Home

1 - ipen.br

Contents

1. 2 x HR OO HN y NNNMNN Woh QONI ror 3 036 9 eooosm op 3 uem data s HOO nmm duo ANNANN UY amar mcm WADI OP 7m A Viesessces goo omooom 2220 OMA ORK ARD P eR EIS D 2 n uiu D et OV 40760 Oecccrrer a re Ot ORK iyi 9 a 4012 SUM RING PASS IMAAL ANCE MASS FLOW AFTER EXACT SOLUTICN OF CONTINUITY EQUATION USING WEDGES RING GEOMETRIES 106 On we om uem ONON ARAN TA ANA OF o gt 234 4 RAD ADA n WAND 2 ea gn MSY AKI PEDET AON cs AOI Ear do RN RUD restent memes Duo MANT o a Kammer 0 4 4 MNNM DS OO uo WOMAN 924 390 INI OAseusevnse 29000050 o o 4 6 MAMAA Qeoccooson AANI LN a He AA WOM FADO SCRATIAORG AR OR Au
2. Z m H 4 X i less ecoo Y o HOMWOMMNO 2 H D _ D Ou DH 6 SLMS lt J ON HemN one um Po 8 0434 V O nensis a a xou Da HAPON uj o UO 5 AFIVOWMTO CO TONNEN code 2 tox AAO 4 O MOOTRRAWO 4 feenocoone Visesescee eecotess gt e ENNIO E E MND Woo rd OK 285502 S 5 REST 2 3 5 NNI 3 ou z ar gt a o J nmenoneno 6 mmap O00 HIMNO 5 0 V emque ae ae a x RADIAL VEL VALUES IN SHROUD TMERWAL VALUES 2 ITERATION WURGEF FASS FLOWS AFTER SOLCTICM OF MOMENTUM EQUATION FOR U VELOCITY ARNIS RONA PAS MMOH 00 JMNM
3. e t o t e ho t Ud JU Ur Ud UU UU Uu P f CUR NMR DORI ANAT 100 00 T7 N 9 0 a IN POOQOAANTIN 3 704 T QUU OO DUDS VA FV eiii mmm 4 332 TIAN errr rer rere eT Tee Tee Peer er tet U Uta ba ig Hag L U U A L Las U a Sh FIERO OON OR P DNDI MOIR vd MOM AUTO PLD IN PMN DOTA 3 PIE Uv DM e AP GIU OS UU ER DEVAN ORK DOWD 444 44 4 9489 9499424 6 NA ARAN 2 227252222212 Leer ee eee ee gt WW WS GJ CL td MAP WOM AIDE Qr OU raj gy 1090 7 COSI YEON DWU CIO NIU PR AMR AM HOO IIS HNN OUI TP O FLOM PONN DONEN SNR DOT Duros TW FM OAD mmm 3 JH q gH d TH4 3 433 test
4. t E tm wens te trud ea ya 1 25 A Eg SOR NS S un c A ME UL P ken k 2 Pa T M AT 1 14 Ke 7 6 Figure 6 6 3 THIRST OUTPUT Composite Plots Mass Flux Distribution 113 STEAM QUALITY CONTOURS AT K 3 AT K 4 AT K amp AT K 6 Figure 6 7 1 THIRST OUTPUT Radial Plane Plots Quality Distribution 114 VELOCITY PLOTS 5 AT K 6 AT K ts 6 7 2 THIRST OUTPUT Radial Plane Plo Figure Velocity Distribution 115 MASS FLUX BT 3 4 AT K 5 AT 6 a eA 5 77 er ee A 1 Af Figure 6 7 3 THIRST OUTPUT Radial Plane Plots Mass Flux Distribution 116 THERMAL HYDRAULIC DATA This chapter details the content and sources of i the thermodynamic property data for light and heavy water and ii the empirical correlations used in the THIRST code Normally the user will not change these However if it is desired to investigate the possible effect of introducing different correlations this may be accomplished by simple coding changes to the routines mentioned below The user can easily insert his own property functions to cover different temperature and pres ure ranges different fluids Pertinent information related to each property or p
5. i44926946 ot AJ GZLu thc COS O 8 i rni HC Oe SORA ON OF OW JF O co O TORUM NO OOM FIL AD MOR FINA UR ODOT DAN MRANNMMMMM mmm io qid Pa Pd gm AORN DE Pee eee eee see esse tees a a tad Le ESOT MUN RAT ANG TH 4 E ORIS INEN 3 9 Ia Pe oU CO UMAR 3 3 OOM OWRD II oue OPM OM OI WOR eee ee eee Ae UII TAO SOC OOS COCR ODA etn We a o F Aa oe GI NMOS ANADIMOQNOOCOTDOWP fs C v 0 2 Puy OU VS AN 5 FINN e SU MINV O TOAN 62 ONE HNN IO OO CI SOU weet el led n TO THIRST OUTPUT Detailed Output Velocity Field 6 5 Figure gt _ _
6. 1 89 The Standard Execution Deck At this point the major effort of preparing the data deck is complete It is now necessary to enter the THIRST job into the computer system Execution control cards can vary between CDC computer installatiors However the following decks are included as examples and operate satisfactorily on the CRNL systen For a full explanation of CDC control cards see references 13 and 14 The decks consist of the following JOBCARD containing job name and account information CONTROL CARDS directing execution 7 8 9 END OF RECORD CARD DATA DECK 7 8 9 END OF RECORD CARD 6 7 8 9 END OF JOB CARD Card Content Explanation 1 JOBNAME BXXX YYY Tttt IOiii JOBNAME 7 char acter job name 2 ATTACH THIRST ID THIRST CY 1 Attach THIRST object code 3 THIRST Execute THIRST code Card 5 to 1 N 2 The code Exec Content Explanation 7 8 9 END OF RECORD CARD N COMPUTER DATA DECK 7 8 9 END OF RECORD CARD 6 7 8 9 END OF JOB CARD above simple execution deck will execute the standard THIRST without reading or saving any RESTART data Advanced ution Decks are discussed in Section 5 5 Job Submission A complete listing of the entire deck is given in Figure 4 5 This t disc may now be submitted to the CRNL system urnaround time for a large job is not particularly fast we uss in Chapter 5 some additional features of
7. lt SOD EN 19 0 8 Y x ITPRINT Dec wast gt NAST toam ONTO 0 15 025 1 IRN m oec LN 1 9 4 Le DAM Dm 24 0 We Os onono WINNS 0 27565 0 4513 0 6209 0 3026 0 978 1 1539 1 3575 0 1 gt 117 gt 225 4 3447004 LJ lt lt SUL 47 0 0 9 9 1 wn O o em 4511 a amo 1 1 E wz ete OL xe Qe Wood 0 Z t0 OD ocu gt u wu 4 440 OU uL mx a a 63 0 um Mo AAD 3125 over on J ozoze D gt 49 s 2 SN CERMASOOUCW EMOTO N LSU D Ou ar t0 1 CU UC OU U m wn QOQON ODOT
8. m QUU Og 3334 E o wo 5 J HII HROTOWNMe 84 m 0o NNAM ov gO t w t gt z SU DD NAM S 2 INO HUN NS 5 Z O 4 44 2 v a o o DOTT LES 5 n On ae A Osemooo op ex JNNNMMNNY e SUMMA 5 hd 2 gt MIO Mm ANON AANA 4 OAO DON O0 OORRTOOMN an HO AON DORK NOUO Z eut 2o Qeecensven iDeeecvoeveo O 849664444 oe 2 59240559 Q 2 O Payapa 274 t ONO 222 OooRNAwwuwunA ta 4 uni 5 x noone 6 B rion auo prods 5 mona 0 ANNANN A
9. 2 lt 5 o o x lt 5 4 r4 XY 2 ON 2 1909 Of DOOW DK HEMRONOND 3 FAMAM MO HOU gt wes 3 vmomovug gt ODANA gt Z COO umo uno RNIN X yews o 259 mre 2 z 5 gt WOKMAZOND 3 wrndtoovtd 7 OO HO Uu AHITOWOWOR O HAOMMOR pem MD 2 MITA IND O Neeserecs O 516466422 OOO b D ORD L OREM ION RO kP DONMA IUT ENT HIORON gt x u er 5 5 PUMPER H Ie NI PCR E H HANM PUO one Z Meri Ti Ai 2 2 2 ot a gt a an a2 o Ter WOODA DA 2 Z seg IN QWwOOQOHUFOR X CORAL O
10. Woe ARMM NASS FLOM AFTER SCLUTION OF ENERGY EQUATICN AND CORRECTION OF GENSITY AP 4 9 210 AO eeouvaom c erstens AA uen 2 EANNAN QU PO qn IOLA qe ASA ONU 44 44 Onna Ah OG Herm IN FOU mors WM DANO Ongs DTS eu E n SORRY qom do HEMOS OD ANI THD ANNAN ADO et RAMANA mese emm bel IUUD t gt mro ey to KAP OM Oa enge a C uwD QR uo vi WR TOP HO VOR ease sts 9 11 a SOURCE 16 5 23 10 T 3640 132 5 2 765373 255 9 am vo On COL 5 5 DE 23 es SHROUD 213094 618 5 8 4 AXIAL VEL VALUES CN HOT S 4 8 ABCVE DIY ME 5 7 28179 72833 RADIAL VEL VALUES 6 SEC HALL TEMP PRIMER 271 28 t 5 THERMAL VALUES AT 25
11. uu c Qoouooo aawa n Hmmm 197 sou 78 53 Source Code Changes THIRST OUTPUT MODIFIED DESIGN 7 1 Figure NO PREHEATE 600 RESTART 1 000000 36 10 AKDIV st 0 9 0 Os C Ce f 9 0 2 Oe 0 0 Oe 0 0 0 9 1 11 116111121212121211222 2222 121111116211121212221212112222222 IJPlOT 91010161Cc1197111111112131212111120 Pus 31 5910 11 12 13 14 15 16 17 18 19 23 21 22 23 24 2 KEYS amp ik 2 25 32 33 35 IFRINT 11 115 00 ELAX 11 PM LUE a 8 491 102 03 06 21 415 42 022 025 3 435 amp 55 b 0 12 23 212 056 ob 78 89 1 1 19 0 21 27566 612 6269 6926 378 1 1639 1 2845 1 3575 t 12 Qa 9 27 5 63 11 99 1174 135 153 171 180 3 Ba 210 CURE 6 DCRNH 7 FEES 2 IF 2 EH 1 10 48FCH 4 XCENTC KCENTH KF TRELT 0 0 AKBR 9 03090 AKWIN 112 T g ggetoc uz 9495696 CONS 1 5878 2 15 16778006 02 a FLOT 0 FL CHC 2550 000 124020006 02 9 679007 PRIMO 1 066130 PSEC Mine n 250000 XWANE 22126053 L MC OATA CALL
12. 1 3f 1981 ae 01021 38 91 see i 4e lt O SD 10 957 911 9 T er D 3 tu 6 Q T 3 T 3 349391 8193 619 G142 T 3T43 T STA 1 51 3 6143 1 STe3 T 01 577 01 37 atasi AN or In 800 5247 19 1534 POIYEZN39 45 009 1y2113H10d AH e ee 24629522 9592494924 OPERATING CONDITIONS PROPERTY AT asa HYPOTHETICAL 609 STEAM GENERATOR PRIMARY PPRI 948790000 N M2 FLOWTU 2484 9300 6 5 SECONDARY PSEC 541000000 N M2 FLONC 305 18000 Koss FLOWRH 23 700000 5 251 60000 CEL VALUES CALCULATED BY THIRST PRIMARY SATURATION VALUES 343 20115 5 348548 5 J KG 13 3 206299 03 3 6 1 59764138 02 M3 KG SECONDARY SATURATION VALUES TSA 65417748 CRL tN 75 79814 KG M3 ES RURI Res STEN 2125241518 02 1225458400 05 SECONDARY SUBCOOLED INLET PROPERTIES 749868 28 J 6 DENC 893 52101 OTHER SUBCOOLED PROPERTIES ARE CALCULATED AS NEEDED CONTROL VARIABLES 1 IVF 1 INPUT VARIABLES AKBR 000000 0900 640090000 AKE
13. SEP 2 2 25 R k k E SE e 2 SEP SEP Aapp Kopp anis DATA VARIABLEI Loss factor calculated for the separated liquid floving from the water level to the last modelled This loss is assumed to be frictional loss CON2 and treated as flow in a pipe 2 2 2 plane in the downc uer MM ns 2 x FLOW _ 2 5 1 _ D 2 D 2 20 2 2 where CON2 amp qi BS XDOWN X L see data 39 Hydraulic Diam Diam Clearance Downcomer Area or Saturation Density of Water 0 316 is based an estimate at the velocity calculated from a recircu lation ratio estimate y RECIR FLOWC _ FLOW Anc 0 56 Parameter used to optimize the This ratio provides the code with an estimate CONS estimate of the recirculation of how the pressure drop through the modelled ratio region changes with recirculation ratio i e total flow The code uses this value to estimate the recirculation ratio needed to balance the pressure loss against the driving head CON4 is set at 2000 If severe convergence problems are encountered other estimates 1 2000 1000 should be tried DATA DATA VARIABLE 57 58 Thermal conductivity of the tube wall material W m C Friction pressure loss for the baffle plate Baffle thickness
14. eoooooo 3 NH lt aamen nure PAG ety a ODOQOOOO FARRAR Interpreted Data Summary of Geometric Parameters THIRST OUTPUT Figure 6 2 3 103 THE WAIN GRIC LCCATICNS THE AXIAL DISTAMCES METERS I 2 1 0002 02 I 3 1 200E 01 I 4 6 T 7 5 600 01 I 8 6 720 01 1 9 4 za 1 13 1 606 00 715 151 2 8602 0 0 1 18 3 210E 00 1519 e 1222 260 220 1 23 4 610 00 124 1 27 5 944 00 ES 1 32 9 4 06 00 1333 5 00 00 1534 13 1060 1 RADIAL DISTANCES IN METERS Jz Jt 2 1 006 01 J 3 2 757 01 J dz 8 026 01 780 021 8 1 154 00 J 9 THE GIRCUMFERERTIAL PCSITIONS IN DEGPEES 2 3 000 0 K 3 2 700 01 K 4 6 84 100 01 K 7 9 9005 101 8 1 170602 K 9 R211 1 7105 02 12 1 800 02 THE CISPLACEO GRIC AXIAL DISTANCE Iz 2 ub 232 8 550060 7 ne RADIAL DISTAMCES 8 903E C1 5 THE CIRCUMFERENTIAL ka 8100 13 2 2 LCCATIONS FOR VELOCITY COMPONENTS IN METERS FOR THE ANTAL VELOCITY 3 1 750 PRIFARY FLUIO FLOW OISTRIGUTION KG S 3 e500E 02 0 a 150 01 9 0 if 5 0 7 235
15. MAR ENDO nuns e F FI 2 3 DN 107 022 MASS FLOW AFYES EXACT SOLUTICN OF CONTINUITY EQUATION USING WEDGES ANO RING GEOMETRIES IMEAL ANCE 55 SUM OF WEDGE we we 3 o ONDD HOO INDAN Devescene SWE NEY D HR ADIN Fn HADDOU 7 INNO Aye oy ADD DI MONS ANNO 2 us PWN Ue ONIS WAND PADS PWN HOO SIND 134 OMMAN ANA a inn 2009020 0 90490 ao 414 nue iHe ANENA CUTWER 440 NUN 00401460 SOAM F LIXIN TS CCLD UWP RMN nint t v COT MEIN OO a MASS FLOW AFTER SCLUTICN OF ENERGY EQUATICK AND CORRECTION OF DENSITY we we HOON Ana Uo 3 2 PUNE AR o WOR PS Dm f MoD APO HE Pe IO D 4 4 420 miha
16. 5 3 11 24 CN COLO SIDES 359 4 Ep AXIAL VEL VALLES CN HOT 3 1 T 9 OIV PLATE 19 5 7 07 193 WINDOR 6 5 85893 10 An SHROUD RADIAL VEL VALUES HEAT FLUX 28856 303 23 NALL TEMP PRIMARY TEMP 273 13 ENTEALPY 18994 94 5 SEC 5 THERMAL VALUES Final Iteration THIRST OUTPUT MODIFIED DESIGN Results Graphical Output 1 3 FIGURE STEAM QUALITY CONTOURS COLO SIDE HOT SIDE som s THIRST OUTPUT MODIFIED DESIGN Final Iteration Results Graphical Output Figure 7 4 1 Quality Distribution VELOCITY PLOTS 1 CM 4 94 M S 5 k on M Aet e SINE EET POM oe gt we a 707 7 7 7 4 PEN UE 4 7 weil 4 7 477 42 4444 utt Ke 2 a 24777 Jagd 2 52 24772 7 y ze 4 aT I 33 Ke 5 kes 2 4 i 8 e 1 H pS xtti 55 2 OR 7 he t 11111 kao nidis e lt 11111 por y 47 c Te Y omm 11111 l rS lt AT 1 32 1 25 aca NOS 444411 UOTE RUE s
17. o gt x Se See p MO E RA 8 4 4 i Ke10 10 0 4 e RUNS tittet 1 PIQUE D NE Ke11 ARUM 414 11 a a 2 E S Race an lt lt Lo ant AT I 21 ZEE AT 1 14 K serene ar KS 4 RS 7X D pr LITEA N t d toate 4 11 K 10 K 10 X 3 p RSA Ke yaad kei Cosa Ke 2 dile Messi gt ta Figure 7 4 2 THIRST OUTPUT MODIFIED DESIGN Final Iteration Results Graphical Output Velocity Distribution MASS FLUX 1 1961 9 2 5 iu Ke esL COLD SIDE HOT SIDE Ta RT e t S 4 4 iiit a tte te ac tye Tras tery 777 AT 1 33 B 2 4 5 EDD EE duree re PETRA ca 272 27 qa ee eee 2 LIP PI SPERA 4 3311111111110 1725 I kac 5 r1 ie 10 3 4 kenn ee ce ee ee gt AT 1 14 Ke 7 K KB KAS 4 7 11 Ql M ae x a Figure 7 4 3 THIRST OUTPUT MODIFIED DESIGN Final Iteration Results Graphical Qutput Mass Flux D
18. or amp z m 2 2 i Ings I TEN EES MODIFICATIONS 7 CONTROL CAPOS 144 ANM FDR 0 OO ANM OUO nim 37376 OUO TOD23023 23390 4 AAAA HA ANYANA atur onse MAGNUS aL dnd unam fui un uw eeek i hip Hi ee bee REPEL EER Eri I cte i Er ere BOE EE auus Lu n re EE Oe ucc uc De nr ac Dc a y Oe D O uc Qe Oc Dc ur ML 4 Qc dl lt A ed ed Lat det tL E o ed 4 a ed et e t lt lt lt P P P REA EASED DIB AAD AMAA AAA UA a a z 2 2 2 gt an gt at 2 2 2 Tu c a 2 525 4 lt gt toe xL gt 2 daw ath enxe Der o vIr gt 4 4 lt lt oet lt Bano gt x 59 o q a AX Xo 2 Ore o0 Mra Z gt r 4 E o 7 gt da gt 2 or 51 Sr egari x AI zo MO OF ae aa
19. HOT 10 23 5 17595 5 IN DCWNCCMER 23 19 2 18 366 5 IN UBEND 32 52110 996 23 5 9 COLO 5 4 U 8EN 5 n 5 8 017208 ABOVE DIV PLATE 21 RADIAL VEL VALUES IM sHRoUD MINDOM 610 5 de 295 45 PRIMARY TEMP 271 10 WALL TERP 6 6 SEC THERMAL VALUES t 25 Summaries THIRST OUTPUT Iteration Iteration 2 Figure 6 4 2 ITERATION NUMBER 58 5 39756 MASS FLOWS AFTER SCLUTION OF MOMENTUR EQUATION FOR U VELCCITY ESTIMATE CF RECIRCULATION RATIO Qa wo WU T TORO Cy 9444 THIF DOW DF QDMA aro DMO 9 AD re 9 eaan 4 ren WW UAI Soeowmwnooo Punto eq OOM 4 26149 apio Jatt mus Ce Fari nuQ Quaderni OO Hem aa ANNANN etd 0 3 DING Go Dpto SOR CIO D o DA etit Hv AN ONO me Qocoomeco NUM par boo ene wanye Dgo OO EX OON DO a Wing 3 QN
20. ITPPD 3 use the homogeneous expression for area change Baroczy s for parallel flow and Ishihara s for cross flow Note that each multiplier is multiplied by DR RHOM DEN This is necessary because all Pressure droo calculations n SOURCU SOURCV and SOURCW are based on the mixture density RHOM instead of the liquid density DEN 621 Two Phase Pressure Drop Correlations Cont d TABLE 7 4 CORRELATION COMMENTS Baroczy aw 2 55 2 gt 500 2 45 65500 0 2 2 22 9 He Pg Ishihara 2 1 8 8 1 1 x OET TABLE 7 5 Void Fraction Relationships CORRELATION 1 The mixture density based on one of three different void fraction relationships is calculated in the sub routine DENS B 5 1 8 user choose of the three relationships by setting IVF as follows IVF 1 homogeneous expression Pa 2 Chisholm correlation x 1 pi l a 3 Smith correlation homogeneous Chisholm Smith S 0 4 0 6 IETT TABLE 7 6 Heat Transfer Correlations CORRELATION parallel flow Single Phase Secondary Side calculated in the function subprogram hd 8 0 4 ENTRY HTFl calculates the parallel flow p d depen dent coefficient HTPL This is done from START once 0 58 0 4 p
21. 6 Ag_ 9 E x x Note that this formulation also automatically handles possible extreme cases in which all flow directions but one are in towards or out away from a control volume Formulation of the Source Terms For stability of the inner iteration it is essential that the coefficients remain positive after the source terms are incor porated Thus in 3 20 SP must be negative Cases in which SP tends to be positive are catered for by artificially augmenting SU For example if 5 one may write SP 2 50 2 SU will then incorporate the old value of V and SP will ensure the formulation is both stable and implicit This section completes the overall description of the model implementation The following chapters contain detailed instruc tions on how to use the code me BG APPLICATION OF THIRST TO ANALYSE THE PROTOTYPE DESIGN Specification of the three dimensional model must include details of all relevant geometrical fluid flow and heat transfer parameters is emphasized that the process of modelling steam generator 1 heavily on diligent assembly of the specifications optimal choice of grid layout and of course correct preparation of the input data This chapter is intended to guide the user step by step through the considerable effort required By means of a detailed example we illustrate the entire procedure required to prepare a
22. momentum Energy i h Sy 2 6 THIRST also solves the primary side energy equation which for a differential length of tube 64 is given by _ _ Adi aE 32 2 7 i where ns are the primary fluid mass flux and enthalpy respectively is the distance along the tube d is the tube outer diameter d is the tube inside diameter and is the heat flux at the outer tube surface The heat flux is calculated from UC TO 2 8 where Tp is the primary temperature Tg is the secondary temperature and U is the overall heat transfer coefficient based on the tube outer area given by 1 4 4 sedo A n 2 2k 2 9 i p Here hp and h are the primary and secondary heat transfer coefficients respectively and its the thermal conductivity of the tube wall material The source term in equation 2 6 is related to the heat flux by S 2 10 where A is the tube surface area per unit volume Modelling Assumptions The governing equations are based on the following assumptions and cetmplifications 1 The flow is steady incompressible and homogeneous 2 The shell and shroud walls are adiabatic 3 The inside shroud wall is frictionless 4 Laminar and turbulent diffusion are negligible in comparison to the frictional resistances and heat source 5 The distributed resistances due to the presence of tub
23. 510 7231 4 16 xN em wo Le IN PRE n 5 ABQVE DIY PLATE 21 12261 5 19 Iteration Summaries IMARY TEMP 288 61 THIRST OUTPUT Iteration 1 WALL TEMP PR 270 77 5 5 4 Figure 6 4 1 25 FTERATION WUMBER MASS FLOWS AFTER SOLUTION OF MOMENTUM EQUATION FOR U VELOCITY wo oe ow os am I1 we ws wee v ae an 2r ain an ae ae on rm ae we ae lt EM cM ro on on on un m WO Ph C ONUS WOR ROOK Octo df d ONDO 200 0v ee ONM PRON DODANA cay OS Afin ha tet H9 ON lt o WON O00 3 m ODIDIN D Am Cete tr Ca ooo NANA nin INN NUYS 3 uim ac IRE RADI WD Ort Diu AO DU POM opu 00 494444 86449 2 ODON DoD ONTO QOO S em nmmne tat Fat POUR FW gt Pow DYN
24. Chen correlation is the only two phase heat transfer relationship used in THIRST Because of their non linear nature boiling heat transfer correlations require considerable coding work to ensure convergence and stability In view of this it is recommended that the user consult with the authors 1f he wishes to insert another boiling heat transfer relationship These correlations are discussed in detail in reference 5 TABLE 7 2 Single Phase Pressure Drop Correlations CORRELATION COMMENTS 3 3 22 8 Smooth Bundle Parallel Flow Pressure Drop for Re 25 172 2 p a 516 calculated in the function subprogram FRIC 0 294 0 132 0 29 3 Re lt 25000 ENTRY calculates the p d dependent coefficient 20 227 PDA 1 This is done from START once per program 0 066 Re gt 25000 run ENTRY FRIC11 calculates the friction factor as a function of Reynolds number This is done as required from SOURCU SOURCV and SOURCW 6 8 28 1 p d Ro Smooth Bundle Cross Flow Pressure Drop 0 92 calculated in the function subprogram FRIC m 0 62 n 77771 ENTRY FRIC2 calculates the p d dependent coefficient PDA 2 This is done from START once per program run ENTRY FRIC3 calculates the p d dependent exponent PDA 3 This is done from START once per program run ENTRY FRIC12 calculat
