/v, cj/,

s ZS .. zo712 6

......--

Electromagnetic Tool
Julian

Levitation--A Research/
_-7 and Robert Hyers

Useful
/ _C'__3

-t,

in Microgravity
Elliot Schwartz,

Szekely,

t_

Electromagnetic levitation is one area of the electromagnetic processmg of materials that has t_ses for both fundamental research and practical apptications. This technique was successfully used on the Space Shuttle Columbia during the Spacelab IML-2 mission m July 1994 as a platform for accurately measuring the surface tensions of liquid metals and alloys. In this article, we discuss the key transport phenomena associated with electromagnetic levitation, the .fundamental relationships associated with thermophysical property measurement that can be made using this technique, reasons for working in microgravity, and some of the results ohtamed from the microgravi_ experiments. INTRODUCTION Electromagnetic levitation, the positioning of conducting materials by electromagnetic forces, has been practiced by metallurgists and materials scientists for several decades.t During the past 1015 years, major advances have been made in our quantitative understanding of the underlying phenomena and, partly as a result, many important practical applications of the electromagnetic levitation principle have evolved. The most important of these are the melting of titanium in induction furnaces and the control of the metal-mold contact in the CREM (name derived from cast-refineelectromagnetic) continuous casting process. zJ These are quite well documented in the general literature and in the specific symposia dedicated to the electromagnetic processing of materials. This article presents a brief review of a rather specific application of the electromagnetic levitation (EML) principle---

microgravity research. There are several motivations for such an undertaking, including the fact that the July 1994 Spacelab [ML-2 (the Second International Microgravity Laboratory) mission, which had an electromagnetic levitation faciliW on board, has yielded very interesting new results. Furthermore, microgravi W experiments require precise planning, because of the great expense and the very long lead times involved. Thus, EML studies necessarily require sophisticated mathematical modeling efforts during the planning the experiments. and evaluation OF of

equations,

which

are given

b_ (I)
r2C

V E=0

_- = 0

f2)

_O =0

(3)

3I" CC

SL

where fields netic

F,, B,

I2I, and

J are

the vector

h, g, b'

for electric flux density,

field intensitv, magmagnetic field in.

THE PRINCIPLE ELECTROMAGNETIC LEVITATION The tation shows bound cally

tensity, and current density, respectively. The time-averaged electromagnetic force and power absorbed per unit volume in the sample are calculated according to the formulae _ = 1 fi x Bdt (5)

principle of electromagnetic leviis illustrated in Figure 1, which the coil arrangement in an earthfacility. A high-frequency (typi100--800 kHz) current is passed

Jo

through conically wound coils, generating an electromagnetic field (also sketched in the figure). If a metaUic specimen is placed between these coils, eddy currents will be induced in the specimen with the following results: A lifting force will be exerted on the specimen due to the coupling between the induced current and the applied electromagnetic field. If this force is sufficiently strong, it can levitate the specimen. The electromagnetic the specimen induce forces within fldid flow in a

where z is the period of the applied alternatingcurrent and (y,_ is the electrical conductivity of the specimen. The governing equations for the electromagneticailydriven fluidflow inside the molten specimen are the NavierStokes equations,which express theconservation of momentum in the fluid. Using art effective viscosity t.t_ that is uniform throughout the fuid, the Navier-Stokes equations, e_pressed as a single vector equation, are given by _

molten specimen, which is turbulent under earthbound donditions. The induced current provides Joule -Vp+g._Vz_ +pg+F

-12 20 10

,.-8

12

heating through Ohmic losses, which may melt, or ev_dn superheat, the specimen. ": As a practical matter tinder earthbound conditions,the abilityto levitate is confined to metallic specimens of moderate size (about 10--15 mm in diameter) for conventional coil designs and moderate to high melting temperatures (above 1,000C). More sophisticatedcoil designs have allowed the levitationof rather largerspecimens. Gamier has re-

:,,,_

where fi is the velocity vector field, which representsthe internalfluidflow in the specimen, p isthe pressure scalar field, _ isthe gravitational acceleration vector,and p is the density of the fluid. The effective viscosity is equal to the sum of the laminar (molecular) viscosity _t and a turbulent viscosity _. The equation of continuity, the statement of mass conservation, is given for the incompressible fluid by _

9t.

