Scribd Logo
Upload

Welcome to Scribd!

  • Naval Reactors Physics Handbook

    Uploaded by

    鄭毅夫
    15 views1,561 pages

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    15 views1,561 pages

    Naval Reactors Physics Handbook

    Uploaded by

    鄭毅夫

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 1561
    NRB-HGR-+ 94002 4 77 NAVAL REACTORS PHYSICS HANDBOOK VOLUME I Selected Basie Techniques HOTRBUTION OF THIS DOCUMENT 1S UNLAATED MASTER DISCLAIMER Portions of this document may be illegible in electronic image products. Images are produced from the best available original document. NAVAL REACTORS HANDBOOKS Liguid Metals Handbook, First edition: Edited by R, N. Lyon, June 1950, Second edition: Edited by R, N, Lyon, June 1952, ‘Third edition: (Sodium-NaK Supplement) Edited by C. B. Jackson. First printing June 1965; second printing (Revised), November 1955 (Available from Superintendent of Documents, Washington, D. C., 20402), Metallurgy of Zirconium, Edited by B. Lustman and F.Kerze, Jr., July 1955 (Published by McGraw-Hill Book Company, New York). ‘The Metal Beryllium. Edited by D. W. White and J. E, Burke, "Waly 1955 (Published ty American Society for Metals, Cleveland, Bibliography of Reactor Computer Codes, Report AECU-3078. Edited "py Hc 8. Brodsky, Dovenber 1055 (Avallaslo trom Superitenent of Documents, Washington, D, C., 20402—This compilation is being opt current by the Nuclear Codes Group Quazterly Newsletter— Inquiries should be sent c/o AEC Computing Facility, New York University). Reactor Shielding Design Manual. Edited by T. Rockwell Il, March T956 (Published Separately by Government Printing Office , MeGraw- Hill Book Company, New York, and D. Van Nostrand Company, Princeton, New Jersey). Gorrosion and Wear Handbook for Water-Cooled Reactors. Edited by ‘D, J. DePaul, March 1957 (Published separately by Government Printing Office and McGraw-Hill Book Company, New York). Naval Reactors Physics Handbook, A. Radkowsky, Chairman of Edi- Sorial Board Mensbooke Volume 1 Selected Basic Techniques. Editedby A, Radkow- sky (Available from Superintendent of Docu- ‘ments, Washington, D. C., 20402). Volume 11 ‘The Physics of Pressurized Water Reactors. Edited by $. Krasik, February 1959 (Classified Edition only, Volume I includes an updated discussion of those portions of Volume ITwhieh are of general interest), Volume I ‘The Physics of IntermediateSpectrum Reactors, Edited by J. R. Stehn, September 1958 (Available from Superintendent ‘of Documents, Washington, D.C., 20402), ‘The Shippingport Pressurized Water Reactor. Naval Reactors, West- ibgtotbe “Eletiris Cosonatoee aoe iquesne Light Company, September 1958 (Published by Addison-Wesley Publishing Company), ‘The Metallurgyof Hafnium. Edited by D. EB. Thomas and E, T. Hayes “(Available irom Superintendent of Documents, Washington,D.C. 20402). Uranium Dioxide: Properties and Nuolear Applications, Edited by US. Bolle, duly 1961 (Available trom Superitendent of Docucnts, Washington D, C., 20402). Neutron Absorber Materials for Reactor Control. Edited by W. K. Anderson and J. 5, Thellacker, January 1962 (Available from Super- intendent of Documents, Washington, D.C., 20402). NAVAL REACTORS PHYSICS : HANDBOOK VOLUME I Selected Basic Techniques Volume Editor A. Rapgowsxr 1964 ‘Naval Reactors, Division of Reactor Development United States Atomic Energy Commission EDITORIAL BOARD This is the first volume of the Naval Reactors Physics Handbook, sponsored by the Naval Reactors office of the Atomic Energy Commission. This volume has been edited by an Editorial Board consisting of the following members: A, Radkowsky, Chairman Naval Reactors, A.E.C. R. J, Breen Bettis Atomic Power Laboratory* R. Bhrlich Knolls Atomic Power Laboratory D. N. Schmoker Naval Reactors, A.E.C, D,S, Selengut Knolls Atomic Power Laboratory J. J, Taylor Bettis Atomic Power Laboratory* Qthers who have served on the Editorial Board for the Naval Reactors Physics Handbook: 4.R, Feldmeier (Formerly, Bettis Atomic Power Laboratory*) H, L, Garabedian (Formerly, Bettis Atomic Power Laboratory*) ALF. Henry: Bettis Atomic Power Laboratoryt H, Hurwitz, Jr, (Formerly, Knolls Atomic Power Laboratoryty 8. Krasie (Formerly, Bettis Atomic Power Laboratory*) @. Mechiin (Formerly, Bettis Atomic Power Laboratory*) H.D. Nine (Formerly, Naval Reactors, A.E.C,) J.R, Stehn (Formerly, Knolls Atomic Power Laboratoryt) P.B, Zweifel (Formerly, Knolls Atomle Power Laboratory!) [OREEMAT By the Wessagtonse Eizcxie Corporation for the Atomic Energy Commission "Operated ty the General lectric Company’ for the. Atemic’ Energy Commission, FOREWORD ‘the Naval Reactors Physics Handbook, seventh ina series of handbooks on basic reactor technology developed under the sponsorship of the Office of Naval Reactors of the Atomic Energy Commission, presents the analytical and experimental phystes techniques which have been developed for the reactors Jesigned in the Naval and Shippingport (PWR) Reactor Pro- grams, Reactor piysics has been strongly emphasized inthese programs, and extensive experimental and computational facilities ‘have been developed. Further, particular attention has been paid to obtaining from reactor operations as much physics information as possible. Prototype reactors have been Carefully instrumented and an extensive series ofphysicstests have been carried out with the twofold objective of confirming ‘and refining the physics calculations used in each particular reactor design and of obtaining increased general understanding of reactors, Power reactors must meet many stringent requirements, such as size limitation, stability, and reliability. Above all, changes cannot be made in a finished core to compensate for design error. The fundamental importance of reactor physics in introducing new concepts and in establishing reactor param- eters such as those relating to fuel loading, startup, control, power distribution, safety, and endurance is widely