ‰uxes. 2.1 General Aspects of Pool Boiling Heat Transfer Boiling heat transfer coe‹cients are order of magni- Following the approach originally developed by tude higher than in single-phase ‰ow. This makes it a Nukiyama2) in his early experiment, heat transfer per- very suitable technique for applications requiring very formance in pool boiling is commonly reported in a e‹cient, compact and lightweight devices. However, heat ‰ux-temperature diŠerence plot (boiling curve). especially at low ‰uid velocity, the vapor removal The curve generally exhibits the trend shown in Fig. 1. mechanisms are in‰uenced by buoyancy forces, which In the boiling curve (solid line in Fig. 1) several heat of course are lacking in the absence of gravity: in such transfer regimes can be identiêd. In a ˆrst zone (AB) conditions the remaining forces (mainly inertial and no boiling exists and heat transfer is by natural convec- superˆcial ones) acquire a dominant role. This might tion. In microgravity, if buoyancy and other driving lead to a diŠerent behavior and even to early heat forces are excluded, natural convection cannot take transfer degradation (the so-called critical heat ‰ux or place and is replaced by transient conduction in the liq- dryout phenomena, i.e. a surface blanketing by a layer uid layer. When temperature at the surface exceeds the of vapor). saturation value of a required amount, bubbles are Although a considerable amount of research has generated in surface cavities by heterogeneous nuclea- been carried out over the last forty years in normal tion and boiling starts. This implies a strong increase gravity conditions, and to some extent with diŠerent in heat transfer performance, and temperature diŠer- gravity accelerations, our predictive capability is still ence is suddenly decreased, path BC. The temperature very limited for conditions that do not exist on Earth. overshoot in point B may be so high as to compromise Generally, a choice is given to develop either heuristic the operation of temperature sensitive equipment, like or mechanistic models for predictive purposes, and electronic devices. Along the path CD (nucleate boil- although the former can be very eŠective, only the lat- ing), more and more nucleation sites are activated and ter can be used to conˆrm physical understanding, the heat ‰ux q! steeply increases with wall superheat which in turn gives opportunity for new and improved DTsat＝Tw－TsaT; this heat transfer mode is termed applications. nucleate boiling and is the most important for industri- A large amount of boiling heat transfer correlations al applications due to its high e‹ciency. However, it were developed in the past to ˆt the needs of industrial cannot be sustained indeˆnitely: beyond a maximum

J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 253 ― 3 Paolo Di MARCO

di‹culties because they retain parameters given on an empirical basis. Thus, prediction of heat transfer performance in nucleate pool boiling still relies on empirical correlations. A large amount is available in open literature. Generally, the dependence on gravity of the so-called heat transfer e‹ciency a? (Straub3))canbe expressed by a power law a  n e＝＝ (2) a0 (0) The exponent is diŠerent if constant heat ‰ux or constant wall temperatures are compared. Straub et al.3,4) reported that n can range from －0.35 to 0.5; a few ex- amples are given in the following. One of the most widely used correlations is Fig. 1 Pool boiling curve. Rohsenow's5), which can be rearranged as q!l 1 q!l －0.67 Nu＝ L ＝ L Pr －0.7 (3) DT k C h h l value named critical heat ‰ux (CHF, point D), it is sat l sf ( l f) suppressed (current theories about this phenomenon which theoretically implies n＝0.83, for the same value will be outlined later). Afterwards, two paths can be of q!. If a constant value of DTsat is retained, n＝0.5 is followed depending on the controlling variable. If this got. Nonetheless, advice is given to consider gravity variable is heat ‰ux, as in electric or nuclear equip- acceleration as a mere dimensional constant in it ment, a small rise in q! causes a sudden jump from (Dhir6)). Thus, the correlation has been successfully point D to E with a very large increase in wall tempera- used to predict pool boiling performance in micrograv- ture which can even lead to the destruction of the heat- ity, by leaving the terrestrial value of g in it (Motoya et er (the so-called burnout phenomenon). Beyond point al.7)). This is consistent with pool boiling performance F the curve has a much slighter slope than CD and the being poorly aŠected by gravity value. However, ques- heat transfer regime is termed ˆlm boiling. In ˆlm boil- tions could be raised about the mechanistic models as- ing, the surface is completely blanketed with vapor sumed to justify the form of Eq. (3). Zhang & Chao8) and at most sporadic liquid contacts may occur: this is proposed to retain the Rohsenow's model also in the reason for the strong heat transfer degradation. An microgravity, supplementing it with the actual bubble unstable vapor ˆlm covers the heater, oscillating with departure diameter in place of the Laplace length. an assigned wavelength; bubbles, spaced one The correlation by Cooper9) wavelength, detach periodically from the ˆlm surface; －－ p 0.12 0.091 ln Rp p 0.55 radiation contributes signiˆcantly to total heat transfer a＝C －0.4343 ln in this regime, especially at high superheat. If heat ‰ux (pcrit) ( pcrit) is now progressively reduced (see the arrows), this ×M－0.5q!0.67 (4) curve can be covered down to point F, the minimum

ˆlm boiling heat ‰ux (MFB), where a further decrease is in dimensional form (Rp is the roughness in mmand takes the system back to G (hysteresis loop). On the M the molecular weight of the ‰uid) but has no gravity other hand, if the wall temperature is the controlling acceleration in it, so n＝0. Several other models have variable, as in heat exchangers, also the unstable path this feature, e.g. Yagov's10), and Cornwell's and DF (termed transition boiling) can be covered, by in- Houston's11). Another well established correlation creasing or decreasing the wall superheat. proposed in VDI Heat Atlas12) (1993) is due to Stephan To summarize, although nucleate boiling is a very and Preusser suitable heat transfer regime, it must be carefully oper- q!D q!D 0.674 r 0.156 ated: Caution must be taken in establishing it from Nu＝ d ＝0.1 d  DT k T k r cold conditions without damaging the equipment due sat l ( sat l) ( l) to temperature overshoot, and heat ‰ux must always h D2 0.371 a2r 0.35 be maintained safely under the critical heat ‰ux value, × f d l l Pr－0.16 (5) a2 sD l beyond which heat transfer degradation takes place ( l ) ( d) which is generally unacceptable. where Dd is the bubble detachment diameter, given by 2.2 Pool Boiling Correlations and Their Extrapola- s tion to DiŠerent Gravity Levels D ＝CF ＝CFl (6) d (r －r ) L Though several mechanistic models of boiling l  phenomena have been developed, they encounter Hence, the gravity acceleration appears through the

