BEGIN NEW DATA CASE C 1st of 18 subcases illustrates Robert Meredith's Type-68 TACS device as C first described in the April, 1998, newsletter. This is AUT5 data case. C test120a.dat (1 HERTZ AT 120% VOLTAGE) C without resistances C ENERGIZED AT ZERO TIME VIA 0.100 OHM RESISTORS -steady state C FREE FORMAT TIME STEP- 1000/CYCLE FOR 1.5 SECOND+ C .000016666666, 1.500 , 60.000 , 0.,,,,,,, { Meredith's original params .000066666667, .100 , 60.000 , 0.,,,,,,, { Much reduced burden on computer C PRN> PLT> NET> SS> MAX> SAVE> NENERG> > 1 5 0 0 1 -1 0 5 5 20 20 100 100 500 500 C $NEW EPSILN, 1.E-10, { Add 24 July 2008 to compensate for protection in REDUCT $NEW EPSILN, 1.E-11, { Add 31 July 2008 to allow solution with MATFUL = 1, too. C Note about preceding line. The default 1.E-8 is too big, and so is 1.E-9. C This change to DCNEW-25 is made at the same time as that to the 2nd subcase C of DC-31. See that file for what happens without a small enough EPSILN. TACS HYBRID C $INCLUDE aut5-tac.inc C $INCLUDE aut5-tck.inc { Uses new Type-68 TACS device, so much smaller file C TACS STEADY-STATE CORE NFLUX INITIALIZATION FOR TRANSFORMER AUT5 C NFLUX IS PROPORTIONAL TO THE INDUCTIVE CURRENT IN THESE SWITCHES: C TYPE 91 TACS SOURCES DERIVED FROM NETWORK SWITCH CURRENTS: C <----A---><----A---><----A---> 91AUT5D1 -1.0 999. 91AUT5D2 -1.0 999. 91AUT5D3 -1.0 999. 91AUT5D4 -1.0 999. 91AUT5D5 -1.0 999. C ZERO-ORDER BLOCKS TO SCALE ABOVE TO NFLUX LINKED (NREF*PHI) WITHIN CORE VOLUME C GAIN IS BASED ON 10000-HENRY FLUX-MONITORING INDUCTANCES. C + + + + + 0AUT5E1 +AUT5D1 1.E4 0AUT5E2 +AUT5D2 1.E4 0AUT5E3 +AUT5D3 1.E4 0AUT5E4 +AUT5D4 1.E4 0AUT5E5 +AUT5D5 1.E4 C Correction of Type-68 TACS device lines on 6 April 2000. VAX ATP error was C traced to the following data cards (one line for each Type-68 device) having C E-field numbers that were not right-adjusted. The first two show how data C columns were misalligned prior to correction: C AUT5T168AUT5E1 13.821 27.642 2.91727E-3 8.44316E-4 C AUT5T268AUT5E1 27.642 62.194 2.43012E-3 2.53199E-4 C The associated data format for 4E16.0 read in GUTS2A was not properly C protected by the addition of BN, so exponents were scaled by a factor of C 100. For example, the first value on the first card was 2.91E-300 instead C of E-3 as intended. After proper allignment, here is the data: C NEW TACS DEVICE 68 MODELS HYSTERETIC LOSS/RESIDUAL MMF COMPONENTS. C 68 99AUT5T168AUT5E1 13.821 27.642 2.91727E-3 8.44316E-4 99AUT5T268AUT5E1 27.642 62.194 2.43012E-3 2.53199E-4 99AUT5T368AUT5E1 62.194 82.925 1.23361E-3 4.76044E-4 99AUT5T468AUT5E1 82.925 103.66 1.22440E-3 5.90543E-4 99AUT5T568AUT5E1 103.66 120.25 1.79633E-3 1.56980E-3 99AUT5T668AUT5E1 120.25 131.34 1.44373E-3 3.08189E-3 99AUT5T768AUT5E1 131.34 139.51 4.95355E-3 2.07092E-2 99AUT5U168AUT5E2 13.821 27.642 2.91727E-3 8.44316E-4 99AUT5U268AUT5E2 27.642 62.194 2.43012E-3 2.53199E-4 99AUT5U368AUT5E2 62.194 82.925 1.23361E-3 4.76044E-4 99AUT5U468AUT5E2 82.925 103.66 1.22440E-3 5.90543E-4 99AUT5U568AUT5E2 103.66 120.25 1.79633E-3 1.56980E-3 99AUT5U668AUT5E2 120.25 131.34 1.44373E-3 3.08189E-3 99AUT5U768AUT5E2 131.34 139.51 4.95355E-3 2.07092E-2 99AUT5V168AUT5E3 13.821 27.642 2.91727E-3 8.44316E-4 99AUT5V268AUT5E3 27.642 62.194 2.43012E-3 2.53199E-4 99AUT5V368AUT5E3 62.194 82.925 1.23361E-3 4.76044E-4 99AUT5V468AUT5E3 82.925 103.66 1.22440E-3 5.90543E-4 99AUT5V568AUT5E3 103.66 120.25 1.79633E-3 1.56980E-3 99AUT5V668AUT5E3 120.25 131.34 1.44373E-3 3.08189E-3 99AUT5V768AUT5E3 131.34 139.51 4.95355E-3 2.07092E-2 99AUT5W168AUT5E4 14.120 28.239 2.69692E-3 7.64028E-4 99AUT5W268AUT5E4 28.239 63.538 2.24656E-3 2.29122E-4 99AUT5W368AUT5E4 63.538 84.717 1.14043E-3 4.30775E-4 99AUT5W468AUT5E4 84.717 105.90 1.13191E-3 5.34391E-4 99AUT5W568AUT5E4 105.90 122.84 1.66066E-3 1.42058E-3 99AUT5W668AUT5E4 122.84 134.17 1.33473E-3 2.79024E-3 99AUT5W768AUT5E4 134.17 142.42 4.58432E-3 1.91911E-2 99AUT5X168AUT5E5 14.120 28.239 2.69692E-3 7.64028E-4 99AUT5X268AUT5E5 28.239 63.538 2.24656E-3 2.29122E-4 99AUT5X368AUT5E5 63.538 84.717 1.14043E-3 4.30775E-4 99AUT5X468AUT5E5 84.717 105.90 1.13191E-3 5.34391E-4 99AUT5X568AUT5E5 105.90 122.84 1.66066E-3 1.42058E-3 99AUT5X668AUT5E5 122.84 134.17 1.33473E-3 2.79024E-3 99AUT5X768AUT5E5 134.17 142.42 4.58432E-3 1.91911E-2 C CURRENTS INJECTED INTO NETWORK FOR HYSTERESIS MODELING AND OTHER COMPENSATION. C PHASE LEGS HAVE CURRENTS INJECTED TO COMPENSATE COUPLING XFMR. MAGNET. AMPS., C FLUX-MONITORING INDUCTANCE CURRENT AND SMALL CORE GAPS: 99AUT5TP =+AUT5D1/( 1.30893E-2)+ 1*(+AUT5T1+AUT5T2+AUT5T3+AUT5T4+AUT5T5) 99AUT5A5 =+AUT5TP+ 1*(+AUT5T6+AUT5T7) 99AUT5TQ =+AUT5D2/( 1.30893E-2)+ 1*(+AUT5U1+AUT5U2+AUT5U3+AUT5U4+AUT5U5) 99AUT5B5 =+AUT5TQ+ 1*(+AUT5U6+AUT5U7) 99AUT5TR =+AUT5D3/( 1.30893E-2)+ 1*(+AUT5V1+AUT5V2+AUT5V3+AUT5V4+AUT5V5) 99AUT5C5 =+AUT5TR+ 1*(+AUT5V6+AUT5V7) C YOKES ARE COMPENSATED FOR MONITORING INDUCTANCE CURRENTS: 99AUT5TS =+AUT5D4+ 1*(+AUT5W1+AUT5W2+AUT5W3+AUT5W4+AUT5W5) 99AUT501 =+AUT5TS+ 1*(+AUT5W6+AUT5W7) 99AUT5TT =+AUT5D5+ 1*(+AUT5X1+AUT5X2+AUT5X3+AUT5X4+AUT5X5) 99AUT502 =+AUT5TT+ 1*(+AUT5X6+AUT5X7) C YOKE & PHASE HYSTERETIC CURRENTS ARE EXTRACTED, EXCEPT AT GROUND: 99AUT5B0 =-AUT501-AUT5B5 99AUT5C0 =-AUT502-AUT5C5 C End of $INCLUDE. File name = aut5-tck.inc C PARTS OF HYSTERESIS COMPONENTS FOR PLOTTING C FORTRAN STATEMENTS; 99= INPUT; 98= OUTPUT; 88= INSIDE C =< FREE FORMAT FORTRAN TO COL 80 -----> C TOTAL OF ALL HYSTER INJECTIONS FOR LEGS A,B,C 99ALLINA =+AUT5T1+AUT5T2+AUT5T3+AUT5T4+AUT5T5+AUT5T6+AUT5T7 99ALLINB =+AUT5U1+AUT5U2+AUT5U3+AUT5U4+AUT5U5+AUT5U6+AUT5U7 99ALLINC =+AUT5V1+AUT5V2+AUT5V3+AUT5V4+AUT5V5+AUT5V6+AUT5V7 C C TYPE 90 TACS SOURCES DERIVED FROM NETWORK VOLTAGES: C <----A---><----A---><----A---> C TYPE 91 TACS SOURCES DERIVED FROM NETWORK SWITCH CURRENTS (1ST NODE) C <----A---><----A---><----A---> 91AUT5AM 0. 999. 91AUT5BM 0. 999. 91AUT5CM 0. 999. C FORTRAN STATEMENTS; 99= INPUT; 98= OUTPUT; 88= INSIDE C =< FREE FORMAT FORTRAN TO COL 80 -----> C NEXT IS SUM OF COUPLING OUTPUT AND ALL INJECTED AMPS - USED BY CORE MODELS 98LEGAMP = AUT5AM + AUT5A5 98LEGBMP = AUT5BM + AUT5B5 98LEGCMP = AUT5CM + AUT5C5 C NEXT IS APPARENT LEG AMPS FOR HYSTER PLOTTING 98NETAMP = LEGAMP - ALLINA 98NETBMP = LEGBMP - ALLINB 98NETCMP = LEGCMP - ALLINC C S-BLOCKS OF ORDER 1 IN COL 2 INTEGRATE VOLTAGES TO GET FLUXES (TIMES 250.0). C + + + + + C N0 & D0> C CD+< IN1> +< IN2> +< IN3> +< IN4> +< IN5> < A >< B >< C >< D >< E > C TACS OUTPUT REQUESTS - TYPE 33 C 33AUT5E1 ALLINA LEGAMP NETAMP 33AUT5E2 ALLINB LEGBMP NETBMP 33AUT5E3 ALLINC LEGCMP NETCMP C 33AUT5T1 AUT5T2 AUT5T3 AUT5T4 AUT5T5 AUT5T6 AUT5 33AUT5U1 AUT5U2 AUT5U3 AUT5U4 AUT5U5 AUT5U6 AUT5U7 BLANK CARD ENDING ALL ATP-SORTED TACS CARDS (from blank.inc) C THIS MUST BE EDITED TO ADD AUT5A5 >> AUT5AM, ETC METERING C $INCLUDE aut5-brn.inc C WSM makes non-Unix by adding a card using DOS editor. C 3-PHASE, 3-LEG SATURABLE XFMR. MODEL FROM MODELAD.FTN - 60. HZ IMPEDANCES C CHANGED AUT5A5, B5, C5 TO AUT5AM, BM, CM ; NOW REQUIRES METERING SWITCHES C AUT5 = IDENTIFYING BUS CODE PREFIX OF INTERNAL NODES. C AUT5AM = NAME OF A METERING NODE ADDED AT AUT5A5 C EACH PHASE LEG HAS 3 WINDINGS WHOSE RATED VOLTAGES AND TURNS ARE: C WINDING 1 (INNERMOST) = 13.130 KV. TURNS = 44.998 C WINDING 1 INCLUDES THE EFFECT OF 0.5670 OHMS/PHASE INTERNAL (DELTA) REACTORS. C REACTOR RESISTANCE/PHASE MUST BE INCLUDED WITH ENTERED WINDING RESISTANCES. C WINDING 2 = 79.674 KV. TURNS = 273.052 C WINDING 3 = 119.655 KV. TURNS = 410.072 C INTERNAL REFERENCE = 29.179 KV. TURNS = 100.000 C *********BEGIN USER COMMENTS ************ C EAST GARDEN CITY AUTOTRANSFORMER*** one *** AT 345/138 KV TAP C ABBREVIATED HYSTERESIS MODEL; SEVEN OF POSSIBLE 11 COMPONENTS C .95 * .7274 SQUARE METER LEG; .97 * .7274 SQUARE METER YOKE C ACTUAL DELTA IS GROUNDED BETWEEN OUTER PHASES AT Y1. C EACH DELTA REACTOR X =.82;OR .5467 WHEN EQUALIZED; REACTOR R = .00427 EACH C ACTUAL DELTA WINDING R = .00510; + .002847 TO EQUALIZE = .0079467 ENTERED C AIR-CORE COUPLINGS FROM COUPCOIL, BASED ON SMIT DIMENSIONS. C B-H CURVE BASED ON ARMCO TRAN-COR H-0 from smit C B-H SHIFTED TO HIGHER CURRENTS IN 1.5 - 1.85 T RANGE C MILD OPEN HEARTH STEEL FOR TANK C TANK HAS 21.65 M CIRCUM; FLUX SHIELD = 3.15 M HIGH C 64-36 EDDY-HYST LOSS RATIO AT RATED; C ********** END USER COMMENTS ************ C COUPLING, LEAKAGE IMPEDANCE & WINDING LOSS MODEL FOR TRANSFORMER AUT5: C < R >< X >< R >< X >< R >< X > 51Y1 Y2 .01016 10124.87377 52EGCT1A 0.000 61435.16312.04458372798.88582 53EGCHEAEGCT1A 0.000 92263.79188 0.000559873.90936.12388840864.72923 54AUT5A1 0.000 22499.42979 0.000136528.47263 0.000205039.49115 0.000 50000.88921 55AUT5A2 0.000 22499.42071 0.000136530.26291 0.000205042.81249 0.000 50000.86902 0.000 50001.52468 56AUT5A3 0.000 22499.42350 0.000136530.70112 0.000205053.04695 0.000 50000.87522 0.000 50001.68516 0.000 50004.18094 57AUT5A4 0.000 22499.25174 0.000136528.91954 0.000205046.26382 0.000 50000.49352 0.000 50001.03270 0.000 50002.52680 0.000 50002.52680 58AUT5AM 0.000 22499.02967 0.000136526.09974 0.000205035.90210 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 51Y2 Y3 .01016 10124.87377 52EGCT1B 0.000 61435.16312.04458372798.88582 53EGCHEBEGCT1B 0.000 92263.79188 0.000559873.90936.12388840864.72923 54AUT5B1AUT5B0 0.000 22499.42979 0.000136528.47263 0.000205039.49115 0.000 50000.88921 55AUT5B2AUT5B0 0.000 22499.42071 0.000136530.26291 0.000205042.81249 0.000 