ORNL-1771 by Jon Morrow

ORNL-1771 This document c o n s i s t s of 197 pages. copy?

of 254 copies. S e r i e s A .

Contract No. W-7405-eng-26

AIRCRAFT NUCLEAR PROPULSION PROJECT QUARTERLY PROGRESS REPORT For Period Ending September 10, 1954

W. H. Jordan, Director S. J. CrQrner, Co-Director R. I. Strough, Associate Director A. J. Miller, Assistant Director A. W. Savoloinen, Editor

DATE ISSUED

-4CL2511grl

OAK RIDGE NATIONAL LABORATORY

Operated by CARBIDE AND CARBON CHEMICALS COMPANY A Division of Unian Corbide and Carbon Corporation Post Office Box P Oak Ridge, Tennessee

3 445b 0250997 0

. , -., ..

,” ?... ,

ORNL-1771 Ptogress

INTERNAL D/STRI BUTlON 1. 2. 3. 4. 5. 6. 7. 8, 9. 10. 11. 12. 13. 14. 15.

G. M. Adamson R. G. Affel C. R. Baldock C. J. Barton E. S. Bettis 0. S. Billington F. F. Blankenship E. P. Blizard G. E. Boyd M. A. Bredig F. R. Bruce A. D. Cailihan D. W. Cardwell C. E. Center R. A. Charpie ?6. G. H. Clewett 17. C, E. Clifford 18. W. B. Cottrell 19. R. G. Cochran 20. D. D. Cowen 21. S. Cromer 22. F. L. Culler 23. L. 6. Emlet (K-25) 24. A. P. Fraas 25. W. R. Grimes 26. E, E. Hoffman 27. A. Hollaender 28. A. S. Householder 29. J. T. Howe 30. R. W. Johnson 31. W. H. Jordan 32. G. W. Keilholtx 33. C. P. Keim 34. M. T. Keliey 35. F. Kertesz 36. E. M. King 37. J. A. Lane 38. C. E. Larson 39. R. S. Livingston 40. R. N. Lyon 41. W. D. Manly

42. L. A, Monn 43. W. 5. McDonald 44. J. I-. Meem 45. A. J. Miller 46. K. Z. Morgan 47. E. J. Murphy 48. J. P. Murray (Y-12) 49. G. J. Nessle 50, P. Patriarca 51. H. F. Poppendiek 52. P. M. Reyling 53. H. W. Savage 54. A. W. Savolainen 55. E. D. Shipley 56. 0. Sisman 57. G. P, Smith 58. L. P. Smith (consultant) 59. A. H. Snell 60, C. L. Storrs 61. C. D. Susano 62. J. A. Swartout 63. E. H. Taylor 64. J. B. Trice 65. E. R. Van Artsdalen 66. F. C. VonderLage 67. J. M. Warde 68. A. M, Weinberg 69. J. 6.White 70. G. D. Whitman 71. E. P. Wigner (consultant) 72. G. C. Williams 73. J. C. Wilson 74. C. E. Winters 75-84. X-10 Document Reference Library (Y-12) 85. Biology Library 86-90. Laboratory Records Department 91. Laboratory Records, ORNL R.C. 92. Health Physics Library 93. Metallurgy Library 94. Reactor Experimental Engineering Library 95-97. Central Research Library

... Ill

EXTERNAL DISTRIBUTION 98. 99 * 100. 101. 102. 103. 104-1 IS. 116. 117. 118-122. 123. 124. 125. 126-128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138-142. 143. 144. 145, 146-148. 149. 150- 157. 158. 159. 160. 161-1 62. 163-166, 167- 168. 169- 1 70. 171. 172. 173. 174. 175- 178. 179. 1 80- 181. 182. 183. 184-185. 186- 188. 189. 190- 1 96.

A i r Force Engineering Office, Oak Ridge A i r Force Plant Representative, Burbank A i r Force Plant Representative, Seattle Air Force Plant Representative, Wood-Ridge American Machine and Foundry Company ANP Project Office, Fort Worth Argonne National Laboratory (1 copy to Kermit Anderson) Armed Forces Special Weapons Project (Sandia) Armed Forces Special Weapons Project, Washington (Gertrude Camp) Atomic Energy Commission, Washington (Lt. Col. T. A. Redfield) Babcock and Wilcox Company Qattelle Memorial Institute Bendix A v i a t i o n Corporation Brookhaven National Laboratory Bureau of Aeronautics (Grant) Bureau of Ships Chicago Patent Group Chief of Naval Research Coiiimonwealth Edison Company Convair, San Diego (C. H. Helms)

Curt i ss-Wr ight Corporation, Wright Aeronautical D i v i sion (K. Campbell ) Department of the Navy - Op-362 Detroit Edison Conipany duPont Company, Augusta duPont Company, Wilrnington Duquesne Light Cornpony Foster Whee Ier Corporation General E l e c t r i c Company, ANPD General E l e c t r i c Company, APS General E l e c t r i c Company, Richland Glen L. Martin Company (T. F. Nagey) Hanford Operations O f f i c e lowu State College Kirtland A i r Force Base K n o l l s Atomic Power Laboratory Lockland Area Office Las Alninos Scientific Laboratory Materials Laboratory (WAQC) (Col. P. L. H i l l ) Nuclear Metals, Inc, Monsanto Chemical Company Mound Laboratory Notional Advisory Committee for Aeronautics, Cleveland (A. Silverstein) Nat iona I Advisory Committee for Aeronautics, Wash i ngton Nova I Research Labor atory Newport News Shipbuilding and Dry Dock Company New York Operations O f f i c e North American Aviation, lnc. Nuclear Development Associates, Inc. Patent Branch, Washington P h i l l i p s Petroleum Cornpony (NRTS)

197-198. 199-208. 209-210. 211. 212. 213. 214. 215. 216-217. 218-219. 220. 22 1-226. 227-238. 239-253. 254.

Powerplant Laboratory (WADC) (A. M. Nelson) Pratt and Whitney Aircraft Division (Fox Project) Rand Corporation (1 copy t o V. G. Henning) Son Francisco Field Office Sylvania Electric Products, Inc. Tennessee Valley Authority (Dean) USAF Headquarters U. 5. Naval Radiological Defense Laboratory University of Cat ifornia Radiation Laboratory, Berkeley University of California Radiation Laboratory, Livermore Walter Kidde Nuclear Laboratories, Inc. West inghouse E Iec tr ic Corporation Wright Air Development Center (WCSNS, Col. John R. Hood, Jr.) Technical Information Service, Oak Ridge Division of Research and Medicine, AEC, OR0

Reports previously issued i n this series are as follows:

ORNL-528

Period Ending November

ORN L-629

Period Ending February

ORNL-768

Period Ending May

ORNL-858

Period Ending August

30, 1949 28, 1950

31, 1950 31, 1950 10, 1950

ORN L-919

Period Ending December

AN P-60

Period Ending March

ANP-65

Period Ending June

ORNL-1154

Period Ending September

10, 1951

ORNL-1170

Period Ending December

10, 1951

ORNL-1227

Period Ending March

ORNL-1294

Period Ending June

OR NL- 1375

Period Ending September

10, 1952

OR NL- 1439

Period Ending December

10, 1952

ORNL-1515

Period Ending March

ORNL- 1556

Period Ending June

ORNL-1609

Period Ending September

10, 1953

ORNL-1649

Period Ending December

10, 1953

ORNL-1692

Period Ending March

ORNL-1729

Period Ending June

10, 1951

10, 1952

10, 1953

10, 1954

FOREWORD T h i s quarterly progress report of the Aircraft Nuclear Propulsion Project a t

ORNL records the

technical progress of the research on Circulating-Fuel Reactors and a l l other the Laboratory under i t s Contract W-7405-eng-26.

ANP research a t

The report i s divided into three major parts:

Il.

Reactor Theory, Component Design and Testing, and Construction,

Materials Research, and

Ill. Shielding Research. The

ANP

Project i s comprised of about

300 technical and s c i e n t i f i c personnel engaged

many phases of research directed toward the achievement of nuclear propulsion of aircraft.

considerable portion of t h i s research i s performed in support of the work of other organizations participating i n the national

ANP effort.

However, the bulk of the

ANP research a t ORNL i s

directed toward the development of a circulating-fuel type of reactor. The effort on circulating-fuel reactors was, u n t i l recently, centered upon the Aircraft Reactor Experiment.

The equipment for t h i s reactor experiment has been assembled, and the current

status of the experiment is summarized i n Section

1 of Part I .

The design, construction, and operation of the Circulating-Fuel Reactor Experiment (CFRE), w i t h the cooperation of the Pratt obiectives.

The

& Whitney Aircraft Division, are now the specific long-range

CFRE i s to be a power plant system that w i l l include a 60-Mwcirculating-fuel

reflector-moderated reactor and an adequate heat disposal system.

Operation of the system w i l l

be for the purpose of determining the f e a s i b i l i t y and the problems associated w i t h the design, construction, and operation of a high-powered reflector-moderated aircraft reactor system. design work, as w e l l as the supporting research on materials and problems peculiar to the (previously included i n the subject sections), â&#x20AC;&#x153;Circulating-Fuel The

The

CFRE

is now reported as a subsection of Section 2,

Reflector-Moderated Reactor.â&#x20AC;?

ANP research, in addition t o that for t h e Circulating-Fuel Reoctor Experiment, falls

into three general categories: studies of aircraft-size circulating-fuel reactors, materials problems associated w i t h advanced reactor designs, and studies o f shields for nuclear aircraft. phases of research are covered in Parts

These

I t II, and Ill, respectively, of t h i s report.

vii

CONTENTS

......................................................... SUMMARY ........................................................... FOREWORD.

PART I . REACTOR THEORY. COMPONENT DESIGN AND TESTING. AND CONSTRUCTION

....................... ......................................... .............. ......................................... ......................................... .............................................. ...................................... ........................................................ ............................................... ...................................... ........................................ .......................................... ....................................................... ...................................... ................................................... ............................................ .................................................... 2 . REFLECTOR-MODERATED REACTOR ..................................... Design the CFRE .................................................. CFRE Component Development Projects ..................................... Reactor Physics ..................................................... ReactorCalculations .................................................. Beryllium Thermal Test ........................................... Beryllium-lnconel-Sodium Systems ......................................... Beryllium-Inconel-Sodium CompatibiIityTests ................................. Mass Transfer Tests in Thermal-Convection Loops ............................. Mass Transfer Tests in Forced-Circulation Loops .............................. 3. EXPERIMENTAL REACTOR ENGINEERING .................................. In-Pile Loop Component Development ....................................... Horizontal-Shaft Sump Pump ............................................ Vertical-Shaft Centrifugal Sump Pump ...................................... Hydraulic Motor Pump Drives ........................................... Forced-Circulation Corrosion Loops ........................................ Inconel ..................................................... Dissimilar Metal ............................................... Gas-Furnace Heat-Source Development ...................................... Study the Cavitation Phenomenon ........................................ Expansion and of Thermal-Convection Loop Testing F a c i l i t i e s .............. 4. CRITICAL EXPERIMENTS .............................................. Reflector-Moderated Reactor ............................................. Reactor .............................................. .

vii

1 ClRCULATING-FUEL AIRCRAFT REACTOR EXPERIMENT The Experimental Reactor System Characteristics of the Fuel and the Sodium Systems During the Water Tests Preparations for Loading the ARE Fuel-Concentrate Injection Nozzle ARE Unloading Experiment ARE Fuel Recovery and Reprocessing ARE Pumps Fabrication and Testing Seal-Gas Pressure-Balancing System Radiation Damage to Pump Drives Radiation Damage to Shaft Seal MotorTest Reactor System Component Test Loop. Operation of Loop Examination of Hairpin Tube Reactor Physics. of

Stress

Loops

Improvement

Supercritical-Water

9 9 10 11 12 12 12 14 14 14 14 15

16 16 16 16

17 17 18 24

24 24 29 29 30 32 34 34 34 37 38 38 38 41 41 41 42 44 44

PART II . MATERIALS RESEARCH

..................................... 53 54 ............................... . . . . . . . . . . . . . . . . . . . 55 .................................. 56 57 ........................................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 ................................. 58 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 ....................................... 60 . . . . . . . . . . . . . . . . . . . . . . . . . . 63 63 ................................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 67 ......................................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 .................................................. 67 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 70 ................................... .................................................. 72 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 72 ................................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 77 ........................................ 78 ................................... .................................................. 79 ................................................ 79 ....................................... 80 ..................................................... 80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

5 CHEMISTRYOF MOLTEN MATERIALS SolidPhaseStudies in the NaF.ZrF,.UF, System Visual Observation of Melting Temperatures in the NaF.UF, System Phase Relationships in UF3-Bearing Systems UF, iin ZrF,. Bearing Systems UF, in NaF-KF-LiF Mixtures UF, in NaF-RbF-LiF Mixtures UF, in the Individual Alkali-Metal Fluorides UF, in Binary Alkali-Metal Fluoride Systems Purification of Rubidium Fluorlde Chemical Reactions in Molten Salts Chemical Equilibria in Fused Salts Stability of Chroniium Compounds in Molten Fluorides Reduction of NiF, by H, in NaF.;ZrF, Systems Reduction of FeF, by H, in NaF.ZrF, Systems Preparation of Various Fluorides Fundamental Chemistry of Fused Salts EMF Measurements Solubility of Xenon in Molten Salts X-Ray Diffraction Studies in Salt Systems Physical Chemistry Production of Purified Molten Fluorides Use of Zirconium Metal as a Scavenging Agent Purification of NaF.ZrF, Mixtures by Electrolysis Preparation of UF,. Bearing Fuels Alkali-Metal Fluoride Processing Facility Production Facility In-PileLoopLoading Chemistry of Alkali-Metal Hydroxides Purification., Reaction of Sodium Hydroxide with Metals 6. COaROSlONRESEARGH

...............................................

......................................... 82 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 ........................................... 85 ............................................. 86 89 ....................................... ................................. 90 ........................................... 94 . . . . . . . . . . . . . . . . . . . . . . . . 95 ........................................ 95 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 98 ................................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9% 98 ..................... ........................................ 99 ......................................... 100

and Seesayv Corrosion T e s t s Brazing Alloys in NaF.ZrF,.UF, and i n Sodium Special Stellite Heats in NaF.ZrF,.UF, and in Sodium Hastelloy R in Various Media Inconel in Molten Rubidium Carburization of Inconel by Sodium. Special Tar-Impregnated and Fired Graphite. Ceramics in Various Media Fluoride Corrosion of lnconel in Thermal Convection I.-aops Effect of UF, .in ZrF4-Base Fuels Effect of UF, in Alkali-Metal Base Fuels Effect of Zirconium Hydride Additions to Fuel Effect of Uranium Concentration Effect of lnconel Grain S i z e Fluoride Corrosion of Hastelloy B in Thermal-Convection Loops Lithium in Type 316 Stainless Steel Fundamental Corrosion Research.

Static

........................................... .......................................... ................................. ............................................

Mass Transfer in L i q u i d Lead Flammability of Sodium A l l o y s Thermodynamics of Alkali-Metal Hydroxides.

Chemical Studies of Corrosion Effect of Temperature on Corrosion of lnconel and Type Corrosion by F i s s i o n Products Controlled-Velocity Corrosion Testing Apparatus

100 102 102 108 108 109 109 110 110

316 Stainless Steel ............. .......................................... .............................. Reaction Between Graphite and Fluoride Melt ................................. Lithium Fluoride Castings ............................................. 111 7. METALLURGY ..................................................... Stress-Rupture Tests of lnconel ........................................... 112 High-Conductivity-Fin Sodium-to-Air Radiator ................................. 116 Investigations of F i n Materials .......................................... 117 Development of Brazing A l l o y s for Use in Fabricating Radiators .................... 118 Radiator Fabrication ................................................. 120 Nozzles for the Gas-Fired Liquid-Metal-Heater System ............................ 120 Special Materials Fabrication Research., .................................... 121 Stainless-Steel-Clad Molybdenum and Columbium. .............................. 121 Inconel-Type A l l o y s ................................................. 122 Nickel-Molybdenum-Base A l l o y s ......................................... 122 Duplex T u b i n g . .................................................... 123 Sigma-Phase A l l o y s ................................................. 124 Boron Carbide Shielding ............................................... 124 Tubular Fuel Elements ............................................... 124 Metallographic Examination of a Fluoride-to-Sodium Heat Exchanger. .................. 124 8. HEAT TRANSFER AND PHYSICAL PROPERTIES .............................. 127 Physical Properties Measurements ......................................... 127 Heat Capacity ..................................................... 127 Density and V i s c o s i t y ................................................ 127 Thermal Conductivity ................................................ 129 Electrical Conductivity ............................................... 129 Vapor Pressure .................................................... 129 Fused-Salt Heat T r a n s f e r . .............................................. 130 Transient Boiling Research ............................................. 130 F l u i d F l o w Studies for Circulating-Fuel Reactors ............................... 131 Heat Transfer Studies for Circulating-Fuel Reactors ............................. 131 9. RADIATIONDAMAGE ................................................. 134 MTR Static Corrosion Tests ............................................. 134 F i s s i o n Product Corrosion Study .......................................... 135 F a c i l i t i e s for Handling Irradiated Capsules ................................... 135 Analysis of Irradiated Fluoride Fuels for Uranium ............................... 136 High-Temperature Check Valve Tests ....................................... 137 Miniature I n - P i l e Loop - Bench Test ....................................... 137 L i f e Tests on an RPM Meter. Bearings. and a Small E l e c t r i c Motor Under Irradiation ......... 137 Removal Xenon from Fused Fluorides ..................................... 140 C,F,.UF, Irradiation ................................................. 140 LlTR Horizontal-Beam-Hole Fluoride-Fuel Loop ................................ 141 ORNL Graphite Reactor Sodium-lnconel Loop .................................. 142 Creep and Stress-Corrosion Tests ......................................... 142 of

................................................ 144 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 ................. 145 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 ............................ 147 10. ANALYTICAL STUDIES OF REACTOR MATERIALS., .......................... 148 148 Analytical Chemistry of Reactor Materials ................................... Determination of Oxygen Fluoride Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Oxidation-Reduction Titrations Fused . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Polarographic Studies in Fused Ammonium ............................ 150 Conversion of UF, and UF, the Respective Chlorides with BCI, .................. 151 Solubility of Tri- and Tetravalent Uranium Fluorides in Fused NaAICI4 . . . . . . . . . . . . . . . 151 Oxidation of UF, with Oxygen. ......................................... 152 Determination of Lithium. Potassium. Rubidium. and Fluoride Ion i n NaF-KF(RbF)-LiF-Base F u e l s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 154 Oil Contamination in ARE Helium ........................................ 154 Petrographic Fluoride Fuels ................................. of Analyses ............................................ 154 Remote Metallography Fission-Fragment Anneoling Studies High.Temperature. Short.Time.Grai n.Grewth Characteristics of lnconel BNL Neutron Spectrum - Radiation Damage Study MTR Neutron-Flux Spectra - Radiation Damage Study

Salts Formate

Summary

Investigations of Service

PART Ill SHIELDING RESEARCH

............................................... 157 ....................... 157 ............................................. 158 ....... 158 ....... 158 . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 ........................................... 162 ........................ 163 12. LIDTANKSHIELDING F A C I L I T Y . ....................................... 164 Reflector-Moderated Reactor and Shield Mockup Tests ........................... 164 Effective Cross of Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 GE-ANP Helical Air Duct Experimentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 13. BULK SHIELDING FACILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 Reactor Radiations Through Slabs of Graphite ................................. 168 Reactor Air Glow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 173 Fuel Activation Method Power Determination of the ARE ....................... 14. TOWER SHIELDING FACILITY .......................................... 175 Fast-Neutron Ground and Air Measurements ........................... 175 Calorimetric Reactor Power Deierrnination ................................... 175 GE-ANP R-1 Divided-Shield Mockup Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 .

11 SHIELDING ANALYSIS Slant Penetration of Composite Slab Shields by Gamma Rays Air Scattering of Neutrons. Single Anisotropic Air Scattering in the Presence of the Ground (Shielded Detector) Single Isotropic Air Scattering in the Presence of the Ground (Unshielded Detector) Formulas for Multiple Scattering i n a Uniform Medium Ground Scattering of Neutrons Focusing of Radiation in a Cylindrical Crew Compartment.

Removal

Section

for

Scattering

PART IV APPENDIXES

........................... ORGANIZATION CHART OF THE AIRCRAFT NUCLEAR PROPULSION PROJECT . . . . . . . . . . .

15 LIST OF REPORTS ISSUED DURING THE QUARTER

xii

183 185

ANP PROJECT

ESS $

P A R T 1.

R E A C T O R THEORY, COMPONENT

D E S I G N AND T E S T I N G , AND CONSTRUCTlQN Circulating-Fuel Aircraff Reactor Experiment T h e water tests of the fuel and the sodium c i r -

c u i t s of the ARE system at room temperature were completed (Sec. 1). The sodium c i r c u i t was pressure-filled w i t h water from the sodium f i l l tanks, w h i l e the fuel system was vacuum-filled ‘io ensure the elimination of gas pockets i n the fuel system. Both systems were found t o function properly, and t h e f i i i i n g , circulating, and draining operations %ere effected w i t h a minimum of d i f f i c u l t y . After the water was drained, the system was dried by heating i t t o approximately 600°F. The e l e c t r i c a l heating system was found t o be satisfactory i n the check thus afforded. The f i n a l system completion work i s now under way, that is, removal o f the sodium system reactor bypass, completion of the fuel-enrichment system installation, completion of thermocouple and insul a t i o n installation, and other minor modifications. When t h i s work i s completed, sodium w i l l Le charged t o the system and the high-temperature checkout phase of the experiment w i l l be initiated. The neutron source was put i n t o the reactor, and the nuclear instrumentation was checked out. Also, mechanical checks were made of the performance of the safety and control rods. The building electric and helium systems were made ready t o accommodate t h e loading facilities, and f i n a l arrangements were completed for attaching the fuel-sampling equipment. Radiation damage experiments indicated the des i r a b i l i t y of providing gamma shielding a t several points where elastomers were used (belts, diaphragms, etc.). Where shielding was impractical, the composition diaphragms were replaced w i t h metal diaphragms. T h e results o f operation of the reactor system component t e s t loop a t K-25 are encouraging i n that the loop has now been operated without major None of the d i f f i c u l t i e s for more than 1800 h r . minor d i f f i c u l t i e s encountered would indicate that serious problems might arise in operation of the

ARE,

T A

Reflector-Moderated Reoctar T h e program for the development arid construct i o n of the Circulating-Fuel Reactor Experiment (CFRE) has been outlined, and many of the de-

Tentative velopment psolects are under way. design data have been compiled and a flow sheet

has been prepared (Sec. 2). T h e f i r s t priority development proiect, the test of beryllium i n contact w i t h sodium and lnconel under thermal stress, has been completed, T h e r e s u l t s of t h i s t e s t were needed i n the determination of the amount of poison t o Le expected i n the reflector. The test indicated that. beryllium w i l l n o t crock under the thermal stresses involved i n the temperature range 1000 to 1300°F. Since corrosion and mass transfer, a5 w e l l as thermal stress, w i l l be important i n the beryllium-lnconel-sodium system, many static and dynamic tests under various conditions have been made. There i s considerable evidence t o indicate satisfactory compatibility in the bery Ilium-lnconelsodium system a t temperatures up to 1200°F. The temperature coefficient of r e a c t i v i t y for the CFRE was computed on the UNIVAC and, for rapid temperature changes, was found t o be -3.5 x lO-’/’F. The c r i t i c a l mass computed for the shomticuboctahedral c r i t r c a l assembly was redetermined because of errors found in the original data. The redetermined value agreed closely w i t h the experimental value, but, since the c r i t i c a l mass i s not very sensitive to errors in detail, further evaluation of the agreement must await additional experimental results.. Experimental Reactor Engineering T h e emphasis in the engineering work is now on development of components for an i n - p i l e loop for insertion in a horizontal beam hole of the MTR and the design and construction of forced-circulation corrosion testing loops (Sec, 3). The in-pile loop for insertion in the MTR i s a joint ORNL and It i s Pratt & Whitney Aircraft D i v i s i o n project. t o circulote proposed f u e l mixtures in the high-flux of the MTR so that the extent of radiation damage t o materials of constructton and the effect of r a d i T w o types ation on the fuel can be determined. of pumps have been developed for in-pile use: a

ANP QUARTERLY PROGRE.S.5 REPORT vertical-shaft centrifugol sump pump for installation external to the reactor shield and a horizontal-shaft sump pump for insertion inside a beam

hole. A turbine-type impeller i s being considered for the horizontal-shaft pump because it would

have the advantage that both the i n l e t and d i s charge could be a t the bottom. Hydraulic motors of suitably small dimensions have been found to be satisfactory drives for these pumps. Two

series of

lnconel forced-circulation cor-

rosion loops for circulating fluoride mixtures ore being developed t o meet the f o l l o w i n g require(1) a Reynolds number of 10,000 w i t h ments: temperature gradients of 100, 200, and 300°F and (2) a temperature gradient of 200°F w i t h Reynolds numbers of 800, 3,000, and 15,000. The maximum f l u i d temperature i s t o be 1500°F. A study i s under way of the cavitotion phenomenon associated w i t h operating l i q u i d metol systems at elevated temperatures, high flow rates, and high pump speeds. A correlation of f l u i d - f l o w n o i s e intensity w i t h pressure data was noted. The number of stations available for convectionloop testing was increased from 18 t o 31 and the basic design of the loops was simplified. Various means of heating the loops and of making operation of them more automatic are being studied.

Critical Experiments The f i r s t step of the present c r i t i c a l experiment program was the construction of a small two-region reflector-moderated reactor t o provide experimental data on a system of simple geometry and materials for use i n checking the calculational methods b e i n g used (Sec. 4). The core c o n s i s t s of olternate sheets of enriched uranium metal and T e f l o n and i s surrounded by a beryllium reflector. The uranium loading can be varied, w i t h i n the specified dimensions, t o make the system c r i t i c o l . The assembly was loaded as prescribed by the inultigroup calculations but was not c r i t i c a l . However, when the calculations hod been corrected t o take i n t o account errors in the original data, a new attempt t o achieve c r i t i c a l i t y w i t h the prescribed loading was made. The corrected prescribed loading was 20.9 to 22.75 Ib of U2,' and the experimental loading was 24.35 Ib of U235. A larger c r i t i c a l assembly of the same shape i s t o be constructed that w i l l c o n s i s t of three regions, w i t h the beryllium island ond the reflector separated by the fuel annulus. A further check on the calculational methods w i l l be obtained.

P A R T II. M A T E R l A L S R E S E A R C H

Chemistry of Molten Materials Studies of the fluoride systems reactor fuels were continued, w i t h phasis being given to systems uranium-bearing component i s the

of interest as particular emin which the less corrosive

UF, or a mixture of UF, and UF, rather than UF, alone (Sec. 5). Recent attempts t o correlate the anticipated reduction of UF, i n the UF3-bearing melts w i t h wet chemical a n a l y s i s for UF, and and results of petrographic examination show

UF',

some surprising Cinoriialies. When UF, dissolved mixtures, or i n N a F - L i F i n LiF, i n NoF-ZrF,-UF, mixtures i s treated under flowing hydrogen at 800°C w i t h excess uranium metal, 90% or more

of the UF, i s reduced t o UF,. However, when t h i s technique i s applied t o UF, in N a F - K F - L i F mixtures, the reduction i s only 50% complete a t 800°C and, perhaps, 75% complete a t 600°C. Petrographic examinations of the specimens reveal n o complex compounds of tetravalent uranium; it i s possible that the UF, i s "hidden" in solid compounds i n solutions or i n complex UF,-UF, which i t i s not a t present recognizable. Solid phase studies of the NaF-ZrF,-UF, system were i n i t i a t e d f o l l o w i n g completion of studies of the NaF-ZrF, system, and t o date n u ternary comA tentative equipounds have been discovered. librium diagram was prepared.

A rnethod for large-scole purification of rubidium fluoride hos been developed thot can be used i f muterial s u f f i c i e n t l y free from cesium cannot be obtained from commercial sources. Fundamental studies of the reduction of NiF, and FeF, by H, i n NaF-ZrF, systems were made as a means of determining possible improvements in p u r i f i c a t i o n Also, methods for preparing simple techniques. structural metal fluorides were studied. Large quantities of purified ZrF4-base fluorides were prepared for engineering tests a t ORNL and elsewhere, and the demand for purified fluorides of other types i s rapidly increasing. Therefore preparation and purification methods have been studied intensively i n an effort t o lower production time and costs. Developments indicate that the p r i c e of purified NnZrF, and NaF-ZrF,-UF, mixtures may be halved i n the next few months. Corrosion Research

The static and seesaw corrosion testing fac i l i t i e s were used for further studies of brazing

PERIOD ENDING SEPTEMBER 70, 7954 alloys, special S t e l l i t e heats, Hastelloy R, Inconel, graphite, and various ceramics i n sodium, fluoride fuel mixtures, and other mediums (Sec. 6). T h e brazing alloy 67% Ni-13% Ge-11% Cr-6% Si-2% Fe-1% Mn was found t o have good corrosion resistance i n fluoride fuels and fair corrosion resistance i n sodium and therefore w i l l be useful for the fabrication of many reactor components. I n the thermal-convection loop studies, UF,bearing fuels were tested in Inconel. Hot-leg attack i s not found i n lnconel loops i n which ZrF,base fluoride mixtures w i t h the uranium as UF, are circulated. A deposit is, however, found on t h e hot-leg surface. Only preliminary information i s available, but it appears that neither attack nor a hot-leg layer i s found w i t h alkali-metal-base Mixtures of fluoride mixtures containing UF,.

UF, and UF, result i n a reduction i n attack from but some attack is that found w i t h only UF,, present, and i n high-uranium-content systems the attack may be significant. Several Hastelloy 8 loops have now been successfully operated i n both the as-received and the over-aged conditions. I n both cases a consider-

able increase in hardness occurs during operation. With ZrF,-base mixtures containing UF,, very l i t t l e attack i s found, even after 1000 hr. Thermal-convection loop tests of molten lithium in type 316 stainless steel were operated for 1000 hr. There were no signs of plug formation, and only a small amount of mass transferred material was found in one loop. A l l o y s of 45% Cr-55% CO,

f i n s for sodium-to-air radiators, stress-rupture and creep tests were made on copper f i n s with various types of cladding at stress levels between 500 and 2000 psi at 15OOOF. The t e s t s show that for a 1000-hr exposure in air, stresses greater than 500 p s i and less than 1000 p s i are tolerable; that is, i n t h i s stress range there i s no indication of brittleness in the core or oxidation of the core due t o cladding failure of type-310 stainless-steelc l a d copper fins. From the over-all considerations of melting point, oxidation resistance, d i l u t i o n of f i n and tube wall, formation of low-melting eutectics, and flowability, i t was found that Coast Metals alloy 52 was the best brazing alloy for use in the construction of radiators w i t h highA sodium-to-air radiator with conductivity fins. 6 in. of type-430 stainless-steel-clad copper highconductivity f i n s was fabricated by u s e of u combination heliarc welding and brazing procedure. Packed-rod nozzle assemblies were fabricated for the 100-kw gas-fired I iquid-metal-heater system, and work was started on the formation of duplex tubing. An attempt i s being made t o prepare tubing that w i l l have good corrosion resistance on the inner surface and oxidation resistance on the outer

Ni-Mo alloys, and the Fe-Cr-base stainless steels have been shown t o be more resistant t o corrosion and mass transfer in l i q u i d lead than are the pure metals. Their resistance to mass transfer can probably be related t o the formation of intermet a l l i c compounds.

surface. Attempts are being made to find new a l l o y s i n the nickel-molybdenum system that w i l l have better high-temperature strength and fluoride corrosion resistance than lnconel has. Hastelloy 6 satisfies these requirements, but it has poor fabrication properties and oxidation resistance; it also loses i t s d u c t i l i t y i n the temperature range of interest for application i n high-temperature circulating-fuel reuctors. Investigations are under way t o find a suitable melting practice and heating treatment that w i l l increase the d u c t i l i t y of Hastelloy B in the temperature range of interest.

Metallurgy

Heat Transfer and Physical Properties

Creep and stress-rupture testing by the tubeburst method has been studied intensively (Sec. 7). I n the tube-burst tests, a tube that i s closed at one end i s stressed with an internal gas pressure. The stress pattern introduced into the specimen i n t h i s test approaches the stress pattern that w i l l be found i n ANP-type reactors. Apparatus for the t e s t s has been constructed, and a theoretical analysis has been made with which a check on the experimental results con be obtained. I n the investigation of high thermal-conductivity

T h e enthalpies and heat capacities of NaF-ZrF,-

UF, (65-15-20 mole %) and of LiF-NaF-UF, (57.638.4-4.0 mole %) were determined (Sec. 8). T h e thermal conductivity of NaF -K F-UF (46.5-26.027.5 mole %) was found t o be 0.7 Btu/hr-ft2("F/ft), that for K F - L i F - N a F - U F , (43.5-44.5-10.9-1.1 mole %) was 2.0, and that for LiF-KF-UF, (48.0-48.04.0 mole 76) was 1.4. A new electrical conduc-

t i v i t y device has been constructed and has been successfully checked with molten salts of known conductivity.

AMP QUARTERLY PROGRESS REPORT A device for studying the ratcs 3 f growth of tube-wall deposits has heen successfully tested w i t h a simple heat transfer salt. Also, a hydsodynamic f l o w system for sti~dying ths r-flectormaderated reactor flow structure h a s bseii tested. A mcsthematical study of the ternperdure structure i n converging and i n diverging channsl systems that duct f l u i d s w i t h volumetric heat S O U ~ C C S and a study of w a l l cooling requirements i n circulating-

fuel reactors were made.

Radbatim Damage Additional irradiations of lnconel CapsuOes cont a i n i n g fluoride mixtures '?awe carried out ir-i the MTR (Sec. 9). Only o n e capsule containing a UF,-bearing fuel h a s been examined, a d it shows no corrosion, i n contrast to that found previously w i t h UF4-bearing fuels. Inspection of the k.ITR in-pile loop, whish failed prior t o startup, disclosed that the failure was caused by o break i n the weld connecting the pump discharge nipple t o the fuel line. Design revisions

and refotaricatiow of some of the ports are in progress. De-velapmental work continued on a sniuller loop for operation i n UIP LITR &piece+ Detailed examination of an Incoilzl loop which circulated sodium at high temperature i n the ORNL Graphite Reactor showed no evidence of radiationMetal lograph ic exam i not i ons induced corros ion. were carried out i n the hot c e l l s oil irradiated fuel plates for Pratt R Whitney Aircraft Division, und studies were made on annealing-out of fissionfragment darnage. Y!mk continued on examination of wirc arid multiple-plate-type u n i t s for

GE-ANP.

Analytical Studies of Reactor u?rla~erta!s

The primary analytical problem continues to be the separation on$ determinntion of trivaIer3t and tetravalent uraniiiiii i n hoth NnF-ZrF, and NoF-

KF-l-iF-base f u r l s (SX.10). A successful P O tentiornett-ic f i t r a t i o n of UF, in m0li-n NaZrF, with metallic zirconium w a s accompl ishcd by '[he sclumeans of polarized platinum c!cc:rodcs, billity of UF, i n NaAlC1, w a s determined to be 18 iiq/g at 200째C, i n contrast t o o s o l u b i l i t y if

of l e s s than 1 mg/g. The:zfora molten NaAICI, i s expected t o dissolve tetravalent uranium selectively Crow the 4ue!s. Calibration measurements have tisen cornplated on the apparatus for thc determination nf oxygen ,. a s metallic oxides in rractor fuels. I he reaction

UF,

involves the hydsofluorination of the oxide and mensuremant s f the increase i n conductivity of l i q u i d MF as a function of the water formed. Investigiaticns are being mode of t h e oxidation of UF, id of UF, with oxygen at elevated tempesatures, An improved method for the determination of l i t h i u m i n NaF-KF-Li F-base fuels was developed. Also, studies were mode of the s o l u b i l i t i e s of potassium, rubidigsm, and cesium tetraphenylborates in various organic solvents to uscertain d ifferanti al so Iu b i I if ie s. P A R T Ill. S H I E L D I N G R E S E A R C H

Sh i a ICli ng Analysis Application of the Monte Carlo method t o the calculation of gamma-soy penetration of crew shield sides has been worked out, and the method appears t o be quite satisfactory for t h i s type o f problem (SEX- B 1). Considerable progress has been made on understanding the Tower Shielding Fac i l i t y (TSF) measurements of ground and air scattering, Expressions have been derived which describe the effect of ground interference with the air-scattered flux, nwd thus it i s now possible to estimate ground-scattered radiation both a t the TSF and in an airplane a t landing and takeoff. Calculations have been set up for evaluating m u l t i p l y scattered radiation i n air, The values obtained W I II be of considerable importance i n studies of the highly asymnietsic but l i g h t shields era1 current designs.

L i d Task Shialdisg Facility Pieparations for a second series of shielding t e s t s for the reflector-moderated reactor at t h e Lid Tunk Shielding Foci l i f y (LTSF) Rave included the c o n s i i u c t i o n of a large tank which will hold a l l the components of the raiockups and the irradisalution which may be ation of the LJF,C,F,, used to simulate the reactor fuel i n mockup experiments (Sec. 12). h4casurenients of the removal cross section nf carbon made i n a continuous carbnr; m5dium hove resulted in a value of ur = 0.750 b a n . I h i s i s to be compared with the prca :ous VUIUSof 0.81 + 0.05 barn measured behind 0 s o l i d slab of graphite. Thermal-neutron f l u x t i ~ e a s u r s m c n t s have been made beyond t w o configurations of GF-ANP helical air ducts, a single duct arid o tricngvlni array of three ducts. Measurements b e y m d a

35-dt~cfarray w i l l begin soon.

PERIOD ENDING SEPTEMBER IO, 7954 Bulk Shielding Facility T h e use of a graphite reflector as a shield component ha5 been investigated at the B u l k Shielding F a c i l i t y (BSF) w i t h measurements of attenuation through various thicknesses of the material (Sec.

T h e fast-neutron spectrum of the BSF reactor from 1.3 t o 10 Mev through 1 ft of graphite was a l s o measured. Removal cross section values for

13).

t h e carbon were determined t o be 0.82 barn for a I - f t slab, 0.84 barn for a 2-ft slab, and 0.80 barn for a 3-ft slab. An experiment has been performed at the BSF t o provide an experimental basis for future e s t i motes of the amount of v i s i b l e light around a nuclear-powered aircraft. Measurements i n an airf i l l e d tube placed against the reactor have indicated thot t h e maximum glow w i l l occur at a pressure corresponding to an altitude of about

30,000 ft.

I n a proposed method for the determination of the power of the ARE, the r e l a t i v e a c t i v i t y of the

fuel samples exposed i n the ARE and i n the known f l u x of another reactor is measured. The method

has been tested at the

BSF.

Tower Shielding Facility Measurements of the ground and air scattering of neutrons have been made at the Tower Shielding F a c i l i t y (TSF) (Sec. 14). A preliminary analysis indicates that the contribution at the maximum altitudes (around 200 ft) i s between 2 and 5% of t h e total scattered neutrons for differential exper iments.

A new procedure for a calorimetric determination of the power of the TSF reactor has been devised which i s based on the relationship between the r a t e of temperature r i s e i n the water of the reactor tank and the reactor power. The resuits of three experiments were consistent to w i t h i n 1%. The next series of tests at the TSF w i l l be on the GE-ANP R-1 divided-shield mockup.

part I

REACTOR THEORY, COMPONENT DESIGN AND TESTING, AND CONSTRUCTION

1. CIRCULAT1NGFUEL AIRCRAFT REACTOR EXPERIMENT E. S. B e t t i s J. L. Meem Aircraft Reactor Engi fieeri ng D i v i s i o n T H E E X P E R I M E N T A L R E A C T O R SYSTEM

A general revision was found t o be necessary i n the ARE i n that shadow shielding had to be installed, at several points where elastomers were used (belts, diaphragms, etc.),to provide protection from gamma fields that would have produced proh i b i t i v e degrees of radiation damage. At some points, where shielding was impractical, the compos i t i o n diaphragms were replaced by metal diaphragms. The entire ARE system has now been completely checked out as far as room-temperature preoperational t e s t s are concerned. Both the fuel and the sodium c i r c u i t s have been operated simultaneously w i t h water as the circulated l i q u i d i n each circuit. T h e moderator volume of the reactor was bypassed for these tests t o keep water away from the ber y l l i u m oxide blocks. The sodium c i r c u i t was pressure-filled with water from the sodium f i l l tanks, while the fuel system was vacuum-filled to ensure the elimination of gas pockets in the fuel system. 90th systems were found to function quite satisfactorily, and the filling, circulating, and draining operations were effected w i t h a lrinimum of d i f f i c u l t y , During these operational shakedown tests, it was found that the gas lines to the sump tanks could not be kept pressure t i g h t because the Swagelok f i t t i n g s developed leaks. A l l these joints have therefore been silver-soldered to correct this situation. The pumps performed satisfactorily, and useful system curves for actual future operation were obtained. After the water tests of the system had been completed, it was necessary to remove all water from the system. T h i s necessitated heating the entire system t o a temperature of approximately 600째F. T h i s operation provided the f i r s t check out of the electrical heating system. Here again the performance was gratifying. Such troubles as arose were corrected rather easily, and, i n general, the heating of the system was effected with very little d i f f i c u l t y . A CO, cold trap was incorporated in the off-gas line while the system was being heated to collect a l l moisture that was driven from the system.

Dry

air was admitted to the f i l l tanks, and both the fuel and the sodium systems were flushed with t h i s dry air, the air leaving the system through the

CO,

cold trap. When the dew point of the e x i t air from this cold trap reached -3OoF, the system was considered to be dry and the electrical heat was

turned off. Before the heat was removed from the system, however, the temperature on the thermal barrier doors around the heat exchangers was raised t o approximately 1000째F and operation of the doors was checked. Even though the doors had been reworked t o eliminate sticking, it was found that. When the barrier the doors were s t i l l binding. doors were removed, it could be seen that the side guides whichrun the height of the door were binding i n the guide ways of the door frames. T h i s binding was resulting from the bowing of the guide r a i l s caused by the thermal gradient which exists i n the door. The door does not require these side guides, since the door housing provides ample guides for the normal functioning of the doors. Therefore the side guide r a i l s are being removed so that there w i l l be no further binding of the barrier doors.

It was also noted while the system was hot that a considerable volume of kerosene was being driven out of the beryllium oxide moderator blocks by the high temperature. The kerosene had been absorbed by the blocks during the cutting operations. To remove the kerosene, a vacuum was pulled on the moderator volume of the reactor. A CQ, cold trap was inserted i n the vacuum l i n e t o the moderator, and w h i l e the reactor was maintained a t a temperature of approximately 45OoF, the vacuum pump was run continuously. T h i s operation wa5 continued u n t i l no more condensate of the kerosenedistillation was being collected i n the c o l d trap. Approximatesy 2 gal of kerosene was removed by t h i s procedure. It i s possible in fact, f a i r l y certain that some residue was l e f t i n the moderator blocks. It was the consensus of the chemists that this would not have an adverse effect on the sodium codant, and therefore nothing further i s to be done amount of tarry residue remaining i n the reactor moderator blocks,

AMP QUARTERLY PROGRESS REPORT

After the system had been thoroughly dried, a l l heat wes turned off and the f i n a l system completion work was begun. The mojsr work i n t h i s category consists i n removing the sodium system reactor b y p s s , removing the flanged f i l t e r pots i n the sodium purification system, and completing the enrichment system tie-in 9.0 the fuel circuit. Thermocouple installation a l s o has t o he completed, as we!/ as thermal insulation of the fuel pump bowls. The decision t o remove the flanged f i l t e r pots and t o substitute weld-senled f i l t e r pots was made several months ago, It was decided at that time, however, t o postpone t h i s substitution unti I after the water test. The original f i l t e r s were useful i n cleaning up the water used t o test the operation of the sodium circuit. Completion of the duel enrichtea test, and ment system i n no way CIffeCtad the hence t h i s work was postponed i n the interest of completing the water t e s t as expeditiously a s possible. The fuel i n j e c t i o n system i s t o be water-tested i ndependentiy of the fuel system. Insulation of the fuel pump bowls was postponed t o a l l o w more rooin for access t o the pumps during the water tests. Since the pump3 hove now been checked out, the insulation iob w i l l be completed. The system should be ready for charging w i t h sodium early in September, and the high-trmperature check out phase of the experiment w i l l be initiated. While the sodium checks are being rui-i, the hot-gas l e & t e s t of the fuel system will be performed. T h i s test involves loading the fuel system w i t h mixture of helium and krypton at a pressure of about 1.5 psi. While t h i s gas i s being circulated a t 1300OF' by the fuel purnp the annulus c i r c u i t s w i l l be monitored by rnoss spectrographic methods for the presence of krypton i n .the annulus. 'The nevtran source W Q S put in the reactor, and the nuclear instrumentation w a s checked out. All three (the PWO regular p l u s the spare) f i s s i o n chambers we?:^: checked, and the count-rate v s The data chamber-voltage curves were plotted, obtained provided the necessary information for establishing the operational plateau for each Also, mechanical chaclts were mode of chambzr. the perdorinnnce of the sofety and control rods. A modification of the rod-cooling c i r c u i t w a s required i n order t o minimize the heat l o s s from the center of the reactoi. T h i s rieccssitated installation of proportioning orifices t o gat correct helium flow c.uruund t h > f i s s i o n chambers and the safety rods.

F i n a l modifications of the sodium and the fuel loading systems were completed, and the building electric and helium systems were made ready t o accommodate the loading f a c i l i t i e s when they are brought into the bui Iding. F i n a l arrangements have been completed for attaching the fuel-sampling system. T h i s sampling system w i l l not be l e f t connected when the power run i s initiated. The engineering prints for the entire ARE installation are now up t o date, and the electrical prints, in particular, were used successfully i n checking out the many heater c i r c u i t s involved in the experiment. C H A R A C T E R I S T I C S ai= T H E F U E L A N D T H E SODIUM SYSTEMS D U R I N G W A T E R T E S T S

The characteristics of both the fuel and the sodium systems while circulating water have been determined, Upon removal of the fuel heat exchanger bypass loop, a glass rotameter was tempor a r i l y installed i n the fuel c i r c u i t between the reactor outlet and the heat exchangers. A direct calibration of the high-temperature fuel rotameter against the glass rotameter was obtained w i t h water as the fluid. Conversion of water flow t o fuel f l o w ' i s made by

where

q 1 = water flow (gprn) , q2 = fuel flow (gpm) ,

Pf2

= water density (g/cm3) =

fuel density (g/cm3)

= flocrt density (g/cm3)

During the period of water operation, data on pump speed vs flow rate were obtained (Fig. 1.1). Also, the pressure head from the pump suction t o The the ieactor inlet was measured (Fig. 1.2). system characteristics, as shown i n these t w o ciJrves, shoirld be the same during operation w i t h f Iuor i des. ' A . L. Southern. D i s c u s s i o n of the R o t n m e t e r U s e d in the A R E Fuel C z r c v i t , A R E F i l e s .

PERlOD ENDlNG SEPTEMBER 10, 7954

r?lsb

ORNL-LR--DWG

2400

3087

2400

e000

7000

40c 0

50 GO FLOW X A l E (yoml 40

Fig. 1.1. Data on Pump Speed Y S Flow Rate Obtained During Water Test of ARE F u e l System.

100

120

140

160

<80

FL0h 3 A K ( y p m l

Fig. 1.3. Data on Pump Speed v s F l o w Rate Obtained During Water T e s t of ARE Sodium System.

gllr

ORNL -LR- DWG 3088

1no

6!)

H!)

FLOW RATE ( y p n i )

Fig. 1.2. Data on Pressure Head from the Pump Suction t o the Reactor I n l e t v s the Flow Rate During Water Test of ARE F u e l System. B y using a calibrated o r i f i c e installed tempor a r i l y in the sodium bypass l i n e around the reactor, it was possible to measure the f l o w rate w h i l e operating the sodium system w i t h water. The pump speed was recorded and a l s o the pressure head from the pump suction to the reactor outlet. T h e system characteristics w i t h water, shown in Figs. 1.3 and 1.4, should be v a l i d while operating During operation w i t h w i t h sodium as t h e fluid. sodium, the flow rate i n each of the two loops w i l l be measured w i t h an electromagnetic flowmeter. The sum of the t w o f l o w rates should equal the t o t a l flow as obtained from the pump speed.

Fig. 1.4. Data on Pressure Head from the Pump Suction t o the Reactor Q u t l e i vs the F l o w Rate During Water T e s t of the ARE Sodium System. P R E P A R A T I O H S FOR L O A D I N G THE A R E

G. J. Nessle Materia Is Chemistry D i v i s i o n

A l l the preliminary preparations for loading the barren and enriched fluoride mixtures into the ARE have been completed. T h e necessary control panels have been assembled, and furnaces, heating units, transfer lines, and vorious complementary equipment are now ready for installation. The exact location of the various operations has been determined. However, because of the large amount of construction work which has been carried out in

ANP QUARTERLY PROGRESS REPORT ARE building around t h ~ s elosotions,

the

made thoroughly familiar w i t h the equipment and i t s operation during t h i s testing period.

FUE

t CO NC E K T R A T E I N J E C T If;

W. 6 . Cobb A i r craft Weac for

NO Z 7 i..F

w. R. Iduntley

E ng i nee: i ng Div I s ion

Pests were rtiade on a resistance-heated fuelconcentrate i n j e c t i o n nozzle t o determine t h e effectiveness of the heating i n thc region where the nozzle w i l l enter the r e l a t i v e l y c o l d pump tank cover plate. These tests showed that a ternperaturc of 1300°F could be maintoitred i n the injection p a high-current tube by using about 140 ~ m from transformw, T h i s nozzle t e s t w a s considered satisfactory t o meet the operating requirements for the ARE enrichment system. A R E UhlLCPsbDlNG E X P E R I M E N T

J. Y. Estabiook Aircraft Reactor Fngineering D i v i s i o n The mockup of the ARE unloading apparatus was tested twice. I n the f i r s t test the storage tank was f i l l e d w i t h 950 Ib of NaZrF,, which was successfully unloaded i n 90-lb increments i n t o aluminum cans. I n t w o cases, however, the Sottoitis of the aluminum cans meltcd and the fluoride mixture leaked out. Modifications made i n the apparatus before the second test included improwenlents i n the pouring spout, the level indicators, and t h e pressure measuring equipment. In addition, small amounts of the fluoride mixture \were frozen i n the bottoms of the aluminum C Q ~ Kto minimize the p o s s i b i l i t y of melting these containers. For the second test the apparatus W Q S f i l l e d w i t h 500 Ib of NaF-ZrF, (SO-SO mole %), arid this material was transferred a t 1200 t o 1250°F t o the aluminum cans i n 90- to 9O-lb incieansnts without d i f f i c u l t y . Attempfs t o transfer at 1100 t o 115OoF, however, led to occasional freezing a t the pouring spout and indicated that ~ m l e p ~much s more r e l i a b l e ternpcsatwt B control could he effected, operation 0: t h i s low teinperature could hoidly bc considered t o be feasible.

AWE F U E L R E C O V E R Y A N D R E P R O C E S S I N G

none of

the parts necessnry for the loading and sampling operations have yet been installed. T e s t i n g of the equipment after i n s t a l l a t i a n and before the f i l l i n g operotion shoiJld require approximately one week. It i s anticipated that ale pcrsonnel to he involved w i t h the loading and sampling operation w i l l be

D. E. Ferguson G. I. Cathers M. R. 5ennett Chemical Technology D i v i s i o n

A fluoride-volatili?y method of processing AHE fuel, based on elemental fluorination of the molten has been investigated becclt~seof NaF-Zi-F,-UF,, i t s attiactiveness from the standpoint of cost, inventory of fissionable ninterial, waste d i sposalr and safC;j” of operation. In preliminary more than 99.95% of the uranium tetrafluoride coiitoiaed i n the f w e l was converted t o the v o l a t i l e hexafluoride w i t h c1 gross beta decontamination factor of 100 to 300. Resublimation of this uranium hexafluoride resulted i n an over-all decontomination facto; of 4000 t o 5000. Since i n the absence of excess fluorine UF, i s soluble i n molten NaF-ZrF,, be dissolved, probably the resublimed product C G ~ accoiding .to the reaction’

UF,

+ NaF

NaF.1!F6

7 ‘

---+NaFBUF, + F, .

The fuel from an aircraft reactor would be reprocessed i n three steps: recovery of the uranium f r o m the used fuel by fluorination, separation of the uranium hexafluoride from excess fluorine by trapping the uranium salt i n a c o l d trap and resubliming the UF, from it, and fabrication of a new reactor fused s a l t by absorbing tlie p a r t i a l l y deconSince uraniumtarninGted hJF, i n molten NaF-ZrF,. zirconium and other a l l o y s can be dissolved a t reosonnhle rates by hydrofluorination in a fused s a l t of t h e No%-zrF,-KF type,5 t h i s a l s o provides a imethnd of processing heterogeneous reactor fuels. A decontamination factor of 4000 t o 5000 i s s u f f i c i e n t i f the fuel is to be recycled directly t o a $usad s a l t reactor, but, i f the uranium i s t o be returned t o a diffusion plant or to a heterogeneous reercqor, more decontaminatisn i 5 needed. In further experimeniol work, u decontamination factor of approximately 20,000 w a s obtained by scrublaing the fission-product fluorides from the

2G. I.