recognized. ‘What is perhaps less well understood is the need for constant interplay of mechanical, metallurgical, and plant considerations with those: of reactor’ physics to provide a well-balanced design, Close coordination of these elements has been stressed in the Naval Reactors and Shippingport Programs from their inception, ‘The physics methods described in these volumesare charac- terized by having been of direct utility in the design of actual power reactors. In addition, basic reactor theory developed in the Naval Reactors laboratories for the Naval Reactors and Shippingport Programs is covered, This includes many areas of reactor physics which have long presented particular diffi- culty, for example, the calculation of thermal neutron spectra, ‘control rod effectiveness, self-shielding and transport effects, temperature coefficients of reactivity, and flux stability. Much of this work is applicable to many other types of reactors, It has been gratifying to note the widespread use of ‘Naval Reactor computer codes throughout the reactor industry. $$ While I am pleased with the advances in physics which have been made in the course of the Naval Reactors and Shippingport Programs, I am not satisfied because I know that a great deal more must be done to improve our understanding of reactors and our ability to design them for longer life, more uniform power distribution, simpler control, greater safety, and more effective utilization of the fuel, T hope that by publishing our available data and techniques in these volumes we will provide impetus toward the solution of many of the basic reactor physics problems still outstanding, ‘My appreciation goes to the many who have given their time to the preparation, review, and editing of this book, HG, Rickover Manager, Naval Reactors Division of Reactor Development 0.8. Atomic Energy Commission EDITOR’S PREFACE ’ ‘The purpose of this work is to present the most pertinent parts of the body of physies knowledge which has been built up In the course of the Naval and Shippingport (PWR) Reactor Programs, with the aim of providing a background of under standing for those interested in nuclear core design, Much of the material has already been published either in sctentific journals or in topical laboratory reports; scientists engaged jn the Naval Reactors Program, as elsewhere, are encouraged to report their work in the normal manner. However, as is usual in rapidly expanding programs, much datahave remained in the form of internal memoranda and other informal com~ munications, Tt was felt to be desirable and timely to review and recapitulate the work which has been done on these pro- grams and to make it available in convenient and readily applicable form, Wherever practicable, the chapters in these volumes have been prepared by authors who actively partic pated in the developments discussed, ‘Volume I of this handbook was planned to bring together topics in the basic theoretical and experimental material ‘which are of especially wide interest, including those common to both thermal and intermediate neutron energy reactor types. ‘The physics design of light water-moderated and -cooled reactors is covered in Volume II (classified), and that of intermediate neutron-energy power reactors in Volume UI. ‘A brief explanation is necessary to explain the order in which these volumes have been issued. Originally, it was intended to publish the volumes in numerical order. However, the termination of Naval Reactors Programs on intermediate spectrum sodium-cooled reactors made it desirable to issue Volume Il first while the team of scientists engaged in the intermediate spectrum work was still intact. Volume I was then issued shortly thereafter to make available an equivalent status of information on the design of light water power reactors, Advantage has been taken of the time lag in issuing Volume I to include an updated discussion of those portions of Volume If which are of general interest. The emphasis in Volume I is thus on light water reactor systems, and as many recent advances in reactor physics of the Naval and Shippingport Reactor Programs as possible havebeen included, Tn connection with the responsibility for these volumes, it should be mentioned that all members of the Editorial Board have participated’ fully in this work by reading all chapters vit EIS 'SCSS and providing detailed comments, Reactor physics today is such a rapidly growing field that future editions of this hand. Pook will undoubtedly be necessary. Criticism and suggestions by the readers will be welcomed and, if practicable, incor- porated in later editions, The writer wishes to express his appreciation to Vice Admiral H. G, Rickover for suggesting the preparation of this handbook and for giving encouragement at all times. Special thanks are also due to Theodore Rockwell, Il, Technical Director, Naval Reactors, for many helpful suggestions. Finally, the writer is most grateful to all members of the Editorial Board and to all authors and chapter editors whose effort was essential to the preparation of these volumes, A, Radkowsky Chairman, Editorial Board Naval Reactors Physics Handbook Chapter 1 Chapter 2 ‘TABLE OF CONTENTS gy Page REACTOR PHYSICS AND ITS APPLICATION TO NUCLEAR POWER REACTORS LL Introduction... +62 eee a 4.2 Reactor Physics and Nuclear Design of Naval Reactors ...... 7 a 3. Numerical Analysis and Digital Computer Programs - 5 6 14 Selection and Verification of Caloulational Models .. +++. +--+ i ’ 8 1.5 Dependence of Design Detail on Performance Requirements .......-e+eeeeeeeees 10 46 Summary .. 02... eee eee 13 ‘THE NEUTRON SLOWING-DOWN PROBLEM 2.1 Introduction. . 6 2.2 Caloulation of Nuclear Cross Sections 2 A, Introduction... peor o B, General Considerations + ++ st . Optical Model. oo D, ‘The Compound Nucleus Model: .11 1.2051 E, The Direct Interaction Model = « « . Fr Unttied Treatment of Optical Model and ‘Resonance Reactions . . . . cry G. Models of Nuclear Bound States - 5 H. Cross-Section Caloulations . 