g 3 4456 03L0S50L 1 Rea 0 RESEARCH A0 DEVELDPHENT mvaet™s & e ’ RATORY RECORDS, ?fi ”?) THE ATOMIC-POWERED AIRCRAFT JANUARY, 1950 v CENTRAL RESEARCH LIBRARY ) b CIRCULATION SECTION : : 4500N ROOM 175 LIBRARY LOAN COPY DO NOT TRANSFER TO ANOTHER PERSON a If you wish someone else 10 see this . Qa £ report, send in name with report and . . ; the library will arrange a loan. .y o — Lov RSy i e k ; *b: : ] e E TR ] ' . _ OAK RIDGE NATIONAL LABORATORY '“{L I b s 2 n‘v. s OAK RIDGE NATIONAL LABORATORY OPERATED BY @ CARBIDE AND CARBON CHEMICALS Dlw@m r,-.- "' UNION CARBIDE AND CARBON CORPORATION l : \ @ ?' POST OFFICE BOX P DAK RIDGE, TENNESBEE - A ¥ s » - v '-—‘i ‘* @ Contract No, W=7405, eng 26 REACTOR 'TECHNOLOGY .DIVISION ' THE ATOMIG-POWERED AIRCRAFT .JANUARY, 1950 Cecil B. Ellis Date Issued: APR 261950 . OAK. RIDGE. NATI%NéLbLABOBATORY opersa CABBIDE AND CARBON - CHEHICALS DIVISION ‘Union Carbide and Carbon Corporation Post 0ffice Box P 0ak Ridge, Tennessee *Prévibusly issued aé CF-50= =42 ORNL 6&/ * This document consists of 36 pages., Copy_ 7 of 121, Series A. TR 3 4y5kL 03L0SO0L L 1b CRNL 684 Reactors INTERNAL DISTRINUTION 1. G. T. Felbeckl§(C&CCC) %. C. E, Lagfon 22, F, L. Steahly 2=3, 706 A Library 15 A. M. Y#lnberg 23, J. A, Lane 4. 706 B Library 16 E. J furphy 24. R. N, Lyon 5. Biology Librar 17. C. @ Center 25. M. M. Mamn 6. Health Physics Mbrary 18, J.A&F. Swartout 26, W. D, lavers (Y-12) 7=8, Training School Wibrary 19, AdFHollaender 27, W. B, Humes (K-25) 9. Metallurgy Libraf 20, 4. H, Gillette 10-13, Central Files 21,4 . B. Ellis 28-41, Central Files (0.P.) EXTERNAL DISTRIBUTION 42=53., Argonne Natio afbratory 54=60, Atomic Energy @miffsion, Washington 61, Battelle Memor MU¥nstitute 62-65, Brookhaven NotiWB1 Laboratory 66, Bureau of Shipg 67. Chicago Operajffiols Office 68=71, General Elecjiific Wompany, Richland 72, Hanford Opegftiorl Office 73-74. Idaho Opergifions ®&fice 75, Iowa StatgfCollege 76=79, Knolls ric PoweM Laboratory 80=-8", Los Alajl: 83, Massaciiisetts Instifte of Technology (Kaufmann) 84-93, Massagusetts Instit@te of Technology 9~95, Natiglal Advisory Cofbittee for Aeronautics 9% . MEPProject 97=-98, Nglf York Operations OMFice 99. Jrth American Aviatiof, Inc, 100, g¥tent Branch, Washing®n 101~115, Mechnical Information Hianch, ORE 116-117, 118=12] University of Californif Radiation Laboratory Westinghouse Electric Ciporation TABLE OF CONTENTS Part I. Conclusions. Part II. Basic Reactor Design Criteria. Part III. Some Viewpoints from the Lexington Report. Part IV. Major Developments Since Lexington. Part V. Sample Calculations with Newer Dats. NOTE: This status report was prepared in January, 1950, in answer to a specific list of questions submitted to the AEC by the Department of Defense. The present reprinting is being circulated with the hope that it may be of gen- eral interest to those engaged in reactor work. The views expressed within are those of the Oak Ridge National Labora- tory only. They do not necessarily reflect the opinions of everyone assoclated with the Aircraft Nuclear Propulsion Program. A. M. Weinberg Research Director April 17, 1950 o THE ATOMIC-PQWERED ATRCRAPT | JANUARY 1220'. PART 1. concmszous | There are nowvnodergfely:good grounds for believing that a supersonie airplane can be rlown-ufider firaniul power. The most crucial questions out- standing today are those ofthe‘achievable aerodynamic lift-to-drag ratio for a large lnperaonig craft; and the attainment of reactor materiales for high temperatures. The nuclear power plant still poses many difficylt umsolved problems. However, for each of these problems two or three promising approaches to a solution can now fie a#etched out. The number of pugsible alteinatives in the design is great enough to suggest that feasible solutions will eventually be found, in one way or another. It is gtill much ng early for accurate prediction of the size and per- formance of the first\huclofir airplane. However, the chief possibilities are discussed in the following sectioné, and some inforwed guesses can be made. It seems likely that é manned craft, bullt as a single unit, could be driven at Mach l.0 at 60,000 Pest altitude with a groes weight in the 350,000 - 450,000 pound range‘-- provided a lift-to-drag ratio of about 9.0 was available. The U°3° content of the reactorrmight be in the neighborhood of 200 pounds. The allowable non-stop flight, time at Mach 1.0 would be at least 100 hours. The continuous holdup of fiseianable material in chemical and metallurgical reprocessing might be kept as léw as 100 pounds per flying alrcraft by special attention to the fuel element design. Of course the “h / logistics of running a squadron of nuclear aircraft would lead to the tie-up of congiderably more material than this in the complete operation. The out- look on other operating characteristics remaing about the same asldiacusaed in the Lexington Rerort. |+ PABT TI. BASIC NEACTOR DESICK CRITERIA The most important compromise to be adjusted in designing a nuclear ailrcraft 1s the balanéa batwéen naximum reactor temperature ind-grbfil veight. Every possible effort must be made to relieve the materials problems by lowering the fuel tempesrature. Likewiss, the grdss welght must be kept as low as possible to reduce-the difficulties of reaching supersonic speeds. These two factors usually oppose one another, since smaller shields mean smeller core volumes and thus higher power densities. The design balance ghould be struck at such a level that the materials difficulties are no worse than the aerodynsmics difficulties. If the desired flight conditions, and the best achievable L/D and machinery efficiency for those conditions, are fixed there still remain two possibilities for lowering the fuel teméerature assoclated with any .gross welght. These are (a) to improve the heat transfer rate within the core, so that the desired power density will not lead to such a high temp- erature, and (b) to improve the shielding art, so that the weight of the | shield necessary for a given core si;e is decreased. It is for these reasons that the most intensive development work on the nuclear aircraft is now being applied toward better heat transfer systems, toward high temperature mater- ials, and toward improved shields. Heat Transfer Mechanisms: The partiéular combination of fuel temperature and gross weight at which the problems on both sides seem most easily manageable will strongly depend on the heat transfer mechanism used, and will surely be different for different mechanisms. R 1. The first question affecting heat transfer is on the physical state of the material in which the heat is inltially developed. If the uranium is dissolved in a liquid, this liquid can be circulated outside the restricted volume of the core and it will therefore not be necessary to accomplish the heat transfer within this small volume at all. There are of course a new group of difficulties brought into the picture by this device, some of which are @iscussed in the following sections. \One of the important gquestions, for example, is the matter of the intermediate heat exchanger somewhere in the shield which would now be necessary. The primary circulating fluid will contain the intensely radicactive fission