Full report

Original page 1

General Disclaimer One or more of the Following Statements may affect this Document This document has been reproduced from the best copy furnished by the organizational source. It is being released in the interest of making available as much information as possible. This document may contain data, which exceeds the sheet parameters. It was furnished in this condition by the organizational source and is the best copy available. This document may contain tone-on-tone or color graphs, charts and/or pictures, which have been reproduced in black and white. This document is paginated as submitted by the original source. Portions of this document are not fully legible due to the historical nature of some of the material. However, it is the best reproduction available from the original submission. Produced by the NASA Center for Aerospace Information (CASI)

Original page 2

NASA Technical Memorandum NASA TM-82532 Y THE VARIABLE POLARITY PLASMA ARC WELDING PROCESS: ITS APPLICATION TO THE SPACE SHUTTLE EXTERNAL TANK — FIRST INTERIM REPORT By A. C. Nunes, Jr., O. E. Bayless, Jr., C. S. Jones 11I, P. M. Munaro, A. P. Biddle, and W. A. Wilson Materials and Processes Laboratory June 1983 (NASA-'" y -82532) THE VARIABLE POLARITY N83-31021 ULASMA APC WELDING PROCESS: ITS APPLI^..drlUN TU THE 3PACF SHUTTLE EXTERNAL TANK IALt-IrA.M Report (NASA) 51 p HC A04 /iF A01 C3.,.. 1.3H Unclas (33/37 28632 4. i PJASA National Aeronautics and Space Administration George C. Marshall Space Flight Center MSFC Form 3190 IRev Mov 1983)

Original page 3

TECHNICA REPORT TAN A TITLE PAGE 1. REPORT NO. 2, GOVERNW NT ACCESSION NO. S, RECIPIENT'S CATALOG NO. NASA TM —82532 4, TITLE AND SUBTITLE 6, RE"T DATE The Variable Polarity Plasma Arc Welding Process: Its June 1983 Application to the Space Shuttle External Tank — First 6. PERFORMING ORGANIZATION CURE Interim Re ort 7, AUTHOR(S) A. C. Nunes, Jr., E. n. Bayless, Jr., C. S. ones P. M. Munafo A. P. Biddle and W. A. Wilson 9. PERFORMING ORGANIZATION NAME AND ADDRESS George C. Marshall Space Flight Center Marshall Space Flight Center, Alabama 35812 12, SPONSORING AGENCY NAME AND ADDRESS National Aeronautics and Sparc Administration Washington, D.C. 20546 15. SUPPLEMENTARY NOTES a.PERFORMI4G ORGANIZATION REPnRr M 10. WORK UNIT NO. 11. CONTRACT OR GRANT NO. 13, TYPE OF REPOR'. Q PFR100 COVERED Technical Memorandum 14, SPONSORING AGENCY CODE Prepared by Materials and Processes Laboratory, Science and Engineering Directorate, 16. ABSTRACT The Variable Polarity Plasma Arc (VPPA) welding process is being introduced as a partial replacement for the Gas Shielded Tungsten Arc process in assembly welding of the Space Shuttle External Tank. This sport describes the technical history of the development of the VPPA process, interim results c- the weld strength qualification studies, and plans for further work on the implementation of the VPPA process. 17 . KEN WORDS I i !i I i I i 16, DISTRIBUTION STATEMENT Unclassified - Unlimited 19, SECURITY CLASSIF. (of Ihlr -port) 20. SECURITY CLASSIF. (of thill 21. NO. OF PAVFS 22. PRK F Unclassified WF r' - Fo rm '1192 (M.r 1969) prga) (unclassified 50 NTIS For mlo br National Tlichnlcal Information Service, Springfield. Virgil"• 211 DI

Original page 4

ACKNOWLEDGMENTS Many people had significant roles in the development of Variable Polarity Plasma Are Welding for application to the Space Shuttle External Tank. Personnel in the Advanced Manufacturing Technology Laboratory, Materials Engineering, and Fabrication and Tank Production of Martin Marietta Corporation's Michoud Assembly Facility in New Orleans, Louisiana, who contributed significantly to the effort were D. Bolstad, C. Brosemer, F. Clover, W. Collins, D. Hartley, J. DeGeneres, T. Hibbard, R. Johnson, L. Johnston, W. McFarlane, W. McGee, S. Motoux, J. Ratcliffe, G. Scott, J. Sexton, C. Thornburg, D. Van Dyke, and all those who supported their efforts. Personnel in the Metals Processes and Metallurgy Research Branches at Marshall Space Flight Center, Huntsville, Alabama, who contributed significantly to the effort were S. Clark, C. Corum (retired), T. Craig, G. DeRamus, Jr., J. dodo, C. Lovoy (retired), N. Tiller, and all those who supported their efforts. In the course of so extensive an effort it is possible that a deserving contributor's name may have been overlooked. The acknowledgment listings must therefore not be taken as exclusive. To all contributors the authors offer their thanks.

Original page 5

TABLE OF CONTENTS Page INTRODUCTION: THE VARIABLE. POLARITY PLASMA ARC WELDING PROCESS .......... I heory of the Plasma Arc ..................................................... 1 Origin and Advantages of Plasma Arc Welding ...................................... 3 Advantages of Straight, Reverse, and Variable Polarity ............................... 3 Origin of Variable Polarity Plasma Arc Welding ..................................... 4 Origin of the Variable Polarity Plasma Arc Welding Program at Marshall Space Flight Center ....................................................... 4 MODIFICATIONS AND DFVFLOPMENT OF THE EQUIPMENT ........................... 5 Computer Controls .......................................................... 5 Power Supply and Automatic Voltage Control ..................................... I 1 The VPPA 'I'orch ............................................................ 14 MODIFICATIONS AND 1)1?Vt_LOPMI?NT OF THE PROCESS .............................. 14 Effect of Weld Parameters on Weld Strength ....................................... 14 Weld Gas Purity Requirements ................................................. 22 IMPLEMENTATION OF TIIF VARIABLE POLARITY PLASMA ARC PROCESS AS A PRODUCTION PR0CFSS ............................................. 24 Phan is Preliminary Study ................................................... 24 Phase 2: Equipment Development and Design Allowables ............................ 25 Phase 3: Demonstration with Production Tooling and Welder Training Program ........... 26 Phase 4: The Pathfinder Project ................................................ 27 MFCHANICAL PROPERTIES OF VPPA WELDMENTS -- AN INTERIM EVALUATION ......... 27 Test Program Design and Data Analysis Groundrules ................................ _'7 ix Results CONCLUSIONS .................................................................. 'K REFt:R} N('I:S ................................................................... 43 iii

Original page 6

LIST OF ILLUSTRATIONS Figure Title 1. Current flow in the VPPA torch, Pilot are voltage and current are approximately 30 V and 20 to 30 A. respectively (0.4 to O.tN M . ................................. ' The MSFC VPPA welding system schematic diagram.. ............... . 6 3. The MSFC VPPA welding system - general view ............................... 7 4. The MSFC VPPA welding system -- torch, cold wire, feeder, and hot block mounted on the manipulator beans .......................................... 8 I S. The MSFC VPPA welding system — computer, weld control panel, power supply and programmer, and plasma control cabinet with nozzle coolant radiator below....... 9 1h. The MSFC VPPA welding system — gas flow schematic diagram .................... 10 7. Simplified schematic of Hobart VP-300-5 VPPA Power Supply .................... I= 8, A typical VPPA wavolorm. Nominal straight polarity settings are :5 A and 30.0 V. The eyrie is divided into IQ msec straight polarity and 4 msec reverse polarity......... 13 u. VPPA welding torch assembly ............................................. 15 10. VPPA .-clding torch - view of disamwinblecl torch parts .......................... 16 II. Used VPPA torch electrodes. `The are emerged from the left end. Upper electrode shows oxygen staining. Both, but particularly lower. show ball pickup .............. 17 12. Tip of used tungsten electrode showing ball pickup .. I .......................... 18 13. MSFC rotating mount for the VPPA welding torch .............................. ly 14. VPPA torch assembly tools ................................................ .0 iv