-20

-0.09 --0.06 .-0.03

0.113 0.06 0.09

Ma_e_c Field(TesJa) Figure 1. Schematic sketch of conical coil arrangement with field strength and gradient along symmetry axis for peak applied current of 405 A. Figure prov_ed by I. Egry et al.

ported levitating copper samples 50 mm in diameter and Japanese researchers have reported the levitationof 20 kg titanium specimens. The electromagnetic phenomena are governed by the well-known Maxwell's

_._=0

(8)

For axisymmetric geometries such as the specimens processed in levitation facilities, the free surface shape at any

50

JOM

* May

1995

|B

I'
::n',e :_ ,ier:ned by the bounda.,v condl:ion, _hichisa baianceotcomvonentsot _tress normal to the tree surface" ' l p - p, = "/' .... ' R. 1 R. 2u., ' Ou an _9) :hermoFh_,-_cai surface tension, cosiW. metals cooled property data heat capaow, _-uch as and wsobserved. The restorm "orce :or __ur:ace The property, database for liqmd in the superheated and understates is incomplete, and the conoscillations ot the sphencai sample or mass m is the surface tension 7, which can be related to the frequency of the oscillations v bv Rayleigh's formula:" 3 7 = - rtmv: 8 112)

where p, is the atmospheric pressure, R. and l-C are the principal radii of curvature, and u Ls the component ot velocity in the direction normal to the free surface. The temperature distribution speclmen is governed by the energy balance equation, given pC. ,! -_-63T "VTj=kV:TQ in the thermal by' (t0)

ventional measurement techniques are not entirelv satisfactory, providing motivation for the develol_ment and execution of the microgravity BRIEF experiments. :3 HISTORY

l) 2) 3) ,4)
tor

in!ly. "tic olrd-

where T is the temperature, and Cpand k are the heat capacity and the thermal conductivity of the specimen. The boundary condition at the free surface of the specimen, which expresses heat losses by conduction to the cooling gas, if used, and by radiation, is given by L -k0T=(ho_a+h,aXT-T:) On (ll)

In order to improve upon the accuracy of existing property, data and work with high-temperature, corrosive metals, researchers have developed methods to measure properties in a containerless mode. A brief, and by no means complete, historical background of thermophysical property, measurements using electromagnetic levitation is presented below. In 1971, Fraser et al. a developed a method to measure the surface tension of liquid metals in a containerless fashion, the oscillating drop technique using electromagnetic levitation. Further developments were made by Soda et al., _ Keene et al., "6 and Egry et al. z7 In this technique, the oscillations of a levitated droplet about its equilibrium shape are

The method has the advantage of eliminating persistent sources of contamination that arise through the use of substrates and / or capillary, tubes associa ted with conventional methods such as the sessile drop, the capillary, rise, and the maximum bubble pressure methods. :" It avoids any contact with a crucible and thus reduces not only svsternatic errors due to surface contamination but also allows deep metal. 30 undercooling of the liquid

Fecht and Johnson Jt developed a noncontact method for measuring heat capacity based on the temperature response of a specimen exposed to a sinusoidally modulated heat source. In an electromagnetic levitation facility, the heat capacity C_ can be determined from the P modulation frequency c0=, the amplitude of the temperature modulation (aT.), the increase in the average sample

where h o_a is the coefficient of heat transfer by conduction to the stagnant inert gas, h, is the radiation heat transfer coefficient, and T is the temperature of the water-cooled coils, which serve as the heat sink. in order to rationally plan the microgravity experiments, we used numerical techniques to calculate Electromagnetic lifting and squeezing forces Induced power Melt circulation Free surface deformation Heat transfer The calculations were verified experimentally in earthbound levitation facilities wherever possible. The calculated melt circulation and free surface shape tated in a pure nickel in an earthbound specimen levifacility and the

(6)

ied
ec-

,)

actual shape are illustrated in Figure 2. The capability to calculate free surface shapes was used to select the value of the applied coil current used to deform the molten specimens in the microgravity experiments. Over the years, notable contributions tothe mathematical modeling of electromagnetic levitation have been made by researchersat MADYLAM in Grenoble, France; u,_2 Cambridge University in England;m( DLR inCologne, Germany; ms theUniversity of Alabama; _7*_s Rice University; _9_ and at the Massachusetts Institute of Technology, n= among others. Because of the need to operate in an efficient manner, mathematical models are being increasingly used to predict behavior and intelligently control industrial materialsprocessing operations. The accuracy of theresultsproduced by such models is limited by the accuracy of

Figure 2. Calculated streamline i_ttem (left side), velocity vector field (right side), and equilibrium free surface shape of a nickel drooler levitated in earthbound electromagnetic levitation facility with conical coil arrangement.