Uraaium

~~~

Cathers, Recouery atiJ Decontamination of $om h‘used Fluoride Fuc!s b y FIuorination,

ORNL-17d9 (June 9, 19541, 3F. N. Biowdgr, G. I. Cathers, D. E. Furguson, and E. 0. Nuriiii, A N P Quar. Prog. R e p . Mar. 10, 1954, ORNb_-!692, p 42, ,H. Martin, A. A l b e r s and H. P. Dust, Z. unorg. a l l g e m , Chern. 265, 128 (1951).

5R. E,

municotion.

LLeuze

and C.

Schilling,

personal corn-

PERIOD ENDING SEPTEMBER 70, 1954 uranium hexafluoride w i t h a molten NaF-ZrF, s a l t bath (at 65OOC) i n which the UF, i s insoluble in the presence of excess fluorine. The uraniurn loss in the scrubbing step was less than 0.001%. Fractional d i s t i l l a t i o n of the UF, from t h i s step, which would be performed i n order t o obtain complete decontamination, could be carried out in only l i g h t l y shielded equipment. T w o experiments on scrubbing UF, w i t h molten NaF-ZrF, were carried out. In one, un acetone-dry i c e cold trap was used between the fluorination and scrub steps; the uranium hexafluoride was condensed (and thus separated from excess fluorine) in the c o l d trap and was then resublimed from it. T h e uranium was recovered from the scrub salt by

TABLE 1-11.

refluorination.

I n the second experiment no c o l d

trap wos used, but the decontamination obtained

It i s thought that was the same (Table 1.1). resublimation of the salt-scrubbed product would achieve additional decontamination. Synethetic ARE fuel was prepared for the two runs by dissosving a 224-day-irrodiated, 30-daycooled, 6-9 miniature uranium s l u g in NaF-ZrF, (56-44 mole %) a t 650째C by passing HF through the molten bath. The f i n a l NaF-ZrF,-UF, mixture had a composition of 54.0-42.4-3.6 mole % and a c t i v i t y of 5 x lo6 beta counts per minute per milligram of uranium. The dissolution and subsequent fluorination were carried aut i n the same n i c k e l reactor. i n each run the scrub s a l t was

DECONTAMlNATlON OBTAINED BY MOLTEN SALT SCRUBBING O F UF6 IN THE FLUORIDE VOLATlLlT'Y PROCESS

Synthetic

ARE fuel;

5 x 10 counts per minute p e r m i l l i 650째C w i t h elemental fluorine g a s at f l o w

beta a c t i v i t y of

gram of uranium; fluorinated at r a t e of 66 rnl/rnin

Run A:

F l u o r i n a t i o n period o f

3 hr; UF 6

p l u s e x c e s s fluorine p a s s e d di-

r e c t l y t o scrub salt f r o m the fluorination step

Run B:

F l u o r i n a t i o n period of 5hr;

UFg

condensed i n cold trap andrcsub-

limed from i t prior to scrubbing and recovered from s c r u b by refluorination B e t a Decontamination F a c t o r s ..........................................

Act; vi ty

Run A

........

............................

Scrub

Over-al

Gross

___ Run B

Scrub

Oves-all

2 x IO4*

x 10,"

750

>7 x I O 4

1 x IO5

ldb

700

2 x IO4

TRE

130

' ~ 3 x 16

>8 Uranium L o s s e s

(X of

l*4

106

IO5

I n i t i a l Charge)

3 x 0.74

In r e s u b l i m a t i o n t r a p

In scrub salt

104

430

I n fluorination s a l t

1.4

0.23

40-3

*Calculated o n t h e gross Ru, Zr, bib, and 'TRE a c t i v i t y rather khan on * o t d activity, w h i c h included much d u e to I This does n o t m a t e r i a l l y a f f e c t comparison with p r u w i o i ~ sgross beta decontcmination datu,

131

ANP QUARTERLY PROGRESS REPORT 67 g of an NaF-ZrF,

mixture of the same compos i t i o n as that used in preparing the i n i t i a l ARE fuel. Both fluorination and scrub operations were carried out a t 65OOC. The scrub decontamination factors given i n Table 1.1 were calculated on the basis of the a c t i v i t y remaining i n the scrub salt. The gross beta decontamination factor of 20,000 obtained i n both runs i s about fivefold better than the 11000 t o 5000 obtained previously, The decontamination factors for individual a c t i v i t i e s are somewhat uncertain because of the high amount of 1131 in the product from the short-cooled slugs. The conversion of UF, t o UF, i n Run A was r e l a t i v e l y poor, with 1.6% remaining i n the fluorination salt. T h i s was due t o a fluorination period of only 3 hr. A longer fluorination period of 5 hr lowered the loss t o 0.003%. The some time was used i n refluorinating the scrub salt of Run 5 t o Since recover 98.5 wt % of the i n i t i a l uranium, less than of the uranium passed through the scrub bath i n the absence of fluorine, quantitative reconversion of UF, to UF, by the reaction

UF,

NaF

--+

NaF-UF,

--+ NaFJJF, + F,

i s indicated. ARE

PUMPS

rication and Testing

W. 6. Cobb

A. G. Grindell

Aircraft Reactor Engineering D i v i s i o n F i v e of the s i x rotary elements fabricated for ARE pumps have passed a l l acceptance tests. The s i x t h rotary element is being used i n the hot shakedown test stand t o test welded-fabricated Inconel impellers and for other miscellaneous testing. Unsuccessful attempts were made by ORNL and by an outside vendor t o lap the bronze bearing wear surfaces of the upper o i l seals for the shaft t o a n acceptable flatness (three wave bands of helium light) w i t h a water soluble scouring agent, Bon-Ami. However, acceptable flatness was attained w i t h convent iona I lapping compounds. Four ARE pumps w i t h welded-fabricated impellers The impellers have been installed i n the ARE. used were four of the s i x welded-fabricated impellers that have passed a l l acceptance tests. One of the impellers i s being used i n the pump i n the reactor system component test loop at K-25 and one i s being held as a spare.

Sea I- Gas

P re ssure- Ba Ionci ng

System

W. G. Cobb Aircraft Reactor Engineering D i v i s i o n

A seal-gas pressure-balancing system was developed and has been installed i n the ARE pumps. T h i s system automatically maintains agas pressure i n the lubricant chamber around the lower mechanical seal that i s s l i g h t l y higher ( $ t o 2 psi) than the pressure i n the pump tank volume i n order t o impede the outward leakage of vapors into the lubricating-oil system. T h e equipiiient involved consists of a differential supply valve, a differential vent valve, and the requisite pipe connections. I n addition t o providing a master-slave relationship for maintaining the pressure differential across the seal, the system has two charact e r i s t i c s which are advantageous t o ARE pump operation, The f i r s t advantage i s that it w i l l provide automatic bleed-off of excessive o i l system pressures created by the gas which will be evolved from irradiation of the lubricating oil. The second advantage i s that i t w i l l maintain a greater 1 t o 2-psi) pressure across the lower seal even when subatmospheric pressures are introduced t o the pump tank, for example, when the ARE fuel system i s being vacuum filled.

(4-

adiation Damage to Pump Drives

A. G. Grindell W. G. Cobb Aircraft Reactor Engineering D i v i s i o n Some alternatives t o the usual V-belt pumppower transmission system were investigated i n ari attempt t o find a system that would not be affected by radiation. The alternatives studied included direct in-line drives, angle-gear-box drives, silent-chain

drives,

and improved V-belt drives.

T h e use of direct drives would have required modification of the concrete shielding above the pump pits, and the use of the angle-gear-box drive would have required development of suitable lubricant seals. The silent-chain drive tested did not have the required mechanical r e i i a b i l ity.6 T w o improved V-belts, a Browning 5-105 Supergrip b e l t (rayon cords in a neoprene matrix) and a Browning B-103 Grip belt (cotton cords i n a rubber matrix), were irradiated, under operating tension, through a short length of the b e l t by a C o 6 0 source t o t o t a l doses of lo6 and 10' rep, respectively. 6A. G. Grindell, M o r s e S z l e n t Chain D r i v e s on A R E P u m p H o t Shakedown T e s t Stand, O R N L CF-54-7-166 ( J u l y 16, 1954).

PERIOD ENDING SEPTEMBER 70, 1954 After irradiation each b e l t was used i n the cold shakedown test stand t o drive a pump rotary element at 1600 rpm w i t h no pump load. After 90 hr of operation, the B-105 b e l t showed no v i s i b l e damage and was returned i o ARE personnel. After 75 hr of operation, the 8-103 belt showed no v i s i b l e

damage reactor a t 800 service length

and was transferred t o the pump i n the component test loop pump t o transmit 3 hp rprn. This belt, which has naw been i n for more than 900 hr, has increased in approximately 1 in. F a i l u r e of the belt

does not appear t o be imminent, and pump operation has not been affected. A n idler pulley i s used to maintain tension. On the basis of these tests, it was decided t o use the cotton-corded, rubber-matrix V-belts on the fuel and the sodium pumps i n the ARE. A s an additional safeguard, the b e l t on the fuel pump drive w i l l be shielded w i t h lead. Radiation Damage t o Shaft Seal

W. G. Cobb J. M. Trummel Aircraft Reactor Engineering D i v i s i o n W. W. Parkinson Solid State D i v i s i o n The rotary subassembly of the ARE pump contains s i x Buna-N O-ring static seals which w i l l receive radiation during reactor operation. T o get information on the expected l i f e of these seals, an O-ring sealing arrangement similar t o two of the pump seals was constructed and was placed i n T h e test assembly included s i x the LlTR flux. radial O-ring seals and three gasket O-ring seals. One side of the seals was supplied w i t h o i l at 75-psig pressure, and the other side was vented t o the reactor off-gas system through bubblers and a catch tank so that it was possible to detect any appreciable leakage by the seals. Since the seals

TABLE 1.2.

were vented i n four groups, it WQS possible to get some information on variations i n seal life. Differences i n the following might produce such vari(1) f l u x density over test assembly, ations: (2) t w i s t i n g and r o l l i n g of t h e rings during assembly for testing, (3) the O-rings, and (4) the machined parts. The gamma radiation intensity a t the test location i n the LlTR was estimated by extrapolation from data taken on the B u l k Shielding F a c i l i t y s 7 T h e neutron f l u x was negligible. T h e results are reported i n Table 1.2, which gives time of irradia t i o n t o f i r s t observed leakage and the total estimated dose t o f i r s t leakage. Examination of the 0 - r ings after irradiation revealed substantia I hardening. No appreciable e l a s t i c recovery of the rings occurred upon removal from the container. T h e O-rings t o be used i n the ARE w i l l not be i n s o h i g h a flux a s t h a t used for these experiments, and they w i l l be under much lower pressures. Since i t i s not essential i n the ARE application that the O-rings retain their resiliency, the results of these experiments do not have immediate significance. It i s evident from these tests, however, that O-rings cannot be used successfully i n future hi g h-power reactor app Iicat ion s, Motor Test

A. 6 . Grindell W. G. Cobb Aircraft Reactor Engineering D i v i s i o n A test of an ARE-type d-c motor operating i n dry helium Q t 130 to 14OOF was terminated after 4000 hr of uneventful operation. Examination of the s i Iver-impregnated graphite brushes used during the test indicated that t o t a l wear was limited to -_--

'1-.

Bulk 1954).

I _ l _ _ _ _ l _

E. Hungerford, The Skyshzne Enperzrnrnts ut the Sbzeldzng Reactor, ORNL-1611, p 24 ( J u l y 2,

RESULTS OF RADIATION DAMAGE TESTS OF RUBBER O-RING SEAL5 T i m e to L e a k a g e

Dose

(hr)

Leakage

(rep)

T w o radial ring s e a l s on bottom piston

128

5 x 108

Two radial ring seals o n middle piston

167

6.7 x lo8

One radial ring on top piston

259

Three gasket seal rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

259 __ ...................

lo9

1 1 ~

.................-__ ......

109 .

ANP QUARTERLY PROGRESS REPORT approximately 1/,6 in. It i s concluded that no problem w i l l be encountered in the use of similarly equipped motors i n the ARE.

R E A C T O R SYSTEM C O M P O N E N T TEST L O O P Operation of Leap

W. 6 . Cobb A. G. Grindell Aircraft Reactor Engineering D i v i s i o n D. 5. Trauger

G. A. Kuipers K-25

Technical Division,

Operation of the reactor system component test Q t K-25 was temporarily halted after 1047 hr of operation t o remove one ARE core hairpin tube for metallographic examination. Shutdown, tube removal, tube replacement, and re-startup were accomplished under a complete helium blanket t o prevent air from contaminating the fluoride mixture or the inside of the lnconel system. A n irradiated V-belt that was installed, as described above, after 875 hr of system operation ha5 operated for more than 900 hr of the approximately 1800 hr of system operation. Plugging of gas line connections t o the pump tank by ZrF4-vapor condensate has occurred periodically. I n some instances the plugs have been removed by heating the gas l i n e

loop'

Examination of Hairpin Tube

G. M. Adamson R. S. Crouse Meta I Iurgy D i v i s i o n Metallographic examination of the hairpin tube showed normal subsurface void attack. The maxiilium penetration was t o a depth of 5 mils. Therefore i t appears that corrosion from impurities w i l l not be a serious problem i n the ARE. Since this loop was not precleaned, the corrosive conditions should have been worse than those i n the ARE. No information on the amount of corrosion that

may be expected from mass transfer could be obtained because the loop i s being operated isothermally. R E A C T O R PHYSICS

W. K. Ergen R. R. Coveyou

M. E.

LaVerne

C. B. M i l l s

A i r c r a f t Reactor Engineering D i v i s i o n R. Bate C. S. Burtnette United States A i r Force The Eyewash code, developed by the ORNL Mathematics Panel for use on the UNIVAC, yielded the following results for the ARE: c r i t i c a l mass, 30 Ib of U235;UZ3' required t o y i e l d the 4% excess

involved and blowing the vaporized ZrF, back i n t o the pump tank. No completely satisfactory solution to this problem has been found to date; t h e use of a large-capacity vapor trap is being studied as a possible solution.

r e a c t i v i t y needed for operation, 35 Ib; k e f f for 40 Ib of U235,1.07; c r i t i c a l mass without controls, safety, or instrument apparatus, 17 Ib (this emphasizes the h i g h cost of these materials); h k / ( h ~ / ~ ) = 0.232; temperature coefficient applicable t o slow

8H. W. Savage e f al., A N P Q u a . Prog. Rep. J u n e 10, 1954, ORNL-1729, p 14.

temperature changes and due t o expansion of the whole reactor and t o thermal base effects, -4.3 x 10-5/0F; reactivity value of U235 in tube bends a t reactor ends, 0.006'in k e f f .

PERIOD ENDING SEPTEMBER 70, 7954

2. REFLECTOR-MODERATED REACTOR A. P. Fraas Aircraft Reactor Engineering D i v i s i o n Four years of work on the ANP Project at ORtdL have led t o the belief that the circulating-fluoridefuel reactor w i t h reflector moderation and a spheric a l - s h e l l heat exchanger can be developed i n t o an aircraft power plant of exceptionally high performance.' T h i s v i e w has been supported i n

recent months by the results of studies by USAF contractors. The question has now became how best t o bridge the gap between the ARE and a prototype aircraft power plant.

I t was proposed about a year ago that a 50-Mw reactor be b u i l t so that the f e a s i b i l i t y of con-

structing and operating a circulating-fuel reflectormoderated reactor could be investigated and i t s performance characteristics, particularly w i t h reference t o control, shielding, heat transfer, and f l u i d flow, could be determined. Preliminary estimates indicated that a power of at least 25 Mw would be required t o disclose the more important control characteristics. Other studies indicated that an output of 50 Mw would be required for the lowest powered aircraft l i k e l y t o be t a c t i c a l l y useful (for radar picket ships, patrol bombers, etc.), w h i l e an output of 150 t o 200 Mw would b e required ( w i t h chemical-fuel augmentation) t o power a strategic bomber. Preliminary reactor test u n i t designs and c o s t estimates indicated that the time and c o s t involved would be roughly proportional t o the rated output of the reactor (largely because t h e s i z e and cost of the heat exchangers, pumps, and heatdump equipment vary d i r e c t l y w i t h reactor power). After much analysis and discussion it was decided that 60 Mw represented a good compromise for the power rating and that such a reactor, t o be c a l l e d the Circulating-Fuel Reactor Experiment (CFRE), should be b u i l t by QRNL, w i t h the a i d of t h e P r a t t & Whitney Aircraft Division. T h i s test u n i t

is t o embody the basic features and proportions that have made the circulating-fuel reactor ott r a c t i v e for aircraft use, but the d e t a i l s of the pumps, radiators, and auxiliary equipment w ~ l not l have t o meet aircraft requirements of size, weight, etc. An operating l i f e of 500 hr, of which a substantial portion should be a t or near 60 Mw, was ' A . P. F r a a s and A. W. Savolainen, O R N L Azrcra/t Nuclcnr Power P l a n f D e s z g n s , ORNL-1721 ( i n p r e s s ] .

considered t o be a desirable goal. The information gained from t h i s project should serve t o provide a sound b a s i s for the design of the f u l l - s c a l e aircraft reactor. A s has been the case with the ARE, the CFRE w i l l require a major supporting effort on the part o f various groups i n the QRNL organization t o obtain much essential information not yet a v a i l able. T h e program on fuel chemistry, for example, w i l l continue along the lines it has followed. The results of research on Loth the basic chemistry and the in-pile and out-of-pile corrosion t e s t s are Metallurgical reobviously v i t a l t o the project. search on promising alloys, welding and brazing techniques, strength properties at various temperatures and i n various ambients, and related work w i l l help greatly t o increase the r e l i a b i l i t y , and p o s s i b l y the operating temperature, of the system. T e s t s under way at the Tower Shielding f a c i l i t y coupled w i t h further L i d Tank Shielding F a c i l i t y t e s t s w i l l provide a more complete basis for the aircraft shield design. Multigroup calculations coupled w i t h c r i t i c a l experiments should provide additional information on the static physics o f the reactor. Much test work in experimental engineering w i l l be required t o validate key features of the design.

DESIGN O F T H E C F R E

A. P. Fraas R. W. Bussard Aircraft Reactor Engineering D i v i s i o n The design of the CFRE i s s t i l l preliminary, but some tentative data can be presented at t h i s

stage. A s presently conceived, t h e CFRE w i l l i nc lude a react or, heat-exchanger, pres sure-she I I , and shield assembly, a s described in previous report^.^^^ A quasi-unit shield structurally and functionally similar t o one suitable for an aircraft w i l l probably be used,4 although the structure w i l l 2A. P. F r a a s , ANP Qunr. Prog. R e p . Mar. 10, 1 9 5 3 , ORNL-1515, p 61. 3 R . W. Bussard a n d A. P. F r a a s , ANP Quur. P r o g . R e p . Dec. 1 0 , 1 9 5 3 , ORNL-1649, p 31. 4 E . P. Blizord and H. Goldstein (eds), R e p o r t u/ the 1953 Sunrmer Thzrldzng Sesrzon, ORNL-1575 (June 11, 1954).

ANP QUARTERLY PROGRESS R E F O R T be simplified

i f simplification proves t o be ex-

pedient. Large blowers coupled t o banks of NaKto-air radiators w i l l be used t o provide heat dumps that w i l l be simpler, more reliable, and easier to control than turbojet engines. While conventional heat exchangers l i k e those of the ARE might be used, experience i n fabricating and testing tubeand-plate finned radiators5 indicates that aircrafttype radiators should cost l i t t l e more and that they w i l l make a much neater, more compact instellotion. Further, they w i l l effect a manyfold reduction i n the NaK holdup i n the system and

will

g i v e a thermal inertia and a NaK transit time es-

sentially the same as those for a full-scale aircraft power plant. The fabricating and operating experience obtained should a l s o prove t o be quite valuable. It appears that d-c motors w i l l be the most practical means for driving the various pumps i n t h e system, but the blowers w i l l probably be driven by a-c motors, w i t h control being effected b y the use of air bypass valves and/or shutters. Key data for the CFRE are presented i n Tables 2.1

and 2.2, and a f l o w sheet for the system i s given i n Fig. 2.1. The maior components are shown approximately to scole on t h e flow sheet, although their arrangement relative t o each other i s purely schematic. The more important plumbing and wiring connections expected t o be mode through the reactor chamber w a l l and t o auxilirrryequipment

are l i s t e d i n Tables

2.3

and

2.4.

CFRE COMPONENT QEVELOPMENT PROJECTS T h e projected design effort for the CFRE has been t i e d c l o s e l y t o a comprehensive series of component development projects. A l i s t of these projects as currently conceived i s given in Table 2.5. It i s f u l l y expected that the need for add i t i o n a l projects will develop as the program evolves, but those l i s t e d should provide the most important information required. T h e f i r s t project was completed recently, and the results are presented below; work on some of The t e s t t h e other projects i s w e l l under way. of beryllium in contact w i t h sodium and lnconel under thermal stress was given high priority because the results were needed in the determination o f the amount of poison t o be expected in t h e With the amount of poison i n the rereflector.

5W. F a r m e r et al., Preliminary Design and Perjonnance o/ Sodiuni-tu-Air Radiators, OBNL-1509 (Aug.

3, 1953),

flectar known, the reactor c r i t i c a l mass, the core size, the fuel concentration, etc. could be determined. In the second test, which i s w e l l under way, the hydrodynamics of the pump-volute, plenum-chamber, and core system are being studied. Work on the same basic problem i s also being done by the Heat Transfer and Physical Properties Group and by *he Pratt & Whitney Aircraft Div i s i o n . It i s hoped that the different approaches o f the three groups w i l l lead t o a thoroughly sound I he objective of the t h i r d project i s t o solution. produce a plastic model of a pump and header-tank

arrangement that will remove xenon from the fuel. Data on the solubility of xenon i n the fuel and on

t h e rate of production of xenon i n the fuel are being obtained by the Fuel Chemistry end Radia t i o n Damage groups; the data thus obtained w i l l provide a measure of the efficiency of xenon removal by the pump. An in-pile test of a f u l l - s c a l e pump may be required eventually. Xenon removal by the pump i s very important, from the control standpoint; in addition, if it can be effected as plonned, the b u l k of the other v o l a t i l e f i s s i o n products w i l l also be removed. T h i s w i l l dispose of a maior potential hazard i n the event of a reactor failure and should simplify the reactor installation design problem. The

infotmation given i n Table

2.5 for most

of the other component development projects i s

largely self-explanatory. T o expedite the work, some of the projects can be undertaken by other organizations; for example, another organization might undertake the entire job of designing, developing, arnd fubricating the fill-and-drain system, the fuel addition equipment, the control rod drive mechanism, or the sample heat exchanger tube bundle. The last project listed, the Zero Power U n i t (ZPU), deserves special mention. Experience w i t h the ARE has demonstrated that many problems w i l l arise i n the fabrication of the CFRE ond that there w i l l be a number of doubtful items, especially welds, that could hardly be trusted i n a high-power reactor. T h i s seems particularly true of the reactor core, heat exchanger, and pressure shell assembly. If it i s accepted that the f i r s t attempt a t fabricating the u n i t w i l l leave much to be desired, planning t o use i t os a hot mockup seems t o be more sensible than spending much time and money trying t o rework and patch it. It can be endurance-tested at tenip2rature without developing

PERlOD ENDING SEPTEMBER ?O, 7954

T A B L E 2.1.

T E N T A T I V E SYSTEM DATA FOR THE CFRE FLUID CIRCUlfS

General

Circuit Fuel

Na K

Fuel-to-NaK heat exchanger Temperature drop (or rise), O F Pressure drop, p s i F l o w rate, c f s V e l o c i t y through tube matrix, fps Reynolds number Over-all heat transfer coefficient, Btu/hr.ft2*OF Fuel-to-NaK temperature difference, O F

3,150 95

400 40

400 35 2.7 8.0 4,600

12.6

36.4

Moderator cooling system Temperature drop (or r i s e ) i n heat exchanger,

Sodium-NaK temperature difference,

Pressure drop i n heat exchanger, p s i Pressure drop i n reflector, p s i Pressure drop i n pressure shell, p s i Temperature r i s e i n reflector, O F

7 36

Temperature r i s e i n pressure shell,

100

32 4 100

Pressure drop i n island, p s i

Temperature r i s e i n island,

100

Power generated i n reflector, kw Power generated i n island, k w

Power generated i n pressure shell, kw

2,040 600 210

1.35

F l o w rate through reflector, c f s

0.53

F l o w r a t e through i s l a n d and pressure shell, c f s NaK cooling system A i r f l o w through NaK radiators. cfrn Radiator air pressure drop, in. H 2 0 Blower power required, hp (total) Blower power required, hp (per fan) T o t a l rodiator i n l e t face orea, ft2

150,000

24 720 180

Shield c o o l i n g system Power generated i n lead layer, kw Power generated i n water layer, kw

132 i4

Pump data Head, f t

Flow,

gpm

Specific speed Suction s p e c i f i c speed impeller speed, rpm Drive, hp

50 600 3,750 13,200 2,700 25

250 430

280 2,600

200

ANP QUARTERLY PROGRESS REPORT T A B L E 2.2.

REFLECTOR-MODERATED REACTOR POWER P L A N T SYSTEM D A T A

.II__.__... ._._..

200-Mw Full-Scale

60-MW C F R E

Aircraft NaK i n rodiator cares, I b Header drums Radiator tubes Outlet lines Inlet lines

128 128 800 900

40 40 240 270

450

135

NaK i n header tanks, I b

2300 500

680 150

T o t a l NaK i n complete power plant, Ib

2800

84 0

12.6

NaK t r a n s i t time through system, sec

1.25

NaK t r a n s i t time through intermediate heat exchanger, sac

0.23

NaK t r a n s i t time through radiator tubes, sec

0.067

F u e l in core c i r c u i t (flowing), f t 3

3.9

Fuel i n header tank, ft3

0.4

3 F u e l i n core, f t

1.8

F u e l f l o w rate, s f s

12.0

2.7

F u e l c i r c u i t t r a n s i t time, sec

Os5

0.7

Intermediate h e a t exchanger NaK f l o w i n g i n circuits, Ib

NaK flow r a t e through rodiotor c i r c u i t s (for a 4OO0F

AT), cfs

Thermal c a p a c i t i e s of power plant, Btu/OF

L i n e s and pumps

300

210

2000

720

600 250

NaK (flowing) Intermediate heat exchanger T o t a l for NaK systems Fuel T o t a l for power plant, Btu/OF

nuclear power. Even i f a leak should develop i n one of the welds, there would be no serious contamination problem. Upon completion of the endurance test, it should not be d i f f i c u l t t o make a something that thorough inspection of the parts could hardly be done after sustained operation at high nuclear power. At f i r s t thought, the ZPU might appear t o delay the program, but closer inspection indicates that i t may actually save a substantial amount of time, because efforts t o get it into operation can proceed at f u l l speed without the inevitable hazards, re-

1700

180 75 180 75 ____ 510

600 250

Radiator cores

straints,

inhibitions,

and hesitations that would

accompany high nuclear power operation. Work on the CFRE can proceed a t a inore deliberate pace, and t h e mistakes disclosed by work on t h e ZPU can be avoided. The degree t o which the ZPU w i l l be a mockup of the CFRE w i l l depend upon the problems that develop i n the course of t h e It i s hoped that a 1 0 4 w component t e s t work. gas-fired furnace can be used, in place of one of the two CFRE heat dumps, i n making some heat transfer tests on the fluoride-to-NaK heat exchanger and in obtaining a temperature gradient

PERIOD ENDING SEPTEMBER 70, 1954

, LL

ANP QUARTERLY PROGRESS REPORT

TABLE 2.3.

WIRING, PIPING,

A N D TUBtNG CONNECTIONS THROUGH REACTOR CHAMBER WALL ....

_.._____I

i___l_

Thermocouples

52 8 8 12

D u p l i c a t e thermocouples t o a l l pipes leaving s h i e l d D u p l i c a t e thermocouples t o main pipes entering s h i e l d Single thermocouples t o o i l i n l e t and outlet l i n e s t o pumps Na and f u e l temperatures

12 12 12 24

Pressure s h e l l temperotures Gamma-shield s h e l l temperatures Shield-water temperatures F u e l and Nu heut-dump system temperatures

8 4 12

Pump d r i v e motor temperatures

Control rod d r i v e temperatures Miscellaneous support structure and leak warning temperatures Total

164

Miscellaneous Instrument Wires

Nuclear instruments

Pump speed Pawer Wiring

Pump d r i v e motors

12 3

(4)

Control rod actuator (1)

2 1

L i g h t s and receptacles F u e l a d d i t i o n machine Total

Tubing Ua and f u e l dump v a l v e octuotors* ( \ Expansion-tank pressure* ( \ L i q u i d - l e v e l gages* Main NaK l i n e s

(1/4

in.

10 2 2 4

OD)

in. OD)

in, OD)

(5 in. OD)

(2 in. OD) Gamma-shield c o a l i n g water (1 in. OD)

4 2

Moderator-cooling NaK l i n e s

Pump c o o l i n g and l u b r i c u t i n g o i l

( 14 in.

4 1

OD)

Xenon vent

Atmosphere vent

H e l i u m supply Total

__.. * T h e s e tubes c a n be replaced by wires.

.........

32 ...____---

PERIOD ENDING SEPTEMBER 70, 7954 TABLE

2.4.

P R I N C I P A L WIRING, PIPING, AND TUBING CONNECTIONS

TO AUXILIARY EQUlPMEMT ...................

Thermocouples

32 8 6 12

D u p l i c a t e thermocouples t o NoK m a n i f o l d a t o u t l e t s of a l l radiator cores D u p l i c a t e thermocouples t o main NaK p i p e s entering radiator cores Single thermocouples t o i n l e t and o u t l e t l i n e s to o i l coolers Radiator air i n l e t and o u t l e t temperatures

Shield woter i n l e t and o u t l e t temperatures

NoK dump system temperatures

NaK pump d r i v e motor temperatures

NaK pump temperatures

Miscellaneous support structure and leak warning temperatures Total

Miscellaneous Instrument Wires

20 6

Nuclear instruments NaK flow meters

4 3 3

NaK pump speed

O i l expansion tank l e v e l i n d i c a t o r s Shield water expansion tank level i n d i c a t o r s

I _

Total Power Wiring

NaK pump d r i v e motors (2) O i l pump d r i v e motors (3) Shield water pump d r i v e motor

6 6 2 8

(1)

Main blower d r i v e motors (4) Moderator-coaling system blower d r i v e motor (1)

Preheat burner d r i v e motors

(3)

Total Tubing NaK dump-valve actuators

(14 in.

OD) OD)

Expansion-tank pressures ( \ in. NaK l i q u i d - l e v e l gages Main NaK lines (5 in.

(14 in.

OD)

Moderator-cooling NaK l i n e s Gamma-shield c o o l i n g water

10 10 2 2

OD)

No and fuel dump-valve actuators

in. OD)

4 4 2

(2 in. OD) (1 in. OD)

Pump c o o l i n g ond l u b r i c a t i n g o i l

1 in. (4

OD)

Xenon vent t o buried charcoal bed

H e l i u m supply to expansion tanks

___

Tmal .....

..............

....

in the ZPU system, While t h i s arrangement would hardly represent normal operation, i t would simul a t e the one-engine-out condition rather well. It i s nut yet clear just what w i l l be done, but every effort will be made to get as much information as possible from the ZPU, consistent with the money and the time available. REACTOR PHYSICS

W. K.

Ergen M. E. LaVerne R. R. Coveyou C. B. M i l l s Aircraft Reactor Engineering D i v i s i o n

C. S. Burtnette United States A i r Force

R. Bate

The three-group, three-region code for reactor statics calculations6 on the ORACLE has been completed. The temperature effects for the CFRE were computed on the UNIVAC for three reactors: a 50-Mw reactor w i t h lnconel coolant tubes i n the reflector, a reactor without these tubes, and a reactor i n which the lnconel in the core shells was replaced by columbium c l a d w i t h 0.010 in. of lnconel The temperature coefficient for r a p i d temperature changes turned out t o be about -3.5 x 10” 5/oF, and the temperature coefficient for slow effects i s -2.6 x 10-5/”F for the heavily poisoned reactor w i t h the lnconel coolant tubes, but it in-

creases t o -3.3 x I O - ~ / O Fatid -3.8 x ~ o - ~ P F , respectively, far the reactor without the lnconel coolant tubes and the reactor w i t h t h e Inconel-clad coluinbium core shells. R E A C T O R C A L C U L A T ION S

M. E. LaVerne Ai rcr a f t Reactor Eng i neer i ng D Iv isi o t i C. S. Burtnette United States A i r Force Results of the parametric reactor study done on the lJNlVAC a t the AEC Computing F a c i l i t y i n New York have been pub1 ished,7*8 and the CurtissWright Corp. has completed, under contract t o the United States Air Force, a series of multigroup, multiregion calculations for the ORNL-ANP Proi6W. K . Ergen, A N P Quur. Prog. R e p . J u n e 10, 1 9 5 4 , ORNL-1729, p 32. ’C. 5. Burtnette, M. E . Laverne, and C. B. Mills,

Reflector-Morierated-Reactor D e s i g n P a r a m e t e r S t u d y , Part I , E f f e c t s of Reactor Proportzcnis, O R N L CF-547-5 (to he published).

‘M. E. LaVerne and C. S. Bwrtnette, A N P Quur. Prog. Rep. J u n c 10, 1954, ORNL-1729, p 32.

ect. A report on the results of the latter calculations i s being prepared and w i l l be issued by the Curtiss-Wright Corp, A redetermination of the c r i t i c a l mass for t h e

rhombicuboctahedron c r i t i c a l assembly (CA-19) was found t o be necessary because of an error in carbon transport cross-section scaling, a difference i n reflector geometry from that assumed for

the original calculations, and an actual Teflon density lower than that used i n the calculations. In addition, a small correction (about 1% of c r i t i c a l mass) was made for the presence of the aluminum l a t t i c e supporting the reactor and for the aluminum control rod sheaths. The eigenvalues of the difference-equation approximation t o the age-diffusion differential equat i o n vary w i t h l a t t i c e spacing. Calculations were therefore made w i t h three different spacings i n order t o evaluate t h i s effect. A f o i l of 0.004 in. thickness W Q S used i n the calculations. The results, corrected as mentioned i n the preceding paragraph, are given i n the following: Space Interval,

-1,

Calculated Critical Mass

(Ib)

0.457

22.75

0.914

2 1.47

1 .a28

20.91

The dotu are plotted i n F i g . 2.2. A Lagrangian extrapolation t o a zero l a t t i c e spacing gives a c r i t i c a l mass estimate of 24.70 Ib. The assembly became c r i t i c a l at a loading of 24.35 Ib of U 2 3 5 . It i s recognized that the agreement between calculation and experiment may w e l l be fortuitous, since the c r i t i c a l mass i s not very sensitive t o errors i n detail. Further evaluation of agreement (or lack thereof) must await additional experimental results. Calculations are now being made for the next c r i t i c a l assemblies, that is, the threeregion assemblies consisting of a beryllium i s l a n d and reflector w i t h a Teflon and uranium-foil fuel annulus. The fuel annulus w i l l be surrounded by core shells; aluminum core shells w i l l be used for one assembly and lnconel for the next. B E R Y L L I U M THERMAL. S T R E S S T E S T

W. Bussard R. E. MacPherson Aircraft Reactor Engineering D i v i s i o n

One of the key questions in the design of t h e

CFRE, as mentioned above, has been that of the

TABLE 2.5. SUMMARY OF CFRE DEVELOPMENT PROJECTS Estimated Proiect Number

Objective

Proiect T i t l e

References

Number of

Status September

1, 1954

Man Hours B e r y l l i u m Thermal Stress

I n v e s t i g a t i o n of b e r y l l i u m under severe thermal stress

2280

ORNL-1515, -1517, and -1721

F i r s t t e s t completed; no cracks appeared i n b e r y l l i u m speci-

conditions

men

Care Hydrodynamics

I n v e s t i g a t i o n of f l a w separation, v e l o c i t y d i s t r i -

3880

ORNL-1515, t o Briant,

Xenon Removal Pump ( p l a s t i c

Development of a t t i t u d e - i n s e n s i t i v e pump and header

Xenon Removal Pump (hat test

Performance and c a v i t a t i o n tests w i t h water,

6/19/53

ORNL-1649, ORNL CF-54-5-1; Memo, Fraas t o file, 8/4/54

T e s t s under way w i t h water

5200

ORNL-1649; Memo, Fraas to f i l e , 8/4/54

D e s i g n completed; f a b r i c a t i o n under way

sodium, and fuel; endurance t e s t

unit) Model-T Component T e s i (isothermal)

Check o f f a b r i c a b i l i t y and endurance t e s t of f u e l

Tests under woy w i t h water

Memo, Wislicenus

3080

tank geometry t o remove xenon and degas f u e l

model)

-1692, -1701; YF-15-11; Memo,

Fraas t o file, 8/4/54;

bution, and perverse particle d w e l l time

8920

Pump, header-tank, and cares h e l l design held pending

pump, plenum chamber, and care s h e l l assembly

r e s u l t s of projects Model-T Component T e s t (100-kw heat input t o care) Second Fuel-to-NaK Heat Exchanger

(4000 kw,

1000-kw input)

lncanel Thermal Stress

Check far f l o w separation; estimate of f u l l - s c a l e

4400

ORNL CF-54-2-35

9280

ORNL-1215,

boundary layer thickness Performance and endurance t e s t a t f u l l - s c a l e pressures, pressure drops, and temperature d i f f e r e n t i a l s

Investigation of effects of severe thermal stress

-1330, -1509, -1729 (P 28); CF-54-1-155; Memo, Ahern t o Fraas, 6/12/54

2680

L e a k Characteristics of Tube-toHeader Welds

F i l l - a n d - D r a i n System

Control Drive Mechanism

I n v e s t i g a t i o n o f p o s s i b i l i t y of self-plugging of weld

D e s i g n of about one h a l f of

components completed; heat

exchanger being fabricated

760

ORNL-1215, ORNL CF-54-1-155

Parts being fabricated

cracks between fuel and NaK Performance t e s t w i t h water, NaK, and fuel

2280

D e t a i l design completed for one proposal

Check of fabricability; shokedown and endurance

3200

F u e l A d d i t i o n System

Check of f a b r i c a b i l i t y ; shakedown and endurance

3200

Woad and P l a s t i c Reactor Mockup

Check of f a b r i c a b i l i t y and t e s t demonstration

960

Sample Heat Exchanger Tube

Check of f a b r i c a b i l i t y and c o s t determination

2240

F a b r i c a t i o n under way

D e t a i l d e s i g n completed far one proposal

Bundle Zero Power U n i t

Prel i m i nary layouts corn pie t e d

for several proposals

(1/2 scoctle)

Preliminary luyouts completed for several proposals

Tests

Same as above

comp leted

tests

and 3

Design of t e s t r i g nearly

c y c l i n g on cracking and d i s t o r t i o n

Check o f fabricability; shakedown and endurance

Preliminary d e s i g n under way

tests

3092

ORNL-LR-DWG

CALCULATIONS EXTRAPOLATION

effects of severe thermal stresses in beryllium at temperatures between 1000 and 1300"F.9~l o There has been serious concern as t o whether cracking or distortion i n the beryllium would prove t o be a major problem. Therefore a test on the beryllium block shown i n Fig. 2.3 was devised to investigate t h e effects of power (and hence thermal) cycling. T h e test loop used i s shown in Fig. 2.4. Heat was generated i n the beryllium block by simple electrical-resi stance heating w i t h a current of about 15,000 amp through the beryllium block. Sodium flowed from an electromagnetic pump t o the block inlet header, upward through the lower

0 2

A r (cml

9R. W. Bussard e t a l . , The Moderator Cooling S y s t e m lor the Reflector-Moderated R e a c t o r , ORNL-1517 (Jan. 22, 1954).

Effect of Space Interval Size on Fig. 2.2. puted C r i t i c a l Mass.

Fig,

2.3.

Com-

'OF. A. Field, Temperature Gradient und Thermal S t r e s s e s in Heat Generating B o d i e s , ORNL CF-54-5-196 (May

21, 1954).

Beryllium Test Block.

ANP QUARTERLY PROGRESS REPORT

UNCLASSIFIED ORNL-LR-DWG 3093

-kvo

TRANSFORMER

\ SURGE T A N K -

111 ' ~ ~ ~ / ~ A I R - C O OCOPPER L E D BUS BAR

ELECl

Fig.

2.4.

Beryllium Thermal Stress T e s t Apparatus.

tube bank t o the corresponding holes in the ber y l l i u m block, through t h e upper tube back into t h e outlet header, out t o t h e radiator section, and back t o t h e pump. Banks of tubes were used to duct the sodium t o and from t h e block t o minimize t h e f l o w of electrical current through the sodium A bypass f i l t e r i n l e t and outlet header tanks. arrangement was provided, a s w e l l as bypass c o o l i n g f l o w t o the sodium-filled lugs connected t o the transformer bus bars. T h e high-power-density volume heat source coupled w i t h transverse sodium f l o w through the d r i l l e d holes gave a high thermal gradient around the holes. During t h e course of the t e s t the power density was cycled regularly by alternating the

operating conditions as shown i n Table 2.6, which a l s o compares the operating conditions for t h e thermal stress t e s t w i t h those for t h e CFRE. T h e t e s t included an i n i t i a l 100-hr period a t a constant high power density followed b y power c y c l i n g at the rate of one c y c l e per day. The changes from one power level t o the other were,

i n a l l cases, performed at a f a i r l y uniform rate i n a 1-min interval. The test was concluded after 1000 hr of operation, including 36 cycles. T h e beryllium sample i s now being examined by the Metallurgy D i v i s i o n for dimensional s t a b i l i t y and evidence of mass transfer, corrosion, or erosion. V i s u a l inspection of the sample after the test indicated that no distortion, cracking, or erosion of the beryllium had taken place.

PERIOD ENDING SEPTEMBER 70, 7954 T A B L E 2.6.

COMPARISON OF OPERATING CONDITIONS FOR B E R Y L L I U M THERMAL STRESS APPARATUS AND CFRE Beryllium Block

CFRE

Power l e v e l

High

Low

Peak power density i n beryllium, w/cm3

Average power density i n beryIltum, w/cm3

1e 5

Cooling passage diameter, in,

0,25

0.25

Cooling passage spacing (minimum), in,

1u 5

1 .o

108,000

1.5 38,000

42,000

1,040

1,060

1,050

1,100

1,080

1,150

1,070

1,200

1,110

C a l c u l a t e d thermal stress (no plostic flow), psi Sodium i n l e t temperature to beryllium,

Sodium outlet temperature from beryllium,

Average sodium temperature through beryllium,

Maximum thermocouple reading i n beryllium, C a l c u l a t e d maximum beryllium temperature,

1,280

T i m e a t power level, hr/doy

14,300

H e a t i n g current, amp

B E R Y LL I UM-IN C ON EL -SOD! U M S Y S T E MS The compatibility of sodium, beryllium, and Inconel has been studied by using whirligig, seesaw, f orced-c ircu I at iontherma I -convect ion- Ioop, and loop tests. The results obtained t o date indicate that it w i l l be possible t o use beryllium unclad in the CFRE i f the temperature of the sodiumberyllium-lnconel region (that is, the reflector and the island) i s kept below 1200째F. The main effect observed thus far i n the t e s t s has been the type of mass transfer usually found when dissimilar metals are used i n a system. In t h i s case the mass transfer has been evidenced by the a l l o y i n g of beryllium w i t h the lnconel w a l l s i n regions where the beryllium and lnconel were close together, w i t h only stagnant sodium separating them.

Beryl I ium-lnconel-Sodi urn Compatibility Tests

E. E. Hoffman

C. R. Brooks

W. H. Cook C. F. L e i t t e n

Metallurgy D i v i s i o n N i n e tests of the compatibility of beryllium, sodium, and lncanel i n dynamic systems have been In t h e w h i r l i g i g t e s t s a completed (Table 2.7). circular loop of lnconel tubing w i t h a beryllium insert was partly f i l l e d w i t h sodium which was For four of circulated at a velocity of 10 fps.

1,150

1,100 1,230

8,450

these t e s t s there was no temperature differential, but for one test there was a differential of 140째F. These tests were of 300-hr durotion. Two therrnalconvection loops of lnconel w i t h beryllium inserts

were tested for 1000 hr w i t h sodium circulating at a r e l a t i v e l y low velocity (approximately 2 to 10 fpm). In addition, seesaw tests were performed i n which beryllium specimens were retained i n the hot zones of o s c i l l a t i n g Inconel tubes partly f i l l e d w i t h sodium which circulated at a v e l o c i t y o f

fpm.

Macroscopic examination f a i l e d t o show attack on the Inconel, but two specimens showed deposits. Black layers were formed i n the annular space between the beryllium insert and the enclosing lnconel sleeve i n the w h i r l i g i g and thermalconvection-loop tests. T h e layer was very adherent and made removal of the insert d i f f i c u l t ; consequently, the weight change data obtained were not considered as being reliable.

A picture of t h i s layer from a w h i r l i g i g loop Microoperated a t 130OOF i s shown i n Fig. 2.5. scopic examination showed that the maximum depth of attack on the inside surface of the lnconel did not exceed 0.5 mil. Chemical analysis of the sodium bath revealed less than 0.04% beryllium i n the sodium. I n a l l tests, beryllium was detected 29

ANP QUARTERLY PROGRESS REPORT TABLE 2.7.

RESULTS OF- BERIILLIUM-SODIUM-IN60raEL COMPATIBILITY TESTS Concentration of Temperature

Test Type

on Surface

(O F l

(CLdcm Whir Ii g i g

300

Re murk s

o f lnconel Tube

1100 (isothermal)

5.3 t o 7.9

T w o dark deposits on lnconel onalyzed h i g h in hery II ium; macroscopically, bery I l i u m showed fine p i t t e d appearance

300

1200 (isotherma I)

N o v i s i b l e d e p o s i t s on Inconel; b e r y l l i u m snwoth

0.05

but discolored

3 00

1300 (isothermal)

0.07

No v i s i b l e d e p o s i t s on Inconel; b e r y l l i u m showed dark discontinuous deposit covering surface

300

270

1400 (isothermal)

39.6 to 127.0

No v i s i b l e d e p o s i t s on Inconel; b e r y l l i u m surface p i t t e d and p a r t l y covered w i t h s m u l l d e p o s i t s

Hot zone, 1200 C o l d zone,

Loose, adherent gray d e p o s i t found i n c o l d zone,

1060

analyzed h i g h i n sodium; b e r y l l i u m smooth but discolored

Ther ma I

1000

H o t zona, 1150

l n c o n e l appeared smooth and showed no v i s i b l e

C o l d zone, 990

convection

deposits; beryl1iurn smooth b u t

d iscolwed

loop

1000

H o t zone, 1300

l n c o n e l appeared smooth and showed no visib!e

Cold zone, 1130

deposits; b e r y l l i u m specimen smooth and l i g h t gray i n color

Seesaw

300

H o t zone, 1400

H o t zone, 14.0

C o l d zone, 710

Middle, 9.9

B e r y l l i u m specimen covered w i t h black f l a k y

Cold zone, 20.6

300

Hot zone, 1100 C o l d zone, 740

H o t zone, 0.07

deposit; surface of Inconel t u b e discolored Similar to above

Middle, 0.02 C o l d zone, 0.01

on the lnconel tube w a l l surface.

X-ray analyses

of the lnconel surfaces indicated that the beryllium

was present as the metal. In t h e thermal-convection loops, a deposit was formed between the contacting ends of the ber y l l i u m insert and the lnconel tube. X-ray analyses of these deposits and those found in the w h i r l i g i g loops showed beryllium, Inconel, and The l i n e s very similar t o sodium bicarbonate. beryllium specimens from the seesaw t e s t s were covered w i t h a black, flaky deposit, which has been identified by x-ray analysis as beryllium oxide and beryllium. Again, lines similar to sodium bicarbonate were observed. Microscopic examination of the beryllium specimen from a whirligig test showed l i t t l e attack on

t h e inside surface, but attack up t o 5 m i l s was observed on the outside surface which was i n contact w i t h the r e l a t i v e l y stagnant sodium in the 0.005-in. annular space (Fig. 2.6).

M a s s Transfer T e s t s in Thermal-Convection bQOpS

G. M. Adamson Meta IIurgy Di vision

A series of thermal-convection loops i s being operated t o determine the effect of dissimilar metal mass transfer in the beryllium-lnconel-sodium system. These studies are being carried out in lnconel loops w i t h short beryllium inserts i n t h e hot legs; sodium i s the circulated fluid. The

PERIOD ENDING SEPTEMBER JO, 1954

Fig.

2.5.

Sodium for

Inside Surface of lnconel Sleeve E n c l o s i n g B e r y l l i u m Insert E x p o s e d in Whirligig T e s t to a t 1300째F. Etched w i t h glyceria regia. 1OOOX.

300 hr

T A B L E 2.8. E F F E C T OF C I R C U L A T I N G SODIUM A T VARIOUS TEMPERATURES FOR 500 hr I N I N C O N E L THERMAL-CONVECTION LOOPS WITH B E R Y L L I U M SPECIMENS IN THE H O T LEGS .... ...... .................... ......_____ -~ . . . . .

Tempera tu re ......

(OF)

.....

. .

...............

___

__ ............

O u t s i d e surface of i n s e r t showed h o l e s to a depth

____

Intergranular penetrations t o a depth o f

lnconel H o t L e g

D e p o s i t opposite the b e r y l l i u m

of 14 mils; i n s i d e surface s l i g h t l y rough

1300

insert m i l s on

No attack or d e p o s i i

outside surface; no a t t a c k o n i n s i d e surface

1100 -

Metallographic Examination

B e r y l l i u m Specimen

1500

9 00

.........

r e s u l t s reported i n Table

N o attack

layer or a t t a c k

No layer or a t t a c k .....................................

2.8

are from tests oper-

ated for 500 hr. Other loops are now i n operation that w i l l be run for longer periods.

Photomicrographs of the outsides of the beBlack, r y l l i u m inserts are shown in Fig. 2.7. f l a k y material was found i n the c o l d legs of t h e loops operated at 1100, 1300, and 1500째F and

-................ ___

or deposit

N o a t t a c k or deposit

. .

.......

i n the hot leg of the loop operated at 1500째F. T h e deposit material has not yet been identified by spectrographic or diffraction studies. The deposit found o n the lnconel opposite the beryllium insert i n t h e loop operated a t 1500째F i s shown i n Fig. 2,8; it has been identified a s a nickel-beryllium alloy.

ANP QUARTERLY PROGRESS REPORT

. .

- 0 3

Fig. 2 ~ 5 "Beryllium Specimen After hirligig T e s t for 30 hr a t 130OoF. ( a ) Surface exposed t o f l o w i n g sodium. ( b ) Surface exposed to relatively stagnunt sodium i n the annular space between the specimen and the lnconel sleeve. Etched with oxalic acid. 250X. M a s s f:ansfer

in FQrced.&irculacion-8oop Tests

1.. A. Mann Airsraft Reactor Engineering D i v i s i o n F. A. A n d e r s ~ n ~ ~

University of Mississippi

A t e s t unit12 for investigating mass transfer of beryllium t o Inconel in f l o w i n g sodium was operated for 1006 hr under conditions nearly the same o s those proposed for the reflector-moderated reThe approximate conditions of operation, actor. w i t h minor fluctuations, were as follows: sodium f l o w rate, 3 gpm; beryllium temperature, 1200째F; nlcsximu~li temperature, 1245" F (sodium entering economizer); minimum flowing sodium temperature, 900" F (re-entering economizer from cooling section); Reynolds number at minor diameter (0.25 in.) "Summer

Research Purticipont.

12L. A. Mann, A N P ORNL-1729, p 2 2 .

Q~ACZT.

Prog. R e p . J u n e 1 0 , 1954,

of beryllium insert, 155,000; minimum Reynolds number (at coldest point i n economizer annulus),

61,000.

Upon termination of the test, the unit was cooled rapidly t o room temperature t o freeze the sodium. Samples of. both the sodium and the Inconel were then taken from each of 24 locations i n the unit. T h e sodium samples were analyzed chemically for beryllium, and none w a s found. Metallurgical examination of the beryllium piece and the Inconel samples indicated that the beryl Iium-lnconelsodium system should g i v e no trouble from corrosion or mrlss transfer under the conditions presently proposed for the reflector-moderated reactor (temperature not over 1200" F, Reynolds number not over 150,000). The above-described test u n i t was essentially duplicated, and a second test w 0 s started w i t h maximum and minimtim temperatures i n the flowing sodiurii of 1300'F (ut the beryllium) and 1000째F, respectively. A powe: f a i l u r e terminated t h i s t e s t

PERIOD ENDJNGSEPJEMBER 10, 7954 after

268 hr of

operation,

Sodium and Intonel

samples and the beryllium insert were removed for analysis and examination, but results have not yet been received. A t h i r d unit has been b u i l t and i s intended to operate for 1000 hr.

R j

Fig. 2.7. Surfaces of B e r y l l i u m inserts i n lnconef Thermal-Convection L o o p s After Exposure t o Flowing Sodium for 500 hr a t ( a ) llOO째F, ( b ) ? 3 W F , (c) 1500째F. Unetched. 250X. Reduced

36%.

Fig. 2.8.

Layer on lnconef Opposite Beryllium

insert i n Tharmal=Convection L o o p T h a t Circul o t e d Sodium at 1500'F for 500 hr. Unetched. 250X. Reduced 36%.

ANP QUARrERLY PROGRfSS REPORT

3. E

TAL ~~A~~~~

H. Savage Aircraft Reactor Engineering D i v i s i o n

The emphasis i n the engineering work has shifted to

research and component development for the CFRE and for in-pile loops, since most of the preoperational ARE tests have been completed. T w o types of pumps have been developed for inp i l e loop use:. The vertical-shaft centrifugal sump pump, which would be installed outside the reactor shield and thus would require auxiliary shielding, i s now being fabricated iii sufficient quantity to meet the demands of the in-pile loop program and the f orced-c ircu lot ion corros ion testing program. The small (4-in.-OD), air-driven, horizontal-shaft sump pump being developed for insertion i n a reactor beam hole was tested w i t h NaF-ZrF, at 1350'F. Some d i f f i c u l f y was encountered w i t h i n i t i a l priming, but operation was otherwise satisfactory. A new, small pump that hos the required s m a l l holdup volume i s being designed. This pump, which w i l l use a turbine-type impeller, has the advantage that both the inlet and discharge HydruiJlic drive motors of can be a t the bottom. suitably s m a l l dimensions have been found t o be satisfactory drives for these pumps. Additional work has been done on the development of forced-c irculation corros ion-testi ng loops for obtaining information on the corrosion of l n t o n e l in high-velocity turbulent fluoride mixtures with large temperature differentials i n the system. TWO series of loops are being constructed t o meet the following requirements: a Reynolds n u m h r of 10,000 with temperature gradienfs of 100, 200, and 300°F and c1 temperature gradient of 200'6 w i t h Reynolds numbers of 800, 3,000, and 15,000. The maximum f l u i d temperature i s t o be !500°F. A forced-circulation loop i s a l s o being developed for t c t t i n g combinations o f structural metals i n contact w i t h high-velocity turbulent l i q u i d metals under high temperature differentials, Soveral exploratory tests were made of gas burners for use w i t h the proposed gas-furnace heat source for h igh-temperature reactor mockup tests. Also, a study i s under way 06 the cavitation phenomenon associated w i t h operating liqiiid metal systems a t elevated temperatures, high f l o w rates, -~

ID. F. SalrnQn, A N P Quar. Prog. R e p . J u n e 10, 1954, ORNL-1729, p 19.

~~~~~~~

and high pump speeds. A correlation of fluid-flownoted. noise intensity w i t h pressure data The number of stations available for convection

loop testing has been increased from 18 t o 31 so that many more long-term tests (2000 hr or longer)

nnd intensive t e s t s of special materials can be made. The basic design of the convection loops has been simplified, and various means of heating the loops and of making operation of them more automatic are being studied.

IN-PILE L O O P COMPONENT DEVELOPMENT

W. 8. McDonald Aircratt Reactor Engineering D i v i s i o n

A n in-pile loop for insertion i n the MTR i s t o be designed, constrvcted, and operated as a joint ORNL and Pratt & Whitney A i r c r a f t D i v i s i o n project. The loop i s being developed for circul a t i n g proposed fuel mixtures so that the extent of radiation damage to materials of construction and the e f f e c t of radiation on the fuel can be determined. Preliminary QRNL work on t h i s project i s concerned w i t h the design and development of compoiients of cr loop t o operate i n a horizontal beam hole. Further developmental work was a l s o done on the vertical-shaft centrifugal puinp for use w i t h both in-pi l e and out-of-pi le forced-circulation loops.

Horizontal-Shaft §ump Pump

d. A. Conlin D. F. Salmon Aircraft Reactor Engineering D i v i s i o n The air-driven horizontal-shaft sump pump described previously' was operated w i t h NaF-ZrF, at 1350OF. It produced a %psi head a t 6000 rpm Initial difficulty in and n flow rate of 1.5 gpm. priming the pump was solved by momentarily flooding the impeller labyrinth. T h i s pump, which was b u i l t as a p i l o t model for checking design principles, hot pump performance, and reliability, C ; ? for in-pile use and has too large o V O ~ ~ J ~holdup w i l l now be used as a laboratory pump.

A new pump has been designed that has the required small pump holdup volume., For priming, t h i s design incorporates a baffle in the sump tank

P E R I O D ENDING S E P T E M B E R 10, 1954 that w i l l permit remote, momentary flooding of the impeller eye and the labyrinth behind the impeller. A turbine-type impeller i s used instead of the conventional centrifugal i mpel ler.

sequence, the pump can be e a s i l y stopped and restarted. A p i l o t model of the pump, similar t o the pump previously described,2 has been designed that incorporates these features (Fig. 3.1) and an

It has been experimentally determined, w i t h water, that the turbine type of impeller w i l l prime i t s e l f if the i n l e t and outlet are placed a t the bottom o f the pump and the fluid i s maintained a t a l e v e l about h a l f way between the bottom of the impeller and the impeller center. T h e horizontal

integral sump tank connected by a passage to the seal cavity.

sump pump principle i s used, w i t h the sump having a minimum volume that a c t s merely as an expansion chamber and a reservoir to replace and catch the f l u i d leaking past the labyrinth seal. Excessive impeller end clearances are necessary for t h i s type of pump because o f the severe temperature gradients present. However, estimated available pump heads (25 ft a t 4500 rpm and 1.5 gpm) are greater than the estimated requirements. Hydraulic e f f i c i e n c y of the pump w i l l be low but performance w i l l be reliable. Further work has been done o n centrifugally sealed and frozen-fluoride-sealed pumps, and i t has been established that the centrifugally sealed pump requirestoo greata volume of f l u i d for sealing for it t o be considered for in-pile use and that the leakage rate of the frozen-fluoride-sealed pump i s excessive. However, the centrifugally sealed pump, which w i l l be useful for other applications, has been modified t o include a system that makes the pump self-priming and deaerating. B y cont r o l l i n g the seal level and using a proper venting

I n Operation, the system would be f i l l e d t o the The seal c a v i t y and sump tank would then be pressurized through the gas i n l e t tube t o the predetermined pressure necessary t o obtain a p o s i t i v e pump i n l e t pressure during operation, and t h e pump would be started. T h e centrifugal force would cause the l i q u i d in the seal c a v i t y t o form a rotating annulus. T h e annulus of f l u i d would develop a small pressure and thus pump f l u i d into t h e discharge tube, through the loop, and back t o the pump. The small bleed hole from the pump discharge to the sump tank would permit deaeration b y bypassing aerated f l u i d from the pump into the sump. T h e bypassed f l u i d would be replaced by clear f l u i d from the seal. When the pump was primed and deaerated, a continuous f l o w of fluid would pass from the bleed hole t o the sump t o remove any accumulation of gas. Throughout t h i s operation, f l u i d entering the loop from the seal to either f i l l the loop or replace the bleed f l o w would be replaced b y f l u i d from the sump, and sufficient f l u i d would thus be in the seal for sealing. Therefore there i s a lower l i m i t on the run level in the

full level.

ANP QUARTERLY PROGRESS REPORT

UNCLASSIFIED O R N L - L R - D W G 3095

-3

, -ROTARY

SEAL UNION

SPARK PLUG PROBE

-\"

II] N

P R F S S J R l Z l N G G A S _INE

LIQUID LEVEL

VOLUME REDUCING

ANTlFLOWaACK S E A L IMPFL LFR VANE'

Fig,

3.2.

Vertical-Shaft Centrifugal Sump Bump (Model LFB),

PERIOD ENDING SEPTEMBER 70, 1954 sump i n that it must be sufficiently high a t a l l times t o provide the necessary flow into the seal.

To stop the unit, the sump tank would be vented at the same time the pump was stopped, but the gas pressure would be maintained u n t i l draining had been completed. This would cause a f l o w of gas along the shaft into the seal c a v i t y and would blow out any f l u i d that might enter along the shaft. It would a l s o maintain a seal c a v i t y pressure somewhat greater than the sump tank pressure and would force the f l u i d out of the seai i n t o t h e sump. IC t h i s procedure were not followed, the sealing annulus of f l u i d would, upon collapsing when the pump was stopped, momentar i Iy f i I I the sea S cavity above the shaft level and cause f l u i d to enter along the shaft. Vertical-Shaft Centrifugal Sump Pump

UNCLDSSIFIED 3096

ORNL -LR-OWG

EP a

._*

I 3 Q

D. R. Ward Aircraft Reactor Engineering D i v i s i o n Developmental work has been continued on the vertical-shaft centrifugal sump pump (Model LFB) for use w i t h forced-circulation loops, both in-pile and out. This i s a vertical-shaft centrifugal sump pump w i t h a side i n l e t and a bottom discharge ( F i g . 3.2). A Graphitar face seal prevents the pressurizing gas from escaping along the shaft. Circulating spindle o i l serves t o lubricate this face seal and t o carry away heat from the seal region. T y p i c a l performance curves for t h i s pump with In a recent test the water are shown i n Fig, 3.3. pump was run for 864 hr at 6000 rpm w h i l e pumping 140OOF sodium a t about 3 gpm. T y p i c a l oil leakage past the shaft face seal was under 0.1 cm3/hr. Developmental work during the past eight months has resulted i n the following improvements i n t h i s pump: the elimination of the tendency for the pump t o introduce gas into the liquid stream a t f l o w rates up to 6 gpm; an improved anti-flowback seal to prevent recirculation of the l i q u i d w i t h i n the pump; an improved means for cooling the face-seal region of the pump; and a reduction of pump volume, which i s desirable when the pump i s used for in-pile loop tests. T h e main difficulty, that of gas entrainment, was primarily caused by vortexing and the passage of gas downward along the rotating shaft. The ad-

FLOW ( g p m )

Fig. 3.3. Performance Curves for VerticalShaft Centrifugal Sump Pump (Model LFB). d i t i o n of vertical baffles adjacent t o the shaft merely broke up the large vortex into several smaller ones, which were just as obiectionable. T h e problem i s solved in the present arrangement by permitting liquid from the high-pressure region of the pump t o be bypassed up through a hole i n the shaft t o an annular space surrounding the shaft. T h i s liquid, i n turn, forms a seal and prevents the gas from following i t s former path. Acceptance tests for these pumps include operating each pump with water to check the p m p i n g Characteristics and freedom from gas entrainment and pumping sodium a t 1400OF for 72 hr a t 6000 rpm w i t h acceptable o i l leakage past the face seal. T h i s hot test a l s o serves t o clean oxide f i l m from the wetted pump parts, as w e l l as i o reveal any mechanical faults that might be present. F i v e of these pumps are being built, and four similar pumps (Model LFA) are now i n use.

ANP QUARPERLY PROGRESS REPORT Hydraulic Motor Bump Drives

J. A. C o n l i n Aircraft Reactor Engineering

D i v is ion

T w o types of hydraulic motors have been found t o be suitable for use as in-pile pump drives. The first, a gear-type motor rated at 1.75 shp a t 3500 rpm, WQS operated a t 6000 rpm for 500 hr under a l i g h t load (about 0.08 shp) that was obtained by coupling the motor PO a small air blower. Upon completion of the test the motor was disassembled and was found to be i n good condition. The rubber l i p shaft seal had worn a s l i g h t groove i n the shaft, and a carbon-like deposit which had accumulated on the shaft just outside the seal caused the seal t o stick s l i g h t l y after it had been i d l e for a prolonged period. However, throughout the test, s e ~ lleakage was i?egligible. The only other indication of wear was a s l i g h t scoring on the gear sides adiacent to the shaft end of the motor. The second motor, an axial-piston type rated a t 3.8 shp a t 6060 rpm continuous duty, was run for

100 hr a t 6000 rpm. Operation W Q S completely satisfactory, and, since this motor i s rated a t 6060 rpm, further testing was judged to be unnecessaiy. The nxial-piston motor was chosen for in-pile s s r v i t e , since it i s considered to be the more r e l i a b l e unit. Throughout these tests, hydraulic a i l w i t h a v i s c o s i t y of 155 SSU (seconds Saybolt wniversal) a t 100OF and a v i s c o s i t y index of 100 wa5 used. It should be noted that during i n i t i a l testing nf the gear-type moto:, o i l w i t h a v i s c o s i t y of 290 t o 325 SSU at lOO'F was used, and, as a result, excessive overheating of the motor occurred t o t h e extent that the shaft turned blue. Despite

heating, a r e s i s t i v i t y value' of 98 p h m - c m was used for Inconel i n computing the necessary length of the resistance-heated tube. However, data obtained during the loop operation indicated that the actual r e s i s t i v i t y of the lnconel was about 70 t o 80 pohm-cm. T o resolve the discrepancy, a duplicate of the heating section nf the loop WQS f i l l e d w i t h NaF-ZrF,-UF, (50-46-4 mole X) and alternately heated and cooled while accurate potential and current measurements were mode. Sixty-two measurements were made i n the temperatwre range of 800 t o 1600°F, and the average r e s i s t i v i t y of the Inconel was found to be 75 pohrn-cm. A program i s under way for the design, construction, and operation of t w o series of forcedcirculation loops for studying the effect o f temperature gradient und f l u i d v e l o c i t y on the phenomenon of mass transfer in lnconel systems containing

Aircraft Reactor Eng ineer i ng D i v i s i on

fluoride mixtures a t elevated temperatures.' The requirements of the .two series of loops are a Reynolds number of 10,000 w i t h temperature gradients of 100, 200, and 300°F and a temperature gradient of 20O0F and Reynolds numbers of 800, 3,000, and 15,000. The maximum fluid temperature i s specified CIS 1500°F, and the Reynolds numbers are t o be evaluated a t that temperature. The loops are t o be constructed of lnconel twbing. The maximum t u b w a l l temperature i s to be 1700°F, and the surface-to-volume r a t i o i s to b e held e s s e n t i a l l y the same for each loop.

University of Tennessee

4W. K. Stair, The D e s i g n of a Small Forced Circulation Cowusion Loop ( t o be published).

this, the motor completed the t e s t satisfactorily.

f 0 R C E D-CI R C Ub A T 10 M CO R R 0 S I 0 N L 0 8 P 5 laaconel Loops

W. 63. Twnnell

W. IC. Stair, Consultant J. F, Bailey, Consultant

-I Re previowsly described, small,

Inconel, forcedc i r c u l a t i o n loop designed t o have a temperature differential of 135OF has been operated for a total (50-46-4 mole X) of 141 hr w i t h NaF-ZrF,-IJF, a t an apparent mass f l o w rate of 400 t o 450 Btu/hr w i t h a temperature differential up to 300°F.3 The

f l u i d velocity i n the loop normally was maintained O S about 1.7 fps a t a temperature gradient of 165OF. Termination of the test resulted from a short i n the motor leads and subsequent rupture of the heated section of the loop. T h e t e s t f a c i l i t y , shown i n F i g . 3.4, i s being rebuilt and i s approximately 80% complete. The f i n a l design conditions are presented i n F i g . 3.5, Reports which give a detailed presentation of the design and an analysis of the i n i t i a l operation are being prepared.4t5 I n the design of the loop for electrical-resistance

3W. C. Tunnoll, W. K. Stair, and J. F. B a i l e y , ANP Quar. P r o g . Rep. l u n e 10, 1954, ORNL-1729, p 21.

'!Y. C. Tunnell, Operation Re ort o / t h e First Small ATV Study Loop ( t o be published?. 61nternatianal Nickel Cn,, Properties o/ Some Metnls

and Alloys.

7W. D,

ature

1954).

Manly,

Gradient

High Flow V e l o c i t y and High TemperLoops, OKNL. CF-54-3-193 (Mar. 18,

P E R I O D ENDING SEPTEMBER

IO, 7954

Fig,

3.4. Inconel Forced-Circulation Corrosion Loop.

ANP QUARTERLY PROGRESS REPORT

ORNL

1 - LR - D W G 3097

TEMPERATURE DIFFERENTIAL, 159.2OF REYNOLDS NO. AT 13OO0F, 3730 REYNOLDS NO. AT 1459"F, FLOW,

4540

0.81 gprn

VELOCITY,

5.29 fps

MASS FLOW RATE, 0.361 I b l s e c LENGTH OF HEATING SECTION, 5 . 1 3 f t LENGTH

OF COOLING SECTION, 5.86 f t

POWER, 17.6 CURRENT,

400 amp

VOLTAGE, 4 4 HEATING SECTION

0.02 5 -in. WALL INCONEL TUBING

Configuration and Design Conditions for lnconel Forced-Circulation L o o p t o Operate a t a Fig. 3.5. Temperature Differential of Approximately 16OOF.