1 2.3, Slowing-Down Theory in a Uniform Medium . 89 A. Introdiction| 88 BL Decomposition of the Transport ‘Equation . ae 80 ©. The Source Due to scattering 91 D, Heavy Moderator Approximations . « 4 E, Expansion Coefficients for Heavy Mass Approximation . .. . - 95, Calculation of Tf) 24 25 26 20 28 4H. Explicit Forms and Recursion Relations for the Quantities 7J,,....-.- 1. Slowing-Down Density ind the Generai- ized Greuling-Goertzel Approximation... 2... J. Slowing Down by Hydrogen | |) Monte Carlo Age Calculations in a Uniform ‘Medium : A. Introduction, (Be Beary ete ere ee ©. Neutron Distribution from a Pisne Fission Source in Water ..... . Resonance Capture asia Uiteodoetion tee ee B. The Resonance Cross Sections | | G. Analytical Models...... D. ‘The One-Energy Problem’ | E. The Practical Width... | Multigroup Diffusion Theory... A. Introduction... . eee B. The Group Diffision Equations | &. Time Dependence and Critieality D. Computation Strategy .... 2-0 1! E. Boundary Conditions: Continuity aé AOS eee eee ee F. Adjoint Equations - 17112111! G. Space and Energy Self-Shieiding | | ‘H. lustrative Example: Multigroup ‘Multigroup Transport Theory A. Introduction... 0... 0011! B. Description and Disoussion of the Method eee eee . Angular Differencing. | D. Lethargy Differenciag . E. Conclusions... ‘Few-Group Approximat A. Introduction. . . BL Single Mode Model | | Ce Crop Byuations een eee D. Special Forms of the Group Equations | E. The Time-Dependent Problem... . F. Computation of Few-Group Constants. _ Fast-Group Fitted Constants in Few-Group jeg treet eerste eee ete A. Introduction? eee B. Fitted Cross-Section Scheme | | ©. Computational Results. D. Variational Procedure for Fast Group Constants, lating | 125, 325 126 132 1s7 138 wa 447 153 154 154 155 161 162 163 18 165 168 115 115 115 ar 182 197 197 197 199 203, 209 2s 218 229 229 230 251 256 3.1 Introduction. 3,2, Inelastic Scattering of Thermal Bae Neutrons. 3.3 34 3.5 36 3.1 3.8 Chapter $. THE NEUTRON THERMALIZATION PROBLEM A B, c D. Introduction. ‘Theory of Scattering - Applications Conclusions . Sensitivity to the Neutron Scattering Kernel .. - A B, Multigroup Theory... +++ A. B c D. ‘Analysis of Thermal Neutron Spectra. ‘Analysis of Thermal Diffusion Length . Introduction... Multienergy Transport Equation Solution ofthe Multigroup Equations ta One Dimension . « ‘Multigroup Monte Carlo Techniques. - Application of Monte Carlo Methods to Neutron Thermalization....... A B, ce D. zB Blackness Theory for Slabs. . A B, c D. EL Fr ‘Introduction. ‘One-Group Thermal Model! "TUT-TS. - ‘Threo-Dimensional, Multigroup Program: TRAM .......+ One-Group ‘Thermal Model with Slowing-Down: TRAC-1.... Normal sod Adjoint Maltithermat Groip ‘Model: MARC-1...... Introduction. . «1... ++ +++ Blackness Coefficients’. . - Matching Conditions Double Blackness : Applications and Resulis << +: Convenient Forms and Conclusions « Approximations to Multigroup Methods ‘A Two-Mode Variational Procedure for Caloulating Thermal-Distusion Theory Parameters . 5 A B, Introduction... ‘One-Group Models : : Conventional Few-Group ‘Approximations... . Variational Approachto Ther: Caloulations Introduction: 2-2. General Formulation. 285, 291 291 293 312 330 332 332 339 351 351 353 365 379 381 381 382 388 496 CC EEEEDIEDEOES=S9~—~~ ©. Procedure for Calculating Therm: Constants... 506 D. Results. . DEI gn Chapter 4 REACTOR DESIGN TECHNIQUES 4.1, Introduction. st 535 Peete re a ssn Shielding ta Plates Leaded With Spherical Poison Particles... 537 C. Generalized Treatment of Particle Sstieatieldiog semitererepva aa D. Representation of Piste Seli-Shieiding | im Diffusion Theory... 2... 505 4.3 Homogentzation Techniques . 20 a icooaetica seer eee 620 B. Methods of Obtaining Homogenized Constants from Cell Theory ear ©. Arrays of Epithermal Self- Shielded | Absorbers . Beeeeorre 636 D. The Validity of Coit Theory for Arrays of Absorbing Plates... ss 645 E. The Validity of Cell Theory for Fuil Gore Cateulations .. 650 F. Cell Calculations with Nonzero-Curzeai ' Boundary Conditions - : 653 ' ©. Summary, 22! os 44 Synthesis Mothods . . + 656 Av Introduction. 22122177 a ace B. Single Channel Synthesis | ft 687 : C. Multichannel Synthesis. a ore : D. Variational Synthesis || m0 45 Gross Fission Product Polsoning ....... 728 : - Introduction... any) Be Analysis 02002 as G. Representative Numerical Results. | 748 D. Discussion of Results... ss...) 448 4.6 Depletion Methods .. 5 751 Ap Introduction. 022022022 122I010 gar B- Rifeotive Microscopic Cross Sections. | 7153 G. Number Density Calculations... 756 D. Depletion of Natural Uranium m7 B, Integrated Depletion Systems 132 Ps 4.7 Burnable Poisons . ‘A. Introduction»... ee) 600) B, Elementary Theory of Homogeneous Burnable Poisons in Thermal Reactors .....- 802 C. Materials Suilabie for Use a5 Burnable Poisons... - 315 D. Self-Shieldad Burnable Poisons . san E, Nonunfform Depletion Effects and Zoning... ese eee ee ee ee ee 831 . Application of Burnable Poisons to ‘Cores with High U238 Content Fuels. 840 Advanced Applications of Burnable Poison 5 843 H, Use of Burnable Poisons in Epitherma Reaotors vee... esse es ee Chapter 5 REACTOR KINETICS 8.1 Introduction......- pee ess 5.2 Neutron Kinetios « eee) Chapter 6 ‘A. Derivation of the Kinetics Equations 855, B, Physical Interpretation of the Adjoint Flux: Iterated Fission Probability 864 C. The Utility of the Kinetics Equations 869 D, Computation of Parameters in the ‘Kinetics Equations. . 7 873 5.8 Reactor System Kinetics . . emt A. Inherent Stability il 878 ‘B, Boiling Induced Reactivity Feedback’ 915 ©. Self-Shutdown of Reactor Power Bxoursions . eee 28) $4 Space-time Kineties. 955 ‘A. Introduction. cece 955, B, The Instantancous ‘iit Method | | * | 987 ©. Nodal Analysis... 2.0.6.0 000 961 D. Modal Analysis «22. : 966 EL Modal Analyses of Xenon Stability : ‘ 977 F. Summarys... 