4 J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 254 ― Review of Reduced Gravity Boiling Heat Transfer: European Research

capillarity length inside Dd, resulting in a very little physical implication, since heaters that are considered value of n (n＝－0.033 for the same heat ‰ux). Once a ``large'' in normal gravity may become utterly reference value aref is determined by experiments, or by ``small'' as gravity decreases. On the other hand, in tables given in VDI Heat Atlas, or by Eq. (5), the value the currently available microgravity experimental of a for other heat ‰uxes can be determined as: facilities, the adoption of very large heaters is prohibit- q! m p 0.3 ed by space, power and weight limitations. Even as- a＝a where m＝0.9－0.3 (7) suming that the present correlations can be extended to ref q! p ( ref) ( crit) reduced gravity conditions, expressions for K (e.g. Di Marco & Grassi13) found a good agreement with Lienhard and Dhir16)) yield a non-trivial dependence of VDI correlation, Eq. (7) for low and intermediate heat critical heat ‰ux on gravity. This means that for very ‰uxes on a wire, both in normal and in reduced gravi- reduced gravity, a simple ``power law'' dependency, ty. like 1/4 or 1/8 is not necessarily valid. 2.3 Critical Heat Flux 2.4 General Aspects of Forced Convective Boiling At present, it is unclear if just one mechanism deter- Heat Transfer mines the occurrence of critical heat ‰ux (CHF) in any In forced convective boiling the phenomenology is geometrical and thermodynamic condition. Regardless more complicated due to additional system eŠects. of modeling, critical heat ‰ux data on ‰at plates have Generally, taking as reference a vertical heated pipe, often been correlated in the so-called Zuber- two kinds of regimes are possible, depending on inlet Kutatelatze form subcooling and ‰ow rate, as sketched in Fig. 2 (the vertical scale is purely indicative). Referring to situation q! ＝Kr 0.5h [s(r －r )]0.25 (8) CHF  f f  A (low ‰ow rate) in the ˆgure, starting from single- In Eq. (8), if the heater is large with respect to the phase liquid heat transfer, in which the heat transfer Taylor wavelength, K (often referred to as Kutatelatze coe‹cient is essentially constant, nucleation is initiat- number) is a constant: K canvaryintherange ed at the wall when the bulk of the ‰uid is still sub- 0.119–0.157, (Grassi14)) and for ‰at plates was as- cooled (subcooled nucleate boiling). Afterwards, a sumed 0.131 by Zuber15) (1959) and 0.149 by Lienhard point is attained in which the bulk of the ‰uid becomes &Dhir16). saturated (quality x＝0), and saturated nucleate boil- Pool boiling on wires and small bodies has been ex- ing takes place. The ‰ow pattern in this region is grad- tensively studied in a number of papers over the ually modiˆed from bubbly to slug and annular, with a 60–70s. A non-trivial dependence of critical heat ‰ux progressive increase in heat transfer coe‹cient. When on the diameter of the wire has been reported. The the liquid ˆlm at the wall is destroyed due to thinning most suitable group to scale the eŠect of the diameter and instability, the liquid deˆcient region is entered, is the so called dimensionless length, i.e. the square- with a sudden decrease of heat transfer coe‹cient and root of the Bond number a consequent increase of wall temperature. The ‰ow regime is now dispersed drop. The droplet evaporation (r －r ) r R?＝ Bo＝r f ＝ (9) s lL The scattering of experimental data in literature is quite high for Boº0.15, allowing Lienhard & Dhir16) and Sun & Lienhard17) to claim that they can no longer be correlated by R? alone. As a result of photographic studies on wires, Bakhru & Lienhard18) concluded that for R?º0.07 the obtained boiling curve exhibits a con- tinuous trend, from boiling inception up to stable ˆlm boiling (provided that this traditional deˆnition still holds) with neither a minimum in heat ‰ux nor a jump in wall superheat. This implies that the very concepts of critical heat ‰ux and minimum ˆlm boiling become questionable. In this case the mechanism leading to complete blanketing of the surface is possibly related to the vapor front propagation on the surface or to the coalescence of the bubble population. Also the properties of the material of the heater may play a role. It is very important to note that a variation in gravity acceleration aŠects Bo as well as a variation in size of the heater: the scaling length in Bond number is Fig. 2 Flow boiling regimes. Case A (left): low ‰owrate, again the capillarity length. If true, this has a strong Case B (right): high ‰owrate.

J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 255 ― 5 Paolo Di MARCO causes an increase in vapor velocity, and in heat trans- ties, so far most of the experiments worldwide per- fer coe‹cient, up to the point where the conditions of formed concerned pool boiling heat transfer. A com- heat transfer to single-phase vapor are eventually mon feature of all the experimental facilities must be reached. In the situation B, at higher ‰ow rate, the cri- stressed: due to the absence of gravity, a free surface sis is reached directly from nucleate boiling conditions: separating liquid and vapor must be avoided. All the a vapor layer is formed at the wall (reversed annular experimental containers were thus initially ˆlled with ‰ow regime). liquid, and connected to a bellows to allow for thermal Both conditions were extensively studied, mainly in dilatation of the ‰uid and for volume compensation the ˆeld of nuclear reactor safety and design, and the due to bubble formation. In this way, pressure and reader is addressed to specialized literature (e.g Del- subcooling conditions could also be varied during the haye et al.19), Collier and Thome20)) in which appropri- experiments. ate correlations for each regime are given. The ˆrst experiments of pool boiling in microgravity In both the outlined situations, nucleate boiling re- were initiated in the late 50s in US. Siegel and cowor- gime cannot be sustained indeˆnitely. A boiling transi- kers (Siegel et al.21–23)) used a 2.5 m-high droptower. tion is eventually reached, which leads to severe heat Since then, experiments were carried out in all the transfer degradation. Usually, this transition is gener- available facilities, namely droptowers and drop- ally termed dryout in conditions of type A in Fig. 2 shafts, parabolic ‰ights, sounding rockets and orbital and departure from nucleate boiling (DNB) in condi- platforms. DiŠerent heater shapes were used. Thin tion of type B in the same ˆgure. DiŠerently from pool wires were used for their low thermal inertia, allowing boiling situation, the crisis in this case depends not for fast transients like in parabolic ‰ights, and for sim- only on local conditions, but on the whole evolution of ple data conditioning (they can be used as resistance the ‰uid before the point of the crisis. In forced con- thermometers). Plates (up to 50 mm diameter) are vective boiling inertial eŠects are expected to dominate more di‹cult to operate, but have more practical sig- over buoyancy in driving phases, except at low ‰ow niˆcance for applications; ˆnally, small heaters were conditions: this threshold, however, has still to be de- adopted to investigate individual bubble behavior. termined in micro-g conditions. Besides, some basic Worldwide studies up to 1990 were surveyed by mechanisms governing bubble evolution in the near- Straub4) andupto2000byDiMarcoandGrassi24). wall region are common to pool boiling regimes. The mechanism of steady state pool boiling in reduced gravity in saturated or slightly subcooled con- 3. European Research ditions was described by several authors, and there is 3.1 Boiling experiments in microgravity substantial agreement on these observations (e.g. Kim In the following, highlights of the experimental ac- et al.25),Leeetal.26),Okaetal.27)). A large bubble re- tivities carried out in the ˆeld by European teams are sides at a short distance from the heater and acts as a given. Their main features are summarized in Table 1; reservoir, engulˆng bubbles forming on the surface, research was mainly focused on pool boiling. In fact, see Fig. 3. This large bubble maintains its size due to due to space and power limitations in available facili- balance of condensation at its cap and coalescence of