products and so cannot ever be allowed to circulate completely outside of the shield. If the heat exchanger between the primary circulating fuel and the secondary fluld, possibly another ligquid metal, can be so designed that embedding it in the shield does not appreciably increase the shield weight, this system will doubtless have much merit. The circulating fuel arrangement has not yet been adequately explored, but it certainly represents another alternative to the solid fuel systems which have been more fisually discussed so far. It has the immense advantage that radiation damage to the fuel elements is unimportant and that many of the questions of thermal stress wilthin the core likewise vanish. It is to be noted that the really fundamental gain expected from the circulating fuel -arrangement is in the increased size of the heal transfer surface available. Instead of having to transfer the heat from those surfaces which can be placed within the small core volume, one can now expose surfaces of the intermediate exchanger which is distributed throughout a presumably much larger volume within the core. However, it is not yet at all certain that o L there will be a net gain, since the méximum femfierature allowable to the intermediate heat exchanger materiasls may not be very -high. | 2. If the fuel is to remain in the C6re, then heat transfer between some solid and the circulating working fiuid will be fiecessary within the restricted core volume. It is this heat transfer process which then must be run at the greatest feasible number of KW/cm® or BTU/sq ft/hr. Of course the processes which will transfer the highest heat flux from a solid to a fluid are probably of the type which cannot be used in the reactor applica- tion. These would be such mechanisms as an explosive reaction occurring at the solid wall, or the use of chemical dissociation or ionization pro- cesses. IExcluding such devices, fhe arrangement which would seem to permit the highest heat flux at temperatures sultable for aircraft work would appear to be convective heat transfer between the solid walls and a swiftly flowing liquid metal. It ies true that the heat transfer pfopérties of liquid metal systems at high temperatfires have not yet been investigated; however, the measurements made so far, which have extended up to the neighborhood of 1200°F, all seem to justify a fairly simple heat.transfer formula. When extrapolated to higher-temperatures and flow velocities and larger film drops, it seems to be a reasonable expecfiation that heat fluxes of the order of 2 million BTU/sq ft/hr can be achieved with liquid metal working fluids. The liquid metal cooled system is the type of aircraft reactor now receiving the most intensive study by all laboratories inter#sted in the ANP Program. The principal disadvantage is the probable difficulty of finding fuel wall and plping materials which will stand the corrosive action of liquid metals at the necessary temperature level. Some suggested sets of specifications for a liquid metal cooled aircraft are listed in Part V. o The liQuid metal cooled aircraft feactor system has been discussed in some detail on previous occasiofis where it was assumed that the system would represent a binary cycle, i.e., the liquid flowing through the reactor core would be led out through the shield to a radiator, where its heat would be transferred to air paséing through a tfirbojet. It is now beginning to seem likely that no such binery system will prove feasible. There are at least three reasons for suspecting that the liquid metal cycle, just as the circulafing fuel cycle, will require an intermediate heat exchanger located within the sghield, and will therefore be a ternary system. Reasons which have been advanced for passing to ternary cycles are: a. It is suggested that it will be impossible to prevent the occurrence of radioactive impurities from the tube walls getting into the primary coolant stream. Thus, the working fluid would carry radio- active material outside of the shield even though the working fluid itself was some relatively inert liquid such as bismuth. b. The vulnerability of & binary liguid metal system to enemy attack might be too great to be tolerated. However, if the unique parts of the system -- the reactor core and its coolant -~ were all kept within the massive shield where they are relatively safe from an enemy projectile, then it would be possible to have the several external turbojet engines served by several independent secondary fluid cycles, any one of which might be lost without jeopardizing the reactor as a whole. ¢c. If it proved desirable to use bismuth or lead-bismuth as the primary liquid metal coolant within the core, it would probably not be feagible to circulate this (even if inert) to possibly a dozen separate . turbojets scattered around the plane. The volume of bismuth necessary would add too much weight to the system. It might, therefore, be de- girable to transfer the heat within the shield to a lighter metal such as ordinary lithium, or perhaps to a molten salt such as NaCH. However, it may be noted that none of these three points has yet been ex- haustively explored, and the binary liquid metel cooled system is not yet completely excluded. |oo L 5 3. Another system of possibly equal heat trangfer capabilities to the liquid metal systeg is one usifig 8 boiling fluid within the reactor core. Such & sgystem has élways been considered impossible for use within a reactor. However, the question has never been settled experimentally and there are growing groupdsqfor feeling that the subject should be re-opened. It is trfie that this §ys;em, being such a radical departure from previous reactor experéénce, will-briné:with it a number of problems on which there 1s yet extremelyvmeager éxperience. One shoul& fherefore only turn to the boiling flfiid-systems if‘their capacity for coping with high heat flux seems to be merkedly higher. than that of flowing liquid metal, or if the liquid metal systems prove impossible to manage from a materials standpoint. The heat‘flUxés:which can be achieved between a solid wall and a boiling fluid in high épeed forced convection have‘never been adequately explored in the laboratory. It is known that the nature of the solid sur- Tace exercises & profound effect upon the boiling phenomena, and it would probably be difficult to mainfiain constant surface conditions. The only really extensive experience is on watef ranging from normal boiling under atmospheric pressure to boiling at pressures as high as 1400 psi or more. With water at atmospheric pressure, the highest heat fluxes obtained are in the neighborhood of 1/2 million BTU/sq ft/hr. This figure really applies to.natural convection; doubtless a high speed flow of the water would produce much greater heat transfer rates than this. At rather high pressures, heat fluxes of 2 million Bry/sq ft/hr have been achieved already. These fluxes are comparable to the fluxes attainable with liquid metals at atmospheriec pressure. If, now, one changed from water to a suitable high boiling fluid, it appears very likely that the heat transfer rate of liquid metals could be matched without going to very great pressures in the system. SNy e 10 The disadvantages of such a cycie are obvious. The material used would be new ard the boiling phenqmenon ig difficult to maintain in a smooth steady state. The problem of controlling a‘nuclear reactor with even minute fluctuationg in the average_core_density has been discussed many times. 