Original page 7

LIST OF TABLES Title Pap able 1.Weld Parameters for Comparative VPPA and GTA Welds in 3/8-in. 2219 Aluminum Alloy................................................................. 21 Ultimate Tensile Strength (ksi) of Comparative VPPA and GTA Welds ............... 22 3. Design Allowable Program Features ......................................... 25 4. 5015 Fixture Demonstration Welds .......................................... 26 M 5, Minimum Required Number of Tensile Tests .................................. 29 6. Minimum Required Number of Tensile Tests for Mismatch ........................ 30 7. Minimum Required Number of Tensile Tests for Peaking ......................... 31 8. Minimum Required Number of Tensile Tests for Peaking Removal .................. 31 9. Minimum Required Number of Tensile Tests for Mismatch Removal ................ 32 10. Minimum Required Number of Fracture and Cyclic Tests ......................... 32 11. 0.125-in. Single Pass Test Data . ........................................... 33 12. 0.1.50-in. Single Pass Test Data ............................................. 35 13. 0.375-in. Single Pass Test Data ............................................. 37 14. 0.375-in. Two Pass Test Data .............................................. 39 15. 0.500-in. Two Pass Test Data .............................................. 41 V

Original page 8

U nl. PAGE M TECHNICAL MEMORANDUM OF POOR QUALITY THE VARIABLE POLARITY PLASMA ARC WELDING PROCESS: ITS APPLICATION TO THE SPACE SHUTTLE EXTERNAL TANK — FIRST INTERIM REPORT INTRODUCTION: THE VARIABLE POLARITY PLASMA ARC WELDING PROCESS This report documents the development of a variable polarity plasma arc (VPPA) welding system and its application to the fabrication of the Space Shuttle external tank (ET) as carried out at NASA's Marshall Space Flight Center and at Martin Marietta Corporation's Michoud Facility. Theory of the Plasma Are The plasma arc is characterized by a jet which projects the hot plasma forcefully against the work piece. Unconstricted arcs used in gas shielded tungsten arc welding (GTAW) processes also exhibit jetting [ I ] caused by magnetic pumping of the arc plasma. In an unconstricted arc the current density ?flowing down the arc interacts with the magnetic induction it set up around the current flow to pro- I + duce an inward directed force J x B per unit volume of arc material. For an unconstricted arc column of radius ra with a uniformly distributed current I (J = I/wra2) a pressure P varying with the radius r according to the relation r2 P = 4g IJ I - z ra (1) is computed where μ is the magnetic permeability of the gases entering the arc. Assuming a permeability about the same as that for free space (p/41r — 10-2 dynes/amp2), the maximum pressure p max (at r = 0) in psi is given by I22 Amax = 7.16 x 10-9 ra (2) where I and ra are given in amperes and inches respectively. The maximum pressure computed for a quarter inch diameter (ra = 1/8 in.), 300 amp arc is 0.04 psi. As the arc expands away from the electrode spot, r a increases and pmax drops. Assuming a conversion of pressure to velocity in accord with Bernoulli's equation, the maximum arc velocity Vmax may be estimated from the relation 2 P Vmax 2 = Pmax (3)

Original page 9

ORIGIN14L E yA19 OF POOR QUAU r where p is the density of the hot arc plasma gas. Although Pmax is small, so is the density of the hot gas, and velocities on the order of 100 m/sec are predicted from equation (2). 0.04 psi i@ equivalent to about an inch of water pressure. The plasma arc jet as shown in Figure 1 incorporates a nozzle which prevents sidewards flow of the arc gases. Thus, when the plasma gas enters the arc plasma at the rear of the nozzle the temperature shoots up and the density drops precipitously. Conservation of matter requires that the exit velocity be high enough to compensate for the reduced density of the gas at the nozzle exit; i.e., pAV is constant along the nozzle duct, where p is the gas density, A is the duct area, and V is the gas velocity. Conservation of momentum for a frictionless, constant area duct requires that P + pV 2 , where P, p, and V are respectively the local pressure, density, and velocity of the duct gas, is constant along the duct. Approximating the plasma nozzle by a frictionless, constant area duct, where the plasma gas supply pressure is AP and where the inlet velocity may be neglected with respect to the exit jet velocity Vj , then p V12 = AP (4) A plasma gas supply pressure of perhaps 25 in, of water would not be unreasonable for the VPPA equipment discussed herein. Plasma arc jet velocities are calculated on the order of 400 m/sec or about four times greater than the maximum velocities expected for the GTAW process. These velocities fall within appropriate ranges cited by Metcalfe and Quigley [ 21: 80 to 150 m/sec for GTAW and 300 to 2000 m/sec for plasma arc welding (PAW). ELECTRODE CONSTRICTING \ N0771F SHIELDING GAS NOZZLE TIORKPIFCE^ i PIASNAARC CURRENT i p = PILOT ARC CURRENT -1 • STRAIGHT (-) OR 0OR REVERSE (s) POLARITY ^ U U ip FOT ARC 4A f^ i Figure 1. Current flow in the VPPA torch. Pilot arc voltage and current are approximately 20 V and 20 to 30 A, respectively (0.4 to 0.6 kW). Bernoulli's equation estimates stagnation pressures of iess than 1 in. of water for the GTAW process example given here and about 12 in. of water for the PAW example. These pressures are equivalent to 0.4 and 4 in. of molten aluminuni. Ignoring surface tension and evaporation reaction effects, the GTAW arc should not be able to depress the surface of a weld puddle more than 0.4 in., while the plasma arc jet might depress the surface up to 4 in. deep. In fact, although the velocity fields in the

Original page 10

GTAW arc can be somewhat enhanced [31, the penetration is not necessarily correspondingly enhanced [41. Neither are significant surface depressions generally seen in aluminum weld puddles under the GTAW torch. On the other hand, PAW often operates in the "keyholing" mode, where the plasma penetrates the entire workpiece; the hole through the workpiece opens in front of the plasma jet and seals up behind it to leave a deep, relatively narrow weld bead. Therein lies one principal advantage of the PAW process: its capacity for doing in a single pass what would require multiple GTAW passes. When operated at still higher pressures, e.g., in the neighborhood of 200 in. of water, a flus ►ang away of the liquid accompanies a powerful cutting action. An open keyhole allows escape of gaseous contaminants. Relative to GTAW, shorter paths for escape of such contaminants through thinner layers of molten metal lining the keyhole may also be significant. The plasma jet has been observed to blow out solid contaminants upon some occasions. The greater tolerance for contamination exhibited by the VPPA process eliminates the need for the scraping procedure and for white glove handling of the aluminum required by the GTAW process. The greater tolerance for contamination is not unlimited, however, and gross contamination would be expected to cause a defect, even for a VPPA weld. The keyhole mode of welding also appears to be associated with reduced peaking distortion even when the geometry is about the same as the multipass GTAW standard for comparison. This is to be expected from the more symmetrical solidification pattern undergone by the PAW keyhole pass. Origin and Advantages of Plasma Arc Welding The plasma torch was introduced by Linde Air Products in 1955 as a cutting, not a welding device 15 1. By 1965 Linde had set up an automatic plasma are welding facility for Westinghouse Electric Corp. Westinghouse used the facility to fabricate 120-in. diameter, 3/8-in. thick D6AC steel rocket cases for the Titan I11-C booster assembly [6,71. The PAW process halved the weld time while retaining properties equivalent to those obtained by the GTAW process. The economy in weld time was effected by replacing the root pass plus three or four fill passes required for GTAW by a root pass plus a single cover pass. In 1965 Thermal Dynamics Corp. also reported the use of reversed polarity PAW to join 1/4-in, thick aluminum plate [81. Advantages of Straight, Reverse, and Variable Polarity The significant current carriers in a welding arc are electrons and positive ions. Electrons carry the bulk of the current, moving rapidly from negative cathode to positive anode. The positive ions drift more slowly through the intereiectrode space. The differential drift rate results in an unsymmetrical heating at the ends of the welding arc in a fixed polarity arrangement. The cathode receives less heat and the anode more heat. The "straight polarity" mode of operation entails a negative electrode (catho.!e) and a positive workpiece (anode). Where the primary object of the weld process is to deliver the maximum heat to the workpiece with minimal deterioration of the electrode, straight polarity is used. "Reverse polarity," i.e., positive electrode and negative workpiece, has the zdvantage that the workpiece is subjected to a cleaning process, "cathodic cleaning," by the impingement of heavy (compared to electrons) positive ions on the workpiece surface. In the case of PAW reverse polarity action appears to condition the surface of the aluminum alloy so that the molten metal flows easily and controllably under the arc. It is conjectured that this fluid control is accomplished through breaking up of