Table 1 Pool boiling in microgravity: main features of the European experimental activities Legend: OF: orbital ‰ight; SR: sounding rocket; PF: parabolic ‰ight; DT: droptower/dropshaft.

REFERENCE FLUID HEATER NOTES

Straub et al., 3 SR R113 Wire 0.2 mm Subcooled conditions

Zell,Straubetal., SR R113 Flat plate 20×40 mm Saturated and subcooled conditions 3, 28, 62

Straub et al., PF R12 Wire, 0.2 mm Several ‰ights 3, 63, 4, 62 0.05 mm Saturated and subcooled conditions Pipe, 8 mm o.d. Flat plate 40×20 mm

Straub and Micko, OF R134a Wire, 0.2 and 0.05 mm Saturated and subcooled conditions 3, 30

Straub at al, OF R11 Hemispherical Saturated and subcooled conditions 3, 64, 65, 66 DT R123 heater 0.26 mm diameter Mainly devoted to study individual bubble behavior Circular heaters, 1, 1.5 and 3 mm dia.

Di Marco and Grassi, PF R113, Wire, 0.2 mm diameter Partial g-level tested 13, 24, 33, 36 FC72 Electrostatic ˆeld applied Di Marco and Grassi, SR FC72 Wire, 0.2 mm diameter Electrostatic ˆeld applied 13, 34

6 J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 256 ― Review of Reduced Gravity Boiling Heat Transfer: European Research

cement (15z) of pool boiling heat transfer was reported for wire geometry. Experiments with R123 and a hemispherical heater (1.41 mm diameter) were also conducted in the BDPU facility, owned by ESA (Straub et al.32), Steinbichler et al.31)) for heat ‰ux up to 300 kW/m2, reduced pres- suresfrom0.04to0.21andsubcoolingupto60K. Also in this case, enhancement of pool boiling heat transfer with respect to terrestrial gravity was encoun- Fig. 3 Saturated or quasi-saturated boiling pattern encountered, decreasing with increasing heat ‰ux. At low sub- tered in microgravity over a ‰at plate. cooling, strong thermocapillary ‰ow was observed around the bubbles attached to the surface. At higher subcooling, this mechanism ceased due to the small new, small bubbles at the base. Lateral coalescence of bubble size and was replaced by the ``pumping eŠect'' bubbles along the surface was observed, with conse- of growing bubbles: the bubble grows, pushing out- quently induced motion in the ‰uid, causing small os- wards the heated liquid in the thermal boundary layer, cillations in the heater temperature. It is inferred that then, once its surface gets in contact with the cooler the dimensions of the surface may aŠect liquid renewal liquid, it rapidly collapses restarting the process. This under the large bubble and thus the possibility of process was too fast to allow the establishment of ther- maintaining steady state conditions. The probability mocapillary ‰ow. of having steady state conditions increased with sub- On the basis of his experiments, Straub3) identiˆed cooling. Both enhancement and degradation of perfor- three basic boiling conˆgurations: saturated or slightly mance with respect to terrestrial conditions were en- subcooled boiling, characterized by bubble detach- countered, as detailed in the following. ment and coalescence; subcooled boiling, with bubbles Straub and coworkers performed extensive tests adherent to the surface acting as heat pipes, with with various organic refrigerants (R11, R12, R113, strong thermocapillary ‰ow along their surface; and R123, R134a) in parabolic ‰ights, sounding rockets highly subcooled boiling, with very small bubbles and orbital facilities from the early 1980s to date, us- growing and collapsing very quickly, acting as little ing various heater geometries (‰at plate, wires, small pumps, with no thermocapillary ‰ow again. It is still circular and hemispherical heaters) in a wide range of an open question whether nucleate pool boiling is en- reduced pressure (up to 0.68). Straub3) recently issued hanced or degraded in micro-g. Straub3) has recently a comprehensive review of his own activity, whose summarized the main outcomes from his research: a) results can only shortly be summarized below. enhancement or degradation are anyway limited (±30 Sounding rocket (TEXUS) experiments investigated z max); b) the heat transfer e‹ciency a? (see Eq. 2) al- boiling of R113 on wires and ‰at plates (Zell et al.28)). ways decreases with increasing heat ‰ux; it can be The same boiling mechanism as above was described greater than one at low heat ‰ux and lower than one at for ‰at plates; and a decrease of heat transfer high heat ‰ux; c) subcooling seems not to in‰uence the coe‹cient was found. For wires, bubble detachment in value of a?;d)a? drops oŠ at high reduced pressure. subcooled conditions was attributed to Marangoni Di Marco and Grassi24) conducted experiments of ‰ow. When vapor production was high enough, sur- pool boiling of R113 and FC–72 on a 0.2-mm platinum face tension forces were no longer able to rewet the wire, in slightly subcooled conditions, both in parabol- surface and a ˆlm boiling situation was maintained ic ‰ight and in sounding rocket. High values of heat throughout. The heat transfer coe‹cient for wires was ‰ux, up to CHF, were tested. Data were also recorded found to be almost independent of gravity level in both in the enhanced gravity phase of trajectory and during sounding rocket and parabolic ‰ight experiments special trajectories resulting in a constant gravity value (Straub3,29)). The onset of nucleate boiling was slightly of 1.5 g for 40 s. Pool boiling data at Martian gravity in‰uenced by gravity acceleration, and tended to occur level (0.4 g) were also collected. Despite a very evident at lower superheating in micro-g due to the lack of free change in bubble size and velocity, no appreciable convection, which mitigates the temperature of the su- eŠect of gravity acceleration on the heat transfer perheated layer. coe‹cient in nucleate boiling on a wire was found. Straub (Straub and Micko30); Steinbichler et al.31)) Critical heat ‰ux (CHF) was clearly detected and it was reported results of experiments on a Get Away Special found to be reduced of about 50z in low gravity; (GAS) payload on the Space Shuttle: pool boiling of analysis of high speed images indicated that the boiling R134a on platinum wires (0.05 mm and 0.2 mm di- crisis was triggered by bubble coalescence along the ameter). A range of heat ‰ux from 50 to 350 kW/m2 wire (Di Marco and Grassi33)). For R?À0.08, the CHF was investigated both in saturated and subcooled con- data obtained in reduced or enhanced gravity lie in the ditions (up to 50 K). For the ˆrst time, a slight enhan- range of most of the other experimental data. Thus the