50000.86902 0.000 50001.52468 56AUT5B3AUT5B0 0.000 22499.42350 0.000136530.70112 0.000205053.04695 0.000 50000.87522 0.000 50001.68516 0.000 50004.18094 57AUT5B4AUT5B0 0.000 22499.25174 0.000136528.91954 0.000205046.26382 0.000 50000.49352 0.000 50001.03270 0.000 50002.52680 0.000 50002.52680 58AUT5BMAUT5B0 0.000 22499.02967 0.000136526.09974 0.000205035.90210 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 51Y3 Y1 .01016 10124.87377 52EGCT1C 0.000 61435.16312.04458372798.88582 53EGCHECEGCT1C 0.000 92263.79188 0.000559873.90936.12388840864.72923 54AUT5C1AUT5C0 0.000 22499.42979 0.000136528.47263 0.000205039.49115 0.000 50000.88921 55AUT5C2AUT5C0 0.000 22499.42071 0.000136530.26291 0.000205042.81249 0.000 50000.86902 0.000 50001.52468 56AUT5C3AUT5C0 0.000 22499.42350 0.000136530.70112 0.000205053.04695 0.000 50000.87522 0.000 50001.68516 0.000 50004.18094 57AUT5C4AUT5C0 0.000 22499.25174 0.000136528.91954 0.000205046.26382 0.000 50000.49352 0.000 50001.03270 0.000 50002.52680 0.000 50002.52680 58AUT5CMAUT5C0 0.000 22499.02967 0.000136526.09974 0.000205035.90210 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 0.000 50000.00000 C PHASE LEG N-L MODELS FOR TRANSFORMER: AUT5 C PHASE LEG CORES ARE 3.60 METERS LONG WITH AREAS OF 0.6910 SQUARE METERS C THE B-H CURVE USED PASSES THROUGH THE FOLLOWING (TESLA, AMP/METER) POINTS: C LINEAR:(0.20, 4.00) AND FULLY SATURATED:(2.00, 15000.00) C FOR SS CALCS EACH LEG HAS REACT. OF 15740. OHMS (AT 100.0 TURNS) C BASED ON SLOPE OF POINT 8 BELOW, HAVING A FLUX DENSITY OF 1.7400 TESLA. C HYSTERESIS CURRENT OFFSET OF 3.115 AMPS IS MODELED AT SATURATION. C LAMINATIONS ARE MODELED BY 8 LAYERS; EDDY LOSSES USE ENTERED CORE RESISTIVITY. C

(--FOR TYPE 98 N/L REACTOR--) P C CURRENT -----> 98AUT5A5AUT5A6 4.912234.877 0.14400 3.45524 0.39528 6.91044 0.81635 15.54839 1.12175 20.73117 1.41336 24.18640 2.00198 25.91428 3.13598 28.16067 4.23399 30.06154 8.81421 31.96448 19.73322 32.83478 56.33246 33.72028 543.11535 34.87714 1083.11535 35.20278 9999 98AUT5A6AUT5A7AUT5A5AUT5A64.912234.877 98AUT5A7AUT5A8AUT5A5AUT5A64.912234.877 98AUT5A8 AUT5A5AUT5A64.912234.877 98AUT5B5AUT5B6AUT5A5AUT5A64.912234.877 98AUT5B6AUT5B7AUT5A5AUT5A64.912234.877 98AUT5B7AUT5B8AUT5A5AUT5A64.912234.877 98AUT5B8AUT5B0AUT5A5AUT5A64.912234.877 98AUT5C5AUT5C6AUT5A5AUT5A64.912234.877 98AUT5C6AUT5C7AUT5A5AUT5A64.912234.877 98AUT5C7AUT5C8AUT5A5AUT5A64.912234.877 98AUT5C8AUT5C0AUT5A5AUT5A64.912234.877 C HIGH RESISTANCE ACROSS EACH UNPARALLELED TYPE 98 ELEMENT ABOVE: C <----><---->RRRRRRXXXXXXCCCCCC AUT5A6AUT5A5 .1E+11 AUT5A7AUT5A6AUT5A6AUT5A5 AUT5A8AUT5A7AUT5A6AUT5A5 AUT5B6AUT5B5AUT5A6AUT5A5 AUT5B7AUT5B6AUT5A6AUT5A5 AUT5B8AUT5B7AUT5A6AUT5A5 AUT5C6AUT5C5AUT5A6AUT5A5 AUT5C7AUT5C6AUT5A6AUT5A5 AUT5C8AUT5C7AUT5A6AUT5A5 C YOKE MODELS FOR TRANSFORMER: AUT5 C EACH TOP+BOTTOM YOKE LENGTH IS 3.40 METERS WITH AREA OF 0.7060 SQUARE METERS C THE B-H CURVE USED PASSES THROUGH THE FOLLOWING (TESLA, AMP/METER) POINTS: C LINEAR:(0.20, 4.00) AND FULLY SATURATED:(2.00, 15000.00) C FOR SS CALCS EACH YOKE HAS REACT. OF 17026. OHMS (AT 100.0 TURNS) C BASED ON SLOPE OF POINT 8 BELOW, HAVING A FLUX DENSITY OF 1.7400 TESLA. C HYSTERESIS CURRENT OFFSET OF 2.943 AMPS IS MODELED AT SATURATION. C LAMINATIONS ARE MODELED BY 8 LAYERS; EDDY LOSSES USE ENTERED CORE RESISTIVITY. C

(--FOR TYPE 98 N/L REACTOR--) P C CURRENT -----> 98AUT501AUT50A 4.636135.606 0.13600 3.52991 0.37332 7.05979 0.77100 15.88444 1.05943 21.17924 1.33484 24.70914 1.89076 26.47434 2.96176 28.76924 3.99877 30.71116 8.32457 32.65503 18.63694 33.54364 53.20871 34.44660 512.94277 35.60585 1022.94277 35.91340 9999 98AUT50AAUT50BAUT501AUT50A4.636135.606 98AUT50BAUT50CAUT501AUT50A4.636135.606 98AUT50CAUT5B0AUT501AUT50A4.636135.606 98AUT502AUT50FAUT501AUT50A4.636135.606 98AUT50FAUT50GAUT501AUT50A4.636135.606 98AUT50GAUT50HAUT501AUT50A4.636135.606 98AUT50HAUT5C0AUT501AUT50A4.636135.606 C HIGH RESISTANCE ACROSS EACH UNPARALLELED TYPE 98 ELEMENT ABOVE C <----><---->RRRRRRXXXXXXCCCCCC AUT50AAUT501AUT5A6AUT5A5 AUT50BAUT50AAUT5A6AUT5A5 AUT50CAUT50BAUT5A6AUT5A5 AUT50FAUT502AUT5A6AUT5A5 AUT50GAUT50FAUT5A6AUT5A5 AUT50HAUT50GAUT5A6AUT5A5 $VINTAGE, 1 C <----><---->RRRRRRRRRRRRRRRRXXXXXXXXXXXXXXXXCCCCCCCCCCCCCCCC P C SHUNT RESISTANCES FOR EDDY LOSS MODELING OF 1/8 LAMINATION THICKNESS LAYERS. C CROSS-FLUX CORE RESISTIVITY IS 12.00 E-8 OHM-METERS. C LAMINATIONS ARE 0.2286 MM THICK; MODELED LAYERS ARE 0.0286 MM THICK. AUT5A5 141050.02696 AUT5A6 70525.01348 AUT5A7 70525.01348 AUT5A8 70525.01348 AUT5B5AUT5B0 141050.02696 AUT5B6AUT5B0 70525.01348 AUT5B7AUT5B0 70525.01348 AUT5B8AUT5B0 70525.01348 AUT5C5AUT5C0 141050.02696 AUT5C6AUT5C0 70525.01348 AUT5C7AUT5C0 70525.01348 AUT5C8AUT5C0 70525.01348 AUT501AUT5B0 152575.09208 AUT50AAUT5B0 76287.54604 AUT50BAUT5B0 76287.54604 AUT50CAUT5B0 76287.54604 AUT502AUT5C0 152575.09208 AUT50FAUT5C0 76287.54604 AUT50GAUT5C0 76287.54604 AUT50HAUT5C0 76287.54604 C COMBINED WINDING 1-2 AND WINDING 2-3 STRAY (EDDY CUR.) LOSS MODEL: C 3-PHASE W1-W2 STRAY LOSSES AT RATED MVA = 636.953 KW. C WITH W1-W2 REF.VOLTS = 2.085 KV.; AND W2-W3 REF.VOLTS = 0.359 KV. AUT5A1AUT5A2 20.47385 AUT5B1AUT5B2 20.47385 AUT5C1AUT5C2 20.47385 C 3-PHASE W2-W3 STRAY LOSSES AT RATED MVA = 127.094 KW. C WITH W1-W2 REF.VOLTS = -0.495 KV.; AND W2-W3 REF.VOLTS = 7.203 KV. AUT5A2AUT5A3 1706.84712 AUT5B2AUT5B3 1706.84712 AUT5C2AUT5C3 1706.84712 C TANK WALL MULTIPLE CHAINED PI MODEL RESISTANCES REFERRED TO 100.000 TURNS: C MODELED SURFACE LAYERS ARE 0.1279 MM THICK; RHOTANK = 15.00 E-8 O-M. C <----><---->RRRRRRRRRRRRRRRRXXXXXXXXXXXXXXXXCCCCCCCCCCCCCCCC P AUT581 1.67957937 AUT582 3.35915874 AUT583 6.71831748 AUT584 13.43663496 AUT585 26.87326992 AUT586 53.74653984 AUT587 80.61980976 AUT504 161.23961952 $VINTAGE, 0 C TRANSFORMER TANK WALL MODEL FOR TRANSFORMER AUT5 C TANK WALL AND VERTICAL BRACING HAS A CROSS-SECTIONAL AREA OF 0.354 SQ. METERS. C ITS INSIDE PERIMETER IS 21.65 METERS. ITS AVERAGE THICKNESS IS 0.016 METERS. C EQUIVALENT HEIGHT AFFECTED BY ZERO SEQUENCE FLUX IS ASSUMED AT 3.15 METERS. C THE B-H CURVE USED PASSES THROUGH THE FOLLOWING (TESLA, AMP/METER) POINTS: C LINEAR:(0.20, 70.00) AND FULLY SATURATED:(2.05, 75000.) C STEADY-STATE CALCS USE THE SLOPES AT POINTS 3 BELOW. (1.0000 TESLA) C

(--FOR TYPE 98 N/L REACTOR--) P C CURRENT ----->< FLUX --------> 98AUT581 21.31036.333 2.20500 3.54468 5.04000 10.63404 10.39500 17.72340 22.05000 23.04042 50.40000 26.58510 110.25000 28.35744 226.80000 30.12978 441.00000 31.90212 1134.00000 35.44680 2362.50000 36.33297 4725.00000 38.03714 9999 98AUT582AUT581 21.31018.166 2.20500 1.77234 5.04000 5.31702 10.39500 8.86170 22.05000 11.52021 50.40000 13.29255 110.25000 14.17872 226.80000 15.06489 441.00000 15.95106 1134.00000 17.72340 2362.50000 18.16648 4725.00000 19.01857 9999 98AUT583AUT582 21.3109.0832 2.20500 0.88617 5.04000 2.65851 10.39500 4.43085 22.05000 5.76011 50.40000 6.64628 110.25000 7.08936 226.80000 7.53245 441.00000 7.97553 1134.00000 8.86170 2362.50000 9.08324 4725.00000 9.50929 9999 98AUT584AUT583 21.3104.5416 2.20500 0.44309 5.04000 1.32926 10.39500 2.21543 22.05000 2.88005 50.40000 3.32314 110.25000 3.54468 226.80000 3.76622 441.00000 3.98777 1134.00000 4.43085 2362.50000 4.54162 4725.00000 4.75464 9999 98AUT585AUT584 21.3102.2708 2.20500 0.22154 5.04000 0.66463 10.39500 1.10771 22.05000 1.44003 50.40000 1.66157 110.25000 1.77234 226.80000 1.88311 441.00000 1.99388 1134.00000 2.21543 2362.50000 2.27081 4725.00000 2.37732 9999 98AUT586AUT585 21.3101.1354 2.20500 0.11077 5.04000 0.33231 10.39500 0.55386 22.05000 0.72001 50.40000 0.83078 110.25000 0.88617 226.80000 0.94156 441.00000 0.99694 1134.00000 1.10771 2362.50000 1.13541 4725.00000 1.18866 9999 98AUT587AUT586 21.310.56770 2.20500 0.05539 5.04000 0.16616 10.39500 0.27693 22.05000 0.36001 50.40000 0.41539 110.25000 0.44309 226.80000 0.47078 441.00000 0.49847 1134.00000 0.55386 2362.50000 0.56770 4725.00000 0.59433 9999 98AUT504AUT587 21.310.56770 2.20500 0.05539 5.04000 0.16616 10.39500 0.27693 22.05000 0.36001 50.40000 0.41539 110.25000 0.44309 226.80000 0.47078 441.00000 0.49847 1134.00000 0.55386 2362.50000 0.56770 4725.00000 0.59433 9999 C FLUX SHIELD N-L MODEL FOR TRANSFORMER: AUT5 C FLUX SHIELDS ARE 3.15 METERS LONG WITH AREA OF 0.3652 SQUARE METERS. C THE B-H CURVE USED PASSES THROUGH THE FOLLOWING (TESLA, AMP/METER) POINTS: C LINEAR:(0.20, 4.00) AND FULLY SATURATED:(2.00, 15000.00) C FOR SS CALCS EACH F.S. HAS REACT. OF 9505. OHMS (AT 100.0 TURNS) C BASED ON SLOPE OF POINT 8 BELOW, HAVING A FLUX DENSITY OF 1.7400 TESLA. C

(--FOR TYPE 98 N/L REACTOR--) P C CURRENT -----> 98AUT503AUT504 2.896673.034 0.12600 7.30340 0.20475 14.60680 0.30870 32.86530 0.39690 43.82040 0.50400 51.12380 0.94500 54.77550 1.73250 59.52271 2.52000 63.53958 6.30000 67.55645 15.75000 69.38230 47.25000 71.20815 472.50000 73.03400 945.00000 75.06272 9999 C RESISTANCE UNDERESTIMATING FLUX SHIELD LOSSES, EQUAL TO LINEAR REACTANCE. C <----><---->RRRRRRXXXXXXCCCCCC P AUT503AUT504 9505.5 C TANK TOP MULTIPLE CHAINED PI MODEL RESISTANCES REFERRED TO 100.000 TURNS: C MODELED SURFACE LAYERS ARE 0.1172 MM THICK; RHOTANK = 15.00 E-8 O-M. $VINTAGE, 1 C <----><---->RRRRRRRRRRRRRRRRXXXXXXXXXXXXXXXXCCCCCCCCCCCCCCCC P AUT591 2.00000000 AUT592 4.00000000 AUT593 8.00000000 AUT594 16.00000000 AUT595 32.00000000 AUT596 64.00000000 AUT597 96.00000000 AUT509 192.00000000 $VINTAGE, 0 C TRANSFORMER TANK TOP MODEL FOR TRANSFORMER AUT5 C TANK TOP EDDY CURRENT PATH IS MODELED WITH A CIRCUMFERENCE OF 15.00 METERS. C AND A WIDTH OF 2.00 METERS; THICKNESS IS 0.015 METERS. C THE B-H CURVE USED PASSES THROUGH THE FOLLOWING (TESLA, AMP/METER) POINTS: C LINEAR:(0.20, 70.00) AND FULLY SATURATED:(2.05, 75000.) C STEADY-STATE CALCS USE THE SLOPES AT POINTS 3 BELOW. C