Original page 11

surface oxide films. Cutoff of reverse polarity during VPPA welding transforms i, weld metal flow which closes smoothly and soundly behind the Keyhole to an irregular, intermittent, globular flow leaving a rough, lumpy bead pushed down below the parent metal surface and protruding jaggedly from the root of the weld. But continuous reverse polarity is not necessary to provide adequate cathodic cleaning action. Alternating current GTAW has been recommended for welding aluminum alloys since the late 1940's [9). Studies of desirable GTAW current waveforms were underway by the end of the 1950's. A 1967 account of Sciaky Brothers' 300 A model SW-3 GTAW power supply [ 101 describes a variable polarity square-wave output. A 10 cycle (167 msec) straight polarity square pulse followed by a one or two cycle (17 or 33 msec) reverse polarity square pulse at a current setting determined by heat requirements was considered optimal. The variable polarity square-wave with unequal straight and reverse polarity times offered a combination of the high heating capability of straight polarity with the cleaning feature of reverse polarity. Adequate cleaning is obtained by incorporating a relatively short (one tenth to one fifth the duration of the straight polarity current) pulse of reverse polarity current into the welding current waveform. Origin of Variable Polarity Plasma Arc Welding By the end of the 1960's B. VanCleave at Boeing Co. was beginning his efforts to combine the variable polarity feature with Keyholing plasma arc welding for joining aluminum [ 11,121. The power supply proved to be a critical item. The first attempts with a pulsed power supply failed because the machine did not supply adequate plasma voltage. In 1972 a Sciaky square-wave generator demonstrated the value of the concept without being able to produce the desired welds. In 1974 Hobart Brothers built, under contract to Boeing, a VPPA power supply that was to prove satisfactory. The currently used Hobart cycle, 19 msec straight polarity followed by 4 msec reverse polarity, keeps about the same straight time to reverse time ratio, but the cycle time is speeded up by an order of magnitude over that of the Sciaky machine. The machine was delivered in 1975. The period from 1975 to 1979, when VPPA welding was incorporated into the Army Roland Missile Production Program, was one of development and 0imination of processing problems, such as a pressure pulsation, for example, that blew out the weld puddle. Origin of the Variable Polarity Plasma Arc Welding Program at Marshall Space Flight Center In view of Boeing's promising work, in 1978 the decision was made to acquire VPPA equipment at Marshall Space Flight Center for evaluation as a potential improvement on the GTA welding system then in use to fabricate the aluminum external tank (ET) of the Space Shuttle. MSFC's first VPPA system was received in January 1979. It centered around a Hobart 400 A plasma torch, mounted on a Progressive Welder and Machine Co. (Pontiac, Michigan) manipulator. Power was supplied by a Hobart Alternate Polarity Cyber-Tig II power supply with an 800 High Pulse Series Programmer. Coolant and torch gas flow were controlled by an HPW-400 Plasma Control Console. Torch to work distance was controlled manually. After a 6-month study it was concluued that: 1) VPPA welds in aluminum were essentially porosity fice as noted by radiography. This characteristic would reduce weld repair costs appreciably with respect to GTA welds. 4

Original page 12

1. The cleaning of faying surfaces by scraping and the draw filing of adjacent surfaces of the joint as required for straight polarity GTA welding could be eliminated. This would reduce joint preparation costs appreciably with respect to GTA welds. 3) A significant reduction in peaking was noticed. This would reduce depeaking costs. Depeaking, usually by magnetic hammer, is required on GTA welds where peaking, measured as the dihedral angle across the weld, exceeds 6.5 deg. These conclusions were so encouraging that the derision was made to develop a production VPPA system to partially replace the GTAW system now in use to fabricate the Space Shuttle ET, with its 36,000 in, of welds over thicknesses ranging from 0.140 to 1.040 in. MODIFICATIONS AND DEVELOPMENT OF THE EQUIPMENT Computer Controls GTAW equipment used to join ET panels incorporates a computer control to manipulate all the welding parameters that have to be varied while making a tapered weld 1131. Because the wall thickness of the tank shells varies (e.g., the liquid oxygen tank nose section varies in thickness from 0.367 in. to 0.140 in.) tapered longitudinal welds are frequently encountered. A computer to control welding parameters would be required for VPPA welding for the same reason; i.e., tapered welds require too high a level of skill and attention to be practical as manual welds in a production environment. Therefore, in October 1979 a request for proposal to modify the then existing VPPA system by the addition of a digital program controller was issued and in March 1980 Hobart Brothers, Inc., entered into a contract to carry out the task. In December 1980 a computer controlled VPPA system was delivered to Marshall Space Flight Center. A second computerized VPPA welding system was delivered to the ET fabrication plant of Martin-Marietta at Michoud, Louisiana, in March 1981. Schematics and photographs of the MSFC VPPA welding system are shown in Figures 2 through 6. The heart of the computerized VPPA welding system is the PAL-100 computer, operating with a Digital Equipment Corp. (DEC) LSI-11/23 microprocessor. The PAL-100 computer sets weld parameters in accordance with set point inputs by the welder operator. The computer can vary the parameters to produce tapered welds. A recent modification of the system closing the control loop on the weld parameters now permits the system to continuously force measured parameters to coincide with desired set points. This has decreased the sensitivity of the system to variations in calibration. The operator of the weld controller inputs a series of choices offered by a menu format software system communicating through a Keyboard-color CRT display. Hard copy can be obtained from an attached printer. Dual floppy disc memory storage is available in addition to the internal 32K memory. The welding system can be operated in the fully manual mode without the computer. Partial override control of the system by the operator through a remote pendant may also be selected. The control system was set up by General Digital Industries, Inc., of Huntsville, Alabama, in the capacity of subcontractor to Hobart Brothers, Inc. The generic name, Process Automation Language, or PAL, of the PAL- 100 computer is applied by GDI to their line of process automation and control computers.