1995 May * JOM

51

:emperamre,_T temperature

L and the sampie IT,) by the relanonship 1 _T,,. oo_,T,. _ T O area

bias :z

tension, cosltv,

heat capac:tv because much

and Fossiblv v:__weaker electro-

etIects

Or

fiu:d

:low

op,

p_'.icie_l[It)2

h,,-,i

Cp = 4-,,2AscT_ where

(13) and

A and are the surface

the emissivity of the specimen, respectively, and c is the Stefan-Boltzmann constant. Lamb s-' suggested that the viscosity, of a spherical droplet of radius R and density p could be related to the oscillation damping constant F bv the formula = -1_ R20F
D

magnetic fields are necessaD' to position the metallic specimens than are required in earthbound experiments. Th!s offers a direct benefit to surface tension measurements because the apparent increase in the surface tension value due to the magnetic field in earthbound ments is thereby eliminated. The noncontact calorimetry used to measure heat capacity, experimethod requires heat levi-

growth. Microgravlty makes it ,heoret. catlv possible to studv the metastabie states of materials obtained through deep undercooling. EXPERIENCE AND RESULTS FROM IML-2 The experiments were conducted using TEMPUS (Tiegelfreies Elektromag, netisches Prozessieren Unter Schwerelosigkeit), an electromagnetic containerless processing facility. TEMPUS uses electromagnetic levitation for containerless positioning and heating of metal samples and can be used under microgravity conditions. A schematic sketch of the experimental arrangement of TEMPUS is shown in Figure 3. TEMPUS was designed to process 8 mm and 10 mm diameter spherical metal samples within two sets of induction coils. The outer (positioning) coils, which operate at a frequency of about 140 kHz and create a quadrupole magnetic field, position the sample. The inner (heating) coils, which operate at a frequency, of about 400 kl'-'I:, and create a dipole magnetic field, provide most of the thermal energy to the sample through induction (Joule) heatmgP _PUS was built by Dornier GmbH under contract from D,_RA, the German space work teams agency. A significant for the participating was the determination part of the experiment of the set of
F; m LIt re

that the only method of specimen loss is radiation, but in earthbound tation quired below

(14)

It has not been possible to use the oscillating drop technique using electromagnetic levitation to measure viscosity in earthbound experiments because of the turbulent fluid flow driven by the strong ISSUES electromagnetic forces. IN MICROORAVITY

experiments cooling gas is retO reach specimen temperatures about 1,100C. The reduction in

Electromagnetic levitation provides the means for processing metallic specimens in a containerless mode, which offers the great benefit of avoiding contamination. Work with very reactive or corrosive liquid metals is thereby possible without concern for falsification of results by chemical reaction with the crucible material. Electromagnetic levitation offers the further advantage of eliminating heterogeneous nucleation on container wails, making it possible to work in the metastable regime of undercooled melts. _ Electromagnetic levitation in a microgravity environment is particularly suitable for measurements of thermophysical properties such as the surface

the necessary magnetic field strength in rnicrogravity makes it possible to process under a vacuum and perform containerless heat capacity measurements on deep eutectic alloys with low melting points. Microgravity makes it possible to perform viscosity measurements using electromagnetic levitation by greatly reducing the magnitude of the electromagnetic Lifting forces needed to overcome gravity. The turbulent fluid flow present in earthbound experiments makes the measurement of viscosity impossible because momentum transfer and the decay of the oscillations are governed by the turbulent eddies rather than by the molecular viscosity. 3s Electromagnetic levitation in a microgravity environment is also attractive for studies relating to undercooling, nucleation, and recalescence. Comparison of the results of earthbound experiments with those of microgravity experiments could provide insight into the

B
V_

T F F
S

d
r k. E

parameters used to configure the generic main process flow for the performance of each particular experiment. The parameter set included such variables as the emissivity for temperature measurement, inert gas pressure, heating and positioning coil voltages, heating times, maximum and minimum allowable sample temperatures, and heating coil voltage and duration. This preparation consisted primarily of a substantial mathematical modeling effort and experimentation with the development model at the Microgravity User Support Center (MUSC) in Cologne, Germany. The microgravity experiments took place July 8--July 23, 1994, aboard the Space Shuttle Columbia as part of the Spacelab IML-2 (Second International Microgravity Laboratory) mission. During that time, we monitored ments from the Payload Command Center (POCC) our experiOperations at NASA's in Huntswe teleand coor-