25. OR ERIN FUNC 6 umo 0 om acta 2 nO COMIS Upewesace TED e ooo CONO NS 0D WR roc Benen Oud 110 0 BROAN E Af UNDO Ww HM Se GENSITY AND CORRECTION ICh SQUAT SCLUTICN OF uceov MASS FLOW AFT WORD lt lt 7 AYNAN Jute Seen DON Dey 12 DV uU wer r NANM g rN Q eR Wood win 4 2 SOC 44 POR eo eu a OD INIA OW vns OMe eus 1 44 m ORREAR AWW Mom NUNN MOOR noQ 44 Noonan ODN TAT OW DOU uus 5 AAA mH yin ecu CX 3 MARGOT DOL rin TOT 700 Qmomqpe 7us Au OS omm uU fO e AUN nmm dw r HEN ec 4 2453 2058 SUM SOURCE MAX SOURCE 10470 assa 446 OUTLET Guar 4 a oo uoce ax ee MF 31 5 RC 4 451 1 b
26. MASS FLOW AFTER I DO epu 040 4060 MOV OH OR OI FUND du 4 QM AH PHP DORON oUW Qmm ap utu Ue 2900 0077 391 145 16 3 766 245 153 323 COLO SIOE ADOT HJ 4 4 94 PARDO DEE QIN CNS ANS LOM duo eo aD OL MM viet FACH e D ae cf m HAMANN NV emi 5 a Wie ART PACING NO 232333223 AQMOOOIN OTS E 5 333 SUM OF RING MASS IPBALANCE MASS FLOW AFTER EXACT SOLUTION OF CONTINUITY EQUATION USING WECGES AND RING GEOMETRIES 105 aa we we N on rf MMO Od enm 2 14 4 Cu faro dias o lj NIS vro Qnin qr Jut D HOO S rrr DON ON DOD HANDOUT Of razz x UD IN HI HORM 060406 7 UND eo 400 HONT Jap et Veecneocnoc EMM NORA OO OI ANUNAY ul d Og D rot OMAMA iO DO up tf 964441 MONIO T
27. Own 4 SOOO NI UU PS COLC SIDE SICE 312 945 0 oon 4 ANM 92 FDA D MA nM Hm UJ vifo CS f OVO 4 0 rt eO 4 Man D Cura toram HooThowow etd idet OOO Sum S 157 10 9 7 90002 MAX SOURCE ee QuRC 4019 sone Qua 15737 AVG OUTL T 26 2319 5 In OCWACCHER 2 36492 5 IN UBEND 3 2 74956 5 5 COLO 510 5 OS ERE 23 5 9 23 5 4 RADIAL VEL VALUES IN SHROUD AXIAL VEL VALLES 99 CN HOT S K 5 SECT 32 66534 8 U BcND 18 re iN PRE n 5 61794 ABCVE DIV PLATE 21 WINDOW 6 5 53325 ae 296 07 WALL TEMP PRIMAR HER IMARY TEMP 47 29 THERMAL VALUES 25 5 SEG 1850 Iteration Summaries THIRST OUTPUT Iteration 6 4 4 Figure 59 ee ITERATION NUMBER 5 39756 SULUTION CF MOMENTUM EQUATION FCR U VELCCITY ESTIMATE OF PECIRCULATICN RATIO MASS FLOWS AFTEF caro cron Ban NOUO ON 4444 TAA AS Oumar ener fin
28. QNIN San ce Orn SOO QUOND AD M Qiu NS ul 982 24 ADO ets i 34D DO HO mero HOON Qoooo Dade WUINOM OR oO eu o 00 Heooooodoo umm Nm ORE AMAA AUFL OO HDD Usowomngy HORN HOOD 3 nnnm 2 ve Wham ud Weaveaces 2 BRS ram 108 0614 Sut RING PASS IMBALANCE MASS FLOM AFTER EXACT SOLUTICH OF CUNTINUITY ECUATION USING HEOGES AND RING GEOMETRIES WDDM HORN 2 o0 AD mon A eura QNM FAN jui GNU NY amy Ie oun CR 2 OTN Ou a aus r ANAM 4 auc Bett AUST o NOWRA Qum oe eo riam omuoo 4 4 MMM 0 OWN MMA MT og INS Qmm mem m p o o uS OM of Qr uM oen ui 2 mme meros WR ANID A Ogg NH 4o sod
29. o AOD T Uv INOW ODIO Yleasecee mo kN de got Qnin mmn or T HII d Nm mmm 267 240 199 E 22 611 785 857 15 671 384 175 3 9 231 ASS FLOW AFTEF K PLANE SOLUTION OF CONTINUITY EGUATICAFOR AND CORRECTION OF U VELOCITY 343 anuo ANNY SIOE 255 325 AMO AN NINN coa ge TR HIN gU OP aC An WOR ef Quin 0400740000607 wes ee v Piaf Om OR P OUR 535230080 HM d or pmo uero gu m gor Alin 2000 mam OD OMIMO H MP Me um NN MMMM mg on f AN 368 494 428 377 426 299 385 399 369 452 L4 D A COLD SIDE HOT SIDE 443 271 IAIN 3 Cu f C3 Qm 2 0 HNGOMA DAA Dame MIND Dao NO ANNA ANVR AN enm HROWdMRAS 942 8 PLANE SOLUTION OF CONTINUITY EQUATICAFOR P ANO CORRECTIGN OF U VELOCITY i eC ISO CE I AMO i Ie IUIS nts o o Mil eap te fe US NOS eu
30. 1 2 V I 1 J K carries the 1 flow out radially U I 1 3 1 K carries the flow un nast the baffle js Am SNR Z NN OG U I 1 3 2 K of velocity blocked by the baffle V I J K carries the flow in radially TYPICAL BAFFLE 1 1 SHROUD PARTITION PLATE The I planes are located so that baffles lie midway between them The location of the J planes matches the baffle cuts for this particular K plane however the cut will not match other K planes and the program is set up to handle this Figure 4 1 Grid Layout at a Baffle Plate SHROUD SHELL 1 2 1 0 1 1 0 1 velocity inside shroud U I 1 3 2 K downcomer velocity 1 1 7 2 Shroud window velocity Z aan TUBESHEET The I and I 1 planes are located so that the top of the window lies halfway between them 9 1 and J 2 planes center the shroud Generally more grid would be located in the shroud window to handle the sudden change in flow direction Figure 4 2 Grid Layout at a Shroud Window 4 2 4 Windows A third example Figure 4 2 shows the grid layout required near the shroud window opening The radial grid lines J 2 and 1 are located to center the shroud The axial velocity at 1 1 3 2 corresponds to the downcomer flow The radial velocity at I 1 J 2 K corresponds to the window flow where the downcomer flow enters the heat transfer area
31. 80000 Lansari 2820822 iM x 3 9 Rly dt 9 97921 2 11 Oott y 113 5 50 3000283 2 31038 05000 eS 41035 MH 5SnIOv 27 373 005 0119 roy 000067 9 3358 090000 7 OLNI 6 Tadd 58836091 31018 20 3000116 2 19 00042 20 32090000 3720 45 0 2 919033 111155 11 14434 113 94014 805 009 BH 2 8013 370008 503 0020022 W30N03 20 30000308 3 433402 25 3095999 5 1 ais Q0 iS xq 5 9 HEU 146 20000457 000 09 33157 6 1033 3 9 HJN32x 1 9493 0 1 t agg it Tg 08 51 92 Aa 003331 103331 ET 909331 4 3 DOT f T SET 171 3 4 42 3 M 2 846577 S 92 1 665177 846 9209 5929 53648 Y 9 CI 5107 206 8 6 95976 2 46 9844 9496 MERZ Taty 3264 2198 TEE 9567 Tete 90 2 9792 Teer 9w 7 T bpe g 29 poet aye ef fe 3 0 Gt 44 S s2 224 ste pe 80 0 8 z 2 6 ep ge TY 00 2 rT o 4 3 al L 285 32 12 9 2 75 Tur 4 4 91235 SE nE TE Ce 62 82 42 92 2 w2 t 12 02 61 at 4T 91 St T 27 IT 3 t 35594
32. RR E CONI HORA D A DOWNA ON ON D T VC AM AP NA DW ROR Qh AM AF ARON ee IM OU DR RENN ONON e NH eo d 009 o ef eee mimm i gg pi POOF Ip PLINY ESET gt dag ss SS a UJ UU UJ f 4 c UU LU UD NAR DOME SO 3090 AD ZPO nS Ft Nm On DONG opu 003 E FINTO HOW ERAR te reer ee eee ee ere WW WU Lug a UU HIA ADIN D ARO mra Im TIO 404 Ser d Iac rte WOO hose rs mpg 2 D ON I DOO e 3 QNIN Be 3 IIO OO OQ 0m 39 33394393 40 nd AABDOSONVANAG lt C oocooocooooooaococoocooooonocooooooo 24444444 424 4544 4444 4 4444 UU Oe I Tad LO
33. KS Ka 10 7 po d 0 25 OT 1535 Ke 7 e 5 Keo 7 17 64 7 7 k 3 10 11 a 0 2 AT 1232 7 5 KB Len KS N 49 7 NK AT 1221 K 9 7 4 x d Ke d we 2 1 HT 1 6 STEAM QUALITY CONTOURS K 7 _ ___ _ KeS COLD SIDE HOT SIDE Koi is j w 20 a 0 s gt 1 22 n 5 2 AT 1 33 1 Ka 7 8 Z n Ke5 ke 0 3 0 2 AT 1 25 Ke 7 K __ _ KS 7 SN 57 N 0 kejo 7 11 2 1 14 5 Ka Z n KS 11 2 8 L AT I 2 Figure 6 6 1 THIRST OUTPUT Composite Plots Quality Distribution OTT 4 09 MS VELOCITY PLOTS 1 esti KM BE Rig LS COLD SIDE SIPE Figure 6 6 2 THIRST OUTPUT Composite Plots Velocity Distribution MASS FLUX 1 1905 1 KG M 2 S K 6 8 C ux COLD SIDE SIDE 47 5 a MM Ee ge 4 eet 2 QUE ot S Ka 2 H 51441 gt QT 1 33 K K 6 ms TAA
34. L Diametrical clearance D Area approach A APP rea gap Ac AP Obtain from mater al property data Also see data no 45 and 52 fL ov AKBL The variable is concerned with the second term the frictional loss 25 f 316 8 L thickness of baffle D diametrical clearance Because this loss s based on approach velocities the area correction is included A 2 Thus FLDB 316 AKBL FLDB RUP APP FLDB 46 DA 5 DATA VARIABLE Friction pressure loss for the See data This variable stored the friction coefficients FLDT thermal plate 58 mentioned in data no 49 and 53 7 2 fL APP BV app da Gt A GAP 2 A FLDT E 22 E _ 2 AKTP FEDT 2 Fouling is assumed to act uniformly over the RFOUL tube surface Resistance due to fouling on external surface of the tube ITEMS 61 69 ARE OPERATING CONDITIONS 61 Feedwater flow rate This is the total steam generator feedwater FLOWC 1 kg s flow rate Reheater flow rate Some designs include a reheater circuit FLOWRA kg s The flow returning from the reheater s assumed by the code to enter the steam generator at the top of the downcomer If there is no reheater circuit set this value to zero Primary flow rate kg s 64 Saturation pre
35. Iteration Summaries THIRST OUTPUT Figure 6 4 5 60 Iteration Po 4 2 AXIAL VELOCITY M SEC ee ae PAAR A SEL 2 10 9191919929909090 3090999 999 30 339990 307000 230323 3273 339 9 203 2 3000 23 23 300 35 1533 3 33 ETT Tree err er Te Tees MUU ule iu a ad a DD dL POM APIO O AMUDO IFN IE 3094 OR WNW DU Pe AOR FIA guo QD SIV F m 60 Tun MTANI CJ PU NI CUNT Y49 33994619 13919 33730 23 O3 39390 9 3 30902 3 GoooOOSoogOooooOoooDooOopnogooooooodogoao toco ttes t e Wwy LO HL UU dd a UJ UU UU LUUD LU LOU UJ UU LU Ld Lu L3 LE DM APU AMINO cO 10 2 QUOS CO HC a OO AMDA VN edo ANM HAD A SON DIDTO DANN SOUL A Seer sever eee PAIA UUM NIA 990 040 tte bet et WW WWE OW fra
36. The order of variable storage TPRINT 1 axial velocity IPRINT 2 radial velocity IPRINT 3 circumferential velocity IPRINT 4 mass flux 5 steam quality IPRINT 6 primary temperature IPRINT 7 tube wall temperature IPRINT 8 static pressure IPRINT 9 density of mixture IPRINT IO local heat flux IPRINT II porosity order ot variable storage RELAX RELAX 1 axial velocity RELAX 2 radial velocity RELAX 3 circumferential velocity RELAX 4 pressure correction RELAX 5 enthalpy RELAX 6 inactive RELAX 7 tube wall temperature RELAX 8 pressure RELAX 9 density RELAX 10 wedges and rings RELAX 11 inactive Allows the user to specify the quality con TCON tours of interest Can have up to 15 values Zero values 7 che end of the arrav are ignored Sets the last execution step On completion LASTEP Last execution step of LASTEP iterations the computation ceases and detailed printing and plott ng starts I9 81 83 Parameter to specify when during the execution plots are to be made Parameter to specify when during the execution the variables specified in IPRINT will be printed out Parameter for overriding the time limit routine Width of the plotting frame when I planes are to be plotted both on the left and on the right of the vertical cut see data 73 IF PLOTO 0 plots are never made IF PLOTO 1
37. for the plotting routine Other data values that deal with the preheater could be altered however the changes made above ensure that these data values are never used An example is AKBC the baffle resistance which is not used because ICOLD never equals 3 or 4 143 These changes were inserted as illustrated in Figures 7 1 and 7 2 Results are summarized in Figures 7 3 and 7 4 Two major prediction changes are evident 1 recirculation climbed from 5 4 to 7 06 2 The heat transfer dropped from 662 to 577 The quality profiles undercut the larger subcooled region on the cold side Mass flux plots indicate a uniform flow distribution across the bundle In concluding this chapter it should be pointed out again that these changes were to illustrate the flexibility of the code and not to compare two design types Each design could be altered to maximize its performance Although the number of changes required to handle this new configuration were small it required a good overall understanding of the code to identify them We therefore stress that when faced with radically different designs the user is advised to consult with the authors 1 PAGE 21 14 01 80 07 08 UPDATE 1 3 598 IDENT UNLABELED CLOFL x x x 2 x NOES 2 On 2 29 c z wz ro o x 4 wae ou a son 9 X
38. 00 1229 0 1 35 24620680 1534 0 IN_METERS FOR THE RACTAL VELOCITY 73 Je 1 0665406 9 112126900 2 40 POSITIONS IN DEGREES FOR THE CIRCUMFERENTIAL K 3 128002201 Kx 3 608901 5 Ks 8 140806402 Ks 9 142606902 10 X RXRXRARmMImM Figure 6 3 lt A 2 SEOCB 55458 955157 55286 55 12 255442 455284 455183 55086 5500 3 1 30 54262 58516 454872 5248 455298 59872 oF 4516 oF e262 56150 L 455290 53498 53899 56466 055154 55156 5465 53893 e 3495 053290 5 452446 52763 53202 54069 55611 455011 59069 53102 52763 52846 6 51716 52055 52722 51679 55670 034870 53679 52722 452055 51715 7 50975 0513575 52166 453298 354729 54729 483298 52160 4851375 450975 MMMM 4 en 8 50 263 450717 5151 52923 55563 054589 052923 51614 50717 50 263 DGE 01 00 51 6 00 10 00 0 06 01 425 62390 10 1 388 06 5 6 3 Cc 0i K710 1c 530E D2 3 049751 050243 451218 52650 54486 2 54486 52650 4612185 250243 49751 THIRST OUTPUT Summary of Grid Locations 1 97108 DEL 23 5 9 SIDE 69189 23 5 5 AN COLO 7
39. 32 04 On ODN Ud oua DOr et HONOR ARO Vleeeveces n ANDOR C9 ADV D o rou 2 0 CTICh CF VELOCITY SOLUTION OF CONTINUITY EQUSTICAFOR P ANG CORR MASS FLOW K PLBNF Qa es mem 7 Nee recess AN OW Apr ERAS a a 020 4 IND Tw TOE PIN 444 Af d LOO orn 2 mm o DAM 1 2 44 HRN FOUR RONAN se oou MOO eom MIN AE WR ad 444 DOR GVO BOO DODO o o tue Dro mero cs GONE 454 or je qu S One QU REUS fer I HANNS DR WENO t Unna ewe Do du C SO DP Ano Cuv Oe conim toi TOD Qro anu rx SOLUTICN OF CONTINUITY EQUATICAFOR ANG CF u vELCCITY 1 MASS FLOW AFTER unaona HOO nonon A Ad FU Oxo OI HO o 2 OND HUS Chea fe ONIN UND
40. 43 4 4 Preparation of the Input Data Cards 64 4 5 Sample Input Data Deck 65 4 6 The Standard Execution Deck 69 4 7 Job Submission s s s s s e b e o o o 70 5 SOME FEATURES OF THE THIRST CODE d v p PA 5 1 The RESTART Feature e 72 5 2 READIN Feature a es ss 75 5 3 Time Limit Feature 77 5 4 Advanced Execution Deck 2 2 78 5 4 1 Job Control Statements s s so s 79 5 4 2 Input Deck s 4 a 81 6 THIRST OUTPUT 8 de cat eue eR VU de 582 6 1 Printed Output Features s s e 82 6 1 1 Preliminary Output soa s gt 802 6 1 2 Individual Iteration Summary 83 6 1 3 Detailed Array Printout 86 6 2 Graphical Output Features s s s 89 6 3 Interpretation of the Output s 92 6 4 Treatment of Diverging Solutions 95 7 THERMAL HYDRAULIC DATA aa e so a 4 116 Thermodynamic Properties s w e s e 116 122 7 1 7 2 Range of Application 7 3 Empirical Correletiona for Flow and Heat Transfer 122 TABLE OF CONTENTS conti
41. Rie ee FUND c n Y moth iui 24 Qu NImmwO eet PAWN OMO erf oco a HTS rinm THO HUNAN WIR OTOL IT OID IN AT Me Ou ANU Deoveveres 7 ONDI WHOKROORY ORS ROR BOM Har OO S CP Nu qf Qaceor teo Ber quor lt AAA Odo trn 60 49 3 3 Dh Ow LIne OPA CO uuu pur reo FUR 146 SUM OF RING MASS 19834 ANCE SGUATION USING WECGES ANI QING GECMETELES MASS FLOW AFTES EXACT SOLUTION OF COATINLITY 2 Quia etm Ae omo nuc 24 4 DD Din ANN JANO at A Or gine AAI Ow www Mra anum me Pou 2 APU Ort trot NID INDE FOr a UND FDI 7 2 OU A0 SOR 4 ME Oe am rea naf t HIN Ha rror mag o ar epar n Ow UM Um e o ena 3
42. U bend supports AKDIV I k or indicate where no obstacles occur 0 These loss factors are used to calculate the pressurd loss relationship for the circumferential velocity between the hot and cold sides 6 to plates or supports the tubes are handled independently If ITPPD 1 THOM used for parallel cross and area change If 2 BAROCZY CHISHOLM used for parallel cross and area change If ITPPD 3 Separate correlations used See Section 7 3 1 homogeneous correlation 2 Chisholm correlation 3 Smith correlation 26 ca DESCRIPTION k shock loss factor for the baffle plate resulting from area change contraction and expansion DATA VARIABLE The loss for the baffle plate 2 L pv fp 2 This data is the AKBL portion which is the pressure drop due to the contraction into the annulus between the drilled plate and the tube It is based on the approach area 2 AKBL DEVICE Also see data no 58 k logs factor for the tube support broached plate based on shock loss due to area change Same as data The tube support plates result in a pressure mo 45 drop due to an area change This value is based on the approach area k loss factor for the larger broached holes in a differentially broached plate Same as data Ino 45 In some design
43. dd pon F dn 4 gt 2000 Sjo eae a TABLE 7 6 Heat Transfer Correlations Cont d CORRELATION COMMENTS 0 8 0 33 hd Gd Cou y 0 023 1 k x ug f Primary tube side Heat Transfer Correlation The parameter A is a two phase heat transfer coef ficient multiplier It is activated when the primary flow is two phase The temperature dependent parameters kf Ur and 65 f are calculated n the function subprogram PROP ENTRY PROP3 as 0 67 0 RCONVA 0 67 120 47 primary enthalpy RCONVA is based on heavy water properties over the temperature range of 245 C to 315 C It is valid in the pressure range of 7 MPa to 11 MPa to an accuracy of 0 54 Note that h is the primary side heat transfer coeffic lent referred to the tube outside surface SET TABLE 7 6 Heat Transfer Correlations Cont d CORRELATION COMMENTS 2k X See d Wall Heat Transfer Coefficient The wall resistance referred to the tube outside diameter RWALL is calculated START CWALL ky the thermal conductivity of the tube wall material is specified by the user in READIN Fouling Resistance RFOUL is specified by the us r in READIN GEOMETRICAL RESTRICTIONS AND POSSIBLE VARIATIONS The basic steam generator geometry as illustrated in Fig
44. plots are made at the end of the job PLOTO 2 plots are made after each iteration Note If PLOTO 2 a very long plot file will be produced Careful selection of values for IPLOTI and IPLOTK are necessary data no 72 and 73 PRINTO is set up the same as PLOTO in data no 78 Note that PRINTO and PLOTO may be reset in the logic to turn the PLOTTING and PRINTING routines on or off THIRST has been set up to print out all the variables make plots and write a RESTART tape if the execution or INPUT OUTPUT time has been reached To suppress this feature set TIMELT to zero Width of the plotting frame when only I planes are plotted on the right side of the vertic cut Height of protting frame PLOTO MI 29 E A 4 VARIABLE 1 1 9 84 Extra integer input locations Data put into these variables is common to ail subroutines 85 Extra real input locations 1 1 9 9 Preparation of the Input Data Cards Once the data specification sheets have been completed it is a straightforward matter to transpose the requisite information into data card form In THIRST the data is all processed through a routine called READIN READIN not only reads the data into core but also performs a detailed check on the completeness and precision of the data supplied The course of execution of the program is directed by the RESTART feature w
45. to fit particular geometrical features Final Assessment This then completes the grid layout One may find that the number of planes in each direction could be juggled to better model the design Once the grid layout has been finalized and the geometry of the design described to the code relative to this grid it is a major undertaking to alter the grid location Thus it is important at this stage to review the grid selection carefully Preliminary Data Specification Having examined the design layout and selected the optimum grid location we must now provide the code with the information required to model the design This section dzescribes the contents of data sheets The specification sheets are included in chart form to emphasize that specification must be completed and verified before any actual input data cards are prepared Each chart is divided into the following columns COLUMN 1 DATA NO for reference purposes COLUMN 2 DESCRIPTION COLUMN 3 DATA VALUES to be taken from specifications COLUMN 4 REMARKS any manipulation of the DATA is described or a summary of options is given COLUMN 5 VARIABLE NAME code name used in THIRST COLUMN 6 FINAL VALUE value to be used as data The data is arranged in functional groups as follows GROUP 1 Preliminary Data Items 1 7 GROUP 2 Geometric Data Entered by Grid Indices Items 8 21 GROUP 3 Geometr c Data Entered by Value Items 22 41 GROUP 4 Cor
46. 11 MPa The properties are 4 calculated in PROPRTY ENTRY PROPI Liquid Specific Volume m kg Steam Specific Volume n kg 811 TABLE 7 1 Fluid Properties and Parameters Cont d FORTRAN ACCURACY OF PROPERTY OR PARAMETER NAME IN POLYNOMIAL PROPRTY EXPRESSION 4 Feedwater Subcooled Properties Feedwater Temperature C ser specified The subcooled feedwater properties are calculated from the user Reheater Drains Temperature C ser specified specified temperatures TINC and TRH The properties are calcula Feedwater Enthalpy J kg at 5 MPa ted in PROPRTY ENTRY PROP1 and at 4 to are valid for the temperature range of 150 C to saturation 6ii Reheater Drains Enthalpy J kg Feedwater Density kg m TABLE 7 1 FORTRAN ACCURACY OF PROPERTY OR PARAMETER NAME IN POLYNOMIAL PROPRTY EXPRESSION 2 Primary Subcooled Properties Heavy Water Temperature C 0 01 at 9 MPa PROP2 0 30 at 7 to 11 MPa Heat Transfer Coefficient Parameter RCONVA PROP ENTRY 0 12 at 9 MPa PROP3 0 50 at 7 to 0 67 0 47 11 MPa 1 1 kgm s 0 33 2 1 905 Secondary Subcooled Properties Light Water Liquid Viscosity keen t s 2 5 at 4 to 6 MPa Fluid Properties and Parameters Cont d COMMENTS The primary temperature and the heat transfer coefficient parameter are calculated from polynomial functions of enthalpy They are valid over the temperature ran