4, From the purely heat transfer standpoint, there seems little reason to congider other 1iquid arrangements than the circulating fuel, the liquid metal, or the boiling liquid systems. All other flowing liquids would yield lower heat transfer coefficients. However, there 1s some possibility of mak- ing use of molten NaOH as primary coolant for the purpose of decreasing uran- ium investment by the addition of hydrogen as & moderator. This would be at the expense of increased gross weight because of the lower heat transfer, and thus larger core volume and shield weight. 5. Another liquid system which has hardly been explored at all is to use a fluid within the reactor core which may be vaporized outside by re- ducing the pressure and ailowing it to run a vapor turbine. Such an arrange- ment offers no improvement of the heat transfer properties of the core volume, but it permits operation of the turbine at considerably lower fluid tempera- tures and so permits lowering the temperature of the reactor materials even without improving the heat trensfer. So far, it has not appeared to be wise to put much time on this cycle since the difficulties_of deviging efficient turbine machinery for such an arrangement would be expected to outweigh the gain from the reduced material temperatures. It is expected that such vepor cycles will be explored only if the problem of reactor material temperatures eventually turns out to be even more acute than it seems now. 6. Continuing in the direction of decreasing heat transfer coefficient, the next step is to a compressed gas. Here, helium at some 2000 psi should S 1 be most suitable far eircraft work. This cycle has so fer received insuffi- cient uttentién,‘ The survey in the‘Lexington Report suggested that even with the decreased hest transfer rate of tbe gas as compared with liquid metal, the resulting sirplane still had a chance.of matching the liguid metal gress weight with no more than sbout 100°F increasé in fuel temperaturs. Such a premium would be quite reasonableito pay for relief from materials corrosion troubles. | T. The system of lowest heat tfansfer coafficient 1s the open-cycle air cooled reactor. Here, air is taken intO'the machine at ambient pressure and compressed to popaibly 20 times this pressure before being sent inte the reactor core. Although thil heat transfer mechanism will lead to a rather large core volume, and so te comsiderably higher reactor temperatures for a glven groes weight, it has to recommend it.the great merit of simplicity as regards handling the working fluid. Whethet this asset is counter-balanced by the added materiels difficultiés arising from high temperature oxidation -and by the heavy machinery weight at highest altitudes, still remains to he settled. Both the NEPA Project and the Lexington Project have explored the air cycle in congiderable detail. Surface to Volume Ratio: The preceding remarks on heat transfer have stressed the choice of the best heat transfer mechanism. There is of course the companion aspect of the problem -~ in order to get the greatest number of KN/cm3 from the core, one should also have the largest possible amount of cooling surface ares per cubic centimeter. The geometry of the core materisl should be arranged to give maximum surface to volume ratio practicablg in view of the requirements of structural rigidity, resistance to thermal stress, resistance to radiation demage, etc. In general, of course, one wbuld achieve higher surface to volume ratios by goiflg‘to pmaller coolant pagsage diameters. The limit in this direction is thé uge Qflporous materials. This has not been explored adequately becaUIe‘of the §ifficulty o% coating inside the pores to prevent escape of fission proiucts. If this problem could be solved, it is certain that very large heat fluxes could be handled by pessing gaseous or liquid coolants through porous materials. High Temperature Materials: If the liquid metal cycle is used, structural and fuel elements must be gought which are resistant to corrosion at high temperatures. Work in this field is now off to a wigorous start with exploration of various solid metals against liquid Bi, Pb and Li. It is also.possible that the fuel elements could be ceremics with metal cladding. .The possibilities for the use of ceramets also remsin to be explored. The materials problem of the circula- ting fuel cycle would appear roughly the same as for the liquid metal cycle, with the exception of the greatly decreased@ surface area to be protected. Pogsibilities of materials with the boiling fluid or vapor cycle have not yet been thought out. With the helium cycle, the material difficulties should be greatly relaxed since the corrosion worries will be negligible. For the alr cycle, the material problems have always seemed most extreme -~ partly because of the presence of oxidation and partly because of the higher reactor temperature necessary. However, as will be discussed in Part IV, the NEPA Division and its subcontractors appear to be having considerable success in developing oxidation-resistant ceramic coatings for temperatures to 2500CF in high-speed air. All of thé;cyCIes will have fo cnntnnd‘to some extent with unknown radiation damngé problens. ‘No aée&nafie experimental data are yet available for the behavior of any material at the tenperatures and the neutron and fission product fluxes applicable'tp any of the»aircraft reactor cycles. waéver, it may be hoped that the rélatively high temperatures of the aircraft system may be a help with regard to radiation damage, since the high temperature may con-. tinuously provide partial annealing and restoration of the damaged areas. It should be nbtedzthat radiation damage problems would be very greatly decreased with the liquid fuel arrangements. Shielding: The shield around the reactor core must protect against three types of radiation: (a) neutrons, which arise almost entirely in the core (except if the shield should contain uranium), (b) primary gamma quanta from the core, and (c) secondary gamms quanta, originating within the shield. The secondary gammas come from neutron capture and inelastic scattering of neutrons. Each of these three constituents of the radiation must be considered separately in the shielding problem. Because of this complexity, it is rather unlikely that a shield of uniform make-up throughout will prove to be the most efficient fram a weight standpoint. Shielding against the gamma radiation is best done by heavy elementé, while shielding against the neutrons is most efficient with light elements which quickly slow the neutrons by elastic collision. Com- plicating features in the problem are that the néutrons are also slowed down by inelastic collision in heavy elements, and that secondary gamma produetion is going on throughout the shield. In order to reduce the total shield weight, the heavy material should be put as close to the center as possible leaving the lighter material to go néarer