(--FOR TYPE 98 N/L REACTOR--) P C CURRENT ----->< FLUX --------> 98AUT591 13.53023.062 1.40000 2.25000 3.20000 6.75000 6.60000 11.25000 14.00000 14.62500 32.00000 16.87500 70.00000 18.00000 144.00000 19.12500 280.00000 20.25000 720.00000 22.50000 1500.00000 23.06250 3000.00000 24.14423 9999 98AUT592AUT591 13.53011.531 1.40000 1.12500 3.20000 3.37500 6.60000 5.62500 14.00000 7.31250 32.00000 8.43750 70.00000 9.00000 144.00000 9.56250 280.00000 10.12500 720.00000 11.25000 1500.00000 11.53125 3000.00000 12.07212 9999 98AUT593AUT592 13.5305.7656 1.40000 0.56250 3.20000 1.68750 6.60000 2.81250 14.00000 3.65625 32.00000 4.21875 70.00000 4.50000 144.00000 4.78125 280.00000 5.06250 720.00000 5.62500 1500.00000 5.76562 3000.00000 6.03606 9999 98AUT594AUT593 13.5302.8828 1.40000 0.28125 3.20000 0.84375 6.60000 1.40625 14.00000 1.82812 32.00000 2.10938 70.00000 2.25000 144.00000 2.39062 280.00000 2.53125 720.00000 2.81250 1500.00000 2.88281 3000.00000 3.01803 9999 98AUT595AUT594 13.5301.4414 1.40000 0.14063 3.20000 0.42187 6.60000 0.70312 14.00000 0.91406 32.00000 1.05469 70.00000 1.12500 144.00000 1.19531 280.00000 1.26563 720.00000 1.40625 1500.00000 1.44141 3000.00000 1.50901 9999 98AUT596AUT595 13.530.72070 1.40000 0.07031 3.20000 0.21094 6.60000 0.35156 14.00000 0.45703 32.00000 0.52734 70.00000 0.56250 144.00000 0.59766 280.00000 0.63281 720.00000 0.70312 1500.00000 0.72070 3000.00000 0.75451 9999 98AUT597AUT596 13.530.36035 1.40000 0.03516 3.20000 0.10547 6.60000 0.17578 14.00000 0.22852 32.00000 0.26367 70.00000 0.28125 144.00000 0.29883 280.00000 0.31641 720.00000 0.35156 1500.00000 0.36035 3000.00000 0.37725 9999 98AUT509AUT597 13.530.36035 1.40000 0.03516 3.20000 0.10547 6.60000 0.17578 14.00000 0.22852 32.00000 0.26367 70.00000 0.28125 144.00000 0.29883 280.00000 0.31641 720.00000 0.35156 1500.00000 0.36035 3000.00000 0.37725 9999 C ZERO SEQUENCE MODELS ARE BASED ON AIR-CORE MODELS, PLUS DERIVED VALUE FOR C X100T OF 3.29690 OHMS/PHASE TANK EFFECT, REFERRED TO 100 TURNS. C THE (UNSHIELDED) TANK TOP IS MODELED AND HAS STRAY LOSSES. C THE TANK WALL IS SATURABLE; BUT ANY STRAY LOSSES ARE NOT MEASURABLE. C THE FLUX SHIELD IS SATURABLE AND LACKS SIGNIFICANT DISTRIBUTED GAPS. C COUPLING REPRESENTING TOP AND SIDE GAP REACTANCES: C PER-PHASE SIDE GAPS = 1.846 OHMS; TOP GAPS = 1.814 OHMS AT 100 TURNS. 51AUT503 0.000 10.978117 52AUT509 0.000 5.4405450.000 5.440545 C SHUNT FLUX PATHS BYPASSING LEGS AND COILS OF EACH PHASE FOLLOW: C <----><---->RRRRRRXXXXXXXXXXXX 51AUT5A4AUT5B0 0.000 7.463166 51AUT5B4AUT5C0 0.000 7.463166 51AUT5C4AUT503 0.000 7.463166 C ZERO-SEQUENCE FLUXES FROM EACH PHASE TO TANK ARE EQUALIZED BY THE FOLLOWING: C <----><---->RRRRRRXXXXXXXXXXXXRRRRRRXXXXXXXXXXXXRRRRRRXXXXXXXXXXXX 51AUT503AUT502 0.000 3296.896640 52AUT502AUT501 0.000 3296.8933430.000 3296.896640 53AUT501 0.000 3296.8933430.000 3296.8933430.000 3296.896640 C NEXT COUPLING SUMS ZERO-SEQUENCE VOLTAGES OF OUTER WINDINGS AT NODE (0,8): C <----><---->RRRRRRXXXXXXXXXXXXRRRRRRXXXXXXXXXXXXRRRRRRXXXXXXXXXXXX 51AUT5A3 0.000 354922.42080 52AUT506 0.000 354922.065870.000 354922.42080 51AUT5B3AUT5B0 0.000 354922.42080 52AUT507AUT506 0.000 354922.065870.000 354922.42080 51AUT5C3AUT5C0 0.000 354922.42080 52AUT508AUT507 0.000 354922.065870.000 354922.42080 C NON-DEDICATED STRAY ZERO-SEQUENCE LOSS RESISTANCE: C PRODUCES 158.213 KW LOSS FOR ENTERED TEST CONDITIONS. $VINTAGE, 1 C <----><---->RRRRRRRRRRRRRRRRXXXXXXXXXXXXXXXXCCCCCCCCCCCCCCCC P AUT508 354.92242080 $VINTAGE, 0 C TIME-STEP-DAMPING BRANCHES ACROSS MAJOR AIR GAPS FOR HARMONIC ORDER: 500.0 C <----><---->RRRRRRXXXXXXCCCCCC AUT5A4AUT5B0 373.16 .00142 AUT5B4AUT5C0 373.16 .00142 AUT5C4AUT503 373.16 .00142 AUT501 164.84 .00322 AUT501AUT502 164.84 .00322 AUT502AUT503 164.84 .00322 C 10000-HENRY FLUX MONITORING INDUCTANCES TO INITIALIZE TACS. C <----><---->RRRRRRXXXXXXXXXXXXRRRRRRXXXXXXXXXXXXRRRRRRXXXXXXXXXXXX 51AUT5A5AUT5D1 3769911.185 51AUT5B5AUT5D2 3769911.185 51AUT5C5AUT5D3 3769911.185 51AUT501AUT5D4 3769911.185 51AUT502AUT5D5 3769911.185 C End of $INCLUDE. File name = aut5-brn.inc $VINTAGE, 1 C <----><---->RRRRRRRRRRRRRRRRXXXXXXXXXXXXXXXXCCCCCCCCCCCCCCCC P C CORNER GROUND DELTA Y1 1000. C ENERGIZE VIA 100 OHM RESISTANCES FOR DAMPING C 0.1 ohm is now essential; 100 is too large relative to xl SRC2A EGCT1A .1 SRC2B EGCT1B .1 SRC2C EGCT1C .1 $VINTAGE, 0 BLANK CARD ENDING ALL ATP-SORTED BRANCH CARDS (from blank.inc) C $INCLUDE aut5-swx.inc C WSM touches using DOS editor to make readable. C MEAS. SWITCHES TO MONITOR CURRENT IN 10000-HENRY CORE FLUX-MONITORING INDUCTS. C 67890123 MEASURING 56789012345678 P AUT5D1 MEASURING AUT5D2AUT5B0 MEASURING AUT5D3AUT5C0 MEASURING AUT5D4AUT5B0 MEASURING AUT5D5AUT5C0 MEASURING C End of $INCLUDE. File name = aut5-swx.inc C ********************************************* STUDY-DEPENDENT CABLE SWITCHING: C SWITCH CARD: COL 1-2 IS 0 FOR ORDINARY & GAUSS. STATISTICS SWITCHES, NOT TACS. C CLOSING, OPENING TIMES AND STATISTICS PARAMETERS BELOW ARE IN SECONDS. C 67890123 MEASURING 56789012345678 P C STATISTICS P C METERS INTERNAL CURRENT OF PHASE LEG A,B,C AUT5AMAUT5A5 MEASURING 1 AUT5BMAUT5B5 MEASURING 1 AUT5CMAUT5C5 MEASURING 1 C ENERGIZED AT ZERO SRCA SRC2A MEASURING 1 SRCB SRC2B MEASURING 1 SRCC SRC2C MEASURING 1 C ENERGIZED AT ZERO PLUS C SRCA SRC2A .0001 999. 1. C SRCB SRC2B .0001 999. 1. C SRCC SRC2C .0001 999. 1. BLANK CARD ENDING ALL ATP-SORTED SWITCH CARDS (from blank.inc) $LISTON C $INCLUDE aut5-src.inc C WSM touches file with DOS editor to make it into a DOS file C TACS-CONTROLLED SOURCES FOR HYSTERESIS/RESIDUAL MMF IN TRANSFORMER AUT5 C *= VOLTAGE IF POSITIVE; CURRENT IF NEGATIVE 60AUT5A5-1 60AUT5B5-1 60AUT5C5-1 60AUT5B0-1 60AUT5C0-1 60AUT501-1 60AUT502-1 C End of $INCLUDE. File name = aut5-src.inc C *= VOLTAGE IF POSITIVE; CURRENT IF NEGATIVE C C 138 KV L-L = 112676.5 V CREST or 1877.94 at 1 hertz C 40 percent = 751.17 C 60 percent = 1126.76 C 80 percent = 1502.35 C 100 percent = 1877.94 C 120 percent = 2253.53 14SRCA 1 2253.53 1. 0. -1.0 999. 14SRCB 1 2253.53 1. -120. -1.0 999. 14SRCC 1 2253.53 1. -240. -1.0 999. BLANK CARD ENDING ALL ATP-SORTED SOURCE CARDS (from blank.inc) C < >< >< >< >< >< >< >< > EGCT1A EGCT1B EGCT1C BLANK CARD ENDING ALL ATP-SORTED OUTPUT CARDS (from blank.inc) 193.01 0.0 0.1 BRANCH TACS ALLINATACS ALLINBTACS ALLINC BLANK card ending plot cards BEGIN NEW DATA CASE C 2nd of 18 subcases. C "FREQUENCY SCAN" use, with subsequent plotting of phasors vs. frequency. C This test case uses uniform frequency spacing. It basically like DC-51, C but involves some variations of output and batch-mode plotting. PRINTED NUMBER WIDTH, 8, 2, { Each column of width 8 includes 2 blank bytes FREQUENCY SCAN, 60., 10., 280., 0, { 60 < f < 280 Hz in 10-Hz increments CHANGE PLOT FREQUENCY { Request to vary frequency of plot points and .PL4 file 5 5 10 3 20 1 { On step 5, change to 5, etc. 0.1 0.0 { dT of col 1-8 is not used; TMAX of 9-16 ==> no simulation 1 1 1 0 1 -1 5 5 10 3 20 1 C Note: 1-punch in col. 80 of the following card was replaced by LIN002 below. C This illustrates a second way to order current output: by branch name. C When the user does not name branches himself, NMAUTO = 1 means the C program will provide default names. The first branch is LIN001, so C this should be used to replace the column-80 punch. But this branch C is in series with the 2nd branch, and has no shunt leakage, so the C same current can be found from the second branch (name LIN002). This C alternative is more interesting since it shows current output seems to C be possible for the 1-phase, constant-parameter distributed model. GEN SEND 3.0 40. -1SEND MID .306 5.82 .012 20. { 20 miles of line from DC-37 1MID REC 34.372457.68.15781 { 10 miles of Pi from DC-3 EARTH 200. 1 C The preceding branch has a 1-punch in column 80 for current output. If a 3 C had instead been used, branch voltage would have been produced, too. Note C it would be equal to the node voltage of node EARTH. No choice of polarity C is possible, this way. More interesting is the "-5" request below, which C allows polarity reversal (a minus sign will be added). BLANK card ending branches C Note: 1-punch in col. 80 of the following card was replaced by SWT001 below. C This illustrates a second way to order current output: by switch name. REC EARTH MEASURING BLANK card ending (here, non-existent) switches 14GEN 100. 60. -1. BLANK card ending sources C Requests for output variables follow. About currents, there is no control C over location in the output vector. Switch currents always precede branch C currents, and both are in natural order (of component number). But order of C voltages is controlled by the user. Note requests for node voltages have C been split, and in between there is a request for a voltage difference. C This is the order in the output vector, note. Yet, all voltage requests C made here (after the blank card ending source cards) follow all that might C have been made with column-80 punches on branch or switch cards. GEN SEND MID REC -5 EARTH { -5 ==> 2A6 name pairs for voltage differences (branch V) -1SWT001LIN002 { -1 ==> Branch/switch current out; use A6 component names EARTH { Node voltage. Note branch voltage 2 lines above has opposite polarity BLANK card ending node voltage outputs PRINTER PLOT C Note the peculiar 60 units/inch on the following plot cards. This is not C a mistake. With 6 lines/inch of the printer plot, & the F-scan frequency C increment of 10 Hz, there will be exactly one line for each freq step: 14660. 