60... 1007 5.5 Kinetics of Low Source Level. . . . 1010 A. Introduction...... ++. 3010 B. Analysis of Low Source Startup 013 ©. Multiplicative Process with Feotback 2 1079 D, Natural Reactor Sourees . .. - 1085, COMPARISONS OF INTEGRAL EXPERIMENTS WITH THEORY 6.1 Introduction... 6.655 143 6.2 6.8 64 65 6.6 87 ‘Highly Enriched Clean Critical Experiments At Ambient Temperatures ......-- A. Clean Slab Experiments with Small Bundle Box Geometry... . . B. Clean Slab Experiments with Rep. Plane Geometry .... 2... ©. Plastic~Moderated Critical iments......... ea Highly Enriched Clean Critical Experiments st Elevated Temperatures............ A thom IT) B. Description of Experiments | C. Caleulational Model .....2./ 1)! D. Results of Core Reactivity Calculations - E. Results of Activation Distribution (Compaxiscns enene sere F. Conclusions. . Jnvestigations of Neutron Peaking in ting Highly Enriched Clean Critical Experiments . A. Introduction is 3B. Experimental Arrangement’ | C. Characteristics of Lattices | D. Experimental Results... . EL Analysis . F. Conclusions | High 0° Consent Critical Ambient Temperature .. - A. Istroduction, 0010205202! + Effective 0238 Resonance Gapiure In- tegrals in Rods and Lattices... . . ©. The Relative U25° Fission Activationas ‘4 Funetion of Energy in Slightly Enriched Uranium-Water Lattices . . D. Relative Pu and U25 Fission Rates in Water-Uranium Reaotor Spectra. . . Pulsed Neutron Source Measurements ... . A. Introduction... eee Be Theory. 202217222) C. Experimental Method 12122121! D. The Far Suberitical Pulsed Neutron Technique... 2.2.2... eee E. Reactivity without Critical Calibration | Low Energy Measurements of the: Neutron Spectrum... : AL Introduction! |! B. Experimental Method |||) | ©. Details of the Chopper Design D. Experimental Correotions ... |. E. Comparison of Calculated and Measured Spectra, . ef F. Conelusions | as nas 1193 3203 ais 2218 1220 1229 1234 1238 243 1245 1245 1248 1252 1254 1288 3271 1275 25 1276 1298 1808 1319 1319 saa 1332 1336 1851 1387 1357 1360 1369 1376 1390 Chapter 7 REACTOR PHYSICS COMPUTATION TA Introduction. ........ A, Digital Computers. << 0... B. Programming Techniqies . | |: ©. Roleof Analysis..... 0.000. 1.2 eration Procedures in Neutron-Diffusion Calculations... are A. Introduction: * | : B, Source Tteraiions «=o... 2. Solution of the Group Equations. + D. Tho Time-Dependent Problem . - 1.8 The-Mathematies of Monte Carlo for Re- actor Calculation ... . . coca A. Introduction. «2222222000007 B. Random Walle Processes |... ©. Analog Processes... ._ Importance Sampling aud isaied ‘Methods. 14 Digital Computer Programs for Reactor Physics Galeulations.-. A.” Introduction. 5 B, Few-Group One-Dimensional Diffusion Theory eee ©. Few-Group Two-Dimensional Diffusion ‘Theory = D. Few-Group Three-Dimensional Diffusion Theory «+++... + E, Few-Group One-, Two-, and Three~ Dimensional Depletion . F, Few-Group Constants Caloulations G. Approximations to Transport Theory: . H, Monte Carlo Calculations... .- I. Theoretical Cross-Section Galculations 3. Perturbation Caloulations ......... K, Modal Caleulations . ... . CONTRIBUTING AUTHORS . Sopaupg INDEX. 1405 1405 a0 1412 wai lai 4418 1426 1439 aan aaa aaa ass 1485 461 1461 146i 1463 473 aa7s 1ae4 1488 1499 1507 1509 1509 1519 EI ’'S=S3r=s 2 NAVAL. REACTORS PHYSICS HANDBOOK: SELECTED BASIC TECHNIQUES depletion effects, other than those due to transient fission Product poisons, could be treated as perturbations. Asa result, ireatments based on age diffusion theory were adequate for the description of such cores, All calculations were performed on desk calculators, At about this time several reactor design efforts of particular Significance for the Naval Reactor program were begun, One of these, the power breeder program at the Knolls Atomic oo wer Laboratory (KAPL), consisted of reactor studies partly Primarily in the neutron intermediate energy spectrum. Tho SAPL program was subsequently reoriented towards the design of @ naval reactor operating in the intermediate spectrum, Ac fier design etfort begun about this time was that of a sumil Hight-water power reactor, the prototype of the first Neve Teactor (STR). The work originated at the Oak Ridge National Laboratory and was transferred to the Argonne Nationt jaboratory. Subsequent STR design work was eartiod owt ot ine «Bettis Atomic Power Laboratory. The development of it ‘aval Reactors and Shippingport programs has continued at Bettis, KAPL, and the Navel eect tors Division of Combustion Engineering Corporation, Zn each of these development programs of high performance Sores the reactor properties have determined the direction and emphasis of the physics efforts, : Feactors. While the Naval Reactors program at the present : Hime utilizes light-water cooling and moderation, the furliosd: £28 lensities are often sufficiently high that a substantial | Fraction of the fissions occurs above thermal neutron enerisice, i Fe ei—imuch of the physics work required for deseribing i the intermediate energy spectrum is useful for lightrvciee i Teactors, Advances in the following areas have been of : particular benefit: Hatistical theory of resonances of fission products (Chap.4). ueory and direct measurement of Doppler broadening of resonances (Chap, 2 and Vol, Il), Interpretation of adjoints and caleulation of reactivity. co- efficients (Chap. 5), Self-shielding of resonances (Chaps, 2 and 4), Critical assembly experimentation (Chap, 6). oo ,— SrttttCO—tstCiCO APPLICATION TO NUCLEAR POVER REACTORS 3 1.2 REACTOR PHYSICS AND NUCLEAR DESIGN OF NAVAL REACTORS In early cores of low fuel and poison loading, the high moderating power of light water produced a neutron spectrum quite closely thermal, anda two-group (or atmost a four-group) freatment was considered appropriate. However, amajor prob- lem was presented as to the proper group parameters to be ‘used; great effort has gone into the solution of this problem. Chapter 