fihe outside. The optimum arrangement and proportion of heavy and light materials has not yet been determined with any finality -- either experimentally or theoretlcally. However, there now exist a number of combinations which mey be fairly close to the best arrsngement possible. It seems likely that a very good shield could be made by using lead as the heavy element and either boron or hydrogen, or a mixture of the two, as the light element. The boron would probably best be employed as boron carbide and the hydrogen might be inserted as water or as a hydrocarbon. Alternative materials are uranium, thorium, and iron and tungsten for the heavy materials: It is clear that the shield must be specifically designed to attenuvate the fastest neutrons, those, for example, over about 1 Mev, and the hardest gemme. quanta, those from 2 to 5 Mev. If the fastest neutrons and the worst gammas are stopped, the other radiation will asutomatically have been taken care of. For the air cycle or the compressed helium cycle there is the question of leakage of fieutrons out of the reactor through the air ducts in the shield. This problem hasg not yet been covered either experimentally or theoretically. The ducts must be expected to add materially to the shield weight. (It is not believed that ducts through the shield will cause important difficulty in the case of the liquid metal cycle.) Calculations have been based, so far, on the so-called "military tolerance" of 25 Roentgens per mission for aircraft crew members. It is suggested that this tolerance, though.doubtless proper for an actual combat mission, may well need to be revised as a design specification for the nuclear aircraft. A tolerance limit of 25 Rdefitgens perfflighf-leaveb room for no more than elght flights and no accidents ifi'a.cféw'mfin's lifetime. Such a situation makes test flying end practice missioné véry.difficult. Certalnly a great deal of exténded flying will be needed in the eafly>days of the nuclear air- craft and will always be needed for'photographic reconnalssance. it 1s sug- gested that the tolerancs per Eh‘hbur day‘be reduced to sofiething in the neighborhood of 5 R. This musf be expected‘to cause an abpreciable increase in shield weight. It may be noted‘that some medical work is mow under way on the problem of artifically increasing man’é tolerance to radiation. Reactor Neutron Properties: In the above discussion of the basic degign congiderations for the aircraft reactor pro?er, the agsumption has been used that heat transfer within the reactor core was almost the prime consideration. This means that in order to keep alrcraft gross weight down, the shield perimeter and so the reactor core diameter ies to be made as small as possible by any practical means; i.e., the core diameter ig to be governéd wholly by heat transfer congsiderations. (Of course the heat transfer arrangement chosen must have been such as to satisfy numerous ;fixiliary requirements on materials, coolant handling, turbine air temperstures ,‘ ete.). It was. then implicitly assumed that sufficient uranium wofild be installed in some fashion within this speci- fied core volume to make the reacfior critical. No limitation was expressed as to the emount of uranium which could be devoted to this purpose. It was also not considered whether the reactor could always be made critical at any desired diasmeter simply by adding enough uraniufi. Naturally, there exists a minimum reactor core diameter set by the possible amount of uranium which can — . \ 16 be arranged in a sfiitable geometry-within the cofe; hdwever, this turns out to be vé:y small and in all interesting cases, smaller thafi the minimum core diameter permitted by the heat transfer consifierations._ It also appears from preliminary calculations that the amount of uranium needed to make any core criticeal whose size is'govérned by the heat transfer situation, will not be greater than a few hundred pounds of U235. Should this amount, as required by an otherwise interesting design, be considered excessive, 1t will usually be poésiblé to decreass the uranium investment in the machine by going to larger aircraft gross weight. This balance between uranium investment and gross weight of the plane is a parallel balance to the compromise between gross weight and reactor materials stressed so far in this gection. It is believed that the mafierials temperatufe is really crucial to the operation of the aircraft and that uranium investment should be given & free hand in design planning, within reason, in order to permit achieve- ment of a feasible fuel element. This situation can be re-evaluated later in the development. The approach of letting the reactor size be governed entirely by the heat transfér requirements can lead in many cases to a fast or intermediate reactor instead of the more familiar thermal type. If this occurs in an other- wise promising design, one must then balance the heat transfer and gross weight gained against the disadvantage of moving into the more unfamiliar nuclear realm. Conclusiong: The above gqualitative sketch of the more important reactor problem has been presented so as to show the relative emphasis now considered important for the various parts of the design. In the following section, a brief survey is given of seleéted parts of the Lb;ingt&n Bepbrt"which anplify the above considerationé to some extent.. A_fiufiber of thé recommendations for future study contained in the Iexipgtoh¢Beport are also listed, to illustrate the large number‘of alternativeé in thé design yetfav&ilable in case a funda- mental block is met in the more obvious schemes. | _ | 18 PART III. SOME VIFWPOINTS FROM THE LEXTNGTON REPCRT During the summef‘of i9h8{ an extensive survey of the possibilities for nuclearlpowered flight was made by the Lexington Project. The Lexington Beportl) wasg g0 stimuiating that the most of the thinking and research on nuclear aireraft durifig the.succeeding 15 months has been devoted to lines suggested therein. . The results from such work as well as from other reactor research throughout the Commission, and from wldespread aerodynamics research, have ndw filled out the picture somewhat. - In a number of respects the‘viewpoints of 1948 have been altered, although not many really new ideas have yet appeared. In the remainder of Part III some of the features of the Lexington Report are outlined as a background against which to consider the newer.material. Conclusions: The principal conclusions of the Lexington Project were: 1. There is a strong possibility that some version of nuclear powered flight can be achieved. The aircraft is expected to be subsonic. A super- sonic plane is not expected withbutnstriking improvements in aerodynamics. 2. The operating altitude will probably not be much above 50,000 ft. for any cycle. 3. The uranium content will be in the range of 20-200 1bs of ye32, k. The manned plane and the tug-tow arrangement are the most interesting. 