A summary of the design dimensions and expected operating conditions for each of the s i x loops i s presented i n Table 3.1, which a l s o indicates the current status of each loop. Three methods of heating were studied: electr ical-res istance, gasfired furnace, and l i q u i d bath. The l i q u i d bath was discounted because it would require pressurization of the l i q u i d rneta1,an excessively long heat exchanger,and a large investment. A gasf i r e d furnace was considered, but lack of operating experience and the high heat f l u x necessary for the loops indicated the use of e l e c t r i c a l heating, w i t h which considerable experience has been obtained. A gas-fired furnace i s being studied as a possible heat source for a large number of corPreliminary calculations of the rosion loops. pressure drop through the loops indicate that the

Model L F A pump8 can be used, and thus it w i l l not be necessary t o develop a pump. For purposes of preliminary design the following assumptions were made: (1) T h e specific resis(50-46-4 mole %) t i v i t y of molten NaF-ZrF,-UF, i s so high that the resistance of the e l e c t r i c a l l y heated lnconel tube may be taken as the parallel Heat transfer to the c i r c u i t r e ~ i s t a n c e . ~ (2) fluoride mixture may be calculated b y relations established for water." (3) A thermal loss equal 'W. 6. McDonald et al., A N P Quar. Prog. Rep. Mar. 10, 1954, ORNL-1692, p 15. n

L. Southern, p e r s o n a l communication. 'OD. F . Salmon, Turbulent Heat Transfer from a Molten Fluoride Salt Mixture to N a K in a Double Tube Heat Exchanger, ORNL-1716 (to be issued). 'A.

PERIOD ENDING SEPTEMBER 10, 1954 T A S L E 3.1.

P R I N C I P A L F E A T U R E S A N D STATUS O F FORCED-C1FtCULAflON CORROSION L O O P S

_ _ . _ . . . . . . _ . . i _ _ . . . . . . ......_.._

.........

. . . ~ ~ ~ . . . . . . I _ _ _

-I____. ____ . ._..--......

L o o p Number

-.-.~~._~_____..-~..__I___-..-~---

..-~

10,000

1 0,000

10,000

800

3000

15,000

Temperature differential, O F

100

2 00

300

200

2 00

200

Power input, kw

27.7

54.2

80,O

6.8

25.5

86.9

H e a t i n g section length, ft*

8.08

22.3

8.08

Cooling section length, ft*

15.64

1.6

15.44

H e a t exchanger length, ft**

18.75

None

Nons

18.75

Pressure drop, p s i

35.12

36-54

47.5

0.06

0.57

1 15.5

Status

Detail design complete

Construction and assembly 80% complete

Detail design

Pre I i mi nary

Construct ion 20% complete

Reynolds number a t

1500'F

"OS-in.-OD, **loo-in.-OD,

0.020-

in.-wal

0.049-in,-wall

complete

design complete; d e t a i l design 20% comp lete

I tubing. tubing used for outer tube

of concentric-tube heat exchanger.

Diosimi lor-Metal Loops

L. A. Mann Aircraft Reactor Engineering D i v i s i o n A loop for testing combinations of structural metals i n contact w i t h high-velocity turbulent l i q u i d metals under high temperature differentials i s being developed, If combinations of materials are found that ore satisfactory for use i n l i q u i d metal systems, greater f l e x i b i l i t y and ease of construction may be attained. A n optimized design, patterned after that used for beryllium-lnconelsodium mass transfer tests (cf. Sec. 2, "ReflectorModerated Reactor"), has been developed and i s to be used f i r s t for testing combinations of lnconel and type 316 stainless steel i n sodium. GAS- F U R N A C E H E A +-SOU RC E DE V E L 0 P M E N T

L. A. Mann

Aircraft Reactor Engineering D i v i s i o n

L. F. Roy"

90%

to 10% of the total heat transfer exists a t the heater section and a t the heat exchanger.

University

Mississippi

Several exploratory tests were made O Q cammercial and QRNL-designed-and-fabricated burners for use w i t h the gas-furnace heat source described T h e heat release intensity desired previously. l 2

of 3000 Btu/ft3.sec, which i s of the same order as that attained i n i e t aircraft burners. The maximum rate of heat release obtained i n the tests, 170 Btu/ft3, was obtained with the i s o n the order

burner designed b y A. P. Fraas. The results of these exploratory tests are presented in Table 3.2. Although the rate of heat release contemplated i n the original design (about 3000 Btu/ft3.sec) was not approached i n these tests, it appears that rates higher than 170 Btu/ft3.sec can be attained by securing more rapid mixing of air and gas, preheating the combustion-air, and providing a positive pressure i n the combustion chamber. Additional tests are to be made.

STUDY O F T H E C A V I T A T I O N P H E N O M E N O N

W. G. Cobb A. G. Grindell J. M. TrummeI Aircraft Reactor Engineering D i v i s i o n

G. F.

Wislicenus, Consultant

One of the problems associated w i t h operating l i q u i d metal systems a t elevated temperatures, high flow rates, high pump speeds, and minimum container weight i s that of local b o i l i n g or cavitation. Because there i s little information available 'Summea ~ c s e a r c hParticipant.

"A. P. F i - ~ i i s , R. W. Bussard, R. E. McPherson, ANF' Qutar. Frog:, K r p . jzine 10, 1954, ORNL-1729, p 23.

ANP QUARTERLY PROGRESS REPORT TABLE 3.2.

EXPLORATORY TESTS OF GAS BURNERS Furnace

Burner

Gas Rate (cfm)

Rate of H e a t R e l e a s e * (Btu/sec.ft

2.1 (max)

ORNL

170

Temperature*

Remarks

(OF)

2000

F l a m e d i d not f i l l combustion space on which r a t e was based

F i s h e r , laboratory type

160

3.5

R a t e based o n estimated flame dimensions

N-4

1.2

Two-in. p i p e with l e t

1.3

National

Maximum a t t a i n a b l e

2900

* T h e s e v a l u e s are the highest that were attained i n t h e t e s t but d o not n e c e s s a r i l y represent the maximum volues attainable, except where indicated.

on the cavitation phenomenon in forced-circulation high-temperature sodium study was initiated,

systems,

preliminary

T e s t s were run on the ARE hot pump test f a c i l i t y i n which pressures at the venturi throat could be varied by varying the pressure of the blanketing gas. Thus it was possible t o vary the pressures without changing the pump speed, the throttle valve setting, or the temperature. Data were taken during eight runs at temperatures i n the range 1200 to 1500’F. Measurements were made of pump discharge pressure, venturi throat pressure, and pump suction pressure as the pump suction pressure was varied. Cavitation was found to begin when the venturi throat pressure ceased t o decrease uniformly w i t h decreasing pump suction pressure. Also, as cavitation began, the loop pressure drop increased and the flow rate decreased. Observation of f l u i d flow noise revealed a correlation of noise intensity w i t h pressure data. The pressure at which cavitation occurred cannot be exactly ascertained, since the pressure a t the middle of the throat of the venturi i s obviously not the lowest pressure i n the venturi. In water cavitation, bubbles form a t the start of the f l o w expansion, and, i n the particular venturi used, the rather sharp radius at the transition from the throat to the diffusion cone probably caused flow separation. The minimum pressures observed a t the throat were within 1 p s i of the vapor pressure of sodium.

Instrumentation for measurement of subatmospheric pressures was required for these hightemperature (up to 1500°F) sodium-cavitation studies. A Moore nul I-balance bel lows-type pressure transmitter and a Moore Model 60-N “nullmaticf’ p i l o t valve modified t o vent the p i l o t volve t o a vacuum source proved t o be satisfactory for this application. When the venting atmosphere or sink pressure i s reduced, the u n i t i s capable of measuring correspondingly lower process pressures. Pressure measurements were made over the range of 20 t o 1.6 psia, w i t h the sodium-filled transmitter maintained a t about 65OOF. The zero shift of the pressure transmitter ranged from 4 . 3 t o -0.2 in. t i g over the range. A second check of the zero shift was made as the transmitter temperature was varied from 800 t o 470’F w i t h a constant system pressure of approximately 16.2 psiu. The zero s h i f t then observed was +0.1 to -0.2 in. tig. In general, the performance of t h i s pressure-measuring device was quite sotisfactory; a more highly refined application of t h i s device might lead PO even smaller zero shift. A report covering t h i s modification i s being prepared. EXPANSION AND IMPROVEMENT O f THERMALCONVECTION LOOP TESTING FAClLlTlES

E. M. Lees Aircraft Reactor Engineering D i v i s i o n

The number of power panels i n s t a l l e d for supp l y i n g thermal-convection loops has been expanded

PERIOD ENDING SEPTEMBER 10, 1954 from

18 t o 31

so that tests of long duration and

extensive tests of "new" materials can be made rapidly. Materials such as the Hastelloy alloys, molybdenum, and nickel-molybdenum a l l o y s are t o be tested t o determine the effects of time, temperature, temperature difference, loop precleaning methods, f l u i d pretreatment methods, corrosion inhibitors, etc. Samples o f newly designed heating equipment have been ordered for study, w i t h the expectation of replacing the present method of i n s t a l l i n g

Fig.

3.6.

Thermal-Convection Loops. ( u )

custom-tailored heating units and insulation on each loop. Split-tube furnaces that would remain in place from t e s t to test may possibly be used. The p o s s i b i l i t i e s of automatizing operation of the loops and o f providing safeguards i n case of shutdowns because of power failures or other emergencies are a l s o being studied. The loops have been redesigned so that they require fewer welds, and the sharp angles of l i q u i d flow have been eliminated. Figure the improved designs.

Old design. ( b ) New design.

3.6

shows the old and

ANP QUARTERLY PROGRESS REPORT

4. CRITICAL EXPERIMENTS A. D. Callihan Physics D i v i s i o n REFLECTOR-MODERATOR REACTOR

D. Scott

ARE

B. L. Greenstreet Division

J. J.

Lynn

Physics D i v i s i o n

J. S. Crudele

J. W. Nooks

P r a t t and Whitney Aircraft D i v i s i o n The f i r s t step of the present c r i t i c a l experiment program was the construction o f a small two-region reflector-moderated reactor to provide experimental data on a system of simple geometry and materials for use i n checking the present calculational The assembly i s approximately spherimethods. cal, 41 in. i n outside diameter, and has a center, or core region, that is about 15 in. in diameter.2 T h e core consists of alternate sheets of enriched uranium metal (4 mils thick) and Teflon, (CF,), It i s and i s surrounded by a beryllium reflector.

possible to vary the uranium loading, within the specified dimensions, i n order to make the system critical. The Teflon and the uranium have been cut to the various shapes necessary to form the subassemblies required to conveniently construct the core. A summary of the materials i n the f i r s t c r i t i cal assembly i s given i n Table 4.1. The aluminum, which i s a necessary structural material, i s present i n the core as a vertical double sheet a t the midplane and i n the reflector as f a i r l y well-distributed sheets and rods. The coating material, which contained 31.8 wt % C, 61.9 wt % F, 5.5 wt % Cr, and 0.83 wt % H, was applied i n a thin layer to the uranium metal to reduce oxidation. The Scotch cellophane tape was used to hold the core subA view of the mid-plane o f assemblies together. the reactor as i n i t i a l l y assembled i s shown i n Fig. 4.1. T h e small rectangular holes below and to the right of the core are the aluminum guide channels for the control and safety rods.

An earlier assembly was loaded with 17.4 Ib of prescribed by the multigroup calculations, and was not critical. I t was discovered, however,

U235, as

that these calculations were i n error because of a numerical error in one of the cross sections and that the predicted c r i t i c a l mass range should have been 20.9 to 22.75 Ib of U235. T h e system was made c r i t i c a l with 24.35 Ib of U235 and had an excess reactivity of approximately 7 cents, with about 20% o f the mid-plane aluminum plates omitted. It has since been shown that the mounting of t h i s plate resulted i n a reduction i n reactivity of approximately 10 cents, and therefore the system as designed and loaded w i t h 24.35 Ib was s l i g h t l y subcritical. Sufficient excess reactivity for subsequent experiments has been gained by the addition of a cylindrical annulus of beryllium reflector 1 7 i in. long and 2T8 in. wide w i t h i t s axis perpenThis addition gave an dicular to the mid-plane. The extra increase i n r e a c t i v i t y o f 65 cents. beryllium at the top of the reactor was removed during f l u x measurements. T A B L E 4.1. COMPOSITION O F CORE AND R E F L E C T O R O F FIRST REFLECTOR-MODERATED REACTOR C R I T I C A L ASSEMBLY Core

1.05 7.48

Volume Average radius Weight of

,A more complete description is g i v e n i n ReflectorModerated Critical A s s e m b l y Experimental Program by D. Scott and B. L. Greenstreet, ORNL CF-54-4-53 ( A p r i l 8, 1954).

in.

Teflon*

58.464 kg (128.62 Ib)

Uranium u235**

11.878 k g 11.067 kg (24.348 lb)

0.1 12 kg 0.085 kg 0.92 kg (2.02 Ib)

Uranium coating material Scotch tape Aluminum Ref lector

21.24 12.9

Volume Minimum thickness

'D. Scott and B. L. Greenstreet, A N P Quar. Prog. Rep. Mar. 10. 1954, ORNL-1692, p 45.

ft3

Weight of

1102.5 16.8

B e r y l I ium Alurni num *Density:

1.96 g/cm3 of core.

** Density: 0.372 g/cm3

of core.

ft3 in. kg kg

(2425.7 Ib) (37.06 Ib)

PERIOD ENDING SEPTEMBER 70, 7954

Fig. 4.1.

F i r s t Reflector-Moderated Reactor C r i t i c a l Assembly.

A N P QUARTERLY PROGRESS REPORT Measurements of neutron f l u x have been made in a vertical direction along a radius by using indium f o i l s with and without cadmium covers. The f o i l s in. in diameare 10-mil aluminum-indium alloy, ter, having an effective indium thickness of 3 x lo-' in., and the cadmium covers are 20 m i l s thick. The results o f these measurements are given i n Fig. 4.2. The values of the indium-cadmium fraction3 and the indium activation produced by neutrons

distribution was made through the core along the radius that was used for the indium f l u x measurements by observing the a c t i v i t y of the fission fragments retained i n aluminum f o i l s placed i n contact w i t h uranium. Similar data, presented i n Fig. 4.3, were obtainedwith the uranium and aluminum enclosed i n cadmium covers. The average uranium-cadmium fraction for the core i s 0.33. The fission rates were measured on opposite sides of

having energies less than the cadmium cut-off are A relative power, or fission-rate, also plotted.

three uranium f o i l s near the reflector, and the results are plotted. In each case the rate was

J T h e cadmium fraction by definition i s the ratio of t h e d i f f e r e n c e between t h e bare and cadmium-covered activ a t i o n s to the bare activation.

higher on the side toward the beryllium. Measurements w i l l be made of the r e a c t i v i t y coefficients, as functions of the radius, of a few

ORNL-LR-DWG

3017

+ 0

5 25

L. 0

+ .-

t 2

4 0

2 4

cLrL 40

5 ,

0 12

RADIAL DISTANCE FROM CENTER OF CORE ( i n )

Fig. 4.2. Assembly.

Radial Neutron Distribution a t Mid-Plane of F i r s t Reflector-Moderated Reactor C r i t i c a l

PERlOD ENDING SEPTEMBER JO, 7954

O R N ! . - e 3 0 , 8

._r

k >

ir 16

ti CI

1 lil

CAYjMIUM FRACTION '

........... ....

RADIAL DISTANCE f FWM CLNTER OF CURVE ( I n 1

Fig. 4.3. Power Distribution of Mid-Plane of Reactor. The paired points on the bare aluminum f o i l activation curve nt 6.4,6.9, and 7.1 in. were obtained from f o i l s placed on opposite sides of the same u r o nium f o i l .

metals, such cis Inconel, nickel, and cadmium. Upon completion of the experiments on this assembly, it i s planned to construct a larger reactor

of the same shape that w i l l consist of three regions,

with the beryllium island and reflector separated by the fuel annulus. The diameters of the beryllium island and the fuel annulus w i l l be about 10.4 and 20 in., respectively. I n i t i a l l y , the assembly w i l l have no structural materials, such as Inconel core shells, and it w i l l provide a further check o f the calculational methods. SUPERCRITICAL-WATER REACTOR

J. S.

Crudele J. W. Noaks Pratt and Whitney Aircraft D i v i s i o n

tained i n stainless steel tubes distributed in an organic l i q u i d i n a pattern designed to give a uniform radial thermal-neutron flux. The liquid, Furfural (C,H,O,), which also serves os c1 neutron reflector on the lateral surface of the cylindrical tube bundle, has a hydrogen density approximately the same as that of water under the temperature and pressure It simulates conditions designed for the SCWR, the nuclear properties of supercritical water as well as can be determined without further knowledge of the effect of binding energies on diffusion and slowing down lengths. Stainless steel was inserted i n the fuel tubes to represent the reactor structure.

The Supercritical-Water Reactor (SCWR) c r i t i c a l assembly was described previously4 as consisting conof an aqueous solution of enriched UO,F,

The loading of the c r i t i c a l assembly was calculated by Pratt & Whitney Aircraff D i v i s i o n personnel,' by use o f a four-group diffusion theory method, ta be 15.34 kg of U235 and 261 kg o f stainless steel in CI 76.2-crn equilateral right cylinder with a

*E. L. Zimmermon, J. S. Crudele, and J. W. N o o k s , ANP Quar. Prog. R e p . Mar. IO, 1354, ORNL-1692, p 45.

'private conlrnunication f r o m George Chase, Project, Pratt S Whitney Aircraft Division.

Fox

ANP QUARTERLY PROGRESS REPORT

side reflector

6.5 cm thick.

The calculated radial

fuel distribution i s plotted in Fig. 4.4 as the smooth curve and i s compared to the stepwise distribution achieved when call tubes are identically loaded, In the experimental study, a quantity of U02F2 aqueous solution,

containing uranium enriched t o

93.14% i n U235, was distributed among the tubes The cat a U235 concentration o f 0.505 g/cm3.

number o f tubes required for c r i t i c a l i t y was rneassured as a function of the Furfural height as i n creasing quantities o f stainless steel were inserted and as the solution was diluted. In t h i s manner a stepwise approach was made t o a loading of uniform

concentration which would be c r i t i c a l at the designed linear dimensions of both f u e l solution and Furfural and which would contain the prescribed mass of stainless steel. T h e s e conditions have not been achieved because the l a s t d i l u t i o n was overestimated and a large part of the solution i s a t a concentration somewhat lower than required for criticality, a deficiency compensated for by locating about 50 tubes of higher concentration fuel i n one peripheral section. The configuration i s described i n Table 4.2. Some neutron-flux measurements have been made by using effectively thin

lom4in.)

indium f o i l s

ORNL- LR- DWG 301 9

0.28

0.26 LT

0 +111

0.24

0.22

0.20

d w E

k z

0 X ~

0.18

0.46

3 LL

2 F

.z 0

0.14

0.12 0.40

(1.

0.08

0.05

0 0 4.

RADIUS (crn)

Fig.

4.4. Radial Fuel Distribution i n Suporcritical-Hater Reactor.

PERIOD ENDING SEPTEMBER 70, 1954 w i t h and without 0.OZin.-thick

cadmium covers.

The data obtained along a radius 36.7 cm from the bottom o f the core and somewhat removed from the high-solution-density perturbation are s h o w n in Fig. 4.5. Similar data from a longitudinal traverse located 2.4 cm from the cylinder axis are given i n Although these results are preliminary, F i g . 4.6. i t i s believed that the apparent nonuniformity o f the radial thermal f l u x i s greater fhan t h e experiA few measures show the mental uncertainty. radial importance o f

U23s to

a l s o decrease from

t h e center. I n one experiment to evaluate the Furfural reflector, annular sheets o f aluminum were inserted adjacent t o the w a l l of the reactor tank, thereby

reducing t h e reflector thickness from 10.0 cm to 3.6 cm. The effect of t h e aluminum, indicated by the c r i t i c a l height of the Furfural, was o s l i g h t increase in the reactivity.

T A B L E 4 . 2 CONFIGURATION O F SCWR CRITICAL EXPERIMENT Experimental

Design

215.0

Height o f U02F2 solution, c m

71.2

76.2

Height of Furfural, cm

69.5

76.2

11.48

15.34

Number of tubes"

M a s s of

U235, kg**

Thickness of reflector, c m Mass of stainless steel, k g

10.0

6.5

220.4

261.0

*Expressed a s equivalent number of tubes 1 in. i n diameter; the a r r a y contains 193 which a r e 1 in. in diameter, 8 which are 3 .in., and 25 which are in.

* * O f this loading, 86.8% w a s in a solution having a of 0.196 g/cm 3, and 13.2% wus in o n e with a U235 concentration o f 0.505 g / c m 3

U235 concentration

slc

ORNL--LR--0WG 302C

Fig. 4.5.

Radial Neutron D i s t r i b u t i o n i n Supercritical-Water Reactor C r i t i c a l Assembly.

ORNL-LR -D'.VG

302<

26 24 22 20

('0

0.8

u 14

0.7

K w

3 4 5 n

6 4

0.5

3 2

03 J

- + - - -

INDILM ACTIVATION MINUS CADMIUM - COVERED INDIUM ACTIVATION

3.1 ~

-.-.-.

~i ~~

L~~ '

DISTANCE FROhl REACTOR TANK FLOOR ( c m )

fig.

4,Q.

_ _ ............................. 1 -

? L

Longitudinal Neutron Distribution i n Supercritical-Water Reactor C r i t i c a l Assembly.

Part II

MATERIALS RESEARCH

5. CHEMISTRY O F MOLTEN ~

W. R, Grimes Materials Chemistry D i v i s i o n Studies of the NaF-ZrF,-UF, system were i n i t i ated after the s o l i d phase studies i n the NaF-ZrF, system were completed, and to date no ternary compounds have been discovered i n the NuF-ZrF,-UF, system. A tentative equilibrium diagram has been prepared, The apparatus for v i suul observation of melting temperatures was used to observe eleven

NaF-UF, mixtures in the composition range 20 to 50 mole % UF,, and a partial phase diagram of the system was prepared. A new apparatus has been constructed that w i l l permit some mnnipulations to be carried out w i t h the fused s a l t s i n an inert atmosphere. Recent attempts to correlate the anticipated re-

duction of UF, i n the preparation of UF3-bearing melts with wet chemical analysis for UF, and UF, and results of petrographic examination show some surprising anomalies. When UF, dissolved i n LiF, i n NaF-ZrF, mixtures, or in N a F - L i F mixtures i s treated under flowing hydrogen a t 800°C w i t h excess uranium metal, 90% or more o f the UF, i s reduced to UF,. However, when t h i s technique is applied to UF, i n N a F - K F - L i F mixtures, the reduction i s only 50% complete at 800°C and, perhaps, 75% complete a t 600°C. Petrographic examinations

o f the specimens reveal no complex compounds of tetravalent uranium; it i s possible that the UF, i s “hidden”in s o l i d solutions or i n complex UF,-UF, compounds i n which i t i s not a t present recognizable. A method for the large-scale Purification o f rubidium fluoride has been developed. All the rubidium fluoride obtained from commercial supp l i e r s to date has contained considerably more than the specified quantity of cesium compounds, and therefore almost a l l of i t has been returned i o the vendors for further processing. T h e purification method developed can be used at a reasonable c o s t on a large scale i f material sufficiently free from cesium cannot be obtained from commercial sources. Values for the equi!ibrium constant for the reaction Cr”

2UF,

t CrF,

in molten NaZrF, have been re-examined by the use UF3/UF, with chromium metal i n the charge. From the data obtained when t h i s

of various ratios of

i n i t i a l ratio i s less than 3, average values of 5 x IO-, a t 600°C and 6 x lom4at 800°C can be At ratios larger than 3 the values incomputed, crease regularly; t h i s r i s e i s probably due to the extreme d i f f i c u l t y i n f i l t r a t i o n and analysis of samples containing 1 to 20 ppm of Cr++. The r a t i o s agree quite w e l l values at low UF,/UF, w i t h the corresponding values of 4.2 x lo-, at 600°C and 4.1 x l o w 4a t 800’C obtained previously when UF, and chromium metal were the charge mater ia1

Fundamental studies were made of the reduction

o f NiF, and FeF, by H, in NaF-ZrF, systems a s a means of determining possible improvements i n Also, methods for prepurification techniques. paring simple structural metal fluorides were studied. Additional measurements o f decomposition potentials o f KCI and o f various chlorides i n molten KCI a t 850°C were made.

Preliminury

measurements of the solubility of

xenon in KNO,-NaNO, eutectic (66 mole 7; NaNQ,) mo6e/cm3 a t show values of 8.5 x l o w 8 and 280 and 3M)”C, respectively, at 1 atm xenon pressure. The all-glass apparatus used has been replaced with CI n i c k e l and glass combination, and measurements of xenon s o l u b i l i t y i n molten fluorides are being made. L a r g e quantities of purified ZrF4-base fluorides are being prepared for engineering tests a t ORNL and elsewhere, and the demand for purified fluor i d e s o f other types for possible reactor fuel application i s increasing rapidly. Consequently, an increasing fraction of the effort i s devoted to production or to research i n direct support of product i o n functions.

Since the consumption o f NaF-ZrF,

and NaF-

ZrF,-UF, mixtures i s expected to reach 10,000 Ib during the f i s c a l year, i t i s important to decrease the cost of production of t h i s material. The HF-H, processing currently used i s adequate from a l l points o f view, but processing times of nearly I00 hr per batch ore now required with the relat i v e l y poor raw materials available. The substitution of hafnium-Free ZrF, and commercially available NaFx.ZrF, for the impure ZrF, now used

should afford a considerable saving.

The

ANP QUARTERLY PROGRESS REPORr proposed use of zirconium metal as o scavenger (to replace most o f the hydrogen processing) h a s shown promise on a small scale; and rapid electrolytic deposition o f iron and n i c k e l from the ZrFd-base mixtures appears to be feasible. It i s anticipated

N a 3 U F -Na3ZrF7, with the high liquidus temperature (8JO°C) at the zirconium compound and the low

liquidus temperature (628OC) at the uranium eomp ~ u w d . In order to study equilibrium relationships i n the region between these t w o s o l i d solution joins, a series af nine compositions equally spaced along

that the p r i c e o f purified tdaZrF, and NaF-ZrF,-UF, mixtures may be halved i n the next few months,

the Na,ZrF,-No,lJF, j o i n was prepared. Phase annlyses on completely crystallized preparations of these compositions indicate that Na,ZrF6 and

S O L I D P H A S E S T U D I E S I N T H E NaF-ZrFA-UF, SYSTEM

C, J.

N a 2 U F 6 do not e x i s t i n equilibrium with each other and with liquid, Instead, the j o i n between Na,ZrF, and N a 2 U F 6 crosses :he two three-phase triangles huving their canimon base o i l the Ma,ZrF7-Na9U,F, l i n e and their apexes at Na,ZrF6 and Na,UF,. Quenching experiments are being made for establ i s h i n g liquidus relationships and boundary curves i n t h i s area.

Barton

R. E. 'I'homa Moore Materials Chemistry D i v i s i o n

R. E,

G. D, White, Metallurgy D i v i s i o n

Insley, Consultant

Studies of the NoF-ZrF,-UF, system were initiated after the solid phase studies I n tho NaF-ZrF,

Four ternary mixtures were prepared i n order to

binmy system were coinpleted. The solid phases present in both the slowly cooled and the quenched samples were studied by x-ray diffraction and petrographic analysis techniques. I'he study of the NaF-UF, system was suspended, except for some visual-observation experiments, iind will be resumed when time permits. Visual-observation experiments (cf. section below 61 on Visual Observation of Melting Temperatures i n the NoF-UF, System"), together with petrographic studies of slowly cooled NaF-UF, compositions,

obtain a preliminary indication of equilibrium relationships i n other parts of the system. The compositions of these mixtures are given i n Table 5.1.

Identification, by x-ray diffraction and petrographic techniques, o f the phases i n completely crystall i z e d preparations uf the mixtures indicated that the ~ F , ~ a s m a l l primary compound N O ~ Z ~ occupies

phase area contiguous t o the primary phase fields o f the NaqZr,F4 ,-NarU8F4 and the ZrF,-UF, It a l s o appears that the solid solution series. solid solution primary phase f i e l d o f the ZrF,-UF, series i s contiguous to the primary phase f i e l d o f solid solution series the Na9ZrgF4,-Na9UBFdl along ths greater part of the boundary curve that enters the ternaiy system a t the Na9U,F,l-UF, melting eutectic (42 mole 5% NaF-58 mole % UF, paint 680°C), follows approximately the 50 mole % N a F ioiri for most of i t s course, and leaves the ternary system at the Na,Zr,F,,-Nn,Zr,F ,9 eutect i c (approximate!y 50 mole % NaF--50 mole 76 ZrF,, melting point 510°C). 7 h e mixtures l i s t e d i n Table 5.1 w i l l be used for quenching experiments

demonstrated that the compound previously designated1t283 as N a U F 5 i s a congruently melting compound Na,U8F, that i s anologclus to the Na,Zr,F, c ~ m p o u n d . ~ Quenching experiments wibh ternary compositions on the join Na,LI,F, l-Nn9Lr8F4 show that t h i s join comprises a completely (or nearly completely) m i s c i b l e series, with liquidus temperatures descending from 722°C at the uranium compound t o 520°C. at the zirconium c o m p o ~ n d . ~ Earlier thermal analysis data and f i l t r a t i o n exh a t there i s another p e r i m e n t ~ ~demonstrated '~ series af complete s o l i d solutions along the join ....

'W. R. Grimes e t a/., ANP Quar. PTog. Rep. Mar. 10, 1951, ANP-60, p 129. 2W. H. Zochariasen, /. Am. Chhpn. SOC. 70, 2147 (1948). 3C. A. Kraus, Phuse Dingroins of Some Complex S a l t s of Urmzium with Halides o the Alkali and Alkaline Earth M c t d s , 14-251 (July 1, 19 3). ,R. E. Thomn, e t al., A N P ua7. Prog. R e p . J u n e 10, 1954, ORNL-1729, Fig. 4.1, p $1, 5 ~ J.. BQF?rJn, 5. A. R o y e r , and R. J. s e i l , A N P ~ m 7 P i o g . Rep. D e c . 10, 1953, ORNL.-1649, pp 50 and 55. 6 C . J. Bartan and R. J, Sheil, ANP qua^. Prog. Rep. Mar. 10, 1954, ORNL-1692, p 54.

Sample

Designation TI .

T3 T4

Composition (mole %)

......

NQF

..................

45 41 42 48

Zr F,

53 55 48

a 4 10 13

PERIOD ENDING SEPTEMBER IO, 1954 V I S U A L O B S E R V A T I O N O F MELTING T E M P E R A T U R E S IN T H E NaF-UF, SYSTEM

i n the near future t o establish solidus and liquidus re! ation sh ips. No ternary compounds have been discovered, to

date, in the NaF-ZrF,-UF, system. A tentative equilibrium diagram, which s h o w s the general relationships believed to exist i n this complex system, i s presented i n Fig. 5.1.

â&#x20AC;&#x2DC;R. J. Sheil and C. J. Barton, A N P QULN.Prog. Rep. June 10, 1954, ORNL-1729, p 42.

M. S. Grim Mater ial s Chem is try

Divi s i on

T h e apparatus for visual observation of melting temperatures, described i n the previous quarterly r e p ~ r t , ~was used to observe eleven NaF-UF, mixtures i n the composition range 20 to 50 mole %

UF,.

Liquidus temperatures obtained

with t h i s

ANP QUARTERLY PROGRESS REPORT apparatus, both from visual observation of crystall i z a t i o n and from thermal effects recorded during the course of the v i sua1 observations, were generally a l i t t l e higher than those indicated by the published diagram for t h i s system.' The data are shown on the partial phase diagram for the NaF-UF, system i n Fig. 5.2. A s o l i d c i r c l e indicates the temperature at which the composition appeared to be completely solidified. T h i s point could not be determined w i t h any degree of certainty with some compositions, uncl the disoppeorunce of the l i q u i d phase w u s n o t always accompanied by B noticeable thermal effect. The liquidus temperatures shown i n Fiy. 5.2 seem to indicate that Na,UF, melts congruently at 629OC, that is, at a temperature only u few degrees higher than the eutectic temperatures on both sides o f the compound. The data also support the r e s u l t s of petrographic studies which indicate that the compound i n the 50 mole % UF, region has the approximate composition Na,U,F, rather than NaUF,. A new appnratus has been constructed that w i l l permit some manipulations to be carried out with the fused salts in an inert utnwsphere. It c o n s i s t s essentially

L u c i t e side w a l l s and top, attached to a stainless steel bottom p l a t e which has a &in. length o f 2-in.-dia stainless steel pipe welded onto i t t o serve C I S a furnace core. The pipe i s surrounded by resistance c o i l s and insulating tape. The box i s equipped w i t h face plates for covering the glove holes so that the box can be partly evacuated before it i s f i l l e d with inert gas, PHASE R E h A T l O N S H l P S IN U F 3 - 5 E A R I N G

C. J.

SYSTEMS

R. J. SheiI A. B. Wilkerson

Barton

L. M. Brntcher

Materials Chemistry D i v i s i o n

G. D. White, Metallurgy D i v i s i o n T. N. McVay, Consultant I t was assuined in the previous studies* o f solub i l i t y relationships for IJF, in various fluoride mixtures that in the presence o f excess m e t a l l i c uranium or zirconium the uranium present i n the fused salt solution WQS trivalent, T h i s assumption seemed justified, since, for the reaction

of a small dry box, w i t h i - i n . - t h i c k Brewer's tabulation' formation y i e l d s ORNL

750

AF"

922

o f standard free energies o f =

-16 kcal

I n addition, petrographic examination of the melts obtained revealed the presence o f red materials tentatively identified as complex compounds o f 700

UF,,

with green compounds of UF, showing only i n traces, i f a t all. Since these observations were those anticipated and since the petrographic exominations had previously been shown to be very s e n s i t i v e in detecting compounds of trivalent uranium i n systems which had been s l i g h t l y reduced, there seemed to be l i t t l e reason to believe that the results o f such examination might be unreliable.

During the p a s t quarter, however, considerable e v i dence ha:, been accumulated by the most accurate wet-chemicol methods yet devised for determining U3+ and U4' i n fluoride mixtures which shows that the reduction of UF, by excess uranium metal i s not complete at 800째C. While the reduction i n

600

550

UF4 h o l e % )

Fig. 5.2. Por4ial W a s e Diagram of the NaF-UF, SyeBcm.

*G. M.

lune

Watson and C. Ah. Blood, A N P IO, 1954, QHNL-1729, p 51.

@UT.

Prog. R e p .

'L, Brewer et ul., Thenodynarfiic P r o p e r t i e s and E y u i libria at High Temperalures of Uranium H a l i d e s , O x i d e s , Nitrides, u r d Carbides, MDDC-1543 (Sept. 20, 1945, rev. Apr. 1, 11947).

PERIOD ENDING SEPTEMBER IO, 1954 NaF-ZrF, mixtures and in LiF appears to be 90%, or more, complete, i n such mixtures as N a F - K F - L i F

only about 50% o f the UF, i s reduced. Toward the end o f the quarter it became evident that several o f the materials such as ''3KF-2UF3" previously believed t o contain o n l y U3' regularly contained Furthermore, large and varying quantities o f U 4'. the crystals containing variable quantities o f U4' are distinguishable as such only with great d i f f i culty, i f at all, by petrographic and x-ray diffraction examination, Con sequent ly the previously reported

data on the s o l u b i l i t y o f UF, i n various systems must be reinterpreted. T h e reasons for the incomplete reduction of the tetravalent uranium are n o t yet completely understood, and i t i s n o t possible at present to define the extent to which the reduction w i l l proceed at various temperatures and i n the various solvents. Accordingly, the significance of much o f the rnat e r i a l presented below i s not completely known.

UF, in ZrF4-5earing Systems T h e s o l u b i l i t y o f UF, in NaF-ZrF, mixtures was previously shown* to increase w i t h increasing temperature and w i t h increasing ZrF,

concentration

o f the solvent over the range 47 to 57 mole % ZrF,. Since the publication of that information it has been shown that the reduction o f

UF,

i n ZrF,-bearing

melts by excess uranium metal i s s l i g h t l y less than 90% complete at 800OC. The temperature dependence o f the reduction i s not yet known for t h i s system, but it i s l i k e l y that the UF, i s more completely reduced a t lower temperatures, An examination of the NaF-ZrF,-UF, system has been attempted by thermal analysis, w i t h petrographic examination of the resulting s o l i d phases.

In these studies, UF, and excess uranium metal are added to the desired NaF-ZrF, mixture before the sample i s heated, and the sample i s stirred constantly while i n the molten state. Therefore the reaction

%UF,

d bUo---UF,

might be expected t o reach i t s equilibrium value at any temperature above the melting point. Consequently, in contrast to experiments in which the m e t a l l i c uranium i s removed by filtration at high temperatures, the s o l i d products i n slowly cooled melts might be expected t o be nearly completely reduced. These studies indicate that UF, i s the primary phase in systems in which the NaF-to-ZrF, ratio i s

l e s s than 5 to 6 and that Na3U,F9 appears as the primary phase iil systems in which the NaF-to-ZrF, ratio i s about 15. It also appears that N a 3 Z r 2 F , , crystals may contain small quantities of UF, i n s o l i d solution. The s o l u b i l i t y data obtained from these and previous studies give l i t t l e reason to expect that UF, can be dissolved in NaF-ZrF, mixtures in sufficient amounts to provide fuel for r e f 1 ec to r-mo d er at ed reactors.

UF, in NaF-KF-LiF Mixtures T h e s o l u b i l i t y o f UF, i n the N a F - K F - L i F eutectic was stated previously" t o b e equivalent t o a t l e a s t 15 wt % at temperatures as low as 525째C. Subsequent careful examination of t h i s system ha5 revealed that when UF, and an excess o f uranium metal are added to the purified N a F - K F - L i F mixture the dissolved uranium species aggregate a t least 22% total uranium in the mixture. However, i t i s obvious that only 40 to 45% o f the soluble uranium i s present as UF, a t 8OO0C, while 55 to 60% may be trivalent at 600째C. Further study o f the system w i l l be necessary before these values can be determined more accurately. Thermal analysis data have been obtained for several mixtures which were prepared from UF, and the N a F - K F - L i F eutectic and then treated with an The data obtained, as excess of uranium metal. shown i n Table 5.2, are in agreement with the data obtained from f i l t r a t i o n studies which showed h i g h uranium concentrations at low temperatures. Petrographic examination o f these materialsrevealed that a t low uranium concentrations a red phase(refractive and index, about 1.44), which is probably K,UF, i s predominant. At high which may contain UF,,

"5. M. Watson and C. M. Blood, A N P Q u u . Prog. Rep. June 10, 1954, ORNL-1729, p 53. T A B L E 5.2.

THERMAL ANALYSlS D A T A FOR

UF3-BEARING NaF-KF-LiF MIXTURES .........

T h e o r e t i c a l Composition

(mole

X)*

NoF

LiF

10.8 10.6

39.5 38.6 37.0 35.3 32.2

43.7 42.8 40.9 39.1 35.7

10.1 9,6 8.8

lJo Used

Thermal Effects

(% o f theory)

(OC)

8.0 120

6.0

200 110 110

16-0 23.3

110 200

475,455 510,460,455 520,455,445 565,475 570,490,470

UF:,

..................

...............

* B a s e d on complete reaction of

U F 4 with U".

ANP QUARTERLY PROGRESS REPORT

uranium concentrations an olive-drab cubic phase (refractive index, 1.46) and a blue b i a x i a l phase (refractive index, 1.544) also appear. There i s evidence that the blue material i s a solid solution

o f Na,lJ,F~,-K,U,F,;

whether i t dissolves

UF, i s

n o t known,

UF,

NoF-RbF-LiF Mixtures

The previously reported

data on the s o l u b i l i t y

o f UF, in the N a F - R b F - L i F eutectic composition were based an the assumption that a l l the uranium For more.recent i n the filtered specimens was UF,. studies, the solubilities were determined by using 200-9 samples of melt, mechanical agitation during

the equi I ibration period, and 20% excess uranium for the reduction step. Samples of the equilibrium mixture were withdrawn with a f i l t e r stick containi n g sintered n i c k e l as the f i l t e r medium. The samples were then analyzed for trivalent uranium and for total uranium. The results of the analyses are presented in T a b l e 5.3. "R. J. Sheil and C. J. Barton, A N P l u n e 10, 1954, ORNL-1729, p 53.

QUUY.

Prog. R e p .

TABLE 5.3- SOLUBlLlTY O F UF3 iN

........

NaF

T h e o r e t i c a l Composition (mole W )

RbF

LiF ..........

14.1

9.4

5.6

37,6

47.0

48.9

42.3

37.6

The significance o f the analyses for UF, i s not entirely clear, Petrographic examination of the s o l i d i f i e d filtrates indicated a reddish-brown maas the only colored terial, presumed to be Rb,UF,, phase, While it i s possible that t h i s material could acconirnodate UF, i n s o l i d solution, i t does not appear l i k e l y that as much as 75% of the total uranium present could be hidden i n t h i s fashion. The good agreement, for some of the mixtures studied, between the theoretical uranium content and the measured total uranium content indicates that the reaction of UF, with uranium metal proceeded nearly to completion. Additional data on t h i s system w i l l be needed before an interpretation of the results can be presented.

F i l t e r e d specimens from preparations in which

UF, dissolved i n LiF was treated at 825 to 850째C

with excess uranium iiietal have shown that more than 90% of the dissolved uranium i s trivalent and Accordingly, i t i s that i t crystallizes as UF,. believed that the behavior observed by thermal analysis of t h i s s y s t e m i s that o f L i F w i t h pure,

NaF-RbF-LiF

Theoretical

__ uF3

u 3 + Content (wt

6.0

39"5

17.7

17.3

6.0

19.3

6.0

...............

i n the Individual Alkali-Metal Fluorides

._____

MlXTURES AT VARIOUS TEMPERATURES F i I t i a t i on Temperature (OC)

793

7 00 6 45 600 550 495 750 675 6 50 6 25

F i l t r a t e Composition

u3+

.........

Total 11 ........

2,92

15.2

1.41

16.9 16.7 14.7 10.7 5.26

4.54

1.17

5.51

17.0 16.9 18.0 16.4

9.18

3,22 9.95

6 00

No sample*

8 00 750 700 6 50 6 25

5.07 4.02 4.07 4.76 4.86

600

(wt

16.8 17.0 17.3 16.4 13.1 ___

6.35

*Filtration wc?s attempted, but the f i l t e r apparently clogged at t h i s temperature and the sample could n o t be

tained.

05-

PERIOD ENDING SEPTEMBER 70, 7954 o r nearly pure, UF,. Recent studies of LiF-UF, mixtures have confirmed the b e l i e f " that no compounds form and that these materials comprise a A partial phase diagram simple eutectic system. for t h i s simple system i s shown as Fig. 5.3. T h e NaF-UF, system has been studied more thoroughly than any other binary alkali-metal fluoride-UF3 system. The best thermal a n a l y s i s data obtained for t h i s system (data obtained with those mixtures which showed the least evidence of contamination o f the melt by tetravalent uranium) obare shown i n Fig, 5.4. The earlier data, tained by mixing N a F w i t h previously synthesized

UF, in sealed capsules, are shown by open circles.

The more recent data, shown by s o l i d circles, were

I2L. M. Bratcher et ul., ANP Quur. Prog. Rep. J u n e 10, 1954, ORNL-1729, p 43. l 3 V . s. C o l e m a n a n d w. c. Whitley, ANP Qum. f r o g . R e p . Sept. 10, 1952, ORNL-1375, p 79. 14W. C. Whitley, V. S. C o l e m a n , and C. J. Barton, ANP Qum. Prog. Rep. D e c . 10, 19S2, ORNL-1439, p 109.

Ir.

OANL-LA-OWG

obtained by reducing NaF-UF,

mixtures with ex-

cess uranium i n open crucibles protected by an atmosphere of helium. Some data, also obtained during the past quarter, were obtained by a method which represents a compromise between the t w o previously mentioned methods. Mixtures were prepared i n open stirrer-equipped crucibles by combinUF,: These i n g dry N a F with freshly prepared" data are indicuted i n Fig.5.4 by the solid triangles.

T h e agreement between the data obtained by the various methods i s considered to be gratifying, in v i e w of the d i f f i c u l t i e s associated with the ease of

oxidation o f U3+to U4'. The identification of the compound i s based chiefly upon the results of petrographic examination of slowly cooled melts. The p o s s i b i l i t y that the compound might be Na,UF, rather than Na,U,F, and the p o s s i b i l i t y that appreciable quantities o f UF, may be "hidden" in these materials have not been completely ruled out by these studies. Very few compositions containi n g more than 50 mole % UF, have been prepared, and the data i n t h i s part of the system are probably of l i t t l e value.

29;

ORhL-LR-DWG

PREPARED

2924

:N O p t 3

UF3 (mole %I

Fig. 5.4. Tentative NaF-UF, System.

P a r t i a l Phase Diagramof

the

KF-UF, and far advanced. It

Phase equilibrium studies of the

RbF-UF,

Fig. 5.3.

UF, (mole Z )

Tentative P a r t i a l Phase Diagram of

the LiF-UF, System.

systems are not so appears that the most probable compositions for the compounds observed are K3UF,, K,U,F, and

Rb,UF, about

md 720째C

"Prepared

that the KF-K,UF, at approximately

W. C.

eutectic melts a t mole % UF,. It

Whitley, July 1954.

ANP QUARTERLY P R O G R E S S R E P O R T seems that i n both systems considerable UF, i s present along with the UF, i n a manner such that i t can escape detection by petrographic examination.

UF, in 5inary Alkali-Metal Fluoride Systemno

A number of thermal analyses followed by pefrographic examination of the solid products have been performed with UF, i n each of the N a F - L i F , These NaF-KF, and LiF-KF eutectic mixtures. specimens were prepared, i n each case, by reducing the desired UF4-alkali-metal fluoride mixture with 10% excess metallic uranium i n n i c k e l crucibles blanketed with inert gas and fitted with stirrers o f nickel. In the N a F - L i F system the lowest liquidus temperature observed was at about 630째C at 10 mole % UF, and 36 mole 96 NaF. In all cases studied (up t o 30 mole 75 UF,) only the free alkali-metal fluorides and the NaF-UF, complex compound (Na,U,F,) were observed, I n the NaF-KF-IJF, system the red compound

K,IJF, appears a 5 the only colored species i n the mixtures containing up t o 10 mole 96 UF,; optical properties are: slightly different from those o f the compound crystallized fromNaF-free mixtures and may indicate s l i g h t solid solution of N a F i n the crystals. Mixtures containing 20 to 40 male % IJF,

show increasing quantities o f a blue phase with optical properties which suggest that i t i s a and K,U,F,. l-iquidus solid solution of Na,lJ,F,

temperatures i n this system seem i o be higher than 700째C in all cases. Solutions of LIF, in the LiF-KF binary system show liquidus temperatures i n the range 485 to 60Q"C, Petrographic examination of thesenaterials in those shows free LiF i n 0 1 1 specimens, K,LJF, containing less than 10 mole 96 UF,, and K,U,F, i n those containing much above 10 mole % UF,. Chemical analyses were not performed on any of these materials, but i t i s l i k e l y that considerable

UF, was present i n these specimens, especially tit

the higher temperatures, Additionnl studies o f these potentially valuable systems w i l l be made. P U R I F I C A T % O NO F R U B l D l l J M F L U O R I D E

C. J. Barton, Materials Chemistry D i v i s i o n D, L. Stockton, QRSQKT A l l the rubidium fluoride obtained

from commercial

suppliers to date has contained considerably more than the specified quantity of cesium compounds, and therefore almost a l l of it has been returned to

the vendors for further processing.

However, the

immediate need for small amounts o f pure material for phase equilibrium studies and for rubidium metal production has necessitated purification of some material at QRNL. Experiments16 have s h o r n the feasibility of separating rubidium froni cesium with the ion exchange medium Amberlite IR-105. T h i s resin was, accordingly, used in a l l the experiments conducted. T h e resin column consisted o f a 3-in.-dia glass pipe containing a 6-ft bed of the resin. The column

was loaded for each run with about 350 g o f crude RbF known to contain about 20% CsF. Elution of

the rubidium was accomplished in about 35 l i t e r s o f solution; the purity o f the product was not appreciably affected by elution rates i n the range 30 to 60 ml/min. Elution sf cesium from the column, which was observed by use o f Cs137, required 65 to 70 l i t e r s

0.5 M (NH,),CO,

o f the (NH4),C03 solution; elution \NOS n o t greatly improved by use of a 1.5 hi solutiori. From a total

o f 1775 g of crude RbF charged in f i v e different runs, somewhat more than 1 k g o f RbF containing l e s s than 1% CsF was obtained. The K F content (0.6% KF) was not appreciably affected by t h i s treatment. It i s apparent that t h i s purification can be conducted on a large scale a t reasonable cost i f inaterial sufficiently free from cesium cannot be obtained froni commercial sources. CHEMlCAL REACTIONS IN MOLTEN SALTS

L. G. Overholser F. F. Blonkenship W. R. Grimes Materials Chemistry Division

Cheinicol Equilibria in Fused Salts

C. F. Weaver Materials Chemi stry Division

J. D. Redrnan

'The apparatus and techniques for experimental determination reactions

constants

CrO

Feo

+ 2UF4 -+ FeF, + 2UF,

and

2UF,eCrF,

i n molten NaZrF, ',I.

equilibrium

for

the

2UF,

were descri bed previously, l7 and

R. Hiygins, Chemical Technology D i v i s i o n , QRNL,

work lo be published.

17L. G. Overholser, J. D. Kedmoo, and C. F. Weaver, ANP Quar. Prog. Rep. ,Mm. 10, 1954,ORNL-1692, p 56.

PERIOD ENDlNG SEPTEMBER 70, 1954 values were given for the "equilibrium constants"

obtained with UF, and the metal as starting materials. During the past quarter, the values have been checked with various ratios of UF, to UF, as starting materials in the NaZrF, solution. The apparatus and d e t a i l s o f t h e experimental technique were the same as i n previous experiments. The

results o f the recent series o f experirnents i n which

2 g of hydrogen-fired (120QOC) Cro and various

UF,-to-UF, ratios were used are shown in Table 5.4. In the table

with the concentrations expressed as mole fractions by using the formula weight NaZrF, = 2 moles. Pt was necessary to calculate the f i n a l UF, coni n the centration, since accurate analyses for

u3+

presence o f Crft could not b e made; any error arising from t h i s calculation would be important only a t l o w UF, concentrations. T h e data show that at both T A B L E 5.4.

600

and

800째C

the

values of K x remain constant at low UF,-to-UF, r a t i o s but increase rapidiy as t h i s ratio r i s e s above 3. T h i s rapid increase in K n may be due to incomplete removal by f i l t r a t i o n of small amounts of chromium metal or to uncertainties in analysis of these trace quantities. Since, as shown above, the UF,-to-UF, ratio i n NaZrF, solution i n equilibrium with uranium metal i s about 9 at 800"C, i t i s not l i k e l y that disproportionation of the UF, i s responsible for t h e increase i n K x .

In any case, the agreement between the mean of four values o f the constant ut each temperature w i t h previously published values i s extremely good. Agreement among the various i n i t i a l UF,-to-UF, r a t i o s seems t o show that a state of equilibrium i s reached i n these experiments. When equilibrium data for the reaction Fe" t

("C)

EQUILIBRIUM DATA FOR THE REACTION Cr" t 2UF,f==!

600

800

UF, Added (moles/kg of melt)

iFeF,

were obtained by the addition of UF,, UF,, and FeF, to NaZrF, i n the same way as for the chramium reaction, t h e results obtained were not so reproducible. The data obtained are shown i n Table 5.5. The mean of the values obtained a t CrF2

2UF3 IN M O L T E N NoZrF5

Equi I i bri um*

Experimental Cond i t ion s Temperature

2UF4&2UF,

Concentroti or?

UF3 Added (moles/kg of m e l t )

KX**

of C r + +

0.358

(PP4

2080

3.0 x

0.267

0.078

9 50

0.178

0.158

150

0.089

0.238

4.4

10-4

0.035

0.287

5.7

lo-,

0.029

0.318

1.1 x 10-2

0.360

5.1 2.7 x

2660

7.4

0.280

0.08 1

1 I25

6.1

0.194

0.165

210

3.3

0.111

0.244

0.064

0.295

0.032

0.325

* V a l u e s shown ore mean of c l o s e l y agreeing values from duplicate experiments.

35 Blank contained

io-,

IO-, lo-,

4.2 x lo-'

1.8

7.9

10-3

200 ppm of Cr

**Kk values previously obtained: at 6OO0C, K x = 4.2 4 W4,a t 8OO0C, K x - 4.1 x lo-,. K by uslng Brewer's v a l u e s for A P : a t 6Q0째C,Kcq - 1; a t 800"C,Keq: c a l c u l a t e d from -l\P= R T I n K eq

0.1.

ANP QUARTERLY PROGRESS REPORT TABLE 5.5.

EBUlLlBRlUM DATA FOR THE REACTION FeO + 2 U F 4 4 2UF3 Experimental Conditions .-......

UF,

Temperature

(OC) -

600

800

_..___.

' q at

Concentration of

0.322

0.053

0.03 19

620

1.2 x 10-6

0.322

0.053

0.0346

735

1.2 x 10-6

0.326

0.083

0,0347

660

9.3 x 10-6

0.326

0.083

0.0347

6 55

9.3 x 10-6

0.321

0.036

0.0 168

995

2.1

10-~

0,321

0.036

0.0168

9 40

1.9

10-~

0.242

0.113

0.206

8600

1.2 x 10-5

0.242

0.1 13

0.103

3210

3.2

0.242

0.1 13

0.103

3270

3.7

0.277

0.077

0.103

41 10

2.7

0.277

0.077

0.103

4050

2,2

0.0318

0.322

0.213

3110

3*4 x 10-6

0.0318

0.322

0.210

3300

2.5

10-~

0.103

0.27 1

0.210

4530

1.1

10"~

0.103

0.271

0.207

4645

3.7

10-5

0.207

0.152

0.213

7940

1.3

10-5

0.245

0.1 15

0.053

495

4.2 x

0.245

0.1 15

0.053

435

2.8 x 10-6

Fe"

in Filtrate (PPld

0.245 __ ..........

6OO0C, K e q = 2 x

0.053 0.1 15 ........~ _ _ _ _ ._ .... 6OO0C, K u = 1.2 x

8OO0C, K r

by using Brewer's values R T In K eq a t 8OO0C, K e q = 1 . 4 ~lo-,.

lom5)agrees w i t h the previously reported value of 1.5 x loa5, although agreement among the individual determinations was not encouraging. The values obtained at 80O0C,however, $ 0 not agree w e l l w i t h the value of 1.5 x reported when UF, and ferrous metal were used as starting materials. In an attempt t o explain t h i s discrepancy, the experiments in which UF, and ferrous metal were The used os reactants at 80OOC were repeated. values obtained, as shown i n Table 5,6, yielded a mean of 0.9 x and agreed more closely with the previously reported values. It appears possible that the reaction of ferrous metal with UF, had not reochedequilibrium i n the reaction time allowed. 62

IN MOLTEN NaZrF5

F e F 2 Added (moles/kg o f m e l t )

calculated from --/\F0 =

6QOOC (1 x

FeF2

UF3 Added (rnoledkg of m e l t )

Added

(moles/kg of melt)

* V a l u e s previously reported:

I _

for

265

....

1.3 ~

10-~

.......____

1.5 x lo-'.

AF0:

T A B L E 5.6. EQUlLlBRlUM DATA FOR THE REACTlON O F FERROUS METAL WlTH U F 4 IN MOLTEN NaZrF5 AT 800째C

UF4 added: 0.360 n i o l e k g of m e l t ..........

-... ~

___

Concentration*

of ~ e + i+n F i l t r a t e

520 390 330 480

1.7 x 5.4 x 10-7 2.8 x lo-' 1.1 x 10-6

(PPm)

*Blank of 100 ppm t o be subtracted from determined v a l u e s i n calculations.

PERIOD ENDING SEPTEMBER 70, I954 Stability of Chrornilam Compounds in Molten IF! uor ides

J. D. Redinan

C. F. Weaver

Mater i al s Chem i s try D i v i s ion

It was shown in previous experiments that CrF, or some complex compound o f divalent chromium

was fhe corrosion product in the ZrFdobase fluoride melts o f interest. However, when N a F - K F - L i F mixtures, with and without UF,, were tested in equipment prepared from al ley s containing chromium, the reaction products found were complex compounds of

CrF,.

Since films that were apparently

NaK,CrF, or similar materials have been observed in some experiments, attempts have been made t o observe the s t a b i l i t y and s o l u L i l i t y o f CrF, and

CrF, in typical s a l t mixtures o f the two types. In the temperature range 600 to 8RQ째C, CrF, i s n o t stable in NaZrF, solution in contact w i t h equipment of nickel. Examination of filtered specimens shows that the reaction

2CrF,

could be based. The Ni" was found to reduce so rapidly that the controlling step was the rate of removal of hydrogen fluoride from the melt; hence the chemistry of the process presents no problem o f consequence from an engineering standpoint. With the completion of measurements at

800'C

during the

past quarter, these studies have been di scontinued. T h e experimentol method used for the measurements at 800째C was the same as that used i n previously

?Oo째C.'8

described experiments at 600 and Figure 5.5 shows a comparison o f the

e f f e c t of temperature on the rates o f reduction for four concentrations. The slopes of the curves show that the apparent activation energy for the rate determining step i s about 7800 cal, T h i s energy of activation i s i n accord with the hypothesis that t h e reaction i s heterogeneous and i s catalyzed by n i c k e i surfaces.

18G. M. Wutsan, C. M. Blood, F. F. Rlankenship, ANP Quar, Prog, Reg). MUT. 10, 1954, QRNL-7692, p 62.

+ Ni ,A NiF, + X r F ,

proceeds to essential completion,

The solubility

6 r F 2 i n the NaZrF, solvent seems to exceed 6000 parts of Cr per m i l l i o n parts of salt at 600%; s o l u b i l i t y at 800째C seems ta be a t least 15,000 ppm of

and may be consideratly higher. In the N a F - K F - L i F system it appears that CPF, i s not stable. Whether the disproportionution reaction

3CrF,-dCr0

i2CrF,

or some other mechanism i s involved 1 5 not evident from the data available. It appears from preliminary

evidence, however, that the s o l u b i l i t y of Cr3' i s several thousand parts per m i l l i o n at 600OC. Reduction of

NiF, by H, in

C. M. Blood

NaF-ZrF,

H. A.

Systems

G. M. Watson Materials Chemistry D i v i s i o n

T h e use of hydrogen as a reducing agent i n the removal of o x i d i z i n g impurities from fluoride melts h a s been routine for almost two years. One of the introduced principal impurities removed i s the Ni as a result of treating the melts w i t h hydrogen Since fluoride in n i c k e l containers a t 80OoC, hydrogen was f i r s t employed in t h i s fashion, k i n e t i c studies have been under way in order to supply data on which improved fluoride production procedures

' fl~ .............

Friedman ____ .................

J.....I

......

.........

Fig.

5.5.

A.-

..............

Effect of Temperature and ConcentraN i f , b y H, in Molten NaF-ZrF1 (53-47 mole %). t i o n on the Rate of Reduction of

AN P QUARTER 1Y PROGRESS R EPORJ The data presented here were a l l obtained a t roughly the same gas flow rate (100 ml/min). However, since the rate o f reduction i s dependent on the flow rate and on the geometry of the a p p e ratus, the aeswlts must be considered merely as representative and are subject to some r e r i a t i o n with conditions; they apply to 3-kg batches in 4-in.-dia pots w i t h hydrogen bubbles r i s i n g from t h e notched end o f a %-in. vertical d i p tube immersed to a depth o f 5 in. Reduction of

FeF, by H, i n NoF-ZrF,

C. M. Blood

System5

M. Wotson

Materials Chemistty D i v i s i o n

Since the reduction of FeF, with hydrogen i s the time-consuming step in purification o f NaZrF, mixtures for reactor use and testing, an extensive study of the equilibrium constant of t h i s reaction has been attempted, The reaction FeF,

+ 2HF

H,+Feo

o f FeF, or CrF,, the HF concentration of the effluent H, was studied as o function o f the H, f l o w rate. The data obtained are plotted in Fig. 5.6, i n which the ratio of HF concentration at a given f l o w rate t o that found at a standard flow rate (210 ml/iTIiil) i s plotted against f l a w rate. As anticipated, the ratio drops at high flow rates; however, the ratio shows no sign o f l e v e l i n g o f f a t low f l o w rates, as would be expected. Although the curve shows no j u s t i f i c a t i o n for the practice, the lowest f l o w rate found to be practicable ( 9 m l h i n ) was considered t o be equivalent to a zero flow rate (that is, the f l o w rate at which the HF concentration could be expected to be at equilibrium), and experimental values found i n subsequent experiments were corrected to the concentration that they would have shown at such flow rate. UNCLASSIFIED 2926 ORhL L7-D:K

i s relatively easy to study, since the hydrogen fluoride concentration in hydrogen t a n be accurately determined, and, after f i l t r a t i o n of h e melt, accurate determinations of FeF, i n the melt can be obtained. Since at these elevated temperatures the a c t i v i t i e s of hydrogen and hydrogen dluoride can be assumed to beequal to their partiul pressures i n the gas phase, then

z 0

2 E z z

STANDARD FLOW R4TE = 210 ml/mln

-,AFo = KT In K c q and

=-,K e 4 Kw

the a c t i v i t y coefficient for FeF, a t low concentrat i o n s i n NaZrF, should be e a s i l y determined. The equilibrium constant for the reaction has been a dynamic estimated by two different methods: method and an equilibration metliocl. In preliminary experiments in was passed through molten NaF-ZrF,

Dynaraic Method,

which

(53-47mole %) 64

contaminated with know1 quantities

Cr Fz

+ H p 5 Fe" +

2tiC

+ 2HF

: a d

200

100

300

400

FLOW RATE (rnl/min)

and

from the re1ati on ships

FeF2

where C F ~ F ,i s expressed i n mole fraction,

Fig. 5.6. Effect of Gas Flow Rote on R e l a t i v e Saturation of E f f l u e n t Gas w i t h HF When Passed Thrcugh Molten NaF-ZrF, (53-47 mole %) Ccntomineted w i t h Known Quantities of FeF, or CrF,. The experiments were performed by adding known quantities o f FeF, to an NaF-ZrF, (5347 mole 76) m e l t in a clean n i c k e l reactor and measuring the HF concentration o f the effluent hydrogen passed at carefully measured f l o w rates. The cumulative t o t a l of HF removed was also determined by c o l l e c t i o n o f die HF in caustic solution and t i t r a t i o n o f the excess caustic.