Marshall Space Flight Center ville, Alabama. From there, commanded our experiments sm sw Chambw dinated experimental TEMPUS scientists.

runs with the other experiments

All of our microgravity

Figure 3. Sketch of the TEMPUS experimental arrangement illustrating the mare sut)systems, including the cylindncaJgeometry of the induction coils,

were performed jointly with Ivan Egry and his group from the German Aerospace Research Establishment (DLR) in Cologne, Germany. Prior to the mission, experiments on the noble metals gold and copper, the congruent-melting alloy

52

JOM

May 1995

."-

.*,a:,

'3

,', :r-'_r'Ja.e ' \ "_.!_ _, r'C' ;2-'_;7 ;- \ -_.Ka,_,aan jnd r T Nacara_a_. -_'teini_'aer',.e _e._[n .,n _.|eit :v_. _'':'.'_'d:,:;_

,IL.

u:

---_av, .n E_ec.'Tz:oma_ne':lc Le_,_ta,':on ",le_[" '.i '_: ! '::_''"J cr:a i'"_ J'%l ._tce_ _,,n. and "t' Bavazlto_iu. tar a Comcai Cod. Ei_ctroma_:nenc bfe!allur_:_at Y'J_*acor or

'2

L_

'.4 - B 5athu_a*h F "co Ca_cuianons

:,,_, 4 24,'_, ic,_a3, pp 737-'48 22 ? '. R Sur-,anaravana and 5_anc Derormanonand External L_ :tared Droniets. ?bus Fluuls 2', N E_-Kaddan and I Szekeiv.

_ Ba_azltoglu. Effect Forces on the Oscillations

.4.3 (-'3) _ lc,_1 i. pp ,t_7-a77 The E!ectromagnetlc Force

5
ca ca CD

'0
<> "g-_esuits

F_e_d. _u_d :ate_ Metal pp

F_o_, Field. and Temperature Profdes in LevlDroplets. ' '_ietallur_tcal Fransactlons 8. 14 1")83) B Li. and [ Szekeiv. ','The Electrodynamlcand v Levitate_ Acta _,fol-

44;-410 J-H ZonsL

!%drodvnamtc

Phenomena

in MagnetlcaU Behavior)," or the Sur#act

09 ,I
iJ.

- Ig-Resu_ts Extrapolatedto 0 g

!en ,_h:opiets--I _teadv State 2b ,o_ .1c_2L pp 43-_--44Q 23 E. 5<hwartz. Measurement na__._e:lcailv-L_'ltatea Droplets Deparn'nem ot Matenals Science 24 M E. Fraser et al., 'Surface y_T) : I 1250 1.149-0.14.104 i [ 1300 (%1064) i i 1350 N/m i 1400 Pure nique," 25. H. L_qu_d Iron and Nickel

Astronautica, of Elec:ro-

Tc,tsion

h .f S )
S

0.8

0- lag-Results from IML-2

, i i 1100 i i 1150 I i 1200 i

in MIcrograwt_. PhD thesis, and En_eenng, MFF 19qSL Tenston Measurements on by an Oscdlanng Drop Techof

0.7

Metailurqlcat Transactions. 2 11971 ). pp. 817-.423. Soda. A. McLean. and W A. Miller, "The Influence Amplitude on Llqmd Surface Tension Measure-

Osc_ilat_on

Temperature(:C)
e

ments 'anth Levtta_ed Metal Droplets. ' Metallur_tcat T,'ansac. hefts. 98 (1978), pp. t45-147. 26. 8.l. Keene, The Use era Fourier Analyser (vr Determination or the Surface Tensmn dr Liqu_ .Metals by the Levuatin_ Tec_maue. NPL Report OMA(A} 56, NISL. Tecldington. (1982). Dro_ UK.