5. A choice of power plant and coolant is not yet possible; however, the three most interesting systems are (a) the open cycle turbojet, (b) the helium- cooled compressor jet, and (c) the bismuth-cooled turbojet. The gross weights 1) LexP-1 -- Nuclear Powered Flighf, A Report to the Atomic Energy Commission by the Lexington Project, September 30, 1948. suggested for a manned plane, operating at Mach 0.9 and 30,000 ft altitude, using these cycles are: Bi Turbojet 525,000 lbs. He-Compressor-jet 650,000 lbs. Air Turbojet 900,000 1bs. 6. Reactor materials development is the most critical need of the program. As lllustration of this, the probable wall temperature of the fuel elements suggested for the most promising cycles are: Bi Turbojet 1840°F He-Compressor-jet 1830°F Air Turbojet 2500°F. 7. ©Shileld welghts are still considerably uncertain. Shielding is of dominant significance. 8. The airframe will be comparatively stralghtforward unless the re- quired gross weights become tremendous. 9. Full-scale testing will be hazardous and expensive. The most striking point in the conclusions from the Lexington Report is that although the nuclear airplane is considered possible, the manned version will be an essentially large craft which is not expected to become supersonic. The principal change in general viewpoint of those working on the nuclear air- craft program since the Lexington Report has been in regard to this point. Although the actual feasibility of a nuclear plane cannot be completely dem- onstrated yet, there are grounds for considerably more optimism in regard to achievable plane weights and speeds. In order to investigate this point, which is of the highest importance for the military end-use, the assumptions upon which the Lexington calculations were based should be most carefully studied. Fundamental Assumptions: A very far-reaching assumption made by Lexington was that the L/D of a gupersonic plane would be 3. This figure is of major importance in the question of supersonic feasibility. Another assumption by lLexington is that the required shield would be a h-foot thick wall of material having specific gravity 6. The resulting weight of shield for a spherical core of k-foot diameter would be 320,000 lbs. It is the combination of such large shield weights with the small assumed supersonic L/D which made supersonic nuclear flight seem improbable. Bismuth Turbojet Cycle: As an example of the nature of the assumptions necessary for designing nuclear aircraft, the optimum Bi turbojet cycle considered by Lexington for a subsonic craft is outlined below: A. Aircraft type is taken to be a "Delta Wing" design. B. The operating altitude is assumed to be 49,000 ft. and the speed 0.9 M. C. TFor exploring the field, a number of possible gross weights of the aircraft are selected a priori. D. Subsonic I/D of 15 ig assumed. From this, the necessary thrust is calculated for each gross weight. E. From gross weight, the total weight of the power plant system, including rotating machinery, reactor, shield and ducting is cal-~ culated. Based on comparisons with existing aircraft, the ratio of total subsonic power plant weight to gross weight 1s taken as 65% for planes over 300,000 lbs; for smaller planes, the ratio is slightly less, dropping to 61% at 150,000 lbs gross weight. F. From the gross weight and the required thrust, the weight of rotating machinery and accessories is estimated. The machinery weight is taken from existing experience to be 2.0 lbs/cu ft/sec of intake air. The amount of air required to achieve the necessary thrust is cal- culated from existing turbojet experience in terms of the figure 40 1bs thrust/cu ft/second air flow, assuming the turbojet air inlet temperature of 1500°F. S 21 —— G. Subtracting machinery weight (plus reasonable egtimates for reactor core weight, ducting and accessories) from the allowed total power plant weight, gives the resulting weight permitted for a shield. H. The permitted U°3? investment is assumed to be 100 1bs and the reactor is assumed to be thermal and moderated by BeO. From these data, a relation between core diameter and free-flow ratio follows automatically. I. The bismuth reactor inlet temperature of 1180°F and reactor outlet temperature of 1656°F are assumed. The efficiency of propulsive parts is taken as follows: Compressor 83.5% (CFR = 6) Turbine 87% Exhaust Nozzle 90% Inlet Diffuser 90% A shell and tube radiator with 1/8" ID air tubes is used. From these assumptions the relation between reactor core diameter and free-flow ratio is calculated which will transfer the necessary power. J. From the above two relations, free-flow ratio may be eliminated s0 as to pick a core diameter which is both adequate for heat transfer and contains no more than the desired amount of uranium. K. Assuming the core to be a right square cylinder, derive the thick- ness of shield around the computed core size which is permitted from the allowable weight derived above. L. Repetition of this process for the various assumed values of gross plane weight provides a curve of gross weight vs. allowable shield thickness. M. From existing shielding data, chiefly on the MO shield, it is assumed that a mixture of light and heavy material with specific gravity of 6 and thickness of 4 ft. is required to give the necessary atten- uvation for 25 R exposure to a crew 10 meters distance from the reactor. N. The last step in the calculation is to pick from the graph the re- quired aircraft gross weight to fly a 4-foot thick shield for the conditions assumed above (Mach 0.9, altitude 49,000 ft.). This gross weight comes out to be 950,000 1bs. Gas Cycles: The calculations for the open cycle air-cooled reactor are similar in spirit although somewhat more involved, because of the necessity of including S a2 the reactor core in the aerodynamic part of the system. The pressure drop through the reactor and shield ducts was assumed to be 30% of the compressor outlet stagnation pregsure. The fuel wall was taken as 2500°F, and the com- pressor ratio was taken as 40. The resulting aircraft weights are very sensi- tive to design altitude. This ariges both from the increased machinery weight needed to handle the required mass flow at low pressure and from the increased core size required to give adequate heat transfer with air of lower density. It was concluded by Lexington that there was little likelihood of being able to carry an acceptable shield above 50,000 ft. altitude with the air cycle. Another sensitive feature is the reactor fuel wall temperature. If this is dropped from 2500°F to 18300F, the gross weight becomes tremendously high, even at 30,000 ft altitude. It was suggested that fuel temperatures lower than 2300°F would not be practical with the air cyecle. Some calcula- tions were made on the effects of bringing the air into the reactor at the center using a split flow. This design will reduce the aircraft gross weight, but at the expense of a strong increase in uranium investment. The third cycle for which extended calculations were made, was the helium compressor-Jjet. The standard helium pressure assumed was 1000 psi. The aircraft gross weight is quite sensitive to this figure, varying at 30,000 £t from 575,000 1lbs to over a million pounds as the helium pressure is changed in the range of 2000 to 250 psi. One may purchase reduced gross welght at the expense of difficulties of handling extremely high gas pressures. Suggestions for Future Work: The Lexington Report made a number of suggestions for alternate designs which should be further explored. The most striking of these wag the endorse- ment of the idea of tug-tow. The tug-tow scheme was expected to require 1/3 to S 23 1/2 the shield thickness of the manned aircraft. This gives less than 85,000 1bs for a shield of specific gravity 6 surrounding a 4-foot diameter core. The greatly reduced difficulties of constructing such a plane must be balanced against the operational disadvantages of the tug-tow systen. it was estimated that a towing cable about 0.6 miles long would be needed. Some of the other suggestions in the Lexington Report which are stlil being consldered actively were for investigation of: 1. 