Original page 13

The entire VPPA system including the manipulator and cold wir t fader as wr' as the torch control and power supply, is subject to computer control. Digital dsplays are used throughout thei system. Controls and displays are accurate within i 1 percent. initial experience with this system has been good. The system has functioned according to design and has been convenient to use. One of the floppy discs has been dedicated as a system memory, the other for weld programs. Thus far about a dozen weld programs have been recorded; there is no limit to the size of the weld program library which might be recorded. At the time of th't writing, it is planned to increase memory capacity through replacement of the floppy disc memory with a hard disc memory. OR1GINAt PAGI+ IS OF POOR QUALITY C!QAOTK N10111M1Tr AWC WIVE M UNI T •^.^ PLASMA CTUAWIe Ca0 1118 ' ow n&n& o • urium ► MINI" 11COMM )4110wl«rT MELD CONTWX S IGNALS CONSOLE TS t^h^t ^ t► •see•s m wigs tie O i IfAT.......... . ,!LOAM On BLOCKTp11CMHOT...... ON" an wetn«s^eeom rs/t Otwm NOaAIIT MELD COiTADL PANEL F w^ MELD LAT0E GENERAL r DIGITAL T COMPUTER lK Figure 2. The M SFC VPPA welding system — schematic diagram. b

Original page 14

I ^ ^I Coro. it I ORIGINAL PAGE f^ OF P JOp, QUALITY I r #s ilm . t r r f w . + 3 Fa m l qj '4 3 ^ Q V) z v M iM t ^ r t t^'^i^ ^ jh c Cili tt- L

Original page 15

Qkctitlyt-l ► ^^r.^e_ ^J OF POOR QUALITY ^.N ^rzJ ^ X1111 .- \ 'a wq 111 i 1 ^♦ ^— ]Pik Lbow, 8 .,r o M G 0 0 0 O • I.w 3 7 O v `o E y S '^ y 3 Q a. a. Ir lam, Ea r. k^ll

Original page 16

ORIGINAL PA GE 15 OF POOR QJALITY G E E m iz 0 a c b C. a al al 3 0 3 a o ^ a ., .o c :^ o ^ ^O ojj OU -^ y :1 N 3 N 0 c .v. t E ^, O a I ^ i (E E U O V ^n C O C U J ^ 3 E Q °' a a U k N 5 ^J H 4J L Vr 9

Original page 17

1 g 1 1 w a ^ z J ^-y ^ r <r. j 0d Z p v 1 d Q Q 1 aJ ORIGINAL PAGE IS OF POOR QUALITY ^ , 1 ^ cd 1 of i # a9b, 00 b 3 ---- -------------------- a -- U 0 ^_ W Y W y^ r li y ^i 10 w N o U O cc ^y F w

Original page 18

Power Supply and Automatic Voitapa Control Power supply to the VPPA torch is provided by the Hobart VP-300-S Variable Polarity Welding Machine. A simplified schematic of the machine is shown in Figure 7. The machine can deliver: o 3 to 299 A straight polarity o 3 to 398 A reverse polarity (using the above plus a S to 99 A supplementary power supply) o An alternate (variable) polarity waveform incorporating 5 to 99.9 msec intervals of straight polarity current alternating with 1 to 99.9 msec intervals of reverse polarity. The current magnitude is controlled by the Cyber-Tig 11 Series 800 Programmer. An Alternate Polarity Programmer incorponated into the VP-300-S Welding Machine controls the straight and reverse polarity interval lengths. Switching is carried out by SCR's. A typical VPPA waveform is shown in Figure 8. The machine runs on 3-phase 480 V power. Some 3-phase ripple can be seen superimposed on the current waveform. The pilot arc current, superimposed upon the main arc current between the electrode and the plasma nozzle, is supplied by a separate power supply located in the plasma control console. The pilot arc polarity does not alternate. The nozzle is held positive with respect to the electrode so that the electron flow is always away from the electrode toward the nozzle. The pilot arc power supply inputs 480 Vac current and outputs 4 to 40 A current at a working voltage in the neighborhood of 20 V. The open circuit voltage rises to 100 V. The main arc is started by application of a high frequency alternating voltage between the electrode and the plasma nozzle so as to first establish the pilot arc from which the main arc proceeds. The high frequency current is controlled by the plasma control console, but is generated by a spark gap oscillator located in the hot block. It has been observed that once the main arc is started it is possible to turn off the pilot arc with little or no effect, at least in some situations. Arc Voltage Control (AVC) circuitry located in the weld control cabinet can be used at the option of the welder operator to hold the arc voltage constant by moving the torch towards and away from the workpiece as required. The AVC takes advantage of the relation between arc length and arc voltage. The purpose of AVC is to prevent weld power variations. Weld power is given by the product of amps, already held constant by the power supply, and volts, held constant by the AVC. Variations in power produce variations in weld consistency. Although for VPPA operation the AVC senses the voltage behind the inverter in the power supply so that the sensed polarity does not change, the changes in voltage magnitude under VPPA conditions were found to cause undesirable weld head motions which left behind a roughened weld surface. This problem was remedied by a cooperative effort between Hobart Brothers, Inc., the Michoud Assembly Facility of Martin Marietta Corp., and MSFC. A circuit was incorporated into the AVC system causing it to ignore the voltage fluctuations produced during the reverse polarity part of the waveform and thus to control the torch from the straight polarity voltage only. With the latter modification the AVC is now operating satisfactorily both in the Keyhole and non-Keyhole modes. 11

Original page 19

ORIGINAL PAGE 19 POOR QUALITY OF ....................................................... e^ 7 -7 i = ^I - - - - - - - - - - - r. {I! CI O Fr O 9p ly . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Original page 20

300 200 100 cc 0a c^ 0 N N -100 U -200 -3m CHICINAL PAGE IS OF POOR QUALITY i o`c g o^ N N J J t h^ 0 CL 0 10 ,'0 30 40 ')0 I I MI (M 11 1 IS[ PONDS ) Figure 8. A typical VITA wa%elorm, Nominal %traipht polarit% .rtlillg% arc "; A and 30.0 V. fh,• odt. i, ili%ulrtt into, I" . •, c'i Nliaight pohtit% and 4 instc it-wi%c la,larit%. 13

Original page 21

ORIGINAL PAGE IS The VPPA Torch OF POOR QUALITY An assembly drawing of the VPPA welding torch is shown in Figure 9 and a photograph of the disassembled parts in Figure 10. The electrode is held in a collet assembly. The coolant passages are sealed by O-rings at part seam interfaces. Coolant leakage into the gas passages is detrimental even at low levels. In the arc, water decomposes to yield oxygen and hydrogen. Oxygen will form a visible film on the electrode as shown in Figure 11. Hydrogen will enter the weld and cause porosity. The torch initially provided by Hobart leaked coolant into the gas passages occasionally. This happened when electrode alignment by means of the three setscrews, then provided for the purpose, disturbed the seating of the O-rings sealing the coolant passages, or when the O-rings were damaged during torch assembly or welding use. Alignment of the electrode is critical. Even if the electrode is well enough aligned to prevent "double arcing," where the main arc proceeds from electrode to nozzle and from thence to the workpiece instead of directly from electrode to workpiece causing the nozzle to melt, it is difficult to prevent a certain amount of assymetry in the arc. With use, the tungsten electrode tip may pick up small balls of melted tungsten, which also causes an asymmetrical arc. An example of tungsten ball pickup is shown in Figure 11 and (enlarged) Figure 12. It is preferred to have the widest part of an asymmetrical arc extended along the weld rather than transverse to it. Therefore at MSFC a rotating torch mount was installed, which permitted rotation of the VPPA torch about its axis so as to align arc pattern asymmetries and thus optimize weld quality. A photograph of the torch rotating mount is shown in Figure 13. The torch orifice geometry is extremely critical for torch performance. Unless gas orifice dimenc'ons, electrode to orifice spacing, and orifice cup to shield cup dimensions are fixed precisely, torch welding characteristics will not be consistent. A special set of tools was developed at MSFC to make it easier for the operator to control the latter dimensions as well as to insert and remove O-rings during torch assembly. These tools are illustrated in Figure 14. A description of the use of tools is not included in this report. Although the torch described has performed well, a further development effort has been funded by MSFC to build a hermetically sealed torch with a self-centering electrode that would not demand the careful and time-consuming procedures required to assemble the present torch. It is presently being evaluated and will be described in a subsequent report. MODIFICATIONS AND DEVELOPMENT OF THE PROCESS Effect of Weld Parameters on Weld Strength In order to assess the effect of the VPPA welding process in comparison with the GTA welding process closely similar VPPA and GTA butt welds were made in 3/8-in. thick 2219-T87 aluminum plate. Both welds were vertical, single-pass welds. Surface preparation necessarily differed to accommodate the GTA process. The square butt edges were left as machined for the VPPA process, but were scraped for the GTAW process. In both cases the surfaces were wiped with solvent to remove grease before welding. 14