The FeF,

concentration a t

specified time could be calculated from the FeF, found by analysis on completion of the experiment and the HF y i e l d after the specified time. any

PERIOD ENDING SEPTEMBER 70, 1954 T h i s procedure, which i s simple, rapid, and capable of high internal precision, yielded the values plotted as the lower curve i n Fig. 5.7. The

logarithm df the HF concentration i s plotted against the logarithm of the FeF, concentration at various flow rates of hydrogen. When these values are corrected to the “zero flow rate” (9 rnl/min) by

reference to Fig. 5.6, they y i e l d a consistent straight l i n e after an i n i t i a l peak. T h i s i n i t i a l peak i s probably due to the reaction of HF at i t s i n i t i a l high concentrations with the clean (hydrogen-fired) n i c k e l apparatus and i t s subsequent re-equilibration

w i t h the more d i l u t e HF-H, mixture. The straightl i n e portion of the curve shows the expected slope

UNCLASSIFIED ORNL-.LR-DWG 2927

io2

1 o4

CONCENTRATION OF FeFz ( p p m F e t f )

Fig. 5.7.

800°C

Equilibrium Concentration for the Reduction of

(Obtained by Dynamic hethod).

FeF,

NaF-ZrF, (53-47 mole %) a t

ANP QUARTERLY P R O G R E S S R E P O R T

0.5 and agrees closely with the slruiyht l i n e

thot o f the effluent. At that time a sample was drown through a sintered nickel f i l t e r and analyzed t o determine the FeF, concentration.

HF and W, are proportional to their partial pressures at 8OOOC and

KFaHF saturated with KF depend only on tempera-

calculated from Brewer’s value o f AFa for the reaction. The values obtained for K x and for Y F ~ on F ~ the assumption

that the a c t i v i t i e s of

that

Al:OFeF2

-45.4 kcal per atom of F

and

LIFO,,

-65.8 kcal per atom o f F

are shown i n Table

5,7,

The QgiYXment of the

values obtained over CI fourfold concentration runge

o f FeF, i s extremely good. It appears that FeF, i n molten N a F - Z r F d i s i n nearly ideal solution.

T A B L E 5,s. EQUlLlBRlUM CONSTANTS AND A C T i V l T Y COEFFICIENTS AT 800OC IN MQLTEN NcsF-ZrF4 (53-47 mole X) FOR THE REACTiOW

t F e F 2 e F e o

2HF

Since

the decomposition pressures of molten

ture,19 definite mixtures o f H, and H F could be obtained while hydrogen gas WQP passed over such mixtures held at a constant temperature. The mixtures of H, thus obtained were used as influent for the reactor containing the salt mixture. I n a typical experiment i n which t h i s method was

used, a batch

(“3

kg) o f previously purified NaF-

ZrF, (53-47 mole %) mixture W Q S loaded i n t o a clean nickel reactor. The mixture was heated t o 80OoC and further p u i i f i e d with hydrogen u n t i l a very low HF concentration (“3 x mote/Iiter) i n the effluen* gas denoted high purity of the melt. The mixture vias allowed to cool to roam temperature, and FeF, and iron wire wereadded to the cold,

frozen mixture under un atmosphere of argon. The mixture was then heated, melted, and brought up t o temperature while i t was sparged with helium gas. The reactor containing the KF*HF or HF generator SVQS heated separately to a convenient i n i t i a l

1598 1513 1458 1372 1318 1233 1122 1067 983 900 8 57 760 606 5.50 48 1

2.15 2.15 2.17 1.98 1.91 2.37 2.07 2.00 2.11 2.18 2.11 2.17 2.19 2.24 2.19 A v 2.13

1.47 1.47 1.48 1.36 1.3 1 1.62 1.38 1.37 1.44 1.49 1.45 1.49 1.50 1.S4 1.50

f 0.22

1.46 f 0.16

Equilibration Method. Since i t was doubtful that the lowest flow rate found to be practicable (9 ml/min) actually represented equil ibriurn conditions, an attempt was rnode to determine the equilibrium conditions directly. A mixture o f HF and H 2 was bubbled through a m e l t containing Fao and FeF, u n t i l the influent HF concentration matched

temperature and connected to the system. Pure hydrogen was passed i n parallel through both reactors - the HF geiuerator and the equilibration reactor in order to clean out any NiF, films on v m l l s and to precipitate n i c k e l impurities. After 1 to 2 hr o f parallel operation, the reactors were connected i n series, and the i n l e t on$ effluent

gas streams were continuously sampled and anaThe temperature o f the furnace o f the lyzed. generator was then adjusted u n t i l the analyzer

showed H,-HF mixtures of the desired composition. The gas mixture was allowed to bubble through the system for extended periods o f time, and samples were taken intermittently for analysis. When i t was considered that the system had approached equilibrium, as indicated by a convergence o f the influent and effluent compositions, the HF generator was removed from thz systern and helium was sub-

stituted for the influent gas mixture. A filtering tube w a s introduced into the melt under an additional argon atmosphere and a f i l t r a t e was obtained. The filtering tubs with the frozen f i l t r a t e was removed and analyzed for iron. This procedure was repeated with appropriate variation of the temperaturf: o f the H F generator and consequent variation o f the influent gas composition, 19G. H. Cndy, /. Am. Chern. SOC. 56, 1432 (1934).

PERlOD fNDING SEPTfhlBER 10, 1954 The data obtained are presented in Table 5.8, which gives the concentration of iron found i n the filtrate, the ratio o f the HF concentrations i n the influent and effluent gases, and the equilibrium constants and corresponding a c t i v i t y coefficients for the F e F As i s evident, the data i n Table 5.8 2: are less precise than those obtained by the dynamic procedure. T h e lowered precision i s probably due t o the more complex apparatus and technique. However, when the data are plotted (Fig. 5.8), the spread of the data i s l e s s than the estimated unThe agreement becertainty i n the A F O values. tween the values obtained by the two methods is remarkably good,

TABLE 5.8. EPUlLlBRlUM DATA FOR THE REACTION FeF2 + H,+Fe' t 2HF AT 800°C IN MOLTEN NoF-ZrF4 (53-47 m o l e %) Concentration of of FeF,

R a t i o of HF in

P2HF

Influent G a s to

H F in

(ppm Fe)

Effluent G a s

912 223 915 7 40 360 185 160 27 5

0.79 0.9 1 1.02 0.98 1.00 0.93 0.9 1 0.99

YFf2F2

0.40 1.08

0.4 1

2.09 2.12 1.71 1.89

1.43 1.45 1.17 1.30 1.47 2.14

0.74

2.14 Av

3.12 1.84

1.25

Preparation of Various Fluorides

E. E. Ketchen Materials Chemistry D i v i s i o n

B. J. Sturm

The simple structural metal fluorides have continued t o be used extensively for numerous purposes. A limited interest also has been maintained i n several of the complex fluorides formed by the interaction of a l k a l i fluorides with structural metal fluorides. Some effort also has been expended on the preparation of other simple fluorides, as well as on the purification of small quantities of the structural metal fluorides. The identity and purity o f these fluorides have been established mainly by chemical analysis supplemented by x-ray and petrographic data i n some cases.

The with an 0.98, was prepared by heating Ag,CB,

ratio mder

a t 150 to 200°C. An attempt to prepare AgF by reacting HF with AgCN was unsuccessful; most o f the AgCN did not react. Several botches of anhydrous ZnF, were prepared by heating ZnF2.4H20 The compound T e F 4 was under HF a t 600°C. prepared by reacting Te and HNO, and reacting the resulting TeO, with aqueous H F to give T e 0 2 F 2 . H 20p which was thermally decomposed to y i e l d Te02,

HF, H20,

and the product TeF,. The TeF, was collected in a condenser held at 130 to 140°C. Small quantities o f Rb,CrF, and Li,RbCrF4 were prepared by heating the proper quantities of CrF3, RbF, and LiF i n sealed capsules a i 900°C. The Rb,CrF, has properties approximating those of a corrosion product found i n a rubidium-base fuel. T h e compound RbCrF, was prepared by heating CrF, and RbF a t 800°C i n a sealed capsule. Numerous batches of structural metal fuel- or coolant-type fluoride complexes were prepared. Among these (prepared by methods previously

K,Ni F, and KNi F,. described) were: K,CrF,, The compounds NiF,, FeF,, and F e F 2 were prepared i n a high state of purity by hydrofluorination of the corresponding chlorides. F U N D A M E N T A L C H E M I S T R Y O F FUSED S A L T S

EMF Measurements L. E. Topol Mater ia1 s Cherni s try Di vi s ion Additional measurements of decomposition potent i a l s of KCI and o f various chlorides i n molten KCI at 8 5 0 T were mude. Morganite (AI 0 ) ? 3 crucibles were employed with a l l the melts studied. Platinum and n i c k e l were used as cathodes, and the anodes were of nickel, carbon (graphite), chromium, and zirconium. The solutions of various chlorides i n KCI were prepared by melting 2 moles of KCI with 1 mole o f the anhydrous chloride i n a sealed, evacuated quartz tube. More d i l u t e solutions were prepared by adding appropriate quantities of KCI to the 33.3% KCI mixtures so obtained. The data obtained recently are given i n Table 5.9.

T h e value of 1.5 v found upon electrolysis o f KCI with an anode of electrolytic chromium agrees with the value previously reported for chromium-plated gold wire i n an atmosphere of helium,20 The more recent value of 1.27 v for electrolytic chromium i n

T. E. Topol, A N P ORNL-1692, 63. -

_ I _ _ _

Quar. P m g . R e p . MUF. 10, 1954,

ANP QUARTERLY PROGRESS REPORT UNCLASSIFIED

0 RNL- LR-- DW G 29 2 B

q0-2

CALCUL ArED A F o FOR

t R O M FREE-ENERGY

FeF2=

FXPtRlMCNTAL

65 4

DATA (BREWER)

'I k c o l

DATA

FROM FREE-ENERGY ----- CALCUC4TLD OFo FOR FeF, = 68 0 k c o l 10-~

1O2

DATA (GI ASSNER)

1 o4

CONCEN.rRATlON FeF2 (ppm Fe++)

Fig, 5.8. Equilibrium Concentration for the Reduction of FeF, b y H, in )41aF-ZrF4 (53-47 800째C (Obtained by Equilibration Method). a hydrogen atmosphere may indicate that a hydrogen electrode i s formed. T h e reaction of UCI, with zirconium metal n ~ i y be ascribed to the following equations:

2UC1, + 3 Z r

3ZrC1,

SZrCI,

I-

21J, !IFo = -10 kcal,

and

4UCI,

3Zr

+ 4U,

A F o - 1-130 kcal,

mole

%) a t

Current-voltage measurements indicate the f i r s t reaction to be predominant. The E o o f 0.4 v for the reaction of

UCI,

2UCI,

with a chromium anode, that is, t

3Cr

3CrCI, t 2U , E o (theoretical) = 0,90

seems to indicate strong complexing 3f the CrCl 2. There seems, from studies by other groups, to be

PERIOD ENDING SEPTEMBER 70, 1954 T A B L E 5.9. DECOMPOSITION P O T E N T l A L S OF VARIOUS CHLORIDES Concentration o f Salt in Solute

Salt

(mole %)

Electrodes Cathode

Anode

Blanketing

Potential 0bserved

Atmosphere

(V)

100.0

1 .so

UCI3

33.3 33.3

Pt Pt

Zr Cr

He He

Spontaneous 0.40

u CI,

33.3 33.3 33.3 3.0 3.0 3.0

C C

H2 He

0.95, 0.83, (0.75) 0.83, 0.95 0.60 to 0.80 1.84 t o 1.87

K CI

Pt Pt

NiCI,

33.3 3.0

Ni Ni Ni

C C C

2UC14 - 2UCI,

( 2)

UCI,

(3)

5UCI,

4UCI,

a,,

2c4,

1.18 v, =

1.9

E O = 1.6 v,

Uranium metal i s deposited on the cathode; so the f i r s t reaction does not occur alone. I f t h i s process takes place i n coniunction with the transformation =

3UC14 +

E o := 0.85 v,

a continuous depletion in the UCI, concentration, which i s low i n i t i a l l y , would result (more UCI, would be oxidized i o UCI, than would be replaced at the cathode). With low concentrations of UCI,

(1.15)

0.34

0.80 to 0.83 0.35 to 0.45

1.45 1.50

He He He

1.00

0.93, 0.82 1.13 to 1.17 1.20

_ 1 _ -

at inert electrodes upon passage of current. Since the measured E ' s are somewhat l e s s than the theoretical estimate of 1.99 v, it i s believed that either some depolarization occurs or that the estimate is too high, or that both are responsible. Concentrated UCI, solutions may undergo the foli o wing electrochemical tran sformati on s at platinumgraphite e Iectrodes:

4uc1,

He He

Ni Ni Ni

UCI, decompose into U and

CI,

H2 He

2.0 1.0 10.0

l i t t l e evidence for t h i s behavior i n fluoride melts. D i l u t e solutions of

C Ni C Ni C C C

1 .o

33.3 33.3

FeCI2

(4)

KCI A T 850'C

l-__l_________...._l________

_ I

(1)

_ I _ _ _

the reaction

(5)

2UCI,

+ 3CI,,

E O

2.2

has been found to predominate. The high value of for reaction 2 eliminates i t s consideration a t these voltages, The decomposition potential of reaction 3 i s high also but may be accounted for that i s by an a c t i v i t y ratio o f (UCI,)/(UC1,)5'4 approximately equal to i f the estimated AFD's of -191 and -183 kcal for UCI, and UCI,, re-

spectively, are correct. The reaction of dilute and concentrated UCI, solutions and nickel anodes may be attributed to

2UCI,

2UC1,

-t

NiCI,,

L-

0.36

D i l u t e UCI, solutions seem to show signs o f reduction i n a hydrogen atmosphere. concentrated and d i l u t e FeCI, solutions also undergo different reactions at. inert electrodes. In concentrated solutions FeCI, i s oxidized to F e c i 3 a t the onode ( E " :0.35 v), whereas in dilute solutions decomposition into the elements occurs (E" = 1.11 v). With nickel anodes, both the fol1 owing reactions probably occu r si ni u It aneou s I y :

FeCI,

NiCI, + F e ,

0.30 v, 69

4NP QUARTERLY PROGRESS REPORT on d

3FeC12 = 2FeCI, + The NiCI, solutions, as

0.35

far a s i s known, undergo the same electrochemical reactions regardless of concentration. Thus the changes i n decomposition potential with concentration may be assigned to differences i n NiC!, a c t i v i t y alone. With the a i d o f the Nernst equation, the NiCI, a c t i v i t y at 33.3 mole % NiCI, may be coniputed to be 100 times that i n the 3.3 mole 76 solution and 300 times that i n the 1 mole % solution. Solubility of Xenon in Molten Salts

R. F.

Newton, Researcli Director's Department D, G. Hill, Consultant

The use of molten salts as reactor fuels offers the p o s s i b i l i t y of removing the gaseous fissionproduct poisons, of which zmon is the rnost important, If, for example, the xenon concentration o f the SO-Mw CFRE can be held at 10% of i t s equilibrium value (1.1 x m o l e of xenon per m i l l i I iter of fuel), important perturbations i n r e a c t i v i t y can be avoided, In on aircraft reactor, however, i t w i l l be nccessary to effect t h i s removal nd xenon with on absolute minimum o f ouxiIiary equiprrient.

T h e available literature appears to contain no measureinents o f she s o l u b i l i t y o f gases i n molten salts. Accordingly, i t h a s seemed desirable to evaluate the s o l u b i l i t y of xenon in molten fluorides over the temperature range o f interest. Preliminary t e s t s indicate that the f o l l o w i n g experimental technique i s satisfactory. T h e apparatus, which i s constructed of n i c k c ! where i t will be i n contact with the fluoridc melt and of glass elsewhere, consists of two sections that can be isolated by freezing a p l t ~ gof fluoride i n the U-tube connecting them, I n one section of the apparatus the m e l t i s allowed to saturate w i t h xenon under predetermined conditions of ternperature and pressure. Melting o f the fluoride p l u g permits the saturated salt t o f l o w i n t o the other section, whi!e the 1J-tube seal prevents entrance o f gaseous xenon into this section; refreezing of the 117 the second p l u g isolates the sections again. section, xenon i s stripped from the salt by repeated circulation o f the hydrogen contained i n the a p p r a t u s through the s a l t and past a liquid-nitrogencooled trap. After the stripping process i s complete, the hydrogen i s removed, $he xenon i s allowed to come to room temperature, arid he amount col-

lected i s ascertained by measurements i n a modif i e d Mcl_eod gage. Preliminary experiments at 600'6 and essentially 1 atm of xenon indicate that the s o l u b i l i t y in the

N a F - K F - L i F eutectic i s n o t greater than l o M 7 mole of xenon per m i l l i l i t e r o f salt. The method and technique proved to be feasible, but minor experimental d i f f i c u l t i e s necesFitatsd modification of the nickel apparatus. More accurate data under these and other conditions w i l l be available i n the near fWtIJre. Meanwhile, i n order to gain familiarity with the technique, the s o l u b i l i t y o f xenon in the KNO,NaNQ, eutectic (36 mole '36 NaNQ,) has been measured i n an apparatus constructed entirely o f glass. The s o l u b i l i t y of xenon i n t h i s melt i s about 8.5 x mole/itil a t 280째C and l o w 7 mole/ml a t 360째C. This l i q u i d apparently shows the p o s i t i v e temperature coefficient o f s o l u b i l i t y expected for gases in molten salts.

%-Ray D i F f r ~ ~ f Studies i~n i n Salt Systems

P. A. Agron

M. A.

Bredig

Chemistry D i v i s i o n

Cesium Halides.

It has been suggested21t22 that

the increase i n volume on melting o f simple binary salts i s indicative ofstructural changes i n the melt, that is, o f a decrease to a lower ionic coordination. Volume increases up to 30% have been indicated for the a l k a l i halides. The x-ruy data23 on the thermal expansion of s o l i d CsBr and Csl gave volume changes of approximately 26%, but an extrapolation o f about 50째C to their m e l t i n g points WQS required for the computation of these changes. It WQS therefore of interest to uscertain whether a solid-phase transition analogous to the one known to e x i s t in CsCl occurs i n the unexplored temperature interval. T h i s would replace the assumed s t r u c t u r a l change i n the melting process by one in the s o l i d state. High-purity CsBr and Csl were intimately mixed with fine nickel powder to provide airs internal stairdurd whose thermal expansion i s k n o w BCCUr a t e ~ y . , ~ ~ a t t i c eparameters a s a function of temperature were obtained on the hi yh-temperature M. A. Bredig, and W. J. Smith, G e m . R e p . D e r . 31, 1952, ORNL-1482, p 32. P ,, A. Agron and M. A. Bredig, C h ~ r n .Semzann. Prog. R e p . J u n e 20, 1954, ORML-1755 ( i n p r e s s ) .

21J. W.

Dzv.

@LOT.

Johnson, ?Tog.

24LsJo;dan and W. H. Swunger, J. Research N n t . Bur. xtandn7ds 5, 123 i ( 1930)-

PERlQD ENDING SEPTEMBER 10, 1954

(50-50 mole %)

x-ray diffraction unit i o within 5째C of the melting points. No transformation was observed i n either s a l t to within 2 or 3OC o f the fusion temperature. The volume changes for these s o l i d s from 25% to the fusion point were 12% for CsBr and 11.2% for

o f the l i q u i d d e n s i t y z 7 of N a F - Z r F 4

with volumes extrapolated for these molten salts2' down to their respective melting points shows that an increase o f approximately 27% tokes place on melting. These data indicate that a structural change, similar to the lowering of the i o n i c coordination from 8 to 6 in s o l i d CsCI, occurs on fusion.

g/cm3. Thus a relatively large density and volume chunge of 26% i s indicated for this complex s a l t on melting. The volume increase may be considered as indicative of a decrease i n ionic coordination in the fused state similar to that observed i n the C s B r and csl soits.

Csi. Comparison o f the molar volumesof the s o l i d s

It was desirable to determine t h e NaF-ZrF,. magnitude o f the volume change found in complex N i c k e l powder was admixed to s a l t s on melting. a sample of a homogeneous phase having the comThe l a t t i c e position of NaF-ZrF, (53-47 mole %)a expansion was measured up to 491째C. The data are l i s t e d in T a b l e 5.10. T h e volume expansion from 25째C to the melting point i s 3.8%. The volume i s an almost linear function of temperature. The thermal coefficient of volume expansion, CL,,, I S

7.75 x

deg-'. Thus the molar volume of the solid a t the melting point of 52OOC gives a density o f 4.00 g/cm3. The latter value i s based on a molecular weight o f 1287.5 g, corresponding to the presence in t h e unit c e l l of 6.77 NaF i- 6 ZrF, molecules26 for t h i s composition, Om extrapolation 2Sfnfernatzonul Cntrcul 'rubles, Vol. 3, p 24. 26P. A. Agron and M. A. Bredig, ANP Quur. P r o g . Rep. June 10, 1954, ORNL-1729, p 47. T A B L E 5.10.

Run No.

The ionic coordination in the s o l i d i s assumed to be similar to that in NaUF;,, for which Zachariasen proposed a fluorite type of packing with equal and Zr4+ of the coordinations for both the Na' order of 8 or 9. It seems possible that the volume

increase on melting results from a lowering in the coordination o f the fluorines around Zr4 ', which might also be considered as radical ion formation around zirconium; t h i s point requires further study. Well-formed single crystals which originated from the central portion o f a large melt of N a F - Z r F 4 (53-47 mole %) have become available. Singlecrystal x-ray analysis has been i n i t i a t e d in an attempt to make a complete structure determination, especially of the positions o f the fluoride ions.

27S. I . Coheri m d T. N. Jones, A Surnmusy of I)ensity Measurements on Molten Fluoride Mixtures rmd n Comelotion Ilsejul zn P r e d i c t i n D e n s i t i e s of Fluoride Mixtures of K n o w C o m p o s i t i o n s , 6RNL-1702 (May 14, 1954).

T H E R M A L EXPANSION OF

Temperature

("C)

at above 6W0C t o 520째C, a value of 3.32 g/cm3 for the l i q u i d i s obtained. An adiusfment of t h i s to the composition o f N a F - Z r F 4 (53-47 mole %), by assuming an almost constant molar volume, lowers the density to approximately 3.2 measured

NoF-ZrF,

Hexagonal C e l l Dimensions (A) ll__....__....-__l

_ I _

(53-47 mole

Mol or \bolumo

Den si ty*

(cin 3

(g/cm3)

II-1

13.79

9.407

309.60

4.158

-4

190

13.84

9.453

313.38

4.108

-2

250

13.87

9.475

315.62

4.079

278

13.89

9.482

316.64

d.066

378

13.90

9.497

3 17.9Q

4.050

395

13.90

9.504

318.14

4.047

A9 1

13.95

9.525

321.18

4.009

321.50

4.005

-7

-3

-6

-8

Extrapol ation

5 20

* T h e density c a l c u l a t i o n i s based on a u n i t c e l l c o n t a i n i n g 6.77 N a F and 6 ZrF4 molecules per unit rhombohedral

c e l l at this composition.

ANP QUARTERLY PROGRESS R €PORT PHnDUCTION OF PURlFlED

P hy s i c a I Chem is try

M O L T E N FLUORBBE5

E, R. Van Artsdalen

F. F. Blankenship G, M. Watson G. J. Nessle

Chemistry D i v i s i o n

The heat o f combustion and the heat o f formation have been determined for boron carbide, The cornbustion process i n a high-pressure bomb i s both incomplete and nonstoichiometric, i t yields some free carbon, but the amount depends upon certain extraneous conditions. However, it i s possible to obtain a f a i r l y r e l i a b l e value for the heat of the reaction

B,@ i-40,---?

2B,O3

+ 20,

by allowing for incomplete combustion and nonstoichiometry as determined by direct analysis o f products. The heat of combustion i s W;,,., -583.5 + 2.2 kcal/mole. From t h i s value and o t t e r established thermal data, i t i s found that the heat = -14.1 rt 2-7 kcal/i-nole o f formation is AH;,,.,, = and thot the free energy of formation i s 3F 2

;,,.,

-13-9 kcaVmole,

The low-temperature heat capacity of molybdic oxide, MOO,, has been measured in the range from 15 to 30OoK, arid the results are in fair agreement w i t h the work of Seltz, Dunkerley, and DeWitt,’, who measured the heat capacify above 6Q°K. How-

ever, low-temperature extrapolations by these 3 authors according to the Debye T law ure i n error.

Molybdic oxide has a layer structure, and i t s heat capacity between 15 and 60°K varies a5 T 2 i n the manner found for certain other crystals with layer l a t t i c e structure, such as boron nitride.30 The f o l l o w i n g values were obtained for Moa3 at 25OC:

298,,6

18.58

eu.

17.93

Similar

cal/molendeg and SO,,,.,, studies are i n progress w i t h

molybdenum disulfide,

MoS,,

another layer com-

pound, E l e c t r i c a l conductivity and density measurements are nearly complete for the molten s a l t systern, potassium chloride-potassium iodide, which, i n a number oi respects, has hem observed to resemble 31 systems discussed previously.

28DetaiIs of this work w i l l be published i n separate reports and a r t i c l e s from t h s Chemistry D i v i s i o n . See also Chem. Senizunn. Prog. R e p . J u n e 20, 1954, Ol?bll.1755 ( i n press).

29H. Seltr, F. H. Dunkerley, and 3. J . DeWitt, J . A m Chem. SOC. 65, 600 (1943). 30A. S, Dworkin, D. J. Sasmor, and E. R. Van Artsdnlen, 1. Chern. P h y s . 21, 954 (19531, 22, 837 (1954), and Cbem. Semzuim. P r o g . R e p . J u n e 20, 1953, ORNL-1587, p 19.

Materials Chemi stry D i v i s i o n Use of Zirconium Metal as a Scavenging Agent

c, M,

Blood F. P. Boody

I-!. A. Friedman F. !4, Miles

G. M. Watson

Materials Chemistry D i v i s i o n The

most

important contaminants

fluoride

can be removed by melts (HF, FeF,, and NiF,) treatment of the melt w t h hydrogen, but t h i s process r e y u i r r s long periods for n e w l y complete re-

Attempts have therefore been moval of the FeF,. made to demonstrate the effectiveness of zirconium metal iri removing these impurities i n a short time, I n NaF-Zrf, Mixtures, Known concentrations of contaminants were added to 3.5-kg batches of previously purified NaF-ZrF, (53-47mole X) i n n i c k e l

containers

and

allowed to

remain overnight at

800°C i n contact with a considerable excess of m e t a l l i c zirconium (30 9); some stirring was assured No experiby continuous sparging w i t h helium. mental d i f f i c u l t i e s were observed either i n e q u i l i bration or in f i l t r a t i o n of the product,

I n one experiment, the 10 g of FeF, added gave an i n i t i a l contaminant concentration o f 1700 ppm Fa. After treatment, the contaminants present i n the f i l t r a t e were 70 ppm Fe, 15 ppm Ni, and 15 pprn

Cr. In a second experiment, 9 g of CrF, was added to give an i n i t i a l contaminant concentratian o f 1650 ppm Cr. The contaminants found i n the f i l t r a t e i n t h i s case were 90 ppm Fc, 2.5 ppm Ni, and 15 ppni Cr. Thus these experiments indicate that the zirconium metal addition i s an e f f e c t i v e means o f preparing pure NaF-ZrF, mixtures. In NaF,KF-L*F Mixtures, The purity of several 2-ky batches o f N a F - K F - L i F eutectic after treatment with hydrogen and with zirconium mctal at 800°C i s given i n Table 5.11. It i s evident thot scavenging with m e t a l l i c zirconium con be s u b stituted for most, i f not all, of the time-consuming hydrogen stripping process, Since the impurities i n fluoride melts are i n i t i a l l y present i n low concentrations, i t i s anticipated tliat the small aniouilt of ’E. f?.

Vun Antsdalen,

1953, ORNL-1649, 1954, ORNL-1729,

p p

ANP

Qunr.

Prog. R e p . DPC. 10,

58; A N P Quar. Prog. R e p . J u n e 10,

57.

PERIOD ENDING SEPTEMBER 70, 1954 TABLE 5.11.

Hydrogen U s e d

( I iters)

HYDROGEN AND ZIRCONIUM METAL FOR REMQViNG REDUCIBLE IMPURITIES FROM NaF-KF-LiF MIXTURES

COMPARISON OF

Zircon iurn Metal

Added

0.0

57 _ I _ _ _ _

1.4

NaF-ErF,

C. M. Blood

F. P. Boody

Ni I p p d

Cr ( w m )

20 20

I _

zirconium introduced wi I1 have a negligible effect on the physical properties o f the melt. However, i n these experiments at 8 0 0 T some a l k a l i metal was detected i n cold regions o f the equipment and some attack on the gaskets and flanges was observed. If large-scale purifications were attempted, i t would probably be necessary to use a lower temperature for t h i s treatment, Purification of

(ppm)

725

1.3

None

Found in Product

2100

0.0

None

300

None

600

___-

Z r (wt %)

(31)

500

Contaminants

Mixtures by Electrolysis

H. A. Friedman F. W. Miles

6.M. Watson Materials Chemistry D i v i s i o n

In the effort to obtain improved fuel purification procedures i t was shown that electrolysis in N a F Zrf, melts at 8OOOC with graphite anodes and n i c k e l cathodes reduces dissolved iron and n i c k e l to relatively low concentration i n a much shorter time than i s required for the customary reduction with hydrogen, Furthermore, the reduced impurities were collected on a removable cathode that could e a s i l y be withdrawn from pots with welded lids. T h i s feature promises to be of marked advantage i n processing large batches i n equipment o f a current design which allows the reduced metal from each batch i o recontaminate subsequent batches and thereby multiply the amount of reduction required, Careful attention was given to current-voltage curves at a l l stages during the electrolysis in the and apparent decomposition hope that “breaks” potentials would permit the progress of the reduction to be followed. F i l t e r e d samples were removed during various stages and analyzed for iron and n i c k e l to determine the actual amount present. To f a c i l i t a t e analysis of the current-voltage curves, the electrolysis pot was f i t t e d with “floating” anode and cathode probes, which were used to determine

the potential between the melt and the “working”

anode or cathode. The probes were made of k-in. n i c k e l rod; their potentials with respect to each other and to both electrodes were followed through many cycles of increasing and decreasing applied voltage, but the results have not been completely interpreted.

McLaren

the Chemical

Technology

D i v i s i o n i s o f the opinion that measurements with the reference electrodes have shown that the depos i t i o n of iron takes place with very l i t t l e overvoltage, A t a current of 1 amp, iron was deposited with a difference of potential between the cathode and the n i c k e l reference electrode o f only 15 mv. After depletion of the iron, the cathode potential increased to 75 rnv. T h e anode potential was found to have a marked discontinuity in the graph of current density and voltage, At some c r i t i c a l current density the anode voltage would increase suddenly. At the beginning of the electrolysis the jump was found only at high current densities, but as electrolysis continued, it was found at lower current densities. The anode lump was on the order o f 0.2 to 2 v, an order o f magnitude greater than the increase found on the cathode.

A l l the electrolyses were carried out i n NaFZrF, (47-53 mole %) mixtures i n n i c k e l containers. The electrodes were suspended from the l i d of a 3-kg capacity purification reactor (4 x 17 in.) with modified spark plugs as insulating connectors; the electrodes were immersed to a depth of about 4 in. T h e anode was a $-in,-dia C-18 graphite bar threaded to a short length of n i c k e l rod which was welded to the spark plug. D i r e c t current was supplied from a Dresser E l e c t r i c Co. 26.v 25amp selenium rectifier; measurements were made with the use o f s u i table potentiometers, voltmeters,

ANP QUARTERLY P R O G R E S S REPORT data from Table 5.12 are deceptive because some o f t h e reduction i s accomplished by hydrogen, which gives high values, and also because the current efficiency for purification becomes very low when the amount o f impurity to be reduced i s very small. When the containerwas used as a cathode, 6 v was

ammeters, and a recording ampere-hour meter. In the f i r s t experiments the n i c k e l container was used as the cathode, and helium gas was continuously bubbled through the melt i n order to stir and t o remove gaseous electrode products, The melt

was f i r s t thoroughly purified by a conventional HF treatment followed by reduction with hydrogen. Then current-voltage curves were taken before and The after the addition o f 1980 ppm F e t + as FeF,, effect o f the added irnpurity was readily noticeable. The purified melt gave a curve l i k e p l o t D o f

the maximum applied potential required to obtain a An HF concentration in the current of 10 amp. mole o f HF per l i t e r of effluent gas of about hydrogen was noted a t 5 amp. I n another series o f experiments a removable cathode, insulated from the pot, was used. The

lom3

Fig. 5.9, while the FeF,-contaminated melt yielded a curve similar t o plot A. Extrapolation of the

x 4 in. cathode was a welded t o a '/8-ins nickel o f 15-mil wire, 25 wires sions were chosen to be

curve obtained from the FeF,-bearing melt gave a decomposition potential o f 0.9 v. An electrolyzing current of 5 amp was employed, but numerous current-voltage curves were also obtained so that at the end of 5 hr, 20.9 amp-hi had passed, T h i s was approximately twice the theo-

for a small-scale model of a 250-lb capacity production reactor. Toobtoin a relatively high level o f i n i t i a l impurity, approximately 20 g of FeF, and 2 0 9 of NiF, were mixed with the starting materials. After the usual HF treatment before electrolysis, a sample was removed to determine the iron concentration. The results of the t r i a l are summarized in the second part o f Table 5.12. At the end o f 12.5 amp-hr the cathode was removed; it had gained 10.3 g of metallic iron (as computed from chemical analysis), representing a 97% yield. A new cathode was then used u n t i l a

retical ampere-hours required to reduce the Fe"; a sample removed at t h i s time analyzed 50 ppm F e and demonstrated that a f a i r l y complete reduction had been achieved with at least 50% current ebfi-

No evolution o f fluorine was noted, nor ciency. was there an appreciable attack on the n i c k e l container. The wall which had been i n contact with the vapor appeared to be covered w i t h a fine dust, presumably NiF,.

total o f 37 amp-hr had passed; t h i s cathode gained g of iron, and thus the percentage recovery had been increased to 99%. A current of 10 amp gave current densities of 11.7 amp/cm2 of anode surface. By doubling the nominal area o f gauze used as the cathode to obtain the total surface area both inside and out, the current density at the

For the next series of experiments a new batch of freshly purified melt was used, hydrogen v m s substituted for helium, and a current of 10 amp was passed. Three purifications were carried out in succession w i t h addition of FeF, and NiF, t o simulate impurities, as shown i n the f i r s t part of T a b l e 5.12. Apparent current efficiencies based on TABLE 5.12,

0.2

E L E C T R O L Y T I C PURlFlCATiON O F N a F - Z r F 4 MELTS WITH HYDROGEN STIRRING Impurities After

Impurities Before Cathode Arrangement .......

E l e c t r o l y s i s (ppm) ____...

N i c k e l container a s cathode

Removable n i c k e l gauze cathode ~

.........

*Unfortunately,

E l e c t r o l y s i s (ppm)

_____-Fe

1980

20 20

15.2

1890

2000

23.6

1980

None

11.8

100

4660

1 2.5

150

37.0

.........

___.....

....

.... ~

a r e l i a b l e analysis of the nickel concentration w a s not obtained.

**Continued without additional impurity.

Ampere-Hours Passed

......___ ........

cylinder of nickel gauze rod; the gauze was made to the inch, The dimeni n the correct proportion

......

N o t sampled

P E R l O D ENDING SEPTEMBER IO, 7954

...

ANP QUARTERLY PROGRESS REPORT cathode was found to be about 8 amp,/cm2 o f nominal gauze area. I n a l l runs under hydrogen, the effluent gas was bubbled through KOH solution to remove HF. A sample of the gas after removal of HF was analyzed by mass spectrometry and found to contain 3.17%

CH, N,

1.25% H,O

(from the KQH solution), 0.73% and only about 0.02% CF,. Evidently the The anodes predominant anode product i s HF. disappeared at a rate corresponding to Q l i f e expectancy longer than 50 ump-hr. The cathode deposits were expected to contain zirconium metal, and t h i s was qualitatively confirmed by the high

zirconium analysis (53.9 wt 5% for the cathode that gained 0.2 g of iron) i n the scrapings sf mixed metal and s a l t from the cathode, However, the zirconium metal could not be distinguished by x-ray diffraction.

The conclusion was reached that electrolysis with removable cathodes wus a practical and efficient means of purifying fluoride melts containing structural metal impurities. Experiments to investigate the removal of sulfates and oxides by electrolysis and an attempt to electralytical Iy reduce

UF,

are i n progress.

A typical set of current-voltage curves i s shown

i n Fig, 5.9. These curves apply to the experiment w i t h ternovable cathodes wnd illustrate the churact e r i s t i c “knees” or breaks in the current-voltage curves as t h e voltage i s increased or decreased. The “knees” are apparently due to a polarization phenora-ienon and show considerable hysteresis. A correlation between the current a t which breaks occur and the ornpere-hours passed i s presented i n

Fig, 5.10.

PERIOD ENDlNG SEPTEMBER 70, 1954 UF, i n ZrF, Systems. About 90% o f the anticipated reduction was obtained when large quantities

Preparation of UF3-Bearing Fuels

C. M. F. P.

Blood Boody

H. A. F. W.

G. M. Watson

Friedman Miles

Materials Chemistry D i v i s i o n Since preliminary results have shown that

UF,-

bearing fuels are l e s s corrosive to lnconel than UF4-beariny fuels, considerable effort has been devoted to studying methods for preparing these materials. The experiments performed thus far have given results that ore not as yet well understood. T h e tabulated values for standard free energy of formation indicate that a t 800OC the reaction

$U0

%UF,+UF,,

4F0 - -16

kcal,

should proceed to essential completion i f the compounds are i n their standard states. There i s ample evidence that UF, dissolved in NaZrF, i s more stable to reduction by chromium than would be e~pected,~ but the difference between the expected and the actual stability i s hardly sufficient to suggest that reduction o f UF, by uranium metal

should be detectably incomplete. The reduction o f UF, to UF, with uranium or zirconium metal has been attempted i n a large number of preparations by the following technique. About 500 g of material containing the predetermined

quantity of UF, i s treated with HF and H, i n n i c k e l equipment i n the usual fashion to render it essent i a l l y free from contaminants. T h i s purified mixture i s cooled below the melting point, and the desired quantity of uranium or zirconium metal turnings i s added. The reaction i s allowed to proceed w h i l e the mixture i n the reactor i s stirred by a stream of hydrogen bubbling through it, Samples o f the melt can be obtained by drawing a sample into a f i l t e r s t i c k of n i c k e l containing a sintered-nickel f i l t e r medium, The specimens removed from the f i l t e r s t i c k are available for petrographic, x-ray, and wet chemical exam inati on. T h e data obtained from several preparations are presented i n Table 5.13. It i s obvious that comp l e t e reduction was not obtained i n any preparation and that reduction i n the N a F - K F - L i F system was much less complete than i n the other solvents. While the number of experiments performed to date may be too small to establish reliable trends, the following pertinent observations cam be made on the basis of examination of these experiments and results.

reducing agent were used to obtain

UF, in

ZrF4-bearing mixtures. I n view o f h e extreme precautions which must be observed to prevent oxidation o f the material during preparation and handling, i t appears safe to state that about 90% o f

UF, can be reduced under optimum conditions. No a l k a l i metal was observed under the operating

the

conditions used. The extent of reduction in these systems could be qualitatively distinguished by petrographic examination.

UF, in Molten LiF. More than 90% o f the UF, in molten L i F was reduced in each case, but complete reduction was not accomplished. It i s not possible t o say whether the difference between 0.96 a t 825OC and 0.91 at 850째C i s real. No alkali metal vapor was detected. Petrographic examination detected the presence of small quontities of

UF,.

UF, in Molten NaF-KF-LiF. In f i v e experiments at 780 to 80OT the maximum reduction obtained i n N a F - K F - L i F was 48% of that expected. The extent o f reduction was only s l i g h t l y affected by an increase i n the reducing agent from 92 to 121% and by an increase i n the reaction time from 3 to 52 hr. At lower equilibration temperatures the extent of reduction was increased. The highest value was obtained at the shortest equilibration time and w i t h the largest excess of reducing agent. I n a l l the experiments at 780 to 8OO0C, a l k a l i metal vapor was evident; however, at 60OoC no It is possible that a a l k a l i vapor was observed. part o f the d i f f i c u l t y in obtaining complete reduction i s due t o reactions o f the type

3KF;=--"UF,

+ 3K0 ,

t26.1

kcal,

L\Fo =

t30.3

kcol,

A F O

and

UF,

KFeUF,

K O

where the metallic potassium being relatively v o l a t i l e is removed from the reaction by the hydrogen stream. It i s also possible that the disproportionation reaction

UF,&$UF,

+ tu,

= + 1 6kcal,

i s responsible. However, it i s especially d i f f i c u l t t o see why this reaction should be more important i n an N a F - K F - L i F solution than i n an LiF solution. Under petrographic examination the products o f the eight experiments with N a F - K F - L i F solution appeared to be identical even though by chemical analyses they showed widely different UF,-to=UF,

P,NP QUARTERLY PROGRESS REPOK'I

___

....................

I n i t i a l M e l t Composition ( m a l 2

.. .....

.......

76)

Ratio

E qu i I i brnti on __

....

ZrF4

UF,

63.7

36.3

47.0

45.9 39.9

39.9

4-0

50.1

4.0

50. i

6.5

4.5

113.8

10.8

4,5

lo,%

4,s 5.5

1.8 1.8

10.8 10.8

--. io I

11.1

95.5

40.8

95.5

43.8

8 25

780

100

A l l w w c r i c h i n re;! crystals (no gr

quontitiss) could be d i s t i n g u i s l - 4 i n t h i s system, although tctravaleiut vcciniwn v.05 found in both thc

LiF and the TrF4-hearing systems

4e:c

duction v a s vastly more mrnplstc.

These r c z v ! t s

FC-

cxamInotiuli

has generally been quite sensitive in a s c e r h i a i n g srrinll aimou~:s of UF, or UT, i n other systzms. Additional studics in these oncl sinillat SjsteAins wi [ E be made,

200

800

a l l ) j s ~ o t das "KT-IJF, ~ c r m p l c xcompound." No t e t r o i o l c n t urmiiiiil (ot!i-- ~!-isn p ~ s s i t ; b = trocc

are quit= surprising, since pct:cgrnphic

600 800

'he

100

600

106

780

600

44.5

175

9 1.8

1.0

0.88

0.15

Found

9.88

1.0

780

44.5

1.0

0.25

Expected

200

43.5

100

850

43.8

128

800

zro

100

800

53.6

4.5

4-5

800

53.6

4.5

800

53.c

4.5

ratios.

LiF

53 I?

47.0

45.9

NoF

Q f u3* i o

Total

0.25 0.15 1.0

0.89

0.14 0.17 0.19

0.12 0.96

1 .o

0.9 1

1.0

0.48

1.0

0.4 1

1.0

0.76

0.9 :8

0.30

0 0

1 .o

Q,5

0.9 78

0.44 0.59 0.33

0.42

Fivoridc ProTczsing Fuciiity

J, F, Eorgu" ly I, Truijt als C h c n i s t r y DivislQn The alkrjl;-mztal fluoridc prsc-cssing f a c i l i t y completed Mc;y 15, 1954 i s used shief!y for thF development of w i t a b l c ~ T D C C S + B for ~ the productican o f alkali-mctr! fli.roride a i x t v r e z containing UF,. There h a w h e m several prepnsitions, howevzr, od specid batches o f othor rarnpssitinnb t o f u l f i l l spccific n e ~ d x . S n g l c batches of appiaximately

PERIOD ENDING SEPTEMBER 70, 1954

2.5 k g can be prepared i n t h i s equipment. During the quarter, 27 preparations that yielded about 70 k g o f material were made. Repeated attempts to prepare N a F - K F - L i F mixtures containing 14 wt % UF, and 1 wt % UF, b y reduction o f UF, with uranium metal yielded mixtures containing 5 to 7 wt % UF, by chemical analysis. These results, along with those reported elsewhere in t h i s document, seem t o show that i n systems containing KF, reduction of UF, by uranium metal i s markedly incomplete. Pending clarification of t h i s point, production of material containing 5 wt % UF, and 10 wt % UF, i s under way, Production F a c i l i t y

J. P.

Blakely F. L. Daley Materials Chemistry D i v i s i o n

During the past quarter, 1760 k g of processed fluorides was produced in the 250.lb f a c i l i t y for external and internal distribution. A breakdown of the production according to composition i s given below: Amount Processed

(ks) NoF-ZrF4-UF4 (50-46-4 mole %) NaF-ZrF, (50-50 mole 76) NoF-ZrF4-UF4 (50-43.56.5 mole %)

the gas i n l e t tubes plugged quite frequently and

t h e HF concentration i n the e x i t hydrogen f e l l much below the value obtained a t lower f l o w rates. The over-all purification time was not appreciably shortened by the fourfold increase in hydrogen

passed. Sublimation of crude ZrF, a t a temperature lower than that previously used b y Y-12 personnel did n o t appreciably improve the purity o f the ZrF, product. Accordingly, the NaF-ZrF, melts produced iri

the

near

future

will

be prepared from the

(NaF)x*ZrF, that i s available i n moderately pure form from a commercial source; hafnium-free Z r F 4 w i l l be used t o adjust the composition as required, since t h i s material i s available in a high state of purity, E l e c t r o l y t i c purification of N a F - Z r F 4 mixtures has been shown to be complete, and the process requires much less time than does the hydrogen stripping process. Accordingly, one o f the u n i t s i n the production f a c i l i t y is being modified s l i g h t l y to test t h i s method on Q large scale. It i s anticipated that the use o f the purer taw materials and the electrolytic purification process w i l l considerably decrease the c o s t o f fuel preparation.

79 1

In-Pile Loop Loading

565 404

J. E. Eorgan Materials Chemistry Division

& Whitney Aircraft D i v i s i o n received 27 k g of NaF-ZrF, (50-50mole %) and 113 k g o f NaFPratt

ZrF,-UF, (50-46-4 mole 96). B a t t e l l e Memorial (50-46-4 I n s t i t u t e received 46 k g o f NaF-ZrF,-UF, mole %) and 68 k g o f NaF-ZrF, (50-50mole %)The remaining processed material was distributed t o various requesters i n the ANP Program. The d i f f i c u l t i e s associated with the long stripping times and large hydrogen volumes required for pra-

duction o f rigorously pure fluoride melts from the r a w materials available have not yet been overcome.32 Attempts to shorten the time required by increasing the hydrogen f l o w rate from 4 t o 15 liters/min and thus decrease labor und maintenance c o s t s were not successful. Under these conditions ._.__i_______.l-

32F, F. Blankenship and G. J. N e s s l e , ANP Qum. Prr,g. Rep. June IO, 1954, ORNL-1729, p 61.

T h e f i r s t in-pile loop was loaded on June

11 w i t h

(62.5- 12.5-25.0 a fuel concentrate, NaF-ZrF,-UF, mole %), prepared from enriched uranium by Y-12 personnel. The loading apparatus and confrols performed satisfactorily, and the transfer o f material to the loop went smoothly. However, the electrical contact which should have indicated when the loop was f i l l e d to operating level was

n o t activated even after 120% of the calculated charge had been added. Subsequent examination o f the loop showed that a weld had given way where the loop was connected t o the pump bowl, and i t seems certain that t h i s leak was present during the f i l l i n g operation. The U235i s being salvaged from the loop. Examination o f the f i l l i n g apparatus revealed that a l l but 68 g o f a total of 5837 g o f the material was transferred t o the loop.

ANP QUARTERLY P R O G R E S S REPORT C H E M l S T R Y O F Ab. K A L I - h i % T A L H Y D R B X I D E S

F. Kertesz Materials Chemistry D i v i s i o n

L. G. Overholser

Pur ifie at i QW

E. E.

Ketchen Materials Chemistry Division Eighteen batches of NaOH were purified by f i l t e r ing the 50 wt '% aqueous solution through a f i n e sintered-glass f i l t e r to remove the Na,CO, and by dehydration at 4QQOCunder vacuum. Approximately one-half the runs yielded products containing less

I and 0.1 wt 76 Na,CO,. thcrn 0.1 wt %H20 remaining runs yielded less than 0.1 wt % H,O,

The but

the values for Na2C0, ranged fromO. 11 to 0.2 wt %. Four 0.5-lb portions of potassium were reacted with water, and the resulting solutions o f KQH were

dehydrated a t 400'C under vacuum. The resulting 0.1 material contained less than 0.1 wt '% K,CO,, wt %

H,O,

and approximately 0,l wt % Na,

Reaction of Sodium Hydroxide with Metals

H. J.

Buttrum

F. A.

Knox

F. Kertesz Materials Chemistry D i v i s i o n

Work has been continued on the reaction o f copper Data were previously r e and sodium hydroxide. ported33 on copper concentrations found in NoOH A recent deteimination at 900째C up t o 8OOOC. shows 2500 pprii Cu i n the melt after exposure. T h i s value i s higher than would be expected from the previous datu taken from 600 to 8OOOC. A p l o t

of lag Cu concentration vs 1/T y i e l d s a straight l i n e from 600 to 8OOOC and thus suggests that true equi librium values were obtained. Extension of t h i s line would give a value of about 1500 ppm rather than the 2500 ppm Cu obtained. 33F. A. Knox, H, J. Buttram, F. Kertesz, A N P Quar. P r o g . Rep. J u n e 10, 1954, ORNL-1729, p 60.

Reactions of copper and n i c k e l w i t h sodium hydroxide were previously investigated by determining the equi librium hydrogen pressures developed. Monel, an alloy composed primarily of copper and nickel, has now been investigated by t h i s method, The Monel reaction tube containing the melt was surrounded by a quartz envelope, to which was attached a manometer far determination o f the hydrogen pressure. Establishment o f equilibrium required long periods of time, and the values obtained were not completely reproducible, Therefore, o n l y a range of equilibrium pressures i s presented i n the following for the sodium hydroxide-Monel reaction: Temperature

600 700

800 9 00 1000

("C)

P r e s s u r e (mrn t-lg)

35 to 112 to 160 to 232 to 280 to

42 130 193 280 360

In another series o f tests nickel oxide (reagentgrade commercial material heuted to 7OOOC in vacuum) w a s added to sodium hydroxide in a jacketed n i c k e l tube, At 350OC a pressure o f 22 mm Hg was found. Mass spectrographic analysis of the gas revealed 5.6% hydrogen and the remainder to be essentially water vapor. The temperature of the t e s t was increased t o 500OC and then, i n l 0 O T increments, to 1000째C. The pressure i n no case exceeded 31 mrn Hg, which i s approximately the saturated ambient water-vapor pressure. Analyses of the gas showed it to be 97 to 98% water vapor, w i t h a l i t t l e hydrogen, Under otherwise identical t e s t conditions, no water vapor was found when

n i c k e l oxide was omitted. The net effect of the n i c k e l oxide i s apparently that of oxidizing the hydrogen formed from the Ni-NuOH reaction to water vapor.

PERlOD ENDlNG SEPTEMBER IO, 1954

6 . CORROSION RESEARCH G. M.

W. D. Manly

Adamson

Metal lurgy D i v i s i o n

W. R. Grimes F. Kertesr Materia I s Chemistry D i v i s ion The s t a t i c and seesaw corrosion testing f a c i i i i i e s

were used for further studies of brazing alloys, special S t e l l i t e heats, Hostel loy R, Inconel, graphife, and various ceramics in sodium, fluoride fue! mixtures, and other mediums. For the fabrication

of many reactor components it may be well t o have a brazing a l l o y that i s compatible both w i t h sodium and w i t h the fluoride fuels. Therefore several corrosion tests were performed on Inconel joints brazed From this w i t h high-temperature btazing allays. work it was found that the brazing a l l o y 67% Ni-13% G e - l l % Cr--6% Si-2% Fe-1% Mn has goad corrosion resistance in fluoride fuels and fair corrosion resistance in sodium. T w o special heats of S t e l l i t e were tested that had a lower carbon content than those tested previously, and it was found that the lower carbon content increased the corrosion resistance i n the fluoride fuels. High-density graphite was tested in sodium and i n fluoride mixtures; It has very poor resistance to sodium attack but seems t o have fair resistance t o the fluoride fuels. Various hard-facing materials, ceramics and cermets, have been investigated froin the corrosion standpoint for use as bearings, seals, MgA1,0,, ZrO,, and and valve seats; AI,O,

Sic

were tested in sodium, a fluoride fuel, lithium, and showed the best corrosion lead. I n these tests,

Sic

resistance. In the thermal-convection loop corrosion studies, emphasis has shifted to fuels w i t h the uranium present as UF, and t o loops of Hasteltoy B. Hotleg attack i s not found i n lnconel loops i n which ZrF,-base fluoride mixtures w i t h the uranium as

UF, are circulated. A deposit is, however, found on the hot-leg surface. Only preliminary information is available, but it appears that neither attack nor a hot-leg layer i s found w i t h alkali-metal-base fluor i d e mixtures containing UF,. Mixtures of UF, and UF, w i l l result in a reduction in attack from that but some attack i s present found w i t h only UF,, and in high-uranium-content systems it may be s ign if icant. The addition of zirconium hydride to the ZrF,base fuel mixtures containing UF, was found to

F, and in a reduction in corrosion. However, t o obtain a complete absence of corrosion, sufficient hydride must be added to cause a loss of wanium. Several Hastelloy i3 loops have now been SUCc e s s f u l l y operated i n both the as-received ond the over-aged conditions. In both cases, a considerable increase i n hardness occurs during operation. With very I i t t l e ZrF,-bose mixtures containing UF,, attack i s found, even after Io00 hr. The mass transfer characteristics OC type 316 stainless steel in molten lithium were studied by using thermal-convection loops. After 1000 hr of loop operation, there was no sign of plug formation, and only a small amount of mass transfer was found i n one ot the loops. The investigation of the carrosion and mass transfer characteristics of materials in contact w i t h l i q u i d lead has shown that certain a l l o y s possess a much greater resistance t o mass transfer than the pure components of those alloys. The probable reason for t h i s resistance t o mass transfer can be related to the formation of intermetallic compounds i n these alloys. T h i s has been proved i n the case of alloys; of 45% Cr-55% Co, Ni-Mo alloys, and t h e Fe-Cr-base stainless steels.

result i n the formation of

The flammability of l i q u i d sodium a l l o y s has been studied, and it has been found that only bismuth and mercury have an effect on the fiammabi I i t y of sodium. It was found that the degree of reaction of No-Bi and No-Hg a l l o y s w i t h air was not s i g n i f i cantly changed when the pressure was varied froin

0.25 t o 1 atm. An investigation has been staried t o determine the amount of hydrogen I iberated when NaOH i s heated t o 900째C i n an inert environment, The container problem has been solved by using a large single crystal of magnesium oxide. A thermodynamic study of the alkali-metal hydroxides has shown that hydrogen i s important in s t a b i l i z i n g NaQH a t temperatures above 600째C. T h i s i s the temperature range a t which mass transfer becomes significant in Ni-NaOH systems. These computations link the formation of hydrogen w i t h the format i o n of unsaturated oxygen ions, which may be necessary for mass transfer to occur.

ANP QUARTERLY PROGRESS REPORT A plausible explanation has been developed for the evidence of a maximum in the corrosion of In-

S T A T I C AND SEESAW CORROSION TESTS

E. E. Hoffman C. R. Brooks

conel by molten fluorides i n s t a t i c tests at 800 t o 900째C and for both void formation and chromium

concentration in the melt being less at 1000째C than at 800째C. Static tests of type 316 stainless steel i n NaZrF,

and i n NaF-ZrF,-UF,

substantiated the

phenomena observed w i t h Inconel. Additional corrosion experiments w i t h fuel mixtures containing simulated f i s s i o n products have shown that the experimental procedure being. fol-

lowed i s entirely unsatisfactory. A controlledv e l o c i t y corrosion testing apparatus has been constructed for the rapid transfer of molten fluoride niixtures through both heated and cooled test sec-

This apparatus was designed to simulate tions. conditions in reactor components. TABLE 6.1.

W. H. Cook C. F. L e i t t e n

Metallurgy

B i v i s ion

Brazing Allays in NaF-ZrF,-UF,

and in Sodium

Specimens of the three brazing OIIOYS described i n Table 6,1 were tested i n NaF-ZrF,-UF, and i n sodium. The specimens were cut from 5-in. T-joints of Inconel which had been brazed in a dry hydrogen atmosphere. The purpose of these tests was t o find brazing cslloy w i t h good corrosion resistance t o both NaF-ZrF,-UF, and sodium, since such ci brazing a l l o y could be used t o back up tube-toheader welds. In use, the brazing a l l o y would be exposed to the fluoride mixture, but resistance t o attack b y sodium would also be required if a fracture occurred i n a tube-to-header weld.

RESULTS OF STATIC TESTS O F BWAZRNG ALLOYS IN NaS-ZrF4-UF4 (53.5-40.6.5 mole %) AND IN SODIUM A? 150OoF FOR 100 hr

Brazing A l l o y Composition

Allay A 67 Ni, 13 Ge, 11 Cr, 6 Si, 2 Fe, 1 Mn on lnconel (brazed 1120째C for 10 min)

Both

We i g ht Change

NaF-ZrF4-UF4

+0.0003

+0.024

Metallographic N o t e s

Very l i t t l e attack; slight surface roughening t o a depth of 0.25 mil

Sodium

-0.0003

- 0.0 17

Attack in the form of subsurface voids t o a maximum depth of 3 to

Alloy 8 65 Ni, 25 Ge, 10 Cr on lncanel

NoF- Zr F4-UF4

-0.0010

-0.077

(brazed a t 1 14OoC for 10 mi")

mils

N o attack; s l i g h t u n i f o r m solution would occount for weight loss

Sodium

-0.0013

-0.1 16

Braze mum

Alloy C 72 Ni, 18 Ge, 7 Cr, 3 p

t o a maxi-

depth of 8 m i l s

Na F- ZrF4-UFq

-0.0005

-5.041

Braze f i l l e t attacked to a depth of 2 t o 3 m i l s in the form of subsurface voids

Sodium

-0.0008

-0.058

Braze f i l l e t ottocked t o a depth of 3 t o 6 m i l s in the form of subsurface voids

lnconel

f i l l e t attacked

PERIOD ENDING SEPTEMBER 10, 7954 Three brazed lnconel specimens were exposed for

100

hr at

1500째F

in \-in.-OD#

0.035-in.-wall

lnconel container tubes to static NaF-ZrF,-UF,

(53.5-40-6.5 mole %), and three were exposed to static sodium. The brazed specimens were cleaned carefully before and after testing so that v a l i d weight-change data could be obtained. The specimens were n i c k e l plated after exposure t o prevent rounding of the edges during metallographic polishing. Presented in Table 6.1 are the data obtained by metallographic examination of the specimens. A l l o y s A and B showed very good resistance to attack by the fluoride mixture, as shown i n Figs. 6.1 and 6.2. A l l o y A i s also shown i n Figs. 6.3 and 6.4 at low magnification in the as-brazed and as-tested conditions. Alloy C was attacked s l i g h t l y by the fluoride mixture. A l l the alloys were attacked

s l i g h t l y by sodium, with alloy A showing the least attack. The maximum depth of attack of 3 to 4 m i l s

on alloy A is not serious, since sodium would be i n contact w i t h t h i s alloy only i n the event of a tube-to-header weld fracture. It may be concluded from the results of these i e s t s that further study is warranted on a l l o y s A and B; both showed good resistance t o attack by the fluoride mixture. Special emphasis w i l l be placed on a l l o y A, since it exhibited superior resistance to attack by sodium.