J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 257 ― 7 Paolo Di MARCO

Bond number (or alternatively its square root, R?,see caused an increase of oscillation wavelength and a Eq. 9) seems to be a suitable parameter to scale the decrease in heat transfer, while the reverse occurred by eŠect of gravity for R?À0.08. The situation appears to applying the electric êld. Beyond an electric êld be drastically diŠerent for R?º0.04: the data obtained threshold, the heat transfer coe‹cient became insensi- in microgravity are very well separated from those of tive to gravity changes, so even in this regime the thinner wires, corresponding to the same value of R?, dominance of electric forces was demonstrated. obtained in normal gravity. This leads to conclude that Recently, Grassi and his coworkers have also per- at low values of the Bond number this parameter is no formed experiments on a ‰at heater 20×20 mm in more suitable to scale critical heat ‰ux, as it could be parabolic ‰ight, imposing an electric êld up to 2 inferred by extending the model of Lienhard and MV/m. Preliminary results (still unpublished) showed Dhir16) to low gravity conditions, and the eŠects of that even in this geometry the electric êld is eŠective gravity and size have to be accounted for separately24). in delaying boiling crisis in low gravity and in reducing Straub3) recently has reported that his own CHF data the detachment diameter and coalescence of bubbles, for wires in parabolic ‰ight and orbital ‰ight were con- thus easing the subsequent condensation of the sistently lower than the corresponding terrestrial produced vapor. values, and could be succesfully correlated adopting a Theresearchcarriedoutsofarworldwideallowedto value of K (see Eq. 8) draw a suitable qualitative picture of boiling in microgravity. Several experimental works lead to con- K＝0.94R?－0.25 (10) clude that the phenomena in the area under the grow- for values of R? as low as 5×10－4. However, no ˆtting ing bubble gives a dominant contribution to the overall of CHF values at diŠerent gravity levels was made. heat transfer in boiling. Sophisticated measurements A cylindrical electrostatic êld around the wire (up with advanced techniques37,38) conˆrmed the above to 10 MV/m at the heater surface) was also imposed speculation. Remarkably, the governing phenomena in by Di Marco and Grassi24) in part of the parabolas. No this region, namely intermolecular forces of adsorp- signiˆcant eŠect was detected on the heat transfer tion, capillary forces, molecular interfacial phase coe‹cient, but the imposition of an electric êld was change resistance and change of phase equilibrium, are found to be eŠective in drastically reducing bubble size independent of gravity. This was experimentally con- and increasing CHF also in microgravity. At high ˆrmed in parabolic ‰ight experiments by Kim et al.25), values of applied voltage, the same value of CHF as in who indicated that what they call ``small scale'' boil- terrestrial conditions was measured, thus demonstrat- ing, i.e. heat transfer in the zones where only small, ing the dominance of the electric force on buoyancy in uncoalesced bubbles exist, is independent of gravity these conditions. However, existing models of CHF in level and subcooling. Straub29,3) also distinguishes the presence of electric êld revealed their limitation boiling mechanisms in primary ones, which are in- when applied in reduced gravity33).Itisalsoimportant dependent of gravity and determined by evaporation to point out that the detachment of bubbles from the and capillary forces in what he calls the microwedge heated surface took place in both the presence and ab- underneath the bubble, and secondary mechanisms, sence of an electric êld, the diŠerence being that in which are responsible for mass and energy transport the former case the bubbles slowed down and stopped away from the surface: they are buoyancy, coalescence at a small distance from the surface and started to processes, momentum transfer due to bubble growth coalesce. The experiments were repeated on a sound- and formation, and thermocapillary ‰ow for sub- ing rocket ‰ight, MASER–834) conˆrming the same cooled states. Gravity thus plays a role in the macro- resultseveninhighmicro-g(about10－5 G) conditions. scale, for convective removal of energy away of the Pool ˆlm boiling on wires was tested by Straub and layer, and can largely be replaced by the other secon- coworkers3 in parabolic and orbital ‰ights. A depen- dary mechanisms. Secondary mechanisms in‰uence n dence of the heat transfer e‹ciency a? on (/0) ,with number and size of the bubbles residing close to the －2 n＝0.16–0.33, valid for /0À10 , was evidenced; surface: when the vapor fraction is larger, vapor may this is in agreement with heat transfer correlations, like adhere to the heated surface leading to substantial heat 35) Bromley's . For lower values of /0, like in orbital transfer degradation and eventually to CHF. Gravity ‰ights, a? is higher than predicted by the relationship may aŠect the two preceding factors, in a way which above and ranges between 0.46 and 0.25. In subcooled depends on subcooling and heater geometry, and this ˆlm boiling, heat is removed by the ˆlm surface by may explain why both enhancement and degradation thermocapillary convection. Di Marco et al.36) per- of heat transfer in microgravity were reported. As a formed experiences of pool ˆlm boiling of FC72 on a consequence, a correct prediction of the nucleation site wire in parabolic ‰ight, in the presence of an electric density and of the bubble detachment diameter, as- êld or less. The variation of the oscillation sociated to a gravity-independent bubble growth wavelength of the ˆlm, due to gravity and electric êld, model (e.g. Dhir39),BaiandFujita40), Stephan and was evidenced. In particular, a reduction of gravity Hammer41)) may lead to a substantially correct predic-