Original page 22

r Z C7 Z ¢ W H Z ¢ W V O Q J N Z W J J O U W O O Q r U W > J W Q Y U Q 2 U Q J a U ¢ O H a O O W CO U_ w Q O ORIGINAL PAGE M OF POOR QUALITY W u W O Z_ w H Z W U E W u O O c¢ r. r- u V W O J W OU Y ^ U ^ 6J O 22 3 u ^ Q 0. J W J Q N U o^ N CO LL i Q Q a S u O P.- U W O J J N w N - 01 Z 15

Original page 23

ORIGINAL PAGE OF POOR QQUALITY 0 a 0 0 ^v v E 0 o^ 0 0 3 I u O .r oc b v 3 Q n. a. c w 16

Original page 24

ORIGINAL PAGE: 13 OF POOR QUALITY 3 0 0 u v I.n a a, x u c .^ J ;,O 3 ^, o L '_ G y 0 3 w- o y T ^V ^V E J V U F. f9 ^ a L 7 L H ♦+ o o x v ob u c V r- > o v 3 04 L1. 17

Original page 25

I ORIGINAL PAGE 1 ► n t t'rY OF POOR Qt

Original page 26

ORIGINAL PALL IS u O a, C 7 v 3 Q 0. 0. v 4.. O w. C 7 O E u, c ro O V LL i M 3 OJ! LL 19

Original page 27

ORIGINAL PAGE IS OF POOR QUALITY i^ 0 0 21 "J :J L O Q 0. y G4 LL

Original page 28

0' ,,- _ a -.a ,^^ QU Weld parameters are listed in Table 1. 1/16-in. diameter 2319 aluminum filler %% iM was fed into the weld to control weld surface contours. TABLE 1. WELD PARAMETERS FOR COMPARATIVE VPPA AND GTA WELDS IN 3/8-in. 2219 ALUMINUM ALLOY VPPA GTA Current: Straight Polarity ( 19 msec) 1-10 A 195 A Reverse Polarity (4 msec) ^200 A Potential: Straight Polarity ( 19 msec) 12.5 V 32.5 V Reverse Polarity (4 msec) —40 V Travel Speed: 8 in./min 6.75 in./min Wire Feed: 38 in./min 35 in./min Shield Gas: 35 CFH-He 100 CFH-He Plasma Gas: 7.5 CFH-Ar — At 35 in. water pressure A VPPA/GTAW pair of as-welded plates was shaved to remove the root and crown reinforcements and another VPPA/GTAW pair left unshaved. Fou- 1-in. wide transverse tensile specimens were machined from each plate for a total of 4 x 4 or 16 test specimens. It is not necessary to reduce the area of tht test specimen at the center ( "dogboning") to control the break site because the weld fusion zone is sufficiently :it ft to localize the fracture there without reducing the specimen section. The crown and root widths of the VPPA and GTA welds were approximately the same (about 7/16-in. and 1/4-in., respectively) indicating about the same heat input to the weld metal in spite of the higher heat input to the VPPA welds 1 141. The VPPA process generated approximately 50 kJ /in. of weld; the GTAW process 23 Win. However, nozzle coolant losses of about one-fourth of the VPPA power hive been measured. and total VPPA power loss (including the nozzle coolant loss) estimates based on thermocouple measurements suggest that another quarter of the power is lost in the keyholing plasma jet that passes out through the backside of the workpiece. Thus, after losses. the VPPA process delivers about the same amount of power to the workpiece as the GTAW process for the :. ituation in question. The ultimate tensile strengths of the samples are shown in Table 2. A statistical analysis (F-test) of this data shows to a 95 percent confidence level that: l) There is no difference in strength between VPPA and GTA welds in the as-welded condition. 2) The VPPA welds are somewhat stronger than the GTA welds when the weld bead surfaces are shaved first. 21

Original page 29

PAGE Is ORIeaqAR Q^ I.ITYOF !POOR TABLE 2. ULTIMATE TENSILE STRENGTH(KSI) OF COMPARATIVE VPPA AND GTA WELDS As-Welded: Mean Std. Dev. Shaved: Mean Std. Dev. VPPA GTA 40877 39993 40885 40212 40481 40125 40141 40579 40596 40227 309 218 38042 37023 38104 37232 38020 36953 37819 36762 37996 36993 107 168 Strengthening effects have been observed in multiple pass welds, both GTA and VPPA. Research into the mechanism by which the multiple pass strengthening effect operates is underway at Marshall Space Flight Center and Martin Marietta Corporation 's Michoud Assembly Facility and will be addressed in a subsequent report. Weld Gas Purity Requirements Gas purity is an essential factor in achieving quality welds in any shielded arc process. The problem of impurities, when manifest, is normally resolved by eliminating the source of the contamination. However, if the gas supply itself is contaminated, serious schedule interruptions can result, particularly where large, manifolded distribution systems are employed. During a period of peak activity in the VPPA program, a means of accommodating transient periods of minor source gas contamination was developed. Centorr Model 2G-100-SS gettering furnaces were installed between the argon and helium supplies and the plasma torch as shown in Figure 6. These furnaces expose the weld gases to titanium at 800°C, which reacts with and removes, or "getters" impurities. Gab contamination affects the VPPA process in several ways. Inside the torch, oxygen from water vapor decomposed in the arc can produce a visible oxide film stain on the tungsten electrode as shown in Figure 11. The plasma jet may sputter and exhibit a green coloration from copper eroded from the nozzle under erratic flow conditions. The weld itself may exhibit a peppery looking weld surface, presumed to be fine porosity due to hydrogen gas from decomposed water vapor. Under more severe conditions the torch may leave behind it a train of irregularities and holes in the weld bead. A rough estimate of the quantity of water vapor tolerable in weld gas can be made on the basis of thF following assumptions:

Original page 30

OF POOR QWI^ LITY 1) A hydrogen content in excess of about 0.1 cm 3 per 100 grams of metal results in porosity. This value comes from sand castings (151 and could be somewhat higher for a weld fusion zone, if cooling is rapid enough to hinder the nucleation of bubbles. 2) All hydrogen in the weld gas dissolves in the weld metal. 3) All water vapor in the weld gas is broken into gaseous hydrogen and oxygen in the welding •irc. If the weld gas volumetric flow rate is Q and the weld gas density is p g, then the mass flow rate gQ. If the weight fraction of water vapor is C, the density of water vapor is p H20, and of weld gas is p the density of hydrogen is pH.,, then the volume flow rate of hydrogen is 1 PH2O PKQ P C H2O PH, or Pg c Q , PH, assuming the amount of water vapor contamination is small. The mass of metal soaking up this hydrogen per unit time is p M AV where pM is the density of the metal, A is the cross section of the weld, and V is the weld speed. The critical concentration of water vapor can be estimated by equating the volume now CCRIT rate of hydrogen divided by the inass flow rate of metal receiving it to the hydrogen content that would cause porosity, i.e., 0.1 cm 31100 grams of metal. It' the hydrogen and weld gas are measured at standard temperature and pressure, and if the weld metal is aluminum, then for argon, the plasma gas, AV C'Ck . 5 (5) and for helium, the shield gas, (6) CCRIT "' S0 AV 23