Special S t e l l i t e Heats i n

bdoF-ZrF,-UF, and

in Sodium

Two special heats of Stellite, heats B and C, were requested from the Union Carbide and Carbon Metals Research Laboratories for corrosion testing. The specimens were sand cast and had Rockwell

C hardnesses as follows: heat B, 39 t o 42; heat C, 42 t o 43. T h e as-received microstructure consisted of Cr,C, carbide particles i n a cobalt-rich matrix. The approximate compositions of these heats were

Ni PLATE

0.001 _ I _

0 0

0.002

3.003

ANP QUARTERLY PROGRESS REPORT

6Hi-Ge-G) Fillet Surface After Exposure far 100 kr a t 152) Fig. 6.2. Brazing Allay UF, (53,5-48-5,5 K I Q ! !%~)a Etched with oxalic acid,

Fig. 6.3. Brazing A!lay A

OR Incane!

in NeF-ZrE,-

T-Joint in As-Brazed Condition. Etched with oxalic acid.

PERIOD ENDING SEPTEMBER IO, 7954

0.0i

0.02 -

0.03

Fig. 6.4.

Brazing Alloy A on Inconel T-Joint After Testing.

Dark areas are not areas of attack.

Etched with oxalic acid. as follows:

specimen surface, being deepest when the network Composition ( w t %)

Heot B Carbon Chromium Tungsten

1 35

S i I icon Nickel

Cobalt

0.30

Molybdenum Manganese

Heot C

0.50 0.50 58

12 0.50 0.50 3 55.7

The specimens were tested i n static mediums for 100 hr a t 1500째F. Inconel container tubes were used and duplicate tests were run. The test results are presented in Table 6.2. Both Stellite heats were heavily attacked along (50-46-4 mole the carbide network in NaF-ZrF,-UF, %), as shown i n Figs, 6.5 and 6.6. The attack varied from a minimum of 7 m i l s to a maximum of 20 mils. T h e depth of attack varied w i t h the orientation o f i h e carbide network w i t h relation to the

was perpendicular to the surface. None of the specimens were attacked i n sodium; however, in heat B there wgre no large equiaxed carbide particles i n a 1- t o 2-mil area adjacent to the surface. The carbide particles i n this area appeared to have transformed to a very fine structure. There was also a large amount of fine precipitate i n the cobalt-rich matrix near the surface, Heat C had s l i g h t l y better resistance than heat B to attack by sodium, and both heats had very poor resistance to attack by the fluoride mixture. Several

of the specimens had shrinkage voids near the surface, one that was 7 m i l s i n diameter, and the attack was deeper in these areas. Hastelloy

i n Various

Media

Hastelloy R centerless ground and annealed rods, in. in diameter and 8 in. i n length, were obtained from the Haynes Stellite Company. Each rod was machined into a specimen 1 x 1 x 3 in.), a container, and a plug. The specimen and the medium to which i f was t o be exposed were loaded into the

(6 4

ANP QUARTERLY PROGRESS R E K I R T TABLE 6.2.

Weight Change ( d i n . 2)

Metallographic Notes

Heat

As received

No F-Pr F4-U F4

.............

Heot

Sodium

___. .........

..........

Bath

RESULTS OF TTATHC TESTS OF SPECiAL STELLITE HEATS BN S AND I H N a F - Z r F 4 - U F 4 (50-46-4 m o l e '7) AT 1500°F FO

Heat B

Heat C

Structure consisted of large, f a i r l y equiaxed p a r t i c l e s o f complex carbide (Cr7Cg) surrounded by a e u t e c t i c structure i n a coboltr i c h matrix

Carbide p a r t i c l e s w h i c h surrounded the dendrites of the cobalt-rich s o l i d s o l u t i o n hod a very fine structure; o n l y a few o f t h e equiaxed carbide p a r t i c l e s seen throughout Heat B c o u l d be found

+0.008

+0.002

No attack; however, no globular

No attack; very l i t t l e difference i n

+0.016

+0.003

S a m e a s above

Same o s above

-0.05 2

-0.045

Carbide network attacked to a depth of 7 to 20 mils, depending on orientation of the network

-0.060

A t t a c k on carbide phose varied from 8 to 19 mils; minimum att a c k of 8 m i l s occurred on specimen where dendrites of cobaltr i c h s o l i d s o l u t i o n surrounded by carbide p o r t i c l e s were oriented p o r a l l e l t o the specimen surfoce; maximum a t t a c k of 19 m i l s occurred where dendrites were oriented perpendicular to ?he specimen surface; large amount o f fine p r e c i p i t a t i o n in matrix during t e s t due to aging

-0.057

Same a s above

S a m e as above

p a r t i c l e s nf the carbide phase i n a 1- t o 2-mil area adjacent PO the surface; carbide p o r t i c l e s in t h i s area completely transformed t o n very f i n e precipitate; precipitnt i o n also occurred i n matrix, esp e c i a l l y near the, exposed surface

container, and t h e plug was welded in place in a dry box under a purified helium atmosphere. The

nominal composition of Hastelloy R i s 45% Ni, 14% Cr, 10% Fe, 6% Mol 3% Ti, 2% AI. The specimens were exposed t o the various media for 100 hr a t 1500°F. The test results are presented i n Table

6.3,

The specimens showed good resistance to attack by NuF-ZrF,-UF, (50-44-4mole %), Fig. 4.7, and b y lithium. Surprisingly, t h i s alloy had better resistance t o attack by lithium than by sodium; the attack by lithium on high-nickel-content alloys (for example, Inconel) a t 1500°F i s usuully heavy and i s greater than the attack by sodium. The sample tested in sodium was attacked slightly, while the specimen tested i n lead was heavily attacked. The specimen tested i n sodium hydroxide was attacked throughout i t s entire thickness.

as-received and t e s t e d structures

.........

Inconel i n Molten Rubidium Tests were candiicted on rubidium metal i n connection w i t h an ORSORT project. Rubidium has a

low melting point, 92"F, and a low boiling point,

129O0F, and therefore i s being considered as a possible reactor coolant, The rubidium metal used i n the experiments was prepared by calcium reduction of the rubidium fluoride s a l t by the Stable Isotope Research and Production Division, An i n i t i a l attempt t o determine d i e vapor pressure vs temperature relationship up t o 1650°C was msuccessful. However, the test equipment has been redesigned and another attempt w i l l be made in the near future, lneonel has been tested i n contact w i t h static rubidium for 100 hr at 1650°F. The attack varied from 0.5 mil i n the vapor zone to a maximum of 1.5 m i l s at the liquid-gas interface, as shown i n Fig. 6.8. The rubidium used was not dist i l l e d and i s known to have contained several per

PERIOD ENDING SEPT€MB€R 70, 1954

f i g . 6.6. Special Stellite H e a t C Before (a) and A f t e r (li) E X ~ Q S Uto W Static NaF-ZrF,-UF, (53.540-6.5 m o l e %) for 100 hr a t 1500°F. The area where the attack terminated i s shown in ( b ) . Etched with

KOH-K3Fe(CN) 6 v

ANP QUARTERLY PROGRESS REPORT TABLE 6.3, RESULTS OF STATIC TESTS OF HASFELLOY R BN VARIOUS MEDlA AT 15QQaFFOR lQ0 hr

Bath

N a F- Zr F4- UF (50-46-4mo I e

Weight

Change

(%I

(g/ i n. 2,

- 0.05

-0.003 7

Metal I ographi c Notes

Specimen unattacked except in one small area where voids extended to a depth of 1 m i l ; no a t t a c k on vapor zone of tube

t7.9

NaOH

+0.608a

Specimen !'-in>

square attacked throughout e n t i r e

thickness

-0.06

Sodium

-0.0021

Specimen and bath zone attacked to a depth of

0.25 to 0.5 m i l i n the f o r m of voids; exposed surface of bath zone of container p a r t l y decarburized t o a depth of

-0.07

Lithium

-0.0047

l mil

Small voids to a depth of 0.25 m i l i n specimen and bath zone of tube; surface p a r t l y decarburized by lithium Specimen h e a v i l y attacked t o a maximum depth of

Lead

3 m i l s (average 2 mils); bath zone a t t a c k e d uniformly t o a depth of 1.5 mils; phase change proba b l y occurred i n attacked areas; vapor zone unattacked

.........

. 0.004 X

0 0

0.002

0.003

Fig. 6.7.

Ha5telloy

w i t h aqua regia.

Exposed to Static NaF-ZrF,UF,

(50-46-4 mole %) for 100 Rr

1500OF. E t c h e d

PERIOD ENDING SEPTEMBER IO, 1954

0,004 I _

0.002 I

0.003 X

3.004

lnconel E x p o s e d t o Static Rubidium for '100 hr o t 1650OF. Note decarburization i n attacked Specimen n i c k e l plated after test t o protect edge. Etched with glyceria regia.

Fig. 6.8.

area.

1500째F i s approximately 0.1%, and therefore the carburization of the nickel Containers in these

cent sodium and some oxygen contamination. Additional static tests are under way w i t h t r i p l e -

d i s t i l l e d rubidium, and an Inconel thermal convection loop i s being operated w i t h b o i l i n g rubidium, Fig. 6.9. T h i s loop has now operated for several hundred hours without difficulty.

tests i s very slight. The A-nickel used for the containers was found by analysis t o contain only 0.05% carbon.

Carburization of Inconel by Sodium

It i s w e l l known that sodium, i n addition t o decarburizing metals, can, i n some cases, carburize them if the carbon concentration in the sodium i s sufficiently high. Therefore, an attempt i s being

made t o determine whether small additions of carbon would prevent decarburization of lnconel specimens

during long-time creep t e s t s in contact w i t h sodium at elevated temperatures. A-nickel containers are being used for static tests so that the r a t i o of lnconel surface area t o sodium volume! w i l l be small. The maximum s o l u b i l i t y of carbon in n i c k e l

The r a t i o of the lnconel surface to the sodium volume i n the tests performed to date was 0.76. The lnconel specimens used were 0.049-in. sheet

reduced t o 0.015 in. by cold rolling, and they were annealed for 2 hr a t 1650째F. The carbon additions (1, 5, and 10 wt %) were made to the sodium in the form of small lumps of reactor-grade graphite.

The nickel containers were loaded with the lnconet specimens, the graphite, and the sodium i n a dry box i n o purified helium atmosphere and sealed As shown i n Table 6.4 and Fig. under vacuum,

6.10, a l l the specimens were very heavily carburized and extremely b r i t t l e after exposure for 100 hr at 1500'F. Also, they were partly covered w i t h a 89

ANP QUARTERLY PROGRESS R E P O R i

Fig. 6.9.

Inconel Thermal-convection Loop far Circulating Bviling Rwbidium.

green surface f i l m that was identified by x-ray T h e oxygen source for formaanalysis t o be Cr,O,. t i o n of t h i s f i l m was, a t first, thought t o be the graphite which had not been degassed prior to the tests; however, a Cr,C, f i l m was later found i n other tests w i t h degassed graphite and standard tests w i t h no graphite addition. Therefore preparations are being made for obtaining oxygen-free sodium for use i n these tests.

'E. E. Hoffman et a [ . , A N P 1954, ORNL-1729, p 70.

Quai. Prog. Rep. J u n p 10,

Special Tar-Impregnated and Fired Graphite

A special tar-impregnated and fired graphite, known commercially as Graph-i-tite, has been tested for corrosion resistance t o sodium and NaF-ZrF,UF, (53.5-40-6.5mole %). Also, a comparison was made o f the carburization of an austenitic stainless steel by contact w i t h Graph-i-tite and w i t h C-18 graphite (reactor grade). The Graph-i-tite was fabricated, as described previously,' by tar-impregnoting and firing graphite 16 times. The repeated tar impregnations and firings produce a high density

PERIOD ENDING SEPTEMBER IO, 1954

UNCLASSIFIED

Fig. 6.10.

Carburization of Incornel Exposed t o Sodium with Carbon Additions for

tested in sodium plus regia. 250X.

100 hr a t 150OOF.

(6) Specimen tested i n sodium p l u s 1 wt % carbon. ( c ) Specimen wt % carbon. Specimens n i c k e l plated to protect edges. Etched w i t h glyceria

(a) inconel specimen before test.

and a “tough skin,’’ which, it was hoped, would reduce penetration of various l i q u i d mediums. An open Graph-i-tite crucible containing sodium and a sealed Graph-i-tite crucible containing NaF-

ZrF,-UF, were tested at 1500°F for 100 hr i n vacuum. T h e sodium completely penetrated the 0.38-in.-thick wall of the crucible, as i n the previous test,’ and caused it t o crack and crumble.

The only macroscopically v i s i b l e sign of attack on the crucible that contained the fluoride mixture was that the inner surface had a higher gloss (Fig. 6.11) after exposure, Metallographic examination of sections of the crucible indicated irregular pene-

tration from 0 t o 5 m i l s deep. No attempt has been made to determine whether reaction products accompanied the penetration.

T h i s attack may be

ANP QUARTERLY PROGRESS REPORT

TABLE 6.4. CARFSURIZATION OF INCONEL BY EXPOSURE T O SODIUM CONTAINING CARBON ADDITIONS FOR 100 hr AT lMO0F Carbon content of lnconel specirnen before test, Carbon Added

Carbon Content*

t o Sodium

of Specimen

Before T e s t

Afrer T e s t

(wt

76)

( w t %)

..........

0.056 w t %

Specimen Weight Change**

(%1 ....

0.882

1.09

1.34

1.24

1.63

*Average of three analyses. **On

Y I 2416

a l l specimens part o f w e i g h t change was due to

formation o f G203f i l m on surface.

INCH

V f '[DENT'IAL Y 12453

compared w i t h the 0 t o 2 m i l s attack by NaF-ZrF,UF, (50-46-4 mole %) i n the previous tests.' The carbon content of the fluoride mixture did not change significantly i n either test.

A comparison of the carburization of nustenitic stuinless steel by tar-impregnated and fired graphite and by C-18 graphite (reactor grade) was obtained

through static tests of cylinders of the same size of each type of graphite in type 304 stainless steel tubes containing equal quantities of sodium i n vacuum. Tests for 100. and SOO-hr periods at

1500째C were made on each type of graphite. Type 3114 stainless steel was chosen because it carburizes more readily than other austenitic stainless

steels. T h e results of the tests are shown in Figs. 6.12 and 6.13. In the 100-hr test, the tar-impregnated and fired graphite curburized the type 304 stainless steel t o a s l i g h t l y greater extent than the C-18 graphite did, and i n the 500-hr test, it clearly carAfter the burized the steel t o a greater extent. 500-hr tests of both types of graphite, steel i n rontact w i t h the sodium showed small subsurface voids; the voids were pronounced in the steel tested w i t h the tar-impregnated and fired graphite. The depth of Carburization of the type 304 stainless steel above the level of the molten sodium covering

the graphite cylinders was 1 m i l in a l l the tests, except the 500-hr test w i t h u tar-impregnated and fired graphite cylinder i n which there was carbur i z a t i o n of the steel to a depth of 2 mils.

F i g . 6.11. Graph-i-tite Crucible Before ( u ) and Aftes ( 6 ) Exposure to HoF-ErF4-UF, (53.5-40-6.5 m o l e %) for 108 hr a t 150t)OF. At the end o f the t e s t the crucible w a s inverted so that the fluoride mixture would flow t o the top. The frozen fluoride mixture can be seen above the shiny inner surface of the crucible,

PERlOD ENDING SEPTEMBER JO, 1954

Carburization of T y p e 304 Stainless Steel Exposed t o Graph-i-tite (a) and to C-18 Graphite in Sodium for 100 hr a t lSOO째F in Vacuum. (a) Depth of carburization was 118 m i l s (concentrated) t o 32 m i l s (traceable). ( b ) Depth of carburization was 16 m i l s (concentrated) t o 32 mils (traceable). Etched w i t h glyceria regia.

F i g . 6.12.

(b)

Fig. 6.13, Carburization of Type 304 Stainless Steel Exposed t o Graph-i-tite (a) and t o C-118 Graphite ( b ) i n Sodium for 500 hr a t 1500OF i n Vacuum. ( a ) Depth of carburization was 40 m i l s (concentrated) t o 49+ m i l s (traceable through the tube wall). ( 6 ) Depth of carburization w a s 33 m i l s (concentrated)

49+ m i l s (traceable through the tube wall). Etched w i t h gllyceria regia.

ANP QUARTERLY PROGRESS REPORT sults of the tests are given i n Table

Ceramics i n Various Mediums Static corrosion screening tests of three ceramic oxides and one carbide stabilized ZrO,, atmosphere at

and

(A1203, MgAI,O,,

sic) were

l5OO0F for 100

calcium-

mode i n a vacuum hr in each of four

med i urns: sod ium, NaF-ZrF,-UF, (53.5-40-6.5molk A fourth ceramic oxide was tested under the same conditions i n sodium and i n

%), lithium, and lead, NaF-ZrF,-UF,

(53.5-40-6.5 mole %).

The

A120,

and MgAI,O, test pieces were cut froin single, synthesized crystals; the Mg0 test pieces were cleaved from a single, synthesized crystal. The nominal dimensions of the test pieces and

the re-

Thickness

Weight

Change

(a)

- 0.5

A1203n

Powder

x-ray examinatians were made of the specimens that were the least corrosion resistant i n each medium, and the results are summarized i n Table

6.6.

The

Sic

specimens were,

in general, the most

resistant to corrosion, and t h e AI,O, specimens These single-run were the next most resistant.

screening tests are qualitative, and t h e results should be used w i t h caution. Larger specimens are to be tested so that more accurate data can be

obtained and metallographic examinations chemical analyses can be made.

TABLE 6.5. SUMMARY OF STATIC TESTS OF CERAMICS IN VARIOUS ~ Material Tested

6.5.

~ A T 1500째F ~ FOR~ 100 hr I

Re marks

(% T e s t e d in Sodium

-0.9

Specimen changed from a colorless transparent state t o a white semitransparent state; a small crack appeared an a n edge

-0.9

Sicb

Specimen e s s e n t i a l l y unaltered; brighter and cleaner; s m a l l part of d i s k edge broken o f f

0.0

M~AI~O,"

O r i g i n a l l y colorless transparent specimen almost comp l e t e l y changed t o a black state; chipped s l i g h t l y on one edge

Zr02

(GO'

stabilized)

MgOC

+8.6 -0.5

Specimen changed from l i g h t

buff t o blue-black; edge

chipped

-0.1

Specimen unaltered except for color change from colorl e s s to l i g h t blue-gray

T e s t e d in

A1203

-3 1

NaF-ZrF4-UF4 (53.5-40-6.5 m o l e W ) Specimen broken and covered w i t h fluoride mixture; brown-red interface present between the fluoride mixture and the specimen

- 1.8

Specimen broken; recovered pieces appeared to be unattacked Recovered p o r t i o n of specimen (more than one-half) covered w i t h f l u o r i d e mixture and apparently comp l e t e l y altered

Z r 0 2 (CaO stabilized)

No v i s i b l e trace of specimen found

Mg 0

No v i s i b l e trace of specimen found Test.ed in L i t h i u m

A1203

and

No v i s i b l e trace of specimen found

Pâ&#x201A;ŹR/OD ENDlNG SEPTEMBER IO, 1954 T A 3 L E 6.5.

Thickness Change

Material Tested

(XI

Sic

(continued)

- . . -

__ -

Weight

Change

Remarks

(XI ... --

Some small p a r t i c l e s found

MgA1204

No visible trace of specimen found

t r 0 2 (COO stabilized)

Specimen in many s m o l l pieces; buff color changed t o charcoal- bla ck Tested i n Lead

0.0

*'2'3

t0.3

No visible attack;

c o l o r l e s s transparent specimen

changed

found on specimen; specimen apparently covered w i t h thin lead film

t o gray; s m a l l s p o t

Sic

No visible attack;

t2.9

MgAI $34

0.0

ZrOZ ( G O stobil ized)

0.0

speciiilen

broken in loading

No visible attack; s p e c i m e n changed from c o l o r l e s s to gray; broken in loading + 0.5

No visible attack; specimen changed from a light buff t o gray

'Disk cut from a single crystal; specimen 0.75 in. in diameter and 0.021 in. thick. bDisk 0.75 in. in diameter and 0.021 in. thick. CCleaved from a single crystal; specimen 0.02 x 0.36 X 0.36 in. T A B L E 6.6. POWDER X-RAY I O E N T I F I C A T I O N OF T H E PHASES O F T H E L E A S T CORROSION-RESISTANT SPECIMENS

Powder X-Ray Data on Specimen

Test Medium

Specimen

~~~

Before Test

.....................

..... ~~

After Test Face-centered cubic, a

Face-centered cubic, a = 7.970

NaF-ZrF4-UFq pattern

Sodium

Primarily face-centered cubic Zr02; secondarily m o n o c l i n i c ZrQ2

Same a s before t e s t

Lithium

Primarily face-centered cubic ZrQ2;

Unidentified, fa re-centered cubi c, a = 4.674

NaF-ZrFd-UF4 (53.5-40-6.5 mote

Zr02 ( 6 0 stabilized)

Face-centered cubic, a = 7.970

Sodium

MsAy34

OF SOME

secondarily monoclinic Z r Q 2

= 7.981

I I _ _ -

F L U O R I D E CORROSION O F I N C O N E L tN T H E R M A L - C O N V E C T I O N LOOPS

G. M. Adamson Metal iurgy D i v i s i o n

Effect of

UF,

ZrF4=Base Fuels

Preliminary results of tneonel thermal-convection loop tests of the corrosive properties of UF3-bearing fluoride mixtures were presented previously2 ~ n d

compared with the corrosive properties of UF,bearing mixtures. Additional t e s t s with UF3-bearing ZrF4-base mixtures have confirmed the reduction or eliminotion of hot-leg attack, along w i t h the formation of Q hot-leg layer, T h e lack of attack and the hot-leg layer are illustrated i n Fig. 6.14.

A N P Q UAR I-EKIY PRO GR ESS R E PORT The p a i t i o n of a hot leg shown in Fig.

6-14 w a s

taken from an Inconel thermal-convection loop which ! UF, in 2.4 wt % had circulated UF, (2,l w t %

total U) i n NaF-ZrF, (53-47 innlc %) for 508 hr a t a hot-leg temperature of 1500째F. The fluoride mixture was drained from the loop at the operating temperature t o ascertain whether the layers previously noted had formed during cooling of the loopa Since the layer was again present, it i s now tonsidered t o have formed during operation. This conclusion was further strengthened when the metal lographic examination revealed e diffusion

zone between the layer and the base metal. An examination of the layer by ii microspark spectrographic: technique3 showed it t o be predominantly zirconium. Since t h i s technique i s not sensitive t o uraniuii~,the layer could also contoin uranium. T h w e i s some indication that some of the LJF, dissociates t o produce uranium metal, which reduces some o f the ZrF, t o produce zirconium metal. T h e accuracy of the analytical determination for zirconium i s not sufficient to reveal ;he reaction products postulated, since some zirconium i s a l s o l o s t by sublimation. In another lncowel loop in which UF, i n NaFZrF, wos circulated for 2000 hs at 1500째F, some voids were found t o a Cllcpth of 4 m i l s upon metallo7

'Examination made by C. Feidmon, C h e m i s t r y Division.

graphic exarninatian. T h e voids did not appear t o be the same Q S those normally found, and it i s thought that they were formed when b r i t t l e intermetallic compounds were p u l l e d out o f the base metal during polishing. No hot-leg attack could be found, as shown i n Fig. 6.15, but there was a layer

0.5 m i l t h i c k on the hot-leg surface. T h i s layer was also reported to be predominantly zirconium.

It has been definitely established that the use of UF, rather than U F , i n ZrF4-base mixtures w i l l lower the corrosive attack and mass transfer in lnconel systems. s-lO\Mevei, it has been found to be impossible to dissolve sufficient UF, i n ZrF4-base mixtures t o obtain fuels o f interest for high-temperature ienctms, T o obtain a fuel w i t h sufficient uranium,

it would be necessary t o use a mixture

Data obtained from loops opero f UF, and UF,. ated w i t h such mixtures and w i t h standard UF,-

bearing mixtures are presented in Table 6.7. The data show clearly that a mixture o f UF, and UF,

produces less attack than

UF, alone, but the mix-

ture does not eliminate the attack, as i s the case w i t h UF, alone.

Effect of UF, i n AJkali-Metal Base Fuels

It has been found that more UF, can be dissolved i n the alkal i-metal-base fluoride mixtures than i n the ZrF4-base mixtures, and therefore additional

tests were made w i t h NuF-KF-LiF (1 1.5-42.0-46.5 mole %) containing UF, UF,, and mixtures of the two.

P I .

. .

F i g . 8.15. Hot-Reg Surface ab Inconel Loop After Circulating UF, in HaF-ZrF4 for 2000 ha OF. Unetched. 250X. Reduced 36%.

PERIOD ENDING SEPTEMBER 10, 1954 T A B L E 6.7. E F F E C T O F MIXTURES O F UF3 AND UF4 I N NaF-ZrF4 ON CORROSION OF INCONEL THERMAL-CONVECTION LOOPS OPERATED A T A HOT-LEG TEMPERATURE OF 1SOOOF

-____..__

Loop

Total

No.

(wt

-~ 473

U as UF3 (wt %)

Operating Time

(hr)

Metallographic Notes H o t - L e g Appearance

Col d - L eg

Appearance

8.7

1.2

500

Moderate intergranular subsurface

Occasional metal c r y s t a l s

v o i d s t o a depth of 4 m i l s

491*

12.8

2.5

500

Moderate to heavy intergranular

D e p o s i t t o 0.3 m i l t h i c k

subsurface v o i d s to a depth o f

7 mils 492*

14.2

1.8

2000

Moderate t o heavy intergranular

Deposit t o

subsurface v o i d s to a depth of

1 mil t h i c k

15 m i l s 4 69

8.5

500

Heavy general a t t a c k and inter-

No deposit

gganular subsurface voids to a depth of 8 m i l s

462

13.9

500

Moderate to heavy intergranular

M e t a l l i c deposit t o 0.3 mil t h i c k

subsurface v o i d s to a depth of

10 m i l s

" T h e s e loops were f i l l e d from the same Latch of f l u o r i d e mixture, and the differences in uranium a n a l y s i s cannot yet be explained.

Two lnconel loops were operated for 500 hr w i t h NaF-KF-UF, t o which UF, and UF, had been added. Because of sampling and analytical difficulties, the relative amounts of UF, and UF, are not known, but it appears that about one-third of the uranium was present as UF,. One loop contained a total uranium content of 10.4 wt %, and after the test the hot leg showed light, widely scattered attack to a depth of 1 m i l (Fig. 6.16); the hot-leg surface was quite rough. The other loop contained 7.4 wt % uranium, and the hot leg showed a v o i d type of attack that varied from l i g h t to heavy, w i t h a maximum penetration of 2 mils; again, the surface was rough. In both loops, a layer was found i n the c o l d leg but not i n the hot leg. These mixtures appear to be superior t o the ZrF,-base mixtures w i t h UF, w i t h respect t o depth

of the subsurface void type of attack, and the hot-

leg deposit found w i t h the UF3-bearing ZrF,-base mixtures did not form. More work i s needed t o determine whether the rough surfaces represent a different type of attack and whether the cold-leg deposits formed are NaK,CrF,, which would interfere w i t h heat transfer.

Fig. 6.16. H o t - l e g Surface of lnconel Loop After Circulating a Mixture of UF, and UF, (4.3 w t % UF, and 6.9 w t % UF,) in NaF-KF-LiF (11.542.0-46.5 mole %) for 500 hr a t 1500째F. Etched w i t h modified aqua regia. 250X. Reduced 36%.

ANP QUARTERLY PROGRESS REPOR7

E f f e c t of Zirconium Hydride Additions t o

Fuel

series of loops fabricated from the annealed p i p e was f i l l e d from the same batch of NaF-ZrF,-UF, (50-46-4mole %) and operated for 500 hr a t 1500째F.

Various amounts o f zirconium hydride were added (50-46-4mole %) as a means of t o NaF-ZrF,-UF, The hydride was added reducing t h e UF, t o UF,. t o small portions of fluorides taken from the same original batch, and f i l t e r s were used when the smal I batches were transferred t o lnconel thermalconvection loops. The data from loops operated w i t h these batches of fluoride mixture are given in Table 4.8. Layers were found in the c o l d legs of a l l loops t o which the ZrH, additions had been made. The data show that to obtain sufficient reducing power by the addition of ZrH, to eliminate corrosion it may be impossible to prevent the loss of some uranium both i n the treatment pot and i n the loop.

Two loops were made from pipe annealed a t 2100째F that had a grain s i z e of 1 t o 1 \ gr/in., at 100 X,

w h i l e the loop fabricated from as-received lnconel pipe and the one fabricated from pipe annealed at 1600째F contained about 6 gr/in.2. Very l i t t l e dif-

ference in hot-leg attack was found i n these loops. Those w i t h the larger grains may have had s l i g h t l y deeper attack, but the attack was heavier and more general and the deep penetrations were concentrated into fewer boundaries. F L U O R l D E CORROSION O F H A S T E L L O Y 6 IN T H E R M A L - C O N V E C T I O N L O O P S

G. M. Adamson

Effect of Uranium Concentration

Meta II urgy D i v is ion

T w o loops were operated w i t h a high-purity NaF(53.5-40-63mole %) mixture, T h i s mixture i s comparable t o the one t o be used i n the ARE and has a higher uranium content than the mixture normally used in thermal-convection loop tests. The heavy hot-leg attack i n both loops was of the usual subsurface-void type w i t h a maximum T h i s i s s l i g h t l y deeper penetration of 10 mils. than the 6 t o 8 mils found with the lower uranium content mixtures. T h i n metal lic-appearing loyers were found in the cold legs o f both loops. The

Loops fabricated from both as-received and over-

ZrF,-UF,

aged Hastelloy B were operated satisfactorily. The operating mortality rate has been reduced from 90% t o 0% i n the last group of four loops. The increase in hardness during operation i s not so great in the loops constructed w i t h over-aged material as i n the loops constructed w i t h asreceived material, but, w i t h proper care, the loops of as-received material can be operated. Very l i t t l e attack was found in a loop which circulated NaF-ZrF,-UF, (50-46-4mole %) for

results obtained w i t h these loops confirm those found previously w i t h similar, but impure, mixtures.

1000

Effect of lnconel Grain Size lnconel pipe was annealed a t two temperatures t o

provide specimens w i t h different grain sizes.

hr a t

1500째F.

The attack appeared as a few

voids to a maximum depth of 1 mil, w i t h possibly some increase in surface roughness. Most of the surface roughness was present i n the as-received tubing, as shown in Fig. 6.17.

T A B L E 6.8. E F F E C T OF ZrH, ADDITIONS TO N a F - Z r F 4 - U F 4 (50-46-4 male %) CIRCULATED I N INCONEL THERMAL-CONVECTION LOOPS A T 1H)O"F FOR SO0 hr

Loop

NO.

ZrH2

(

469

Uranium Content

Hot-Leg Attack

Added .~~

Heavy general attack

Before T e s t ...............

and intergranular voids to

a depth of

~.~ ..........

8 mils

___

8.5

(X)

After T e s t

a. 8

459

0.2

Moderate to heavy attock to a depth of

6 niils

8.6

a. 5

470

0.5

L i g h t to moderate attack to a depth of 3 m i l s

7.5

7.1

440

0.9

T h i n hot-leg deposit; no attack

5.4

5.1

471

2.0

Hot-leg layer to 1 mil thick; no attack

4.0

PERIOD ENDING SEPTEMBER 70, 7954 in which lithium was circulated. These loops were constructed of 0.840-in.-OD, 0.147-in.-wal I pipe. The hot and c o l d legs were 15 in. i n length, and the 15-in. connecting legs were inclined at an

angle of 20 deg. T h e welding and loading operations on these loops were performed in a dry box in a purified helium a t m ~ s p h e r e . ~ A t no time during these tests was there any indication of plug fotmation. The operating conditions are given i n Table 6.9. Macroscopic examination revealed no differences between hot- and cold-leg surfaces in loops 1 and 2. Only loop 1 has been examined completely; loops 2 and 3 h w e been sectioned and have been examined macroscopically. Loop 2 was very similar in appearance to loop 1, w i t h no crystal deposition. Loop 3, however, revealed mass-transfer crystals attached t o the cold-zone walls. These crystals did not plug the loop or noticeably affect the circulation. The crystal deposition was heaviest on the major radius of the This exposed loop-bend w a l l in the cold zone. loop has not yet been examined metollographically.

T A B L E 6.9. OPERATING CONDITIONS FOR T Y P E 316 STAINLESS STEEL THERMAL-CONVECTION LOOPS WHICH CIRCULATED LITHIUM FOR 1000 hr .......__. . . .

Loop

No.

Fig. 6.17. As-Received Hastelloy B ( u ) and HotL e g Surface of Hastelloy B L o o p (b) After Circulating NaF-krF4-UF, (50-46-4 mole %) for 1080 hr a t 15OOcF. Etched w i t h H , C r 0 4 Reduced 36%.

-1-

HCI.

250X.

Chemical analysis results now available for the

Hastelloy B loop previously operated4 for 500 hr confirm the low attack rate found metallographical Sy, since neither the nickel nos the molybdenum content in the fluoride mixture increased. L I T H I U M I N T Y P E 316 S T A I N L E S S S T E E L

E. E.

W. H. Cook Hoffman C. F. L e i t t e n R. Brooks Metallurgy D i v i s i o n

Tests have recently been completed on three type 3 16 stainless steel the rma I-convect ion IOOPS

G. M.

- _..

Adomson, ORNL-1729, p 77.

__.~ ~ ~ .4NP @ ~ 2 7 . PTOg. R e p . J u n e 20, 1 9 j 4 ,

Hot-Zone

Cold-Zone

Temperature

Differential

(OF)

1490

1220

270

1472

1355

117

1301

1094

207

Loop 1, which was operated at the highest temperature and w i t h the highest temperature differential, had no mass-transfer crystals in the c o l d zone, and the maximum attack i n the hot zone was 1 t o 2 m i l s (Fig. 6.18). Chemical analyses of the lithium and the amounts of crystals recovered are presented in Table 6.10. A t present it i s not understood why so l i t t l e mass transfer occurred in loops 1 and 2, since i n a11 three loops the same procedures and

testing techniques were used and a l l were f i l l e d from the same batch of lithium. Some as yet undiscovered factor seems to have an effect on the rate of mass transfer. al., Met. D I U . 10, 1954, ORNL-1727, p 37.

T - E r - H o f f m a n et 4Jr.

Serrizmn.

Prog. R e p .

ANP QUARTERLY PROGRESS REPORT

Fig. 6.18. at

Hot-Leg Surface of Type 316 Thermal Convection Loop After Circulating Lithium for 1000 hr 1490OE. Specimen nickel plated after test t o protect edge. Etched w i t h giycerio regia.

F U N D A M E N T A L COR Ri9 S I 0 N R E 5 E A R C H

TABLE 6.10.

IN TYPE 316 STAINLESS STEEL THERMALCO N v E CT IO Id L O O P S

G. P. Smith Metallurgy D i v i s i o n

Mass Transfer i n Liquid L e a d

compounds i n these alloys.

To test t h i s hypothesis

6J. V. Cathcart, A N P Q u a . Prog. R e p . J u n e 10, 1954, ORNL-1729, p 79.

100

__.__~______

Other Than

Lithium, as-cast,

Weight of Crystals Removed from

Lithium (ppm)

Analyzed

As previously

sistance t o mass transfer of materials such a s the 400 series stainless steels might be related t o a tendency towsrd the formation of intermetall ic

~~. .~

Materia I

Metallurgy D i v i s i o n

It has b e e n suggested6 that the increased re-

Components

J. V. Cothcart the investigation of corrosion and mass transfer i n l i q u i d lead has indicated that certain alloys possess much greater resistance t o mass transfer than their pure components. For example, the time required for small thermal-convection loops containing types 410 and 446 stainless steel t o plug was from two t o f i v e times longer than that required far coinparable loops containing pure iron or pure chromium.

ANALYSES OF LlTHlUM CIRCULATED

L ithi urn (g)

(10

160

150

<10

0.010

120

<10

0.001

before test L i t h i u m fromloop

1, after test L i t h i u m fromloop

2, after test ~

............ ~.~ ....

further, a loop was operated which contained specimens of Q specially prepared 45% Cr-55% Co alloy. T h i s a l l o y composition corresponded t o a two-phase region in the cobnlt-chromium phase diagram, one phose being the intermetallic compound CoCr.

The

PERIOD ENDING SEPTEMBER 70, 1954 loop plugged after 768 hr of operation w i t h hotand cold-leg temperatures of 820 and 510째C, reA transverse section of the hot-leg spectively. specimen i s shown in Fig. 6.19. A survey of the a l l o y s which have been tesfed revealed that the plugging time for a l l those i n which intermetallic compound formation i s a possib i l i t y (types 410 and 446 stainless steel, 2% Si-14% Cr-84% Fe, Hastelloy B (5% Fe-28% Mo67% Ni),25% Mo-75% Ni, 45% Cr-55% Co, and 16%

Ni-37% Cr-47% F e (austenite plus sigma)] was in excess of 500 hr, the only exception being the loop containing the 16% Ni-37% Cr-47% F e specimens which plugged i n 456 hr. The plugging times for loops containing the pure constituents of these a l l o y s were: Ni, 2 hr; Co, 80 hr; Cr, 100 hr; and Fe, 250 hr. No mass transfer was observed i n a loop containing molybdenum specimens after 500

hr of operation. On the other hand, loops w i t h specimens of alloys for which there i s v i r t u a l l y no tendency toward compound formation (types 304 and 347 stainless steel, Inconel, and nichrome) a l l plugged in less than 150 hr. On the basis of these data, it was concluded that resistance t o muss transfer in l i q u i d lead i s considerably greater in a l l o y s in which intermetallic compound formation i s possible. Future work w i l l include tests of two alloys: one

w i t h the composition of 50% Cr-50% Fe and the other w i t h approximately 50% Fe-50% Mo. These compositions correspond c l o s e l y t o the compositions of intermetallic compounds i n types 410 and 446 stainless steel, and i f the ideas presented above are correct, both should show r e l a t i v e l y high resistance t o mass transfer i n l i q u i d lead.

Fig. 6.19. Transverse Section of 45% Cr-55% CQ Specimen Exposed to Liquid H o t - L e g Section of a Quartz Thermal Convection E m p .

bead

82Q째C i n

the

101

ANP QUARTERLY PROGRfSS REPORT

Flammability

Sodium

Alloys L, L. H a l l

M. E. Steidlitz G. P. Smith

Metallurgy D i v i s ion The studies of the flammability of l i q u i d sodium alloys, whish were reported p r e v i ~ u s l y ,were ~ ex-

It has been found that the degree of retended, action of jets of sodium-bismuth and sodium-nercury alloys i s not significantly changed when the pressure of air i s varied from 0.25 t o 1.0 atm. Additioncrl data on the reactivity of sodiumbismuth solutions in dry air were obtained, and

action, w h i l e the rest showed slight reaction. The cross-hatched band demarks the approximate I imits w i t h i n which the l i n e separating the region of reaction from the region of no reaction lies. Figure 6.21 shows how the rate of reaction was found t o vary w i t h composition a t constant temperature. T h e rate scale i s arbitrary, w i t h no reaction given the value zero and that of pure sodium given the value 4. Further work i s being done t o determine the flamnrability temperature for pure bismuth,

the results are summarized in Figs. 6.20 and &21. Figure 6.20 Show5 the l i m i t s of the region of no reaction on a temperature-composition diagram. A careful calibration of the flammability apparatus has shown that there was a substantial error in the temperatures previously reported i n these tests. In t h i s respect the earlier data have been corrected.

The c i r c l e s represent the temperature-composition values a t which tests were conducted. The numeral by each c i r c l e gives the number of tests. The unshaded c i r c l e s represent tests which showed re-

7G. P . S m i t h

ond

M. E.

0 3

Steidlitz, A N P Quar. P T O ~R. e p .

r-' '

F i g . 6.21.

-1

O R L I -1 R D\"i(l 3030 ~

1 -

I O

7QQT

Function of

Ternparatuse.

%?rrnodynarnics 2 0

0 6

R e l a t i v e R e a c t i o n R a t e of Sodium-

Bismuth Solutions at

Y -

0 5

MOLE FR4CTION OF SODIUM

D e r . 10, 19.53, O R N L - 1 6 4 9 , p 96.

UNC.ASSlF~-Il

of Alkali-Metal Hydroxides

G. P. Smith

C. R. Boston

Metal lurgy B i v i s ian

iLi E

c 3

c -

700 -

Q d

I w

Free Energy of Formation Data. Numerical data for the standard free energies of formation of the alkal i-metal hydroxides and related compounds are nn essential guide in studies of hydroxide chemistry.

Therefore the free energy data hove been computed a function of temperature for the hydroxide, the peroxide, and the superoxide of sodium. The data are presented in Fig. 6.22, together w i t h the data far the oxide and water which are used in subsequent

NO REACTION

CIS 600

F i g . 6.20. Temperature-Corrrposition Diagram Showing Regions Within Which Jets of SodiumBismuth Solutions Did and Did Mot React with Dry Air.

102

Values of the standard free energy of formation at 25째C are available8 for sodium oxide, sodium peroxide, and sodium superoxide. The value for sodium hydroxide may be computed readily from the values for he standard heat of formation of

'I-,

Brewer, Chern. Reu.

52, 6 (1953).

PERIOD ENDING SEPTEMBER 70, 1954 UNCLASSIFIED ORNIL-LR-DWG 3032

r-1

slope at t h i s point i s determined from values of the standard entropy of formation by using the relation

These entropy data are obtained from the standordstate entropy values. T h e change i n slope at each transition point is determined from the transition entropy, which, of course, i s readily computed from the heat of trans it ion. Recently obtained d a t a l 2 on the heat capacity of l i q u i d sodium hydroxide w i l l permit a more precise evaluation to be made of the slope of the free

energy curve from the melting point up t o 1000째C. The heat of fusion of sodium peroxide (TLM = 980째K)

300

700

500

900

1100

1300

TEMPERATURE ( O K )

Fig. 6.22. Free Energy of Formation Data as a f u n c t i o n o f Temperature for the Hydroxide, the Peroxide, and the Superoxide of Sodium. sodium hydroxide' and the entropies of sodium hydroxide," hydrogen, oxygen, and sodium in their standard reference states. The computation shows = -91.0 kcal/mole. A t higher that AF;,,(NaOH)

temperatures, values of the standard free energy of formation hove been given i n the literature for only sodium oxide,' Howeverp values have been esti-

mated for sodium peroxide, sodium superoxide, and sodium hydroxide in the following way. The curves of the standard free energy of formation vs temperature for the class of compounds dealt w i t h here are known t o a f a i r l y good degree of approximation t o consist of segmented straight lines w i t h discontinuous changes i n slope at the transition points of the compound and of i t s component elements. One point on each of these curves can be fixed from the values for the standord free energy of formation data a t 25"C, mentioned above. T h e 'Unless otherwise specified, literature values used herein are taken f r o m F. D. Rossini, Selected V a l u e s o/ Cb emzcal The mzodynumzc Prope rt ze s, Notion al Bureau of Standards, Washington, 0. C., 1952. 'OJ.

C. R. Kelly

7 3 , 41 14 (1951).

and

P. E.

Snyder,

J. Am. Chern. Soc.

"F. D. Richardson and J. H. E. Jeffes, J. Iron Steel Insf. (London) 160, 261 (1948).

was not available, but it should introduce only a small change i n slope, which may be considered to be negligible for the computations which follow. Decomposition of Hydroxides. I n considering any solvent as a reaction medium it i s important t o determine whether the solvent tends t o decompose t o an appreciable extent, particularly if any of the products of the decomposition take part in the reaction. The classic example is, of course, water. Here the covalent water molecules decompose s l i g h t l y into ions. The resulting equilibrium fre-

quently exerts a controlling influence on the course of aqueous reactions. The following is an outline of the results obtained t o date in a theoretical search for the significant decomposition equilibria It should be i n fused alkali-metal hydroxides. emphasized that the work presented here i s i n no sense d e f i n i t i v e or complete. The most important result of the treatment given here i s the evidence for the possible occurrence of appreciable quantities of peroxide as the result of decomposition equilibria. The way i n which t h i s peroxide may play a decisive r o l e i n mass transfer and corrosion w i l l be shown in a later report. In a fused alkali-metal hydroxide in an inert environment such that the volume of any gas phase i s small compared w i t h that of the l i q u i d phase, the fused hydroxide may be regarded thermodynamic a l l y as being i n equilibrium with every chemical species composed of nothing more than oxygen, hydrogen, and the a l k a l i metal. In practice, many of these chemical species can be ignored for either "W. D. Powers and G. C. Blalock, E n t h a l p i e s m d S p e c z f t c Heats of Alkali and Alkaline Earth Hydroxzdes at f f z g h Temperatures, ORNL-1653 (Jan. 7 , 1954).

1 03

A M P QUARTERLY PROGRESS REPORT of two reasons. First, the equilibrium concentrat i o n of the speries in question maybe insignificant. Second, the rate of formation of the species i n question may be so slow that it does not approach equilibrium in any time period which may be considered. Since i n the following discussion the thermodynamic

approach i s used,

only the f i r s t

reason wi I I be treated. There are three particularly useful theoretical nwosures of the significance of a decomposition They are the equilibrium constant, the reaction. standard free energy change, and the concentration

of the products of decomposition. The t h i r d measure i s the most directly related t o experiment and hence, at the present stage, the most useful. The only concentration i n a fused hydroxide which can be measured at temperature by present methods i s the partial pressure of a gaseous product. Hence, considerable emphasis i s placed on possible decomposition reactions which produce gaseous products and on an expression of their equilibrium concentrations i n terms of the decomposition pres S ure. The decomposition pressure of a pure substance w i t h regard to a gaseous product i s referred t o here as that partial pressure of the gaseous product which, when i n equilibrium w i t h the condensed phase, i s required t o maintain the over-all composition a f t h e condensed phase equal t o the composition of the pure substance. According to the phase rule, the pressure so defined is unique for a given several gaseous products, temperature. I f there a ~ e there i s a decomposition pressure for eoch such product. The standard states used here are those conventionally chosen in defining the standard free energy of formation. Although t h i s is not the usual choice o f stardard states in the case of solutes, it i s d e f i n i t e l y the most convenient choice for the computations which follow. The four possible decomposition reactions which give a gaseous product a t the temperatures a t which the hydroxides are I iquid are considered here. They are as follows:

(4) 2MOH

= 4 MQ,

(3) 2MQH 104

K y 3

MQ,

M i H,

A1;:

AF;

a l l substances are partitioned among a l l present and where K i and Ab': are the

where phases

equilibrium constants and standard free energies of reaction respectively, The sparsity of r e l i a b l e inforniotion on the a l k a l i metal -oxygen cornpounds raises serious questions about the application of all the above e q u i l i b r i a to

all the a l k a l i metals.

T h i s subject i s treated b y

Brewer' and w i l l not be repeated here. However, it i s the essential point of t h i s treatment to de-

termine whether unsaturated oxygen ions such as the peroxide and superoxide ions may occur a s decompos it ion products, in addition t o the saturated oxide ion. The existence of soline sort of unsaturated oxygen ion i s certain for a l l the a l k a l i metals, and it i s known that the s t a b i l i t y of the unsaturated species increases i n going from lithium t o cesium. Therefore, the basic competition referred t o below between hydrogen and the mono-oxide for water should be a characteristic of a l l hydroxide decomposition equilibria and should lead t o a waterhydrogen equilibrium in the gas phase of the type discussed. T h e data for sodium compounds are reasonably re1 iable. Hence, sodium hydroxide wi I I be treated separately. 8 y definition of the equilibrium constant,

where

ax i s the a c t i v i t y of substance X in the

hydroxide phase and N X and yx are the corresponding mole fraction and a c t i v i t y coefficient, respectively. Let the mole fractions be those which are in equilibrium w i t h the decomposition presswre. 'Then N M z 0 = N H Z O I and it i s possible t o write yM20 ......

yH20 Likewise, for Eqs.

hf7;

+ MH + bH2 ,

2 through 4, y,u202 I

YH2

PERIOD ENDING SEPTEMBER 70, 7954

(G2

r3-

I',

2 uizl Q H

three values of the decomposition pressure w i t h

Y M 0 yhf

respect t o hydrogen. Except over unique temperature ranges, one of these three decomposition pressures w i l l be much larger than the other two and hence, by the law of mass action, w i l l s t a b i l i z e

Y M O , YMMH

= r4 - - l

- 4----"

I'4

4 O H

The above equations are exact.

However, a rrtore

useful form can be obtained by assuming that the extent of decomposition i s very small. As w i l l be shown later, t h i s assumption involves a negligible error for a l l four reactions. From t h i s assumption two conclusions can be drawn: (1) the a c t i v i t y of the hydroxide as used in the above equations may be taken t o be unity because of the choice of standard states; (2) the partial pressures of water

the others. Thus, Qnly one of the three pressures w i l l be the decomposition pressure of ,llOH w i t h regard t o hydrogen. Because of the lack of thermodynamic data on a l l the alkali-metal hydroxides except sodium, it is necessary, in most applications, to use ratios of decomposition pressures. Thus, it I S assumed that r e a c t i o r ~ s1 and 2 are simultaneously a t equilibrium and that they determine the gaseous water and hydrogen pressures, respectively. By d i v i d i n g Eq. 56 into 5a, the decomposition pressure ratio becomes

and hydrogen in the gas phase may be taken t o be equal t o the fugacities in the gas phase and hence equal t o the fugacities i n the l i q u i d phase. Finally, by definition, the fugacity of the gases in the standard reference state i s unity. Substituting these conclusions into the above equations, solving for the partial pressures, and applying the reaction isotherm g i v e

2/5

where ( p H [

)

..2Ii

where

T h e concentrations of the oxide in reaction 1 and of the peroxide in reaction 2 are not independent but are related through the equilibrium

4"s" i s t o be read as the partial p r e s s w e

of hydrogen provided that no other hydrogenproducing reaction i s operative save the Ith. From the original choice of m o l e f r ~ c t i ~ n s it may be noted that the ubove partial pressures are the decomposition pressures and that there ure

Furtkrrrrore, the concentrations of oxygen in t h i s equilibrium atid of hydrugen in reuction 2 are related through t h e equilibrium

(8)

AF,.

ANP QUARTERLY PROGRESS REPORT

6.11 were obtained. The pressure-

The existence of these equilibria i s coinpatible

shown i n ? a b l e

ha, as may be seen from the fact that AF; - AF;, K , , , = K ~ / K ~ Hence . it woucd have been possible t o derive Eq. Sa by using equilibria 7 and 3. T h i s i s because the t w o sets of equilibria 1 and 2 and equilibria 7 and 8 are

a c t i v i t y coefficient products in t h i s table give the

w i t h Eq.

AFY

pressures which would be obtained i f the solutions were ideal. Figure 1.23 shows the relative im-

portance of equilibria

1 and 2 i n terms of the loga-

rithm of the pressure r a t i o as a function of the Figure reciprocal of the absolute temperature.

both equivalent t o the one equilibrium

6.21

gives the free energy differences (AF" and 1,2 as functions of absolute temperature. 1,3 Examination of these data shows the following: (1) Apart From unexpectedly large deviations from ideality, equilibrium 1 w i l l predominate a t a l l

AF" ) From

Eq. 9 it w i l l be noted that the pressure ratio

i s determined by the competition (Pt+20'PH2 ) 1 , 2 between the oxide and hydrogen for oxygen, w i t h the one trying t o form the peroxide and the other trying t o form water. In l i k e manner, an expression analogous t o Eq. 6a can be derived under the assumption that equil i b r i a 1 and 3 determine the water and hydrogen pressures, respectively. T h i s expression i s

temperatures.

(2)

Assuming

ideality,

hydrogen

evolution from fused sodium hydroxide in an inert container w i l l be small a t all temperatures, but it should be measurable at 906째C. Likewise, the peroxide concentration at al I temperatures may be small, but it should be appreciable a t 900째C. (3) Deviations from ideality of a moderate amount could make hydrogen and peroxide formation either insignificant or greatly important a t high temperatures.

UNCLASSIFIED ORPll -l-R-DNG

3033

K1,3 =---

K3'l3

30L 60

Fig. 6.23. for NaQaH.

B y using the free energy data computed above for sodium hydroxide, the representative values

T A B L E 6.11. Temperature

("a

AFY (kcal)

10 0 420 14 0 (!/TEMPERATURE) x io4 (OK)-'

160

Decomposition Pressure Functions

DATA ON THE DECOMPQSlTlON O F SODIUM HYDROXIDE

At.;

(kcal)

~ . ~ ~ ~ . . . ~ ~

106

_. ..~.

35.3

76.0

600

21.4

64.4

4 . 7 x 10-6

900

13.5

59.9

10 -26

x1~-3

8 x

(mm Hg)

~~~.~ .~ ..........

10-56

10-95

10-18

10-27

lC-18

1.7 42.0

(mm Hg)

.........-

7.4 x 10-6 2.1

PERIOD ENDfNG SEPTEMBER lo, 1954

UNCLASSIFIED

ORNL-LR-OWG .....................

3034 I

The degree of approximation can be checked by using the data for sodium compounds. In t h i s case AFT,2 i s -35.5 kcal, when using the approximation,

........

500

700

900

TEMPERATURE

Fig.

....

...

300

1100

!300

(OK)

6.24. F r e e Energy Differences as Functions

of Temperature,

as compared w i t h the correct value of -40.7 kcal. i s -94.2 k c a l compared w i t h the Likewise, The approximate correct value of -94.9 kcal. values of and A F ; , 3 at 25째C are tabulated i n T a b l e 6.12 for a l l the alkali-metal hydroxides. Table 6.12 shows that hydrogen and the peroxide should be produced most readily i n potassium hydroxide and least readily in lithium hydroxide. Nevertheless, equilibrium 1 i s predominant, and equilibrium 3 and hence equilibrium 4 are insignificant. T A B L E 4.12. AND

There are no satisfactory heat of formation data for any of the alkali-metal hydroxides other Therefore, only the relathan those for sodium. t i v e importance of the decomposition equilibria Some of the entropy values for can b e studied.

the higher oxides have been estimated and are

However, the heats reported in the literature. of formation for rubidium and cesium oxides are quite uncertain, and it i s presently believed that the accuracy of the AFYIZ and AF';,, values given i n Table 6.12 would not be improved by use o f these entropy estimates. Therefore, only enthalpy values w i l l be used, but it w i l l be shown that t h i s restriction is not so serious as it might at f i r s t seem t o be. Comparing equilibria 1 and 2 shows

All

the data required i n the above equation are available, except those for the entropy term. Since t h i s term represents a difference i n entropy for t w o similar substances, it may be supposed that i t s omission w i l l not be serious. A similar situation obtains for AFY,3. Hence, the following approximotions w i l l be used.

APPROXIMATE A T 25OC

AF;,3

A";, 2 (kcal)

A q ,3 (kcal)

-47 -36 -25

-76

-35

- 70

-34

- 94

- 73

Various investigations o f corrosion and mass transfer hove shown the fundamental role played by hydrogen. The above computations show that hydrogen i s important i n s t a b i l i z i n g sodium hydroxide a t temperatures above 6QO"C, as shown in

-__Table 6.11. T h i s i s the temperature range a t which mass transfer becomes significant. These computations also l i n k the formation of hydrogen w i t h the formation of unsaturated oxygen ions. As w i l l be shown in a subsequent report, the presence of these unsaturated ions may be necessary for mass

107

ANP QUARTERLY PROGRESS REPOR7 transfer t o occclr. A s may be seen in Table 6.12, hydrogen and the unsaturated oxygen ions are most important for the hydroxides of potassium, rubidium, and cesium. Furthermore, these computations provide the f i r s t estimate of the thermal s t a b i l i t y of the alkali-metal hydroxides (other than LiOH). These estimates confirm the frequently made assumption that the alkali-metal hydroxides are b y far the most thermally stable hydrogen-containing liquids known.

It must be emphasized that these computations are given as a progress report in a continuing investigation. Thus, many of the values obtained, such as the standard free energies of formation, do not i n themselves provide direct information about hydroxide corrosion, but they are essential values for any theoretical investigations of corrosion. Experimental Studies (with M. E. Steidlitz, MetaIlurgy Division). A n investigation i s under way t o determine experimentally whether a significant amount of hydrogen i s liberated when NaOH i s heated t o 900°C i n an inert environment. The preliminary method being tested i s to use a pair of automatic Toepler pumps to concentrate any gaseous products liberated from the hydroxide. The collected gas i s analyzed w i t h a mass spectrometer. The most d i f f i c u l t aspect of t h i s research has been the endeavor t o provide an inert container for the NaOH, which reacts w i t h a l l metals studied a t 900°C (including the most noble metals) t o produce hydrogen. It reacts strongly w i t h nearly a l l teromics except magnesium oxide. Unfortunately,

nonporous crucibles of pure MgO are not available because of the d i f f i c u l t y i n sintering t h i s material. Ordinary M g 0 crucibles, which usually contain a small amount of binder such as silica, break down due t o attack on the binder. The problem was solved b y machining out large single crystals of pure magnesium oxide prepared by the Ceramics Group.

In the experiments which have been performed t o date, appreciable quantities of hydrogen have been found in addition to much ,larger quantities of water. However, there was evidence of contaminat i o n by organic material and the hydrogen may have come from t h i s source. These studiesare continuing,

13H. J. Buttram e t ai., A N P Q u m . Prog. R e p .

1954, ORNL-1729,

108

63.

J7mp

10,

C H E M I C A L STUDIES O F C O R R O S I O N

F. Kertesz Materials Chemistry D i v i s i o n

Effect sf Temperature o n Corrosion of Inconel and Type 316 Stainless Steel

H. J.

Uuttram

R . E. Meadows N. V. Smith

Materials Chemistry D i v i s i o n The effect of temperature on the corrosion of lnconel by molten fluorides in static tests, as indicated by the extent of void formation observed and by chromium concentration of the melt, was discussed in a previous report.13 In those studies there was some evidence that a maximum in the

corrosion could be observed a t 800 t o 900°C and that both v o i d formation and chromium concentration i n the melt were less a t 1000°C than a t 800°C. While it i s possible t o rationalize the decreased v o i d formation on the basis that a high rate of diffusion for chromium i n the metal may minimize the formation o f voids, the decreased chromium concentration in the melt is, hawever, more d i f f i c u l t to explain. During the past quarter NaZrF, and NaF-ZrF,-UF, (53.5-40-6.5 mole %) were tested i n type 316 stainless steel under static and nearly isothermal conditions for 100 hr a t 100°C intervals over the range 600 t o 1000"C, as a check of the findings w i t h Inconel. Metallographic examination of the capsules exposed a t 600°C revealed light intergranular penetration up t o 1 m i l i n depth. At 1000°C the attack was s t i l l light, but one isolated case of intergranular penetration t o a depth of 18 mils W Q S noted. When the chromium concentration of the melt was plotted against t e s t temperature for the NaZrF,, [I slight maximum i n the curve appeared

between 800 and 900°C; a similar p l o t for the UF,bearing mixture shows a plateau between 700 t o 1000°C. However, more chromium appeared in solution i n the NaZrF, samples than i n the UF,bearing mixtures. The tendency toward less corrosion a t 1000°C than a t 800°C never appears t o be pronounced, but it has been sufficiently persistent i n these tests t o require explanation. A plausible explanation can be evolved i n the following manner. ?he NaZrF, melt which has been used in these experiments gave r i s e to abnormally high chromium concentrations (1000 ppm) compared w i t h equilibrium values (200 ppm) obtairred i n measurements carried out w i t h carefully

PERIOD ENDING SEPTEMBER IO, 7954 handled melts and pure chromium metal. Itappears, therefore, that impurities were predominantly responsible for the amounts of corrosion measured. The most l i k e l y impurity i s HF, which can result from hydrolysis of adsorbed water or hydrated water in salts containing Na3Zr4F,, or from inadequate pur if ication.

If HF were responsible, the ascending portion of the curve could be due to an increasing rate o f reaction w i t h increasing temperature. (There i s l i t t l e reason t o believe that equilibrium conditions The are reached i n static capsules in 100 hr.) descending portion of the curve may be ascribed t o a marked decrease in corrosiveness of HF a t increasing temperatures. Standard free energy e ~ t i m a t e s 'show ~ that Ai:' for the reaction of HF w i t h Ni t o form NiF, becomes positive

near 5OO0C, w i t h F e t o form FeF, w i t h Cr to form CrF, a t 1400째C.