2. - fast reactors, split flow in the air cycle, separaved shields, placing the heavy gamma shield material chiefly around the crew instead of the reactor, so as to get the benefit of the ilnverse square law attenuation over the gepsration distance, mixed reactors containing zones in the core of different moderators which could operate at different temperatures -- the aim being to reduce core size without increasing uranium investment, reactor cores containing adjacent insulated regions of fuel and moderator at considerably different temperatures; the aim belng So get hydrogenous moderator into a high temperature reactcr so as to reduce the uranium investment for a given size, shadow shielding; i.e., making the shield thinner on the side away from the crew (however, no more than about 107 of the shiecld weight was expected to be saved by such a device), study of the possibility of using extremely small chamnels in the core, even going to porous solids as a device for increasing the heat transfer rates, vapor cycles using a condensable vapor such ag steam or mercury, possibly in a ternary system with bismuth in the ccre itself, studies of control mechanics so as to achieve completely integrated systems. (It was emphasized that mechanical devices will be diffi- cult to arrange which will maintain fast precision movements in the yresence of aircraft accelerations.) Ay | ol PART IV. MAJOR DEVELOPMENTS SINCE THE LEXINGTON REPORT During the time since September, 1948, theré have been a number of gpecial groups which have investigated features applicable to the nuclear poweréd aircraft. There has also been continued work in various labora- tories along all of the lines involved in the problem. The overall super- vision of the nuclear aircraft work has been vested in a joint Commitiee of the Atomic Energy Commission, the Department of Defense, and the National Advisory Committee for Aeronautics. Radiation Damage The field of radiation damage, which was not studied in detail by Lexington, has been thoroughly surveyed by the AEC Committee on Effects of Radiation on Materials. The report of this Committeel) described the funda- mental factors involved in radiation damage to both metals and non-metals. 1t stressed the amount of research and engineering testing yet remaining to be done before any materials can be considered thoroughly suitable for z high power reactor. The principal reason for optimism on radiation damage in the alrcraft reactor is indeed the very high temperature involved, which should lead to partial annealing of the damaged regions continuously. Of specisl value in settling these unknown radiation damage questions will be the new Materials Testing Reactor now being constructed at the new Reactor Proving Grounds of the Atomic Energy Commiss;on in Idaho. Doth the establishment of this proving ground and the construction of a high-flux materials testing reactor were strongly recommended by the Lexington Report as requisite to the nuclear aircraft development. 1) AEC-500, "Survey of Effects of Radiation on Materials", by B. L. Averbach, D. S. Billington, J. W. Irvine, Jr., W. E. Johnson, A. R. Kaufman, A. V. Lawson, Jr., J. R. Low, S. Untermyer, and J. C. Slater, September 30, 1949. L g 25 A large program of radiation damage measurements on many types of materials is now underway throughout the AEC. This extensive work involving both reactor irradiation énd accelerator bomberdment is certain to provide much needed fundamentél information from which some of the radiation effects to be expected from the aircraft reactor may be deduced. However, experiments at the simultaneous high fluxes and high temperatures to be met in the air- craft case probably cannot be performed until the first prototype ANP reactor operates on a test stand. | Shielding: Information on shielding has progressed to a considerable extent during the last 15 months. On the theoretical side, Bethel), and Tonks and Burwitz have analyzed the shielding problem and concluded that a shield to adequately surround a k-foot diameter core might be bfiilt at 220,000 1lbs instead of the 320,000 lbs assumed by Lexington. The weights for smaller reactors would be proportionately less. Further theoretical work2) was also done by the Summer Shielding Session held in Oak Ridge in 1949. This work provides a firm basis for analyzing the forthcoming new experiments. The principal feature of the newly developed theory is the proposal of a principle governing the best proportion 6f heavy and light materials in the inner region of the shield. It was shown that the ratio of heavy to light material over a considerable range of the thickness should be adjusted so as to lead to equal neutron attenuation length and gamme attenuation length. This is the so-called "matched" section of the shield. It is contemplated that a shield would consist of an inner thin layer of perhaps boron to stop T ‘ ) ORNL Central Files No. 49-6-149, "Report on the Status of Shielding Informa- tion for the NEPA Project", by H. A. Bethe, June 10, 1949. 2) ORNL-415 - ORNL-440, inclusive; TID-256. SRay 26 some of thé;neutrons, immediatelj follofied‘by a layer of pure heavy material such as lead, to quickly reduce the primary gemma radiation to a level com- parable tovthe fast neutron flux. Then would come the matched section, in which the level of primary.afid secondary gammas and fast neutrons would be simultaneocusly reduced. On the outside would be a region of pure capturing ligfit material such as boron cerbide, to stop the remaining neutrons. Recent theoretical designs by NEPA have contemplated replacing the boron carbide region by water or by gasoline. The latter has a good hydrogen density and might well be convenient as an emergency fuel for landing the aircraft on chemical engines in the event of a nuclear stoppage. A further theoretical development by the NEPA Project is in the realm of the "separated shield", the scheme in which the shielding around the reactor core is predominantly light material for neutrons only. The crew is then placed at the extreme end of the aircraft and the crew quarters are surrounded by a relatively thin layer of the heavy material for protection against the gamma radiation. The distance of separation between crew and reactor, as contemplated at present, is 100 feet. It is believed that this device will produce a very marked saving in weight, especially for larger core diameters. For reactor core diameters less than about.2,5 ft the separated shield is not believed at present to lead to a great deal of advantage. Further theoretical work has been