2 of this volume treats the calculation of parameters for fast neutrons, and Chap. 3 for thermal neutrons. "im the case of the fast group itwas krown that the continuous slowing-down model in use for heavy moderators would be inaccurate for hydrogen moderators. More suitable models for hydrogen slowing down wereutilized. This work was supple- mented by experimental determinations of the slowing-down distribution in light-water metal mixtures, A second difficulty was that, as a result of the high leakage in the cores being designed, the fast neutron spectrum was quite different from that in an infinite medium. In this case, since the fast neutron spectrum does not vary rapidly from point topoint in the core, it is a good approximation for the purpose of computing param- eters to assume a sinusoidal space dependence of the flux with a geometric buckling equal to that ofthe average material ‘buckling in the region of interest, The multigroup treatment then becomes mathematically the same as the normal mode spectrum calculation for an infinite medium. The calculation yields the neutron spectrum appropriate to the region and, by averaging the energy-dependent cross sections over the spectrum, the fast group parameters can be obtained. ‘The importance of inelastic scattering by heavy elements in slowing down was recognized early, but accurate treatments were held back by lack of experimental data, In recent years a great many measurements have been made and, furthermore, an extensive program for calculating such cross sections has been developed. The calculation program is covered inChap. 2 for elements of particular interest to the Naval Reactors program, in the thermal energy region a formidable problem was presented by the rapid change in hydrogen binding and cross section with neutron energy. The Radkowsky prescription furnished a phenomenological method of obtaining the thermal transport mean free path from the measured variation of the hydrogen cross section with neutron energy. Since these core types had much greater thermal macroscopic absorption than the natural uranium graphite cores, it was necessary to take into account thé deviation of the thermal spectrum from a SESS ~=~=CSS OTTO, 4 NAVAL REACTORS PHYSICS HANDBOOK: LECTED BASIC TECHNIQUES Maxwell distribution due to absorptions occurring before the neutrons could come into thermal equilibrium with the modera- tor. This was accomplished by application of the Wigner-Wilkins formalism and improved models, Because of the small slowing-down length in ight-water cores and the large amounts of reactivity associated with transient poisons (xenon and samarium) and temperature change, control requirements in these cores were relatively stringent and necessitated a close spacing of control rods. An associated problem was that the small thermal diffusion lengths resulted in large flux distortions due to heterogeneities, suchas control rod water channels. Thus, toavoid excessive water-hole peaking it was necessary to utilize control rods consisting of thin plates in a relatively complex geometry, for example, cruci- form. The overall result was a core which was quite complex from a physics standpoint, An accurate description of the flux peaks and the control rodworth requires an explicit representa- tion of the control rods, structural members, water regions, fuel regions, and other features ofthe core affecting the neutron distribution. To provide sucha description it has been necessary to carry on investigations in such areas as self-shielding and cell theory (Chap. 4) and blackness theory (Chap. 3), and to make full use of the best mathematical tools as they have become available, including successively desk calculators, small digital computers, manually adjusted electric network simulators, and large scale digital computers ultimately lead- ing to large two- and three-dimensional diffusion theory com- puter programs (Chap. 7). Jn addition to the development of methods for the static description of the cores, a great amount of work has been re- ' quired to obtain suitable treatments of kinetics, safety, and stability problems (Chap. 5). The characteristics oflight-water cores which are significant from a kinetics standpoint differ in many essential aspects from those of cores previously studied: for example, the light-water cores have much larger moderator and void negative temperature coefficients; much smaller metal reactivity coefficients; much smaller thermal neutron lifetimes; different potential accidents, suchas the cold ' water accident; and different transient behavior occasioned by the need for rapid and flexible maneuvering. In the Shippingport program, as an approach to economic power generation, the emphasis has been on use of fuels of high 0238 content in order to minimize diffusion plant costs, This has led to the study of ight-water uranium lattices which, owing to the relatively close spacing of the uel elements, have unique characteristics: large interactions, particularly in U 298 fission (fast effect) andin resonance absorption (Dancotf effect), APPLICATION TO NUCLEAR POFER REACTORS 5 and a large contribution to slowing down by the uranium in- elastic scattering, To understand and describe such lattices,. a joint program was set up between the Bettis Atomic Power Taboratory and the Brookhaven National Laboratory in which» sets of slightly enriched uranium fuel rods were studied, al- ternately, at Brookhaven in a subcritical or exponential as- sembly and at Bettis in a critical assembly. Some of the associated techniques are described in Chaps. 