47. 3 6 or 5 3 13 Substitution of 3 13 into 3 14 gives N N 1 5 4 1 171 U P P or oN 5 i i U P Relax when M N 1 Sy Sy 1 0 Apo 3 14 This pre relaxed equation can obviously be solved using the identical techniques already discussed In THIRST all equations are pre relaxed in this manner except for the pressure corrections and density calculation Equation 2 23 returns a pressure correction rather than the pressure itself Pressures arising from this correction may be relaxed according to 3 12 but this is not usually necessary Density may also be relaxed by 3 12 Upwind Biased Differencing It is well known that symmetric central difference representation of first derivative terms in transient equations leads to unstable numeric behaviour 10 11 Stability is usually ensured by incorporating one of two devices in the numeric scheme The first artificial dissipation adds an artificially large viscous term to the equations The second upwind differencing uses difference formulae which are asymmetrically weighted towards the upwind or approaching flow direction Both devices stabilize the computation and in fact it can be shown that they are numerically equivalent 11 Central differencing has the same destabilizing effect in steady state and computations can be stabilized by the sam
48. 3991990 39909999399 WWW aa ui IDO 40 tU ANN DRD TORO N ROR SIE A AAI ADF APNI A DMO DANS SOMMNAIP ON DUM f 3 PID ANAM MM MM NNER AYA reetaetoereerienre Atty rt ttt 993909009 33390 369 30900 WWW WO Ju bg UI a L yd FIOM OMNI HOUR NOM SOMME TWN D HOON fO e TOV ARUN m clar ADMD ODIO O0 0 ON DANON ONORA I DIVA ORO AAI 97 0 60 7 U 400o0200000500050020000009200020003500 PORES M orla a ODD ADMO H ADF UUM A DOW OR MON J ADAU SOS TINT BONO 3 MP PO TOORA CH RADI ID AHANY ADDOT ANON 03944464 THO non nt monton enum TUN gt
49. 4 1 to 6 4 5 During the progression of the solution to convergence the following information is summarized on one page for each outer iteration a Iteration At the beginning of each iteration prior to any further calculation the EXEC routine prints the outer iteration number b New Estimate of RECIR only after the ninth step After the ninth iteration tho program begins to calculate the R CIRculation ratio Because the solution technique is iterative the value will change until the solution approaches convergence c Mass Flows at Planes of Interest The mass flows are calculated at I planes selected by the user The user can choose any or all of the I planes by using the IMASSF d parameter see data no 71 Table 4 1 The mass flow information is preceded by a line indicating the point within the iteration step at which these calculations were performed The mass flows at designated I planes are plotted in five columns of eight entries each for a maximum of forty positions if required The mass flows are given for both the hot and cold side The calculations are made tn MASSFLO MASSFLO is called whenever the axial velocity or local density is changed Summary of Overall Performance Variables and the Convergence Indicators At the end of an iteration step summary of the overall performance variables and the The user has no direct a convergence indicators are printed control over this fo
50. 494 MAMANE eT 2 wane FAO OO uie DDI 5 QUATICN ANG CORRECTICN CF GENSITY GY MASS FLOW SCLUTICN OF E we ne 9 5 Day gt wena tern gt 4444 MOAR r3 nsns Cua O fnaoe oro AMIN SOR nw FIND 9444 ANIM AA Qaod monn OEM goeooooorc scr ROR DSR MAD 2 WOR AR 3 US eO NI Oc HON OO NNI Wau NIIS 2 pori MPO ED SIDON 7 9 11 5 5 2 74394 IN UBEND 32 5 9 AXIAL VEL VALUES 4 DIY 10 5 RADIAL VAL VALUE PLATE 21 7 81779 d WINDOW 83353 EAT FLUX 17256 IMARY TEMP Hi 296 07 MP PR 273 38 ENTHALPY WALL T 8044 3 1 5 SEC 5 TWERHAL VA
51. I must be specified as follows If IcOLD 1 1 no plates ICOLD I 2 normal tube support ICOLD 1 3 outer baffle plate see data no 23 ICOLD 1 4 inner baffle plate see data no 22 ICOLD 1 5 thermal plate ICOLD 1 6 differentially broached plate usually first plate on hot side Location of all baffles tube support plates etc on the hot side See Layout This array is the same as data 8 except that it applies on the hot side Shroud window height on the cold The last axial plane lying side inside the window on the cold side Su Top of the feedwater distribution bubble Feedwater inlet window lower limit Feedwater inlet window upper limit Height of the preheater The last axial plane lying inside the window on the hot side t Last axial plane passing through the distribution bubble First axial plane lying inside the feedwater window FEEQWATER w Last axial plane lying inside the feedwater window Last axial plane inside the preheater 1 1 VARIABLE TDOWNH IPEEDU 9 Eg DATA VARIABLE Effective elevation where the The code treats the conical ISHRD downcomer annulus expands section as a change in porosity halfway through the expansion DASHED LINE IMDICATES CODE TREATMENT 325 RPM Starting elevation of the U bend The I planc located at th
52. Sen eee M w a gt Gn OM 74 ex 2 n 4 ae 1 HO oo arvo NO FO v Tuxxeludeg Ze se 21 v MIC ae w Oro WZ J4 ROI 26 Zar 2 ezo 5 Widest 4 4 nzu uo 255 Pee Down OIIX v ALOG IN Ou WOU 2 HHOOROT MOR Rane 2 amp u wi ZI lt w w E a 2 a z lt gt c o oo o Lj 23 QC Uc az Qr a OC ac uc CE uc Qc 020 02 Qc ac ac ac Oc oc Qc Ur De Qc Qc Cz Qr Qc 0 0 IN dL E 16 1 ab t 4 ale tet lt lt lt lt lt lt lt lt Cai aal i t en qi ian JH ou qui gU PD eie Pn uy pes je ir P pen par n PET eo m ROV Aa aoa Wage qYYzrz FLUGHC FELOWRH FLONC h gt SUBE H FLCHC F LOWH x n a S oM
53. T ow eer _ oor fu Pme smx x i UL 3 gt HO a SRO OX O d O U U Ot je 9 P Execution Deck Figure 4 5 SOME FEATURES OF THE THIRST CODE While the instructions included in the previous chapter are sufficient to prepare an input deck an understanding of some additional features of THIRST is required for advanced use of the code This chapter describes some of the THIRST input and output options The RESTART Feature RESTART has been introduced very briefly in chapter 4 This feature was included in THIRST to reduce the repetition required in making several runs with only slight data modification and to enable results to be stored so that later printouts and plots can be made without re executing In THIRST all the variables are initialized in START with at best a rough guess Using RESTART the variables can be initialized with results saved from a previous run resulting in improved convergence RESTART has been set up so that a user can stop at say the twentieth step examine the results and then proceed further if desired The user can also run the code examine the results make logic or input changes and continue running the progran or merely change the plotting or numbering parameters and
54. THIRST analysis of a particular steam generator design We assume the user is familiar with the fundamentals discussed in Chapters 2 and 3 and now discuss Design Specification the hypothetical steam generator Grid Selection arrangement of optimal grid layout Preliminary Data Specification procedure for assembling the data specification sheets Preparation of Input Data Cards Sample Input Deck Execution Deck assembly of a THIRST job and submission to the computer 35 Design Specification The particular case chosen for this example is the hypothetical steam generator discussed in Chapter 1 and shown in Figure 1 2 Design parameters used in the current example are summarized in Table 1 2 A large number of variations of this design can be investigated using the standard THIRST code by specifying parameter variation through input data Designs which deviate from the hypothetical model in major aspects may require code modifications These are considered in Chapter 5 Grid Selection The first task is to describe the geometry of the design to the computer This is accomplished by superimposing a cylindrical coordinate grid onto the design and by specifying the location of flow obstacles in terms of this grid THIRST accepts a maximum of 40 axial planes 20 radial planes and 20 circumferential planes however due to a storage limitation the maximum number of nodes must not exceed 4900 In order
55. arranged to provide finer division in the region where steep gradients are expected for example near the tubesheet Following the now classical grid arrangement introduced by Harlow et al 9 scalar variables such pressure density and enthalpy are centered at the points of intersection of the grid lines or nodes As pressure is the driving force pressure differences generate velocities between nodes thus velocities are centered between nodes The resulting grid arrangement is shown in Figures 3 1 to 3 4 Velocities are considered positive in the direction of the coordinate vector The Control Volumes Finite difference approximations to the partial differential equations may be derived in many ways However the control volume integral approach has proved particularly successful in fluid modelling This is principally because it easily incorporates variable mesh size yet rigorouslv enforces continuity It does however introduce additional complexities as the finite difference form of each equation must be integrated using a control volume centered on the primary variable concerned Thus scalars are cons dered to be constant over control volumes centered at grid points while the ax al momentum equation is integrated over a control volume centered on the U velocity and the radial and azimuthal momentum equations are centered on V and W respectively Typical control volumes for each of these four cases are also shown in
56. brackets after the MAX SOURCE This information can be useful during debugging to indicate trouble areas If the location remains fixed and the imbalance fairly high one should examine the region for a modelling error The last three lines on Figure 6 4 5 contain local values of thermal and hydraulic variables These values should be compared with earlier iterations to ensure that they have The positions shown in brackets have been selected to monitor variables because they are particularly sensitive areas When we are content that the solution has converged we should examine the printouts to check numbers against intuition and then examine the plots to verify hat flow and quality patterns are consistent These outputs will not provide additional evidence of convergence but will enable the user to intuitively verify the results For instance the quality profiles on the I 2 plane could be superimposed over the velocity vectors to verify that the velocities are concentrating near the point of highest quality Thevelocity vectors in the U bend should indicate an outwards radial flow to the lower resistance regions The flow around the baffles should be well defined Having examined the output we conclude that for this example the solution has converged Treatment of Diverging Solutions If the solution has not converged we should either restart the program and continue for more iteration or examine the mod
57. concentrate on the last several iterations to verify that a converged solution has been obtained Later in this section we will discuss handling runs that terminate in execution errors Comparing iterations 58 59 and 60 Figures 6 4 3 4 and 5 we observe that RECIR has converged to the fourth significant figure which indicates that the pressure distribution has also converged Examining the mass flows at various stages in the iteration step we see that the values are basically stable Minor changes can be expected due to the nature of the finite difference technique however a swinging from value to another would indicate an inconsistency in the modelling between stages If the swinging is significant further debugging of the logic should take place In the middle of the mass flow printout is the SUM OF RING WEDGE MASS IMBALANCE As explained in earlier sections continuity is enforced simultaneously over groups of control volumes Control volumes are grouped alternatively into wedge and ring geometries The MASS IMBALANCE should approach zero at convergence however the level indicated is considered acceptable At the end of the run we also should be satisfied that the pressure drop value is stable that the heat transferred out of the primary PRIM H T is equal to that absorbed by the secondary SEC and that the source terms are sufficiently 11 location of the maximum imbalance is given in
58. d ROS PERRO 0 qm 0 023 C ENTRY HTF2 calculates the parallel flow Reynolds number dependent portion of the Nusselt number This Brosse ERA is done from SOURCH as required 0 36 ENTRY HTF3 calculates the cross flow Reynolds number dependent portion of the Nusselt number This is done from SOURCH as required Chen Correlation Saturated Boiling Heat Transfer Secondary Side hip h hab calculated in the subroutine SOURCH The saturation pressure and temperature dependent terms are calcula ted in the function PROP These terms are valid for light water in the range of 4 MPa to 6 MPa saturation pressure 281 TABLE 7 6 Heat Transfer Correlations Cont d CORRELATION ENTRY PROP1 AKBO and XTTK are calculated Chen Correlation Cont d 1 x Gd 0 58 0 79 0 45 0 49 E F 5 parallel flow AKBO 0 00122 0 5 0 29 0 24 0 24 H p fg g 0 6 to wl max 36 cross floy 1 Ue k 0 1 i Ve ENTRY PROP9 calculates dP dT 59 that be calculated is the derivative of the saturation pressure versus saturation temperature relationship TABLE 7 6 Heat Transfer Correlations Cont d CORRELATION 0 79 0 45 0 49 oe 062 goed 0 29 0 24 0 28 2 f fg g n 0 24 0 75 AP Se lt 2000 D ME zd 1 17 1
59. in the plane of interest are added vectorfally The resultant vector is printed as an arrow with its length indicating magnitude end angle indicating direction Mass flux contours are determined by multiplying the velocity vector calculated earlier by the local density Two plotting formats are available to the user Full Diameter Horizontal Cut Composite b Vertical Cut Compostte Full Diameter Horizontal Cut Composite This composite includes plots of values of the 2 and K N planes which lie next to the line of symmetry These are put out by the plot routine automatically Included on this frame are up to eight horizontal cuts through the modelled region corresponding to eight axial lines specified by the IPLOTI parameter The selection of horizontal cuts is made by the user in the input deck by specifying the number of desired I planes maximum of eight A negative sign in front of the specified I plane positions the plot on the left ot the Full Diameter Plot otherwise the plot appears on the right of the Full Diameter Plot No more than four I plots for tne left and four for the right may be specified If only four i piancs are specified all the plots should appear on the right as che plotting routine will redure the frame size Examples of this composite are given in Figure 6 6 which depicts quality velocity and mass flux profiles consecutively Vertical Cut Composite The second plot format is a composite of fo
60. iv compute dP from 2 23 v compute U V W from 2 20 and 2 21 If the equat on set were linear steps iv and v would complete the solution However the linearized equations contain some remnants of the initial estimate so steps iv and v must be repeated several times 16 Finally the energy equation must also be incorporated vi compute h from equation 2 15 vii compute from equation 2 16 The sequence i to vii is now repeated to convergence The iteration sequence may be summarized as follows repeat compute U from equation 2 12 ii compute V from equation 2 13 111 compute W from equation 2 14 gt iv compute dP from equation 2 23 repeat v compute dU dV dW from equat 29 2 21 vi compute h from equation 2 15 vii compute 0 from equation 2 16 In the THIRST program the outer iteration sequence is orchestrated by the executive routine which calls a separate routine to perform each of the above steps Thermal Hydraulic Data Fluid Properties and Parameters s mentioned n Section 2 2 equations of state for both the primary heavy water and secondary light water fluids are required the analysis These are incorporated in the THIRST code using relationships derived from standard tables Full details of these are given in Chapter 7 2 5 2 Empirical Relationships In assembling the terms of the differential equations any thermal hydraulic code must rely on e
61. local mass flux 921 TABLE 7 2 Single Phase Pressure Drop Correlations Cont d CORRELATION Downcomer Window Loss Factor K is determined by the user and is stored as AKWINDH for the hot side and AKWINDC for the cold side It includes the downcomer to window contraction shock loss plus the 90 elbow due to change in flow direction shock loss K is based on the window area TABLE 7 3 Separator Pressure Loss CORRELATION Cy CON1 CONI is calculated by the user and read in as input The separator loss factor Kg and the total separator area A should be available as design specifications for the steam generator of interest separator loss factor total separator throat area total mass flow through separators 1 homogeneous mixture density TABLE 7 4 Two Phase Pressure Drop Correlations CORRELATION COMMENTS All two phase pressure drop multipliers are calculated in the subroutine TWOPH and called from the SOURCU SOURCV and SOURCW subroutines Three types of pressure drop are calculated in the program area change i e expansion and contraction losses cross flow and parallel flow The ponding two phase multipliers are TWOA TWOC and TWOP The user can use the listed correlations as follows ITPPD 1 use Thom s correlation for all three pressure drops ITPPD 2 use Baroczy s correlation for all three pressure drops
62. number is ARATB Reynold s number in in baffles area See also data no 58 Aap SAA ORARATBEV te R oc app Gap area u u Diametrical where ARATB a clearance 53 Area ratio multiplier to determine data Reynold s number in in thermal 52 ap plate 65 DATA DATA VARIABLE 54 Loss factor calculated for the Separator two phase flow from the last oss Factor modelled plane inside the shroud k to the separator exit Keep is sep normally given by the manufacturer based on V is calculated actor ark sep by user and is generally much total than k low area before enter separator total sep Beparator Brea To calculate the recircu lation ratio the flow from the last modelled plane inside the shroud to the last modelled plane out side the shroud is modelled one ddmension 1 t aly is combination of the loss factors for the two phase mixture It is based on the total flow as shown below CON2 data 55 is the loss factor for the saturated liquid flowing out of the separators WATER LEVEL 1 From 1 to 2 area contraction into separators PYSEP where velocity in 7 SEP separator FLOW Vsep 2 k 2 Iur te maw P sgp 2p From 2 to 3 separator loss 2 22