done on the possibility of "shadow shielding"”. This is the arrangement in which the shield on the side away from the crew is made thinner than the shield on the side nearest the crew. The savings to be gained by this method do not yet seem extreme, but it may prove somewhat useful. SNy o7 Additional calculations have been made by NEPA personnel on the ad- vantages to be gained fram a shield built almost entirely of uranium hydride. It appears that the weight savings would be considerable; however, the diffi- culties of handling this material at the high temperatures prevailing in the inner region of the aircraff shield might be prohibitive. In the absence of adequate éxperimental shielding data, it is not yet possible to really define shield welght with accuracy; however, the large number of possibilities for at least partially reducing the weight which are mentioned above, give grounds for optimism that a 220,000 1b shield is adequate for a 4-foot reactor running at several hundred thousand KW. On the experimental side, the new shield testing facility of the Oak Ridge National Laboratory Reactor, the so-called "1id tank", has now yielded definitive measurements on water as a shielding agent. This represents the beginning of an extended program of measurements which will include numerous heavy metal, boron, and water combinations. Another new shield testing facility is being proposed for construction, in which full-scale samples of aircraft reactor shields -~ including ducts -~ could be tested at the full design attenuation. Reactor Materials: The materials picture now looks somewhat brighter than in 1948. For reactors containing principally metals, there is the added flexibility to be gained from the use of zirconium. This element is available in a ductile form, and it has been found to have quite useful properties in general. Its alloys have yet to be explored. _ Ay 28 For a ceramic reactor, gs_in the air-cooléd cycle, the NEPA Project guotes evidence of considerable success with coatings to withstand oxida- tion at 2500°F. Several types have been found so far which will stand up for 100 hours at 2500°F in still air. Some have stood up for considerable tifie with alr flowing past at approximately Mach 1.0. One of the best coat- ings contaifis a mixture of iron, titanium, chrofiium and aluminum, which has protected the berylliumécarbide underbody from oxidation for more than 1000 hours at 2500°F., It may be noted in passing that a coating which will pro- tect the underbody against oxidation in rapidly moving air might also be expected to prevent diffusion of fission products from the inside outward. It is to be hoped that the fission products will not diffuse outward more rapidly than oxygen will diffuse inward. A materials matter on which there 1s yet no real progress but considerable new calculations is the question of separated LiT isotope for use as a primary | reactor fluid. Design and cost estimates are now being prepared on the possi- bility of large scale Li 1sotope separation by various agencies of the Atomic Energy Commission, using any of several'separatioh procegses. It is hoped at present that the expense will be sufficiently low as to render Li a primary coolant material of interest. In any case materials studies on liquid Li gystems are being carried forward, since Li may well be the best secondary working fluid in a termary cycle system. Intermediate Reactors: Although little work has been done on the possibility of passing from a thermal reactor for the aircraft, to an intermediate or a fast reactor, the feasibility of making reliable intermediate reactor calculations has been greatly improved by the recent critical experiments at the Knolls Atomic Power Laboratory. aamny - Another uncertainty in epithermai reactor calcuiations hag now been decreased by the hew;measurements on the Xel35 ebsorption band at the Cak Ridge National Laboratory. This is also important for thermal reectors since it bears on the added U235 investment which must be assumed to over- come poigon during operation of the aircraft at the extremely high neutron fluxes which will be required. Tug-Tow: The Lexington suggestion for a tug-tow system is now being investigated from the operational standpoint by the Air Force. It would not appear to be necessary to carry out reactor development aimed specifically at this system until it is proven operationally sound. Any reactor which will power a single- unit manned plane will be more fihan adequate for a tug-tow system. Reprocessing: The question of uranium hold-up in the reprocéssing of the aircraft reactor fuel elements was not investigated by Lexington, although it was suggested that the continuous hold-up might run to as much as 10 ~ 20 times the uranium content of a reactor core. However, two techniques now appear to be within sight which would greatly reduce this. These are the use of fuel elements something like the General Electric pin type, which might per- mit running to at least 15% depletion, and the use of remote metallurgy to refabricate the material with no more than about 10 days cooling time. It seems likely that such methods might keep the amount of fuel continuously undergoing chemical and metallurgicel processing as low at 100 lbs per flying ailrcraft. Of course considerably more than this amount must be tied up in a complete operation of a task force of nuclear airplanes, because of logistic recagong. Ny 30 Miscellaneous: In additiofi to the above larger pileces of work, advances in numerous details of calculation and laboratory experiments are being made at NEPA, North American; RAND, Battelle Memorial Institute, KAPL, Bureau of Standards, and othér sites active in'étomic:energy work. Among the new items now being surveyed which offer possibilities of giving extra degrees of freedom to the design may be mentioned: a) The use of molten NaOH as a primary reactor coolant. Some rough experiments indicate that its heat transfer properties at high temperature are very similar to those of room temperature water. The reason for using such a material is to get hydrogen into the reactor and so, because of its moderating action, to reduce the uranium investment needed for a given core diameter. b) The possibility of obtaining self-controlling core materials which do not require moving control mechanisms is being explored. The line of attack now under way is to include in the core a liquid at a temperature and pressure not too far below its critical point. The resulting swift change of density with temperature would have a strong regulating effect on the neutron flux. c) An attempt to experimentally check the thermal relaxation times in- volved in the integrated control of reactor plus power plant is now under way on a moderate scale. It may well turn out that the handling of the thermal time lags, due to the large heat capacity of the extensive cir- culating systems and of the heat exchangers, may prove as difficult for the aircraft operation as the actual control of the pile neutron flux. It may be most desirable to control the aircraft thrust by some auxiliary means of wasting turbine power at times, rather than by making any changes in the heat production level. ' d) ©Some consideration is being given to the possibility of using liquid metal alloy coolants having melting points considerably above room temp- erature. It is thought, for example, that the radicactive heat in a core which had once been operated would be sufficient to keep many alloys melted for quite & long time. If such things could be arranged, it might be possible to find liquid metals less corrosive than Bi, Pb, or 1i. PART V. SAMPLE CALCULATIONS WITH NEWER DATA 2 The NEPA Project has made many detailed calculations of possible nuclear aircraft chardcteristics; lThese.arecontinually being revised as new date and new viewpoints appear. Neither NEFA fior any other ANP group yet feelgs that it has sufficient fundamental date to seriously compare the different poséible cjcles on an equelly informed basis. Also, it is by no means possible yet to accurately specify the performance characteristics | obtainable from any of the cycles. However, as an illustration of the way some of the current thinking is running, a partial list of the design figures for four recent NEPA suggestions are given below.