Original page 31

where ORIGINAL PAGE PS ('GRIT = critical water impurity level in ppm OF POOR QUALITY A = weld cross sectional area in in.' V = weld speed in in,/min tr1 = weld gas flow rate in ft;/hr. Representative values A = 0. IS in. 2 , V = 8 in./min, and Q - b ft ;/hr for the argon plasma gas, S() ft ;/hr for the helium shield gas yield critical contamination levels of I ppm for both argon and helium. It' only a tenth of the hydrogen produced enters the weld, the rest escaping, then about 10 ppm water vapor contamination could he acceptable. At these critical concentration levels problems with gas purity are indeed to be anticipated. Hydrogen enters the weld puddle at the surface, at different rates depending upon the temperature. At the puddle edge, at the melting temperature of the metal, (pure) aluminum call 0.69 cc of hydrogen gas for every 100 got of metal 1101. Closer to the spot where the an impinges oil puddle surface, more hydrogen, up to 20.1) cc per gni of m 1171 ca be dissolved. Circulation ofcall call molten metal within e weld puddle portions of the puddle containing very high quantities of hydrogen into the rest of the metal. This condition call to higher than equilibrium hydrogen concentrations in some parts of the weld 118,191. A high level of hydrogen supersaturation call severe porosity. Conversely. if the hydrogen content in the weld gases call he kept to sufficiently low levels, reiectable porosity would not normally be expected. IMPLEMENTATION OF THE VARIABLE POLARITY PLASMA ARC PROCESS AS A PRODUCTION PROCESS Phase 1: Preliminary Study l'he steps by which the VPPA process will partially replace the GTA process on the external tank were betiun in January 1974 with a preliminary study carried out oil simplified welding system. The system centered around a 400 A plasma torch powered by a Hobart Alternate Polarity Cyber-TIG 11 all Power supply with high Pulse Series Programmer. lissentially a hand welding system, it was modified by attaching the torch to a power manipulator, but full automation was lacking. No automatic voltage control was Provided; the torch/work spacing was controlled by hand. It was established by the end of a six month study of the VPPA process that very significant cost reductions ought to be possible by replacing the GTA welding process currently used to assemble the Spare Shuttle external tank by the VPPA process. The cost reductions expected were to conic from the ease with which sound, porosity free welds are made with the VPPA process. Joint preparation could be simplified by elimination tit' scraping and filing procedures presently required for GTAW, and weld quality still remained so high as to greatly reduce repair costs when compared to GTA welding. Further, a ill and a consequent reduction in depending costs was indicated by the preliminary reduction study. 111timately, the exist significant rust savings may be derived by reducing the 100 percent X-ray inspection now inillosed and replacing it with a sampling plan to verify weld quality. 4

Original page 32

Pheee 2: Equipment Development end Design Albwebles Following up the promising results of the preliminary study, concurrent programs for equipment development and establishment of design allowable properties were instituted. A complete VPPAW system was set up at Marshall Space Flight Center beginning in late 1979. Over the ensuing time period improvements were introduced into the system as follows: 1) Computer controls 2) A modified automatic voltage control 3) A modified torch and torch assembly procedure 4) Getter furnaces to remove weld gas impurities. During this period studies of the heat flow in the VPPA welding process and of the differences between VPPA and GTA weldments were also carried out. The design allowable study carried out during this phase showed that the new welding process would produce acceptably strong weldments. If the VPPAW process would have required redesign of the external tank to reduce weld stress, the process would have been unacceptable and would have been abandoned at this point. Phase 2 came into full operation in December 1980 with the delivery of the computerized VPPA welding system to MSFC and was extended to the Michoud Assembly Facility of Martin Marietta Aerospace with the delivery there of a second computerized VPPAW system in March 1981. Table 3 describes the main features of the design allowables test program. TABLE 3. DESIGN ALLOWABLE PROGRAM FEATURES o Welding (Thicknesses = 0.125, 0.250, 0.375, 0.500, 0.650, 0.800, 1.00 in.) o Test specimens (both tensile and fracture toughness, including notching and cycling the latter) o Weld surface planishing (some specimens) o Mismatch and peaking removal (some specimens) o Mechanical testing (tensile and fracture toughness) • Analysis of test data • Endurance test of equipment (3-hr continuous operation) 25

Original page 33

The endurance test cited in Table 3 consisted in a continuous operation trial of the system for a period of 3 hr and 6 min at a current load of 220 A. To obtain a sufficiently long plate the weld was spiraled up the side of a 7 ft diameter cylinder. Measurements of temperatures during this test indicated that about one-quarter of the torch power was lost to the plasma gas nozzle, and suggested the desirability of increasing the surface area of tha heat exchanger which dissipates the nozzle coolant heat. Phase 3: Demonstration with Production Tooling and Welder Training Prcgram Upon receipt of favorable results from the A-allowable testing program, VPPAW equipment was demonstrated with actual production tooling and a training program for Michoud Assembly Facility welders and welding engineers was set up. The 5015 weld fixture, which holds the panels comprising a 15 ft long barrel section of the external tank, was selected for the demonstration. Assembly of the barrel from panels requires 10 vertical position, longitudinal welds as described in Table 4. The demonstration was carried out in July- August 1982 [ 201 . TABLE 4. 5015 FIXTURE DEMONSTRATION WELDS Number of Welds Thickness (in.) Number of Passes 4 0.320 2 1 1 0.320 0.320/0.550 Taper 2 2 1 0.320/0.650 Taper 3 1 0.320/0.650 Taper 2 The 5015 fixture differed from previously used experimental fixtures in two ways. The higher heat sink effect of the 5015 fixture required an extra 10 to 15 A weld current. The chimney effect in the 5015 weld channel required baffles at top and bottom to prevent disturbance of the shield gas by rising air currents. All the weld strengths developed on the 5015 fixture were comparable to corresponding GTA welds made on the same fixture and were judged successful. Fracture toughness testing is nearing completion and will be reported in the second interim document. During this phase, three welding engineers and seven welders from the Michoud Assembly Facility were trained in the theory and use of VPPA welding equipment in the welding research laboratory at Marshall Space Flight Center. The main emphasis in the training program, which covered theory, equipment, and program (A-allowables), was hands-on operation of the equipment. Welder trainees were asked to diagnose deliberately set equipment maladjustments and to restore correct operation. Length of training varied from one to four weeks, depending upon the release time permitted by the trainees' work schedules. Additional problem solving time for welding engineers was and is arranged when necessary in the context of general operations. 26

Original page 34

Phan 4: The Pathfinder Project Pathfinders are defined as subassemblies produced on a production fixture, to demonstrate the acceptability of a major tool, hardware, or process change. They normally become flight articles if, after inspection and associated witness panel testing, it can be demonstrated that all engineering requirements for the article have been met. In the external tank VPPA weld demonstration project, the mPjor weld tools will be divided into three groups based on the attitude of the weld torch and the relative motion between the torch and part. A VPPAW demonstration pathfinder will be cycled through one tool in each group with the results considered applicable to all other tools in the same group. The groups consist of: a) A fixed part with horizontal torch attitude and vertical torch movement. b) A part rotating about its horizontal axis with a fixed horizontal torch. c) A fixed part with a moving torch with attitude change from horizontal to vertical. The first introduction of variable polarity plasma are welding into external tank production will occur with Group (A) tools. The T04A5015 Vertical Barrell Assembly Fixture has been chosen for initial pathfinder demonstration. MECHANICAL PROPERTIES OF VPPA WELDMENTS — AN INTERIM EVALUATION Test Program Design and Date Analysis Groundrulas Tables S through 10 list the (minimum) numbers and types of tests planned for in the design allowables test program. The test program was designed in accordance with the following considerations: 1) Data analysis was to be based on the guidelines presented in MIL-HDBK-SC, Chapter 9. 2) One data set (0.25 in. thick, single pass, room temperature) would include 300 specimens and would be used to establish "normality." 3) Basic room temperature allowables for other thicknesses and numbers of weld passes would be based on data sets of 50 or mere specimens, with A-basis allowables calculated using one-sided -5C. Table 9.6.4.1. Tolerance Limit Factors from MIL-HDBK 4) Aloowables for combinations of mismatch and peaking, and for various weld repair procedures, were to be established by structural analysis methods ("BOSOR") using a maximum allowable fiber stress based on weld certification requirements. For purposes of comparison with GTAW requirements, and for establishing standards for the preparation of repair specimens, individual mismatch and peaking "allowables" for each thickness would be based on sets of 50 or more specimens tested at room temperature. Subsets of approximately 10 specimens of equal mismatch or peaking would be analyzed separately by MIL-HDBK-5C methods to establish "A-basis" allowables for each set. These "allowables" were to be plotted to establish the permissible mismatch and peaking values, based on a strength requirement of 31 ksi. (Note: Current external tank propellant tank weld thicknesses are based on a net 27