63. out the DU DV and DW arrays after the CALCW subroutine These arrays can be printed using the logic in FPRINT They contain all the resistances in the model and can be checked to see if any resistances are out of line The pressure correction generated in RINGS1 and WEDGES can be printed out to identify trouble spots Printing out the pressure corrections for the CALCPK and CALCPIJ is more involved since the control volumes are not grouped 1 m When RESTART is set to 1 there should be no control card that attaches a RESTART tape to the program An ATTACH Statement is necessary when RESTART is set to 2 and 3 an ATTACH statement is present when RESTART equals 1 the program will run and output will be printed including any plots but the RESTART tape requested by negative value of RESTART will not be made and you wili get DMPX The dayfile will indicate ILLEGAL 1 0 REQUEST the FILENAME and FET ADDRESS as well as WRITE NOT AT EOI ON PERMANENT FILE Finally the user should document convergence problems and their solution so that future problems will be easier to track down 99 aoa cocoa 2 em acc ndug jo Azewung 1041100 LSYIHL 2779 5330 WA 135343 5358 23361 81 34 323 1211 2 ON avadv 303 2371 2 1 0 ON 11 91202491
64. pressure correction is done using an alternative iteration sequence In this sequence a further standard pressure correction is performed in CALCPI This imposes continuity over the I planes following the modified sequence used in CALCW 155 TABLE 3 THE PRESSURE AND VELOCITY CORRECTION SEQUENCE Routine or CALCPK CALCPI ALTERNATE STEPS WEDGE 1 CORRECT WEDGE 2 Purpose Solve continuity equation for pressure corrections by K planes Apply velocity corrections and compute new pressures Solve continuity equation for pressure corrections by I planes Apply velocity corrections and compute new pressures Solve continuity equation for pressure corrections by rings Apply velocity corrections and compute new pressures Adjust velocities tor continuity in neighbouring rings Solve continuity equation for pressure correction by wedges Apply velocity corrections and compute new pressures Adjust V velocities for continuity in neighbouring wedges 7156 Finally continuity is imposed on alternate iteration steps over wedges and rings In the latter two cases the resulting equations are not solved by alternating direction tridiagonal iteration but by direct solution of the banded linear equation set This is done fully by Gaussian elimination using the decomposition and back substitution routines MATSET and SOLN Auxiliary Routines The routines th
65. set up to 1 IMASSF which the hot side and cold side culate the mass flow in the axial direction mass flow will be calculated and for selected planes This information is printed out printed out any time the axial velocities or densities are adjusted Any number of I planes may be specified up to NI Selection of the K planes to be This variable allows the user to select any IPLOTK plotted number of the circumferential planes for plotting Note the K 2 and K N planes are automatically plotted to give the tirst frame and should not be requested again Selection of the I axtal planes to Tte plotting routine is set up to plot up to IPLOTI be plotted a maximum of 8 horizontal cuts This variable is used to specify the I planes of interest For example Yf IPLOTI 10 the 10th plane vill be plotted to the right of the vertical cuts see Section 6 5 for more details IPLOTI 10 the 10th plane vill be plotted on the left of the vertical plot Note there must be only 4 specified for the left side negative number and 4 specified for the right side positive number 75 76 Select the variables to be printed out Relaxation factors Contour intervals for the plotting routines 77 This parameter allows the user to trim the ITPRINT output down to variables cf specific interest IPRINT 1 the variable is printed IPRINT 0 the variable is skipped
66. standard version of the THIRST code which has been written for analysis of a hypothetical steam generator containing many features common to CANDU designs Figure 1 1 In particular it is a natural circulation steam generator with the following features integral prebeater tube matrix with round U bends annular downcomer with re entry through specified windows in the circumference Geometrical specifications and nominal operating conditions of Such a hypothetical design are listed in Table 1 2 for a typical 600 MW thermal steam generator A simplif ed diagram of a natural circulation steam generator with integral preheater is given in Figure 1 2 The area inside the shroud is completely filled with tubes except for the central tube free lane between the hot and cold legs and the annulus between the outer tube limit of the bundle and the Shroud The surface of the outer limit of the bundle in the U bend is spherical The primary fluid enters the right side of the sketch flowing up inside the hot side tubes transferring heat to the secondary fluid en route The tubes turn through 180 in the U bend region and the fluid returns down the cold side The secondary fluid enters as subcooled water through the integral preheater where baffles force the flow to cross the tube bank in a zig zag pattern to enhance heat transfer At the preheater exit this flow now raised to saturation temperature mixes with flow recircu
67. the NINJNK card In such cases of multiple definition the definition encountered earliest in the deck takes precedence so the new value will be used 4 5 Sample Input Data Deck The data deck sheets in Table 4 2 have been prepared from the specification sheets of Section 4 3 according to rules outlined in Section 4 4 I as 256 2 CODE 133 5 VIVO LAdNI LSYIHL 67 4 1 M j 4 43365 LNdNI il HIN LSYIHL THIRST INPUT DATA SHEET a gt o 9 gt gt 2 w un 0 1012 o 2fafs H 5 z c 4 1 lt
68. to appreciate the selection of grid locations the user should understand the staggered grid arrangement used in THIRST described in Chapter 3 Essentially velocities are centered between grid lines in their corresponding direction and centered on grid lines in the other two directions as shown in Figure 3 1 An axial velocity for example has a control element with boundaries as shown in Figure 3 2 The top boundary corresponds to the plane the bottom to the I 1 plane The left side boundary is located midway between the J and J 1 planes The radial velocity has a control element that extends between J planes and straddles I and K planes And similarly the circumferential velocity extends between K planes and straddles the I and J planes Baffles Figure 4 1 shows how the code handles flow around a typical baffle We observe a radial flow to the left under the baffle an axial flow around the baffle followed by a radial flow to the right above the baffle Note that the baffle lies in the middle of the U velocity control element and the radial control elements lie on either side of the baffle We can see that axial grid lines must be located such that the baffle plates lie midway ustween them Partition Plate Figure 4 1 also shows the code treatment of the partition plate The circumferential velocity W corresponding to the K plane 1s blocked by this partition plate which is centered between the K and K 1 planes
69. uv o FON uma SA ATP co emo uw ANNAN 100 OD 4 QUI cn eO umet 4 mamme mm Jaana es riot 4420 OM egenos Weesseces rind Oococoocoo WARN 2 7 tH p NO oe UMW ORS QaMomcons mo pann GA ESEE HO FOR SOLUTION OF CONTINUITY EQUATICSFCR ANG CORRECTICN OF U WELOCITY MASS FLOW AFTER K PLANE apap ea ODN PRON lt NO 22 OQ NNommemm WIN OU Qun couv Uu e o3 OE HDO AVM aru Ann 00 094444444 Qe mmm mm Jenn tet o Ein nnn geo ema use gt MASS FLOW AFTER I FLANE SOLUTION OF CONTINUITY EGUATICAFOR F ANC OF VELCCITY 32 gen afmmon go DI Cur tht 3 3 0 r AUT NUMM TOR cm
70. will correct the modelled tube surface a large discrepancy may indicate an error in treating thetube free lanes or in the location at the start of the U bend IUBEND The main grid lucation Figure 6 2 and particularly the displaced grid locations should be checked to ensure proper modelling of flow obstacles For instance the displaced grid at I 13 for the axial velocity should in this case correspond to the elevation of the first inner baffle The primary fluid flow also included on this page ts distributed to reflect the different tube lengths Scanning the distribution one should see a drop in primary flow along the K 2 plane with increasing J When satisfied with the validity of the input one can proceed to examine the iteration by iteration output Figure 6 4 Of prime importance is the line bounded by asterisks Part of this line contains the overall parameters which must converge on single values RECIR PRESS DROP PRIM and SEC HEAT TRANSFER and QUAL The last two terms are the sum of the absolute value of the mass imbalance over all control volumes and the maximum mass imbalance at an indicated control volume These should approach zero at convergence By examining the line of printout as the solution proceeds one can assess whether the solution is converging to a single solution or oscillating slowly about the solution Assuming thet the run has completed normally without any execution errors we will
71. 2555629 AECL 7254 THERMAL HYDRAULICS IN RECIRCULATING STEAM GENERATORS THIRST Code User s Manual LSECONDIRY TY Sita 1 SEPARATORS i Ittttteer erggt mE Model of steam generator used far analysis THIRST code results Profi es of mass flux and s eam quality CARACTERISTIQUES THERMOHYDRAULIQUES DES GENERATEURS DE VAPEUR A RECIRCULATION Manuel de l utilisateur du code THIRST M B CARVER L N CARLUCCI W W R INCH April 1981 avril ATOMIC ENERGY OF CANADA LIMITED THERMAL HYDRAULICS IN RECIRCULATING STEAM GENERATORS THIRST Code User s Manual M B Carver L N Carlucci W W R Inch Chalk River Nuclear Laboratories Chalk River Ontario 1981 April AECL 7254 L ENERGIE ATOMIQUE DU CANADA LIMITEE Caract ristiques thermohydrauliques des g n rateurs de vapeur 8 recirculation Manuel de l utilisateur du code THIRST par M B Carver L N Carlucci et W W R Inch R sum Ce manuel d crit le code THIRST et son utilisation pour calculer les 6c n ements tridimensionnels en deux phases et les transferts ae chaleur dans un g n rateur de vapeur fonctionnant l tat constant Ce manuel a principalement pour but de faciliter l application du code l analyse des g n rateurs de vapeur typiques des centrales nucl aires CANDU Son application d autres concepts de g n rateurs de vapeur fait l objet de commentaires On donn
72. E 139086588 130000000 ARATTC 0 CON 94950000 ONS 11560000 CHALL 16 100008 FLDE 21000 000 LOT 1600 0000 AKDIV 1 1 9620 Is 2 1 06 10 3 1 0 10 4 1 0 10 6 1106 10 7 1 06 15 8 1198215 119 dns En Iz 12 1106215 13 1106215 fas 16 1 06 15 I 17 1106215 I 18 1108215 19 1 0 15 0 I 22 0 Ia 23 0 24 0 26 9 27 9 28 0 29 0 ta 31 2 11600 33 118690 34 6 Figure 6 2 1 THIRST OUTPUT ALATU QENSH AMU PRE ENHH 4 40000000E 02 176 67000 ese 5 3600000 1177907 6 J KG 903825 8900090 05 KG M S 5645620 219056 02 J KG N M2 1094156 3 J KG wwe Interpreted Data Summary of Operating Conditions 001 101 IWPUT CUTPUT ANC CCNTFOL FEATURES FESTART 1 0900006 6 604058 INCHES YL 1 000000 TIMELT 6 250 002 INCHES 1 0000000 PLCTC 5 INCHES TCON X 50 22 25 39 23 RELAX woo ad gt gt gt zzz zz gt gt ec zz Du gt gt gt 222 IIFLCT Sw Hi tes 2 IEEE MTM 22223223 uaaaaann IFASSF 18 19 2t 21 22 17 un m IPLCYK 10 5 IPLCTI 14 21 25 32 33 235 THI
73. E READIN then checks the value of RESTART and If RESTAKT 1 READIN terminates the run If RESTART 2 3 READIN uses the values stored on the tape from a previous run and continues executing thus only variabies to be changed are required as input data Time Limit Feature If the code senses that insufficient time remains to complete another iteration step and to print anu plot the output it will automatically call FPRINT for a printout and call the WSTART routine to write a RESTART tape The user can subsequently attach the RESTART tape and continue executing with additional time Both execution time and input output time are monitored but the time limit feature can be suppressed by setting the parameter TIMELT to zero If TIMELT is not set in input or is set to 1 in the input deck time remaining is checked at the end of each iteration The statements IFE RL NE O JUMP CATALOG ENDIF JUMP should be included in the job control deck to catalog a RESTART tape when a time limit is encountered Advanced Execution Deck The simple execution deck introduced n Chapter 4 is sufficient to run a standard HIRST job in which no RESTART tape is read or saved For more advanced use we now include an execution deck which will permit the use of a RESTART tape and also permit certain code changes to be made using the CDC program library editor UPDATE This advanced use deck contains
74. ED ARRAY DATA CALLED FCF ARRAY Figure 7 2 THIRST OUTPUT MODIFIED DESIGN GENERATO SOROS EERE EAA 05041400 A5 888 RN 54 GSES De Ge 9 9 0 0 Ge 0 0 0 Lo 0 1141111 113111 1431106 27 28 29 35 31 32 33 34 35 1 45 1 81 2 16 2 54 2 66 3 21 3 56 3 91 3 17 1 26 1 5 194 5 78 6 36 EEOL 12 2 BENO 36 1 2 LASTEP 60 1 500030 AKBRL 1 6006Cu 5 OF is 9800 ARATE 5 0 1156509 CONS ZLOCe CCL 1 04 EGNORM EDSHROX M 306 2803 FLOWRH 23 7025 FLONTU 2 2 PITCE 2 4511253 02 FLOTO 1 5 140000 PSHRC 1 10 02 3 2e 827 0005 05 RSHELL 1 445255 TINC 1 04190000 15 6456 3 6 300003 Data Summary ente 514 9 o p o 60 ITERATION NUMBER 7 06239 ESTIMATE GF FECIRCULATION RATIO EQLATION FOR U VELOCITY SOLLTICN CF MOMENTUM MASS FLOWS WD tor tma FAI 2 FINAL aTa OR REP Re S ee 2 Att PUNY AOE Oww sou S mmm mmm WENO AC OE Fi De MLD JA gii obo CONO HMO ANNANN Wh ANOLD NAD Cuna A PONN HOON
75. Each function is valid over the pressure range of 4 MPa to 6 MPa The properties are calculated in the function subprogram PROPRTY ENTRY PROP1 Liquid Density kg m 0 00 Steam Density kg m 0 01 Saturation Temperature C 0 03 Enthalpy of Vaporization J kg 0 01 DIDP is the derivative of the TSAT Liquid Saturation Enthalpy J kg 0 01 versus PSEC expression Liquid Viscosity kg m 1 575 0 05 BHDP is the derivative of the ENSS 21 versus PSEC expression Liquid Specific Heat J kg C 0 00 Liquid Prandtl Number 0 02 Steam Viscosity kg m 1 0 00 Surface Tension N m Change in Saturation Temperature per Unit Change in Pressure C Pa LTT TABLE 7 1 Fluid Properties and Parameters Cont d FORTRAN ACCURACY OF PROPERTY OR PARAMETER NAME IN POLYNOMIAL PROPRTY EXPRESSION X Change in Saturation Liquid Enthalpy per Unit Change in Pressure The Chen correlation parameters are Chan Correlation Parameters defined in Section 7 3 The two paraMeters are expressed as functions of various saturation properties Primary Fluid Saturation Properties Heavy Water Pressure MPa user specified Primary Fluid Saturation Properties are expressed as polynomial Saturation Temperature C functions of the user specified primaty saturation pressure PPRI Liquid Saturation Enthalpy J kg Each function is valid for heavy water over the pressure range of Enthalpy of Vaporization J kg 7 MPa to
76. Figures 3 1 to 3 4 2202 J I J x r PLANE x PLANE Figure 3 1 Grid Layout showing Scalar and Vector Locations 21 K J r 8 PLANE K 1 K K 1 I J x r PLANE I K x 8 PLANE Figure 3 2 Control Volumes for Scalar Quantities 22 I J x r PLANE I K x 8 PLANE Figure 3 3 Control Volumes for Radial Velocity Vectors t 23 I J PLANE 1 x PLANE Figure 3 4 Control Volumes for Circumferential Velocity Vectors 2j X The Control Volume Integral Approach Although the equations to be solved are integrated over different control volumes the procedure in each case is completely the same Thus each equation may be written in the form of equation 2 1 and integrated s rdrd dz 0 3 1 Although the integration is done formally by use of Gauss theoren SS ff as 3 2 v 5 the result is intuitively obvious from first principles It is cron A04 z 8 eoe Bpud rAgdz 85 dv 3 3 h 1 The quantities obviously represent the flux through the appropriate control volume face and the quantities represent the flux imbalance in each coordinate direction Integration of the Source Terms The source terms are frequently non linear in Integration of these terms is accomplished term by term The result can be linearized with respect to and stat
77. GENERATOR Thermal Rating 600 MW Primary Inlet Temperature 315 C Primary Inlet Pressure 10 7 MPa Primary Inlet Quality 9 034 Primary Flow Rate 2500 kg s Feedwater Temperature 180 C Steam Pressure 5 M a Steam Flow Rate 310 kg s Recirculation Ratio 5 5 Downcomer Water Level 15 m Number of Tubes 4850 Tube Bundle Radius 1 3 m 0 0125 m Tube Diameter 3 STEAM OUTLET ON N SS 3 4 SEPARATOR BEDS BEND RADIUS OF TUBES TUBE BUNDLE DOWNCOMER ANNULUS TUBE SUPPORTS PRIMARY FLOW IN TUBES SECONDARY FLOW iN SHELL PREHEATER SECTION FELCO WATER NOZZLE aan a gt DIVIDER PLATE HEAVY WATER OUTLET HEAVY WATER INLET Figure 1 1 Cutaway View of a Steam Generator 4 STEAM OUTLET 4 1 1 LIGUID RETE gt SWELL DOWNCOMER 35 PREMEATER p Poh RECIRCULATING FECOWATER FLOW IN 1 PRIMARY FLOR OUT PRIMARY FLOW COLD 51061 WOT 5 01 1 Seed eS Figure 1 2 Simplified Model of the Steam Generator The Hypothetical Prototype Steam Generator Although steam generators develore by different manufacturers share a number of common features it would be a prohibitive task to attempt to write a computer code which would comprehensively include all possible designs The bulk of this manual therefore describes the
78. IN SE woog EROR OI nen 04444 eU OMA 23 ANNP eT 5 mm COO ORA HORDAMaNS Od ORIS NMM JUD MASS FLOW AFTEF SCLUTICN OF ENERGY EQUATION AND CORRECTION OF DENSITY 4 UY ve Quod DN IND Em ANON name WADA HAD 4 Q4 0330 Dd OUO 4 4 MMI fur y rn uu Umm mmm Ce ttu boc 933 OIL 252 WORT TARP eroi D HO mL OO PONO Hn erm ANNAN 0 953353650 WOR iA IMS CX 4 0 ACHT ID OO ARAMOMNOO AE eceoe NI D M OOo Vieonsoo teo 2 uie IU COR S fO OQ Nace roo 508995885 eoo 4 m m a oo 25 L SUM SOURCE 3 gt 7162676 PRIM os AXIAL VEL VALUES
79. LUES AT 25 Summaries THIRST OUTPUT Iteration Iteration 8 6 4 3 Figure 59 ITERATION NUMBER 5 39673 MASS FLOWS AFTER SOLUTION CF MOMENTUM EQUATION FOR U VELOCITY ESTIMATE OF RECIRCULATICN RATIO os o we ou HILO 4 AO Nc AIA 2 2 qa a eio Doo 44 44 Sur rg IIO CHO INI DM met MOURN 4 M HOT 5105 OON mph e ried HINAAN ocooo ooo Cm mm re M res oo Cu OnOo tet Qomoooooo DR E MH iQ vt 0m ON UM 644 Q4 OR QOR N Neceervos WD a HASS FLOW AFTER K PLANE SOLUTION OF CONTINUITY EQUATICKFCR P AND CORRECTION OF U vELOCITY on we Amp un on MAP nouo is eun nus 10 24 A HAN 100 FIN DMI EHE ar dr
80. Produce new output without executing the program a The RESTART variable have six values three positive three negative RESTART 1 is used for a new run in which all variables and arrays are initialized with rough guesses by the program 11 routines are executed to attain a solution RESTART must always be set to 1 1f the number of gr ds or the grid layout is changed RESTART 2 is used to continue an old run 11 the variables in common blocks are set to values calculated in a previous run which has been stored on a RESTART tape New values for selected variables can be entered by including these in the input deck they then replace any stored values Variables which are not to be changed are omitted from the input deck The new run is then executed until a total of LASTEP iterations have been completed RESTART 3 is used mainly to obtain new output from previously completed run With this option all parameters except those read in to specify the type of output required are set to their values from the RESTART tape The EXEC routine then passes the control directly to the OUTPUT routine for the summary of the last step and a printout of all arrays requested through the IPRINT NN parameter Finally if PLOTO 0 the plot routine generates the plots requested by IPLOTI and IPLOTK After this output the program terminates no further execution is attempted 5 writes a RESTART