* ¥ The specificationas for these four preliminary design ideas were provided for this report in advance of publication through the courtesy of the NEPA Division, Fairchild Engine and Airplene Corporation. More details will be available in NEPA Quarterly Reports. 32 . ~ Bi-Li=Alr Bi-Li-Air He-Air Open Air Cycle i - Turbojet - Turbojet Compressor-dJet Turbojet Alrcraft Degign flight Mach Ho._ 0.8 1.5 1.5 1.5 Design altitude (ft.) 45,000 k5,000 45,000 45,000 Design point L/D 19.0- 6.67 6.5 6.5 Component Wengts(IbLl Turbojets and air . ; - : ducting 34,700 - 80,000 gL, 000 140,000 Radiators 6,300 30,000 96,000 Helium machinery 11,000 Ligquid pumps and lines 6,000 5,000 Reactor and shield 110,000 200,000 176,000 265,000 Airfreme and equip. - 118,000 175,000 213,000 235,000 Payload | 10,000 10,000 10,000 10,000 Gross Weight 255,000 500,000 600,000 650,000 Turbo jets Number used 6 12 12 Total design point ‘ thrust (1b.)*: 16,460 75,000 92,300 100,000 Total design point , ' | air flow (1b./sec.) 366 2kT0 2150 Total frontal area (£t.<) Th.1 162 190 165 Alr Compressor Pressure ratio 5 L b 20 Air Radiator Air inlet temp.(°F) - 289 433 430 Coolant inlet , temp. (OF) 1525 1600 1753 Total frontal area (£t.2) 59.3 242 370 Air Turbine Inlet temperature (°F) 1400 1400 1500 2100 Inlet pressure (psi) 13.2 23.5 20.8 93.8 ¥ Some of these designs include extra thrust for emergency use over and above the figure which would be gotten by dividing gross welght by L/D. 33 'Bi-Li-Alr Bi-Li-Alr He-Alr Open Air Turbo jet Turbo jet Compressor-Jet Turbo jet Intermediate Heat Ex-~ changer (at design point) Bi flow rate (b/sec) 8,800 118,200 Bi inlet temp.(OF) 1,58% 1,740 Bi outlet temp. (°F) 1,28% oko 1i flow rate (1lb/gec) 206 390 Li inlet temp. (OF) 1,225 800 Li outlet temp. (COF) 1,525 1,600 Reactor Flow arrangement Straight Straight Split Split Core diameter (ft.) - 1.98 3,28 3.2 5.93 Reflector thickness (ind 2.5 2.5 6 3.1 Moderator and reflector - material BeoC+1/3C BepC+1/3C BepC+1/3C BeoC+1/3C Free flow ratio 0.35 0.30 0.30 0.40 Tube hydraulic diameter (in.) 0.1 0.24 0.105 0.17 Heat transfer area {f£t.2) T34 1,660 2,910 15,6%0 Bi velocity (ft/sec) 13.3 Uranium investment -~ 90% enriched (1b) ~ 200 ~100 75 180 Medlan energy for fina%on)(ev) »~ 1000 | 8o.1o 0.7 6’0'2 Power (XKW 111,000 558,000 730,000 90,000 Virgin flux (N/ow2/sec) 2.3x101% 2.hxiolt 2x10M4 8.3x1013 Max. power density | (KW/in.3) - 28. 20. 37.2 6.32 Max. heat flux (BTU/£%.2/hr.) 900,000 1,600,000 1,110,000 198,000 Mex. wall temp. (©F) 1,601 ~ 1,750 2,500 Shield Type Unit Unit Separated Reactor-Crew Separation (£t.) 35 100 100 Reactor shield wt.(1lb.) 117,000 Crew shield wt. (1b.) 45,000 Comparison among these figures shows some of the changes in periormance characteristics which can be expected by altering the design assumptions. It ) = is especlally to be noted thgt‘by accépting a moderately fast reactor the welght of gfl5on1y-s1igbtly subsoniC'mediumwéltitude plane might be brought down to th?.B-SG ¢laés, with.unlihited renge of cofirse. The design.suggeations sfioyn.above vere worked out for a plane cruising at the rated gpeed and gltifude; The NEPA Division has also made calculations on-ltnding‘copditions. If fhe‘nuclear aircraft actually has to land under nuclear pdwer and carrying the full shield weight, extra performence above the cruleing specifications must be built into the design at several points. Asauming = 1anding‘apeed of 150'mi/hr; and a maximum alloweble sinking apeed durifig normal landing of 10 ft/sec, the required turbojet thrust for the Mach 0.8, 45,000 ft., Bi-cooied exemple risés_ from 16,460 1bs to 32,200 lbs. The resctor power is also doubled, to become 222,000 KW, and the heat flux in the core rises to 1,800,000 BTU/ftz/hr. The power density becomes 56 KW/in3. At the Oak Ridge National Laboratory some much more qualitative calcula- tions have recently been made to illustrate the point stressed in Part I of this report -- the interchangeability of fuel element temperature and aircraft grogs weight. The following weight estimates are for Mach 1.05 at 60,000 ft., with liquid Bi cooling. Although an L/D of only 7.0 was assumed -- instead of the 10 now belleved eventually pdssible for these conditions -- the results are probably somewhat optifiiétic. This is partly due to an assumed gross welght of only twice the shield weilght; possibiy 3 times the shield weight would be more realistic at 60,000 ft. altitude. - 2 Core Aircrafy . Reactor Power ‘Max. PFuel Diameter = Gross Weight ° Power ~ Density Wall Temp. 0.0 £t." 136,000 1bs. 135,000 KW -- “e 0.5 . 165,000 - 156,000 68 Wi/emS 6730°F 1.0 195,000 194,000 13.1 2kl 1.5 1 229,000 227,000 .6 1706 2.0 266,000 2k9, 000 2.2 1439 2.5 3031000 - 307,000 1.3 1315 3.0 352,000 350,000 87 1247 g-g 399,000 398,000 .63 121h 451,000 450,000 AT 1164 Those abovafio(ctors of core dianetér‘lesa'thnn about 2.5 ft. uofiid have to be epi-thnrunl to fast. | It should be enphaaized agein at this point that new experimental knowledge in shielding may produce a rolativoly great effect on the aircrutt gross weighta derived rrou.purelj'theorgticnl calculations. It woulfiiinénod be rash to Qndtg definite performence predictions for any nuclear aircraft at the present time. About as far as one could go would be to estimate that the region of Mach 1.0 and 60,000 ft. might be reached with & gross veight im the range 350,000 - 450,000 lbs., provided an L/D of Qbout 9 vere available. In general, it does appear that -- nainly becafise of -the fact that higher L/D and lower shield welght seem more likely now than at the time of the Lexing- - ton Report -- there i§ no reason to believe that supersonic flight cannot be achleved with nuclear power. | ‘In spite of the theoretical feasifility of nuclear flight, the fact that not one kilowett of mechanical power has as yet been extracted from uranium fisaion presents a serious psychological barrier.to the whole development. It is felt that the enormous techmnical problems which must be overcome in developing a power plant of hundrede of megawatts will be approachéd most realistically by building some non-flying, lower performance, power reactors as a first step.