Original page 35

strength requirement of 31 ksi ultimate.) Individual mismatch, peaking, and repair "allowables," other than thou described above, were to be generated mainly for information only and would be based on sets of 20 or fewer specimens. Results The data presented in Tables I 1 through 15 for 0.125, 0.250, 0.375 (one and two pass) and 0.500 (two pass) in. thicknesses represent the compilation of mechanical property data available as of February 1983.The data are final; it is not planned to add to any of the sets presented here. For any given thickness a significant change in allowables is unlikely, but minor changes to accommodate grouping of sets into thickness ranges are possible. Therefore, the design allowables listed here should be considered interim design allowables. A-basis allowables, where appropriate, are shown in the right-hand columns. The allowable peaking and mismatch values described above have been indicated in parentheses on the +350°F, -320°F, and -4231' lines. In some cases the mismatch and peaking allowables have been conservatively adjusted to oil account fir scatter, based from other thicknesses and taking producibility into account. Based on the data developed to date, weld tool certification requirements will be 38 ksi minimum up to 0.40 in. thickness and 36 ksi minimum above 0.40 in. thickness. CONCLUSIONS The Variable Polarity Plasma Arc Welding system which: (1) reduces weld repair costs through virtual elimination of porosity; (2) reduces joint preparation costs through elimination of the need to scrape or draw file faying surfaces: and (3) reduces depeaking costs by reducing the amount of peaking to be eliminated. will allow, based on statistical analysis of data from 0.125, 0.250, 0.375, and 0.500 in. thick test welds, weld tool certification requirements of 38 ksi minimum for thicknesses up to 0.40 in. and 3(1 ksi minimum for thicker sections. Data for 0.650, 0.800, and 1.00 in. thicknesses ii in preparation and the stress levels cited are to be considered interim levels at present. I'quipment development stages have largely been completed and have been chronicled in this report Welder training programs have been implemented and demonstration welds have been made on a production fixture (5015). Tlie Pathfinder Project, in which a pilot production operation initiates final implementation of the VPPA process into the Space Shuttle External Tank assembly process, is scheduled to get under way as this document is completed. As a result of the successful process development program and initial implementation effort to date, plans are underway to incorporate the VPPA process on several other fixtures in a timely manner so as to increase the output potential of the major weld fixtures. In addition, the high quality of VPPA welds made to date merits strong consideration for reducing the 100 percent X-ray inspection now impowd oil welds of' the external tank. '8

Original page 36

ORIGINAL PAGE 18 ..r rnnnn ni 1 A 1 IT/ C E Q hh {/^Q hh hQhh ^hh I I I Q c c('^ .C^,,V cR^'1l Nh rte. ►^ ^„ ^hh h0hh hMV1 V9 O V1 V1 ^ M ^i h c71 to R c a a H Ohh hO hh hOh V1 Qhh ^C Q T ^a w a O a M M M M M M M M M M M M M ^ ^• .7 , z ^ M M M M M M M a z a V' v 3 3 Q, o •d o^ 0 0 0 "' E \o o^ erM^ Q M'^t'^! M^ M'ef Mef E M^^ '^!M std 000o g oo go 00 00 ^, o00 0000 a E w ^ .^ ^ ^ '••^ ^ ^ r-. /1 r••^ i••^ air 0000 $ oo go go go $oo 0000 w 79 ,'^., N N .^ N .^ N .••+ N 'd I; E a 0 8 00 0000 000 00 00 00 •° 000 M N V' 1 N N h N N ^/ Y^ V ^^ h^ h N N N N N F 3 N O O M O M E""NM MN E"NM 8 0 r O M O O M "p I ^p ^., w E"NN hE NN v1F"NN HNN (0 W n Ma engT d t^ vY ,.., h aen IT 8 a:.a ^ .92 3 O0 h v 4 Q u ee E V N M ^D 00 S 0 0 0 0 a:^ 0 0 h ^-+ N M ^ 29

Original page 37

o o ow) H 00 c o o o 0 o 0 0 a 0 V) x o 0 0 o O 0 w C w O H o ^ w 0 a z 0 0 0 w 0 0 0 0 H w C O C4 w C) ca Mo a 0 0 5 C z W O O O O O 0 O r-J a w x 0 o 0 z M O ^ G W ..1 CC O a O H C7 p O O di o 8 f J h O ^G C r l 00 `-' O O O O O O F^ 30 ORIGINAL PAGE 19 OF POOR QUALITY g c g 0 a o 0 ^ 0 0 O 0 ^^ 0 O O ^ O O ^-^ H

Original page 38

ORSG!NAL PAGE 18 OF POOR QUALITY TABLE 7. MINIMUM REQUIRED NUMBER OF TENSILE TESTS FOR PEAKING Thickness (in.) 20 0.125 10 0.250 10 0.375 10 0.500 10 0.650 — 0.800 — 1.00 — Totals 40 Degrees of Peaking 40 60 8° 10° 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 — 10 10 10 — 10 10 10 — 10 10 10 — 70 70 70 30 TABLE 8. MINIMUM REQUIRED NUMBER OF TENSILE TESTS FOR PEAKING REMOVAL Reworked Peaking Angle Cryogenic Tests* Thickness As Welded (in.) Peaking Angle 00 0.125 10° 10 0.250 10° 10 0.375 100 10 0.500 80 10 Totals 40 20 40 60 -320°F -423°F 10 10 10 5 5 10 10 10 5 5 10 10 10 5 5 10 10 10 5 5 40 40 40 20 20 *Note: The reworked peaking angle for the cryogenic tests will be determined based upon the results of the room temperature peaking tests. 31

Original page 39

As Welded Reworked Mismatch (in.) Cryogenic Tests* Thickness Mismatch (in.) (in.) 0.020 0.125 0.070 10 0.250 0.070 10 0.375 0.100 10 0.500 0.100 10 T(,i.(Is 40 0.044 0.060 -320°F -423°F 10 10 5 5 10 10 5 5 10 10 5 5 10 10 5 5 40 40 20 20 *NO'1'1-: The reworked mismatch for the cryogenic tests will be determined based upon the room temperature tests. TABI-F 10. MINIMUM REQUIRED NUMBER OF FRACTURE AND CYCLIC TESTS Thickness Temperature (°F) Fracture Cyclic 0.125 RT 0.250 RT -320 0.375 RT -320 -423 0.500 RT 0.800 RT 1.000 R'I' 8 -- 12 6 12 6 12 6 12 6 12 -- 8 5 -- 5 Grand Total = 110 32

Original page 40

v1 O ^ ^ N P r+ ^ v'1 v7 ^ i H Pr! 00 N ^.y .r et M go I I I t± n ^ ^ H 0 E^ ^ ^ to'E 0 N ^-' M I v I ^ i a v ... I a u V .tea ° d '^ z N M O - ORIGINAL PAGE IS nF POOR QUALITY O; NN d; op n o0 CJ; N '^' N Q^ ^ M M M f^f M N Q 00 00 ^1 OR M M tit N V1 N ^D N %0 00 g g M i I I I I I i O^ ... M C^ t+ O ^ ^D N O t v O H'f .^. N Y1 {/1 N N ^'... Q N v ^'.. err , , v , I I I a~ I h O^ CO M ^ ° ^ d '^ -- vi M v^ ^n v^ N M 4 t^ M en O O O ^O ,^O ,^O ^O ,.., ,^O ^O ,.,0 r +v+ ,^ p O C ^ C ^ O N Q ^ O O hO C ^ C ^ O N_ S O I I O I O I O I I O I O I O I g l O O o M o -' N O O O O v NO O O O S O O O^O C C C ..1 O ^ ^ ^ o O o yy Z en ^ E V'1 V't 1/1 1f1 C N N N tV 5 " C C O O C G O O O O O 0 0 o a o o a d a o^ d d ;' eti ^ A A e4 V1 V1 V1 Vi V1 V'i N N N N N N N O O O C C O O 33