2 and 6 of this Jolume, and in Vol. IL? Anewphase in the program began with the investigation of seed-blanket cores, which differed con- ceptually from the types of cores considered previously. The pasic idea of the coupling of two quite dissimilar core regions, or of driving one core regionby another, led to a deeper insight fnto the meaning and measurement of reactivity. Tt was neces~ sary to treat such novel constructs, as the extremely high neutron flux gradients existing between the seed and blanket regions, the seed-blanket power sharing, and the concept of a large core having essentially the kinetics properties of a Small core (the seed). Furthermore, the depletion character- {atics of these cores necessitated study of the properties of plutonium isotopes in ight water and of fission products in very high burnup fuels inwhichmostof the energy is derived from the fissioning of plutonium. ‘The advent of burnable poisons had a major effect on light- water reactor designs, In the first place, the much greater core endurances made possible by the use of burnable poisons resulted in depletions so large that the core ‘parameters changed greatly throughout life. This made it necessary to examine the adequacy of the core from reactivity and thermal standpoints at many times during core life, so that in effect the reactor designer was faced with the problem of designing many cores, As a result of the initial neutron flux spatial variations, the cores, even if they were homogeneous to begin with, became heterogeneous both radially and axially after ap- preciable depletion. ‘This made it imperative to develop codes to provide two- and even three-dimensional descriptions of the core, not only initially, but throughout life as well (Chaps. 4 and 7), The previously used rough approximation of stable fission product absorption as a constant number of thermal barns per fission was quite inadequate for accurate calculations of endurances with such high fuel depletions. Tt was found necessary to go to a detailed study of the data on individual fission products, taking into account: ‘successive transmutations and epithermal absorptions. The great increase in fuel loading concomitant with the increase in endurances resulted in blacker cores, L.e., cores with much larget macroscopic thermal absorption cross 6 NAVAL REACTORS PHYSICS HANDBOOK: SELECTED BASIC TECHNIQUES sections. This, in turn, necessitated a much more careful examination of neutron thermalization and the spatial variation of the neutron spectrum, The powerful variational technique was found to have many applications in the field. The extended use of burnable poisons led to their utilizationoutside the fuel, in lumped self-shielded form, both macroscopically (e.g., plates and rods) and microscopically (particles). This again Introduced the need for extensive calculations and experiments to determine accurately the self-shielding of the poisons and, in some cases, of the fuel, both initially and as a function of fuel depletion. In many cases, conventional analytic methods have been inadequate and it has been necessary to utilize Monte Carlo techniques (Chaps, 3, 4, and 7). The desire to obtain optimum reactor performance has also resulted in more complex designs. Local and gross zoning of fuel and burnable poisons has been studied to maintain power distributions of specified shapes and to minimize the number of mechanically moving control elements, In this connection, it is of interest to note that, as a core fuel loading increases, the emphasis of the problems changes, For example, in lightly loaded cores the reactivity associated with transient xenonand samarium is very largo; secondly, it is often difficult to obtain a large, negative temperature coefficient of reactivity, On the other hand, in cores having a high fuel loading the xenon and samarium reactivity effect is small and the temperature coef- ficient large and negative, but major problems are presented in power peaking near geometric discontinuities and in shutting the core down with adequate’margin. It is also noteworthy that the designer must aim at a balanced core design since changes in the core to improve one characteristic may adversely affect another. Thus, improvement of the power distribution often leads to a reduction in the effectiveness of a given set of control elements. 1.3 NUMERICAL ANALYSIS AND DIGITAL COMPUTER PROGRAMS To provide detailed descriptions of these complex and ad- vanced designs taroughout core life, substantial effort has been placed on numerical analysis. Chapter 7 discusses some of the progress made in that effort, particularly in the treat- ment of the few-group two-dimensional nuclear diffusion equations, the transport equations, and Monte Carlo methods, Faster convergence rates in the numerical solutions have been obtained over the years by utilizing more sophisticated over- relaxation techniques. In addition, a more rigorous foundation for the basic numerical processes has been established, : i APPLICATION TO NUCLEAR POWER REACTORS : ‘This work in numerical analysis stimulated fundamental mathematical investigations in the nature of the spectrum of the multigroup diffusion operator and finite difference ap- proximations to it. The results provided a mathematical basis for the concepts of criticality, multiplication factor, principal distribution, and importance function innuclear reactor theory. ‘The existence of expansions, similar to those in quantum mechanics, of the solution of the diffusion problems in terms of the characteristic functions associated with the diffusion operator was established. This, again, provided a firm basis for handling many problems in reactor theory; e.g., involving perturbation theory and pulsed source techniques. ‘As a result of the extensive theoretical analysis advantages have been gained in computer utilization which canbe exploited either for reduced design time or greater design capability. Some measure of the gain can be indicated by reviewing the effects of numerical analysis studies on successive programs of the PDQ (two-dimensional diffusion theory treatment) series of digital programs. QED-1 was the first two-dimensional few- group program for the IBM-704, Its successor, PDQ-1, gained. in speed as well as convenience to the user