60.280. GEN SEND MID REC 19660. 60.280. SEND MID EARTH REC BLANK card ending plot cards BEGIN NEW DATA CASE C 3rd of 18 subcases illustrates IF-THEN-ENDIF for conditional data C assembly as described in the July, 1998, newsletter. Environment C variable RESISTOR must be set prior to use, in order. to avoid C an error stop. This can not be set internally, unfortunately. PRINTED NUMBER WIDTH, 9, 2, POWER FREQUENCY, 50., .0001 .020 50. 1 1 1 1 1 -1 5 5 20 20 C 1st of 2 identical, disconnected networks uses manually-defined branches: C IF ( RESISTOR .EQ. 'ONEHALF' ) SET VARIABLE { Execute OS command SET RESISTOR=ONEHALF IF ( RESISTOR .EQ. 'ONEHALF' ) THEN { If user wants to use resistive alternative: GEN TRAN 0.5 { 1st of 3 alternatives is resistor ELSEIF ( RESISTOR .EQ. 'ZERO' ) THEN { If user wants to use resistive alternative: GEN TRAN 0.0 { 2nd of 3 alternatives is illegal resistor ELSE { Alternatively (if environment variable RESISTOR is not YES): GEN TRAN 0.5 { 3rd of 3 alternatives is inductor ENDIF { Terminate 5-line IF statement that uses RESISTOR to choose the model TRAN 0.5 BLANK card ending program branch cards. BLANK card terminating program switch cards (none, for this case) 14GEN 1.414 50. 0.0 -1. BLANK card terminating program source cards. GEN TRAN BLANK card ending program output-variable requests. BLANK card ending all plot cards BEGIN NEW DATA CASE --- NEWSORT --- C 4th of 18 subcases illustrates new /-card sorting as described in a story of C the January, 1999, newsletter. Correction of trouble sorting /INITIAL was C involved as explained by Prof. Juan A. Martinez Velasco of the Polytechnic C University of Catalunya in Barcelona, Spain. This is a modification of his C FILE4.DAT DELTAT is increased and # of steps is reduced. Blank cards no C longer are required, so have been removed. A requirement of the new logic C is that all data after miscellaneous data cards must be sorted. This was C not true of the old logic (see DC-8). Note that some data classes are not C being used (no problem). Note the NEWSORT declaration on line 1, which C was patterned after the use of NOSORT in DC-36. In effect, the NEWSORT C request will make sure SZBED carries a minus sign. Not illustrated is C parallel OLDSORT to make sure SZBED does _not_ carry a minus sign (this C would force use of the old sorting logic). It is assumed STARTUP remains C as it was (no minus sign applied to SZBED) in order to handle DC-8. .000100 .010 1 1 -1 5 5 20 20 /OUTPUT BUS-0 BUS-1 BUS-2 /SOURCE 14BUS-0 1.0 50.0 0.0 -1.0 1.0 /BRANCH BUS-0 BUS-1 10.0 10.0 BUS-1 10.0 { Illustrate an in-line comment C Comment cards are allowed anywhere, of course. Add one here after C the first two branches just to show there is no problem. /SWITCH BUS-1 BUS-2 .00035 1.0 1.0 1 /BRANCH BUS-2 10.0 /INITIAL 2BUS-2 0.5 BEGIN NEW DATA CASE C 5th of 18 subcases is an improvement over DC4AT5.DAT of SPY @5. But that C required SPY. Beginning 23 August 1998, parametric studies can be run C using higher-level batch-mode commands. See October, 1998, newsletter. C Following request carries params MAXKNT IOPCVP NOSTAT POCKET CALCULATOR VARIES PARAMETERS 5 1 0 { Loop five times PRINTED NUMBER WIDTH, 11, 2, { Request maximum precision (for 8 output columns) .0001 .020 1 -1 1 1 1 -1 2 5 5 20 20 $PARAMETER { This will be serviced by CIMAGE just as any other $-card would be C __MILLIHENRIES__ = 1000. - ( KNT - 1.0 ) * 100. { L =1000, 900, 800, 700, & 600 mHenry C MILLIH = 1000. - ( KNT - 1.0 ) * 100. { L =1000, 900, 800, 700, & 600 mHenry VALUE = 1000. - ( KNT - 1.0 ) * 100. $$ { L =1000, 900, 800, 700, & 600 mHenry __MILLIHENRIES__ = VALUE { 1st of 2 uses of preceding intermediate variable MILLIH = VALUE { 2nd of 2 uses of preceding intermediate variable C Note about preceding. There are two uses of the same inductance value. The C 2nd is 6 columns wide for use with narrow-format series R-L-C data. The 1st C is for wide-format, which allows 16 columns. The underscore is required as C the first character if it is not a letter. Note that the second symbol is C contained in the first, so order is mandatory. If order were switched, an C error message would terminate execution following a warning message: C ++++ Notice. Symbol 1 is contained within symbol 2. C Swap these two and try again. C Intermediate variable VALUE was introduced 29 Dec 98. Before that, each C of the two symbols had the right hand side of VALUE. For this trivial C case, the difference is unimportant. But for other cases, the formula is C much too long to be contained on a single line. So, it is broken down C into components, with an intermediate variable used for each. Note the C $$ on the right of the intermediate variable. By definition, such a C variable is used within the pocket calculator only. The name is not to be C searched for and replaced in later data cards outside of the $PARAMETER C block (terminated by the following blank card). For industrial-strength C use of intermediate variables, see Orlando Hevia's induction motor model C in Rule Book documentation of $PARAMETER. BLANK card ends $PARAMETER definitions that are processed just b4 branch cards C Series R-L-C, narrow: RRRRRRLLLLLLCCCCCC TRAN 10.0 3 GEN TRAN6 5.0MILLIH { $PARAMETER will define this inductance $VINTAGE, 1, { The preceding used old, normal, narrow format. Switch to wide C Series R-L-C, wide: RRRRRRRRRRRRRRRRLLLLLLLLLLLLLLLLCCCCCCCCCCCCCCCC TRAN6 10.0 3 GEN TRAN 5.0__MILLIHENRIES__ { $PARAMETER will define this inductance $VINTAGE, 0, { Return to old, normal, narrow format BLANK card ending program branch cards. BLANK card terminating program switch cards (none, for this case) 14GEN 70. 50. -1. BLANK card terminating program source cards. GEN TRAN TRAN6 BLANK card ending program output-variable requests. CALCOMP PLOT { Needed to cancel PRINTER PLOT of preceding subcase 194 2. 0.0 20. -.40 .40TRAN BLANK card ending batch-mode plot cards -1 10.TRAN { Statistical tabulation of branch voltage (TRAN, TERRA) BLANK card ending all statistical tabulation request cards BEGIN NEW DATA CASE C 6th of 18 subcases illustrates data that created the compiled TACS code C that is built into the UTPF. As used in RUN.BAT the output will be C CODETACS.INC and DECKTAC*.INC --- files put in the UTPF. The logic C shown below was reproduced 9 times to produce the 60 variables mentioned C in the January, 1998, newsletter (see TAL explanation). Below will be C found the basic pocket calculator formulas that were used to illustrate C how slow Dube's MODELS and TACS are for such basic mathematics. C BENCHMARK DC-18b --- original source of formulas to be found below. COMPILED TACS MAKE { Alternative to SET COMPTACS=MAKE used because later subcase SINGLE STEP IF MAKE { Program will reduce TMAX to DELTAT if SET COMPTACS=MAKE PRINTED NUMBER WIDTH, 10, 2, { Request maximum precision (for 8 output columns) ABSOLUTE TACS DIMENSIONS 10 90 100 20 30 400 350 60 .00002 2.0 1 -11 0 0 1 -1 5 5 20 20 100 100 1000 1000 10000 10000 TACS STAND ALONE 99TEST1 = 10.0 * ( 1.0 + TIMEX ) ** 2 + 50. 99TEST2 = 1.E2 * COS ( 2.0 * 3.14159 * TEST1 / 100. ) 99TEST3 = 10.0 + 5.5 * TIMEX * SQRT ( ABS ( TEST2 ) ) 99TEST5 = 2.0 * PI * TEST1 / 100. 99TEST6 = 2.0 * SIN ( TEST5 ) 99TWOX = 2.E-4 * ( TEST2 * TEST2 + 2500. * TEST6 * TEST6 ) C 99TEST1 = 10. ---- This line is replacement for preceding to approximate removal of math 33TEST1 TEST2 TEST3 TEST5 TEST6 TWOX BLANK card ending all TACS data cards CALCOMP PLOT { Negative IPLOT = -11 will ignore the following plot cards 143 .2 0.0 2.0 TEST1 TEST2 TEST3 143 .2 0.0 2.0 TEST5 TEST6 TWOX BLANK card ending plot cards BEGIN NEW DATA CASE C 7th of 18 subcases illustrates use of pocket calculator to vary the frequency C of a distributed-parameter transmission line within the FREQUENCY SCAN loop C by means of TO SUPPORTING PROGRAM to provide LINE CONSTANTS rederivation. C This use of LINE CONSTANTS for each frequency of the F-scan will not be C seen in the .LIS file as proof of use unless /OUTPUT is added to the C request that declares "TO SUPPORTING PROGRAM" (see illustration on comment). C This subcase added 9 November 1999. This does work. Prior to adding the data C here, there was an attempt to add it to end of DC-52, but that failed. It C seems there is some conflict with LMFS (complicated). The $PUNCH statement C was not being executed right. Ignore that problem by moving data here. C FMIMFS--DELFFS--FMAXFS--NPD----- C 11111111222222223333333344444444 FREQUENCY SCAN 2.0 200. 