81. RST OUTPUT Interpreted Data Summary of Output Parameters Figure 6 2 2 GEOMETRY LCCATICNS 6 me KCENTH FEEGL JBRCH m on Zor zur INPUT 52265 AND LOCATIONS M M M M gt mm c gt gt a nmm x Miner u OQuiXu m D 0 e ooaoodo suo 2 gt 2 rer Pd D6Ocoooo eoooooo NR sooo Poe erat ECT CALCULATEC SIZES AN LOCATIONS XUBENT 52170800 AIMC 4 HEAT TRAMSFEF AREA AS MODELLED 81 THIRST 4531 7619861 18 4 8 7500066 102 1 0640 24451791 ATO ACTUAL HEAT TRANSFER AR MOCELLEO N TO FOR FACTOR REQUIRE 50112 TORR SUPPORTS ON HCT SIOE LOCATICN CF TUB IDIOT porti LLIT 5555555 nn ANN IIIIILIIL ed ANC THERFAL FLATE COLD SIDE TUBE SUPPORTS PAFFLES LOCATICM proe nt
82. S Pu Orso nOn Hoo o Lapa papal OO gt MASS FLOW AFTER K PLANF SOLUTION OF COMTINUITY EQUATICAFOR AND CORRSCTICN OF U VELOCITY MAM IND AW 4 30 2 Oo 2 JAIN RIMIS UA 13 209 Ft APN 00 4 A cim Oud rus r Hm MONT ow ett tA eate CMR NIRA 444 Qoo ceni HOO Dw ANNAN WAM CORK TH 2909099000 Qmm m mamn 14154 tet et tet WIPO HOM Qaa oue iem nO 4 OMIIN Trib GNE Te Hon IN 44 Man INNOD D AANA 2 MOIS If 02 Eo Ot waehopo dup eqs MASS FLOW AFTER I PLANE SOLUTION CONTINUITY EQUATIONFOR P ANO CORRECTIQN OF U VELOCITY on we on nn wa on X WOR OOo ARON OAM 37 ganana WN PRINS SLOT eo
83. Thus the location of the shroud and the location of the top of the window governs the 1 1 1 J 2 grid selections Axial Layout I Plane When allocating the grid the user is advised to start with the axial planes Figure 4 3 shows the axial grid layout on the vertical cut of the hypothetical model One can see the appropriate selection of the axial grid location around the preheater baffles The tube support plates cannot always be located midway between planes because of the limit on the number of axial grid lines availcble In such cases support plates will be effectively seen at lower or higher elevation than their actual location However this will not unduly influence the model because the tube support plates do not redirect the flow but simply add to the pressure drop Two axial grid planes I 7 and I 8 are positioned so that the top of the shroud window on the hot side is located midway between them The tor of the shroud window on the cold side is lower and thus the 1 6 plane located such that the I 6 and staggers the top of the cold side shroud window 39 AXIAL GRID LAYOUT 1 L H J 1 Et etm 5 pes ee Sii cm nn COLD SIDE awd Axtal Grid Layout Figure 4 3 When the axial planes have been al
84. an r HUP DOO ACU Ma tee DO Uf ON PUMA IBN 44 QOOOnM Oo 2 JAAHANS SICE MOO ANU 900940000070 et ttl 94 e uoo Pe e mmm Qooccococ eii ninm ei g TN io NOCO ON HN Vis eserves Q ve 17 a MASS FLOW AFTEF I PLANE SOLUTION OF CONTINUITY EQUaTIOAFOR F AND CORRECTION OF U VELCCITY ao ww anu mo ne ow CD gp QR Cao OO Re Duy 9 NAAR D DH ONM ft 1 7 5 BANNA wm gogoan iu DT 39min D 9 744 0 40 UN aou oo Ou 344 ND OMNI NNNMMMMA FINN DIP 94 One OPO 2 DUO etn gg HOT 510 OO Nano 303039 FAD HH OU OIA Qr meme Qnin ete SST INE 01 uw Aeesseces TOM HAM ty OND Han AH WAN DDP OND ONAT 2 HCW NONI
85. ant mathematical formulation and numerical solution tecnniques These are summarized in chapters 2 and 3 which follow 2 FOUNDATIONS OF THE MODEL The THIRST code computes the steady state thermai hydraulics of a steam generator by solving the well known conservation equations in three dimensional cylindrical coordinates This chapter states the equations involved outlines the overall solution procedure and lists the assumptions used to formulate the model and the thermal hydraulic data required The Governing Equations The THIRST code solves secondary side transport equations having the following general form ac Stove 1 dS Bows 30 Boup 85 2 v and u are the velocity components n the r 9 and z directions respectively 8 is the volume based porosity p is the mixture density 5 is the source term corresponding to the transport parameter The latter two for each of the five transport equations are listed in Table 2 1 In the table P is the pressure flow resistances per unit volume offered by the tubes baffles and are the and other obstacles h is the secondary fluid enthalpy 5 is the rate of heat transferred per unit volume from the pr mary to the secondary and g is the acceleration due to grav ty 10 TABLE 2 1 Transport 5 Equation Equation Number i Continuity 1 0 Radial I
86. arameter calculated in PROPRTY is listed in Table 7 1 Thermodynamic Properties Heavy water and light water saturation and subcooled properties as well as property related parameters are calculated in the function subprogram PROPRTY Saturation properties are computed from polynomial functions of user specified saturation pressures whereas subcooled properties are calculated from polynomial functions of temperature and or enthalpy The user can easily insert his own property functions to cover different temperature and pressure ranges or different fluids Pertinent information related to each property or parameter calculated in PROPRTY is listed in Table 7 1 Heavy water primary properties are based on an AECL proprietary program Light water secondary properties are based on the ASME steam tables THIRST is set up to handle a two phase primary fluid For Steam generators which have a subcooled primary fluid enter ng the tube bundie the primary inlet pressure is Specified rather than a saturation pressure The subcooling is specified by defining a negative thermodynamic quality TABLE 7 1 Fluid Properties and Parameters FORTRAN ACCURACY OF PROPERTY OR PARAMETER NAME IN POLYNOMIAL PROPRTY EXPRESSION 2 Secondary Fluid Saturation Properties Pressure MPa specified by Secondary Fluid Saturation Proper user ties are expressed as polynomial functions of the user specified Secondary saturation pressure PSEC
87. ariable name and its value from the data cards As each piece of data is associated with the computer variable name READIN can a ensure that all the data required are provided and determine which data are missing b allow the user flexibility in choosing the order of the data initialize the variables with values the restart tape READIN is set up to accept the title card with the RESTART value as the first card The title can be set up to 40 characters in columns 1 40 word RESTART 14 be located in columns 50 58 and the RESTART value in colucns 60 68 the RESTART name and value are left off READIN assumes a value of 1 If RESTART is set to 1 the second card must specify NI NJ and NK These variables specify the array sizes for all the variable arrays except IMASSF IPLOTI and IPLOTK If NI NJ and NK are not specified an error message is sent and the run erminates It the RESTART is set at 2 or 3 the run will continue from a point reached an earlier execution Because the values are stored in matrix arrays the number of grids in each direction must remain fixed Therefore any attempt at re specifying the number of grids changing the value of NI NJ or NK is ignored All the remaining data can be introduced in any order our examples we have elected to group the data according to its usage geometric correlation related operating co
88. array of values for variable NV is printed where NV is selected as follows NV 1 axial velocity NV 2 radial velocity NV 3 circumferential velocity NV 4 mass flux NV 5 steam quality NV 6 primary temperature NV 7 tube wall temperature NV 8 static pressure NV 9 density 10 heat flux 11 porosity This printout is generated by the FPRINT subroutine The PRINTO parameter calls FPRINT as follows PRINTO 0 the FPRINT array is never called this would be used where the user is interested in the plots only If PRINTO 1 che FPRINT array is called after exit from the iteration loop at the end of the run PRINTO the FPRINT array is called every N 1 iteration steps This tends to create large output files and thus is only used for debugging Careful selection of the IPRINT NV parameter is suggested Graphical Output Features The plot routines have been set up to produce quality contours b velocity vectors c mass flux vectors for any planes of interest Quality contour values are specified by TCON in the input deck Up to 15 contour intervals are allowed If less than 15 contours are desired then set the remaining position of the TCON array to zero and the plotting subroutine ignores them Velocity vectors are determined by first interpolating each velocity component to the grid nodes The two velocity components 1
89. at Transfer All the fluid flow and heat transfer correlations used in THIRST are given in Tables 7 2 to 7 6 The secondary side smooth bundle frict on factors and heat transfer coefficients are calculated in the function subprograms FRIC and HTF respectively These relationships are valid for equilateral triangle tube bundle arrays with pitch to diameter ratios ranging from 1 3 to 1 7 The user can easily insert his own correlations if those coded are unsuitable for his application Tube support plates and baffle plates are assumed to resist the flow only in the axial direction The tube support plate pressure loss is assumed to result entirely from the sudden area 123 change through the plate friction resistance is ignored The baffle plate pressure loss is a combination of shock loss plus frictional loss in the reduced area The value of the loss factors are determined by the user The method for calculating these factors is shown in the data sheets Two phase pressure drop correlations in the form of multipliers are coded in the subroutine TWOPH The user can choose various combinations of these multipliers by setting the appropriate value for the index ITPPD The mixture density distribution is calculated in the subroutine DENS The user has the option of calculating density using the homogeneous the Smith or the Chisholm void fraction relationships by setting the appropriate value of the index IVF
90. at form the inner and outer iteration sequences call a number of auxiliary routines which have not yet been described They are listed here Routine Function RSTART Read Restart tape SOMOD Find maximum source term FRIC Multiple entry routine for all single phase frict on factors HTF Multiple entry routine for all single phase heat transfer PRPRTY Multiple entry routine for all fluid thermodynamic properties TWOPH Multiple entry routine for all two phase pressure drop correlations VOLL Compute control volume parameters in tube filled regions BCUT Compute fraction of control volume in the tube free lane or occupied by a baffle 157 APPENDIX B REFERENCES AND ACKNOWLEDGEMENTS References 1 2 3 4 5 6 7 S V Patankar Computer Analysis of Distributed Resistance Flows 1 Introduction to the DRIP Computer Program CHAM Report B262 Combustion Heat and Mass Transfer Ltd 1975 R H Shill Private Communication September 1977 W W R Inch and R H Shill Thermal Hydraulics of Nuclear Steam Generators ASME Nuclear Engineering Division Confer ence San Francisco August 1980 W W R Inch D A Scott and M B Carver Steam Generator Thermal Hydraulics Analytical and Experimental AECL 6885 presented at the 5th Symposium on Engineering Applications of Mechanics University of Ottawa 1980 L N Carlucci Thermal Hydraulic Analysis of the Combustion Engineering Syst
91. ates of U U Dy 1 eil Ey Fy P 2 17 11 and 111 operate similarly on equations 2 15 and 2 14 to give 2 18 Fy P Fy P 2 19 The new values of the U V W matrices have thus been computed from the initial estimates using the momentum equations If the original estimates of all the variables were correct the values would satisfy the continuity equation 2 11 Invariably however they will not satisfy 2 11 but will generate a mass imbalance residual As pressure is the dominant variable in the momentum equations it is logical to adjust the pressure matrix in a direction that will reduce R to zero A logical method of adjusting pressure is to assess its effect on the veloc ty components by differentiating equation 2 17 with respect to pressure dU _ p dP U Thus we can write dU Fy aP av Fy aP 2 20 dW 15 Now if the pressure adjustment matrix dP in 2 20 is correct the new velocity matrix Unew dU will satisfy the continuity equation 2 1 Substituting 2 21 and similar equations for Vypyw and in 2 11 then gives rise to the equation BVorp CWorp AFg BFy dP 0 2 22 Or more simply R F dP 0 2 23 Equation 2 23 thus illustrates the pressure correction matrix dP required to eliminate the mass imbalance generated by the old velocity values Thus the reievant steps are
92. d its flexibility and reliability have been illustrated by extensive application the time is now appropriate to present the code in a formal manner It is our intent in this manual to present sufficient details of tne THIRST code to permit a new user to run the code and to obtain parameter survey studies based on variations of a reference hypothetical steam generator design Suggested approaches to other basic designs are also included THIRST Thermal Hydraulics In Recirculating STeam Generators BOSS BOiler Secondary Side DRIP Distrihuted Resistance In Porous Media CANDU CANada Deuterium Uranium Before presenting details of the code implementation and discussing the input data quired some background knowledge of the nature and function of steam generators must be established Steam Generator Thermal Hydraulics The steam generator is a cr tical component in a nuclear power plant because it provides the interface for heat exchange between the high pressure reactor primary coolant circuit and the secondary turbine circuit The integrity of this interface must be maintained to prevent mixing of fluids from the two circuits while thermal interaction must be maximizea for efficient transfer of energy to the turbine from the reactor Figure l l is a cutaway view showing the salient features of a typical CANDU steam generator The hot primary fluid from the reactor circulates through the network of tubes h
93. de a subroutine called RECIR estimates the recirculation ratio which will balance the driving head against the flow dependent pressure losses Recirculation ratio is defined as RECIR FLOW OF SATURATED LIQUID OUT OF THE SEPARATORS INLET FLOW where the inlet flow is the sum of feedwater and reheater drain flows If we add the feedwater flow to this liquid separator flow we have the flow in the downcomer annulus DOWNCOMER FLOW RECIR INLET FLOW FEEDWATER FLOW In terms of code variables we have FLOWH RECIR FLOWC FLOWRH FLOWC The code then determines the velocity at the boundary by dividing the new downcomer volumetric flow rate by the annulus area The downcomer enthalpy is calculated by summing the individual flows coming into the downcomer multiplied by their enthalpy values and divided by the total downcomer flow 141 FLOW FROM SEP ENTH PREHEATER FLOW PREH ENTH FEEDWATER FLOW FEEDWATER ENTH ENTIS TOTAL DOWNCOMER FLOW In THIRST the liquid saturation enthalpy is set to zero and all other enthalpy values are relative to this zero level Thus the above expression reduces to the following form in terms of code variables FLOWRH SUBRH FLOWC ENC SUBH FLOWH The enthalpy value at the I plane is set to this value and thus the boundary conditions handle the introduction of feedwater into the downcomer The code changes required to incorporate thes
94. e Baffle cuts must be parallel to the divider plate Code modifications are required if other types of cuts 1 normal to divider plate or other baffles 1 triple segmental are considered Tube Supports The user can specify any number of horizontal tube supports up to the start of the U bend The code can handle a vertical U bend tube support if it is located midway between the hot and cold sides Downcomer Windows The downcomer window heights on the hot and cold sides can be specified independently Once specified each window extends over the full 90 circumferential arcs on the hot and cold sides Separators three dimens onal modelled region can be extended to just below the separator deck The separators are treated as a one dimensional resistance Remembering that only of the steam generator is modelled 139 ADAPTATION OF THIRST TO A NEW DESIGN As discussed in earlier chapters THIRST has been generalized to accept minor geometric changes and most sizing changes As the user becomes more familiar with the code alterations to handle radically different designs will become easier to make Initially the user is advised to return to the authors for advice on preparation of modification decks to handle radically new designs An example of such modification is now considered In order to eliminate problems that can arise with a pre heater several steam generator designs introduce the
95. e changes are In START initializing subroutine D START 112 FLOWH RECIR FLOWC FLOWRH FLOWC this statement initializes the downcomer flow rate to include the feedwater flow D START 114 FLOTOT FLOWH this statement tells the program that the total flow 15 equal to the downcomer flow as all the inlet flows occur at the top of the downcomer 0 START 159 also D START 260 SUBH FLOWRH SUBRH FLOWC ENC FLOWH this statement initializes the downcomer enthalpy value 142 In RECIR calculating the recirculation ratio D RECIRC 65 RECIRC 66 FLOWH FLOWC FLOWR1 FLOWC FLOTOT FLOWH D RECIRC 67 SUBH FLOWRH SUBRH FLOW ENC FLOWH We now have introduced the feedwater in the top of the downcomer Our next task is to eliminate ihe preheater and the feedwater inlet For the most part we vill ieave the data the same if it is not related to the preheater fhe following chart contains the essential changes to remove the preheater Data No Name Reason for Change New Values 8 ICOLD Set plate loss locations 7 1 6 3 1 to the same as in IHOT 5 1 2 2 1 7 2 6 1 12 IFEEDB Remove preheater bubble by 1 reducing its height to I l 13 IFEEDU Set upper limit of feedwater 2 window to the 1 2 14 IFEEDL Make the lower limit feedwater 10 window greater than the upper limit so that no control volumes lie between the two 15 IPRHT Set the top of preheater to 1
96. e devices Consider for example a one dimensional central difference statement of equation 3 5 6 0 N 5 c NU Md So 0 3 15 This can be reduced to G a 25 3 16 5 C approaches C the denominator becomes very small generating undue excursions values particular if exceeds C very siightly a small increase gives large decrease in an impossible situation However if we add diffusion terms which involve the second derivative the resulting equation can be shown 13 to be D _ D C_ o 28 bp 22228 55 n 3 17 D D cC n s 8 Note that 3 17 will always be stable providing the diffusion influence D Dg is large enough Similarly on physical reasoning alone may consider that is swept primarily in the direction of flux The simple upwind statement of 3 16 already introduced in section 3 5 is Coes x 20 This reduces to bp s 20 3 18 which will always be stable Equation 3 18 is the simplest possible upwind formulation and is equivalent to adding excess viscosity Its use has been eriticized because it can lead to diffusion of the solution particularly when the flow direction is not normal to the grid axes 14 15 A number of higher order difference schemes which can be used to give more accuracy may be developed 10 12 and some of these may be i