8 J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 258 ― Review of Reduced Gravity Boiling Heat Transfer: European Research tion of boiling heat transfer in any gravitational êld, time dependent, with development of secondary rolls. at least for low to intermediate heat ‰uxes. However, The periodic 3D spatio-temporal structure of the ther- this is far an easy task to accomplish, and debate still mocapillary rolls was visualized by shadowgraphy and exists about the dominating heat removal mechanism: tracer particles. It was shown that in the present con- as outlined by Stephan42) several theoretical works lead ˆguration under reduced gravity, the stationary ther- to conclude that evaporation in a tiny area under the mocapillary roll developed down to the bottom of the bubble where the liquid-vapor interface approaches test chamber, due to the absence of the counteracting the wall (the so-called micro-region) gives the eŠect of buoyancy, and the convective heat transfer in dominant contribution to the overall heat transfer in the ‰uid was increased, while in normal gravity (for boiling (e.g. Wayner43); Stephan and Hammer41), the same bubble size and thermal gradient) the contri- Straub3)). Conversely, according to Demiray and bution of thermocapillary convection was very small, Kim44), although microlayer evaporation plays a role due to the presence of secondary rolls. in bubble heat transport, the heat transfer occurs Betz and Straub48) have set up recently a model for mainly through transient conduction and microcon- the study of Marangoni convection around a gas bub- vection during liquid rewetting after bubble departure. ble either in microgravity or not, conˆrming that a 3.2 Ancillary experiments in reduced gravity for boil- strong enhancement of single-phase convection ing heat transfer around the bubble can be obtained in microgravity, if As already said, boiling heat transfer involves a the cell size is comparable with the bubble diameter. blend of a large amount of physical mechanisms. The The model compared well with experimental results individual understanding of these mechanisms of and the relevant dimensionless parameters were identi- course helps the comprehension of the whole êd. The authors appropriately point out that Maran- phenomenon. As this review is focused on boiling, goni convection is strongly in‰uenced by surface con- these experiments are deˆned as `àncillary'' here, even tamination, so the purity of the ‰uid or its degradation though each of them has of course its own role and sig- in time may substantially aŠect the experimental be- niˆcance. In particular, experiences concerning ther- havior of the system. mocapillary convection (a form of Marangoni convec- Studies on bubble dynamics in reduced gravity are tion) and bubble dynamics will be outlined herein. important to understand bubble behavior, and in par- The role of thermocapillary convection in boiling ticular to asses the role of dynamic forces in bubble has still to be completely assessed, and it is expected evolution in the absence of thermal ones. that this mechanism becomes more important in the Pamperin and Rath49) performed experiments of in- absence of buoyancy. Straub points out that ther- jection of air bubbles in water with injection oriˆces of mocapillary ‰ow around the bubbles was observed in 0.39 and 0.80 mm diameter in the droptower of reduced gravity in subcooled boiling, but never in satu- ZARM, Germany. By assuming that the bubble rated conditions: the origin of this ‰ow has still to be detachment in microgravity is ruled uniquely by a understood, and Marek and Straub45) suggestthatitis balance between surface tension at the bubble neck originated by the gradient of dissolved gas concentra- and inlet gas momentum forces, they developed a tion along the bubble interface. Straub46) also stresses detachment criterion based on a modiêd Weber num- that quantitative measurements indicated that the con- ber tribution of thermocapillary convection to overall r u 2d boiling heat transfer is small, however he gives large We*＝  o o＝8(11) s evidence that thermocapillary ‰ow is an important mechanism for transporting energy away from the li- They claimed that their experimental data are in agree- quid-vapor interface into the bulk ‰uid (and not from ment with the derived criterion, as they observed the heated surface to the ‰uid). In the absence of detachment only for We*À10 during the 4.74 s of buoyancy and subcooling, like in reduced gravity satu- microgravity available in the tests. rated boiling, the energy transport into the bulk of the Di Marco et al.50) investigated the process of nitro- ‰uid uniquely relies on bubble motion. gen bubble formation injected in FC72 from an oriˆce Reynard et al.47) experimentally studied ther- of 0.1 mm diameter drilled in a horizontal tube. The mocapillary convection around a single air bubble in- experiment was operated in the Japanese dropshaft of troduced under a heated horizontal wall in a silicone JAMIC, at a level of about 10－4 G. It was also possi- oil layer. Experiments were performed under normal ble to apply a non-uniform electric êld around the gravity and microgravity conditions, in parabolic tube, thus obtaining a geometrical conˆguration simi- ‰ight. For liquids with a low Prandtl number, diŠerent lar to the one adopted by the same authors for boiling states of thermocapillary convection exist. When a experiments in reduced gravity: this allowed to study critical threshold is exceeded, an oscillatory state fol- dynamical and electrical eŠects separately from ther- lows the steady one and the initially axisymmetrical mal ones. Bubble size, detachment frequency and ‰ow, in the shape of a roll around the bubble, becomes velocity were measured by digital processing of high-