8 C Following request carries params MAXKNT IOPCVP NOSTAT POCKET CALCULATOR VARIES PARAMETER 1 1 0 C Preceding MAXKNT is not used since FREQUENCY SCAN is involved. ATP will C internally calculate the correct # of steps from FMINFS, FMAXFS, etc. C Note we scan from 2 Hz to 200 Hz --- 2 decades, with 8 points/decade C for a total of 16 plus the starting point, so 17 total. The output C looks like exponential explosion as resonance around 300 Hz (would the C 138 miles be 1/4 of a wavelength at 300 Hz? Probably) is approached. C The 8 points/decade are needed to see the right edge accurately. C 30 July 2001, preceding NOSTAT is added. This is a binary switch C to suppress statistical tabulation ("no statistics") if value is 1. C But for F-scan, there is no statistical tabulation, so value is C immaterial. $PARAMETER C FRECU_____ = 0.01* ( EXP( LOG(10.0) /10.0 ) ) ** (KNT-1.0) C The preceding was Tsu-huei's original, single desired line. But when first C tried, there was an error (see April, 2000 newsletter). For about half C a year, the following lower-order equivalent was required: C EXPON = KNT - 1.0 $$ C BASE = EXP( LOG(10.0) / 10.0 ) $$ C FRECU_____= .01 * BASE ** EXPON C Preceding was not nice for plotting. Switch to 8 points/decade beginning C at 2 Hz (below which nothing much happens): C EXPON = KNT - 1.0 $$ C The preceding line worked fine. But it was not as demanding of pocket C calculator logic as the following. As mentioned in the newsletter, the C leading "-" on the leading constant was mishandled prior to 13 June 2000: C EXPON = -1.0 + KNT $$ C BASE = EXP( LOG(10.0) / 8.0 ) $$ C FRECU_____= 2.0 * BASE ** EXPON C The preceding 3 lines were used successfully until 13 October 2007. Then, C new logic to handle extraneous parens was added to POCKET, and it seemed C that this also should handle the original problem. To prove it, comment C out the preceding 3 cards and replace by the equivalent original. The .LIS C file is esstentially unchanged. WSM, 13 Oct 2007 : FRECU_____ = 2.0 * ( EXP( LOG(10.0) /8.0 ) ) ** (KNT-1.0) BLANK card terminates $PARAMETER definitions EXACT PHASOR EQUIVALENT { Switch from lumped-R to exact Pi-equiv. of distributed C The preceding is not really necessary. It should make the representation more C accurate, of course. At the highest frequency, if we are approaching 1/4 of C a wavelength, there should be a difference. But that difference seems small. C Consider the output vector at the ending frequency, with and without EPE: C With EPE: 17 200. 303000. 0.0 371408.7751 -8.40003412 583671.5409 -30.2933187 1684.552878 49.89669008 C Without : 17 200. 303000. 0.0 371386.1316 -8.41590291 583780.5064 -30.3165151 1685.249616 49.82908871 PRINTED NUMBER WIDTH, 13, 2, { Request maximum precision (if 8 output columns) -1. -1. 60. 60. { Note dT and Tmax are ignored for FREQ SCAN 1 -1 1 0 1 SSSS1AAAAA1A 15. 1 SSSS1BBBBB1B 15. SSSS1CCCCC1C 15. C Add /OUTPUT anywhere on right of following request to produce output of C the supporting program. Without this, all output will be suppressed. C < TO SUPPORTING PROGRAM (NEXT) > /OUTPUT { Add to see LINE CONSTANTS output < TO SUPPORTING PROGRAM (NEXT) > { Request for a jump to a supporting program C The following data, through the blank card, is from 2nd subcase of DCNEW-3. LINE CONSTANTS { This is the supporting program that is to be executed in-line BRANCH AAAA1AAAAA2ABBBB1BBBBB2BCCCC1CCCCC2C UNTRANSPOSED { Request for the line to be represented as untransposed C LINE CONSTANTS DATA FOR JOHN DAY TO LOWER MONUMENTAL 500-KV LINE. 1.3636 .05215 4 1.602 -20.75 50. 50. 1.3636 .05215 4 1.602 -19.25 50. 50. 2.3636 .05215 4 1.602 - 0.75 77.5 77.5 2.3636 .05215 4 1.602 0.75 77.5 77.5 3.3636 .05215 4 1.602 19.25 50. 50. 3.3636 .05215 4 1.602 20.75 50. 50. 0.5 2.61 4 0.386 -12.9 98.5 98.5 0.5 2.61 4 0.386 12.9 98.5 98.5 BLANK CARD ENDING CONDUCTOR CARDS OF "LINE CONSTANTS" CASE 100. FRECU_____ 1 11 1 138. 1 BLANK CARD ENDING FREQUENCY CARDS OF "LINE CONSTANTS" CASE $PUNCH, sup.pch ! { Output the just-created branch cards to a disk file BLANK card ending LINE CONSTANTS data cases { Last of supporting program data C End of data for supporting program. Branch data input now resumes: $CLOSE, UNIT=7 STATUS=KEEP { Disconnect file containing KC Lee branch cards $OPEN, UNIT=7 STATUS=UNKNOWN FILE=dumpch.lis FORM=FORMATTED $INSERT, sup.pch, { Dynamic connection of disk file created by $PUNCH above AAAA2A 1000. { Resistive load at far end of 1st of 3 phases BBBB2B AAAA2A { 2nd of 3 phases is terminated by the same R CCCC2C AAAA2A BLANK card ending branch cards BLANK card ending switch cards POLAR OUTPUT VARIABLES { 1st of 3 alternatives for output gives mag, angle C Preceding is one of 3 alternatives. The other two are, after commenting: C BOTH POLAR AND RECTANGULAR { Request for (in order): mag, angle, real, imag C RECTANGULAR OUTPUT VARIABLES { 3rd of 3 alternative outputs gives real, imag 14SSSS1A 303000.FRECU_____ 0. -1. 14SSSS1B 303000.FRECU_____ -120. -1. 14SSSS1C 303000.FRECU_____ 120. -1. BLANK card ending source cards SSSS1AAAAA1AAAAA2A { Names of 3 nodes (all a-phase, note) for V-node output C Column headings for the 4 output variables follow. These are divided among the 3 possible FS variable classes as follows .... C First 3 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 1 output variables are branch currents (flowing from the upper node to the lower node); C For each variable, magnitude is followed immediately by angle. Both halves of the pair are labeled identically, note. C Step F [Hz] SSSS1A SSSS1A AAAA1A AAAA1A AAAA2A AAAA2A SSSS1A SSSS1A C AAAA1A AAAA1A C 1 2.0 303000. 0.0 303004.5676 -.028517799 301911.4209 -.185323276 301.7646507 1.720504686 C New parameter values follow: 1) 2.6670429 C 2 2.6670429 303000. 0.0 303008.121 -.038018478 301917.2789 -.247632726 301.7897693 2.294000532 C New parameter values follow: 1) 3.5565588 C 3 3.5565588 303000. 0.0 303014.4401 -.050681287 301931.0223 -.330715516 301.8820814 3.058524978 BLANK card ending output variable requests (here, only node voltages) C New parameter values follow: 1) 84.339301 C 14 84.339301 303000. 0.0 311692.5772 -1.38233484 333737.5916 -8.33491659 541.8603856 48.84191334 C New parameter values follow: 1) 112.46827 C 15 112.46827 303000. 0.0 319327.0284 -2.11141826 362848.3744 -11.7652408 709.4891996 53.86022223 C New parameter values follow: 1) 149.97884 C 16 149.97884 303000. 0.0 335125.0318 -3.6799135 425532.3722 -17.6205065 1015.840142 55.61763042 C New parameter values follow: 1) 200. C 17 200. 303000. 0.0 371408.7751 -8.40003412 583671.5409 -30.2933187 1684.552878 49.89669008 F-SCAN COMPONENTS MAG MAG { Access "mag" and "mag" next (only 1st used) 197.25 0.0 2.5 0.02000SSSS1AAAAA1A Magnitude of i-aAmperes F-SCAN COMPONENTS ANGLE ANGLE { Access "angle" and "angle" next 147.25 0.0 2.5 -40. 0.0AAAA1AAAAA2A Angles of v-nodeDegrees BLANK card ending plot cards BEGIN NEW DATA CASE C 8th of 18 subcases illustrates two separate $PARAMETER blocks within a C PCVP (POCKET CALCULATOR VARIES PARAMETERS) loop over time simulation. C The data comes from Prof. Juan Martinez as should be explained in the C April, 2001, newsletter. Prior to correction on 19 December 2000, symbol C RESIST was not being correctly handled on the 2nd pass. It should have C value 5 ohms on the first pass and 15 ohms on the second pass. Note that C it is in the 1st of two blocks. Prior to correction, memory of this was C being overlaid by memory for the 2nd block. Now, there is no overlaying. C 25 December 2000, add variable HOLD01 to verify the new non-volatile C storage (any variable HOLDxx that begins with the 4 letters HOLD). C Gabor Furst asked for this extension as should be explained in the C April, 2001 (or newer), newsletter. HOLD01 begins with value zero, C so has value 1 on first pass and 3 on 2nd pass of loop over KNT. The C variable is not really being used for anything in this illustration, C but it certainly could be. Except for its non-volatile nature, it is C a normal variable in every other respect. C Following request carries params MAXKNT IOPCVP NOSTAT POCKET CALCULATOR VARIES PARAMETERS 2 0 1 .000100 .030 20 -1 $PARAMETER RESIST = 5.0 + (KNT-1.0)*10.0 HOLD01 = HOLD01 + KNT { Non-volatile variable because 1st 4 letters "HOLD" BLANK card ending 1st of 2 $PARAMETER blocks $PARAMETER AB_TON = 0.01 ABTOFF = 0.01 + ( KNT - 1.0 ) * 0.01 GND__K = KNT SERIALIZE 'GND___' BLANK card ending 2nd of 2 $PARAMETER blocks GND__2 100. 0 GND__1 RESIST 0 SOURCAMIDD_A RESIST 0 BLANK card ending branch cards MIDD_AGND__K AB_TON ABTOFF BLANK card ending switch cards 14SOURCA 0 1.5E5 50. -1. 