800째C, and

at In

the experiments under discussion here, the melt a t +t with the end of a 100-hr test contains mostly Cr

very l i t t l e Fe" and Ni". There i s some evidence that a t the lower temperature a11 three elements are attacked rather s l o w l y and indiscriminately by HF and that the resulting Fe" and Ni" are replaced by Crt+ i n a fast

secondary reaction. The effect of increasing temperature i s t o increase the degree of approach t o eqoilibrium in 100 hr in a static capsule; however, a t higher temperatures the effect of a more unfavorable equil ibrirrm constant becomes manifest. Iron and n i c k e l are r e l a t i v e l y unreactive toward HF a t IOOO"C, and the amount of Cr++pickup in 100 hr could w e l l be less than that noted a t 800째C for three reasons: the free energy change for the reaction o f chromium w i t h HF i s smaller, fewer F e f + and Nit' ions are present for reaction w i t h Cr", and there i s mechanical interference by the unreacted n i c k e l and iron. I n other words, chromium i s most readily oxidized from an a l l o y by HF if the accompanying a l l o y constituents are also attacked. Corrosion

H. J. Buttram

by Fission Products

R. E. Meadows D iv is ion

Materia I s Chemis try

previously reported experiment,"

a fuel

mixture containing simulated f i s s i o n products a t 1000 times the concentration expected in the AWE

was Corrosion tested. Very heavy attack was observed; t h i s was expected, not because normal amounts of f i s s i o n products are particularly corr o s i v e but because oxidizing agents such as RuF4, MoBr,, and elemental tellurium were used i n sufficient concentration t o cause excessive corrosion in any case. Additional experiments were performed w i t h t h e use of the same additives, individually, i n amounts as small as possible i n an attempt t o measure the relative a c t i v i t i e s as corrosive agents. Preliminary results, such as those previously reported which showed YF, t o be extremely corrosive, and additional t r i a l s during t h e past quarter, which showed

that such compounds as CsF were also very corrosive, made it obvious that the techniques employed were unsuitable and that very misleading indications were being obtained. Free energy considerations, as w e l l as general experience w i t h fluoride systems in closed capsules, make it apparent that impurities such as HF and water were responsible for the effects noted. Hence it must be concluded that none o f the experiments performed t o date on corrosion by simulated f i s s i o n products have been satisfactory for the intended purpose.

Controlled-veiocity Corrosion Testing Apparatus Smith F. A. Knox Materials Chemistry D i v i s i o n

N. '4.

I n order to more nearly simulate corrosion conditions of molten fluorides circulating through reactor components, a control led-velocity corrosion testing apparatus has been constructed. T h i s apparatus, similar to one previously described,16 allows the rapid transfer of molten fluorides through both heated and cooled test sections. The apparatus consists of t w o 4-in.-ID, 24-in.-high lnconel c y l i n drical pots connected by means o f three sections of fh-in. inconel tubing w i t h a W Q I I thickness of 0.035 in, The two outside sections are 5 ft long and the center section, which was i n i t i a l l y 3 ft long, WQS later increased t o 3'/, ft, T h e center section i s cooled b y either air or water. Suitable furnaces and heaters permit holding the pots and transfer Each inconel pot lines at desired temperatures. has a gas i n l e t which allows application of helium

"L. Brewer et (21." 7'?ie Tbeririodynamic P r o p e r t i e s w t d Equilibria a t N i g h Temperatures of Vrunium Ualides, Oxides, Nitrides, rrnd Curbzdrs, MDDC-1543 (Sept. 20, '1945, rev. Apr. 1, 1947).

109

AM? QUARTERLY PROGRESS REPORT at predetermined pressures to transfer the melt. A prohe indicates when the desired liquid level i s reached in the second pot and activates a rejoy which closes the helium i n l e t to the f i r s t pot and opens a corresponding one in the second pot t o reverse the cycle. Velocities obtained with the apparcstus have reached 5 fps, which corresponds to a Reynolds number of only about 1500. It i s hoped that art improved design now being assembled w i l l make possible turbulent flow. A 200째C gradient across

the cooling section was desired; however, it has not yet been possible t o achieve o differential of more than hQ"C. T h e transfer c y c l e consists i n having the melt contained i n either pot a t 700"C, being heated to 8W"C in the f i r s t section of the

transfer line, cooled in the iiiiddle section, brought back t o 700째C in the third section, momentarily

stored i n the opposite pot cat 7QO"C, and then the c y c l e i s reversed. A complete cycle, that is, from one pot to the other and back, required approximately 5 min. It i s believed that this apparatus w i l l y i e l d corrosion data under conditions not easily obtainable by presai-at means. Weactiom Between Graphite nad Fluoride Melt

F. A.

Knox Materials Chemistry D i v i s i o n Graphite has been used extensively as n CORtainer material during preparations of fuel and

110

coolant materials; therefore, an investigation WQS made of possible reaction between it and a fluoride melt. In the procedure used an lnconel reaction tube was loaded w i t h either NeF-ZrF,-UF, (53.5-

40-6.5 mole %) alone or w i t h the fluoride mixture

plus graphite which had been treated under vacuum for degassing. The reaction tube was connected t o a manometer and then held a t 1000째C u n t i l the gas pressure became constant. It was found that the rather large gas pressures Analysis of obtained were due chiefly t o SiF,. the graphite indicated the presence of nearly 1% Si, The use of spectrographically pure graphite in this apparatus h a s yielded pressures very nearly the same as the pressure of the pure fluoride. No HF or CF, w a s detected during these tests.

1- ithi urn FiMor ide Casti rigs N. V. Smith F. Ksrtesz Materia I S Chemis try D i v i s i o n

A t the request of the Instrumentation and Controls Division, high-density lithium fluoride cylinders were cast from powdered material for use i n an These cy1 inders were prex-ray spectrometer. pared by charging the powder t o graphite crucibles

and heating in an inerf atmosphere to 900째C. T h e graphite liner containing the melt was then removed from the furnace and a strong air i e t was applied t o i t s periphery, The resulting cylinders were uniform and acceptable.

PERIOD ,ENDING SEPTEMBER 10, 1954

7. METALLURGY W. D. Maniy Metallurgy D i v i s i o n Creep and stress-rupture testing have, i n the post, been performed primarily on u n i a x i a l l y stressed specimens.

T e s t s of t h i s type are continuing,

but

the interest i s now centered on tube-burst tests i n which a tube that i s closed a t one end i s stressed w i t h on internal gas pressure. The stress pattern introduced into the specimen i n t h i s test more nearly approaches t h e stress pattern that w i l l be found i n ANP-type reactors. The ratio o f the tungential stress to the longitudinal stress i s quite important in the d u c t i l i t y of the metal being tested, ond therefore equipment i n which the longitudinal and the tangential stresses can be varied has been constructed. A theoretical analysis of a stressed cylindrical pressure vessel has been made w i t h which o check on the experimental results can be obtained. Also, to determine the effect of tangential and compressive stresses on the corrosion rate o f metal i n contact with fused salts, on opparatus for loading a sheet specimen i n bending was designed and constructed; t e s t s are now being made. In the investigation of high-thermal-conductivity materials for f i n s for sodium-to-air radiators, stressrupture and creep tests were made on copper fins w i t h various types of cladding at stress l e v e l s be-

tween 500 and 2000 psi a t 1500째F. The tests have shown that for a 1000-hr exposure in air, stresses greater than 500 p s i and less than 1000 p s i ore tolerable; that is, i n t h i s stress range there i s n o indication of brittleness in the core or oxidation of the core due t o cladding failure. Thermal conduct i v i t y measurements of a 6% AI-94% Cu aluminum bronze were made in the temperature range 212 to

1562OF.

T e s t s o f brazing olloys have shown that i n brazing high-conductivity fin materials t o lnconel tubing there are four a l l o y s that can be used: low-melting-

point Nicrobraz (LMNB), Coast Metals alloy 52, electroless nickel, and an Ni-P-Cr alloy. From the over-all considerations of melting point, oxidation resistance, d i l u t i o n o f f i n and tube wall, formation o f low-melting eutectics, and flowability, it was found that Coast Metals alloy 52 was the best alloy for the construction of radiators w i t h highconductivity fin s.

A sodium-to-air rodiator w i t h 6 in. of type-430 stainless-steel-clad copper high-conductivity f i n s was fabricated by using a combination heliarc welding and brazing procedure. The tube-to-fin sections were assembled and brozed w i t h Coast The tube-to-header joints were Metals alloy 52. made by using the semiautomatic welding equipment and then back-brazing. The header sections were closed by manual heliarc welding. In the construction o f a 100-kw gas-fired liquidmetal-heater system, packed-rod nozzle assemblies were needed as the i n l e t s for a i r and gas. These devices were made by brazing 1G-in.-dia stainless steel rods i n a t i g h t assembly with a ductile, oxidation-resistant alloy o f 82% Au--l8% Ni. Work has started on the formation o f duplex tubing i n an attempt to prepare alloy composites that have good corrosion resistance on the inner surface and oxidation resistance on the outer surface. Composites of copper and type 310 stainless steel, lnconel and type 310 stainless steel, Inconel and Hastelloy B, and Hastelloy 5 and type 310 stainless steel have been prepared. Tubing produced from stainless-steel-clad molybdenum and columbium and from several special Inconel-type a l l o y s was fabricated into thermal convection loops for fluoride corrosion testing. Several new a l l o y s were produced for the corrosion tests w i t h l i q u i d lead. Attempts are being made to find new alloys in the nickel-molybdenum system that wi It have better high-temperature strength and fluoride corrosion resistance than lnconel has. Hastelloy B satisfies these requirements, but it has poor fabrication properties and oxidation resistance, and i t loses i t s d u c t i l i t y i n the temperature range of interest for appl ication i n high-temperature c ircu fating-fuel reactors. Investigations are under way i n an effort to f i n d a suitable melting and heating treatment that w i l l increase the d u c t i l i t y of Hastelloy B in the temperature range of interest. Investigations of methods for producing boron carbide shield pieces of the required density and shape indicate that the pieces should be molded w i t h nonmetallic bonding material by cold pressing followed by sintering. The bonding materials being

111

A M P QUARTERLY P R O G R E S S REPORT

studied include sodium silicate, nitride, and boric oxide.

silica,

The fluoride-to-sodium intermediate changer, which failed i n a l i f e test after

silicon

heat

1680

exhr of

the following equations:

(1 )

service i n the temperature range 1080 to 15OO0F, was examined, It i s probable that the

cyclic

- POTO

- r 2.

PIT2

2 2 Tire (p,

*..

r 2.

r02

failures i n the tube-to-header welds were caused by unequal thermal expansion, which caused stress concentration at the roots of the tube-to-header welds. These stresses, combined with expansion and contraction of the tubes, would tend to propagate cracks through the walls. These failures emphasize the extreme desirability of using backbrazing as a means of rniriirnizing the notch effect i n tu be-to-header welds. STRESS-RUPTURE TESTS O F INCONEL

R. B.

J. H.

Oliver D. A. Douglas J. W. Woods DeVan Metallurgy D i v i s i o n

The tube-burst test for obtaining information on the stress-rupture properties o f lnconel has been studied intensively. The stress pattern introduced into the specimen i n this test simulates to some extent the stress pattern that w i l l be present i n circulating-fuel reactors. The test consists of stressing a closed-end tube with internal gas pressure. In tests o f this type reported previously,' i t WQS observed that lnconel specimens stressed in this manner showed less d u c t i l i t y and much shorter rupture l i f e than the uniaxially stressed specimens, and therefore an intensive study o f the multiaxial W. Jordan o f the stress system was initiated. Mechanics Department of the University of Alabama i s investigating t h i s problem. P a r t o f the investigation consists of a study of the theory of stresses i n cylindrical pressure vessels, w i t h particular attention to the variations in stresses calculated by the thin-wall formula vs the stresses determined by the niore exact Lame' theory. The stresses under discussion are those i n the w a l l s only, with no consideration given to the end closure shape, except that the ends are assumed to be completely closed. The three principal stresses a t a given point are the radial stress (G), the tangential (hoop) stress (ut), and the longitudinal (axial) stress ( u j . The theory of e l a s t i c i t y y i e l d s

'R, â&#x201A;Ź3. O l i v e r , B. A. Douglas, and J. W. Woods, A N P Qrtnr. P r o g . Rep. ] m e 10, 1954, ORNL-1729, p 89.

112

where r i s the radial distance to the point at which the stress i s desired, r z and ro are the internal and

external radii, and p i and p , are the internal and external pressures. For the special case of the cylinder subjected to internal pressure only ( p , =O), the equations reduce to:

(4)

u . = YZ

- 0 ,

r 2.

u UZ. = ua*

___

r2 0

r2 1

where the additional subscripts i and o an the stress terms indicate stresses at the inner and outer surfaces, To simplify these equations for application t o cylinders w i t h thin walls, it i s fre-

quently assumed that the tangential stress does not vary across the w a l l of the vessel. T h i s simplification results i n the following equations for internal pressures only:

(7)

where the letter s i s used to denote stress and I i s the thickness o f the wall ( t = ro - r.). A simplified expression for radial stress i n thin-walled cylinders i s not commonly used.

PERIOD ENDING SEPTEMBER IO, 7954 UNCLASSIFIED ORNL-LR-DWG 2848

In order t o determine the error involved i n using the formulas for thin-walled vessels, ratios o f the stresses computed by the exact and the approximate theories are given below.

1.28

I 24 1.20

s,i

T .

24.-

It6

1.12

T .

108 U

1.04

T .

4 .oo

0.36

0 32

Equations

9 through 11 are plotted i n Fig. 7.1.

The so-called "exact" formulas (often c a l l e d the Lam; formulas) are v a l i d only w h i l e the stresses are elastic. The approximate formulas are based on equilibrium considerations and the assumption that the stresses are distributed uniformly across the wall. Under i n e l a s t i c (including creep) conditions, i t i s believed that the tangential (hoop) stress distribution l i e s somewhere between those assumed in the Lam; and i n the approximate e l a s t i c theories. With increasing stress, increasing temperature, and increasing time, the actual stress distribution w i l l depart from the Lame' assumption and approach the thin-wall theory assumption. Thus the two theories g i v e the upper and lower bounds for the actual stress distribution. For the longitudinal (axial) stresses, the Lam: and the thin-wall theories both use the assumption that the stresses are uniformly distributed over the cross section. Since the Lam; theory uses the correct value for the cross-sectional area, while the thin-wall theory uses an approximate area, the Lam; theory i s believed t o be more correct i n both the e l a s t i c and the i n e l a s t i c ranges o f stress. The majority of the tube-burst tests made to date have been of 0.010- and 0.020-in.-woll tubing, and thus the approximate formula applies. Specimens are now being tesfed which have 0.060-in. walls. Since i t i s predicted that under the t e s t conditions the

0.M

o a4

STRESS COMPIJTED BY THIN-WALL

THEORY

l--_l_.._

f -

Fig. 7.1. Comparison of E l a s t i c Stress in Cylinders a s Computed by Lam6 Theory and by Thin-Walled Pressure Vessel Theory. stress w i l l be uniform across the w a l l thickness the approximote formula w i l l be used for calculating t h e stresses, ond the s l i g h t error thus introduced wi II be neglected. It i s w e l l known that in the case o f o thin-walled closed-cylinder pressure vessel the ratio of the tangential (hoop) stress to the longitudinal (axial) stress i s 2 : l . Some investigators have shown that t h i s stress distribution results in the minimum d u c t i l i t y for a given material. Therefore, i n order to better understand the criteria for predicting failure o f tubular specimens and i n order to study the effects o f anisotropy of the tubular material, it would be desirable t o be able t o test a series o f specimens w i t h varying stress ratios. Jordan has devised the apparotus described below w i t h which it i s p o s s i b l e to stress a tubular specimen purely tangentially, purely axially, or i n ony desired r a t i o of these stresses.

113

ANP QUARTERLY PROGRESS REPORT The most obvious method of changing the stress ratio i s by loading a tubular specimen axially w i t h

(Note that the diamefer n of the piston rod does nat enter into the calculation of the stress ratio.)

pressure acting on the pistons i s borne by the rod connecting the pistons. T h i s w i l l reduce the axial stress i n the cylinder w a l l to zero (assuming no friction between piston and cylinder) without afIn actcsol practice, fecting the tangential stress. one piston can be replaced by a fixed end. By using a piston o f larger net area than the opposite cylinder end, compressive axial stresses can be introduced into the cylinder w a l l s without altering

2d ( d d2

r a t i o i s given by the equation

i n Fig. 7.3. By using thin-walled cylinder approximations, the stress ratio i s given by the equation ut

...............

In the event that it i s necessary to extend the piston rod through the top end of the cylinder (to provide Q thermocouple well, for example), the arrangement shown i n Fig. 7.4b may be considered. The stress

the tangential stresses. T h i s arrangement i s shown

......

2d ( d t t )

2d (d

_ -

+t)

...............

- b2

- D2

UNCLASSI FlED ORNL-LR-DWG 2850

IJN C L A S S I FIE D

ORNL-LR-OWG

PISTON

__--

_---

__--

2849

SPECIMEN

- PISTON ROD

-PISTON

F i g , 7.2. Apparatus for Producing Tangential Stress in a T u b u h r Specimen.

114

Fig.

7.3.

Apparatus for Producing Compressive

A x i a l Stresses in a Tubular Specimen Altering the Tangential Stresses.

PERIOD ENDING SEPTEMBER IO, 7954

U NGI.AS S IFIED ORNL-LR--DWG 2851

DIMENSION D MAY BE GREATER OR LESS THAN d

Fig. 7.4. Apparatus for Obtaining a Pasitive Stress Ratio of u,/ca in o Tubular Specimen. ( a ) out thermocouple well. ( b ) With thermocouple well.

With-

ANP QUARTfRLY PROGHESS REPORT

As before, the stress ratio i s independent of the diameter a of the piston rod. Since the piston rod

UYCLASSIFIED ORNL-LR-DWG 2852

i s i n compression, however, i t should be made large enough to prevent buckling. The bleed lines shown i n Figs. 7 . 4 ~ and b should be kept open

during testing to assure that no pressure leaks from the large-diameter chamber to the smaller one. A t e s t r i g has been b u i l t that imposes the condition rrt/oa = no. Strain measurements made w i t h a strain-gage bridge fastened to the gage area showed that the friction in the system was negligible and that the only active stress was tangential. A specimen i s now being tested at 1500째F under t h i s stress cond it ion. There has been considerable speculation as to the p o s s i b i l i t y of a difference i n the rote of corrosive attack of fused salts i n contact w i t h lnconel under tensile stresses as compared w i t h lnconel i n compression. In order to observe any difference, it is desirable that the compressive and tensile stresses be imposed on the same specimen and be of the same magnitude. One approach to t h i s problem i s to test a specimen under pure bending conditions. T h i s would ensure that the magnitude o f the maximum tensile stress would always be equal t o that of the maximum compressive stress during a given test. A n apparatus which would produce t h i s effect was also designed by Jordan, and a specimen A i s now being tested in o fused s a l t medium. drawing of the specimen and the loading apparatus i s presented in Fig. 7.5. If it i s assumed that no f r i c t i o n occurs between pins C and and the specimen, all the force P wi I I be transmitted across

section 1-1 by the l i n k I_ and no axial force w i l l e x i s t in the specimen. Likewise, since l i n k L i s loaded axially, it cannot transmit any bending moment and therefore the specimen must r e s i s t any

The bending moment existing across section 1-1. force i n l i n k L w i l l be equal to P to satisfy the equilibrium of forces. The bending moment i n the

specimen will then be equal to the product of P and the horizontal distance between the centers of pins A and 6. In order to minimize variations i n t h e bending rrioment (and therefore stresses) i n the specimen during a test, the horizontal distance between A and B should be held as nearly constant as possible. Since the specimen deformsunder load, the ends will rotate and some change i n the distance AB must occur. In order to minimize the change, the pins A and B are placed with their centers s l i g h t l y above and below, respectively, a horizontal

116

SPECIMFN

F i g . 7.5. men

Apparatus for Testing a Sheet Speci-

Under P u r e

ending C o d it ions.

diameter. T h i s w i l l allow some rotation to occur with very l i t t l e change i n the horizontal distance. The apparatus shown in Fig. 7.5 has been constructed. Strain gages were placed on the tension and compression sides of a specimen, and load was In the range of loads contemplated i n a applied. creep test, the tensile and compressive strains were essentially of the same magnitude; thus very l i t t l e friction existed between the p i n s and the specimen. HIGH-CONDUCT I V I T Y -F IN SODIUM-TO - A I R R A D ! A T 0 R Developmental work has continued on a sodiumto-air radiator w i t h f i n s o f a high-thermal-conductivity material. The developmental effort includes investigations of materials with high thermal conductivity, the development and testing of brazing

PERIOD ENDING SEPTEMBER IO, 1954 a l l o y s for use i n fabricating the radiator, ond the fabrication of radiators for service testing.

Investigations of Fin Materials

J. H.

Coobs H. lnouye Meta IIurgy D i v i s i o n

Stress-rupture and creep tests o f type-310 stainless-steel-clad copper fins that were 8 m i l s t h i c k

were made at 1500째F a t stresses between 500 and psi. The material was obtained from the General P l a t e Div. of Metals & Controls Corp. and

2000

had an average cladding thickness o f 1.87 m i l s per side. The maximum cladding thickness was 2.3 m i l s per side. The results of the several tests

mode show that some alteration of the copper OCcurs even though the specimen does not fracture. T h i s was indicated by the brittleness of the normally ductile composite. For long-time exposures (1000 hours), stresses greater than 500 p s i and less than 1000 psi are tolerable and there i s no indication o f

brittleness in the core or oxidation due t o cladding failure. I n a l l t h e specimens which ruptured, o x i dation o f the copper at the point ot fracture was complete, and the copper which did not appear to It was also noted that be oxidized was brittle. when the strain i n 2.5 in. was lo%, the copper remained ductile, w h i l e strains above 20% (regardless of exposure time) produced brittleness. The data obtained are given in Table 7.1. The aluminum bronzes are among the materials being considered as cladding for copper fins. Although extensive diffusion occurs between these alloys and copper, increases in the thermal conductivity of t h e fin can be realized b y cladding the copper with these alloys. The composition gradient e x i s t i n g i n a diffused composite results i n an increase i n the conductivity of the cladding material (because of the outward diffusion o f aluminum) and a decrease i n the conductivity of the copper core. Thermal conductivity measurements of two aluminum bronzes were made by the Solid State D i v i s i o n and are reported i n Table 7.2. T h e values given for measurements a t temperatures up to 392째F compare favorably w i t h the datu of Smith and Palmer,2 but for temperatures above 392째F they exceed the Also, the values extrapolated from their data.

2C. S. Smith and E. W. Palmer, Am. Inst. Mining Met. Engrs., Metals Technol. Tech. Pub. No. 648, 1-21 ( 1935).

T A B L E 7.1. STRESSRUPTURE PROPERTIES OF T Y P E-310 STAINLESLSTE EL-CLAD COPPER FINS A T 15OO0F & m i l cladding on each side of 4-mil copper sheet

Stress

Rupture Time

(hF)

(Psi)

Elongation in

2.5

in.

(%I

Remarks

2000

Capper embrittled

1800

336

Copper embri ttled

1650

52 8

Copper embrittled

1500

887

1400

1220

Copper embrittled I n progress

1000

27.5

T e s t terminated a t 1000 hr; specimen o x i d i z e d throughout

1000

T e s t terminated a t 500 hr; oxide stringers on surface of specimen

500

T e s t terminated a t 1000 hr; no indication of failure

500

Test terminated a t 500 hr; no indication o f failure

T A B L E 7.2. THERMAL CONDUCTIVITY OF ALUMINUM f3RONZESATVARlOUS TEMPERATURES T h e rma I Conductivity Temperature (OF)

[col/sec.cm2

("C/C"S1

.................................................

For

6.2% A1-93.8%

For

8.4% AI-91.6% Cu ..........

212

0.178

0.176

30 2

0.240

0.210

3 92

0.280

0.250

1156

0.55

1562

0.77

estimated value of s i x times the thermal conductivity o f stainless steel at 1500째F was exce=ded by the experimental value for the &2% AI-93.8% Cu alloy, The values given i n Table 7.2 were not corrected for the volume expansion of the a l l o y with temperature and are therefore accurate only to w i t h i n

*5%.

117

ANP QUARTERLY PROGRESS REPORr Development of Brazing Alloys for U s e i n Fabricating Radiatars

K. W. Reber

and the change i n core chemistry from copper to a copper-nickel alloy would seriously reduce the heat transfer efficiency o f the fin. Other claddings, such as types 310 and 430 stainless steel, have been

P. Patriarca G. M. Slaughter

found t o be relatively immune t o diffusion effects,

Metallurgy D i v i s i o n

J. M.

Cisar

Aircraft Reactor Engineering D i v i s i o n Several a l l o y s were investigated for brazing copper f i n s clad w i t h Inconel, type 310 stainless steel, or type 430 stainless steel in order to establish an optimum combination for fabrication o f a sodium-to-air radiatoi. The evaluation of these alloys was based on metallographic examination of tube-to-fin specimens before and after exposure to s t a t i c air for periods up to 600 hr. T h e alloys tested are l i s t e d i n Table 7.3, TABLE 7.3.

BRAZING ALLOYS FOR

HIGH-CONDUCTIVITY CLAD COPPER FBNS

Alloy Name ..........

Composition (wt %I ~....~ __ ~~

(Or=) ..... .

Coast Metals alloy

89 Ni, 5 Si, 4 €3,

1840

Electroless nickel

88 Ni, 12 P

1800

Ni-P-Cr

8 0 Ni, 10 P, 10 Cr

1800

copper, appears to be due to the formation of a complex eutectic resulting from the presence o f an iron-base cladding material rather than from a

Siini Iar intergranular penenickel-base material. tration was found when the 88% Ni-12% P a l l o y

Saazing

1920

Nicrobraz (LMNB)

penetration, undetected in the case of Inconel-clad

UNCLASSIFIED

Temperature

80 Ni, 6 Fe, 5 Cr,

Low-melting-point

however, and Q number of brazed joints were evaluated to determine the optimum a l l o y for brazing each of these materials to Inconel tubing. A joint of type-310 stainless-steel-clad copper brazed to lnconel tubing w i t h the 88% Ni-12% P alloy i s shown i n Fig. 7.7. It may be seen that f i n dilution and intergranular penetration o f the resulting This brazing alloy into the tube wall occurred.

.-~.

0.02

5 Si, 3 B, 1 C

2 Fe

A microsection i s shown i n Fig. 7.6 of an lnconelclad copper fin (2 mils of lnconel on each side o f 6-mil-thick copper) ioined to 3/,6-in.-OD, 0.025-in.-

w a l l Inconel tubing with 88% Ni-12% P alloy and then exposed to static air a t 1500°F for 600 hr. Solution o f the tube w a l l during the brazing c y c l e was negligible, but rather severe solution of the It i s evident, however, that the f i n l i p occurred. oxidation resistance of the joint was not affected. Examinations of specimens brazed w i t h the other alloys l i s t e d i n Table 7.3 revealed similar results, that is, minimumsolution o f tube w a l l s and excellent oxidation resistance of the braze joints. T h e effects o f diffusion, however, were in evidence after each test. Color changes i n the copper core, along w i t h the formation of voids, indicate that the use of Inconel-clad copper i n a radiator would not be advisable, since the void formation

118

F i g , 7.6, Inconel-Clad Copper Fins Brazed t o Inconel Tubing with 88% Ni-12% P Alloy and Exposed t o %tic A i r a t 1500° for BOO hr. Note

d i l u t i o n of fin, hole formation in copper due t o diffusion, and the excellent oxidation resistance of the bronze joint. As polished. 50X. Reduced

1.5%.

PERIOD ENDING SEPTEMBER

was used w i t h type-430 stainless-steel-clad copper,

os shown i n Fig. 7.8.

In t h i s case, the intergranular

penetration was completely through the tube wall.

flow point cf

1925째F

?O, 1954

i s so close to the melting

point of copper that it cannot be used w i t h ease. Since the oxidation resistance of the Coast Metals alloy 52 is comparable to that o f LMNB and the

T h e e f f e c t oaf the intergranular penetration on the physical properties of the tubing has not yet been determined, but it is believed that it would be advisable to use a brazing alloy that does not react w i t h the iron-base cladding material.

effect o f boron diffusion i s the s a m e for both alloys, the Coast Metals alloy was selected for the fabrication of the f i r s t high-conductivity-fin sodium-to-air

The tube-to-fin joints shown i n Figs. 7,9 and are, respectively, type-310 stainless-steel -clad

T h e problem of edge protection of the clad copper fins has been resolved. Aluminum is applied to the

and type-430 stainless-steel-clad copper fins that

edges o f f i n s by dip coating or spraying, or it i s painted on as aluminum powder in (I Nicrobraa

7.10

were brazed ivith Coast Metals alloy 52 to lnconel and then exposed t o static a i r for 400 hr at 1500째F.

A negligible amount o f d i l u t i o n and a r e l a t i v e l y

minor amount of boron diffusion into the lnconel tubing occurred. Although low-melting-point Nicrobraz

(LMNB)has

been used successfully to braze a l l the proposed cladding materials t o Inconel,

i t s r e l a t i v e l y high

F i g . 7.7. Type-310 Stainless-Steel-Clad Copper f i n s Brazed t o Inconel T u b i n g w i t h 88% M1-12% P Alloy. Note d i l u t i o n of fin and intergranular penet r a t i o n i n t o tube wall. As pofished. SOX. Reduced 15%.

radiator.

cement slurry. approximately

The fin i s then heated at 80OoC for T h e diffusion of the aluminum

'/2 hr.

into the copper core results i n an aluminum bronze which exhibits excellent resistance to oxidation when exposed to static air for 500 hr a t 1500째F. It appears that the depth o f bronzing, which con3 tinues during oxidation testing, i s about /32 to in. i n 500 hr. The feasibility of simultaneously bronzing large numbws of fins i s to be determined.

F i g . 7.8. Type-430 Stainless-Steel-Clad Copper F i n s Brazed to lnconel Tubing w i t h 88% Ni-I2% P Alloy. Note dilution of t h e f i n and severe intergranular penetration into the tube wall. ished.

50X. Reduced 19%.

A5 pol-

119

ANP QUARTERLY P R O G R E S S R E P O R T

UNCLASSIFIE D Y -12601

0.01

0.0 4

0.02

0 02

Lo 0

0.03

0 03

0.04 -

004...

0.05

0 05

I _

O.Ot

._.._I

0 06

0.0;

0_ 07 _..II

z -

0.01

0 08

F i g , 7,9. Type-310 Stainless-Steel-Clad Copper Brazed t o Inconazl Tubing with Coast Metals Alloy 52 and Exposed t o Static Air a t 1500OF for 400 hr. A s polished. 50X. Reduced 16%. Radiator Fabrication

Patriarza

G. EA. Slaughter

Metallurgy D i v i s i o n

Ai r c m f t

J. M.

Cisar

Reactor Engineering D i v i s i o n

A sodium-ta-air radiator witti 6 in. o f type-430 stainless-steel-clad copper f i n s was fabricated by using a combination heliarc welding and brazing procedure. The tube-to-fin section was assembled

and brazed w i t h Coast Metals alloy 52 at 1020째C. The s p l i t headers ,were then assembled and heliarc welded by using the seniiautomatic equipment available i n t h e Welding Laboratory. A photograph of the radiator arid welding torch after the welding of the 108 tube-to-header joints i s shown i n Fig.

F i3. 7. IO. T y p e 4 3 0 Stai aal es s-Steel-CI ad Copper B r a ~ e dw i t h Coast Metals Alloy 52 and Exposed t o Static Air a t 1 5 O Q O F for 400 kr. A s polished, 5QX. Reduced 17.5%. back-brazing operation satisfactorily sealed the leaks o n d r e m o ~ e dthe effects of notches as sources of stress concentrations. The completed unit, as shown i n Fig. 7.12, was helium leak-tight after the rebrazing operation. This u n i t i s to be used in the 100-kw g os- f ired 1 iqui d-metal- heating sys tern. N O Z Z L E S FOR THE G A S - F I R E D I IQUID-ME T A L - H E A T E R SYSTEM

K. W.

P. Patrinrca G. M. Sluughter Reber Metallurgy D i v i s i o n

J. M. Cisar Aircraft Reactor Engineering D i v i s i o n

7.1 1.

One of the design features of the 100-kw gas-fired, liquid-metal-heater system now being constructed i s the packed-rod nozzle assembly to be used for inlet control o f the air and the gas flow. Test data a i d

The remaining header sections were then manually heliarc welded by u t i l i z i n g the complete penetrafion technique. Although a pressure t e s t indicated that s i x tube-to-header joints were not leak-tight, a

theoretical considerations indicated that &-in.-dia stainless steel triangularly packed sods would be satisfactory for use in t h i s nozzle. It was expected that these rods could be joined i n a r i g i d bundle

120

PERIOD ENDING SEPTEMBER 70, 7954 U)JCLASSI F ItD Y 12916

Fig. 7.1 1. Sodium-to-Air Radiator Showing Semiautomatic He1iarc Tube-to-Header Welds. w i t h a minimum of plugged interstices by furnace brazing w i t h a ductile oxidation-resistant high-

temperature brazing a l l o y such as 82% Au-18% Ni. Experiments revealed that proper jigging was extremely important i n obtaining the completed airnozzle blank (Fig. 7.13). Each layer of rods was tack-welded for rigidity, and the layers were then tack-welded to form the assembly. Since the determination of the exact amount of brozing alloy required was difficult, i f not impossible, sufficient material was used to ensure good bonding at a l l capillary joints. T h e excess material wa5 then removed from the nozzle interstices by placing a flat cross section against a metal sheet on which

Fig. 7.12. Completed Sodium-to-Air Radiator Showing T u be4c-F in Mani fa1d 'Construct E on. many fine lines had been scribed. Upon rebrazing, these fine scratches octed as capillaries and the excess materia! was removed. T h e design of the natural-gas nozzle specified that a regular pattern of interstices should be

selectively closed, T h e closures were obtained by inserting wires into the designated holes and brazing them to the rods.

S P E C I A L MATERIALS FABRICATION RESEARCH

J. H. G o b s

H. lnouye

Metallurgy D i v i s i o n

Stainless-Steel-Clad Molybdenum and Columbium Two thermal-convection loops constructed o f stain 1es s-s teel-cl ad mo I y bdenum were assem bled by

121

ANP QUARTERLY PROGRESS REPORi

1041,

UNCLASSIFIED Y-12934

Inconel-Type Alloys

Eight tlierrnal-convection loops of Inconel-type alloys were fabricated. I he compositions and room temperature properties of the alloys are given i n Table 7.4. T h e alloys are being studied as a part of the effort to obtain a l l o y s w i t h better hightemperature corrosion resistance than that of Inconel. I

T A B L E 7.4. P R D P E R T I E S OF HIGH-PURITY INCON EL- T YP E Ab QQYS * Nomina I

-l e n s i l e

Yield

Composition

Strength

Point

( w t %)

(psi)

18 Cr, 7 Fe, 75 NI

85,000

41,500

10 Cr, 15 F e , 75 Ni

77,700

38,800

5 Cr, 20 ke, 75 Ni

72,700

34,200

76,000

36,000

5 Cr, 10 k e , 75 Ni, 10 Mn

Heat No,

F i g , F.13. Brcassd A i r - N o z z l e BBawk, Note bonding at the points of tangency of the stainless steel rods.

using threaded molybdenum joints, whish were then protected from oxidation by the U S C of welded cover plates. h t h loops f a i l e d early i n corrosion t e s t s hemuse of ftac:urc of the C O Y W plates and leakage f r c m the joints. I t i s believed :hot overheating of the cover plates

occurred while the loops were being i n i t i a l l y heated Since no metallurgical by o l e c t r i c u l resistance. bond existed between t h e molybdenum and the cladding, separation became inore pronounced w i t h increased tsnipernfUre, unti I fincpl l y a1 I the current was being ctxrried by iha rather t h i n COVE?' plates, Future molybdcnum Ptre;inol-convection loops w i l l which i s weldable, and the heavy nickel electrodes required f o i resistance hcating will be welded directly to the molybdenum tuhc instead of to the cladding.

be made from arc-melted molybdenum,

Columbium clad w i t h type 310 stainless steel has been fabricated into a theriiicrl-convection loop by heliarc wc!dli?g i n air. l h i s lsop w i l l be tested as soon as thc accessory parts are available.

122

G, 7 Fe, 83 NI

E longati on

(%I

* h t a obtained from t h e Superior T u b e Company.

Nick91-Molybdenurn-Bnse Alloys

The nickel-molybdenum-base a l l o y s are a l s o b e i n g studied as port of the effort to obtain a l l o y s w i t h better high-temperature strength and corrosion resistance than are offered by Inconel. These requirements are met by Hastelloy B; however, some d i f f i c u l t i e s have been encountered i n attempts to fabricate t h i s material. In castings and i n fusion welds, o whole range of composition results because of severe coring during freezing, and thus the properties of the fabricated specimen vary from one area to another. Even long-time high-temperature aging trentinents of such material w i l l not produce on equilibrium structure. The aging treatment also results i n a great increase i n hardness w i t h a corresponding decrease i n ductility. At temperatures between 1200 and 1800"F, the wrought material exhibits i t s poorest d u c t i l i t y because of hot shortness, which cun arise from trace impurities or the precipitation of an age-hardening constituent, which Thus the i s probably the case for t-lastelloy 8. temperature range of poorest d u c t i l i t y of Hastelloy R coincides w i t h the temperature range o f contemplated u s e i n high-temperature circu lating-fuel reactors. Hastelloy B i s not recommended for use i n a i r at temperatures above 1400째F becaose of

PERIOD ENDING SEPTEMBER 70, 1954 excessive oxidation. A potential source of trouble exists because of the molybdenum content; molybdenum oxide has been reported to cause catastrophic oxidation of surrounding structures, especially the

steels. The solutions to these problems are being approached i n two major ways. First, attempts are being made t o find suitable melting and heating treatments of Hastelloy 5 that w i l l reduce the degree o f age hardening so that both the cast and wrought materials wi I I possess acceptable properties i n the temperature range o f interest. Second, Hastelloy-type alloys are being studied. Both phases of the investigation w i l l involve attempts to extrude seamless tubing, since, a t present, Hastelloy

B tubing must be made b y welding strips. Aging studies o f the following materials are now i n ptogress at temperatures between 1200 and 1700°F:

commercial, wrought, Hastelloy 5 plate; vacuummelted Hastelloy B, as cast; cast and wrought 20% M0-80% Ni alloy; wrought 24% Mo-76% Ni alloy, Some of the properties of the 20% Mo-80% Ni alloy already obtained indicate that t h i s alloy w i l l be easier to fabricate than Hastelloy B from an extrusion standpoint; sound tubing could be made, whereas none was obtained from Hastelloy B. T h e recrystallization temperature of a 6Wo coldworked sheet o f the 20% Mo-80% Ni a l l o y was ap-

proximately 1900°F for an annealing time of 30 min. However, t h i s alloy oxidizes in a i r a i 1500°F at a faster rate than does Hastelloy 5, although

under a temperature c y c l e between room temperature and 1500°F the oxidation rates are similar. T h e A scale formed s p a l l s near room temperature. sheet, 0.065 in. thick, could be heiiarc welded without producing porosity, but an argon arc produced porosity. I n strength tests at 150O0F, the 20% M0-80% Ni a l l a y had a rupture l i f e of 90 hr a t 8000 psi; when tested a t SO00 p s i for 600 hr, it had a 3% elongation; during a short-time test the a l l o y had a t e n s i l e strength of 40,000 p s i and an elongation of 10% in a 2-in. gage length. Oxidation and oxidation-protection studies are in progress to find suitable methods o f protecting the nickel-molybdenum-base a l l o y s during tests i n thermal-convection loops and creep-testing apparatus. T h i s work w i l l be the ground work for further studies of the protection of reactor components. The coatings being studied are those that can be applied by methods readily available, that is, chromium electroplate, chromium p l u s nickel electroplate, nickel

p l u s aluminum spray, and aluminum spray. In constant-temperature tests, a chromium electroplate was found to be beneficial. In oxidation tests involving c y c l i n g between room temperature and 1500"F, these coatings eliminate the spalling characteristics of Hastelloy B. The chromium electroplate and the chromium p l u s n i c k e l electroplate afford the best protection. Duplex Tubing Attempts have been made t o fabricate duplex seamless tubing that wi II have good corrosion resistance on the inner surface and oxidation resistance on the outer surface.

In the f i r s t ex-

periments, attempts were made to deep draw duplex blanks 4 in. in diameter. The blanks were made by hot pressing and a l s o by hot rolling, The following materials were combined by hot pressing 0.025-in.thick sheets under an argon atmosphere w i t h Al,O,coated graphite dies at

1500 psi.

Pressing

Te rnperature

("a

310 s t a i n l e s s

1000

Inconel plus type 310 s t a i n l e s s

1200

Copper plus type steel

steel Inconel p l u s H a s t e l l o y Hastelloy

B plus

type

310 s t a i n l e s s

1200

steel

Copper was also combined w i t h type 310 stainless steel by hot r o l l i n g a t 1000°C. F a r this, a three-ply composite w i t h capper in t h e center and cover plates of the stainless steel was used. One o f t h e cover plates was oxidized at 1100°C for 2 hr t o prevent bonding during rolling. Included i n these experiments were some tests i n v o l v i n g the fabrication of composites o f molybdenum and Inconel. T h e purposes of these tests were t o determine the conditions necessary t o obtain a metallurgical bond b y hot rolling, to determine t h e extent of directional properties due io the molybdenum, and t o determine whether such a combination could be deep drawn. T h e f i r s t r o l l i n g experiments were made on an evacuated capsule at 1225OC w i t h reductions of 10% per pass. The total reduction was 79% i n thickness. No bonding was obtained b y using n i c k e l as an intermediate layer between the

123

ANP QUARTERLY PROGRESS REPORT lnconel and the molybdenum, l n a second series o f experiments the r o l l i n g temperature was reduced to lOOO"e3, with reductions o f 40% per pass and a

T h e density o f the proposed shield pieces i s t o equivalent to 1.4 be about 1.8 to 1.9 g/cm3 of R,C,

m e t a l l i c compounds to dorm at the interface. Bonding was achieved by using nickel as the intermediate

an integral number of pieces. Additional tests of the compatibil ity of lnconel w i t h pure boron carbide are being made. Previous tests indicated the formation o f a diffusion layer

The lower total reduction of 85% i n thickness. r o l l i n g temperature reduced the tendency for inter-

layer, and evidences of intermetallic compounds were found only by metallographic examination at high magnifications. The find thickness of t h e composite w a s 0,010 in. o f molybdenum on 0.040 in.

of Inconel. In future experiments w i t h t h i s combination, the maximum ratio o f molybdenum to Inconel wi tkout the directional properties o f wrought molybdenum wi II be determined,

All

the composites mentioned above were successfully reduced from 4-in. blanks to 21/2-in. cups by the "Guerin" forming process (male parts of d i e are hard rubber) at room temperature. Attempts to deep draw the composites are to be made.

ma- F'ho se A! lays

For the continuing study of sigma-phase alloy corrosion by l i q u i d lead (cf. Sec, 6, "Corrosion Research"), some additional tubing was made. Two ingots o f a 48.2% Cs-51.8% F e alloy and two ingots of o 45% Fe-40% G-1% Ni a l l o y were vacwum cast into l>,-in.-dia ingots 6 in, long. These ingots were then hot r o l l e d i n an air atmosphere at 1250째C to rod in. i n diameter. After annealing at 11Qo"C for 2 hr and quenching, the rod was readily machinable into tubing. T h e transformation from the ferrite to the sigma phase w i l l be accomplished b y aging at 1425째F. Two of the tubes (Fe-Cr) were cold swaged 12% so that the effect of cold swaging on the distribution o f the sigma phase could b e determined,

oron

Carbide Shielding

The experimental fabrication o f b-in.-thick shield pieces by warm pressing boron carbide bonded w i t h copper or silver has been discontinued i n favor of t h e fabrication of pieces molded with nonmetallic bonding material b y cold pressing followed b y sintering t o develop the desired properties. Nonm e t a l l i c bonding materials that would be suitable from nuclear considerations and would be stable at the expected service temperature include sodium silicate, silica, s i l i c o n nitride, and boric oxide. Arrangements are being made to secure samples o f these materials for compatibility t e s t s with lncanel under operating cot~ditions,

124

to 1.5 g/crn3 of boron. They would be molded i n t h e form o f equilateral triangles or diamond shapes, 2 to 3 in. on a side, so as fa cover a sphere with

5 m i l s thick i n ature of

a 100-hr test at t h e operating temper-

1500째F.

Twelve tubular fuel elements are being prepared for further drawing experiments at Superior Tube Company. 'they are being assembled with highfired UO, inathe 30- to 5 5 - p p a r t i c l e s i z e range, as

well as w i t h very fine UO,, and w i t h prealloyed and elemental stainless steel and iron core matrixes. Plans are cilso being made to t r y hot swaging o f the tubes i n an attempt to reduce stringer formation i n t h e cores by reducing the amount o f cold working. H o t swaging on a mandrel may prove to be successful i f bonding to the mandrel can be prevented by an oxide scale or coating.

M E T A L L O G R A P H I C E X A M I N A P I O H OF A FL UO R i D E - Y O - S O D I U M H E A T E XCMWNG E R

R. 9. Gray

P. Patriarca

G. M. Slaughter Meta I lurgy D i v i s i o n

T h e fluoride-to-sodium intermediate heat e x changer, which i n a l i f e test failed after 1680 hr i n c y c l i c service i n the temperature range 1080 t o

1500"F, was examined metaIlographically. All the

tube-to-header joints were manually heliarc welded, and tests showed the heat exchanger to be heliuni leak-tight before assembly into the test rig. Visual examination indicated the presence of 19 fissures i n the tube-to-headcr welds i n two of the three adjacent headers a t the hot sodium inlet end. T h e probable progressive propagotinn of these fissures i s shown in Figs. 7.14, 7.15, and 7,16. T h e fissure shown i n F i g e 7.16 probably extends to the surface of the weld. If so, there BWS ~1 leak i n the system at t h i s point. It seems l i k e l y that differential thermal expansion betwcrn t h e tubes and the casing caused stress concentrations at the roots of the tube-to-header welds. These stress concentrations wou Id tend to propagate cracks through the w e l d s i n the course of thermnl cycling,

fig* 7.14. Tube-to-Header Joint Without Crock Extending into the Weld. Etched. 1OOX.

fig.

7.15. Tube-$-Header

Joint

witha Slight Crack

Extending into the

Reduced 16%.

Weld. Etched. IOOX. Reduced 16%.

125

ANP QUARTERLY PROGRESS REPORT

Fig. 7.16.

Tub-to-Header Joint Showing Severe Cracking.

particularly since the columnar dendrites, which arc typical of a weld structure, are aligned i n such a way as to a i d parallel fractures. T h i s investigation emphasizes the extreme des i r a b i l i t y o f using bock-brazing as n means for minimizing the notch effect and for reducing the p o s s i b i l i t i e s of leaks developing i n the system.

126

Etched.

lOOX,

Corrosion tests in static sodium and i n fluoride mixtwrcs indicate that a 67% Ni-13% Ge-11% Cr-6% Si-?% Fe-1% Mn alloy niay be useful for t h i s application, T h i s alloy, which flows w e l l at 2050"F, was found to be v i r t u a l l y unattacked i n the fluoride fuel arid corroded to a maximum of 4 m i l s i n 100 hr at 1500째F in sodium.

PERIOD ENDING SEPTEMBER 70, 1954

8. HEAT TRANSFER AND PHYSICAL PROPERTIES H. F. Poppendiek Reactor Experimental Engineering D i v i si on T h e enthalpies and heat capacities of NaF-ZrF,(65-15-20 mole %) were determined; the heat capacity i n the solid state over the temperature range 90 to 614°C WQS found t o be 0.17 cal/g."C, and the heat capacity i n the l i q u i d state over the temperature range 653 t o 924°C was found to be 0.20 cal/g."C. T h e enthalpies and heat capacities of LiF-NoF-UF4 (57.6-38.4-4.0 mole %) were also

P H Y S I C A L P R O P E R T I E S MEASUREMENTS

UF,

obtained; the heat capacity i n the s o l i d state over t h e temperature range 97 t o 594°C was found to b e 0.23 cal/g.OC, and i n the l i q u i d state over the range 655 t o 916°C it was found t o be 0.53 cal/geoC. T h e thermal conductivities of three fluoride mixtures i n the solid state at normal temperatures were determined. T h e conductivity o f N a F - K F - U F 4 (46.5-26.0-27.5 mole %) was 0.7 B t u / h r s f t 2 ("F/ft), that for KF-LiF - N a F - U F 4 (43.544.5-10.9-1.1 mole %) was 2.0 Btu/hr.ft* ("F/ft), and that for LiF-KF-UF, (48.0-48.0-4.0 mole %) was 1.4 Btu/hnr.ft2 (OF/ft). A new electrical cond u c t i v i t y device has been constructed and has been successfully checked with molten salts of known conductivity. Additional forced-convection heat transfer measurements of molten N a F - K F - L i F (1 1.5-42.0-46.5 mole %) have been made. Measured thermal con-

d u c t i v i t i e s and thicknesses of insoluble deposits on the inner walls of lnconel heat transfer tubes made i t possible to calculate the thermal resistance of the deposit. The calculated value was

i n good agreement with values previously deduced A device from the heat transfer measurements, for studying the rates of growth of tube-wall dep o s i t s has been successfully tested w i t h a simple A hydrodynamic f l o w system heat transfer salt. t o be used for studying the reflector-moderated reactor f l o w structure has been tested. A mathematical study of the temperature structure i n converging and diverging channel systems that duct f l u i d s w i t h volume heat sources has been made. A study of wall-cooling requirements i n circulating-fuel reactors was made.

'W. D. P o w e r s and G. C. Blalock, Meut C a p u c z t z p s o/ Conipositims No. 39 and 101. O R N L CF-54-8-135 ( t o be issued).

Heat Capacity

W. D. Powers Reactor Experimental Engineering D i v i s i o n T h e enthalpies and heat capacities of two fluor i d e mixtures have been determined w i t h Bunsen i ce ca Ior imet ers :

(65-15-20 mole 70)

NaF-ZrF4-UF, Solid

(90 t o 614°C)

- 3 + 0.17T C p = 0.17 5 0.01 L i q u i d (653 t o 924°C) H T - H o g C = 22 + 0.20T cp = 0.20 5 0.02 LiF-NaF-UF, (57.6-38.4-4.0 mole %) Solid (97 t o 594°C) H T - Hoot -= G.22(7)T 4 0.0G017T2 C p = 0.22(7) + 0.000331' L i q u i d (655 t o 976°C) H r - H O o C = -68 + 0.53T C p - 0.53 -t 0.04 [IT

fiOoC =

I n these expressions H i s the enthalpy i n cal/g, i s the heat capacity i n cal/g."C, and T i s the P temperature i n "C. The experimental enthalpy data for L i F - N a F - U F 4 (57.6-38.4-4.0 mole %) are plotted in Fig. 8.1. Automatic Simplytrol u n i t s have been added t o the calorimeter systems t o more accurately control t h e furnace temperatures. T h i s modification has increased the precision of the experimental data. Currently, enthalpies of a fluoride mixture are being measured w i t h the copper block calorimeter.

Density and V i s c o s i t y

S. I. Cohen Reactor Experimental Engineering D i v i s i o n Prel iminary density measurements were made on rubidium metal, and the density was found t o be represented by the equation p(g/cm3)

1.52 - 0.00054 (T

39°C) , 127

ANP QUARTERLY PROGRESS R€PORT

450

400

350

300

150

to0

0 200

400

600

TEMPERATURE

Fig. 8.1.

calculated value for the l i q u i d density at the melting point, given i n the Liquid Metals [fundbook,2 f a l l s w i t h i n 3% of the value yielded by t h i s study. T h i s experiment was carried out at the request of the QRSORT group currently studying the use of rubidium vapor in an aircraft reactor cycle. Measurements were mode on a sample of metal 'R.

June

128

1000

Enthalpy vs Temperature for biF-NaF-UF, (51.6-38.4-4.8mole %IO).

where T i s the l i q u i d temperature i n "C. MeasureA ments were made by the buoyancy principle.

N. L y o n (ed), Liquid Metals Handbook, p 52, of the N a v y , N A V E X O S P-733 (rev), 1952.

AEC-Department

800

("GI

produced by the Stable Isotope Research and Production D i v i s i o n by reduction of a high quality fluoride which was prepared by the QRSORT group from RbF furnished by t h e Materials Chemistry D i v i s ion. In order t o obtain more accurate v i s c o s i t y data it was decided t o i n i t i a t e a program of refining viscometry techniques. In particular, the influence of f l u i d density on calibrations of bath the r o tational and the capillary devices i s being studied. E a s i l y handled fused salts are being uSed a s calibrating fluids. The present studies indicate that some of the old calibration curves, which were

PERIOD ENDING SEPTEMBER 70, 1954 based

low-density calibrating

liquids,

were

shifted by about 30%. T h i s means that some of the previously reported v i s c o s i t y measurements are about 30% high. Because of significant improvements made i n fluoride preparation and handling, the mixtures which are now being obtained from the Materials Chemistry D i v i s i o n are extremely pure, and they no longer cause the d i f f i c u l t i e s encountered in the early stages of the project, such as fouling of equipment and variations i n physical characterNo doubt some of the earlier v i s c o s i t y istics. measurements were high because of the impurities i n the materials.

Thermal Conductivity

D. Powers

Reactor Experimental Engineering D i v i s i o n T h e thermal conductivities of three solid fluoride mixtures at normal temperatures were determined by the transient cooling technique:

of polarization was further substantiated by the small observable increase of conductivity w i t h frequency i n contrast t o the usual large dependence of conductivity upon frequency, a s observed i n the low-resistance conductivity c e l l . The measurements on KNO, were obtained during a single run without the necessity of having t o replatinize the electrodes after each determination, A refinement of t h i s current-potential c e l l which w i l l permit measurements of greater accuracy i s now under way. Preliminary measurements of several fluorides, the conductivities of which were determined previously by an alternate method, were found to be in good agreement within the inherent limitations of accuracy. A summary of the preliminary nieasurenients of the conductivities of K F - L i F - N a F - U F 4 (43.5-44.5-10.9-1.1 mole %), NaF-ZrF,-UF, (50.046.0-4,O mole %), and N a F - Z r F 4 - U F 4 (53.5-40.06.5 mole %), as w e l l as the measurements of several other salts, has been compiled in a separate report. 4

T hemal Conductivity

Vapor Pressure

[E+u/hr.ft2 ( O F / f t ) l

NoF- KF-UF4 (46.5-24.0-27.5 mole

K F -L i f -N o F - U F , (43.5-44.510.9-1.1 miole %)

2.0

L i F-KF-UF4 (48.0-48.0-4.0

1.4

m a l e %)

thermal conductivity of s o l i d LiF-KF-UF, mole %) was measured by the steadystate flat-plate technique; a value of 1.5 B t u h r - f t ’ (OF/ft) was obtained. T h e ratios of corresponding I iquid t o solid thermal conductivity measurements for the fluoride mixtures studied t o date are near unity. The

(48.0-48.0-4.0

Electrical Conductivity

N. D.

R. E. Moore Materials Chemi stry D i v i sion

0.7

Greene

Reactor Experimental Engineering D i v i s i o n T h e experimental current-potenti a I conductivity c e l l has been successfully tested and standardized. The electrical conductivity of a KNO, melt was determined to w i t h i n a few per cent of the values reported in the literature. As predicted, the effects of polarization w i t h i n t h i s c e l l were greatly reduced by the use of the alternote m e t h ~ d . ~T h i s reduction in the effect

Vapor pressure measurements on the mixture N a F - Z r F 4 (25-75mole %), which were begun last q ~ a r t e r , have ~ now been completed. T h e method and apparatus, originally described by Rodebush and Dixon,‘ were discussed i n previous reports.7fB T h e data, given i n Table 8.1, can be represented by the equation

l o g l o P(mm Wg) =

9368

-- - t 10.57 , T ( O K)

from which the calculated pressures and the heat of vaporization (43 kcal/moie) were obtained. T h e approximate

I iquidus temperature,

obtained from

3N. D. Greene, ANP guar, Prug. Rep. J u n e 10, 1454, ORNL-1729, p 100, ‘N. D. Greene, M e a s u r ~ r i i c n t s vf the Electrzcnl Conducttvzty of Molten Fluorzdrs, O R N L CF-54-8-64(to be issued).

5R. E. Moore a n d C. J. Barton, A N P Quar. Prog. R e p . June 10, 1954, ORNL-1729, p 101.

6W. H. Rodebush a n d A. L. D i x o n , P h y s . R e v . 26, 851

(1925). ’R. E.

Moore and C. J. B a r t o n , ANP Quar. F r o g . Rep. Sept. 10, 1 9 5 1 , ORNL-1154, p 136.

‘R. E. Moore, A N P Qunr. Prog. Rep. D e c . 10, 1951, ORNL-1170, p 126.

129

ANP QUARTERLY PROGRESS REPORT T A B L E 8.1. _

VAPOR PRESSURE IN NaF-ZrF4 (25-75 m o l e W )SYSTEM _ ~ ~

T e t i l pe ra tu re

("C)

Ob s e r v e d

Calculated

Pressure

Pressure ( m m i-lg)

(mnl

ng)

754

769

7 76

786

796

805

811

828

115

844

152

155

the intersection of the vapor pressure curve w i t h that of pure ZrF,, i s about 760째C. FUSED-SALT H E A T T R A N S F E R

ti. W. Hoffman Reactor Experimental Engineering D i v i sion Additional data for N a F - K F - L i F ( 1 1.5-42.0-46.5 mole 75) flawing for short-time periods i n an elect r i c a l l y heated type 316 stainless steel tube have been analyzed. The results are i n agreement w i t h the general turbulent f l o w correlations for ordinary fluids. The thermal conductivity of K,CrF, was found t o be 0.13 Btu/hr.ft2 ("F/ft)9 in the temperature range 100 t o 2000F. Measurements from a photomicrograph, Fig. 8.2, of a section of the tube used i n the lnconel experiment show a f i l m thickness of about 0.4 mil. The thermal resistance calculated from these results, 0.00025 hr-ft'." F/Btu, i s i n agreement w i t h the measured thermal resistance obtained from the heat transfer experiments i n the lnconel system. T h e forced-convection apparatus," which i s t o be used for studying the rate of growth of w a l l deposits, as well a s for heat transfer meaiuremerits of fused fluoride mixtures, has been operated successfully for more than 100 hr w i t h the heat W. Hoffman and J. Lones, A N P Qunr. f r o g . R e p . J u n e 10, 1 9 5 4 , ORNL-1729, p 101. "1-1. W . Hoffman a n d J . Lones, A N P Quar. f r o g . R e p . Mai-. 1 0 , 1954, ORNL-1692, p 98. 'H.

130

FILM, 0 4 mil

Fig.

8.2.

K,CKF,.

incone!

Tube with F i l m

Deposit of

transfer s a l t NaN0,-NaN0,-KNQ, (40-7-53 wt %) as the circulated fluid. The system i s now being modified so that resistance-heated, as w e l l as externally heated, test sections can be inserted i n t h e system. With resistance-heated sections, larger axial f l u i d temperature r i s e s can be a t tained. It i s planned t o study next a zirconiumuranium fluoride salt mixture in Inconel.

TRANSIENT BOlLlNG RESEARCH

M. W. Rosenthal Reactor Exper irnenta I Eng ineer i ng Biv i s i on A study of the heat transfer phenomena which occur when a water-cooled and -moderated solidf ue I-element reactor sudden ly becomes super c r i t i c a l has been undertaken. The time required for generotion of steam in water which i s i n i t i a l l y below the b o i l i n g point w i l l be determined. The effects of water temperature, rate of power increase, air concentration, and surface condition w i l l be investigated.

Construction of the equipment for generation of power surges by electrical means in aluminum f i l a ments has been completed, and the entire experimental apparatus has been assembled and tested. Satisdactory operation has been achieved. The transient-power generator consists of 100 t h y A different b i a s ratrons connected i n parallel. can be put on the g r i d of each thyratron, and the system can he set so that a signal of increasing voltage applied simultaneously t o a l l tubes fires them i n a sequence that approximates the desired function. A peak current of about 500 amp can be produced, w i t h the duration of the power surge ranging from 30 t o 500 msec. T h e instantaneous values of voltage drop across the test filament and current through it are recorded by n fast-response Hathaway oscillograph. A Fastax high-speed motion picture camera i s

PERIOD ENDING SEPTEMBER JO, 1954 used t o photograph the event, and illumination i s supplied by an Edyerton stroboscopic flash u n i t A repetition rate synchronized w i t h the camera. of 6000 tranies per second i s obtainable w i t h a f l a s h duration of 2 psec. C l o s i n g a single s w i t c h i n i t i a t e s the operation of a l l the equipment; a time delay i s provided in which the camera can accelerate t o full speed before current i s passed through the filament.