J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 259 ― 9 Paolo Di MARCO speed images. The results showed that, in microgravity which will be housed in the European Columbus and in the absence of electric êld, bubble detachment laboratory of the International Space Station (ISS). did not take place at low gas ‰ow rate at least through- FSL is a multi-user facility for conducting ‰uid physics out the duration of the drop; a bubble diameter up to 5 research in microgravity conditions56).Itcanbeoper- mm was obtained. Conversely at higher gas ‰ow, the ated in fully- or in semi-automatic mode and can be dynamical eŠects were su‹cient to induce bubble controlled on-board by the ISS astronauts, or from the departure. The value of detachment diameter was low- ground in the so-called telescience mode. Each experi- er than predicted by available theoretical models and ment (or experiment category) will be integrated in an the role of surface dynamics in promoting detachment individually-developed Experiment Container (EC), was evidenced in the high-speed images. For this well- which will be separately transported on the Space wetting ‰uid, the detachment occurred for values of Shuttle within the Multi-Purpose Logistics Module We* far lower than prediction of Eq. (11); this was (MPLM) and stored on board. With a typical mass of also conˆrmed in the experiences of Herman et al.51). 25–30 (max. 40) kg, and standard dimensions of 400× The application of electric êld proved eŠective in 270×280 mm3, the EC provides space to accommo- providing a force able to remove the bubbles away date the ‰uid cell assembly, including any necessary from the oriˆce and in promoting bubble departure at process stimuli and dedicated electronics. A very com- diameter values greater than, but of the same order of plete set of diagnostic instruments is integrated within magnitude, as in normal gravity. For the higher values FSL. It includes a set of cameras oŠering high-speed, of the tested electric êld, the detachment diameter high-resolution, infrared, and color recording, illumi- was almost the same as in normal gravity. In this way, nation with either white or monochromatic light the eŠectiveness of the electric forces in promoting sources, particle image velocimetry (including imaging bubble detachment and their progressive dominance of liquid crystal tracers for simultaneous velocity and over buoyancy force was experimentally demonstrat- temperature mapping), thermographic mapping, inter- ed. ferometric observation along two perpendicular axes utilizing a combination of state-of-the-art convertible 4. Forthcoming European Activities: Experiments interferometers. The Microgravity Vibration Isolation and Facilities Subsystem (MVIS), developed by the Canadian Space 4.1 European Experimental Facilities Agency, will be implemented in the facility to isolate Several European facilities are available for –via magnetic levitation– the experiment from space microgravity experimentation in ‰uid physics. Among station g-jitter perturbations. them, the A–300 ZERO–G52),thelargestaircraftever 4.2 Research Programs used in parabolic ‰ights, with a test cavity of 20×5× The CIMEX (Convection and Interfacial Mass Ex- 2.3 m3, diŠerent models of sounding rockets change) research program57) aims to investigate (MASER, TEXUS, MAXUS53)) which can attain over processes involving mass transfer through interfaces, 750 s of microgravity with MAXUS, and the droptow- and their coupling with surface-tension-driven ‰ows er of ZARM in Bremen, Germany54),which,afterthe and instabilities. During the ˆrst two-year phase of this closure of the Japanese dropshaft of JAMIC, current- MAP (Microgravity Application Promotion) project, ly holds the record of the longest available micro-g funded by ESA, four experiments are being prepared time (9.5 s) in drop facilities. for subsequent ‰ight onboard the International Space Concerning the orbital facilities, the Fluid Physics Station (ISS), using the Fluid Science Laboratory Facility (FluidPac)55),tobeinstalledonboardofthe (FSL). The main focus is on ‰ows and instabilities with retrievable Russian capsule FOTON, is a multi-user evaporation, and though only CIMEX–3 involves boil- platform, originally conceived for observation of sur- ing directly, all the experiments are expected to pro- face tension and thermal phenomena along a gas-liq- vide useful information on boiling phenomena. Both uid interface, which can be used also for diŠerent ‰uid single component and multi-component ‰uid systems physics experiments, including boiling. The platform are studied, in collaboration among several European hosts 13 diŠerent optical diagnostic tools; up to 4 ex- teams, an industrial partner, and with the advice of periment containers can be located on its caroussel and non-EU collaborators. On a fundamental point of operated one at a time. Data collected in a 15-day or- view, progress is expected in the understanding of bital mission can be stored on board and partly trans- diŠerent regimes of interfacial mass transfer processes, mitted to ground via telemetry. A new launch of Fuid- in the presence of several eŠects (inert gas, Marangoni Pac, in FOTON–M2 mission, is foreseen in 2005: it convection, micro-regions or triple lines, surfactants). will carry four diŠerent ‰uid-physics experiments, and On the applied point of view, both direct and prospec- among them ARIEL, to study pool boiling in the tive researches are conducted, aiming to optimize heat presence of electric êlds. pipes, thin-ˆlm evaporators, two-phase ‰ow and The most promising European facility for two-phase boiling technologies in normal and reduced gravity. In experiments is the Fluid Science Laboratory (FSL), particular, CIMEX–1 studies evaporative convection

10 J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 260 ― Review of Reduced Gravity Boiling Heat Transfer: European Research in liquid layers, and CIMEX–2 is devoted to the coe‹cients for saturated and subcooled nucleate physical understanding of capillary phenomena in pipe pool boiling (up to medium heat ‰uxes) do not grooves, for optimization of heat pipe performance diminish appreciably with gravity reduction, contra- using advanced structures. In CIMEX–3, a versatile ry to the prediction of most heat transfer correla- loop system will be developed for the study of micro- tions. Occasionally, even enhancement up to 30z at gravity two-phase ‰ows and heat transfer issues. This low heat ‰uxes was observed. part of the program foresees the development of A su‹cient evidence was gained that the phenome- transparent swirl evaporators and a high-e‹ciency, na taking place in the micro-region under the bub- low-pressure-drop condenser, to be used in a ble, which is the major contributor to boiling heat mechanically-pumped loop, the characterization of transfer rate, are independent of gravity accelera- ‰ow patterns and void fraction in reduced gravity by tion; gravity rules the mechanisms of heat and vapor using diŠerent working ‰uids: the results may prove removal from the surface in terrestrial gravity, the viability of mechanically pumped two-phase loops aŠecting free convection and bubble-departure size and will be useful and for thermal-gravitational and velocity, but may be replaced by diŠerent scaling1).InCIMEX–458), thermocapillary ‰ows mechanisms in micro-g. This may expain why boil- around vapor bubbles will be investigated using optical ing is less sensitive than expected to gravity accelera- and thermal diagnostics; parabolic ‰ight precursory tion. experiments have been already performed, as previ- In microgravity, diŠerent pool boiling regimes ously outlined47). were identiˆed with increasing subcooling: ESA supports the creation of European networks, saturated/slightly subcooled boiling with bubble called `Topical Teams' (TTs), to consolidate, optimize departure and coalescence, subcooled boiling with and promote plans for research programs in the space no bubble departure and thermocapillary convec- environment in anticipation of future announcements, tion, highly subcooled boiling with rapidly growing also involving conventional industry for transfer of and collapsing bubbles acting as micro-pumps. space technology. The TTs are composed of European Critical heat ‰ux is reduced in low gravity, scientists, but they are open to the (unsupported) however it is higher than predicted by the extrapola- cooperation of non-EU members. At present, a TT on tion of correlations well assessed on earth. The scal- Boiling and Two-Phase Flow is active. This TT is cur- ing of CHF in microgravity based on Bond number rently deˆning the requirements for future experiments revealed unsatisfactory so far, and a separate depen- of boiling, bubble dynamics and two-phase ‰ow in dence on gravity and heater size seems to exist. microgravity, aimed to the realization of one or more The experiments in ‰ow boiling are still insu‹cient ECs for FSL. to elaborate ‰ow maps or to identify the minimum As reported by Delil (2003), European teams are ‰ow above which the role of gravity is still sig- currently involved in the development of two near-fu- niˆcant. ture mechanically pumped, two-phase heat transport The research so far conducted worldwide allowed to systems applications in space, namely: gain su‹cient qualitative knowledge about boiling The two-phase ammonia thermal control system heat transfer process in microgravity. Some of this of the Russian segment of ISS59,60), knowledge was also useful for clarifying the fun-