1. BLANK card ending source cards MIDD_A BLANK card ending names of nodes for voltage output BLANK card ending batch-mode plot cards BLANK card ending all statistical tabulation request cards BEGIN NEW DATA CASE C 9th of 18 subcases is copied from 5th, but MODELS data has been added C to process and display extrema. This service begins 1 January 2001. C Optional col. 32: MEXTR 0 or 1 ==> minimum extrema; 2 ==) all extrema MODELS PROCESSES EXTREMA 0 { Use MODELS in simulation but only for extrema C Following request carries params MAXKNT IOPCVP NOSTAT POCKET CALCULATOR VARIES PARAMETERS 5 1 1 { Loop five times PRINTED NUMBER WIDTH, 11, 2, { Request maximum precision (for 8 output columns) .0001 .020 1 -1 1 1 0 -1 5 5 20 20 C Note integer miscellaneous data card has extrema flag MAXOUT = 0 as proof C that the user does not need to remember to set it. PCVP automatically will C set this to unity. After that, MEXTR will suppress such output for passes C 2 onward unless MEXTR = 2 (which will show all extrema). MODELS output C will always be seen, although note that such output will disappear if all C write( statements are removed from the MODELS data. In the following C data, INPUT variables are special. In place of the usual variable types, C one of 4 new ones must be used: Xmax, Tmax, Xmin, and Tmin corresponding to C the 4 rows of extrema output. The name within parentheses ("dummy") is not C used, so can be any character string beginning with a letter. It can not C be omitted, however (unfortunately). Next comes the fixed in-line comment, C which can begin in any column. But, to the right of the colon, format is C fixed. It is assumed I2, 1X, 2A6 information is involved. The integer is C the same one used on a batch-mode plot card: 4 if node voltage, 8 if branch C voltage, and 9 if branch current (or an S.M., a U.M., or a TACS variable). C The two names that follow correspond to the printed heading. One minor C enhancement is use of symbolic BLANK, which ATP interprets as real blanks. C Each INPUT variable must be on its own line (see the 4 lines just below). C 2 January 2001, WRITE( statements are left-adjusted to column 1 to take C advantage of new logic that will hold the user's case within text (as C delineated by apostrophes). C 3 January 2001, add an initial $PARAMETER block that has no use. This is C inspirated by Prof. Martinez's data TESTA1.DAT which inspired correction C to allow such practice. $PARAMETER { This will be serviced by CIMAGE just as any other $-card would be UNUSED = 5.0 { Following data does not involved character string "unused". BLANK card ends $PARAMETER definitions that are processed before MODELS data MODELS INPUT tran1 {Xmax(dummy)} -- Plot type and variable names: 8 TRAN6 BLANK INPUT tran2 {Tmax(dummy)} -- Plot type and variable names: 8 TRAN6 BLANK INPUT tran3 {Xmin(dummy)} -- Plot type and variable names: 8 TRAN6 BLANK INPUT tran4 {Tmin(dummy)} -- Plot type and variable names: 8 TRAN6 BLANK MODEL m1 -- Begin definition of a model named m1 VAR pass, sum -- Verify that values are kept between passes INPUT row1 {dflt: 0} -- Connects with Xmax, the 1st of 4 extrema rows INPUT row2 {dflt: 0} -- Connects with Tmax, the 2nd of 4 extrema rows INPUT row3 {dflt: 0} -- Connects with Xmin, the 3rd of 4 extrema rows INPUT row4 {dflt: 0} -- Connects with Tmin, the 4th of 4 extrema rows INIT pass:=0 -- Begin with the pass number equal to zero sum:= 0 ENDINIT -- Begin with the sum of Xmax equal to zero EXEC -- For each new set of extrema, we want to do this: pass:=pass+1 -- Count the pass number of the loop over simulations sum := sum + row1 -- Sum all Xmax as trivial demonstration of math write(' ************ In MODELS, pass = ', pass, 'sum =', sum ) write( ' Four rows Xmax, Tmax, Xmin, Tmin = ', row1, row2, row3, row4 ) write( ' ' ) -- Blank line separates this pass from following one C 3 March 2001, to satisfy need of Prof. Juan Martinez in Barcelona, add the C following copy of the preceding 3 lines. This produces a 2nd copy of the C preceding output in disk file MODELS.1 that is connected to I/O unit 37 C as used by $CLOSE, $OS, and $OPEN commands immediately before plotting. write1(' ************ In MODELS, pass = ', pass, 'sum =', sum ) write1( ' Four rows Xmax, Tmax, Xmin, Tmin = ', row1, row2, row3, row4 ) write1( ' ' ) -- Blank line separates this pass from following one ENDEXEC -- End of operations for each new set of extrema ENDMODEL -- End of the model named m1 USE m1 as m1 INPUT row1 := tran1 INPUT row2 := tran2 INPUT row3 := tran3 INPUT row4 := tran4 ENDUSE ENDMODELS $PARAMETER { This will be serviced by CIMAGE just as any other $-card would be C __MILLIHENRIES__ = 1000. - ( KNT - 1.0 ) * 100. { L =1000, 900, 800, 700, & 600 mHenry C MILLIH = 1000. - ( KNT - 1.0 ) * 100. { L =1000, 900, 800, 700, & 600 mHenry VALUE = 1000. - ( KNT - 1.0 ) * 100. $$ { L =1000, 900, 800, 700, & 600 mHenry __MILLIHENRIES__ = VALUE { 1st of 2 uses of preceding intermediate variable MILLIH = VALUE { 2nd of 2 uses of preceding intermediate variable BLANK card ends $PARAMETER definitions that are processed just b4 branch cards C Series R-L-C, narrow: RRRRRRLLLLLLCCCCCC TRAN 10.0 3 GEN TRAN6 5.0MILLIH { $PARAMETER will define this inductance $VINTAGE, 1, { The preceding used old, normal, narrow format. Switch to wide C Series R-L-C, wide: RRRRRRRRRRRRRRRRLLLLLLLLLLLLLLLLCCCCCCCCCCCCCCCC TRAN6 10.0 3 GEN TRAN 5.0__MILLIHENRIES__ { $PARAMETER will define this inductance $VINTAGE, 0, { Return to old, normal, narrow format BLANK card ending program branch cards. BLANK card terminating program switch cards (none, for this case) 14GEN 70. 50. -1. BLANK card terminating program source cards. GEN TRAN TRAN6 BLANK card ending program output-variable requests. C 3 March 2001, to satisfy need of Prof. Juan Martinez in Barcelona, add the C following 6 $-cards. Together with preceding WRITE1 of MODELS, this C demonstrates the needed ability for MODELS to create a disk file that will C be disconnected and renamed at the end of a PCVP pass. When execution is C complete, TEMPFILE.LIS will contain the 5th and last of 5 successive C contents, of course. Result of the TYPE command will not be detected in C the .LIS file, unfortunately, but it can be seen on the screen during C execution, if there is any question about content of the preceding 4. Of C course, the WRITE1 statements are the same as WRITE, so in fact the C output of TYPE is indirectly documented in the .LIS file. $CLOSE, UNIT=37 STATUS=KEEP { Disconnect file connected to MODELS WRITE1( $OS, DEL TEMPFILE.LIS { Make sure this file does not exist so RENAME is legal $OS, RENAME MODELS.1 TEMPFILE.LIS { Save MODELS WRITE1( file to a new name $OS, TYPE TEMPFILE.LIS { Display on screen the MODELS WRITE1( file contents C $OS, PAUSE { For interactive use on a fast computer, allow time to inspect C the output produced by the preceding TYPE command. $OPEN, UNIT=37 FILE=models.1 FORM=FORMATTED STATUS=NEW ! { Connect new file for WRITE1( BLANK card ending batch-mode plot cards C The following request for statistical tabulation should work fine. But it C was demonstrated in the 5th subcase, so why repeat the burden here? C -1 10.TRAN { Statistical tabulation of branch voltage (TRAN, TERRA) C 30 July 2001, add preceding NOSTAT = 1. Since there is not going to be C any statistical tabulation, why burden associated storage unnecessarily? C Because NOSTAT = 1, in fact the following card will not be read. BLANK card ending all statistical tabulation request cards BEGIN NEW DATA CASE C 10th of 18 subcases is related to 8th. It illustrates the use of C character string substitution as well as numeric substitution within C an IF block. This need, too, comes from Prof. Martinez. Since the C network is purely resistive, there are no real dynamics other than C the switching. Note the first pass involves the switch to NAME2 C whereas the 2nd involves the switch to NAME1. On the first of 2 C passes, RESIST = 5 whereas on the second RESIST = 10. Increase C step size from 100 usec to 2 msec to speed simulation. Normally, C node names of character strings will be a full 6 bytes long. But C just to illustrate that shorter names are handled correctly, 5-byte C NAMEK is used in this subcase. WSM, 6 Jan 2001. C 26 January 2001, switch the order of the two parameters within the C IF block. Of course, the answer should not be affected. But that C error first reported by Prof. Juan Martinez would mistreat data of C which the first assignment was character rather than numeric. Switch C the order to demonstrate that the problem has been cured. C Following request carries params MAXKNT IOPCVP NOSTAT POCKET CALCULATOR VARIES PARAMETERS 2 0 1 .002 .030 1 -1 $PARAMETER { Begin block that defines numeric RESIST and character NAMEK C IF( KNT .EQ. 1.0 ) THEN This standard form involves THEN on right, note. IF( KNT .EQ. 1.0 ) { This is equivalent (THEN on right is not required) NAMEK = 'NAME2' { Beginning 26 January 2001, character line appears first RESIST = 5.0 + ( KNT - 1.0 ) * 10.0 { Previously, this numeric line was first ELSE { Alternatively (if not the first pass, so for KNT = 2 onward): NAMEK = 'NAME1' RESIST = ( KNT - 1.0 ) * 10.0 ENDIF { Terminate 7-line block that defines symbols RESIST and NAMEK BLANK card ending 1st of 2 $PARAMETER blocks NAME2 100. NAME1 RESIST SOURC MID RESIST BLANK card ending program branch cards. MID NAMEK 0.009 0.020 BLANK card terminating program switch cards 14SOURC 0 100. 