F L U I D F L O W STUDIES FOR C I R C U L A T I N G F U E L REACTORS

J. 0.

Bradfute

L. D.

F. E. L y n c h

Palmer

Reactor Exper imental Engineering D i v i s i o n T h e flow system for the hydradynamic studies of t h e ref lector-moderated reactor core, shown schematically in Fig. 8.3, has been fabricated, assembled, and tested. A detailed sketch of the t e s t section i s also shown i n Fig. 8.3. Since it w i l l b e necessary to insert, photograph, and then remove a coordinate g r i d from time t o time, t h e system was designed so that it could be readily assembled and disassembled. The cores are POsitioned and supported by a brass sleeve, which was attached t o the t e s t section support along i t s center l i n e before it was machined. The sleeve was carefully bored and lapped so that the shafts protruding downward from the cores would f i t w i t h minimum clearance and yet loosely enough so that they could be removed by hand. The upper surface of the t e s t section support flange was machined perpendicular t o t h e center l i n e of the brass sleeve. T h i s arrangement permits accurate alignment of the core w i t h i n the P l e x i g l a s test section without supporting members in or ahead of the t e s t region. Attempts are currently being made t o obtain some preliminary f l u i d f l o w information for the reflector-moderated reactor core for the straight entrance condition shown in Fig. 8.3 by photographing particle f l o w through the P l e x i g l a s t e s t sect ion. T h e phosphorescent flow-visual ization system, w h i c h was described previously,’’ has been further modified. The ZnCd sulfide phosphor has been replaced by a ZnS (Cu-activated) phosphor whose luminosity was found t o be greater after excitation. T h e duct work in the f l o w c i r c u i t has been im-

“ L . D. Palmer and G. M. Winn, A F e a s i b i l i t y Study o/ € l o w Visualization U s z n g a Phospharescent P a r t i c l e M e t h o d , .ORNL CF-54-4-205 (Apri I 30, 1954),

proved so that higher Reynolds numbers are now attainable, and a diverging L u c i t e channel has been constructed. Some of the flow features, such a s asymmetry, f l o w separation, and the effect of entrance lengths, are t o be observed for different entrance configurations in such channel systems. T h e phosphorescent flow-visualization method i s a l s o t o be used i n the reflector-moderated reactor core experiment. H E A T T R A N S F E R S T U D I E S FOR C I R C U L A T I N G -

FUEL REACTORS

ti. F. Poppendiek L. DoPalmer Reactor Experimental Engineering D i v i s i o n The c l a s s i c a l hydrodynamics study, conducted by Nikuradse,” of converging and diverging channels was reviewed. It i s f e l t that some of t h e fundamental f l o w features observed i n converging and diverging channel systems v y i l l a l s o The e x i s t in the ref lector-moderated reactor. r e s u l t s obtained by Nikuradse indicate that for the case of turbulent flow in convergent channels t h e flow becomes more stable or l e s s turbulent i n nature and that the eddy d i f f u s i v i t i e s are s i g n i f i c a n t l y lower than in a parallel-plate channel system. For the case o f turbulent, symmetrical f l o w i n diverging channels the t l o w i s tound t o be more turbulent than in the parallel-plate system. When the total channel angle of a divergent system is increased beyond 8 deg, asymmetrical f l o w results. For a 10-deg angle, a thin, low-velocity layer of fluid flowing i n a reverse direction t o the main stream i s noted near one of the channel w a l l s . T h i s separation phenomenon becomes more pronounced a t larger channel angles. Some radial temperature distributions were calculated i n the established-flow region of a few converging and diverging f l o w channel systems cont a i n i n g volume heat sources w i t h i n the ducted fluid. The experimental velocity and eddy diff u s i v i t y data obtained by Nikuradse were used i n the analyses. The r e s u l t s indicated that w a l l - f l u i d temperature differences con be significantly higher i n converging flow channels than i n parallel-plate systems; t h e converse i s true for symmetrical diverging #low channels. Asymmetrical divergent f l o w passages are d i f f i c u l t t o analyze a t t h i s t i m e because the d i f f u s i v i t y structure i s not known. 12J.

(1929).

Nikuradse,

Forschung.sarbaiten V D I 289, 1-49

131

ANP

132

PERIOD ENDING SEPTEMBER 70, 1954 However, it appears reasonable t o conclude that high temperatures probably would e x i s t i n the thin, low-velocity layer of fluid i n the 10-deg divergent

channel because the flow there i s probably laminor

i n nature. A report i s being prepared which w i l l be useful t o designers for predicting wall-cooling requirements i n circulating-fuel reactors which have pipe, parallel-plate, or annular fuel-duct geometries. The temperature solutions, which were presented

p r e ~ i o ~ s l y ,have â&#x20AC;&#x2122; ~ ~ been ~ ~ tabulated i n detail so

that the superposition process outlined i n those reports can be quickly effected. Cooling requirements and resulting f l u i d temperature distributions have been determined for t y p i c a l circuloting-fuel reactor systems 13H. F. Poppendiek and L. D. Palmer, Forced Convection Heat Transjeer in P t e s with Volume fleut Sources Wzthrn thp Fluids, ORNE-1395 (Nov. 5, 1953). 14H. F. Poppendiek and L. D. Palmer, F o r c ~ dConv e c t z o n Heat Transfer B e t w e e n Parallel Plates and zn Annuli wzth Volume Neat Sources Within the F l u i d s , ORNL-1701 (May 11, 1954).

133

ANP QUr4RTERLY PROGRESS REPORT

J. B. T r i c c Solid State D i v i s i o n A. J. Miller ANI‘ Proiect Additional irradiations of lnconel capsules containing fluoride fuels were carried out i n the MTR. Only one capsule containing a UF3-bearing fuel has been examined, and it shows no corrosion. T h i s is in contrast t o the capsules containing UF,-benring fuel i n which minor increases i n penetration have a t times been observed in the irradiated specimens as compared w i t h the out-ofp i l e controls. However, i t should be eniiphasized that even in the UF4-bearing capsules, the appearance of small deviations from the controls could be due t o experimental d i f f i c u l t i e s , Inspection of the previously described L l T R inp i l e circulating-fuel loop, which failed prior t o the startup, disclosed that the failure was caused by a break i n the weld connecting the pump discharge nipple to the fuel line. Design revisions and refabrication of some components are i n progress. Developmental work continued on a smaller loop for operation i n an LlTR A-piece. Work on a loop for insertion i n the MTR is described i n Sec. 3, I1 Experimenta I Reactor Engineering. ”

Development of equipment continued for in-pile stress-corrosion tests on lnconel in contact w i t h fluoride mixtures, and testing of the MTR tensilecreep r i g neared completion. Detailed examination w a s completed of an lnconel loop i n which sodium w a s circulated at high temperature i n the ORNL Graphite Reactor, and no evidence of radiationinduced corrosion was found i n the Inconel. Metal lographic examinntions were carried out i n the hot c e l l s on irradiated fuel plates for the Prutt & Vhitney Aircraft Division, and studies were made on anneal ing-out of f i s s ion-fragment dorncsge. Work continued on examination of wire and multipleplate type units for GE-ANP. M f R S T A T I C C O R R O S I O N TESTS

G. W. Keilholtz Solid State D i v i s i o n

W. E. Browning

A new series of capsules was irradiated i n the MTR t o compare NaF-ZrF, fuels containing LJF, w i t h those containing WF,. The property t o be compared

134

i s the effect of f i s s i o n on static eor-

rosion of lnconel by the fuels at 1500OF. The series includes f i v e capsules containing 1.79 mole 76 UF,, 49.3 mole % ZrF,, and 48.9 mole % NirF and f i v e capsules containing 1.74 mole % UF, 48.2 mole % ZrF,, and 50.1 mole 96 NaF. Analyses are being run t o confirm the purity of the

UF, w i t h respect t o UF, contamination. For every capsule, there w i l l be an out-of-pile control capsule. Each fuel generates 1100 w/cm3 i n the A-38 position in the MTR. Six capsules have been irradiated for the specif i e d two-week period. F i v e have been received

a t ORNL after irradiation; three of them have been opened and t w o have been examined -- one conBoth capsules taining UF, and one containing UF,. had been irradiated w i t h the metal-liquid interface a t 1500 k SOOF. The UF3-bearing capsule had experienced one temperature excursion to 1660 t 5OoF for less than 30 sec, but the temperature of

the UF,-beariiig capsule never exceeded 1550 5 50‘F. The fuel i n each capsule was frozen t w o or three times during its history. No evidence of corrosion or f i l m could $e seen i n the UF,-bearing Inconel capsule, There was subsurface v o i d formation i n the UF4-bearing lnconel capsule t o a depth of about 2 mils. No intergranular corrosion or concentration of voids at the grain boundaries w o s evident. A n a l y t i c a l samples of the irradiated fuel were taken w i t h two sizes of d r i l l s . Theuranium concentrations i n the central and outer samples were identical, w i t h i n experimental error (13%),and were of the value expected after burnup correction. Analyses for iron, chromium, and n i c k e l gave the following results: Fe, 0.03 t o 0.14 wt %; Ni, 0.01 t o 0.05 wt %; Cr, less than IOe4 wt %. There were no significant differences i n the final iron, nickel, and chromium compositions between the two capsules. Analyses of the unirradiated fuels have not been completed. I t would be unwise to draw conclusions on the basis of only one pair of samples; however, if these f i r s t results arc confirmed i n future capsules, [IF, fuels may be considered quite desirable as far as radiation-indiiced corrosion i s concerned.

P E R I O D ENDING SEPTEMBER IO, 7954 T w o UF3-bearing capsules and two UF4-bearing capsules i n t h i s series w i l l be irradiated for s i x weeks each t o amplify the apparent differences i n the effects of UF, and UF, on Inconel. It is interesting to note that the UF, capsule i n t h i s series i s the f i r s t one run i n the MTR and the f i r s t one run a t 800 w/cm3 or greater that has not shown a tendency toward intergranular corrosion. Whether the difference between earlier UFq-bearing capsules and the one i n t h i s series i s due t o improved temperature control or t o uranium concentration effects w i l l be determined by additional irradiations. FlSSlON P R O D U C T C O R R O S I O N

C. C.

STUDY

G. W. K e i l h o l t z Webster Solid State D i v i s i o n

Steps are being taken t o perform corrosion tests out-of-pile w i t h irradiated fuel in order t o separately study the effects on lnconel of fission-product concentration and of lnconel irradiation. A water-cooled f a c i l i t y for irradiation of s o l i d fuel was constructed and has been installed in position C-46 of the LITR. T h e unperturbed thermal f l u x i s on the order of 4 x l o T 3 .An lnconel tube has been constructed for casting solid bars of f u e l 1 in. long and 0.1 in. i n diameter. The s o l i d bars of fuel w i l l be transferred t o lnconel capsules for irradiation. T h e capsule has been constructed s o that it can be welded shut without the bars being melted and can be reopened i n the hot c e l l for transfer of the fuel t o the corrosion t e s t capsule. Upon completion of an out-of-pile heating cycle, the t e s t capsule w i l l be opened in the hot cell. T h e fuel w i l l be d r i l l e d out and divided into three samples: a portion for petrographic study t o detect any oxidation or reduction, a portion for mass spectrographic study t o determine burnup, and a portion for chemical analysis of the fuel. T h e capsule w i l l be s l i t on the remote s l i t t i n g machine described below for a metallographic study. At least two concentric samples can be taken from a capsule of t h i s s i z e by remote means. F A C I L I T I E S FOR HANDLING IRRADIATED CAPSULES

C. C.

G. W. K e i l h o l t z Webster Solid State D i v i s i o n

It was observed metaIIograpliicaIly, i n some cases, that when several transverse sections were taken from a single fluoride fuel corrosion

capsule the corrosion pattern was not consistent over the length of the capsule but appeared t o be a function of the longitudinal temperature gradient. One method of obtaining a longitudinal section would be t o grind away half of the capsule on the m i l l i n g machine. However, by such a method it would be d i f f i c u l t t o ascertain that a section through the diameter of the capsule had been made. Also, the m i l l i n g operation would create a serious contamination problem. The method being used consists i n passing the capsule longitudinally by a 10-mil-thick s i l i c o n carbide fine-grit wheel and thus s l i t t i n g the capsule down the center. T h e capsule i s clamped i n t o a v i s e and then fed into the wheel by means of a mechanical linkage which i s to be replaced by a variable-speed motor and a contact cutoff switch. When the capsule i s put into the gripping adapter in the vise, it i s automatically aligned so that the wheel w i l l c u t i n the plane of a diameter through the capsule. B y making different gripping adapters, various sizes of tube can be s l i t lengthwise. Thus far the apparatus has been used on only MTR-type fluoride fuel capsules (O.lOO-in.-ID and 0.200-in.OD); it has worked very satisfactorily. Since the abrasive wheel has a rubber base, it must be kept cool. The area around the cut m u s t a l s o be kept cool to prevent any high thermal sfresses from affecting the corrosion results. The wheel and specimen are over a tray and are covered w i t h a splash shield s o that a l i q u i d coolant can be used. The coolant level i s always kept above the bottom edge of the abrasive wheel so that by cutting upward into the specimen the coolant i s carried i n t o the cut; therefore the specimen need not be i n the coolant. Carbon tetrachloride i s used as the coolant because i t does not react w i t h the fluoride fuel a t high temperatures. It i s forced i n t o the tray from outside the hot c e l l through Tygon tubing by air pressure, T h e tray and splash shield are attached t o the mount far the v i s e and they move w i t h the specimen. Once installed w i t h i n the hot cell, the remotely operated tube-slitting machine can be used w i t h ease i n cutting longitudinal sections from various sizes of tubing. Thicker abrasive wheels can be used for heavier walled twbing. The advantages of t h i s method of cutting are that it permits visual observation of the amount of fuel l e f t on the tube walls, i t permits observation of constrictions without the necessity of removing or disturbing the

135

ANP QUARTERLY PROGRESS REPORT material causing the trouble, it gives a well-exposed surface for chemically removing the lost traces of fuel, it permits a c u t to he made through the bottom p l u g of the capsule so that an examination can be made for crevice corrosion, and i t leaves nearly one h a l f of the capsule for further and more complete study, i f desired. After the fuel i s removed chemically without attacking the metal wall, the specimen can be electroplated so that the fuelmetal interface w i l l not be rounded during the meta I lograph i c pol i s hing opsrat ion. The s t a b i l i t y of the fluoride fuel under irradiation and the extent of the corrosion of the container by the fuel are determined, i n part, by analyzing samples of the irradiated fuel for the concentration of the original components and for the maior constituents of the container materio!. T h e fuel samples are put into solution w i t h i n the rnasterslave hot c e l l s and ore then transferred t o the A n a l y t i c a l Chemistry D i v i s i o n for analysis. Work on the necessary transfer shields has been coordinated w i t h the design and construction of the lead harrier being provided i n the a n a l y t i c a l chemistry hot c e l l s t o improve transfer conditions and t o reduce personnel exposure. T w o transfer carriers weighing 1150 Ih each and g i v i n g 2 in. of lead shielding have been constructed and are now i n use. Each carrier holds four 30-mI bottles that are held i n place by spring clamps. A motor-driven decapper for removing the bottle cap and a remotely controlled pipet that can be lowered i n t o the bottle are provided. The bottles may be inserted into the carrier w i t h i n the c e l l s and can be l e f t i n the carrier during a l l sample removal operations. The whole u n i t may then be returned to the hot c e l l s for reuse. T w o lead storage units weighing 450 Ib each and g i v i n g 3 in. of lead shielding were a l s o b u i l t and are available for use when a l l the transfer carriers are i n use, The lead storage units have a capacity of four 50-ml volumetric f l a s k s and can be loaded w i t h i n the master-slave calls. ANALYSIS O F I R R A D I A T E D F L U O R l D E FUELS F O R URANIUM

sampling operations, described previously,lt2 which have never been f u l l y resolved. I n the c u t t i n g and d r i l l i n g operations it i s possible for small c h i p s of metal, either froin the capsule or the drill, t o get into the fuel sample. To minimize t h i s possibility, d r i l l s used recently have been carefwlly honed to remove burrs. A search i s a l s o being conducted for new d r i l l materials. In the early use of the present sampling technique,’ it was possible for d i r t from the hot c e l l s t o inadvertently enter the fuel samples during c u t t i n g and drilling. Phis p o s s i b i l i t y has been minimized i n thc newer equipment.2 The transfer operations, i n several of which the s a l t i s exposed to contaminotion from hot c e l l dirt, pieces of rubber from manipulator grips, and the like, are all possible sources of contamination. T h e observation of opaque material, i n some cases i n large amounts, during petrographic examination of irradiated fuel samples3 indicates that at one or the other of the stages discussed above foreign material may have been introduced into some of the samples. Such contaminants moke the resulting uranium analyses I‘

IOVb.”

The samples obtained for chemical analysis have, a t times, been v e r y small, often being as small as 10 mg. They are weighed in tared weighing bottles on a conventional chain-type a n a l y t i c a l balance, and, since they are so small, the precision of weighing i s about t4% instead of the usual negligible amount. T h i s lack of precision results i n an abnormally high uncertainty i n the uranium analyses. After being weighed, the samples are dissolved and the solutions are made up t o volume i n the hot cells. The resulting solutions often contain as l i t t l e as 500 ppm of uranium. Such high d i l u t i o n i s undesirable, s i n c e the t i t e r may change appreciably on long standing. It seems reasonable t o conclude that the uncertainty i n uranium analyses may be as high as +5% or more, even for the usually precise (52%) potentiometric titration. An increase of ten times i n sample s i z e (easily attainable) would restore t h i s technique t o i t s normal precision.

M. 7. Robinson

G. VI. K e i l h o l t r Solid State D i v i s i o n

One of the regular steps i n in-pile static corrosion testing of fluoride fuels has been the analy s i s of the irradiated fuel for uranium. However, there are several maior problems connacted w i t h

136

’J. G. Morgan e t ai., Solzd State Diu. Semzann. P r o g . Rep. Aug. 3, 1953, ORNL-1606, p 40.

2C. C. Webster and J. G. Morgan, Solzd S t a t e D i u . Semzann. Prog. R e p . F e b . 28, 1954, ORNL-1677, p 27.

3G. D. White and M. T. Robinson, Solzd State D z v . Srmrat711. Prog. R e p . F e b . 28, 1954, ORNL-1677, p 28.

PERlOD ENDING SEPTEMBER 10, 19.54 Similar remarks apply, i n part, to the mass spectrographsc analysis. T h e samples taken are weighed by the analysts on a microbalance t o a l l e v i a t e the e f f e c t of small sample size. However, the handling of the very small crucibles i n the hot c e l l s i s not so delicate an operation as might be desired. The p o s s i b i ? i t i e s of contamination entering the sample are high. No highly useful conclusions have emerged from much of the analytical work done in the past, probably because of t h e d i f f i c u l t i e s described above. Recently, however, mar ked improvement has been made as a r e s u l t of continued efforts t o improve sampling techniques, sample sizes, und handling procedures, and uranium analyses have been carried out in c1 satisfactory manner. Since the analytical results obtained recently show that no meosurable segregation of uranium takes place, the procedure of taking more than one sample froin each capsule w i l l be abandoned. This w i l l allow very much larger samples t o be obtained, w i l l make the analyses more sure, and w i l l minimize the effects of foreign additions t o the salt. N I G H - T E M P E R A T U R E CHECK V A L V E TESTS

R. W i l l i s

G. W. K e i l h o l t r

Solid State D i v i s i o n

As part of the development program for a s m a l l

in-pile pump loop for operation i n o beryljium A-piece i n the LITR, an attempt was made t o deve Iop a h igh-temperature c Iiec k-va Ive pump. Such a pump has the advantage of being a compact, completely sealed unit.

A test r i g was run w i t h Inconel, stainless steel, and S t e l l i t e 2 5 check bolls. None of these materials gave successful tests. Time of operation before failure varied from 30 min for the Inconel to 2 hr for the stainless steel. Except w i t h the stainless steel balls, operation of the valves was not continuous; they would occasionally s t i c k i n both open and closed positrons. When failure occurred, it was abrupt, and therefore s t i c k i n g of h e valves was indicated rather than a gradual stoppage of the system. MINIATURE

IN-PILE L O O P

BENCH TEST

G. W. Keilholtz J. C. Morqan Solid State D i v i s i o n I

A n exploded v i e w of the in-pile pump described previously4 i s shown in F i g . 9.1. The RPM meter c o i l measures the rotation of the shaft on which

i s mounted die motor armature, The Graphitar bearing i s mounted in the bearing casing above the impeller housing. The assembled loop i s shown i n Fig. 9.2. The pump contains an extra reservoirt o enable metered vo/vmes of s a l t t o be pumped into the surge tank for calibrating the venturi. The furnace for heating the pwmp and reservoir i s a l s o

shovvn. From performance tests with this setup, pump speed v s head and pump speed Y S flow have been determined. A t 5000 rpm the flaw was 4 Cps, corresponding to a Reynolds number of 3000 i n ea 200-mil tube. At the 540 w/cm3 generated in-pile, d i e temperature differential i s expected to

10PF.

be about

L I F E T E S T S O N A N R P M METER, B E A A N D A SMALL E L E t f R l C M O T O R U N D E R ~ ~ ~ I O NA ~ l ~

J. G., Morgan M. E. Robertson G. W. Keilholtr

Solid State D i v i s i o n

Tests were conducted on a n RPM meter, bearings, and a s m a l l electric motor as a part of the development of n loop fox insertion i n an LdTR A-

piece, In order t o drive the pumpr it i s desirable t o USE a variable-speed electric motor and to a l i g n

the s h a h w i t h lubricated bearings. Also, a record of shaft speed would be desirable. The entire motor assembly w i l l be just above the upper grid guide i n the LBTR sand i n a n estimated thermal neutrons/cm2.sec. Hole 53 of the f l u x of 5 fi lQ1 ORNL Graphite Reactor provided a comparable f l u x intensity and was used fur these tests. The in-pile apparatus, w i t h a single lower bearing A as i n the second test, i s shown i n Fig. 9.3, motor was mounted i n a horizontal position w i t h

In the upper and lower armature shaft bearings. f i r s t test two lower bearings w e r e used. The motor and RPM meter were canned and inserted into The variobles the center l i n e of the reactor. measured were resistance of t h e motor winding, resistance of the RPM coil, the voltage applied t o the motor, the current drawn by the motor, the temperature of the lower Fafnir bearing, the temperature of the motor field, and the speed of the motor shaft. The Delco ac-dc motor u s e d is rated a t 115 v, 13.9 amp, and hp and has insulated wires. Some substitutions of materials were made. In t h e f i r s t

ANP QUARTERLY PROGRESS REPORT

UNCLASSiFIE D

PHOTO 12739

R P M M E T E H COIL

MOTOR ARMATURE

BEARING CASING

I M P E L L E R HOUSING

F i g . 9.1. In-Pile Pump.

UNCI AISIFIED PilOTO 12138

'4 E N T U R I

Fig, 9.2, Bench 'Test Loop.

138

PUMP AND E X T R A RESERVOIR

PERIOD ENDING SEPTEMBER IO, 7954 UNCLASSIFIED 20814

ORNL-1.R-DWG

'HFARING, MOTOR

FAFNIR DVJ-6

\THERMOCOUPLE

RPM

COIL

' N S T R U M F N T LEADS

DELCO,AC-DC, '115 hp, SFRVICE NO 5047431

SlATOR TAPE

GLASS

B R U S H HOLDER P L A T E R P M COIL

MICA

N O 36 A W G C C R O C ,

BFARING GREASE

RRG-2

(NL.GI

200-5500 T U R N S GRADE 2 )

F i g . 9.3. Motor and

RPM

test a brush holder plate of mica backed w i t h a stainless steel plate was used, I n the second test fired L a v i t e asbestos covered w i r e was used from the brush holder t o the f i e l d coil. Glass tape was used around the f i e l d coils. The rest of the construction features were unaltered; that is, fish-paper-coil supports, p l a s t i c (paper-backed Bake1ite) commutator segment spacers, and cardboard armature shaft spacers were used. The upper bearing (New Departure 77R6) and lower bearing (Fafnir DW-6) were packed w i t h California Research Corporation grease, identified a s RRG-2 [NLGl Grade 21.

The RPM meter consists of a magnet fastened t o the shaft which w i l l rotate inside the gas-tight motor shell. A c o i l wound around a C-shaped iron core i s mounted outside the shell. A s the shaft rotates, pulses aregenerated i n the c o i l and counted remotely by a frequency meter calibrated in rprn.

T h e c o i l was wound w i t h 36-9 Ceroc 200 magnet w i r e covered w i t h glass tape impregnated w i t h Glyptol. Summary data on the t e s t conditions are given i n Table 9.1. The motors were examined, after irradiation, in t h e * hot c e l l . Assistance was given i n hot c e l l examination by J. Noaks and J. Crudele, Pratt & Whitney Aircraft Division. A withdrawal cask and i t s f a c i l i t i e s were loaned by L. E.Stanford, General E l e c t r i c Company.

Meter Test Assembly.

TABLE 9.1.

CONDITIONS OF TEST MOTOR AND RPM METER ASSEMBLY Test

No. 1

T e s t No. 2

Motor load

60 0.45

d - c voltage Current, amp Shaft speed, rprn F u f n i r bearing temperature,

' C

Motor field p l a t e temperature,

33 0.20

4500

100 84

648

T i m e of operation

4500 rprn,

I n flux,* hr * T h e r m a l flux,

5 x 10" neu+rons/cm2-sec; 1; gumma f i e l d , 4.4 r/hr.

613

r a t i o of f a s t

to thermal neutrons,

I n the f i r s t test, motor load was abnormally great from the beginning and the f i e l d plate temperature was high. When the motor failed after 100 hr, the stator temperature rose t o 182OC w h i l e the power was kept on. The suspicion that there was a cocked upper bearing plate was confirmed upon postirradiation examination (Fig. 9.4). The sensing c o i l for the RPM meter functioned well; there was no resistance change or v i s u a l deterioration.

139

ANP QUARTERLY PROGRESS REPORT

Fig. 9.4.

Motor Irradiated in the F i r s t Test.

In the second test the motor operated as long a s it would be expected t o under normal conditions, However, postirradiation inspection showed that the commutator segments had been loosened as a result of the deterioration of the plastic commutator segment spacers under irradiation (Fig. 9.5). The bearing grease proved t o be adequate and W Q S essentially unaffected, as evidenced b y the constant power demand by the motor a t constant rpm. The RPM c o i l a l s o showed no resistance change or The brush wear v i s i b l e breakdown (Fig. 9.6). appeared t o be normal. A high-temperature motor i s being constructed for testing that w i l l have g lass-insu loted-w ire, m ica, and h igh-a It itude brushes. W E M Q V A L O F XENON F R O M FUSED FLUORIDES

T. Robinson

G. 14â&#x20AC;&#x2122;. Keilholtz

F i g . 9.5. Ccornniwtator Segments from Motor Irradiated in the Second Test. purging. A stainless steel capsule w i l l be constructed i n CJ manner that w i l l allow bubbles o f helium t o pass through a f i l t e r plate arid through the fissioning fuel i n hole HB-3 o f the LITR. P l a s t i c mockups of the capsule are being used t o A study the characteristics of the helium flow. four-channe! gamma-ray spectrometer has been constructed by the Instrument Department t o monitor the X e 1 3 5 concentration in the flowing stream of helium. C 7 F 1 ,-U F 6 1R R A D lA T I O N

W. E, Browning

G. W. Keilhoitz Solid State D i v i s i o n

Solid State D i v i s i o n

D. E. Guss United States Air Force

W . A. Brooksbank Anatytical Chemistry D i v i s i o n

A program has been initiated to study removal of from fluoride fuels by means of helium Xe13â&#x20AC;&#x2122; 140

A n experiment i s being conducted to determine i n a very rough fashion the effect of p i l e irradiation on a 20% solution of UF, i n C,F,,. T h i s material i s of interest t a t h e ShieldingSection of the Physics D i v i s i o n for simulating l i q u i d fuels in dynamic shielding experiments, and the experiment i s being conducted

w i t h the a i d of personnel from that

PERlOD ENDING SEPTâ&#x201A;ŹMBER 70, 7954 unirradiated capsule contained no residue except for a faint chalkiness on i t s inner wall. Additional capsules are being irradiated for 1/100 of the accumulated exposure.

LIITR H O R I Z O N T A L - B E A M - H O L E FLUORIDE-FUEL L O O P

W. E. Brundage C. D. Baumann

F. M. Blacksher R. M. Carroll J. R. Duckworth

C. E l l i s M. 1.Morgan A. S. Olson W. W. Parkinson 0. S'I sman

Solid State D i v i s i o n

A loop for circulating fluoride fuel was inserted i n hole HB-2 of the LlTR and f i l l e d w i t h fluoride (62.5-12.5-25 mole X) during a fuel NaF-ZrF,-UF, protracted reactor shutdown. The pump was started as soon as sufficient fuel had been added, but circulation of the fluoride mixture became erratic after 5 t o 10 min and f i n a l l y stopped completely. Fluoride fumes i n the off-gas l i n e from the jacket around the pump indicated a leak,5 and therefore the loop was removed before the reactor was started.

Fig. 9.6. division.

The

Motor Irradiated in Second Test.

Three nickel capsules were charged w i t h

the two components to make a 7-9 solution i n each. Normal uranium was used. T w o capsules were irradiated in appropriate secondary containers i n a water-cooled hole i n the ORNL Graphite Reactor at a thermal flux of 6 x 10'' neutrons/cm2.sec for 63 hr. This exposure corresponds t o 25 times the f i s s i o n density expected i n the shielding experiThe capsules were ment w i t h enriched uranium, opened i n a special gas manifold, and the gas pressures were determined. The pressure inside the irradiated capsules was shown t o have been greater than 50 and less than 100 psia. T h e pressure i n the unirradiated capsule was not perceptibly different from the vapor pressures of the two components (about 4 psi). T h e gas i n the capsules was analyzed by infrared absorption spectrometry at K-25 and was found t o contain CF,, C2F,,

C,F

,6, and other fluorocarbon campounds not identified. No UF, could be detected i n the gas sample. Pressure measurements a t -80 and -190T were consistent with these analyses. T h e irradiated capsules were f u l l of o nonvolatile (at 25OC) greenish powder. The powder W Q S insoluble

C,F,,.

It w i l l be analyzed for uranium.

loop was carefully disassembled to avoid

loss of uranium, since over 5 kg of enriched fuel

The

mixture had been used t o charge the loop. Recovery of the fuel from the "nosepiece" at the in-pile end of the loop had t o be carried out i n a hot c e l l because of neutron activation acquired when the u n f i l l e d loop was exposed to reactor f l u x before the fuel was added. The leak was found i n the weld ioining the loop tubing t o the discharge nipple at the bottom of the pump bowl. Complete breaking of the weld occurred probably during disassembly of the loop or during cutting of the pump jacket, rather than while the pump was i n operation. Before insertion i n the reactor the jacket for enclosing the loop and the pump in an inert atmosphere was tested for gas tightness. The loop, w i t h the pump superstructure replaced by a blank flange, was vacuum tested a t 800 t o 1000Â°F w i t h a helium leak detector. After the Joop was inserted in the reactor, i t was pressure tested at 1200 to 1500OF a t 16 psig, and there was no perceptible loss i n pressure over a I-hr period, even though the pump superstructure w i t h i t s rotating shaft seal was i n place,

'w. E. Brwnda e at ai., ANP Quar. Prog. Rep. l u n e 10, 1954, ORNL-7729, p 107. 141

ANP QUARTERLY PROGRESS REPORT

The connections of the loop tubing a t the pump have been redesigned for the second loop t o provide a f i l l i n g and drain line. T h i s w i l l permit charging the loop w i t h a non-uranium-bearing fluoride mixture, circulating i t a t operating temperature, and draining i t before the loop i s inserted into the reactor. 'The danger of tube rupture because of expansion of the fuel on melting necessitates the draining and r e f i l l i n g feature, i f operating tests are t o be carried out. The necessary changes in the pump shield to accominsdate the drain and f i l l l i n e are being made. TheMetaIlurgy D i v i s i o n w i l l supervise a l l c r i t i c a l welding and has provided specifications for an improved weld joint developed for lnconel tubing. The components fabricated previously for the second loop are being changed t o conform t o the specifications for the new weld joints, A new flange for closing the rear of the jacket has been made in order t o give more space for making pipe connections t o the large air lines required for the heat exchanger. During installation of the loop in the reactor an unexpectedly long time w a s required t o connect heater leads, thermocouple leads, and air, water, and helium piping i n the confined space available between the shielding and the instrument panel, The piping a t the reactor dace i s being revised t o f a c i l i t u t e connection when the second loop i s ready for insertion s o that installation nwy be performed more rapidly. Modifications are 0lso being mode t o remove lag in t h e flow alarms for the jacket cooling water, The o i l systamfor cooling the pumps i s being rebuilt t o incorporate more r e l i a b l e punips and improved f l o w and alarm instruments. A hydraulically operated door to match the withdrawal shield has been installed i ; i a w a l l of II hot c e l l i n preparation for handling the loop after irradiation, A horizontal band S Q W has been modif i e d for nlnnost complete remote operation and i s being installed i n the hot cell. Other remote kand l i n g tools for disassembling the loop have been made and are being tested.

0 R Ed h G R A P H BY E R E A C T OR S O D iLs $4-I NC 8 N E L.

W. E. Brundage

LOOP

W. 'A'. Paikinsori Solid State Divisiori

Metallographic examination W Q S completed on un lnconel loop which had circulated sodium i n hole

142

58-H of the ORNL Graphite Reactor for 218 hr with t h e irradiated section at 1500°F. The unperturbed thermal f l u x i n t h i s hole i s about 7 x and t h e f l u x above 1 ev is a factor of 10 lower. There w a s n o evidence of radiation induced corrosion of t h e Incanel. The temperuture of the pump c e l l section was held between 1025 and 1100"F, and that of the l i n e entering the reactor was somewhat higher; the temperature of the l i n e emerging from

lo",

t h e reactor was about 1500°F. Before insertion i n the reactor the loop had been operated at 800 t o 1000°F for 80 hr while sodium was circulated

through the loop and through a bypass f i l t e r c i r c u i t connected t o t h e loop. C R E E P AND S T R E S S - C O R R O S I O N TESTS

w. w.

N. E.

DClVi§ J, C. Wilson Hinkle J. C. Zukas Solid State D i v i s i o n

Development of Q suitable apparatus for i n - p i l e stress.corrosion experiments has bDen slow because of d i f f i c u l t y i n securing constant heattransfer rates from the fuel t o a heat sink where In the design t h e f i s s i o n heat can b e removed. shown i n Fig. 9.7, f i s s i o n heat i s conducted from t h e fuel through the stressed specimen wall t o a

sodium-filled annulus. The sodium a c t s as an e f f i c i e n t he& conductav that exerts no constraint upon thc defo:mation of the specimen tube as it creeps under applied stress. Part of the f i s s i o n heat i s removed by gaseous convection at the outer surface of the sodium-containing chamber and part by conduction through a metal r i n g t o a water iacket. Simpler designs i n which the sodium was contained in u constant-diameter annulus did not permit good temperature control. T h e temperature gradient that resulted from simulated f i s s i o n heat input at the bottom and heat abstraction at the top caused heated sodium t o r i s e discontinuously i n the annulus, and temperature excursions of 100°F i n 1 sec were not uncornrnon. A substantial temperature gradinnt (500 t o 800°F per in.) i s required to meet headioum requirements i n t h e experiniental hole and t o prevent sodium d i s t i l lation out of the annulus. Several designs o f the sodium annulus were tried, such as baffles and spirals, and r a t i o s of outs ide-to..i nsicde sodiuni-nnnu lu s d i ameters were varied but did not result i n a constant temperature I he appa'ratus shown i n Fig. 9.7 appears gradient. t o operate satisfactorily, but more operating time

PERIOD ENDING SEPTEMBER 70, 7954 IJNCLASSIFIEO

O R N L - L R - D W G 3tn9

x,-,,

L'JULAIU I

LHAfvlBtK

r----

THERMGCOL!PLE WELL

m u s t be accumulated before it w i l l be certain that sodium d i s t i l l a t i o n i s not taking place in sufficient quantities t o cause trouble. In the apparatus a tubular Inconel specimen cont a i n i n g 0.2 g of NaF-ZrF,-UF, (53.5-40.0-6.5 mole 5%) in t h e fuel chamber attached t o t h e fixed base

i s eccentrically loaded through a lever arm by a weight. A stress pattern i s produced across the horizontal c r o s s section which varies from a maximum compressive stress. Surrounding the specimen i s a sodium coolant chamber w i t h baffles which reduce the magnitude of the thermal-conv e c t i o n currents and allow closer control of the specimen temperature. A furnace helps t o marlatarn the temperature at 1500째F throughout the test. Two thermocouple w e l l s prolect into the sodium

chamber t o a position near the w a l l s of the speci-

A double-walled, conmen at the neutral axis. centric, sodium f i l l line, w i t h only the outer tube welded t o the chamber, and a vacuum exhaust t u b e complete the apparatus. All surfaces in contact w i t h the fuel and sodium ore made of Inconel. T h e assembly i s hydrogen f i r e d for 1 hr a t 1750"F, f i l l e d w i t h fuel, and welded i n a dry

box under purified helium. The other parts of the container are welded into subassemblies, hydrogen fired, and then welded together, w i t h helium backing up the welds. The lever arm and weight are adlusted and spot welded onto the specimen,

and t h e chamber covers ore welded into place. Before the rig i s f i l l e d with sodium, it I S f i l l e d w i t h NaK and heated t o

1000째F

to remove any

143

ANP QUARTERLY PROGRESS REPORT

o x i d e remaining on the metal surfaces. The r i g i s then washed with butyl alcohol, which reacts w i t h t h e NaK and leoves the surfaces clean and ready t o b e f i l l e d w i t h sodium. The sodium i s metered out into a tube that has f i r s t been cleaned w i t h NaK, sealed with Swagelok fittings, and provided w i t h a helium valve. A micrometallic f i l t e r sealed i n t o the l i n e remov~ts any oxides of sodium that may be formed when the connection t o the test r i g i s made. Once the connection i s made, the apparatus i s evacuated and t h e sodium i s melted and forced into the chamber by helium pressure through the valve; a combination of high surface tension and back pressure made f i l l i n g the chamber sonlewhat more d i f f i c u l t without t h i s procedure. For the bench tests, i n which s t i c k s of sodium were used, the f i l l e d chamber was heated w h i l e a vacuum was pulled on the system, and a r i s e i n pressure was noted a t about 525”F, presumably f r o m dissociation of sodium hydride. It remains t o be seen whether such a procedure is necessary t o ensure the purity of the sodium now available, The dry-box work i s completed by cutting the outer f i l l tube, removing the inner (contaminated) tube, and crimping and welding both the fill and exhaust tubes. T h e entire assembly i s welded i n t o a stainless steel water jacket furnished w i t h Kovar seols and a capillary tube through which helium f l o w s throughout the t e s t . A probe beneath the weight c u t s o f f the current t o the furnace at a predetermined deflection of the lever arm and indicates t h e time required for a given degree of deformation. A gravity “ti It” indicator a i d s i n plumbing the specimen i n the exposure can. A base p l a t e w i t h a conical cup furnishes a well for the sodium i n case of rupture during operation. A suitable safety system i s being developed, Approval by the L-ITR Experiment Review Comm i t t e e i s being withheld u n t i l additional t e s t s are made on the compatibility of sodium w i t h the various component materials it might contact i f a rupture occurred at the t e s t temperature. Several t e s t s aimed at a qualitative determination of the speed of reaction and the heat generated by such reactions should be completed w i t h i n a short time. After an irradiation period of from two to s i x weeks i n h o l e HB-3 of the LITR and a suitable decay period, the specimen i s to be sectioned by t h e remote metallography group and examined for stress-corrosion effects.

144

Bench t e s t s on t h e specimen tube alone are under way w i t h the outside surface i n air. The i n s i d e of t h e tubes contains either air, helium, or

(50-46-4 mole %). A n atmosphere NaF-%rF,-UF, chamber has been b u i l t for t e s t i n g four specimens simultaneously i n vacuum or any desired atmosphere. One 200-hr t e s t has been completed on (50-46-4 mole %) at 1500” F barren NaF-ZrF,-LJF, and about 1500 psi. No stress dependence of corrosion was found; another specimen was tested A for 400 h r but has not y e t been examined. melting and overturning technique for removing fluoride fuel from the section that i s t o be examined metallographically has been successful; t h i s w i l l greatly simplify handling o f irradiated capsules. T h e MTR t e n s i l e creep r i g i s undergoing f i n a l t e s t i n g and i s scheduled for irradiation t e s t s on lnconel i n October. REMOTE METALLOGRAPHY

M. J.

Feldmon

R. N. Ramsey

W. Parsley A. E. R i c h t

Solid State D i v i s i o n

Work was continued on the examination of the solid-type fuel sandwiches of interest t o the Pratt & Whitney A i r c r a f t Division. I n addition t o the two experiments mentioned previouslyI6 s i x other specimens have been examined. The r e s u l t s on three o f them (capsules 1-4, 1-6, 1-7) have been p u b l i ~ h e d . ~ A report covering t h e other three capsules (1-5, 1-8, 1-9) i s being prepared. Examination of the eight capsules processed t o date has indicoted greater resistance t o cracking and core-clad separation upon bending of the stainlesssteel-matrix samples than upon bending o f t h e iron-rnatrix samples, better resistance t o cracking upon bending of the large

UO, p a r t i c l e s i z e core

than upon bending of the smaller p a r t i c l e s i z e core, and a greater increase i n hardness upon irradiation of t h e small p a r t i c l e s i z e core than upon irradiation of the larger p a r t i c l e s i z e core. Work on solid-type fuel elements for the GE-ANP group has continued. Examinations of two n i ...

......

6A. E. Richt, E. S r h w o r t z , and M. J. Feldman, Solid S t a t e Dzw. Semiann. Prog. R e p . F e b . 28, 1954, ORNL1677, p 9.

’M. J. F e l d m a n et al.. Metalloerabhic AnaIvsis o l Pratt and Whztney C a p s u l e s 1-4, r-6: and 1-7,’ ORNL CF-54-5-41 ( M a y 5, 1954).

PERlOD ENDlNG SEPTEMBER 10, ‘1954 chrome V fuel elements have been completed.8 Examinations o f two single-plate fuel elements9 and one two-stage m u l t i p l e fuel p l a t e t e s t assembly” were a l s o completed. k o r k w i l l continue o n both wire and multiple-plate elements. Expansion o f remote metollograpkic f a c i l i t i e s i s proceeding.

F ISSION-FRAGM E N T ANN E A L l N G S T U D I E S

M. J.

Feldmon W. Parsley Solid State D i v i s i o n

Because of interest in t h e p o s s i b i l i t y of t h e removal o f the radiation damage t o fuel plates by annealing, a preliminary study was undertaken. T h e eight samples i n P r a t t 8, Whitney capsule 1-9 (four were small p a r t i c l e sire, ( 3 - p UO,, and four were large p a r t i c l e size, 15- t o 44-p UO,) were selected for t h e study. One sample from each group (large and small p a r t i c l e size) was examined (1) as irradiated, after t h e following treatment: (2) 900°F anneal for 24 hr, (3) 1100°F anneal for 24 hr, and (4) 1400°F anneal for 24 hr. A graph showing the results of t h i s preliminary study i s shown in Fig. 9.8. The conclusion drawn fron. t h i s i n i t i a l study i s that a portion of the neutron damage to t h e samples has been annealed out. Since the f u l l annealing temperature for type 347 stainless steel i s about 1800°F (for complete removal of the effects of c o l d work), a complete anneal for neutron damage a t 1400°F b a s not expected. For the core, where the maior portion of the damage i s by f i s s i o n fragments, it i s f e l t that the curve shows a reduction in hardness because of a partial anneal o f the neutron darnage

w i t h n o reduction in t h e fragment damage. An attempt w i l l be made t o determine the anneal necessary t o remove a l l the neutron damage from t h i s t y p e o f material and, i f possible, the f i s s i o n fragment annealing temperatures. HIGH - T E M P E R A T U R E , SH 0 RT - T IM E, G R A I N GROWTH CHARACTERISTICS OF I N C O M E L

M. J.

Feldman

A. E. Richt

Solid Stat;

stock used for t h e t e s t s was made. information i s available i n the literature on the usual times and temperatures for lnconel grain growth, but because o f the nature of t h e experiment, w i t h i t s p o s s i b i l i t i e s o f local short-time hot spots i n the f u e l , t h i s study was iniliated. F i g u r e 9.9 i s a graph of t h e data obtained. Of major interest were t h e definite temperature dependence of t h e carbide s o l u b i l i t y and t h e extremely short times (relative t o the corrosion t e s t times) a t which large grains could be produced. D e t a i l s o f the experiment have been published.” BNL N E U T R O N S P E C T R U M - R A D I A T I O N DAMAGE STUDY

J. B. T r i c e

P. M. Uthe

Solid State D i v i s i o n

f?.

Bolt

J. C.

N. Shiells

Carroll

C a l i f o r n i a Research Corporation

V. \Val sh Brookhaven National Laboratory A series o f measurements were made i n hole E-25 of the Brookhaven reactor t o determine neu-

tron f l u x energy distributions. The project was a cooperative effort with California Research Corporation and Brookhaven National Laboratory. Information obtained from these measurements i s to b e used to correlate neutron f l u x intensity and energy distribution w i t h radiation damage to cer-

t a i n lubricants and organic compounds that have been irradiated in E-25 and exomined by CRC. Data for t h e irradiated samples have been coded for the ORACLE by the ORNL Mathematics Panel so that the large number of arithmeticol comput a t i o n s usuolly required b y o threshold detector experiment can b e performed in a shorter period of time than i s usual.

W. Parsley

Division

A s an a i d i n the onolysis of the s t a t i c corrosion capsules, a study of the short-time, high-femperoture, grain-growth characteristics of the lnconell

‘NI. J. Feldman e t al., Metallogtuphical A n a l y s i s of G.E. Wire Fuel E l e m v n t No. 1 , ORNL CF-54-4-8 ( A p r i l 1, 1954).

‘A. E. Richt and R. N. Ramsey, Metallographiral A n a l y s i s of S i n g l e Plate F u e l E l e m e n t s G E - A N P 30 and 3C, QRNL CF-54-3-42 (March 9, 1954).

’OM. J.

al., MrtalIogtaphic A n a l y s i s of Test Speczrpen C E - A N P - 1 0 , QRNL CF-54-7-77 (July 9 , 1954). Tnuv-Stage

Feldrncrn e t

MTR

’M- J. Feldman e t al., Fhort 7‘zme-High-Temperuture Grain Growth Charucteristzr of litconel, ORNL CF-546-70 (June 8, 1954).

145

ANP QUARTERLY PROGRESS REPORT

UNCLASSIFIED

.""'.,.

600

ORNL-LR-DWG

2687A

55 0

500

450

400

5 n 0

350

s cn

300

13 K

250

200

150

(00

I RR AD IATE D 6 3 3 hr TO-i2.4qb

U"'

___ e-

BURNUP

Fig. 9.8. Preliminary Annealing Study of Stainless Steel-UB, Fuel Plates i n $ r a n and Whitney C a p s u l e

1-9.

116

PERlOD ENDING SEPTEMBER IO, 1954 UNCLASSIFIED ORNL-LR-DWG 1552A

...............,---

MTR

NEUTRON-FLUX SPECTRA - RADIATION DAMAGE STUDY

T. L.

Trent

P. M.

Sol i d State

Uthe

J. 5.

Trice

Division

J, Moteff

General E l e c t r i c Company

F i s s i o n probes w i t h different

.-E

w 8

z+-

used included U238, Y2361 Np237, and Th232, whose f i s s i o n threshold energies span the energy

'/2

region between and 2 MeV. The data from these traverses w i l l be used w i t h the r e s u l t s from a c t i vations i n H5-3 of other, nonfi ssionable, threshold detectors that are i n t h e process of being irradiated.

0.5

1800

Fig. 9.9,

conel.

fissionable rna-

t e r i a l s were used t o make neutron f l u x traverses i n H5-3 of t h e MTR. The fissionable materials

1900

2000 ~ 1 0 0 2200 TEMPERATURE (OF)

2300

2400

Grain Growth Characteristics of In.

After measurements have been completed i n HB-3 w i t h t h e hole empty, several inches of beryllium w i l l be introduced into t h e hole i n such a way as

t o change t h e energy distribution of t h e transmitted neutrons. T h e new energy distribution w i l l then be measured, and materials will be irradiated i n both moderated and unrnoderated neutron spectra.

An attempt w i l l then be made t o correlate spectra and radiation damoge. 12

"J. B. T r i c e and P. M. Uthe, Solid Stute Semiunn. Prog. R e p . Fe6. 28, 1954, ORNL-1677, p. 14.

147

ANP QUARTERLY PROGRESS REPORT

10. ANALYTICAL STUDIES OF REACTOR D, Susan0 Anal y t i c a l Chemi stry Di v i s i on

J. M. Warde Metallurgy Division

The primary analytical problem continues to b s the separation and determination of trivalent and and NaF-KFtctravalent uranium in both NaZrF,. L i F - b a s e fuels. A successful potentiometric titration of UF, i n molten NaZrF, with metallic zir-

conium was accomplished by means of polarized platinum electrodes, The observed end point occurred at a weight o f titrant which corresponded to one-half an squivolent of zirconium metal per

UF,. T h i s stoichiometry indicates reducPolarot i o n of UF, t o a mixed oxidation state. mole of

giaphic studies of uranium fluorides i n reactor fuels

dissolved i n molten ammoniirni formate at 125OC Preliminaiy experiments reve"aled that tetravalent uranium was not reduced a t the dropping-mercury electrode. Molten ammonium formate appears to be o useful solvent for fluoridebase fuels containing various oxidation states of uranium. The conversion o f UF, to UCI, i n fluoride fuels was shown to be quantitative when the fuel was heated with boron trichloride at 40O0C for

were initiated.

min.

Trivalent uranium fluoride i s converted to

UCI, under these conditions. The s o l u b i l i t y of UF, i n NaAICI, was found t o be 18 mg/g at 200OC os contrasted to o solubility of UF, of less than 1 mg/g. Molten NaAICI, i s expected to dissolve

telkavalent uranium selectively from the fuels. Calibration measurements have been completed on the apparatus for the determination o f oxygen a s metallic oxides i n reactor fuels. The reaction inv o l v e r thc hydrofluorination of the oxide and measurement of the increase i n conductivity of l i q u i d HF as a function o f the water formed, Investigations were also made on the oxidation

UF, with oxygen at elevated temperatures. At 30OoC, UF, i s converted quantitatively to UF, and UO,. The mechanism of the oxidation of UF,

with oxygen i s being studied.

An oxide residue of

11% of the original material was found when UF, was heated in oxygen a t 3OOOC for 2 hr. An improved method for the determination of

about

lithium in NaF-KF-LiF-base fuels W Q 5 developed, The chloride s a l t i s extracted with 2-ethylhexano1, and the chloride i s titrated in the nonaqueous mediums,

148

Studies were also made on the s o l u b i l i t i e s o f potassium, TZI bi d iom, and cesi urn tetrapheny i borate s i n various organic solvents to ascertain differential The f e a s i b i l i t y of a spectrophotosolubilities. metric titration of fluoride i n NaF-KF(RbF)-Li Fbase materials by decolorization of zirconium or

aluminum complexes was investigated. Determinat i o n of the concentration of o i l in helium was mode by absorbing the hydrocarbons in petroleum ether and weighing the residue obtained after evaporation. It was found that prior bleeding o f the l i n e effect i v e l y eliminated contamination by oil. ANALYTBCAL CHEMISTRY O F REACTOR MATERBALS

White

Analytical Chemistry D i v i s i o n Research and development were continued on the problem of determining UF, and UF, in NaZrF,and NaF-KF-LIF-base reactor fuels. A number of n e w opproachss to t h i s problem were investigated, including redox titrations o f the fluoride mixtures in the molten state, for which polarized platinum electrodes were used, and polarographic studies of fluoride mixtures i n molten ammonium formate. Work was continued on the determinotion of oxygen in fluoride fuels and on the determination of a l k a l i metals and fluoride ion i n NaF-KF-LiF-base mat e r i a l s, Determination

of Oxygen

A. S, Meyer, Jr.

Fluwide F u e l s

J. M, Peele Analytical Chemistry D i v i s i o n

T h e apparatus for the determination of oxygen as

oxides i n fluoride fuels, which w a s described briefly in a previous report, has been constructed. A schematic drawing of t h i s apparatus i s shown i n Fig, 10.1. T h e hydrogen fluoride purification s t i l l , which was fabricated from m i l d steell i s equipped with a 24 x 1 in. packed column with an estimated efficiency of 25 theoretical plates. Conductivity

' A . S. Meyer, Jr. and J. M. Peele, ANP Quar. h o g . R e p . l u n e IO, 1954, ORNL-1729, p 110.

PERIOD ENDlNG SEPTEMBER 70, 1954 UNCLASSIFIED OHNL-1.H-DWS 1724

MONEL NEEDLE VALVFS

MONEL BELLOWS VAI>/FS

I.---

ANHYDROUS HYDROGEN FLUORIDF

20 % KOH

Fig. 10.1. Apparatus far ~

measurements are carried out in a 25-ml fluorothene c e l l f i t t e d w i t h platinized platinum electrodes. T h e reaction vessel, a modified Parr bomb, i s s i l v e r plated on the inner surfaces, as are the l i n e s and the f i t t i n g s in the reactor-conductivity c e l l section. In the procedure now being tested the sample i s placed i n the reaction vessel and the system i s evacuated to valve 5 (see Fig. 10.1). After the c e l l has been f i l l e d with a measured volume of t h e distillate, the conductivity is measured and t h e a c i d i s transferred to the reaction vessel by distillation. When the oxides have keen converted t o water, the solution of water in hydrofluoric

SODA-LIME TRAP

of r Oxygen ~ i ~in Fluoride ~ t i SoJts. a ~

a c i d i s d i s t i l l e d to the c e l l where the measurement of conductivity i s made. Transfers of the a c i d are repeated u n t i l constant readings are obtained. T h e d i s t i l l a t i o n apparatus has been tested and has been found to deliver anhydrous acid at a rate s u f f i c i e n t to fiIJ the cell in less than 1 min, When t h e s t i l l was charged with 99% HF, the conduct i v i t y of the d i s t i l l a t e corresponded to a maximum concentration o f water of approximately 1 ppm. Weighed samples of anhydrous LiOH are being taken through the procedure outlined above as Q means o f calibrating the apparatus. Calibration measurements have been completed over the range 0.03 to 0.25% H,O in HF,

149

ANP QUARTERLY PROGRESS REPORT Oxidetioar-[Reduction Titrations i n Fused Salts

A. S. Meyer, Jr. .4nalytical Chemistry D i v i s i o n

Experiments were carried out to determine whether the equivalence point of a titration i n which UF, dissolved i n a molten fluoride fuel solvent i s reduced by the addition o f small increments of an electrodeposition metal could be observed by one of the conventional electrametric techniques. The f i r s t system studied involved the reduction o f

in NaZrF,

by zirconium metal according to proposed reaction

4UF,

Zio.----+ 4UF, t ZrF,

UF, the

Since no insulating material which i s compatible

with molten NaZrF, was available, i t was not possible to devise a reference electrode for measurement of the potentials o f the solutions. Accordingly, polarized platinum electrodes were selected CIS indicating clectrodes. In order to obtain more inlgrmation about the nature of the polarization phenomena, the area of one of the electrodes was made approximately 80 times that o f the other. TAiis, since the current density at die s m a l l e r electrode W Q S 80 times as great n s that at the larger, the potential difference between the electrodes was approximately equal t o that of the smaller electrode. By reversing the direction of the polarizing current, it was p o s s i b l e to determine the potential a t the cathode and at the anode. T h e solution of UF, i n NaZrF, was contained i n a 25-ml platinum crucible whish was placed in the helium-filled quartz liner of a pot furnace, Agitation was accotnplished by means af a mechanic a l l y driven platinum stirrer which was constructed o f 0,05in.-dia platinum wire, Additions to the solution were made through a quartz tube which extended from the top of the liner to a point about 1 in. abave the nie!t.

In pure NaZrF,

the anode was inore strongly polarized than the cathode. At polarizing currents of 80 pa the potential of both electrodes exceeded 1 v. Approximately hr was required for the electrodes t o reach equilibrium. Platinum electrodes could not be appreciably polarized in NaFKF-LiF-base fuels. When UF, was added to he molten NaZrF5, the solution which potential decreased rapidly. For

2E. L. Colichmno, Polaro raphy in Molten Ammonium Fomzate, LRL-117 ( A p r i l 1959).

150

contained 1.25 mole % UF, the potential of both electrodes wets approximately 300 mv, w i t h n o significant difference between the potentials of the cathode and anode. During the early part of the titration of the molten fluoride mixture with zirconium metal, the potential of the cathode decreased to values less than 10 mv and remained constant throughout the titration. At the same time the potential o f the anode increased rapidly and attained potentials i n excess o f 1 v before the end point o f the titrution was reached.

When an equivalence point was reached, the potetit i a l of the anode decreased rapidly to about SQ mv and then decreased gradually on further addition of reductant. Only one break in the potential v s The curve i s shown t i t r a n t curve wa5 observed, Fig. 10.2.

_--

ORNL-LR-DWG

2 725

EQUIVALENTS OF ZlRCOYlUM PER MOLE 0'

UF4

Fig. 10.2. Titration of UF, i n Molten NaZrF, with Zirconiun1 Metal. Theobserved end points of two titrations occurred

at a weight of titrant which corresponded to one-

h a l f an equivalent of zirconium metal per mole of

T h i s stoichiometry indicates the reduction of t o a mixed oxidation state that corresponds t o the postulated formula U,F,.

UF,. UF,

Poiarogaaphic Studies in

Fused Ammonium Formate

D. L. Manning Analytical Chemistry D i v i s i o n

A. S. Meyer, Jr.

CoIichmun2 reported that when solutions of UCI, in molten ammonium formate were reduced a t the droppingmercury electrode, three we1 I-defined reduction waves w i t h half-wave potentials of -0.1 t o -0.2, 4 , 7 5 to -0.80, and -0.95 v (vs pool) were

PERIOD ENDlMG SEPTEMBER 70, 1954 Conversion of UF, and UF, to the Respective Chlorides with BCI,

obtained. The following reactions were postulated to explain these waves:

D. L. Manning Analytical Chemistry D i v i s i o n

A. S . Meyer, Jr. and

H e also reported that anhydrous hexavalent uranium gave no reduction wave u n t i l after an aging period during which the compounds were reduced by the ammonium formate to a lower valence state, probably quadrivalent. Since it appeared that t h i s technique could be appiied to the determination of tetravalent uranium i n the presence o f lorge quantities of trivalent uranium,

experiments were carried out to obtain

In preliminary polarograms of solutions of UF,. experiments i t was found that UF, and NaZrF,. and N a F - K F - L i F-base fuels dissolved readily i n molten ammonium formate at 125째C to form green solutions containing as much os 1 mg of uranium per m i l l i l i t e r . Although UF, alone was found to be much less soluble, UF, which had been fused with NaZrF, was reodi I y soluble. Similar relationships were found when molten ammonium acetute was used as a solvent. No U(IV) reduction w w e s were obtained when polarograms of solutions of UF, were recorded immediately after dissolution, and no reduction waves were observed in the polarograms of a sample of N a F - K F - L i F-base fuel that contained When these solutions were both UF, and UF,. aerated, a single reduction wave w i t h a halfwave potential of approximately -0.2 v was observed. The potential and the diffusion current of t h i s wave corresponded closely to those recorded immediately after the dissolution of an equivalent

in molten formate. Since the quantity of K,UO,F, U($V) waves may have been shifted to potentials outside the range of the solvent (0 to -0.9 v) by the c m p l e x i n g action of the fluoride ions, additional polarograms were taker] o f solutions conExcept for a small reduction wave t a i n i n g UCI,. at about -0.2 vp n o evidence? of reduction was found. The diffusion current of t h i s wave was increased by aeration to values corresponding t o those obtained For hexavalent uranium, and therefore &e increase was attributed to the oxidation 04 &IC/,, It appears that i n molten ommoniurn formate IPO oxidation state of uranium other than hexavalent i s reduced at he drapping-mercury electrode.

It was reported previously3 that, on the basis of free-energy calculations, UF, UF,, ZrF4, and NaF are quantitatively converted to the chlorides when at equilibrium with boron trichloride. Since t h e exchange reaction in organic solvents at moderate temperatures was found to be too slow for analytical application, tests were carried out with

gaseous BCI, at elevated temperatures. T h i s technique offered on additional advantage i n that the resulting chlorides could be separated by sublimation. When a sample of NaZrF,-UF, fuel was heated i n a stream of BCI, and helium at 350째C, the bulk of the zirconium was sublimed as ZrCI, within 90 min, and a l l but traces were removed after heating for an additional hour at 450OC. It has been reported* that uranium can be sublimed as UCI, by

treating UF, with BCI, at 600 to 65OoC, but in these tests only a small fraction of the UCI, was v o l a t i l i z e d when the UCl,-NaCl residue, which fused at about 45OoC, was heated to temperatures as high as 750째C. Analyses of the residues showed that when the conversion was carried out o t 40OOC for 90 min the uranium in the NaZrF,-UF, samples was quantitatively converted tu UCI, and that a l l the uranium remained i n the residue. When the conversion was carried out at 5oO째C, 2% of the uranium was volatilized, while at 750Dc, 4% was volatilized.

When UF, samples were converted to UCI, by t h i s method, significant fractions of the uranium were oxidized to UCI,, It i s believed that t h i s oxidation w i l l be eiiminated when a l l traces of o x i d i z i n g contaminants are removed from the gaseous reactants,

Solubility o f Tri- and Tetravalent Uranium FBuorides i n Fused NaAsCl

A, S. pjleyer, Jr.

J. Ross

Analytical Chemistry D i v i s i o n

In the continuation of experiments to determine

t h e optrmum conditions for the conversion of UF, _

_ ~ __~'A. S. Meyer, Jr., I3. I. Manning, and W. J. Ross, ,4NP @an Prog. Rep- June 10, 1954, 0RNL.-1729, p 110. 'V, P, Calkin5, Drrect Conversion 01 T F 6 to TCL,, CEW-IEC C-0.350.4 (Septa 23, 1986).