f 1
q

Figure 4. Surface tension data for liquid Au obtained from Spacelab IML-2 microgravity experiments, wfth linear fit of the data points. For the purpose of companson, the "1 g-results" are uncorrected data obtained by Saueriand et al. z4in earthbound levitation expenments. The "1 gresults extrapolated to 0 g" are surface tension values obtained using the Cummings and Blackburn correction formula, which accounts for the al_arent increase in the surface tension value caused by the magnetic field. AuCu, and nickel had been planned. The experiments on gold and AuCu were performed successfully based on the preparation before the mission, and a successful experiment was also conducted on a sample of eutectic ZrNi previously used for heat capacity measurements. The experiment was planned dunng the mission based on the experience and teamwork of the groups. Using electromagnetic levitation in microgravity, we measured the surface tensions of a pure metal (Au), a congruent-melting alloy (AuCu), and a eutectic alloy (ZrNi). The surface tension measurements for gold, which were in the temperature range of 1,225-1,330C, are shown in Figure 4. The filled points on the plot are data points obtained from the microgravity experiments. The different symbols represent different experiment cycles. The "1 g-results" indicate uncorrected surface tension values obtained from the earthbound levitation experiments "1 g-results the surface of Sauerland et al., _ and the extrapolated too g" indicate tension valu_ obtained usthe TEMPUS forthcoming team will papers. be presented in

27. [. Egry et aL. "Sm'face Tension Measurements of Liqmd MetaLs Using Levita_on, Micrugrav_w, and Image Processing." lnternatu_nal Iournal of Thermo_ysics, [ 3 (1) (1992). pp. 65-74. 28. J.W S. Rayleigh, "On the Capillary Phenomena of lets.' P re_ee_lrt_$ of the Royal Socle W of London, 29 ( 1879L pp. 7 1-97 29 ]'. Iida and R.LL Guthrie. The Physical Protx'_ttes ot Ltqt_Ld ._e*a/s (Oxlord, 30. R. Wiilnecker, "Contamertess Physics Ijztt_'_, 31 H.J. Fecht England: Clarendon D.M. Heriach. Press. and 1988). B. Feuerbacher,

ACKNOWLEDGEMENTS The authors acknowledge the financial support provided by, NASA under contract numbers NAG8-815, NAG8-970, and NAG8-1078. References
1. W.A. Peeler, "l_,,ntatlon Melting--A Survey, of-the-Art," Journal of Metals, 17 (S) (1965), pp. of the State487-.493.

Processing of Bulk Fe-Ni Melts." .4_l_e_ 49 (20) (1986), pp. 1339-t341. and W.L Johnson, "'A Concep_al Approach

f
_t e

for Noncontact Calorimetry m Space," Rev_ #Sctenttfic Instruments. 62 (5) (1991L p. 1299-1303. 32. R.K. WunderEch e_ al.."Pnnapies of Non-Contac_ A. C. Calonmet_'y," cations, ed. PA: "['MS, W. I993), Contamerless Hofmesster pp. 51-56. of a Viscous Liqmd Globe." Society, 13 (I) (I 881), pp. 51Processmg: Techm4ue_ ana A_'pliand R. Schiffman (Warrendale.

t f

33. H. Lamb, "On the Oscillations Precedings of #u_ London Math. 66. ]4. S. Sauerland, lemtttrten fluesslgen

Messung der Oberflaechens_annun_ an Metalltro'pten. Ph.D. theses, DLR Cologne / Germany Szekely, (1993}. "The Measurement of

2. C. rives and B. Forest, "CRF.M: A New Casting Process Part I--Fundamental Aspect," Light Metals 1987, ed. R. Zabs_mik (Wammdale, PA: _, 19873, pp. 769.-778. 3. I.P. Riquet and I.L Part II--[ndugtrial Zabrezn_ (W_dale, Meye, "CREM: A New Casting Process Aspects," Light Metals Z987. ed. R. PA: "['MS, 1987), pp. 779-784. Mas"net_r_gnarm_ PA: TMS, 1992). Processing o_ Matenals m Process

Umvev_|t'y of Aachen, 3S. I. Egry and J.

Thermophysical Properties magnenc Levitation," Adv. 266.

in Microgravtty Using Electr_ Space Res., I 1 (7_ (l 991 ). pp. 263the OscillaPrec,,stag: and R. of of Ffutd

4. J. Sz_a_ty _ al., ed., Metallurgy (WarmndaJe, 5. S. A.saL ed.. Electrmnagnttu: [ap,l_ ISl], tgq4). 6. L-H. Zon$, Expm'unents sorption m

36. S, Sauerland et al., "Magnetic Field E.f fec_ on tion Spectrum of Levitated Dr'opa." Contalnerless Teckn_ues and Applicatwns, ed. W. Hofmeister 37. D.L (Warrendale, Cummings 224 and PA: TMS, 1993L D.A. Blackburn, pp.