50. -1. BLANK card terminating program source cards MID BLANK card ending program output-variable requests. BLANK card ending batch-mode plot cards BLANK card ending all statistical tabulation request cards BEGIN NEW DATA CASE C 11th of 18 subcases illustrates a PCVP loop over time simulations, with C MODELS used as a part of the simulation. This is unlike the preceding C 9th subcase, which involved MODELS PROCESSES EXTREMA. The 9th subcase C uses MODELS not within the dT loop, but rather only upon completion of C the simulation. The present use within the dT loop became operational C 17 May 2001 when SUBR1, TACS1, and TACS2 were corrected to properly C handle MAGVOLT.DAT from Orlando Hevia in Santa Fe, Argentina. C Following request carries params MAXKNT IOPCVP NOSTAT POCKET CALCULATOR VARIES PARAMETERS 3 1 0 { 3 passes, minim print $PARAMETER { This will be serviced by CIMAGE just as any other $-card would be VOLTAG = 25.0 $$ REACTV = 0.5 + 0.5*KNT - 0.5 $$ ADMITC = (REACTV/VOLTAG**2)*1000000. POWERR = REACTV BLANK card ends $PARAMETER definitions that are processed before MODELS data C 2.E-5 40.E-3 60.0 60.0 C 100 1 1 1 1 C Preceding miscellaneous data cards of MAGVOLT are modified to simplify and C speed this illustration. The PCVP card involves 3 passes whereas MAGVOLT C had 61. Reduce the number of time steps from 2000 to 50, and do not save C the .PL4 files (ICAT = 1 in column 64 creates .001, .002, etc. files). .0002 .010 60.0 60.0 1 1 1 1 0 -1 0 5 5 MODELS HYBRID { Begin MODELS data that is to participate in the simulation INPUT voltage {v(bus2a)} node0 {v(busx)} tmax {atp(tmax)} C Comment added 18 September 2001. The preceding line illustrates the ATP( C function of MODELS. In this case, it is used to read the value of variable C TMAX, which is the ending time of the study. Beware of loop counter KNT C of PCVP use. Inside the $PARAMETER block, KNT is used, note. But if C accessed using the ATP( function, a bad value will be obtained. This is C because KNT is not the name of the variable used to count the PCVP loops C within the program. Rather, MNT is. So, what the user might guess would be C ATP(KNT) instead must be ATP(MNT). The need for this warning was inspired C by a report of trouble by Marta Val Escudero of ESB International in Dublin, C Ireland. As should be reported in a future newsletter, "PCVP loop index KNT C conflicted with the energization number of STATISTICS prior to separation of C the two as summarized in the July, 2000, issue." MODEL maxim INPUT voltage parame tmax VAR maxim kvoltage INIT maxim:=0 ENDINIT EXEC kvoltage:=ABS(voltage) IF (kvoltage > maxim) THEN maxim:=kvoltage ENDIF IF (t > tmax) THEN C Only on the final time step, one line is written to the C MODELS.1 file. This is a little inconvenient for test cases. C We would like to redirect this output to the .LIS file by changing C WRITE1 to WRITE. But, since we are minimizing out (see the PCVP C declaration, output within the dT loop of the 2nd or later pass is C being suppressed, and only a single such output (for pass 1) would C be seen. So, we leave WRITE1. For the record, content should be: C DCN25, 11th subcase. PARAME = 0.5 MAXIM = 1.04828954 C DCN25, 11th subcase. PARAME = 1.0 MAXIM = 1.22325171 C DCN25, 11th subcase. PARAME = 1.5 MAXIM = 1.41404906 C Note we have added labeling and preserved lower case C by left-adjusting the WRITE1 command: C write1(parame, ' ', maxim) write1( ' DCN25, 11th subcase. PARAME = ', parame, 'MAXIM = ', maxim ) ENDIF ENDEXEC ENDMODEL USE maxim AS sensit INPUT voltage:=voltage parame:=node0 tmax:=tmax ENDUSE ENDMODELS BLANK card ending MODELS data. Next come branch data: BUSX POWERR BUS0A BUS1A .4948 1.979 BUS1A BUS2A 5.562 22.25 BUS2A 800.0 BUS3A ADMITC BLANK card ending program branch cards. C BUS1A BUS3A 5.E-3 1.0 Switch card of MAGVOLT used multiple of dT BUS1A BUS3A 5.001E-3 1.0 { Perturb T-close so problem with roundoff BLANK card terminating program switch cards (none, for this case) 11BUSX -1 1.0 0.0 1.0 14BUS0A 1.0 50.0 -90.0 -1.0 1.0 BLANK card terminating program source cards. BUS0A BUS1A BUS2A BUS3A BLANK card ending program output-variable requests. BLANK card ending plot cards (none for this data) BLANK card ending statistical tabulations (none illustrated here) BEGIN NEW DATA CASE C 12th of 18 subcases illustrates lack of difference in the first 6 bytes C of two variable names of $PARAMETER usage. This originally was disk C file DISABLE0.DAT from Prof. Juan Martinez in Barcelona, Spain. It C was sent to BPA on 14 September 2001. Because BLOCK_A1 & BLOCK_A2 C are the same through byte 6, execution fails. The first sign of trouble C is incorrect evaluation of BLOCK_A2 (note BLOCK_A1 is correct): C Parameter 2 defined. Value = "$ENABLE " ----- result of evaluation of the preceding IF-THEN-ELSE-ENDIF block. C Parameter 3 defined. Value = 1.500000E+00 ----- result of evaluation of the preceding IF-THEN-ELSE-ENDIF block. C Of course, parameter 3 should be character just as parameter 2 is. It C as not numeric as this indicates. As a result of this error, later data C is bad, and execution should end with "KILL = 4. The last card is a C series R-L-C branch having zero impedance. ..." $PARAMETER OPTI_1=1.0 BLANK $PARAMETER IF(OPTI_1.EQ.1.0) BLOCK_A1='$ENABLE ' BLOCK_A2='C ' ENDIF BLANK .1e-6 .2e-4 50 1 BLOCK_A1 BUS_10BUS__1 1.0 0 BLOCK_A2 BLANK card ending program branch cards. BLANK card terminating program switch cards (none, for this case) 14BUS_10 1.5E5 50. -1. 1. BLANK card terminating program source cards. BUS__1 BLANK card ending program output-variable requests. BLANK card ending plot cards (none for this data) BEGIN NEW DATA CASE C 13th of 18 subcases is a correction of the preceding subcase. The only C change is replacement of BLOCK_A2 by A2_BLOCK (2 places). Then the C two symbols are different through the first 6 bytes. The subcase now C is handled properly. Note that it represents a variation of the 10th C subcase, which had a note dated 26 January 2001. If the 10th subcase C no longer ends its $PARAMETER block with a character variable, this C present subcase does. Also, it illustrates the use of 2 of them within C a single block. $PARAMETER OPTI_1=1.0 BLANK $PARAMETER IF(OPTI_1.EQ.1.0) BLOCK_A1='$ENABLE ' A2_BLOCK='C ' ENDIF BLANK .1e-6 .2e-4 50 1 BLOCK_A1 BUS_10BUS__1 1.0 0 A2_BLOCK BLANK card ending program branch cards. BLANK card terminating program switch cards (none, for this case) 14BUS_10 1.5E5 50. -1. 1. BLANK card terminating program source cards. BUS__1 BLANK card ending program output-variable requests. BLANK card ending plot cards (none for this data) BEGIN NEW DATA CASE C 14th of 18 subcases illustrates dynamic as opposed to static $DISABLE C use. This originally was disk file DISABLE3.DAT from Prof. Juan C Martinez in Barcelona, Spain. It was sent to BPA on 14 September 2001. C Prior to a change to CIMAGE on 17 Sept 01, execution failed. $PARAMETER OPTI_1=1.0 BLANK $PARAMETER C IF(OPTI_1.EQ.1.0) BLOCK_A1='$ENABLE ' BLOCK_A2='C ' C ENDIF BLANK card ending $PARAMETER block $PARAMETER C IF(OPTI_1.EQ.1.0) BLOCK_B1='$DISABLE' BLOCK_B2='$ENABLE ' C ENDIF BLANK card ending $PARAMETER block .1e-6 .2e-4 50 1 BLOCK_A1 BUS_10BUS__1 1.0 0 BLOCK_A2 BLOCK_B1 BUS_10BUS__1 1.0 0 BLOCK_B2 BLANK card ending program branch cards. BLANK card terminating program switch cards (none, for this case) 14BUS_10 1.5E5 50. -1. 1. BLANK card terminating program source cards. BUS__1 BLANK card ending program output-variable requests. BLANK card ending plot cards (none for this data) BEGIN NEW DATA CASE C 15th of 18 subcases shows that a FREQUENCY SCAN with multi-part output C is possible with no output voltage. This addition is made 16 June 2002 C to demonstrate that a problem, first observed by Orlando Hevia, has been C corrected. The change to HEADL4 was simple: (N2 .EQ. NV) was replaced C by (N3 .EQ. NC). Before this correction, execution died in plotting as C ATP attempted to read too much of the .PL4 file. Now, everything is OK, C as the following shows. The plot illustrates a mixture of magnitude and C real part on the same plot. At the peak, which occurs at 160 Hz, both C have the same value, of course. 160 Hz is the resonant frequency of the C series R-L-C branch. The variation is nice and smooth and simple using C 10 Hz steps. Here, using 20 Hz, the curves are slightly lumpy. PRINTED NUMBER WIDTH, 10, 2, { Each column of width 10 includes 2 blank bytes FREQUENCY SCAN 20. 20. 200. { F in Hz = 20, 40, ... 200 1.E-6 .001 1 -1 1 1 XX 1. 1.0 1.E3 1 BLANK card ending branch cards (just this one) BLANK card ending switch cards (none) BOTH POLAR AND RECTANGULAR { Any of the 3 possible multi-part outputs works 14XX 100. 60. -1. BLANK card ends source cards C XX { Prior to correction of HEADL4 on 16 June 2002, node voltage was needed BLANK card ending names of nodes for node-voltage output F-SCAN COMPONENTS MAG REAL { Access magnitude and real parts next 19620. 0. 200. -40.120.XX XX { Plot the magnitude and the real part BLANK card ending plot cards BEGIN NEW DATA CASE C 16th of 18 subcases is added 18 February 2003 to demonstrate that the C complaint by Prof. Juan Martinez has been corrected. The preceding day, C he had sent REPLACE.DAT to demonstrate that a $PARAMETER block within C a dynamic $DISABLE block was being mishandled. The 3rd pass, execution C was being stopped by an error message within TACS (which is not even C being used). C Following request carries params MAXKNT IOPCVP NOSTAT POCKET CALCULATOR VARIES PARAMETERS 3 0 $PARAMETER C IF(MNT.GT.1.0) -- Original line until 16 April 2009 involved erroneous MNT IF( KNT .GT. 1.0 ) { Correct MNT --> KNT to agree with all other DC*.DAT KNT1____='$DISABLE' ELSE KNT1____='$ENABLE ' ENDIF BLANK card ends 1st $PARAM block .0002 .02 50 -1 $PARAMETER { SIMPLE PARAMETER DEFINITION, JUST TO ILLUSTRATE THE CASE PARAM2=1.0 BLANK card ends 2nd $PARAM block KNT1____ { This will be replaced by $ENABLE on 1st pass, & $DISABLE thereafter $PARAMETER { THIS BLOCK IS CAUSING CERTAIN DAMAGE WITHIN PCVP ROUTINE PARAM3=100.0 BLANK card ends 3rd $PARAM block VOLTSA PARAM3 $ENABLE { End of either $DISABLE or $ENABLE block that begins with KNT1____ $PARAMETER { THIS COULD BE ANY MATHEMATICAL COMPUTATION TO OBTAIN THE CONDI TIPO_F=MNT BLANK card ends 4th $PARAM block $PARAMETER { $PARAMETER BLOCK TO COMMENT OR UNCOMMENT NEXT BRANCH CARD IF(TIPO_F.LT.1.0) OPTION=' VOLT' ELSE OPTION='C VOLT' ENDIF BLANK card ends 5th $PARAM block OPTIONSA PARAM2 { BRANCH CARD RELATED TO PREVIOUS PARAMETER VOLTSA 1.0 BLANK card ends branch cards VOLT_AVOLTSA MEASURING 1 BLANK card ends switch cards 14VOLT_A 0 1. 