by the insertion of automatically calculated overrelaxation factors in the iterative method. The time savings have varied with the experience and ability of the designer. However, since the convergence rate is very sensitive to the choice of overrelaxation factor, the im~ provement is substantial. PDQ-2 utilized Chebyshev poly- homials to carry out the numerical method of solution and was estimated to decrease the overall time by a factor of 1.6. PDQ-3 utilized single line rather than point overrelaxation and increased the speedofa given problem by an average factor of 1.4 over PDQ-2 (some design problems actually showed improvements of a factor of 5). Finally, PDQ-4, produced for the Philco-2000, further decreased the average time bya factor of 1.4 by using double line overrelaxation (not including the increase in speed provided by the machine). Tn addition to the improvements in iterative processes, ~ improvements in the iteration strategy, ic., the choice of the number of iterations performed ina single groupbefore moving to the next group, provided an increase in the overall con- vergence rate. Such advances have, for example, been in- corporated in KARE. with the Peaceman-Rachford iteration procedure, Insertion of symmetry conditions of various kinds for half core, quarter core, and eighth core periodicity also introduced substantial factors of gain in speed for a specific problem, However, probably the most important result of the Work in numerical analysis has been the development of an understanding of, and a sound theoretical foundation for, the 7 TE EN'T':C 8 NAVAL REACTORS PHYSICS HANDBOOK: SELECTED BASIC TECHNIQUES mathematical processes involved, providing insight into the approaches which should be taken for specific problems and the accuracy which can be expected (Chap. 7). Reference 4 is. a review paper of this work and contains a rather complete bibliography for further reference, ‘The need for an accurate and detailed description of reactor behavior has had a marked effect on the development and veri fication of calculational tools used by the nuclear engineer in the design effort, The nuclear designer is faced with making @ compromise between the accuracy and detail with which he desires to describe a reactor and the cost and limitations of computer machinery available to him. Since practical con. siderations dictate that extensive approximations be made in the techniques used to obtain design information, dependence on more basic calculations and comparisons with experiment for evaluating the accuracy and deficiencies of the design tools have been essential. For these reasons, the formulation and tse of more exact calculations have been very important. The studies made in this area, inadditionto yielding standards fo which more approximate calculations may be compared, have supplied insight into the reactor physics phenomena of interest and guided the formulation of more approximate descriptions that retain the important features of the theory. ‘The use of the more basic theory itself has also identified many areas of sensitivity and guided the reactor physicist in further While it is possible to develop definitive techniques in reactor theory, they are generally available only in a re- stricted sense. They may incorporate a detailed description of the energy behavior or the angular dependence but be limited in the spatial description available. For instance, the use of multigroup Monte Carlo has been extensive, but the tractable application of this technique has been limited to small cells in either the resonance region or the thermal neutron range, 14 SELECTION AND VERIFICATION OF CALCULATIONAL MODELS While these basic techniques are accurate in their sphere of application, they are not general enough to obtain all the information about the reactor behavior that is desired, The essentials of these different techniques must be brought together in a self-contained nuclear design model that will describe the entire configuration. This usually requires drasticapproxi- mations in all of the variables, Practical examples of this are the use of few-group diffusion theory, performing spatial caleu- lations at diserete points rather than continuously, representing ee APPLICATION TO NUCLEAR POWER REACTORS 9 a group of fuel plates and water channels as a homogeneous region, and restricting the calculations to a few selected times during the core lifetime. Identifying this framework does not specify the calculational model but only outlines the ground rules within which the model willbe formulated, The few-group cross sections, transport corrections, geometrical detail re- quired, and spatial mesh necessary for the adequate description of flux shapes and depletion effects must be specified in such a manner that the essential features of the configuration are Gescribed accurately. This will often be supplemented by including self-shielding factors (for particles, plates, reso- nanees, ete.) in the few-group parameters, using blackness theory to describe the effect of strong absorbers, or by adjust- ing the few-group parameters to get agreement with a more exact calculation. The basic theory is then used to evaluate the aecuracy of the approximations and to identify the required improvements, ‘While many of the features of the model may be tested against standards, these comparisons must usually be made in a re~ stricted sense, while other features are less susceptible to detailed comparison. Coupled with this is the lack of complete and accurate nuclear data with which to perform the calcula~ tions, The comparison and testing of the derived calculational model (using the best nuclear data available) with experimental measurements on critical reactor configurations are, thus, an essential step in further evaluating the limitations of the model. Especially important have been the clean critical assemblies ‘which provide a simplified geometry especially designed