The hybrid two-phase CO2 loop of the Tracker damental mechanisms of boiling. It has now become Thermal Control System of Alpha Magnetic Spec- quite clear that e‹cient boiling can be sustained in trometer (AMS–2)61), an international experiment microgravity, especially in subcooled conditions, and searching for anti-matter, dark and missing matter, that vapor removal from the proximity of the heated planned for a 5-years mission as attached payload surface is the key problem to be solved to avoid early on ISS. degradation of heat transfer performance. However, we are still far from the elaboration of quantitative 5. Conclusions models, for which extensive experimentation is still re- In this paper, the fundamental characteristics of quired. boiling heat transfer and its advantages for high 6. Nomenclature demanding heat removal applications were outlined. The most signiˆcant European research in the ˆeld was A area (m2) recapitulated, while the activities carried on in US and a thermal diŠusivity (m2/s)

Japan were described in detail in companion papers. Bo Bond number, l/lL The main results achieved so far can be summarized C generic constant as follows do oriˆce diameter (m) Although bubble dynamics and transport mechan- D diameter (m) isms are diŠerent in micro-g, bubbles detach equally  actual vertical acceleration (m/s2)

from the heated surface and the heat transfer hf saturation enthalpy ( J/kg)

J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 261 ― 11 Paolo Di MARCO k thermal conductivity (W/mK) 6) V. K. Dhir: in ``Handbook of phase change–boiling and con- K Kutatelatze constant, (see Eq. 8) densation'', ed. by Kandlikar S., Shoji M., Dhir V. K., Taylor l, L length (m) and Francis, 1999, sec. 4.4, p. 86. 7) D. Motoya, I. Haze and M. Osakabe: Proc of ASME HTD, s 364–1 (1999) 303. lL Laplace length, (m) 8) Zhang N.: Int. Comm. Heat Mass Transfer, 26, 8, (1999) 1081. (rl－r) 9) M. G. Cooper: Saturation nucleate pool boiling–a simple cor- M molecular weight (kg/kmol) relation, IchemE Symp. Ser., 86 (1984) 786. n power law exponent, (see Eq. 2) 10) V. V. Yagov: Heat transfer with developed nucleate boiling of liquids, Thermal Engineering, 35 (1988) 65. Nu Nusselt number 11) K. Cornwell and S. D. Houston: International Journal of Heat p pressure (Pa) and Mass Transfer, 37 SUPPL 1 (1994) 303. Pr Prandtl number 12) VDI Heat Atlas, ed. Verein Deutscher Ingenieure, (English q! heat ‰ux (W/m2) version), VDI–Verlag, Dusseldorf, D, 1993. th Q heat rate (W) 13) P. Di Marco and W. Grassi: Proc. of the 5 ASME/JSME Joint Thermal Engineering Conference, San Diego, CA, USA, r radius (m) March 1999, paper AJTE99/6275. R? dimensionless radius 14) W. Grassi: Proc. 40th ATI Heat Transfer Nat. Conf., Trieste, Rp surface roughness (mm) I , June 1985, pV33. T temperature (K) 15) N. Zuber: Atomic Energy Commission report AECU–4439, 1959. uo gas velocity at oriˆce (m/s) * 16) J. H. Lienhard and V. K. Dhir: J. Heat Transfer, Trans. We modiêd Weber number, (see Eq. 11) ASME, 97 (1973) 152. a heat transfer coe‹cient (W/m2K) 17) K.H.SunandJ.H.Lienhard:Int.J.HeatMassTransfer,13 a? heat transfer e‹ciency, (see Eq. 2) (1970) 1425. 18) N.BakhruandJ.H.Lienhard:Int.J.HeatMassTransfer,15, DTsat wall superheat, Tp–Tsat (K) h dynamic viscosity (Pa s) 3 (1972) 2011. 19) J.M.Delhaye,M.GiotandM.J.Rietmuller(eds.):Thermo- 3 r density (kg/m ) hydraulics of Two-Phase Systems for Industrial Design and s surface tension (N/m) Nuclear Engineering, Hemisphere Pub. Corp., NY, 1980. F contact angle (rad) 20) J.G.CollierandJ.R.Thome:ConvectiveBoilingandCon- Subscripts densation, 3rd ed., Oxford Univ. Press, 1996. 0 referred to terrestrial gravity 21) R. Siegel and C. Usiskin: J. Heat Transfer, Trans. ASME, 81 (1959) 245. crit critical 22) C. Usiskin and R. Siegel: J. Heat Transfer, Trans. ASME, 83 CHF critical heat ‰ux (1961) 243. d detachment 23) R. Siegel and J. R. Howell: Critical Heat Flux for Saturated  gas Pool Boiling from Horizontal and Vertical Wires in Reduced l liquid Gravity, NASA Tech. Note TND–3123, 1965. 24) P. Di Marco and W. Grassi: Int. J. Th. Sciences, 41, 7 (2002) ref reference 567. sat saturation 25) J.Kim,J.F.BentonandD.Wisniewski:Int.J.HeatMass w heated wall Transfer, 45 (2002) 3921. 26) H. S. Lee and H. Merte: Heat Transfer 1998, Proc. of 11th Int. 7. Acknowledgements Heat Transfer Conference, p. 395. 27) T. Oka, Y. Abe, K. Tanaka, Y. H. Mori, H. Yasuhiko and A. The author is strongly indebted to all the persons Nagashima: JSME International Journal, Series 2, 35, 2 (students, researchers, professionals, ESA personnel) (1992) 280. who assisted and steered him in carrying on his activity 28) M. Zell, J. Straub and A. Weinzierl: 5th European Symp. on in the êld of microgravity boiling for more than ten Material Science under Microgravity, Schloss Elmau, 1984, p. years, and in particular to Prof. Walter Grassi, who 327. 29) J. Straub: Proc. 3rd Int. Symp. on Heat Transfer and Trans- introduced him in the êld of microgravity research port Phenomena, Beijing, 1992, p. 16. and originally conceived the European research line 30) J. Straub and S. Micko: Proc. EUROTHERM Seminar n.48, related to the application of electric êlds to boiling Pool Boiling 2, Goren‰o D., Kenning D. B. R., Marvillet C. heat transfer. Eds., Paderborn, D, Sept. 1996, p. 275. 31) M. Steinbichler, S. Micko and J. Straub: Heat Transfer 1998, References Proc. of 11th Int. Heat Transfer Conference, 2, Seoul, Korea, Aug. 1998, p. 539. 1) A. A. M. Delil: AIP Conference Proceedings, 654 (2003) 3. 32) J. Straub, G. Picker, M. Steinbichler, J. Winter and M. Zell: 2) S. Nukiyama: Trans: JSME, 37 (1934) 357 (reprinted in Int J. Proc. EUROTHERM Seminar n.48, Pool Boiling 2, Goren‰o Heat Mass Transfer, 9, p. 1419). D., Kenning D. B. R., Marvillet C. Eds., Paderborn, D, Sept. 3) J. Straub: Boiling Heat Transfer and Bubble Dynamics in 1996, p. 265. Microgravity, Adv. Heat Transfer, 35 (2001) 57. 33) P.DiMarcoandW.Grassi:Proc.12thInternationalHeat 4) J. Straub, M. Zell and B. Vogel: Proc. 9th Int. Heat Transfer Transfer Conference, Grenoble, F, August 18–23, 2002, 3, p. Conference, Jerusalem, Israel, Aug. 1990, p. 91. 617. 5) W. M. Rohsenow: J. Heat Transfer, Trans. ASME, 74 (1952) 34) P. Di Marco and W. Grassi: AIP Conference Proceedings, 608 969. (2002) 172.