97. e le d tail des hypoth ses employ es pour formuler le mod le et pour appliquer la solution num rique Laboratoires nucl aires de Chalk River Chalk River Ontario KOJ lJO Avril 1981 AECL 7254 ATOMIC ENERGY OF CANADA LIMITED THERMAL HYDRAULICS IN RECIRCULATING STEAM GENERATORS THIRST CODE USER S MANUAL by Carver L N Carlucci W W R Inch ABSTRACT This manual describes the THIRST code and its use in computing three dimensional two phase flow and heat transfer in a steam generator under steady state operation The manual is intended primarily to facilitate the application of the code to the analysis of steam generators typical of CANDU nuclear stations Application to other steam generator designs is also discussed Details of the assumptions used to formulate the model and to implement the numerical solution are also included Chalk River Nuclear Laboratories Chalk River Ontario KOJ 1JO 1981 April AECL 7254 i TABLE OF CONTENTS INTRODUCTION 1 1 Steam Generator Thermal Hydraulics 1 2 The Hypothetical Prototype Steam Generator 1 3 The THIRST Standard Code and its Intended Application uos Tinte frei 1 4 The Use of This Manual et de ui ces FOUNDATIONS OF THE 2 1 The Governing Equations gt gt 2 2 Modelling Assumptions 2 3 Boundary Conditions s g
98. e start of the IUBEND curvature of the U bend The radial distance from the center to the effective line dividing the reduced broached side from the normal broacbed size for differen tiallv broached plates In some designs the first tube support plate the hot side is differentially broached to induce flow into the center of the stcam generator The last radial grid line cor responding to the larger diameter holes is used to identify this point J gt JBRCH K plane on the cold side next to K plane near the ECENTI the 90 angle center of the SHROUD region en tw cola side BUBBLE X CKCERTC ENIM lt SHEEL K plane on the hot side next to the As 14 but on 72 gt 90 1 hot side ee ee TT PART Ti d PLATE Angle at which the feedwater K plane that lies just inside the iter distribution bubble starts distribution bubble 7 Distance from the partition plate to the edge of the inner baffle m Distance from the partition plate to the edge of the outer baffle Distance from the partition plate the edge of the inner baffle at the exit of the preheater half of the width of tube free ane between the hot and cold side m Used to determine which control volum
99. eating the secondary flow which evaporates as it rises inside the shell Failure of any one of the tubes would lead to expensive downtime for the station The most likely causes of such tube failure are corrosion and fretting of the tube material Corrosion can be minimized by regulating secondary fluid chemistry and by optimizing secondary side flow to minimize flow stagnation areas where corrosion tends to be highest of tube surfaces due to flow induced vibrational contact can aiso be analysed and 12231 flow conditions can be computed with sufficient accuracy The location of tube supports which minimize vibration can then be specified either case a detailed picture of the flow patterns under operating conditions is required The THIRST code provides such a picture 0 1 0 uU Babcock Babcock Babcock Babcock Babcock Foster Foster Combust Combust Combust Westing TABLE 1 1 STEAM GENERATOR DESIGNS ANALYSED Nuclear Thermal Power Manufacturer Plant Rating MW amp Wilecx Pickering CANDU PWR 150 amp Wilcox G 2 CANDU PWR 515 amp Wilcox Pt Lepreau CANDU PWR 515 amp Wilcox Cordoba CANDU PWR 510 amp Wilcox Darlington CANDU PWR 660 Wheeler Dariington CANDU PWR 670 Wheeler Wolsung CANDU PWR 515 ion Eng Maine Yankee US PWR 845 ion Eng System 80 US PWR 1910 ion Eng Series 67 US PWR 1260 house Model 51 US PWR 850 TABLE 1 2 PARAMETERS OF A TYPICAL CANDU STEAM
100. ed in general form S v Sy 3 4 Here the term Sp normally contains all coefficients of and Sy contains remaining terms which are generally but not always unrelated to gt Reexamining the equations in Table 2 1 it is apparent that the greater part of the programming in the THIRST code is involved with formulating and integrating the resistance components of the source terms using the appropriate empirical correlations This is done in subroutines with the generic name SOURC Integration of the Flux Terms It 18 apparent from equation 3 3 and figure 3 2 that values at for example control volume face n can be obtained to first order accuracy by upwind approximation for any variable A which assumes that the velocity vector convects scalars from upwind only Thus if all velocities are positive inlet flows convect neighbouring scalars outlet flows convect the control volume Scaiar Denoting the coefficients of by C and using the upwind approximation equation 3 3 is reduced to the orm T Cuty Su 5 3 5 where Cy is the flux evaluated at control volume face i 26 Collecting terms gives A MU Aids Sy n s e w h 2 3 6 A Ca etc 5 1 Once the coefficients A have been computed equation 3 6 is the standard linear equation set 3 7 which can be readily solved alg 3 8 Actually the size of the matric
101. elling for errors It may be necessary to use lower relaxation factor to promote convergence If the solution terminates on an execution error or will not converge the user will be required to debug the model The efficiency of the user s debugging methods will improve only with experience To assist in debugging the following potpourri of examples is included a If the program has terminated before completing one iteration it is likely that insufficient input data has been given or that the array sizing doesn t match the arrays referenced One can identify the line in which the error occurred and generally find the error using an OPT 0 on the FTN card b 4 the program fails after the eighth iteration examine the RECIR subroutine because it is called after the eighth iteration If the program terminates with an error message ARGUMENT LESS THAN ZERO this is most likely generated by quality values greater than l arising from a very high pressure gradient the user should refer to DENS to see how pressure affects quality The high pressure gradient generally occurs when a gross inconsistency in the treatment of flow obstacles occurs between various stages of the iteration Large pressure corrections are required to procedure maintain continuity The stage within the iteration that contains the inconsistency can be determined by examining the mass flows printout Large swings in
102. em 80 Steam Generator EPRI Report NP 1546 project 130 1 June 1980 W W R Inch Thermal Hydraulic Analysis of the Combustion Engineering Series 67 Steam Generator EPRI Report NP 1678 project S 130 1 Jan 1981 M B Carver Thermal Hydraulic Analysis of the stinghouse 5i Steam Generator EPRI report in press proje 5 130 1 March 1981 158 81 S V Patankar and Spalding A Calculation Procedure for Heat Mass and Momentum Transfer in Three Dimensional Parabolic Flows Int J Heat Transfer 15 p 1787 1972 9 Harlow and J E Welch Numerical Calculation of 1 Dependent Viscous Incompressible Flow Physics Fluids 8 p 2182 1965 10 Carver and H W Hinds The Method of Lines and Advective Equation Simulation 31 p 59 1978 11 Carver Pseudo Characteristic Method of Lines Solution of the Conservation Equations J Comp Physics 35 1 p 57 1980 Acknowledgements The authors wish to acknowledge the early work by R H Shill on steam generator codes much of which laid the foundations of the current THIRST code N M Sandler has been of invaluable assistance in the computer programming among other contributions he designed the plotting and read in sections D G Stewart and C Taylor have also contributed towards restructuring and rationalizing code content The monumental task of deciphering typing and revising th s manuscript might have fou
103. es and other solid obstacles are calculated using standard friction factor correlations Similarly primary to secondary side heat transfer rates are calculated using empirical heat transfer correlations 2 2 6 Reductions of flow due to the presence of tubes and other obstacles are accounted for by defining a volume based porosity 7 The primary temperature distribution is calculated from the enthalpy distribution by using a polynomial curve fit see Chapter 7 8 Secondary subcooled values of temperature viscosity etc are calculated by using polynomial curve fits of each parameter expressed as a function of the secondary enthalpy see Chapter 7 Boundary Conditions Boundary and start up conditions such as primary flow and temperature secondary feedwater flow and temperature downcomer water level etc are described in detail in Chapter 4 Overview of the Solution Sequence The numerical solution sequence apart from some variations discussed later follows the techniques outlined by Patankar and Spalding in reference 8 A fair understanding of the mechanics of the technique is required for advanced use of the THIRST code and Appendix A contains details of the overall formulation At this point however we present a brief exposition of the philosophy of the method including only a minimum of mathematics THIRST solves the five secondary side transport equations 2 1 in three dimensions t
104. es contain the baffle plate Control volumes which are partially exposed to the baffle partly filled have a weighed impedance Used as above OUTER BAFFLE ap 12 Es LIP 9 VARIABLE DIA DIAIN Shell inner EDFEED BUBBLE EDFEED liam Outer bubble EOFEED 0 diam SHELL surovo irc ads EDNORM suROUD EDNORM 77 SHEL Shroud outer diam D no EDSHRDX EDSHRDX D Outer diameter of the tubes Inner diameter of the tubes m Hydraulic equivalent diameter n the downcomer annulus at the feed water bubble m Hydraulic equivalent diameter for the normal downcomer anaulus below the con cal section m 30 Hydraulic equivalent diameter for the downcomer annulus above the conical expansion zone applies at 1 planes greater than ISHRD see data no 14 m Upper shell inner diam USHELL Pusumoup Upper shroud outer diam Pysuroup 31 Total heat transfer area 32 Distance between the outermost tube SEELE OGAP and the shroud inner surface SHROUD OUTER TUBE LIMIT 6 feedwater bubble stance between tubes PITCH Porosity in the downcomer annulus above the expansion region Inner radius of the shroud area is en
105. es prohibits direct solution 80 iterative methods are used and equation 3 8 is solved by an inner iteration The Inner Iteration The matrices of equation 3 7 are too large to permit direct Solution of the equation set by means of 3 8 even when sparse matrix techniques are considered so an iterative technique 1s used It is well known that the solution of equation sets in which the matrix A is tridiagonal can be performed extremely quickly as the algorithm reduces to recursive form 27 Equation 3 7 can be converted to tridiagonal form by including for example only the coefficients along the r direction on the left hand side 3 9 Similar expressions can be written for the 6 and 2 directions 9 Sp 3 10 j 5 Anon A191 5 9 Sp 3 11 j n s e w A A one dimensional problem can be solved directly by 3 9 two dimensional problem is solved by an alternating direction iteration ADI method This involves solving 3 9 and 3 10 three dimensional solution requires the solution of 3 11 in addition This For example 3 9 and 3 10 could alternately until the solutions converge creates several possibilities be solved for a number of iterations for each time 3 11 is solved The most suitable strategy depends on the nature of the flow problem The THIRST code has a number of different strategies designed to promote convergence in three di
106. es the friction factor as function of Reynolds number This is done as required from SOURCU SOURCV and SOURCW The pressure gradient is related to the source term by AP A TABLE 7 2 Single Phase Pressure Drop Correlations Cont d CORRELATION 64 8 lt 2000 0 25 0 316 gt 2000 e e _ 24 Pg tube support loss factor based approach area COMMENTS Downcomer Annulus Pressure Drop calculated in the function subprogram FRIC ENTRY FRIC13 calculates the friction factor as a function of Reynolds number This is done as required from SOURCU and SOURCW Tube Support or Broach Plate Pressure Drop the loss factor is stored as AKBR in code based on the contraction into the support plate and the expansion out of the plate dt is based on the approach area before the contraction TABLE 7 2 Single Phase Pressure Drop Correlations Cont d CORRELATION Baffle Pressure Drop is the shock loss factor based on a contraction into the baffle and expansion out of the baffle Tt is based on the approach area 2115 the friction factor which varies with Reynolds baffle loss factor based on approach The constant portion is stored in FLD see data sheets for discussion of FLD calculation number area approach area local area baffle thickness diametral clearance 0 25 Ge 0 316 local friction i factor
107. espectively The energy equation solution CALCH also uses the same sequence as Table A 2 152 TABLE A 1 THIRST LOGIC STRUCTURE EXECUTIVE ROUTINE ITERATION SEQUENCE Routine Called Purpose Equations READIN Read all data Initialize all array pointers Compute all geometry and initial correlations RECIRC Compute recirculation Compute U vector from axial 2 3 momentum equation Compute V vector from radial 2 4 momentum equation CALCV Compute W vector from azimuthal 2 5 equation CALCW Force exit axial velocities to be positive Compute pressure and velocitv 2 2 corrections from contina ty equat on Compute enthalpy distribution from 2 6 energy equation CALCH Compute dens ties from equation of state Compute axial mass flows MASSFLO BOUND OUTPUT N FPRINT WSTART Write tape for Restart Impose exit plane boundary values Output summary Repeat unless time or no of iterations is about to expire Final output of 153 TABLE A 2 GENERAL SOLUTION OF TRANSPORT EQUATIONS Routine or Purpose Compute K 1 boundary flux Start next K plane Call appropriate SOURCE routine to evaluate resistances and assemble Sy SOURCE and Sp terms Compute all flux terms to complete definition of Equation 3 6 Incorporate under relaxation Set up the system Ap B using tridiagonal in x and solve using forward and bac
108. feedwater through a distribution ring located at the top of the downcomer annulus below the liquid free surface The colder relatively dense feedwater mixes with the saturation liquid coming from the separators and flows down the annulus to the shroud windows The average density in the downcomer is increased thus the recirculation ratio increases The log mean temperature difference LMTD cf the units is reduced however and thus we would expect a drop in overall heat transfer without the preheater This section considers a design which does not contain a preheater but introduces the feedwater at the top of the All dimensions remain the same as the original This unit downcomer values All operating conditions remain the same may not be well designed since the basic layout normally would be altered when feedwater is introduced at the top However it will serve to illustrate the extent of code modifications Altering the code to handle new geometries requires both data and code logic changes To simplify the logic changes we will locate the last I plane just below the feedwater distribution 140 ring The downcomer flow rate is thus inereased by the amount of the feedwater flow The downcomer enthalpy is also reduced because the feedwater is subcooled Both of these parameters serve as boundary conditions to the model To illustrate the changes required we must look at how the downcomer flow is deterrined In the co
109. ge of 245 C to saturation The heat transfer coefficient parameter RCONVA is defined in Section 7 3 All subcooled properties ENTRY PROP4 to PROP8 inclusive are cal culated as polynomial functions of enthalpy and are valid over the temperature range of 150 C to satura tion OZI TABLE 7 1 Fluid Properties and Parameters Cont d FORTRAN ACCURACY OF PROPERTY OR PARAMETER NAME IN POLYNOMIAL PROPRTY EXPRESSION 2 Temperature C PROP ENTRY PROP5 Liquid Specific Heat J kg 1 96 ENTRY PROP6 Liquid Prandtl Number PROP ENTRY PROP7 Liquid Density kg m IPROP ENTRY PROP8 Derivative of Saturation Pressure PROP ENTRY with Respect to Temperature for PROPS Chen Correlation Pa C l Section 7 3 7 2 422 Range of Application The thermal hydraulic data in the THIRST code is limited to the following range of operating conditions Primary heavy water 7 MPa to 11 MPa inlet pressure subcooled to two phase inlet the overall temperature drop should be such that the outlet temperature is not less than 245 C Secondary light water 4 MPa to 6 MPa mean pressure feedwater temperature range of 150 C to saturation If it becomes necessary to investigate different fluids or conditions outside of the above ranges the user must redefine the appropriate property polynomial functions in the PROPRTY subprogram Empirical Correlations for Flow and He
110. hich is described in Section 5 2 The input cards are assembled from the variable names and values already detailed in the last two columns of the charts in Section 4 3 immediately preceding The cards must adhere to the following rules 1 The first card must contain the title 1 to 40 columns and the RESTART value word RESTART in columns 50 to 59 and value in 60 to 69 If the RESTART name and value are not included READIN assumes a RESTART value of 1 2 All succeeding cards are read with the following format statement FORMAT A9 6 49 1X 49 65 The input cards for data arrays or single variables are 50 60 70 1 0 20 0068 6 8 4 1 3456789 180040 7 NN is the number of entries in the array called ARRYN It is only required for arrays IMASSF IPLOTI and IPLOTK 3 The second card must contain the number of grid points NI NJ NK selected for each direction to provide READIN with the counters for checking array data 4 From this point onwards the data may appear in any order since the variable name is always included with the data READIN treats each variable name and the corresponding data as a variable set 5 is possible that after a data deck is prepared some tem porary changes are found necessary this case a data item may be changed in situ in the deck or a single card with the changed variable may be inserted immed ately after
111. ic Pa ameters THIRST OUTPUT Summary of Grid Locations THIRST OUTPUT Iteration 1 THIRST OUTPUT Iteration 2 THIRST OUTPUT Iteration Iteration o a 5 Iteration Summaries Summaries Summaries Iteration 58 THIRST OUTPUT Iteration Summaries Iteration 59 THIRST OUTPUT Iteration Summaries Iteration 60 5 Sa da May THIRST OUTPUT Detailed Output Velocity Field s 56 THIRST OUTPUT Composite Plots Quality Distribution sa es THIRST OUTPUT Composite Plots Velocity Distribution THIRST OUTPUT Composite Plots 8 o y o Mass Flux Distribution EE 5 5 THIRST OUTPUT Radial 1 Plots Quality Distribution sas a o THIRST OUTPUT Radial Plane Plots Velocity Distribution THIRST OUTPUT Radial Plane Plots Mass Flux Distribution VAS Gre xu ies Page POR 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 Figure Figure Figure Figure Figure Figure LIST OF FIGURES continued THIRST OUTPUT MODIFIED DESIGN Code Changes THIRST OUTPUT MODIFIED DESIGN Data Summary s s 9 w THIRST OUTPUT MODIFIED DESIG Final Iteration Results Graphical Output THIRST OUTPUT MODIFIED DESIGN Final Iteration Results Graphical Output Quality Distributio