totest particular aspects of the theory (Chap. 6). Originally, the ex- periments were carried out at room temperature. Since the parameters of light-water cores change appreciably from room to operating temperatures, pressurized critical facilities have been provided to permit studying the nuclear assemblies over a wide temperature range. In addition, the use of pulsed source techniques has permitted the study of the reactivities of sub~ critical assemblies. This has allowed the calculations to be checked at many points rather than merely at critical and has yielded reactor physics information from smaller amounts of material and smaller experimental effort. ‘The adequacy of the model for use indesign problems is still dependent on the degree of detail and accuracy required by the design objectives, If these are extensive, the final stage of design must be paralleled with the testing of the model against a detailed mockup of the core. The amount of information re~ quired for design of the core does notallow the obtaining of all this information experimentally. Nor will it yield information on lifetime performance, but 2 careful analysis of the mockup OE 10 NAVAL REACTORS PHYSICS HANDBOOK: SELECTED BASIC TECHNIQUES will supply a valuable normalization point and final verifica- tion of the ability of the design model to describe the core configuration, Finally, prototype cores have been heavily instrumented and extensive test programs conducted to verify the design calculations, 1.5 DEPENDENCE OF DESIGN DETAIL ON PERFORMANCE REQUIREMENTS It may be fitting to close this chapter with a concrete example of the necessity for extending reactor physics calculations to greater detail as a result of increased core performance re- quirements. In early reactor designs only knowledge of the overall neutron flux and resultant power distributions was essential. As higher power densities were demanded, it was important to determine what has been called by analogy with quantum mechanics the fine structure, i.e., the power output of each individual fuel element. Finally, it has developed in the design of the second Shippingport core, as wellas in other high performance cores, that it is necessary to go even beyond this point to what might be called the hyperfine structure, ie., to the determination of the power distribution within a fuel ele- ment. In the second Shippingport core the design maximum power density in the seed is almost 500 watts/cc, more than twice that of the first Shippingport core. As shown in Fig, 1.1 each seed subassembly contains highly enriched oxide fuel wafers, 0.250 in. wide, embedded in zirconium plates. In order to reduce the power peaking, a zoning ofeach seed subassembly in 2 zones of fuel was adopted, basedupon the calculated peak- to-average value of the hottest wafer inany zone to the average in the subassembly. This gave a calculated value for the peak- to-average of about 1.4 which was in reasonable agreement with experiment. However, examination of the power distribution indicated that there was a sharp gradient in the vicinity of the hottest wafer. A special experiment was devised to study this gradient, using four quarter-width wafers inplace of the hottest wafer. The induced gamma ray fission product activity of each quarter wafer was measured and an 18 percent rise in power found across the quarter-inch width of the wafer (Fig. 1.2). As a result of this unanticipated peaking, it was estimated that the thermal output of the core would be reduced by 10 percent, To rectify the situation a new zoning scheme was adopted (Fig. 1.3) with 3 fuel zones, Figures 1.4 and 1.5 compare the power peaks as a function of seed fuel depletion for the original and the new zoning schemes. This is an example in which refinement of reactor physics’ studies has led to a significant ‘economic benefit in plant performance. APPLICATION TO NUCLEAR POWER REACTORS u SS orensoe FIGURE 1.1. Shippingport Core 2 Seed 1 Original Seed Zoning. FIGURE 1.2. Shippingport Core 2 Seed 1 Distribution of Power within Seed Fuel Wafer Located Adjacent toa Control Rod Water Channel. nnn 12 NAVAL REACTORS PHYSICS HANDBOOK: SELECTED BASIC TECHNIQUES ED coe me nce ont tae ane Fen xen FIGURE 1.3. Shippingport Core 2 Seed 1, New Zoning of Seed to Optimize Power Distributions, ‘Taking into Account Intrawafer Flux Gradients. vere SetD RSENS FIGURE 1.4. Shippingport Core 2 Seed 1, Peaking Factor Lifetime Variation as Calculated for Original Zoning (A, B, C, E, F, and G ‘represent various points within the fuel assembly shown in Fig. 1.1). _ | APPLICATION TO NUCLEAR POWER REACTORS B CORE AGE AS INDICATED BY AVERAGE FUEL FRACTION REUAINING 1SEED| FIGURE 1.5. Shippingport Core 2 Seed 1, Peaking Factor Lifetime ‘Variation as Calculated for New Zoning (A, B, E, F, and G represent various points within the fuel assembly shown in Fig. 1.3). 1.6 SUMMARY ‘The reactor physics work inthe Naval Reactors and Shipping- port programs has been guided by the objective of a detailed description of complex reactors designed for optimum per- formance, The need for accurate calculational and experimental techniques to meet design requirements has required the de- velopment of systematic procedures and appropriate tests of their adequacy. The following chapters describe a number of useful techniques and procedures that have evolved in the course of this program, Finally, it should be pointed out that, asis inevitable in works of this kind, some very recent developments could not be included in this volume. As the work of the Naval Reactors and Shippingport programs continues to advance at a rapid pace into areas of great practical importance as well as of inherent physics interest, the editor must establish a cutoff point in time if a book is to be published. REFERENCES 1. 8, Krastk, Ed., “Naval Reactors Physics Handbook, Vol. Ul; The Physics of Pressurized Water Reactors,” 1959, 0 SNS ’d'S

    You might also like