(Nagoya,

65--o9. "Oscillatmns

I. Szekely, and G. Lohoefer, "C.a]cuJatmns and Concernmg the Lifting Force and Powe" Abthe TEMPUS EML," Acre Astronaut_a. 29 (5) Computat_n of the glecm3mag-

M agn_ically Methanes,

L_-,vitated AsphencaJ (1991), pp. 395--416.

Droplets,"/ournal

ABOUT THE AUTHORS duliln Szekely earned his Ph.D. in chemical engineering at Imperial College in 1961. He is currently a professor in matenals science and engineering at the Massachusetts Institute of Technology. Dr. Szekely is also a Fellow of TMS. Elliot Schwartz earned his Ph.D. in materials engineenng at the Massachusetts Institute of Technology in 1995. He is currently a postdoctoral research associate in the Materials Science and Technology division at Los Alamos National Laboratory. Dr. Schwartz is also a member of TMS. Rol_l_rt Hyers earned his B.S. in materials science and engineering at the Massachusetts Institute of Technology in 1992. He is currently a doctoral candidate in the same department. Mr. Hyers is also a member of TMS.
Foe m_re InformlUon, NMIomll New 8&.'i-4272; contact Laboratory, Mexico e-mill 87545; E. Mall (505) Schwartz, Stop Los G755,

(1993), pp. 371-378. 7. E. Schwan2 et al., "The neti Force Fields Metall/c Droplets toh_rodynaraics (Wat'tm'tda_. 8. E Schwartz D'mptew side'ru_

and Transport m the Microgravity

l_,_mena m Levltated Env'tronment," _g_,J. Szek_y of I.iqu_ M_kods et ,!1.

in Process Metallurgy, ed. PA: TMS, 1992), pp. 81--8_;. and i. Szekely, "The S'hl,_e

Metal Confor

m E]eceromagnmic Levita,lon Exits [ntm'nai Fluid Flow," xpemraental

M_ Matemals Scmace. ed. R. Sd'uffman (Warmndale, PA: TMS, 1994), pp. 73-79 9. I. Szekely, Flu ,a Flow Phenomena i4_t, tals Processing (New York: Academac Pt_.ss, 1979), 10. E. Schwm'_ and J. Szekely, Eseani_l p. 208. " "Mathehustical Modeling: An

ing the correction formula of Cummings and Blackburn. 37It can be seen in Figure 4 that the data points from the microgravity experiments and the "1 g-results extrapolated to 0 g" fall very close to a single regression line, illustrating the accuracy of the results from the microgravity experiments. We confirmed our hypothesis that microgravity experimentation would eliminate the need for the Cummings-Blackburn formula, which is used to account for the effect of magnetic field on the surface tension value in earthbound levitation experiments. The microgravity surface tension measurements and the results of

Component of the Design Proce_mg m the Computer P-I-Kaddah, and S.P. Marsh (W_dale. 147-161.

of Space Exparu_ts/' Age II, ed. V.R. VoUet, N. PA: TMS, 1995), pp.

II. A. Gag1_ud, J. Etay, and M. Gamier, "The Levitation Met_ Process Using Cold Crucible Tec.M'uque," Tra_tmns ISIJ, 28 (19_8), pp. 36*40. 12, A. C.agnoud EIec_omagnetic mS," 268. IEEE and L Ledercq, "Free Boundary, Problem in Levitation Melting and Continuous Caston Ma_eflcs, 24 (l) (1988), pp. 256MeaLs," A_l:_.q.'m

Transa_wns

13. A.L Mes_l, "Magn_ic Levitation of Liquid Journal ef Fluid Mechanics, 117 (1982), pp. 45--70. 14. A.D. St'_yd and H.K. Moffatt, "Fluid_c#l of the Levitation-Melting 117 (1962), pp. 45-70. Process," Journalof

Fluut Mechanics,

15. G. Lohoelet, "Theory of an Elect_magneticaUy Levitated MetaJ Sphere---q) Absorbed Power," 51AM ]ournalo(A_ied ,Vtat_nnat_cs, 49 (19_9), pp. 567-581. 16. G. Lohoefe_, "Force and Torque of an Bec_'_n_gne'dcaUy Levitated Metal Sphere," Quarterly ofApplie_ Mathematics,Sl (3) (1993), pp. 495-$18.

Alarnol Loll flUl Allmol, (505)

6(_P6782;

_dllot@tanl.gov.

1995 May JOM

53