12 J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 262 ― Review of Reduced Gravity Boiling Heat Transfer: European Research

35) A. L. Bromley: Ind. Eng. Chem., 44 (1952) 2966. 53) http://www.ssc.se/esrange/ 36) P. Di Marco, W. Grassi and F. Trentavizi: Ann. N.Y. Acad. 54) http://www.zarm.uni-bremen.de/ Sci., 974 (2002) 428. 55) P. Baglioni, R. Demets and A. Verga: 51st International As- 37) S. Bae, M. Kim and J. Kim: Exp. Heat Transfer, 12 (1999) 265. tronautical Congress, Oct. 2000, Rio de Janeiro, Brazil, paper 38) P. Stephan, C. Hohmann and J. Kern: AIP Conference IAF–00–J.3.01 Proceedings, 608 (2002) 163. 56) http://www.space‰ight.esa.int/users/ 39) V. K. Dhir: AIChe J., 47 (2001) 813. 57) J. C. Legros, P. Colinet, L. Joannes, P. Stephan, G. Bekaert, 40) Q. Bai and Y. Fujita: Thermal Science and Engineering, 7 G. Lebon and P. Cerisier: Proc. 1st Int. Symp. on Microgravi- (1999) 45. ty Research and Applications in Physical Sciences and 41) P. Stephan and J. Hammer: Heat and Mass Transfer, 30 Biotechnology, Sorrento, Italy, September 10–15, 2000, p53. (1994) 119. 58) P. Arlabosse, N. Lock, M. Medale and M. Jaeger: Phys. 42) P. Stephan and J. Kern: Proc. of the 5th Int. Conf. on Boling Fluids 11 (1999) 18. Heat Transfer, Montego Bay, Jamaica, May 4–8, 2003. 59) Y. I. Grigoriev, E. I. Grigorov, V. M.Cykhotsky, Y. M. Prok- 43) P.C.Wayner,Y.K.KaoandL.V.LaCroix:Int.J.HeatMass horov, G. A. Gorbenko, V. N. Blinkov, I. E. Teniakov and C. Transfer, 19 (1976) 487. A. Malukhin: AIChE Symp. Series–Heat Transfer n. 310, 44) F. Demiray and J. Kim: Proc. AIAA/ASME Thermophysics (1996), Ed. M.S. El–Genk, p. 9. and Heat Transfer Conference, St. Louis, MO. June 24–26, 60) E. K. Ungar and T. D. Mai: AIChE Symp. Series–Heat Trans- 2002. fer n. 310, (1996), Ed. M.S. El–Genk, p. 19. 45) R. Marek and J. Straub: Int. J. Heat Mass Transfer, 44 (2001) 61) A.A.M.Delil:Proc.12thInt.HeatPipeConf,Moscow,Rus- 619. sia, 2002. 46) J. Straub: Ann. N.Y. Acad. Sci., 974 (2002) 348. 62) M. Zell and J. Straub: 6th European Symp. Material Sciences 47) C. Reynard, R. Santini and L. Tadrist: Proc. of 12th Int. Heat under microgravity conditions, Bordeaux, 1986, p. 155. Transfer Conference, Grenoble, F, Aug. 18–23, 2002, vol. 3, 63) J. Straub, M. Zell and B. Vogel: Proc. 1st Europ. Symposium p. 467. Fluids in Space, Ajaccio, France, Nov. 1992, ESA SP–353, p. 48) J. Betz and J. Straub: Ann. N.Y. Acad. Sci., 974 (2002) 220. 269 49) O.PamperinandH.J.Rath:Chem.Eng.Sci.,50, 19 (1995) 64) G. Picker and J. Straub: Heat Transfer 1998, Proc. of 11th Int. 3009. Heat Transfer Conference, Seoul, Korea, ed. by J.S. Lee, v.2, 50) P. Di Marco, W. Grassi, S. Hosokawa, G. Memoli, T. 1998, p. 109. Takamasa and A. Tomiyama: Int. J. Multiph. Flow 29, 4 65) G. Picker and J. Straub: Microgravity Fluid Physics and Heat (2003) 559. Transfer, Kahuku, Hawaii, USA, Sept. 19–24, 1999. 51) C. Herman, E. Iacona, I.B. Foldes, G. Suner and C. Milburn: 66) J. Straub: Int. J. Therm. Sci., 39 (2000) 490. Experiments in Fluids, 32 (2002) 396. 52) http://www.novespace.fr/ (Received Aug. 21, 2003)

J. Jpn. Soc. Microgravity Appl. Vol. 20 No. 4 2003 ― 263 ― 13