50. -1. 1. BLANK card ends source cards VOLTSA BLANK card ends names for node voltage output BLANK card ends batch-mode plot cards BLANK card ends requests for statistical tabulation BEGIN NEW DATA CASE C 17th of 18 subcases is added 11 May 2003. This data comes from Ricardo C Tenorio of ABB in Vasteras, Sweden. The 2 symbol names are not distinct C through 6 bytes. Unlike the 12th subcase, ATP did not automatically C shut itself off with an unrelated error message. Mr. Tenorio found that C execution continued, and results were wrong. In E-mail dated May 5th, C he explained: I've tried to simplify the case as far as I could. The C case involves an EDM module designed to apply generic faults, i.e 1-ph, C 2-ph & 3-ph, to a named bus." Data interpretation provided the first C sign of trouble: "Only the first variable BLOCk_11 is correctly C interpreted as a string! The others are interpreted as number ... C Line 22 ends this IF-THEN-ELSE-ENDIF block. |ENDIF C Parameter 1 defined. Value = "$DISABLE" ----- result of evaluation ... C Parameter 2 defined. Value = 0.000000E+00 ----- result of evaluation ... C Yes, clearly wrong. Most serious was continued execution. So, a special C error trap was designed for MATDAT. Beginning May 11th, it halts this C case as well as the 12th subcase (which now is stopped sooner, and with C a relevant message for the user). What is needed (in addition to the C ambiguous names)? An IF( block and text (as opposed to just math). C This uses the pocket calculator, which in turn requires distinct names. C The first 6 bytes of each is BLOCK_ so data must be rejected. C EATS testing of RUNEATS.BAT required a modification. Immediately C below the $PARAMETER block will be found the two uses. Note that C these two cards have been commented out. The problem of EATS is C simple: switch cards must be read for counting, yet $PARAMETER is C skipped during counting. So, if not commented out, execution would C die in OVER4 with an illegal decode (BLOCK... is not a switch card). C So, we modify to avoid death at the hands of EATS. For normal (non- C EATS) use, the change makes no difference. Execution will be halted C before any data below the $PARAMETER loop is read. It makes no C difference what data is down there. 50.E-6 10.E-3 50. 50. 1 -1 1 1 1 -1 5 5 20 20 51SOUR1AEQUIVA 1.003 9.5900 52SOUR1BEQUIVB 0.319 6.9810 53SOUR1CEQUIVC BLANK card ends branch data $PARAMETER IF( 3.0 .EQ. 1.0 ) BLOCK_11='C ' BLOCK_12='C ' ELSEIF( 3.0 .EQ. 3.0 ) BLOCK_11='$DISABLE' BLOCK_12='C ' ENDIF BLANK card ends $PARAMETER block C BLOCK_11 { This symbol is to be replaced by either $DISABLE or a comment card C Note: the preceding card was commented to prevent death in OVER4 using EATS EQUIVA 0.0100 9999. C BLOCK_12 { This symbol is to be replaced by a comment card (either condition) C Note: the preceding card was commented to prevent death in OVER4 using EATS EQUIVAEQUIVB 0.0100 9999. EQUIVA 0.0100 9999. EQUIVB 0.0100 9999. EQUIVC 0.0100 9999. $ENABLE { End possible preceding $DISABLE after last switch card BLANK card ends switch data 14SOUR1A 428660.705 60. 0.00 -1. 1. 14SOUR1B 428660.705 60. -120.00 -1. 1. 14SOUR1C 428660.705 60. 120.00 -1. 1. BLANK card ending source cards C Since there will be an error message, remaining data is immaterial; it C will not be used, anyway. So let's avoid waste by eliminating remaining C data through the end of the subcase. BEGIN NEW DATA CASE C 18th of 18 subcases is added 8 October 2006. This data involves a PCVP C loop that produces 5 time-simulations with appended power and energy output C due to column-80 punches > 4. Original complaint was from Orlando Hevia of C Santa Fe, Argentina on 18 Aug 2006. Prior to correction in October of 2006, C column-80 punches above "4" were unusable with "FIND" code of statistical C tabulations. Punches 5-16 (use hex if above 9) serve to append power P or C energy "E" or both. The printed dT-loop output was believed to be correct. C But the subsequent "FIND" command offered no class for power & no class for C energy. Now it does. As for UTPF changes, look for WSM06AUG since the C work began in August. For uniformity, no SEP or OCT were used. The C vast majority of changes are confined to DICTAB. WSM, 6 October 2006 C Controls of following request card: MAXKNT IOPCVP NOSTAT { Loop 5 times with POCKET CALCULATOR VARIES PARAMETERS 5 1 0 { reduced printout $PARAMETER { Each loop KNT=1, 2, ... MAXKNT will involve different TIMECLOSEX TIMECLOSEX = (KNT)*0.0004 C 13 October 2007, the preceding line is modified by the addition of C extraneous parentheses around the KNT. Orlando Hevia had sent data to C show that RTR6__= 0.684+0.054*(KNT)**1.75 is wrong (Salford ATP dies) C whereas RTR6__= 0.684+0.054*KNT**1.75 works fine. This resulted C in a correction to POCKET. To build verification into a test cases, WSM C now simply adds unused parens () to this 18th subcase. Answer is unchanged. C However, this did _not_ stress the new logic as intended. Well, keep the C preceding addition but add a second, unused variable by copying Orlando : RTR6__= 0.684+0.054*(KNT)**1.75 { Orlando's formula that was fatal b4 Oct. 2007 BLANK card ending parameter block C DELTAT TMAX 1.E-4 .200 { Orlando Hevia had had 20 usec and .25 sec here C IOUT IPLOT IDOUBL KSSOUT MAXOUT 200 11 1 C Replace Orlando's J. Marti line by 42 constant-parameter miles from DC-38: -1FASEA1FASEA2 .305515.8187.01210 42. 0 { 42-mile, constant- -2FASEB1FASEB2 .031991.5559.01937 42. 0 { parameter, 3-phase -3FASEC1FASEC2 { transmission line. C For list of all column-80 punches for power or energy, see DC-37, subcase 6. C In the following, "4" is old whereas "G" and "D" date to 1999. Outputs C will be as shown farther below: FASEA2 1.0E01 G FASEB2 1.0E01 D FASEC2 1.0E01 4 GENEA OPENXX RTR6__ { Do-nothing branch will confirm value of RTR6 BLANK card ending branches GENEA FASEA1TIMECLOSEX 0.05 3 GENEB FASEB1TIMECLOSEX 0.05 0 GENEC FASEC1TIMECLOSEX 0.05 0 BLANK card ending switches 14GENEA 100000. 50. 0.0 -1.0 14GENEB 100000. 50. -120. -1.0 14GENEC 100000. 50. 120. -1.0 BLANK card ending sources FASEA1 C Column headings for the 9 EMTP output variables follow. These are divided among the 5 possible classes as follows .... C First 4 output variables are electric-network voltage differences (upper voltage minus lower voltage); C Next 3 output variables are branch currents (flowing from the upper node to the lower node); C Next 2 output variables are either power or energy or both as a pair (column 80 punches > 4). C Remember that 1 of the voltages really are powers, and currents are energies; the 4-punches will overlay. C Step Time FASEB2 FASEC2 GENEA FASEA1 GENEA FASEB2 FASEC2 FASEA2 FASEB2 C TERRA TERRA FASEA1 FASEA1 TERRA TERRA TERRA TERRA C End actual dT-loop output. For clarity, let's label each of the preceding C output variables, showing the variable class and how it was requested: C Variable class bran V power bran V node V bran I bran I energy energy energy C Produced by: "D" "4" "3" A6 name "3" "D" "4" "G" "D" C Summary class <----- 4 voltages appear first -----> <--- 3 currents next ---> <-- appended P,E C Note that power and energy of the 4-punch are counted among voltages and C currents whereas any due to punches 4-16 are appended on the right in a C separate, final variable class that here has 2 entries. 9 = 4 + 3 + 2 BLANK card ending node voltage outputs C The user who does not want to see any batch-mode plotting can preserve the C following 3 lines (no need to remove) while setting misc. data IPLOT < 0. PRINTER PLOT { Character mode always works. CALCOMP PLOT might not (unknown) 193.01 0.0 .06 BRANCH { Show the 3 energy signals on a printer plot FASEA2 FASEB2 FASEC2 BLANK blank card ending plot cards C BASE-------BUS1--BUS2-- ... -4 1.E6 FASEA2 { Compartments 93-95 will be used for base energy = 1.E6 -4 1.E6 FASEB2 { Compartments 103-105 will be used for ... -4 2.E6 FASEC2 { Compartments 44-49 will be used for base energy = 2.E6 -3 2.E8 FASEC2 { Compartments 18-24 will be used for base power = 2.E8 C Note about preceding. Three of the 4 tabulations lost a compartment when C the number of shots was reduced from 10 to 5. If the user wants to see a C smoother tabulation, he can increase shots to 10 or even 20. Of course, C TIMECLOSEX = KNT*0.0004 then would have its final factor correspondingly C reduced to .0002 or .0001, respectively. Orlando used 20. -1 1.E4 FASEB2 { Compartments 101-104 will be used for base voltage = 1.E4 -2 1.E3 GENEA FASEA1 { Boxes 90-97 will be used for base current = 1.E3 FIND { Search all extrema vectors to find the worst of some set of variables ALL { Rather than list variables separately, search a set of "all something" 6 { In response to "all what?" this answers: "all energy" outputs EXCLUDE { Repeat the search after excluding the worst shot, which is shot # 4 DISK /LIST { Create deterministic data file for last excluded shot, which is # 4 C Note about preceding. If no EXCLUDE, shot #4 still would be punched. If C EXCLUDE use, it is data for the worst excluded shot that will be created. BLANK card ending statistical tabulations BEGIN NEW DATA CASE BLANK