112. istribution 150 APPENDIX A LOGIC STRUCTURE OF THE THIRST CODE This appendix discusses the logic structure of the code including the outer and inner iteration sequences the pressure correction and the function of the subroutines The Outer Iteration Sequence The executive subroutine orchestrates the outer iteration sequence computing in turn each velocity component from the associated momentum equation to obtain velocity and pressure corrections as described in Section 3 computing the enthalpy from the energy equation and finally obtaining new densities from the equation of state The Inner Iteration Sequence used in CALCU CALCV CALCW and CALCH Because it is possible to set up all the conservation equations in general transport form the solution of each equation follows the same general sequence The problem is to solve the matrix equation 3 6 241 SU i Cs etc As tA LS This is accomplished by setting up the alternating direction tridiagonal solution in a plane as described in Section 3 4 The general sequence used in CALCU and CALCV is given in Table A 2 However as W 18 8 velocity the K planes must be incorporated more implicitly CALCW sequence in CALCW is identical to Table A 2 except that it is done by I plenes and uses routines SOLVE3 and SOLVE4 which set up the tridiagonal systems in KJ and I r
113. kward sweeps through r Repeat tridiagonal in r and sweep in x Assemble coefficients in the plane preparatory for a K block solution Set up the system B using tri diagonal in 0 Perform one solution using coefficients assembled above thus correcting the above results for K variation SOLVE 2 154 Pressure and Velocity Correction Routine CALCP The pressure and velocity correction obtains pressure corrections by embedding the velocity corrections in the continuity equation as described in Chapter 2 The sequence is shown in Table A 3 First the continuity equation is solved for the embedded pressure corrections as in Section 2 4 Then each velocity is corrected following equation 2 21 and finally the new values of pressure are computed As mentioned in Section 3 unlike the other variables pressure is under relaxed if necessary after the 1 equation solution rather than before The solution of the embedded continuity equation in routine CALCPK is performed exactly the sequence of Table 2 except of course there are no source terms to evaluate in the continuity equation In standard applications of the Spalding and Patankar technique this pressure correction would be performed several times and then the sequence would pass on to the energy equation as shown in Table 1 However CRNL experience has shown that convergence can be promoted more rapidly if further
114. lated from the hot side The resulting mixture undergoes partial evaporation and rises as a two phase mixture through the remaining bundle section into the riser and up into the separator bank Here the two phases are separated The steam leaves the vessel to enter the turbines while the remaining saturated liquid flows through the annular downcomer to the bottom of the vessel Here it re enters the heat transfer zone through windows around the shroud circumference The downcomer flow entering through the windows on the hot side partially penetrates the tube bundle before turning axially to flow parallel to the tubes the cold side the downcomer flow must pass under the preheater to the hot side before it can turn axially Thus the downcomer flow converges the center of the hot side tube bundle As this fluid rises through the hot leg it absorbs heat from the tube side fluid Quality develops very rapidly because the downcomer flow is very close to saturation Above the top of the preheater this mixture mixes with the fluid from the preheater The tubes are supported by broached plates located along straight portions at the U tubes Further lattice supports are located in the U bend The baffles in the preheater are drilled plates In this design no feedwater leakage through the thermal plate floor of the preheater or the partition plate is allowed Ail the feedwater must exit at the top of the preheater The primar
115. located to satisfy the axial flow obstacles such as baffles tube support plates window openings etc the user should then examine areas which are critical to the aralysis and ensure that a sufficient number of Brid planes are located in these areas For instance the region just above the tubesheet at the shroud window is particularly important The I 2 plane is located just above the tubesheet The I 3 to 1 5 are added to this region to provide more detail I 22 plane is added above the preheater to handle the migration of hot side flow to the cold side To enable the tracing routine used to calculate the heat transfer in the U bend an axial plane must be located at the start of the U bend curvature At least 3 additional axial planes should be located in the U bend to ensure the accuracy of the routine which calculates the pressure drop and heat transfer in the U bend 1 the last plane should be located very close to the second last plane so that the axial boundary values which are based on the last internal values can be calculated Radial Division J Planes In our example we have used 36 axial planes We have now 4900 36 136 more nodes available to share between the radial and circumferential directions Figure 4 4 shows a horizontal cross sectional cut of our design Note that only one half of the steam generator is modelled as the design is symmetric about a line d viding the hot and cold sides The bundle b
116. mensions These are discussed in Appendix Stability of the Solution Scheme The outer iteration scheme discussed in Chapter 2 normally proceeds to convergence in a stable manner and converges rapidly providing each inner iteration is stable To promote stability of the iterations three principal devices are incorporated in THIRST The first that of under relaxation is common to most iteration schemes The second upwind weighted differencing is frequently used to stabilize both steady state and transient thermal hydraulic calculations 10 The third concerns the formulation of the source terms to ensure stability Under Relaxation Because the solution is obtained by iteration there is a strong likelihood that variable values may fluctuate unduly during the initial stages It is common practice to stabilize these fluctuations using under relaxation Thus if is calculated from 3 9 to 3 11 using previous values pu it is then replaced by 1 0 57 gt 3 12 Relax Calc old Relaxation factors for each equation solution are supplied with the THIRST code but may be changed by data input if necessary In practice it is possible to impose under relaxation before attempting the linear equation solution instead of after its completion This is preferable as it minimizes the chances that the linear equation solution itself may generate unlikely values Recall that the equation to be solved is
117. mpirical correlations to approximate a number of phenomena which cannot be prescribed analytically These empiricisms include correlations for single and two phase heat transfer and pressure drop in rod bundle arrays and for void fraction All correlations used in the THIRST code are summarized in Chapter 7 IMPLEMENTATION FUNDAMENTALS The previous chapter has discussed the governing equations developed a suitable solution philosophy and mentioned the thermal hydraulic data required to complete the specificat on of the model This chapter is concerned with the manner in which these general pr nciples are implemented in the THIRST code This involves the establishment of the computational grid the conversion of the partial differential equations to discrete node equations by means of control volume integration and the technique used to perform the inner solution of individual equations The control volume integration and equation solution are of course built into THIRST but in order to choose an effective grid layout the user needs some feeling of these procedures The Coordinate Grid A three dimensional cylindrical coordinate system is used for obvious reasons The entire flow domain between the tubesheet and the separator bank is subdivided by planes of constant r 2 and 0 The grid arrangement is chosen to suit the geometry and expected flow patterns of the steam generator Thus it is usually not uniform but is
118. mplemented in schemes similar to that used in THIRST 15 in the THIRST code the simple formulation is retained however The large flow resistances and heat sourres due to the closely packed tube bundles in the steam generators dominate the computation to such an extent that the differences which would be caused by higher order methods are believed to be minor 582255 3 6 Notation used THIRST Finally we have up to here been using single subscripts n s etc for simplicity The code however is written in cylindrical coordinates and uses terms such as AXM to denote Ay On this basis equation 3 6 becomes Ase 194 sy 3 19 where _ Ag Ag _ DIVG SP The upwind formulation can be implemented to consider flow direction automatically in the following manner E cim 0 5 2 3 Ci pav 3 face area _ 3 20 r 2 2 2 r r 6 90 29 2 2 3 Coy etc Cra mass flow through control volume face ri depending on the transport parameter v are either defined at that face or interpolated to that face 33 a a t a DENG r 9 8 net accumulation of mass in the control volume n The table below defines and for each i i Ay Ont r AL ae Ag
119. n THIRST OUTPUT MODIFIED DESIGN Final Iteration Results Graphical Output Velocity Distribution c Er THIRST OUTPUT MODIFIED DESIGN Final Iteration Results Graphical Output Mass Flux Distribution 144 145 146 147 148 149 INTRODUCTION The THIRST computer code is the latest in a series of three dimensional steady stato computer codes developed at CRNL for the detailed analysis of steam generator thermal hydraulics The original code designated BOSS arose from the DRIP program of Spalding and Patankar 1 and was adapted for application to CANDU type steam generators 2 Although the equations to be solved remain the same extensive changes have been made to the program structure the numerical computation Sequence the empirical relationships involved the treatment of the U bend and the numer cal and graphical presentation of results The code has therefore been renamed THIRST In conjunction with these developments the program has been used to successfully analyse the thermal hydraulic performance of a number of different steam generator designs from CANDU to American PWR nuclear plants The program has also been used for extensive design parameter surveys Some results of these analyses have been released in publications 3 7 Steam generator designs already analysed are summarized in Table 1 1 As the structure of the THIRST code is now well established an
120. ndered but for the efficient and cheerful efforts of Mrs M L Schwantz ISSN 0067 0367 ISSN 0067 0367 To identify individual documents in the series Pour identifier les rapports individuels faisant we have assigned an AECL number to each partie de cette s rie nous avons assign un num ro AECL chacun Please refer to the AECL number when re Veuillez faire mention du num ro AECL si questing additional copies of this document vous demandez d autres exemplaires de ce rapport from au Scientific Document Distribution Office Service de Distribution des Documents Officiels Atomic Energy of Canada Limited UEnergie Atomique du Canada Limit e Chalk River Ontario Canada Chalk River Ontario Canada 1 0 KOJ 140 Price 9 00 per copy Prix 9 00 par exemplaire 1167 81
121. nditions and input output parameter selection READIN contains lists of ali the variables required by the code As the data cards are read READIN searches through the list to match input variable names vith the ones on its list If the i subroutine can make the match it stores che data in the variable and removes the variable name from the list If READIN cannot match an input variable to one on its list it issues the following message X CANNOT MATCH X VARNAME This message contains the input var able name and its value so the user can trace the nature of the error This error could result from a misspelling of the variable name from reading the same variable twice or using improper data This error does not result in a termination of the run If a variable appears twice READIN stores the first value and disregards the second If a variable is mispelled READIN ignores the variable and its value and thus the intended variable name will not be removed from the list When the end of the input deck END OF FILE is encountered READIN checks that all the variables on its list have been initialized Some variables may not be stroked off because they are either mispelled or simply left out If a variable name remains but has not been initialized READIN issues one of the following messages THE FOLLOWING VARIABLE S HAVE NO INPUT DATA VARNAME or THE FOLLOWING ARRAY S HAVE NO INPUT DATA VARNAM
122. nued GEOMETRICAL RESTRICTIONS AND POSSIBLE VARIATIONS 8 1 Tube Bundles 8 2 Preheater 8 3 Tube Supports 8 4 Downcomer Windows 8 5 Separators ADAPTATION OF THIRST TO A NEW DESIGN APPENDIX A Logic Structure of the THIRST Code APPENDIX B References and Acknowledgements 137 137 137 138 138 138 139 150 157 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure iv LIST OF FIGURES Cutaway View of a Steam Generator Sirplified Model of the Steam Generator Grid Layout showing Scalar and Vector Locations Control Volumes for Scalar Quantities Control Volumes for Radial Velocity Vectors 4 Control Volumes for Circumferential Velocity Vectors E xe cw m Grid Layout at a Baffle Plate 3 4 34 Sod gis Grid Layout at a Shroud Window Axial Grid Layout Radial and Circumferential Grid Execution Deck dai THIRST OUTPUT Summary of Input Data THIRST OUTPUT Interpreted Data Summary of Operating Conditions THIRST OUTPUT Interpreted Data Summary of Output Parameters oe E THIRST OUTPUT Interpreted Data Summary of Geometr
123. o compute distributions of the dependent variables w h and The mixture density is calcula ted from the equation of state p h P The variables are 13 stored three dimensional arrays of up to 5000 prid points This generates about 30 000 simultaneous non linear differential equations Obviously this requires some form of technique which permits the solution to concentrate on portions of the equation set rather than attempting a simultaneous solution This is accomplished by considering each of the transport equations separately and then iterating through the full set of equations The solution of any given transport equation itself involves developing a finite difference statement of the equation and solving it in an inner iteration but we will delay considering this until later Suffice it to say that the transport equations can be reduced to a set of linear matrix equations and written as follows Continuity BEV 0 2 11 Momentum 09 Ey 0 2 12 DyV E FyP 0 2 13 EQ 0 2 14 Energy Hh 0 2 15 State f P h 2 16 The coefficient matrices A to G are functions involving first estimates of the dependent variables v W p h We wish to solve equations 2 11 to 2 16 in a sequence that will eventually lead to all six equations being satisfied This is accomplished as follows i solve equation 2 12 to get new estim
124. oJ GJ DRAW WR HO 2 HOD UNICO UMS TIU CU IN OQ D AND FUN ID AMADINAVONNA 0 ANTE DOT DAFN OAM QDO Gui UMW OAT AIS PTAA ANNOM PU 164 hehe eo eT eee ee ee eee er BW IW UUW tt DP FAL UMP MOA fot ODL Omne ip OTOL DN UOI QNO v AMI QUUM RIN AVE etes ROMS ODOT DE P ee IFO 4 4 449 6499 66404460 640246 LAAN 3 44909343 did dg d 33444408 TU Ocoococoooooooceoocogooocooooc uen OGoOocOocoooononucoooooeoocoogoogoooogcooo 714 4454244 4 24522 24 24 11 t Gu UG UE Lu d MONN GC C2 4 fu o PON IDOLON A AEO HS 0000 OPW ID IO P et PO HOA A OUS OM 3 UU OQ qu vf TRUS Seen I AAO Y 3222742 I PIA DORON ANP PWN O v EPOD 4E D Pe uu viride einn IUINIU PO I 109 10 21303 55559 55 20 190 9
125. oundaries and baffle plate edges are marked as dashed lines The shroud and shell locations are shown as sol d lines RADIAL AND CIRCUMFERENTIAL GRID SHELL DOWNCOMER ANNULUS FEEDWATER K 10 Figure 4 4 Radial and Circumferential Grid 4 2 8 Figure 4 4 also shows the radial grid layout 1 corresponds to the center point The second radial position J 2 is located very close to the 1 point because it is the First active point in the radial grid pattern The J 9 and J 10 points are located so as to center the shroud inner radius as discussed previously The J 3 to J 8 points are positioned at equal intervals as specific locations are not dictated by special geometrical features Circumferential Division K Planes We have now used 36 x 10 360 grid planes and we have 4900 360 13 grid planes left to be allocated in the 1 ential direction To simplify the layout we will only use 12 with equal numbers on the hot and cold side The code can accept unequal numbers of grid planes on the cold and hot side if the geometry requires it K 6 and K 7 planes are located such that they straddle the partition plate The K 2 and K 11 planes the first and last internal planes are located fairly close to the boundary points as they are the first active points inside the boundary The remaining points are spaced equally however this is not a requirement and spacing may be adjusted
126. ove the preheater in the downcomer etc Three sets of variables AXIAL VELOCITY CROSS FLOW VELOCITIES and THERMAL VALUE are printed The location of each variable is described including its 1 J K coordi nates If the user wishes to change the locations to be printed the OUTPUT subroutine must be altered The overall values 4 and local values are printed out in OUTPUT OUTPUT can be called at any point in the execution if the user desires to At present it is called at the end each iteration step Detailed Array Printout Figure 6 5 The last type of printed output again under user control is the complete printout of selected variables at every active node in the medei The format for the printout is XXXXXXXXXX VARIABLE NAME 1 XXXXXXXXXX u il i w c M 1 1 l 1 1 1 1 1 1 1 1 i 1 1 1 i 1 1 1 il M i lt 0 XXXXXXXXXX VARIABLE 2 XXXXXXXXXX etc This printout can be very long depending on how many variables are specified for printout Figure 6 5 shows the first page of a detailed array printout of axial velocity obtained by setting IPRINT I to 1 Each selected variable takes a similar format and each generates five pages of output for 12 so the feature should be used with caution Variables to be printed may be selected by the input variable IPRINT IPRINT NV is entered non zero the
127. relations and Resistances Items 42 60 GROUP 5 Operating Conditions Items 61 69 GROUP 6 Utility Features Items 70 85 Items within each group are arranged alphabetically for ready reference VARIABLE ITEMS 1 7 PRELIMINARY DATA Controls the use of the restart RESTART 1 0 new run no RESTART tape used option see Section 5 1 as input RESTART 2 0 continue executing from point reached in a previous run RESTART 3 0 attach the data stored on tape from a previous and print and or plot the data RESTART 1 or 2 or 3 proceed as above but write the final results on a restart tape Number of axial planes Must be an integer number Number of radial planes Must be an integer number Must be an integer number Number of circumferential planes rw Distance from the secondary side of the tube sheet surface to each axial plane in meters Distance from the center point to each radial plane in metres The angle in degrees from a line passing through the center of the hot side to each circumferential plane 77 DATA DATA ITEMS VARIABLE 8 21 ARE GEOMETRIC DATA ENTERED ACCORDING TO GRID LOCATION USING GRID INDICES Location of all baffles tube support plates and thermal plates on the cold side See layout This array is set up to indicate which axial velocities are passing through a plate resistance Each axial plane