151

A N P QUARTERLY PROGRESS REPORT and UF, t o the corresponding chloride^,^ i t was noted that UF, was dissolved much more readily

than UF, by fused NaAICI,. Measurements o f the s o l u b i l i t i e s were carried out by equilibrating sainp l e s of the uranium fluorides with molten NaAICI,

i t s uranium content, which i s assumed to be UF,. I n these tests the reaction time was held constant a t 2 hr and the temperature was varied from 200 t o 600oC, with the results shown i n the following:

Solubilities of

and

mg of

2OOOC. Since UCI, was also found to be sparingly soluble at the lower temperatures, the above values are indicative of the true solubility of trivalent uranium i n t h i s medium rather than o f incomplete conversion of the fluoride to chloride. At temperatures lower than 200'6, molten NaAICI, i s expected to dissolve tetravalent uranium selectively from UF, which hos been complexed by potassium fluoride i n the fuel. Since a l l available UF, samples are contaminated w i t h tetravalent uranium, it has not been possible t o carry out accurate solubility measurements at t h e lower temperatures by relying on the determinat i o n of total dissolved uranium. An apparatus has been constructed for the determination o f microgram quantities o f trivalent uranium i n the filtrates by hydrogen evolution. T h i s apparatus, which i s based on the measurement of evolved gases at reduced pressures, i s now being calibrated against coulonietrically generated hydrogen. In order to carry out dissolutions at lower temperatures, LiCl-AICI fluxes which melt at approximately 120OC have been prepared. Oxidation of

UF, with Oxygen

W. J.

A. S. Meyer, Jr.

Ross

Analytical Chemistry D i v i s i o n

Preliminary investigations have been initiated to determine the extent of the oxidation of UF, by oxygen at elevated temperatures. The reaction i s assumed to be

4UF,

solution, -

+ O,+

3UF,

UO,

and the soluble portion i s analyzed for -

'W. J. Rodd and J. L, Mattern, ANP Quur. Prog. Rep. Mar. 10, 1954, ORNL-1695 p 111.

152

(OC)

(%I

200 3 00 500

UF, per gram of

NaAICI, were found at 200 and 175OC, respectively. However, the s o l u b i l i t y o f UF, decreased from 5 iiig/g of NaAICI, at 225OC to less than 1 mg/g at

Reactivity

Ternperoture

for periods up to 5 hr, under an inert atmosphere, followed by the separation o f the undissolved s o l i d s by f i l t r a t i o n and calculation of the s o l u b i l i t y from the concentration of uranium in the filtrate.

100 91 25

600

At 300째@ the reaction proceeds quantitatively as written. At higher temperatures, however, the concentration of soluble uraniuiii (UF, in ammonium oxalate) decreased rapidly from the theoretical content with increasing temperature, probably as u consequence of a further reaction of oxygen with

UF,.

The experiment was then repeated with UF, and At 2OOOC the entire sample was soluble oxygen. i n ammonium oxalate, as expected, which indicated no oxide formation. At higher temperatures, a black

insoluble residue was formed upon dissolution o f t h e sample i n oxalate solution. The amount o f residue increased at higher temperatures from 11% at 3OOOC t o 27% at 600OC. The major portion of the sample dissolved rapidly, and a y e l l o w f i l t r a t e was formed. It i s to be noted that Kraus6 reported that the following reaction occurs a t 8 O O T :

UF,

+ O,+

UF,

UO,F,

The yellow color of the filtrate substantiates t h i s reaction i n part, but there is no way o f accounting

for the black residue, obviously an oxide o f uranium. it i s known,7 however, that UF, i s converted quantitatively to U,Q, when ignited a t 900째C i n air. Petrographic and x-ray studies of the material w i l l be made. Further tests are planned at various reaction times and temperatures. Determination of Lithium, Potassium, Rubidium, NoF-KF(RbF)-LiF-Base Fuels

and Fluoride !Qn i n

J. C. White

G. Goldberg B, L. McDowell

Analytical Chemistry D i v i s i o n

A c r i t i c a l review of the onalvtical methods for

6c. A.

Technical Report 107 the Months 01 A p n l , 1943: Chemical Reseurch - General, !?17?$ 1 r 3 9 , 1944. 'J. J. Katz and E. Rabinowitch, T h e Chemistry o f U r m i u m , p 374, MeGraw-Hill, New York, 1951. Kraus,

PERIOD ENDING SEPTEMBER IO, 1954 ond fluoride ion i s being made to ascertain h o s e applicable to the analysis of N a F - K F ( R b F ) - L i F base reactor fuels.

Lithium. An indirect volumetric method for the determination of lithium i n the presence o f the large concentrations o f sodium and potassium and/or rubidium which are found in NaF-KF(RbF)-LiF-base

fuels was developed. The method was designed specifically t o provide a rapid, and yet precise, means o f determining lithiuni. The a l k a l i metals are f i r s t separated from uranium and then converted t o chlorides. L i t h i u m i s extracted as the soluble chloride s a l t by heating a t about 135OC with 2-ethylhexano18 u n t i l the s l i g h t l y soluble salts of sodium and potassium become "free-flowing" and n o longer c l i n g to the w a l l s of the vessel. These salts are removed by filtration, and then the f i l t r a t e i s diluted w i t h ethanol. The chloride i n the ~ the f i l t r a t e i s titrated by the Volhard m e t h ~ d ,and I ithium concentration i s caiculated. T h e coefficient o f variation i s 0.5% for 1 to 50 mg o f lithium. A topical report of t h i s work i s being written. Potassium and Rubidium. Since the concentration o f either potassium or rubidium i n NaF-KF(RbF)L i F - b a s e fuels i s o f the order of 30 wt %, the determination of potassium or rubidium by means o f the flame photometer has been neither sufficiently accurate nor precise. The conventional method for t h i s determination i s the perchlorate separation of potassium or rubidium perchlorate from the sodium and lithium salts in ethyl acetate. The method i s entirely satisfactory but requires considerable operator time. Therefore the opplication o f possible substitute procedures which are specific for potasA prosium and/or rubidium was investigated. involving the precipitation o f potassium cedure' bitartrate i n ethanol-water and t i t r a t i o n o f the solution o f the s a l t with standard base was tested. The method i s extremely simple ond rapid, and the bitartrate s a l t i s readily filterable. Rubidium behaves similarly to potassium, but cesium does not precipitote as the bitartrate salt under these conditions. Further work i s necessary t o determine the effect of acid and s a l t concentrations on the

8E. R. Caley md H. D. Axilrod, 2nd. Eng. Chem. Anal. Ed. 14, 242 (1942). 9 1. M. Kolthoff and E. B. Sandell, Textbook of Qumtitatzve

1947.

Inorganic A n d y s r s , p 477, MacMillan, New York,

"lbid., pp 414-15. ' ' A . F. levinsh and Yo. 8, 57 (1953).

Ozoi, j . A n d . Cbem. USSfE

s o l u b i l i t y of potassium bitartrate before o decision can bereached as to the a p p l i c a b i l i t y o f t h i s method t o NaF-KF-LiF-base fuels. A second method under consideration i s the use o f sodium tetraphenyl boron12 as a reagent for the direct determination of potassium, rubidium, and cesium. The potassium s a l t i s precipitated i n aqueous acidic solution, filtered, dried a t 1 10째C, and weighed as potassium tetraphenylborate. The method has the advantageof being simple and rapid, and i t has a favorable weiyht factor for potassium. The main interference in the use of tetraphenylboron i s the ammonium ion. The reagent i s expensive, but it i s readily available a t the present time. Preliminory work on the s o l u b i l i t i e s of potassium, rubidium, and cesium tetraphenylborates in nonaqueous solvents has shown some interesting differences in solubilities. The p o s s i b i l i t y of exp l o i t i n g differential s o l u b i l i t y as a means o f separating the salts i s being explored. Acetylacetone, diethylketone, and methyl i sopropylketone are three possible solvents for t h i s separation. The order o f s o l u b i l i t y in organic solvents has been estoblished as: KTPB > RbTPB 1 CsTPB (TPB = tetrapheny Iboron). Fluoride Ion, The determinatian o f fluoride ion i n fluoride fuels was previously conducted by the pyrohydrolysis l 3 procedure which was particularly fuels. A l k a l i metal successful for NaZrF,-base fluorides, however, are resistant t o pyrohydrolysis, and thus considerable time, of the order of 3 to 4 hr, i s required t o pyrohydrolyre completely 100 mg of N a F - K F - L i F - b a s e material. T w o alternative methods have been studied, therefore, for application to the determination of fluoride ion i n NaF-KF-LiF-base materiols. A procedure, which was formulated by ChiIton,14 was tested. In t h i s procedure the fluoride ion i s titrated w i t h a standard solution of aluminum (potassium alum), and the resulting change i n acidity i s recorded. Considerable d i f f i c u l t i e s have been encountered, the foremost of which i s the extremely small change in pH at the equivalence point. The method in i t s present form does not appear to be applicable to this work. T h e f e a s i b i l i t y of a spectrophotometric titration o f fluoride ion based on the decolorization of a

~..__.._ ......_._i_.

"W. Ceilmann and W. Gcbuukr, %. m a l . C,%ern. 139, 161 (1953). I3J. C. Wurf, W. D. Cline, and R. D. Tevebaugh, h a l . CfJeVL. 26, 342 (1954). I4J. A. Chilton, personal communication.

153

zirconium or aluminum complex, or lake, i s also being investigated. In t h i s method, the solution of fluoride i s titrated with a colored zirconium complex, and the change in absorbance o f the solution i s plotted as a function of the concentration of zirconium i n the complex. Zirconium alizarin sulfonate and zirconium eriochrome cyanine both react rapidly w i t h fluoride ion and w i l l receive further study. Preliminary results indicute that the reactions of these compounds with the fluoride ion A simple titration c e l l i s being are reproducible. designed.

O i l Contamination i n ARE Helium

G. Goldberg Analytical Chemistry D i v i s i o n T h e contamination o f the helium for the

systems LiF-KF-UF, or LiF-NaF-UF,, quarternary system Li F-NaF-KF-UF,.

fluoride was found to form a complex with

ARE with

360 pg/ft3 was found. A tower containing o f charcoal was placed i n the gas line, but i t had l i t t l e effect i n eliminating the oil. Further tests on the same tank car showed that bleeding the I ine before samples were taken markedly lowered the contamination level. For example, i n three helium,

70 in.,

successive samples of 15.5 ft3 of gas eech, the contamination decreased from 100 t o 10 pg/ft3. These tests indicate that bleeding the l i n e with 50 to 100 ft3 of gas before the equipment i s coupled i n t o the system w i l l effectively remove o i l from f i t t i n g s and valves. PETROGRAPHIC INVESTIGATIONS O F FLUORIDE F U E L

In was an NaF-ZrF, (53-47 mole 76) coinpound. addition, several loops containing N a F and ZrF, with reduced UF, have been examined, and no appreciable oxidation of the UF, could be found.

W. F. Vaughan Williams Analytical Chemistry D i v i s i o n

J. C. White

C. R.

Most o f the samples received during the past quarter were from NaZrF,and N a F - K F - L i F-base materials. Other types of samples that were analyzed included silver fluoride, chromium fluorides, iron fluorides, beryllium, sodium, and rubidium metals, Inconel, end alkali-metal carbonates. A revised procedure was established to determine UF, and lJF, i n NaF-KF-LiF-base fuels. Also, the method of White, Ross, and Rowonâ&#x20AC;? was adapted successfully to the determination of oxygen and metallic impurities in rubidium metal. A total o f 1,165 samples, involving 10,702 deA breakdown of the terminations, was analyzed. work i s given in Table 10.1. 15J. C. White, W. J. Ross, and R. Rowan, Jr., Anal. Cbem. 26, 210 (1954).

TABLE 10.1. SUMMARY O F SERVICE ANALYSESREPORTED

G. D. White, Metallurgy D i v i s i o n T. N. McVay, Consultant UF, i n alkali-metal fluo-

r i d e systems have shown that two phases e x i s t i n One i s red and i s near the KF-UF, system. 3KF*UF,, and the other i s blue and i s near KF-UF,. Only one binary phase i s present i n the NaF-UF, Lithium system, and i t i s probably 3NaF.2UF3. fluoride has not been found to combine w i t h UF, i n either the binary system LiF-UF,, i n the ternary

154

UF,,

w i t h a composition near ~CSF-UF,. Some work was done on the systems L i F - Z n F , and NaF-CsF, and no binary Compounds were found, Part o f a loop i n which NaF-ZrF,-UF, (50-46-4 mole %) had been circulated was received from Rattelle Memorial Institute, Samples tuken from the w a l l contained about 50% ZrO,, and the balance

SUMMARY O F S E R V I C E A N A L Y S E S

o i l from valves and f i t t i n g s was checked befare the gas a t the s i t e was used. The method used for these tests was to pass the gas through dual scrubbing towers contain i n g petroleum ether and then t o evaporate the ether to dryness and weigh t h e residue. In the i n i t i a l test on a tank car of

Petrographic studies of

ar i n the Cesium

Number of Samples Reactor

Chemistry

Fuel Production

Experimental Engineering Miseel Ioneou s

Number of Determinations

679 38 379 69

6,694

1,165

10,702

...__

395 3,406

207

Part 111

SHIELDING RESEARCH

11. SHIELDING ANALYSIS J. E. H.

P. B l i z a r d

Faulkner

E. Hungerford

F. H. Murray C. D. Zerby

Physics Division

Consolidated

H. E. Stern

Vu1tee Aircraft Corporation

Theoretical analyses have been made so that an understanding of the air and ground scattering measurements at the Tower Shielding F a c i l i t y could be obtained and the attenuation in the side w a l l of an aircraft crew s h i e l d could be determined. In connection w i t h the air and ground scattering analyses, special attention was given to the calculation of the reduction of air scattering due to ground interruption and to the effects o f multiple scattering. The gamma-ray slant penetration has been calculated for the side w a l l of an aircraft crew shield, and the variation of radiation intensity throughout the crew volume has been briefly explored.

T h e parameters and their variations that have been investigated to date are l i s t e d below: Thickness of Compton T h i c k n e s s of lead, in. Initio1 photon energy,

Carlo methods. The problem was programed for solution on the ORACLE and was arranged for investigating the effects o f the variations of a l l the parameters involved. T h e shield was taken as a composite slab o f two materials comprising a thick layer of Compton scattering material' followed by a thin layer of

lead. In each of many cases the i n i t i a l incident photons were taken as monoenergetic and incident

on the slab a t a particular angle w i t h the normal t o the slab. The stochastic process used the exacf physical analog o f the probability laws known to govern the l i f e o f a photon.

the range of energy considered, the Compton scatterer has approximately the same physical charoct e r l s t i c s r e l a t i v e to a photon passing through i t as does coricrete or polyethylene (CH2).

'In

3, 9 , 15

bo, 3/10,

2, 6

0, 30, 60

A l l combinations of these parameters were investigated. T h e following solution o f a typical problem indicates part o f the information obtained: Initial Conditions

C. D. Zerby

obtain a fundamental, but practical, knowledge o f the gamma-ray penetrations, the problem has been treated theoretically by using stochastic, or Monte

3 m c 2 units

I n i t i a l angle of incidence, deg

SLANT P E N E T R A T I O N O F COMPOSITE S L A B S H I E L D S B Y GAMMA R A Y S

The total radiation dose in the crew compartment of an airplane i s partially dependent on the gammaray f l u x penetrating the compartment shield. To

in.

T h i c k n e s s of Compton scatterer

3 in.

T h i c k n e s s of lead

>,o

I n i t i a l photon energy

6 mc

I n i t i a l angle of incidence

in.

deg

Results F r a c t i o n of i n i t i a l energy

1. r e f l e c t e d

0.0234

2 . absorbed in the Compton scatterer owing

0.318

0.0568

to scattering collisions

absorbed in lead owing t o scattering collisions

4. absorbed In l e a d awing

5. 6.

cat I i s ions

t o absorption

0.0515

penetrating shield

0.550

penetrating shield without being degraded i n energy

0.435

Energy build-up factor

1.265

The energy spectrum for the reflected photons was also obtained in t h i s typical problem. The spectra for the photons penetrating the shield were obtained in the angle intervals 0 to 15, 15 to 30, 30 to 45, 45 to 60, and 60 to 90 deg. These spectra w i l l be available i n a forthcoming report. _____ ...... ............. 2The electron density of the Compton scatterer was taken as the same as that for polyethylene. 3 M u l t i p l y mc 2 units by 0.5108 t o obtain MeV.

157

ANP QUARTERLY PROGRESS REPORT

where

A161 S f A T T E R l N G O F N E U T R O N S

g(a

I n an effort t o understand the air-scattered neutron intensities which have been measured at the Tower Shielding F a c i l i t y , a number of calculations have been carried out w i t h a variety of conditions and assumptions. The predominant difference between the current work and that which WQS done prev i o u s l y 4 i s that the interference of the ground has been taken into account. The ground Scattering i s calculated separately. I n an attempt t o obtain a better fit to thc data, such aspects a s anisotropy of scattering and slant penetration at the detector tank (crew shield) have been taken i n t o account i n some of the work. Furthermore, measurements have been made o f the ongular distribution of the radiation leaving the reactor shield, and these data have been used i ~some ? o f the calculations. Because the importance of the multiply scattered neutrons has oppeared to be greater than was anticipated, considerable effort lins been made to extend the calculations to beyond the singly scattered case. Some results are reported f a r two scatterings, and a general method has been developed for a l l orders of scattering.

/(a)

+ 0.3

= a,

cos

1; b, cos a;

/3)

c, cos2

Murray

Single !sotrapis Air Scattering in the Presence of the Ground (Unshielded Detector)

J. E. Foulkner

A calculation was made of the reading of an isotropic detector caused by first-scattered neutrons i n air as the distance above the ground was varied. The f o l l o w i n g assumptions were made: 1. The source and detector have the same a l t i tude,

2. 3. 4. 5.

T h e ground i s an i n f i n i t e plane, The source is isotropic. The scattering i n air i s isotropic. There is no energy degradation on a i r scatter-

6. The attenuation i s pure inverse square. With these assumptions the reading on the detector may be shown to he proportional to

the at-

1 D

F o r a convenient machine cnlculation,

where D i s the separation of source and detector, h i s the altitude, and / may be expressed i n closed 4See, for example, H. Goldstein, Chop. 2.9, p 831, i n F. Hogerton and R. C. Grass, Technical Information Service, AEC, 1953.

Reactor Handbook, ed by J. 5 -I-. H.

Single Anisotropic Air Scattering of in the P r e s e n c e of the Ground (Shielded Detector), 'ORNL CF-548-1041(to be issued). Neutrons

Murruy,

for

158

The angle i s the angle between the vertical plane taken as the face of the crew compartment and the plane triangle formed by the source, the scattering volume, and the detector, The results of t h i s calculation5 w i l l be published later.

tenuation i n air WQS neglected. The detector was placed w i t h 10 cm of water between it and the face of the crew shield which, when extended, contained the reactor source. The beam from the source was about an axis through the detector, i n theory, and contained teinis cos a and cos2 q c1 calculation W Q S made separately for the terms 1, cos 4 and cos2 a in i ~ i eexpansion of the source flux. If D i s the reactor-detector distance and h the height of both above the g~ound, the flux at the detector i s expressed as the sum of the terms:

and

Ing.

Single Anisotropic Air Scattering in the Presence of the Ground (Shielded Detectcab

F. H.

+ /3)

O 5 4 5 n/2,

PERlOD ENDING SEPTEMBER 70, 1954 form in terms of the Spence functions.

A plot of

formulas,

dw, do, represented by a general formula, S ( Z o , for the probability that a particle arriving w i t h i n a -* u n i t volume a t P and moving in a direction W, w i t h i n the s o l i d angle d u o w i l l be scattered into the s o l i d angle d o in the direction Z. L e t Q and P be the points ( x o , y o , z Q ) and (x,y,z), Fig. 11.2, and l e t f,(Q,Z,) do, dvQ be the number of particles which have suffered their mth c o l l i s i o n in t h e volume element d V g = dx, d y , dzQ ond emerged i n the s o l i d angle doo w i t h a direction Go. If the next c o l l i s i o n occurs in a volume element d V , and i f n i s the total cross section, the number

Formulas for M u l t i p l e Scattering in Uniform Medium

H. Murray

Method for Anisotropic Source and Scattering. T h e Fourier analysis can be applied t o multiple scattering i n such a manner a s to lead to e x p l i c i t U N C L ASS1 FI E D

ORNL

- 1.R - 0 W G

but

which do net require a knowledge of group theory for their application. Here the scattering law i s

/ ( h / D ) has been made (Fig. 11.1) by using the reA report that sults of ORACLE computations. gives details o f the calculation i s being prepared.

similar to some derived by Wignerâ&#x20AC;&#x2122;

2645

6J. E. Faulkner, Single Isotropic A i r Scattering o f Neutrons in the P r e s e n c e of t b e Ground (Unshielded D e t e c t or), ORNL CF-568-96 ( t o be issued).

â&#x20AC;&#x2122;E.

P. Wigner, P h y s . Rev. 94, 17 (1954).

UNCLASSIFIED ORhL- LR -0WF ?646

INTENSITY NORMALIZED TO 1 AT INFINITY

-.I

0.2

0.f

.................

...............

1 .O

2 .0

3 0

2 h/D

4 .O

Variation of Radiation f n k n s i t y w i t h Fig. 11.1. A l t i t u d e for a Constant Source-Detector Separation Distance.

Fig. 11.2. culation.

Geometry for Multiple Scattering Cal-

159

ANP QUARTERLY PROGRESS REPORT of c o l l i s i o n s in d V , for which the p a r t i c l e emerges i n the s o l i d angle d a w i t h direction fmtl

(P,&) d V , rtw

$$$ X(d0,&9e - a r r 2 dr dw,

where r =

[ ( x - xo)2

Here ,Amo i s the s o l i d angle subtended at

0 by

- y o )2 +

- z o )2

2 i s equal t o

f,,,(QId,) 6oo

d V , or by d V , / r 2 ; by putting r 2 & dfijo = ~ V Q then ,

L e t ySt(i;)bethe normalized spherical surface harmonics. For convenience, these functions can be ordered

in a single series Y c :

Yo,o; Y ,

Y 1,-1

'. Y 2

Y1,o; Y,

Y1,,; Y,

Y2,-2; Y5

Y2,-1;

...

Let fm(Q,

From Eg.

z) =

f,,,(Q)

yc(d)

Introducing a Fourier transformation of both sides and u s i n g

k,x

k2y

k,z

= kr cos (k,t)

With

gives

where

dV,

= d(u

- xo) d(y - y o ) d(z - zo)

The direction 2 , i s the direction from Q t o P , s o that, as indicated, the transform o f the r i g h t s i d e of Eq. 1 breaks up into a sum of products of Fourier transforms. The last integrals on the r i g h t can evidently be evaluated independently of the position 0. B y expanding S i n the form

SG,,

A q p Y p ( G o )Y , ( 3

P E R I O D ENDING SEPTEMBER 70, 1954 these integral 5 become

--and

T h e source function may be ani sotropic; however,

i f an isotropic source i s present or i f the scattering

Let I

and then

Z: Fm+1,? yi(0)+

r),

F,,z,c

1 A y p l P c Y , ( w-+) P

Equating coefficients of identical Y ' s gives I

where

law possesses axial symmetry, some simplifications occur. When the scattering density has been obtained by a Fourier inversion, one further integration gives the flux into an arbitrary u n i t volume from any direction; i f the detector s e n s i t i v i t y depends on direction, the total scattering to be counted at any point P ' i s represented by o convolution and can be calculated by another Fourier transformation and inversion. The method i s to be applied to air scattering measurements taken a t the TSF and w i l l make possible subsequent calculations for unusual reactor shield shapes,

Thus, i f Fm,c i s represented as a vector (F,,l,0,F,72,, , F m , 2 ,

.. .) for al I rn,

Method for isotropic Source and Scatfering. If the scattering i s isotropic after each collision, the formula to be used for calculating the total flux at the point P after the nth c o l l i s i o n i s a convolution:

F p ) = us J j-

1Fn- ,(P) d V Q +(Q,P)

161

AM? QUARTERLY PROGRESS REPORT w i tll

For isotropic re-emission from the ground, -aRpQ

N ,A

+(Q,P) = e

4.nR2pQ

87i2b2

/(a,)s i n a da

Since the Fourier transform of a convolution i s the product of the transforms of the individual functions, for an isotropic source w i t h no energy l o s s a f t e r each c o l l i s i o n the transform TF-n becomes

..........

with

........._

...............

For cosine re-emission from the ground,

In taking the inverse of 'i'F,, the path of integration i s deformed i n t o t h e closed curve about the negative -0. Calculations of these scattering functions for the first- and secondscuttered f l u x g i v e

real a x i s to the left of

F1 =

1.635

F 2 = 0.19167 ,

t o t a l source strength (neutrons/sec), reflection (a1bedo),

These figures are to be m u l t i p l i e d b y 1/4n2 A D , for X = 130 meters, to obtain t h e number o f f a s t neutrons a t a distance D feet from the source. GROUND S C A T T E R l N G

A. Simon

OF N E U T R O N S

H. E. Stern

tained:

162

the

ground

relative source strengfh per steradian at ungle a between emergent ray and source axis,

h = height of source and receiver above the ground,

A n investigation i s being carried oiut t o derive

formulas for the ground-scattered neutron f l u x t o be obtained froin Q general source distribution. The model used i s that of a source whose axis, or direction of maximum intensity, forms an arbitrary angle 0 w i t h the source-receiver axis but l i e s i n The the same horizontal plane as the latter. differential source strength i s assumed to be a function of only the angle between the emergent ray and the source axis. The receiver i s isotropic under the assumptions of no air attenuation and a constant albedo for the ground. The f o l l o w i n g expressions for the f l u x at the receiver are ob-

coefficient

sin' a c o s 2

-2

o =

(? -

cos O e o t a s e c

angle between source axis and sourcereceiver axis,

D = separation distance between source and receiver.

PERIOD ENDING SEPTEMBER IO, 7954 Alternatively,

if a single-scatter approach w i t h attenuation by a emov oval" mechanism i s used, the fol-

lowing formula i s obtained:

where

scattering cross section, removal cross section. A detailed report o n t h i s c a l c u l a t i o n is being prepared.

FOCUSING O F R A D I A T I O N I N A CYLlNDRlCAL

ylr

......

ORNL-LR-DWG

2647

CROSS SECTION OF CREW COMPARTMENT

C R E W COMPARTMENT

J. E.

Faulkner

A study of the focusing of radiation in a cylindric a l crew compartment has been made, w i t h particular attention being given t o the case w i t h the following conditions : 1. The cylinder i s infinite. 2. The surface radiation density i s constant along the inner cylinder walls. 3. The angular distribution of the radiation depends only on the angle between the direction of emergence and the normal t o the cylinder. 4. The attenuation inside the crew space i s pure inverse square. 5. The detector i s isotropic. Under these conditions the angular d i s t r i b u t i o n may be represented as an isotropic component plus a series i n odd powers of the cosine. The contribution of each term at a given point in the component in the series may be expressed i n an even polynomial i n x (see Fig. 11.3), where x i s t h e

x=h/R

1 .o

STERADIAN IN T h E

NORMOL DIRECTION

0.8

0.6

6 b-

0.4

0.2

distance of the given point from the a x i s . o f the cylinder divided by the radius of the cylinder. For pure cosine distribution the reading of an isotropic detector i s independent of position in the crew compartment. 'A. Simon and H. E. Stern, Some Cuiculutionul Methods jor Air m d Ground Scattering, O R N L CF-54-8-103 (to be i s sued).

0 0

0.2

0.6

0.4

0.8

i.0

Fig. 11.3. Variation of Radiation Intensity w i t h Distance from Center of Crew Compartment.

163

ANP QUARTERLY PROGRESS REPORT

G. T. Chapman d, M. M i l l e r W. Steyert D. K. Trubey Physics D i v i s i o n

.J. 5. Des

Pratt. and Whitney Aircraft D i v i s i o n Preliminary

work

for

further

circulating-fuel

reflector-moderaked reactor shield experiments at the L i d Tank Shielding F a c i l i t y tinued w i t h irrudiution in the Reactor of a sample of the

(LTSF) has conORNL Graphite

UF,-C,F,,

mixture

that may be used to simulate the reactor fuel i n the shield mockup. Also, a new effective removal cross section for carbon has been obtained for the case iii which the carbon i s distributed uniformly throughout the shield. Thermal- and fast-neutron measurements have

been mode around an array of three of the GE-ANP h e l i c a l a i r ducts, and a 35-duct array i s being assembled far further measurements. D i f f i c u l t i e s i n fobricoting the enriched uranium plate for the new LPSF SOUPCE have delayed the completion of t h i s project; however, it appears at present that It i s these d i f f i c u l t i e s hove beer1 surmounted. anticipated that the installation w i l l be completed w i t h i n a month. R E F L E C Y O W-MOB E R A T E D W E A C T O W A N D SWIEI-D M Q C K U P TESTS

D. K. Trubey

J. 6. Dee

\N. Steyel't

A second series sf mockup tests for the circulating-fuel reflector-moderated reactor (RMR) and

For t h i s s h i e l d is being initiated at the LTSF.' series Q larger tank (approximately a 6-ft cube) has

been constructed t o hold a l l the dry components of the configurations, as w e l l as ow expansible p l a s t i c bag for containing borated water. The beryllium blocks t o be used i n the mockups w i l l also be placed i n p l a s t i c containers for additional protection. The dry tank face adjacent to the source p l a t e has a k-in.-thick Inconel window, which corresponds t o the RMR core shell. In order to determine the effect of the lnconel on the gamma-ray dose, 1-

F o r f i r s t s e r i e s s e e C. L. Storrs et al., A N P Quur, Prog. Rep. S e p t . 10, 1953, ORNL-1609, p 128.

164

gamma mmsurements were taken in pure water in the tank. The dose W Q S higher by Q factor of 2 than thrrt normally observed in the LTSF, and the increase agrees closely w i t h the calculated 9-Mev capture gamma-ray dose from the Inconel.

In addition to die presently planned s t a t i c f i s s i o n

source t e s t s for RMR designs, a dynamic f i s s i o n source t e s t i s being considered for measuring the

sodium a c t i v a t i o n from delayed neutrons released i n the heat exchanger and the attenuation of gamma rays from short-lived f i s s i o n products. A l i q u i d contairiing 20 wt % Uc-',. being considered i s C,F,, In cooperation w i t h the Radiation Damage group of the Solid State Division, Q sample containing natural uranium was irradiated in a fluorinated nickel capsule in the OHNL Graphite Reactor for an integrated f l u x of 1617 nvt, which compares w i t h an integrated f l u x of I O i 5 nvt for the enriched fuel. After the v o l a t i l e pioducts hod been removed by vacuum d i s t i l l a t i o n , a precipitate containing most of the radioactivity and a large part of the uranium was found i n the irradiated c ~ p s u l e s ;thus, t h i s fuel mixture could not be used for long exposures. A similar capsule t e s t i s being prepared for a shorter exposurer and alteriiate solutions are being explored. E F F ECTllVE R E M O V A L CROSS SECTION O F CARBON

TriJbcy

Measurements of the removal cross section of carbon have been made in a continuous carbon

medium obtained by dissolving sugar (C,,H,,O, i n water, The solution (density = 1.312 k 0.001 g/cm3) contained 44.2 wt % sugar, which gave 0.354 g/cm3 of carbon and a hydrogen and oxygen

density that W Q S 95% of that o f p l a i n water. It was contained in ei large tank that had n &-in.-thick lnconel window on the source side. The therma I-neutron flux, the fast-neutron dose, and the gamma-ray dose are shown as functions of distance from the source in Figs. 12.1, 12-2, and

PERIOD EMDlNG SEPTEMBER 70, 7954

Fig. 12.1.

tion.

I, DiSrANCE

80 IO0 120 FROM SO?IRCC ( c m )

160

140

Thermal-Neutron Flux in Sugar Solu-

12.3, respectively.

P l a i n water curves are

also

10-3

shown for comparison. The gamma-ray water curve i s higher than the normal LTSF water curve because of i h e high-energy capture gamma ray from the lnconel window of the tank (see preceding discuss ion). Since the medium contained almost as much hydrogen and oxygen as does p l a i n water and contained them in the same ratio, nogeometric correction was made in calculating t h e effective removal cross section, ur. The average uT for the range of

140 cm from the source was 0.750 barn. T h i s compares w i t h a value of 0.81 i 0.05 barn from an LTSF

measurement behind a slab of graphite that contained 51.3 g/cm2 of carbon, which corresponds to

Fig.

............... .... 80 90

______.._.....

100 1.10 (20 I , DISTANCE FROM SOURCE (cm)

12.2. Fast-Neutron Dose

430

in Sugar Solution.

the sugar-water solution a t 145 cm, In order t o observe the neutron attenuation in a medium removed from the source, the sugar-water solution i s being placed i n a 36-in.-long aluminum tank located 48.2 cm from the source. Thermalneutron, fast-neutron, and gamma-ray measurements The measurements at w i l l be made in the tank. the interface w i l l indicate the difference in neutron age between water and the sugar solution for the more penetrating o f the f i s s i o n neutrons.

165

ANP QUARTERLY PROGRESS REPORT helical air ducts a t the LTSF. The ducts (Fig. 12.4) were fabricated from f l e x i b l e steel conduit shaped around a 9-in. core. After removal of the

CE-ANP HELICAL AIR D U C T E X P E R I M E N T A T I O N

J. M. Miller Thermal-neutron measurements have been made around 3-in.-dia b y 4-ft (developed length) GE-ANP

core, the ducts were stiffened by Fibcrglas wrapping.

sprr 2 6 5 1

ORNL-LR-OWG

1 i

1 60

~ ~~

100

140

120

130

I,DISTANCE FROM SOURCE (ern)

Fig. 12.3.

-1

110

150

160

I 0-1

Gamma-Ray D o s e i n Sugar Solution, 5

I20

130

I40

I, DISTANCE FROM SOURCE

F i g . 12.4.

166

GE-AWP Helical Air Duct.

I50

160

(Cm)

Fig. 12.5. Thermal-Neutron F l u x Measurements Beyond One and T h r e e GE-ANP H e l i c a l Air Ducts H o r i z o n t a l Traverses.

PERIOD ENDING SEPTEMBER 70, 7954 Two configurations, a single duct and a triangular

array of three ducts (5 in. from center to center), were placed i n the LTSF with one end adjacent to A comparison of the thermalthe source plate. neutron f l u x beyond the two configurations is given i n Fig. 12.5, and Figs. 12.6 and 12.7 are thermal-

neutron traverses parallel to the source plate. An array of 35 ducts is to be placed in a medium of

steel Raschig rings (35 vol %) and borated water for a study of the effect of the addition of ducts to gamma shietding.

5 1......

10-1 -70

50 70 y , VERTICAL DISTANCE FROM SOURCE AXIS (cm) -50

Fig. 12.6.

-30

-10

Thermal-Neutron Flux ffleasurements Beyond One GE-ANP H e l i c a l Air Duct Vertical Traverses.

-70

-50

. j ........ j

-30

!.I..

-IO

y , VERTICAL D l S T A N L t FROM SOCINCL AXIS [ c m )

Fig. 12.7. Beyond T h r e e Traverses.

Thermal-Neutron

GE-ANP Helical

Flux Measurements

Air Ducts

- Vertical

167

...

ANP QUARTERLY PROGRESS REPORT

SHIELDING FA R. G.

Cochran

F. C. Maienschein

G. M. Estabrook K. M. Henry E. B. Johnson J. D. Flynn T. A, L o v e M, P, Haydon R. W . P e e l l e Phy s ics Div i s ion At the Bulk Shielding F a c i l i t y ( W F j mensurements were made of reactor radiations through t h i c k slabs of graphite. I n t h i s experiment the fast-neutron spectrum through graphite and Q removal cross section of carbon were a l s o determined, T h e l i g h t to be given o f f from a nucleor-powered airplane has been further investigated, and some quantitative measurements are reported, I n addition, ci method of determining the power of the ARE by means af fuel activation techniques i s

'Thin graphite shims were added t o f i l l the tanks, which were sealed by heliarc welding and were pressure tested for leaks. The total aluminum w a l l thickness for each tank was 1.3 cm.

described.

it replaces. Since the presence of the graphite perturbed the neutron f l u x i n the reactor cote rather severely, the neutron f l u x distributions and thus the reactor power were determined i n the usual way by means of eobalt foils.

R E A C T O R WAPiATlONS T H R O U G H

aIF: G R A P H ~ T E Cochran J. D. F l y n n SCABS

R. G.

K. M. Henry

G. M. Estabrook

Measurements have been completed for determining the attenuation of large thicknesses of graphite next to c( reoctor. These measurements are of interest for evaluating a graphite reflector a s n shield component, and they also provide a direct comparison w i t h LTSF determinations of the carbon removal cross section. Graphite thicknesses of 1, 2, and 3 ft were used, and the usual gamma-ray, thermal-neutron, and fast-neutron dose measureIn ments were made hehind each slab thickness. addition, the fast-neutron spectrum (above 1.3 Mev) through 1 ft of graphite was measured. A graphite slab 1 ft thick and one 2 ft thick were

constructed, and the 3-ft thickness was obtained by strapping thesc two slabs together. When strapped together, there wos no p o s s i b i l i t y o f watoa getting between the slabs, Each slab wos constructed Inrge!y o f graphite blocks, 4 in. x 12 in. x 5 ft, stacked in aluminum tanks in such a way that there could be no streaming of radiation through cracks extending through an entire slab thickness.

' T h e d e t a i l s of t h i s experiment will be reported i n a memorandum by R. G. Cochran e t a/., Reactor Radz-

atzons Through Slab7 o/ Graphite, O R N L

(to be issued).

CF-54-7-105

Bulk Shielding Reactor loading No, 26 (Fig. 13.1)

was used. T h i s configuration i s very similar to looding 22, which has been studied i n considerable detail for other experimentsI2 but loading 26 uses

l e s s fuel to compensate for the presence of the graphite, which i s a better reflector than the water

2R. G. Cochran e t al., R e a c t i v i t y Measurements w i t h t h e Bulk Shielding Reacior, ORNL-1682 (to b e issued).

E X T O 7 GR D PLATE

W b E R PEkCECTOR

Fig. 13.1. Reactor,

?t4

ELEMENT

CONTRO.

ROD

26 e ~ fthe Bulk Shielding

P E R I O D ENDING SEPTEMBER 70, 7954

T h e thermal-neutron f l u x (Fig.

13.2)

was measured

behind each of the three graphite slab thicknesses to a distance of 300 crn from the reactor face, Measurements were made w i t h a 3-in. U235 fission

chamber, an %in. BF, counter, and a 124-in. 1 BF, counter, and the data were normalized t o indium foil data. Fast-neutron measurement5 (Fig. 13.3) were made w i t h a three-section neutron dosimeter. Gamma-ray measurements (Fig. 13.4) were taken

behind each of the three graphite slab thicknesses, both to determine the magnitude and relaxation lengths of the gamma-ray dose through graphite and to ensure that none of the neutron detectors was being used i n a too high gamma-ray field, which would have caused their readings t o be high. The fast-neutron spectrum from 1.3 to 10 Mev was measured with the BSF fast-neutron s p e ~ t r o m e t e r . ~ A spectrum w i t h reasonable s t a t i s t i c s could be obtained only through the 1-ft graphite slab; the neutron intensities through the 2- and 3-ft sections T h e spectrum measured with the were too low, -.

3R. G. Cochran and K. Henry, A Protun Kecozl T y p e Fast-Neutron Spectrometer, ORNL-1479 (April 2, 1953).

F i g . 13.2. Behind

Thermal-Neutron Flux Measurements

AGHT Graphite.

.....!. 1

.....

ose Measurements Be-

169

ANP QUARTERLY PROGRESS REPORJ 41%

ORNL-LR-DWG 2657

5 0 - c C ION

Fig. 13A,

Gamma-Ray Dose Measurements

hind AGHT Graphite.

1-ft graphite thickness between the reactor face and the end of the spectrometer collimator i s shown i n Fig. 13,5. The spectrum measured w i t h the end of the collimator against the reactor face i s also s t1own. Removal cross sections have been calculated for each thickness of graphite by use of a method suggested by E. P. Blizard.4 The resulting cross sections are 0.82 barn/atom for the 1-ft slab, 0.84 b a d a t o m for the 2-ft slab, and 0.80 barn/atom for the 3-ft slab. These removal cross sections are i n goad agreement w i t h measurements made on graphite a t the LTSF. R E A C T O R A I R GLOW

R. G. Cochraii K. M. Henry

T. A. L o v e F. C. Maienschein

R. W. P e e l l e

Attempts t o theoretically determine the amount of v i s i b l e l i g h t which may surround a nuclear-powered airplane in f l i g h t have resulted i n widely differing

~ a l u e s . ~ 'Therefore ~ an experiment was performed at the BSF t o provide an experimental basis for future estimates. The end of an a i r - f i l l e d aluminum periscope tube was placed a t the reactor face, and the amount of l i g h t produced in the tube was measured by a photomultiplier w i t h spectral response similar t o that of the average human eye. Other relative measurements were token w i t h a photomultiplier which was sensitive chiefly in the blue and nearultraviolet range. The latter measurements are plotted i n Fig. 13.6 as functions o f the air pressure i n the tube. Measurements w i t h argon i n the tube demonstrate that nei ther the approximate amount of l i g h t produced nor the exact spectrum emitted i s strongly dependent on the atomic number of the gas. It i s also interesting t o note that the l i g h t production i n a given volume of air appears t o have a maximum a t a pressure corresponding to an altitude of about 30,000 ft. It i s demonstrated i n Figs. 13.7 and 13.8 that t h e air glow i s largely caused by gamma radiation rather than by neutrons. Figure 13.7 shows the decay of the l i g h t plotted along w i t h the decay of the reactor gamma ion chamber current just after reactor shutdown. The attenuation by water of the radiation which produces the air glow i s shown i n Fig. 13.8. T h i s attenuation rather c l o s e l y follows that o f gamma rays. The photomultiplier was used t o compare the quantity of l i g h t given off i n the air-glow tube w i t h that from a small tungsten lamp mounted a t the reactor end of the ti~be, T h i s comparison showed

that 7.2 x lom5 lumen was given off b y the glow for c1 reactor power of 100 kw and atmospheric pressure. Presumably, the amount of light should be proportional t o the integral of the gamma-ray dose rate over the volume of the air in t h e measuring 10 tube. T h i s integral was estimated to be 1.1 x 10 (r.cm3)/hr. Therefore the effective l i g h t production J ~air ~ is per u n i t V O ~ L of

6.5 x

(Iumen/cm3)/(r/hr).

If i t i s assumed that all the l i g h t i s given off a t 4E. P. B l i z a r d , Procedure /or Obtaining E / / e c t i v e Rernoval Cross T e r t z o n s f r o i n L i d T a n k Data, QRNL CF-54-6-164 ( J u n e 22, 195.1)5T. A. Welton a s quoted by C, E. Moore, Visual Detectability o/ Azrcralt at N i g h t , LP.C-15, p 24 (Aug. 14, 1953). 6J. E a Faulkner,_Vrsible 1.1ght Produced in A i r Around Reactors, O R N L Lt -54-5-99 (to be i s s u e d ) .

PERlOD ENDING SEPTEMBER JO, 7954

1 O6

E-7

I il

........

...............

C0L.I.I f

................

...........................

___-

.......

M A TOR A G A I N S T REACTOR

...............

........

1 o4

to3

Fig. 13.5. Fast-Neutron Spectrum of the Bulk Shielding Reactor

bseaved Through

1 f t o f Graphite.

171

ANP QUARTERLY PROGRESS REPORT

ORNL-LR-DWG

2659

EQUIVALENT A L T I T U D E ( f t )

60,000

10.‘

10-2

40,000

20,000

1‘ 0

PRESSURE (cm Hg)

Fig. 13.6. Variation of Glow with Pressure.

4400

K, which i s i n the blue region,

7.2 x lO-’lumen

1,8 x 1 0 - 7 w a t ~o f v i s i b l e light.

The amount of air

glow can then be characterized a s the fraction of energy transformed into v i s i b l e I ight:

/ = -

w a t t s o f v i s i b l e l i g h t ii) tube watts of energy dissipated in tube by gamma r a d i o t i a n

1.9 x -

3.2 x 10-2

6 x

for atmospheric pressure.

T h i s experimental valus nf l i g h t y i e l d i s i n disa5 yreament w i t h the theoretical estimates of Welton (0.0.5) and of Fcaulkner6 (2 x In neither of these calculations considered.

172

was

the

effect of pressure

T h e value o f L may be used t o calculate the l i g h t emission from a t y p i c a l nuclear-powered

aircraft. For a reactor power of 200 Mw and a reactor shield of 147 cm of water (similar t o the s h i e l d design of the 1950 ANP Shielding Board’), the BSF measurements indicate that the gamma-ray dose just outside the edge of the shield would be 1.36 x lo6 r/hr with ri relaxation length of 22 cm. The dose integrated over spoce outside the shield M u l t i p l y i n g t h i s by L, i s 1.2 x 10’’ (r.cm3)/hr. t h e luminosity would be 75 lumens, which would have roughly the equivalent brightness to the eye o f CI 10-w (+7 w) incandescent light i f the aircraft were far enough away t o appear CIS o point source ....... .............. ...7 R e p o r t v/ the Shielding Board / o r the Aircralt Nuclear Propulszon P r o g r a m , ANP-53 ( a c t . 16, 1950).

PERIOD ENDING SEPTEMBER 10, 7954

m2 6 6 i

ORNL-LR-DWG

TIME (rnin)

DISTANCE FROM HtACTOR ( c m l

Decay of Air Glow and Reactor Gamma

Fig. 13.8. Attenuation in Water of Radiation Causing Air Glow.

o f light. T h i s large uncertainty i s primarily attributable to the low s e n s i t i v i t y o f the eye-equivaA more detailed discussion l e n t photomultiplier. of these preliminary values w i l l be given i n a forthcoming report. 8

T h e a c t i v i t y resulting from the irradiation of fissionable material i s not easy to predict because o f the large number of isotopes for which calculations would be required. However, for identical flux, exposure time, and waiting period after exposure, the specific a c t i v i t y o f two samples should be the same. Furthermore, a t low neutron fluxes the a c t i v i t y should be proportional to the flux.

Fig. 13.7. Rays.

FUEL A C T I V A T I O N M E T H O D F O R POWER DETERMINATION O F THE ARE

E. B.

Johnson

When the ARE goes i n t o operation, many measurements o f i t s performance w i l l be made. Among them will be the power level a t which it operates, which will be measured i n more than one way. A method suggested by J. L. Meem i s based on the measurement of the r e l a t i v e a c t i v i t y of fuel samples exposed i n the ARE and i n a known flux i n another reactor (BSR). An experiment has been initiated a t .the BSF t o implement t h i s suggestion.

8F. C. Maienschein et al., iMeusurc7nents o/ u l?t?acto~Induced A i r Glow, O R N L CF-54-9-1 (to be issued).

It i s contemplated that the ARE w i l l be operated a t a nominal power level of approximately 1 w for 1 hr, that it w i l l then b e shut down, and that a sample of the irradiated fuel w i l l be withdrawn for counting and analyzing.

The

ARE experimental

progrom w i l l then proceed as scheduled. A decay curve w i l l be run on t h i s fuel sample from the ARE and compared w i t h the decoy curve obtained on a similar fuel capsule irradiated for the same length of time i n the BSR i n 01 known neutron f l u x o f about the same magnitude. From the r e l a t i v e a c t i v i t y of t h e two samples, the operating power o f the ARE

173

ANP QUARTERLY PROGRESS REPORT can be determined readily. parison i t will be necessary content of the sample and of the energy per fissione9 A

For the power comto know the uranium the ARE, as w e l l as s m a l l correction w i l l

9J. L, Meein, L. B. Hollond, and G. M. McCummon, Detemznatzon of t h e P o u e r o/ t h e Bulk Yhzeldrng R e a c t o r - Par! 111. M e a s u r e m e n t of the k n e r g y R e l e a s e d p e r Fzsszon, ORNL-1537 (Feb. 15, 195.2).

174

be necessary for the local depression of flux b y

the uranium sample. T h i s factor has been estirtiated Details for the applib y W. K. Ergen t o be 0.80. cation of t h i s method are given i n a separate 10 report.

''E. R. Johnson, Fuel Actzvatzori Methods f o r P o w e r Determination o/ the A R E , ORNL CF-54-7-11 ( i n press).

PERiOD ENDING SEPTEMBER 70, 1954

14. TOWER WELDING F

T. V.

L. 6 .

C. E. C l i f f o r d Blosser J. L. H u l l Holland F, N. Watson Physics Division

Q. L. Gilliland,

General E l e c t r i c Company

M. F. Valerino, National Advisory Committee for Aeronautics J. Van Hoomissen, Boeing Airplane Company

T h e Tower Shielding F a c i l i t y (TSF) experimental program has, thus far, included measurements of ground- and air-scattered fast neutrons and the development of a new procedure for the determination of the power of the reactor. T e s t s on the GE-ANP R-1 divided-shield mockup have been started. FAST-NEUTRON GROUND AND AIR SCATTERING MEASUREMENTS

T. V. Blosser D. L. G i l l i l a n d

Van Hoomissen

F. N. Watson

T h e performance of neutron and gamma-ray air scattering experiments that are free from an excessive background of ground-scattered radi ation Therefore, i s a primary objective of the TSF. measurements of the scattered fast neutrons as a function of reactor-detector altitude were necessary t o determine the contribution o f groundscattered radiation t o the total flux, particularly at the maximum altitude. These measurements w i l l h e l p to indicate t h e magnitude of the yroundscattered neutron background t o be expected in future differential experiments, and they w i l l a i d i n an understanding of the variation of ground and air scattering as the ground i s approached. Measurements of t h e thermal -neutron distribution were taken i n the detector tank, which i s ess e n t i a l l y a 5-ft cube of water and which was, for these experiments, situated 64 ft from t h e reactor tank. The reactor wos placed at an angle 8 of 330 dey from the d axis (Fig. 14.1), and a BF, counter was moved along a lrne normal t o and near the r i g h t side of the detector tank. In t h i s region, contributions from other faces of the tank were negligible; thus the neutrons detected by the counter were t h e air- or ground-scattered f a s t neutrons which entered the side w a l l and were thermalized i n the water near the detector. The reactor and detector tank altitudes were varied simultaneously, in discrete steps, from 0 to 195 f t -

k composite p l o t of the measurements (Fig. 14.2) indicated only small difterences i n slope but appreciable differences in magnitude between the A p l o t of the curves for the various altitudes. f l u x v s the altitude (Fig. 14.3) showed o pronounced peak i n the region between the 15- and 20-ft altitudes, which indicated the importance of ground-scattered neutrons in t h i s region, I t should be noted that the reading at zero altitude was obtained w i t h the reactor half-irnmersed i n t h e ground pooi, the upper half being shielded as before; so essentially h a l f of the source was occluded, as WQS hQlf of t h e scattering medium. Thus t h e air-scattered neutrons should be no more than h a l f of the valwe at f u l l altitude and no less than one quarter. Because the intensities at the 150- and 195-ft altitudes were the same, i t w a s concluded that the ground-scattered contribution at these altitudes was small. In a preliminary analysis of the data, i t %as estimated that t h i s contribution was 2 t o 5% of the t o t a l scattered neutrons, While the detailed calculntians of ground and air scattering being undertaken i n connection with this investigation are not yet completed, a preliminary comparison of the data w i t h calculations carried O L J ~at the Boeing Airplane Company for Q not too dissimilar situatforr has been made. The qualitative agreement i s very good, although the peak i n the total flux, as measured, seenis to be at a lower altitude, A report on t h i s experiment i s being prepared.â&#x20AC;&#x2122; A REACTOR

DEK E ~ I_. G i i l i l a n d

B ~O W E R

L. 13. Holland

I_, H u l l

A grocedtire has been developed and tested for

CI calorimetric

determination of the power of the

175

....-....,......_____...

ANP QUARTE RLY PROGRESS 9 EPOR7 UNCLASSIFIED ORNL-LR-DWG 843

DETECTOR TANK

[.EFT SIDE

R E A C T O R TANK

TOP VIEW

SHOWING

d - distance from the center of the reactor tank t o the

0, 0, 0

point.

of both the 0, 0, 0 point and the center of the reactor.

- altitude

8 - horizontal angle between p and d axes, Q

- detector

orientation angle, horizontal angle between d axis and a perpendicular t o the braad side of the t r i p l e t B F , chamber,

DETECTOR TANK

DETECTOR ORIENTATION ANGLE 0

point location - on the outside of the detector tank 2 3 ft above the inside bottom of the tank.

0, 0, 0

x , y , z .-- coordinates of geometric center of detector. Since the center of detection of the various counters w i l l shift w i t h the location of the detector

i n the tank; the x , y , z coordinates indicate the geometric center of the detector.

p - thickness of water between outer reactor face and reactor tank wall, measured along a radius.

Fig. 14*1. Geometrical Canvention Adopted for Tower Shielding F a c i l i t y Experiments,

176

PERIOD ENDING SEPTEMBER 10, 1954

io3

ORNL-LR-DWG ~~

........ __ ....... _._

............

0 = 330 deg a = -90 deg

h = VARIABLE d = 64 f t

p = 45.2 cm

......._I

._._I_-

.--w

U 0 0

2 "

t m

" a

,Ti ...................

-J

I-

CURVE 1

2 3 4 5 6

7 8

ALTITUDE

50 ft 96 f t

.......

........

197 f t 0 ft

---I

V----r---t - - - t

...................

.--__

................

I-

._-

1....................

...................

~ _ _ _ _

J !

0.0 c m ............

....

s? 3 W

-................

...

2772

CENTER OF TRIPLET BF, CHAMBER AT: x = VARIABLE y = 79.6crn I

. . .

.........

-...........

.....

.....................

- 75

-65

- 55 -45 x , HORIZONTAL DISTANCE F R O M d AXIS TO DETECTOR CENTER (cm)

Fig. 14.2.

-35

Attenuation of Scattered Neutron Flux in Detector Tank a t Various Altitudes.

177

ANP QUARTERLY PROGRESS REPORT

30 beg

d-zdft p =

a5 2 cm

a = 30a e g

CFNTCR OF -RIPLEI x = SEE 4B7b

BF3 CHAbABCR AT In IRCN @ALL A-

5 ern'

y = 79 6 c m

1-@3cm

La--

f i g . 14.3.

w i t h Altitude.

-~ 96

I20

44@

168

192

h AlTlTUDE (ftl

Variation of Scattered taeka~ari Flux

GE-ANP R - 1 DIVIDED-SHIELD MOCKUP T E S T S

TSH. Thermocouple measurements of the b u l k water temperature i n t h e 12-ft-dia renctor tank are made w h i l e the water i s being stirred suff i c i e n t l y to ensure that no significant difference i n temperature e x i s t s throughout the tank. The r a t e of temperature r i s e of the water when the reactor i s started i s then proportional t o the power of the reactor.

Small corrections must be made

for the power input due t o the mixing and for the heat loss. T h i s procedure was f i r s t used for a 5 by 6 fuel element loading i n the TSR. The reactor tank wos lowered i n t o the drained handling pool, and four mixers were placed nt various positions w i t h i n the tank. The turbulence created by the mixers was sufficient for thorough mixing of the water, b u t the power added by the mixers was not enough t o g i v e a measurable increase of the water temperature. Before the reactor was started, the temper-

1 78

ature remained constan1 for more than 1 hr w i t h the mixers running. Some heat insulation was obtained by creating, w i t h a double layer of canvas, a dead-air space above the water and nrnund the reactor tank. The temperature r i s e during a reactor run w a s measured The b y 12 thermocouples placed i n the tank. thermocouples had previously been caJ i t r a t e d at temperatures of 0, 25, and 50°C and agreed to w i t h i n 0.1"C. The f i r s t run of the experiment was started a t 24"C, and the t o t a l temperature r i s e was approxiThe experiment was repented at a mately 14°C. higher temperature level but w i t h the same heat The i n i t i a l input and t o t a l terriperature rise. temperature far t h i s run was approximately 15°C above the ambient temperature, and the r i s e was ngain 14°C. The cooling rate o t the conclusion of t h i s run w a s 6.3°C/hr. In a t h i r d run the i n i t i a l temperature was brought 15°C below the ambient b y cooling the reactor tank w i t h i c e before the run. T h e results of the three runs were consistent t o w i t h i n 1%. A detailed report of this experiment i s being prepared.2

r. v.

~i~~~~~

I. G i l i i l a n d

J. Van Hoomissen F. N. Wfiiitsnn

T e s t s on the Gk-ANP R - l divided-shield mockup (Figs. 14.4 and 14.5) w i l l begin w i t h traverses around the reactor-shield section i n the handling pool for compat ison w i t h measurements made cat t h e PI SF.^ hSensur=metits w i l l then be made in the LSF detector tank, which w i l l be situated 84 ft from the reactor shield at an a l t i t u d e @f 195 ft. It i s nnticipated that the datector tank w i l l be replaced by the crew-compartment iflockup i n midSeptember, -

2C. E. Clifford et al.,

Cuiorirneinc Powel- I)etermznation o/ t h e Tower Shielding Reaclor, O R N L CF-54-8105 ( t o b e i s s u e d ) . '!-+.E. tiungerford, U u i k Shielding Fnrilzty T e s t s on the G E - A N P R - 1 D i v i d e d Shzeid Mochup, OKNI. CF-54-89 4 (to be i s s u e d ) .

P E R I O D ENDING SEPTEMBER 10, 1954

Fig.

14.4. Reactor-Shield Section of the GE-ANP R-1 Divided-ShieldMockup.

179

ca-

Fig. 14.5.

180

I Crew-Compartment Section

(Ends Removed) of the GE-ANP R-1 Divided-Shield Mockup.

part IV

APPENDIX

15. LlST OF REPORTS ISSUED DURi G THE REPORT NO.

T I T L E OF REPORT

QUARTER AUTHOR

DATE ISSUED

Aircraft Reactor Experiment

1. CF-54-7- 11

Fuel Activation Method for Power Determination of the ARE

E. 6. Johnson

7-3 1-54

CF-54-7-143

ARE Operating Procedures, Part I, Pre-Nuclear Operation

W. 6. Cottrell

7-20-54

CF-54-7- 144

ARE

J. L.

Meem

7-27-54

CF-54-8-171

Critical Mass of the ARE Reactor

C. B. Mills

8-28-54

A. H. Fax e t al.

6-25-54

R. M. Spencer

8-26-54

H. J. Stumpf R. E. MacPherson J. G. Gallagher

9-19-54

Operating Procedures,

Part I I ,

!I.

Nuclear Operotion

General Design

CF-54-6-201

A Reactor Design Parameter Study

CF-54-7-1

The Effect of the RMR Shield Weight of Varying the Neutron and Gamma Dose Components Taken by the Crew and Camparison of the RMR Shield Weight to that for an Idealized Shield

CF-54-7- 187

T e s t Results and DesignCornporisons for Liquid Metal-to-Air

Radiators C F -54 -8-2 00

High Conductivity F i n Test Results

111.

H. J. Sturnpf

J. E. Ahern

8-2 7-54

Experimental Engineering

CF-54-7-166

MorseSilent Chain Drive Test on ARE Pump Hat Shakedown Test Stand

A. G. Grindell

9- 16-54

CF-54-8-199

Aircraft Reactor Experiment Fual-Removal System Mockup

J. Y. Ertabrook

8-26-54

IV. C r i t i c a l Experiments CF-54 -7- 159

The F i r s t Assembly of the Small Two-Region Reflector Moderated Reactor

Dunlop Scott

9-26-54

CF-54-8-180

The Second Assembly of the Small TwwRegion Reflector Moderated Reactor

Dunlap S c o t t

8-26-54

R. S. Craiise G . M. Adamson

8-5-54

S. 1. Cohen et nl.

6-21-54

W. D. Powers

8-24-54

CF-54-8-2 7

Metallurgy

Preliminary Metallographic Examination of K-25 Heat Exchanger Loop

VI. Heat Transfer and Physical Properties CF-54-6- 188

Physical Properfy Charts for Some Reactor Fuels, Coolants, and Miscellaneous Materials (4th Edition)

CF-54-7- 145

Measurement of the Thermal Conductivity of Molten Fluoride Mixture No. 10-4

S. J. Claiborne

183

REPORT NO.

TITLE OF REPORT

AUT HQR

~ ~ - 5 4 - a - 1 0 Meosurement of the Density of Liquid Rubidium

VII.

DATE ISSUED

S. 1. Cohen 7'. N. Jones

8-4-54

M. T. Robinson

6-2-54

W. 0. Powers G. C. Blalock

8- 17-54

Radiation Damage

C F-54-6-4

Release of Xenon from Fluoride Fuels: Proposal for perimental Program

CF-54-8-135

Heat Capacities of Compositions No.

Ex-

39 and No. 101

VIII. Shielding CF-54-6- 164

Procedure for Obtaining Effective Removal Cross Sections from L i d Tank Data

E. P. Rlizord

6-22-54

c F-54-8- 9s

Preliminary Study of Fast Neutron Ground and Air Scattering a t the Lower Shielding Facility

Tower Shielding Group

8-23-54

CF-54-8-97

Calculation o f Fission Neutron Age i n NeZrF5

J. E. Faulkner

8-3 1-54

c F-54 - 8-9 8

Age Measurements i n L i F

E. Faulkner

8-3 1-54

I X . Miseellaneous CF-54-6-216

ANP Research Conference of May 19, 1954

A. W. Savoloinen

6-25-54

CF-54-6-217

ANP Research Conference of June 8, 1954

A. W. Sevolainen

6-25-54

CF-54-7-35

ANP Research Conference of June 29, 1954

A. W. Sovolainen

7- 2-54

CF-54-7-162

ANP Research Conference of July 20, 1954

A. W. Savolainen

7-26-54

184