128. rmat The informaticn provided includes RECIR Recirculation ratio used for this iteration PRESS DROP is the pressure drop between the average in Pa pressure at the last I plane L plane inside the shroud and the average pressure at the last I plane L plane outside the shroud in the downcomer PRIM H T is the net amount of heat given up by the in MW primary fluid e SEC H T is the amount of heat picked up by the in MW secondary This includes the heat required to raise the feedwater and reheater drain flows to saturation plus the heat absorbed in evaporating the secondary liquid NOTE PRIM should equal SEC when convergence has been achieved AVG OUTLET QUAL average outlet quality SUMSOURCE is the summation of the absolute value of the mass imbalance for each control volume normalized by dividing by the total flow This indicator should approach zero with gence MAXSOURCE 2 7 11 is the largest mass imbalance normalized by dividing by the total flow in the modelled region The location is given in the brackets as 1 2 J 7 Kall the location remains fixed and the imbalance is significant the use should examine the region for a possible error in that area Summary of Local Values The last section of the iteration by iteration printout summarizes local values at strategic locations in the model The locations are fixed in the code at such points as window inlets ab
129. s the first plate on the hot side has smaller broached holes near the shroud and larger broached holes near the center to encourage flow penetration This factor is for the area change in the central larger holes k loss factor for the smaller broached holes in a differentially broached plate Same as data Shock loss for the outer small broached holes See data no 18 for the radial position where the hole size changes For some designs the tubes are not rolled into the thermal plate and leakage through the plate may occur The pressure loss relationship is AKTP 5 e a shock loss and a fric tion loss This data no deals with the shock loss Again it is based on the approach area EG DATA VARIABLE ea indow area 251 Shock loss factor for the shroud j This pressure loss relationship is based on a AKWINDC window on the cold side 90 flow direction change and an expansion from the downcomer annulus into the shroud window Both and are based on A 2 AKWINDC stk 90 Aan 51 Shock loss k factor for the shroud Because the shroud window height may differ AKWINDH window on the hot side between the hot and cold side a second loss factor may be required Area ratio multiplier to determine Approach The local Reynolds
130. s 1 this card will have no useful purpose and should be omitted THIRST2 PL 30000 Execute the job and set the print ng limit at 30000 pr nt lines IFE RI NE 0 JUMP If a RESTART tape has been written either through a time limit or a negative value of RESTART then R1 is set to one If the pro gram has not written data RESTART tape then Ri O and the execu tion jumps to ENDIF JUMP Thus this card controls the se guence to CATALOG only when the data RESTART tape exists Explanation CATALOG TAPE60 THIRSTDATA Catalog the RESTART ID THIRST tape ENDIF JUMP Point to which the IFE card directs control End of record card To enable the computer to allocate storage the size of the grid layout must be specified in the EXEC routine The following correction may be used to change this allocation and is included for example purposes Card No Explanation 16 D EXEC 4 Delete the fourth card in EXEC 17 COMMON F 4320 13 Reserve 13 arrays 111 variables plus 2 ing spaces Each array contains NI NJ 36 10 12 4320 storage places 18 7 8 9 END RECORD 5 4 2 Input Deck Unless the changes made above incorporate new input data no form changes in the deck of Chapter 4 are required THIRST OUTPUT In this chapter we present the basic output obtained from the THIRST code Possible variations of output are also discussed Output from THIRST i
131. s in both printed and graphical form The following paragraphs refer to sample output which appears consecutively at the end of this section starting on page 99 Printei Output Features Preliminary Output After the program logo THIRST prints out the values in the input deck arrays are printed first the single integer values second and the single real values last All the error messages related to the input are printed out in this section Figure 6 1 The next section of the printed output contains a summary of all the input received by the code for this run and a summarv of the propert es which THIRST has calculated from curve fits Figures 6 2 1 to 6 2 3 contain a Operating Conditions Primary Secondary b Properties as Calculated by THIRST using Curve Fits Primary Saturation Values Secondary Saturation Vaiues Secondary Subcooled Inlet Properties c Output Selection and Control Parameters d Geometrical Parameters 6 1 2 Figure 6 3 contains a The Grid Locations for Scalar and Vector Components The Axial Positions in Metres The Radial Positions in Metres The Circumferential Positions in Degrees b Primary Fluid Flow Distribution per Typical Tube in kg s All the above output is generated in START before the iteration procedure begins The user has no control over the format without altering the logic Individual Iteration Summary Figures 6
132. ssure of the primary MPa Flow rate for the whole unit FLOWTU Used to calculate prinary properties PPRI 86 DATA DATA VARIABLE DESCRIPTION VALUE REMARKS VALUE No 65 66 67 Saturation pressure of the Used to calculate secondary proverties PSEC secondary MPa Take tne value at the normal water level Inlet quality of the primary fluid For a two phase mixture it is the actual OLTU quality For a subcooled primary fio this value is cilculaced using Enthalpy of Liguid Saturation Latent Heat Initial estimate of recirculation he recirculation ratio is not adjusted for RECIR ratio c the first 9 steps to allow the flow to settle out This value serves as initial condition pues of the reheater return Temperature of the feedwater C TINC TRH flow ITEMS 70 85 ARE UTILITY FRATURES AVAILABLE TO USER 70 horizontal lines of data which areas where planes are concentrated IIPloT are to be included on the vertical one may decide ro leave out some T lines from cut plots vertical plots so that the plotted arrows do not overlap Normally all the lines would he plotted IF 1 I plot the Line skip the line Note TIPLOT T must have NI entries 6S DATA No 1 73 VARIABLE Selection of the I position for A subroutine MASSFLO has been
133. t e e 2 4 Overview of the Solution Sequence 2 5 Thermal Hydraulic Data 2 5 1 Fluid Properties and Parameters 2 5 2 Empirical Relationships IMPLEMENTATION FUNDAMENTALS a 3 1 Coordinate Grid p e s 3 2 The Control Volumes gt 3 3 Control Volume Integral Approach 3 3 1 Integration of the Source Terms 3 3 2 Integration of the Flux Terms 3 4 Iteration 6 3 5 Stability of the Solution Scheme 3 5 1 Under Relaxation we ews 3 5 2 Upwind Biased Differencing Notation used in THIRST Formulation of the Source Terms a e ee TABLE OF CONTENTS continued 4 APPLICATION OF THIRST TO ANALYSE THE PROTOTYPE DESIGN 34 4 1 Design Specification 35 4 2 Grid Selection e 46 e Bow Xo o7 e Bo oe 35 4 2 2 Baffles d e 22 Wo 9 3 36 4 2 3 Partition Plate 4 6 4 Ww ov 36 4 2 8 Windows Se fen we INS Wl der elena s 6 4 2 5 Axial Layout 1 Plane ee Wow ret ane m e we JO 4 2 6 Radial Division J Planes Ss PRS de Ww AD 4 2 7 Circumferential Division Planes 42 4 2 8 Final Assessment 42 4 3 Preliminary Data Specification e
134. tape on completion of a run absolute aluo of n i e 1 2 3 determines the initial conditions for the run On completion tape is written through the WSTART routine ready for subsequent reading with RESTART gt 1 Saving and Accessing a RESTART Tape In WSTART a special global parameter R1 is set to 0 when the tape is written A set of statements can be included in the execution control cards to catalog the tape if this global parameter is set IFE R1 NE 0 JUMP CATALOG TAPE50 THIRSTDATA ID THIRST ENDIF JUMP If R10 control jumps past the CATALOG card to the ENDI JUMP card If WSTART has set 1 0 the CATALOG card is executed and the RESTART tape is stored under the name THIRSTDATA ID THIRST and CY n the lowest available number Thus each time a RESTART tape is made a new THIRSTDATA file is created The user will have to exercise strict management of these files to avoid confusion and the creation of unnecessary files To RESTART from this tape the user includes the card ATTACH TAPE50 THIRSTDATA ID THIRST before the execution card in the JOB control statements The information stored on this file will then be used in the RESTART routine to initialize the arrays and the variables providing RESTART is set at 2 or 3 in the input deck The READIN Feature To assist the user with the entry of data into the code subroutine called READIN has been written READIN extracts the v
135. the code Our output will appear in Chapter 6 n x re J om an ou e ar 2 Me u ee re a wa ou ere re qe nea m mw we a ro D an OLOA a o om om due et m Bo Oe T aFirasjgorae xea ul lle 3905 4 0 x 1 0 ou z NOM Je 006 gt ok ILI mx a Izo N IuUnlI rreReezOoc t OI HO0 JO0U0r Oc 54 4 gt 4 1 4044 Jm OD Em aa vac oe aa on ta ud NK faa Ada AM GENERATOR 1 laa A 3T NJ cuu pM ooo mAr 6 0 HYPOTHETICAL t00 M4 77879 1 0 1 1 1 0E1 NI i c umen A ama ANN mamya w mAmmwa c T cane met ANNN IPLOT Lal aio A m anna Ao in QA Nanum N N Two LI mao Omme WANN m NOD AMAA
136. thermal values generally indicate a problem in the heat transfer subroutine source terms especially if a new correlation has been introduced Reduce the relaxation factor for Ty to promote stability If the solution is not converging and the reason is not clear it may prove useful to call for plots for several The plocs should then be super In this succeeding iterations imposed to identify regions that are oscillating way the region s of possible modelling errors be pinpointed One could also call for FPRINT output for several succeeding iterations g h 1 j k If the FPRINT array is called and columns of zeros appear in the output or if a mode error occurs check that the common card in EXEC hich sets the size of the F array has been dimensioned correctly If the PRIM H T is different than the SEC H T the problem is most likely located in the SOURCH rcutine where the heat transfer source terms are calculated Check that the no tube regions are bandled correctly and that any new correlations are used correctly If the results seem to oscillate between two sets of values check the wedge and ring routines to ensure con sistency of treatment These routines are used on alter nate steps If the flow oscillates between the not and cold side shroud windows exam ne the treatment of flow obstacles in the downcomer If resistances appear to be incorrect one can print
137. three major segments 1 job control statements 11 update correction set 111 input data Because the function of each section in the execution deck is different they will be explained separately It is assumed that the reader has a basic understanding of the job card Sequence and the update routines available through the computing system A listing of the execution deck without explanations is Shown in Appendix C 5 4 1 Job Control Statements THIRST with any code changes in the associ ated correction set and list on disc Card No Explanation THIRST B652 EXAMPLE T500 10100 2 ATTACH OLDPL THIRSTPL ID THIRST Attach the code Stored on file name THIRST UPDATE C DISC Update the file 4 FTN I DISC B THIRMOD Compile the file THIRST from DISC Store compiled file on THIRMOD ATTACH THIRST ID THIRST Access standard THIRST code COPYL THIRST THIRMOD THIRST2 Merge modifications and standard code to crcate new program THIRST 2 ATTACH PLOTLIB Attach library plot ting package LDSET LIB PLOTLIB SUBST PLOT PLT Explanation 60 THIRSTDATA This card is required ID THIRST only when the RE START option is used ABS RESTART GT 1 The data catalogue from a previous run under file name THIRSTDATA with ID THIRST and for 1 will be attached and used to initialize the variables RESTART i
138. tirely open For the reg on around the bubble one has to alenlate a porosity hich when multiplied 2 2 SHELL 7 RBUBBLE gt 221741 Revert 7 eunoUD 06 2 2 R R SHELL SHROUD yp PSHRD SHELL SHROUD 0 Lower shroud puter rad RsuRown 0 DATA DATA VARIABLE gt DESCRIPTION VALUE REMARKS VALUE Calculated inner radius of the shell m Height of thermal plate above level tubesheet m Inner radius Inner radius The code ignores the thickness of the shroud To maintain rhe correct downcomer area the 2 2 RADIUS Gron Qr ANE inner radius of the shell has to be reduced to RSHELL TPLATE Tubesheet thickness TUBSHET 40 Height of the downcomer water above the tubesheet m 41 Height at which the two phase mixture can be assumed to be separated relative to tubesheet XDOWN This is used to calculate the gravity heaa inside the shroud Generally one coula take the elevation halfway along the separator 26 mew RR ITEMS 42 60 loss factor for the centerline between the hot and cold side AKDIV I Parameter for selecting two phase Imultipliers Parameter for selecting void fraction correlation See layout CORRELATIONS AND RESISTANCES This array is used to indicate the location of the partition plate AKDIV I 1 0 15 the
139. ur vertical cuts corresponding to circumferential planes The number and indices of K planes to be plotted are specified by the parameter IPLOTK There is no limit on the number of K planes to be selected Examples of this composite for quality velocity and mass flux profiles are given in Figure 6 7 In some grid layouts axial planes are grouped together to provide greater detail Unfortunately when velocity or mass flux vectors are plotted they tend to overlap To ensure clarity of the plots an additional plot parameter called I PLOT has been introduced If IIPLOT I 1 the values on that I plane are included on the vertical cut plots If IIPLOT I 0 the corresponding I plane values are left off the plot The user has control over the plotting frame size the first composite the width is specified by 1 and XL2 If horizontal plots are made on the left and on the right of the vertical cut the routine uses the wider plotting frame Specified in 1 If other horizontal plots appear only on the right the routine uses the narrow plot XL2 The height for all plots is YL The length to width ratios of the plots may not be in proportion to the actual design as the width may be increased to add clarity Scaling factors are determined by the code The plotting routines can be called at any point in the code by the statement CALL CONTOUR The parameter PLOTO has been introduced to control the calling of the plot ro
140. ure 1 1 has obvious geometric restrictions Foremost is the restriction to cylindrical coordinate geometry However a number of minor geometrical changes can be made quite readily enabling the code to accept a wider variety of designs Tube Bundles The tube bundle is U shaped with a spherical U bend Porosities and control volume centroids for the U bend region are calculated in the subroutine VOLL If the design of interest has non spherical U bend geometry i e square elliptic major modifications of NEW as well as some changes in SOURCU to SOURCH will be necessary The user is advised to consult the authors before such modifications are undertaken The user can specify any tube bundle outer diameter and tube free lane width There are no provisions in the code to handle cylindrical tube free areas in the centre region however Porosities and single phase fluid flow correlations are based on an equilateral triangle pitch arrangement The user should modify the correlations FRIC and HTF if other arrangements are of interest If the arrangement is square ATR in the subroutine START must be redefined as ATR 0 5 PITCH 2 Preheater The preheater geometry is defined by specifying the following thermal plate elevation top of preheater elevation feedwater inlet opening and baffle plate cuts The feedwater inlet 138 opening extend over the full 90 circumferential arc on the cold sid
141. utines If FLOTO 0 the plot routine is never called This 5e used where the user wants only a printout If PLOTO 1 the plot routine is called at the end of the program If PLOTO 2 the plot routine is called at the end of each iteration This leads to a very long plot life PLOTO is set in the input deck PLOTO and PRINTO can be reset in the program to initiate the plotting and printing function Interpretation of the Output Having discussed the layout of printed output we now turn again to the printed output Figures 6 1 to 6 5 to examine its content and its significance The first page of printout Figure 6 1 contains a summary of all the data introduced through the input deck No error messages of consequence were issued and a comparison with the data sheets indicetes that the data has been introduced correctly The second third and fourth pages Figure 6 2 contain input values and calculations made with the input The operating conditions should be checked against the information sheets Property values generated by the code should be checked against values in standard tables Correlation data should be verified The input output parameters are simply informative Finally the geometric data should be verified against drawings or data sheets The modelled heat transfer area should be examined to ensure that it is not radically different than the prescribed value Although the correction factor
142. y fluid heavy water enters the tube bundle from the reactor circuit as a low quality two phase mixture The primary mass flow distribution is determined by the code although the quality distribution is assumed to be uniform at entry The secondary fluid is light water It enters the preheater at subcooled conditions is assumed to enter the preheater at a uniform velocity The driving force for natural circulation is provided by the heighc of water in the downcomer annulus The THIRST Standard Code and its Intended Application The THIRST computer code as evidenced by Table 1 1 can be readilv adapted to a number of steam generator designs the numerical method is extremely robust The standard THIRST package however pertains to a hypothetical steam generator The program models a region extending from the face of the tubesheet n Figure 1 2 up to the separator deck including the downcomer annulus Symmetry permits analysis of only one half of the vessel The Use of This Manual The THIRST package is designed to make modelling of steam generator thermal hydraulics as straightforward as possible Thus a seasoned user of the code will normally consuit only chapters 4 and 5 of this report which outline in detail the procedures required to layout the computation grid and prepare the input data However to properly accomnlish these tasks the user must first understand the fundamental principles of the relev

Download Pdf Manuals

Related Search

Related Contents

Philips FB541/07/0U  Nel-Tech Labs Messager USB installation user manual  Final Report - BME - University of Connecticut  Alois - Vermeiren  Philips 1.5 m Stereo Audio Composite A/V cable  Proview VX755 Computer Monitor User Manual  English  IE-703114-MC-EM1 UM - Renesas Electronics  

Copyright © All rights reserved. Failed to retrieve file