Original page 41

^a oo c^ a 06 N M M Q y v 00 N 00 r- Ca M o^ v 1 .c ^ 'U -^ O ... t+'1 'L O^ vi p M M N ^D ^' d d d M M O^ O d d d d ^ ^ I I I oG ^- M o o et d M M G M N 4 00 "S7 v O b ^ i I et I °^ I 00 I o f ORIGINAL PAGE 18 (W POOR QUALITY ry v^ ^c ^, o^ 1 I I ^N,.l N ..! N o d M I I I N t+t d N 00 N d N P ^O l^ 00 00 ^ l., d d d v+ t^ d ^D N -n pp d ^O v1 v1 I I I i I I d .- M oo ►n M M V'1 M d d lA ^ y N v r I r I 0 ... ^ ^^ ... N a v ? o `^ ^ p •^ N .^ .. U 1-. V vM.r vl N Q^ ON Gil O O O O J O .- C O G O Q ^,^ I I I I s v o S 8 ^ C C C O Cv O r. 7--. !n v iz 0:. D: Ot. Q:i 1! I'- DD 00 00 00 a„ L C'^. C v! C.. ^ ^ 3!, rJ !0 ^' !0 tCa ~~ a a, a a =^ ^!! V^ V1 V1 fl N CA N -19 O O O O 34 16 M ^c r C O O^ S C O 0 O M O I i I po r4 ci 8 S 8 op op C O O C O O O O O M M M rn CA r4 Q: ^I ^• r OC 00 00 .^ .^ r. !ti !0 td R1 ^ !b ^ d L a a a °C °^ ^ h h L/'1 lP) to V! N N N f-4 N N ..r 17 17 O O O O O O

Original page 42

• s N 61 t^ O N [ .M.+ ... M ^ °V1 h M pppp ^p M ON 00 00 --^ G CA .+ N^+t V'; O; v vi Q M g M .- O O OR r N ^D ^D v M ^ v^ ^O M ^ I I I I I p^ ... r d --; O H r d a M M H vn M 0 r- I-^ r r4 a r l I I I I a w a ,^ .a c v z h ORIGINAL PAGE IS OF POOR QUALITY O %n N v1 a^ M f'9 M M M N Q Q pp V e^ m O C J O O O N N r T tt ... N -+ O^ M O^ 00 06 r ^D .^ — M M M M M M M %n %A M ^O O^ ^O O oc ^" ON 00 n r ^+ ^ M M M M M vii h I i I I I I I I ^+ '^ ^O Q O; ... 00 of 06 06 r-: 16 4 C M M M M M M q N° M v .i ^ I •^ ^ I ^„ ° d d N r O ^ ^ N N d p O O° 00 M O sC 0 0 M e+ MC v10 O N ^ ,c^ I I o i o f o f o f o f a i o f o f G I 0 '~`-' i O GT Om O^ CIA ^ O O O w v ,.., Gq vi a a O p^ g ° w .-r E" CA Z y 1 ar v CA IT rn v W in C ^., O O O O O V1 Y1 %n N W) P^ N N N N U G O C O O 00 ON ev °O rr► °Om S O r'+ O O O O O C C O C C O G ° ... r O O O O O en IT 'n Ja 12 10 kn O O O O O O C? V1 1A N V1 V1 1A try N N N N CA N N O A O O C G G O 35

Original page 43

'A ^ O ... er N 7 N M M M M M N 0 N Q\ ^0 00 C7 - O O C O h r. 00 O --^ Q^D O .__. w; Vi 00 \D (y rc M M M ^O 00 to M O^ i- N 0 0 00 ^O V1 M [f M M M M C ^ I i I I I A M M M M pn M .., .^ t^ ^ ^ M M O M, M1 tw .." N N 00 Chi 'v b V C O U ri a 00 W 8 O S J ^ O ^^ o p p O o 0 E I I I I I I .S ,n N 8 g o g$^ C] O O O O O v E o I^ O O^ r' =1 a. Z N F, 7 ^^ + a ai a a a U H ba W [ ^^ $ a a a a a O O 0 o o O W) kn vi W) W) t ,^ N N N N N N ... O C O O p O 36 nC amp ni IALffl/ M ^O i I I d 00 N 1 1 ! 1 M N 00 v^ 00 tl^ ... N %n %A 00 f'^1 M Vl vl ^O 00 ^O Wi ON ^O N M v) v1 M M V 1 v1 i I I I I iT ^? M M in v) ,.., 00 ^ M N ^n ^.. , ^ ^ r. d `,. .. 00 O N ^ O O O $ 8 p o O I I $ 000 I g o $° o 8 ., M M M M ;' + M o4 S a a o 0 0 0 0 0 wl W) +n vi N N N N N N C C G O O O

Original page 44

vt O^ 00 ^O oO M M d' OF PUQR QUALITY 00 1' ^O O^ N N H ^ t^+t ^ ^ ^^'1 M M M M 1 a4 C d ^' N a^ H ^ p o0 N 00 u v^ ON Tr ^+1 M to ^0 00 Q ^O t+1 fn 110 eq i I I i C ^ p ^O M N 00 N O^ M M Vs to Q Q H ^ 0 wN oq E M d a ^ I w a v z M C; O O N M ^" ^^ N -+ ^ O O^ ^O M M M t^+1 M M tf V1 00 .-, 14 M N N .^ en fn en I I i I I I I I v1 t+1 O Q O O ^ ^ h v1 V1 M M M M M VI M M r4 ... N ..r I I I I I I i I QQ^ QQ^ ^ d d ^O M O en c ,^ c 0 0 0 0 0 ,^ o „ o ,^ o ^a o 0 0 0 o Nc No h I o f M o I I I I C I o, l r. p 0 I O N 0 O O S M O Q C) N O O N o0 N N z a z^ 0 n M M .. d 9I 9I t/1 v1 V1 V1 1• h h h M M M M `" O O O O N M 00 NO O p O O d O O C O O C O O O ON O O O O O ^" ^„ ^, ^ ^ • ^• •-•^ •-• + t t Ij E :^ to V1 v1 4'1 V1 ^!1 vl V1 h h h h h h h !+1 M M M M M M M O O O O d O O c 37

Original page 45

N M N 00 ON M It Yn w YYn ^ V^1Y n n n 00 ^ ( I I ► M M M M Y ^ ^tA h ..±h 000 M N 00 h a, 1 I I I `, b O -n O ^ O C M N N y F. N O h N v1 O^ N M ^O 00 P ^O M 00 O; h M M 0^ Vt 00 M M M M M M It h M M h ^ M N ^O .-- h h V1 p^ oo o M g M M I I I I I N 00 N h ° h ^ Q^ 00 kn en en I I ► ^M N00 M ^ ^/1 O 00 N h fM+1 Q H M M h M M M M M M C ^ ^ N ^ h ^ b I I I I 00 _ °n I I 1 GDI u C b ^ N M O v M v W ^ Q ..a ro ^ O 0A Q E 5 I I I I I H g o C I °r, I I Q^ ^ N M N O C I I C i I I 1 g C c 0. -. .. ^ r. .^ M N v a.+ N IA M M N F^ oN N en en !^ clu a a^', ci, ^y a, °G oG pG ~ a a 01 W) %n in ^ n r N r r h h V M M M M th M M t+y M M M M ^- 0 0 0 0 0 0 0 0 0 o d o 38

Original page 46

^ .-- ^ u°^ ^ %n ^+ M fJ 00 M M Vl H1 M M Y ^ ^n I^ to O M ^' x b O O ^ K1 O O V N N of M M 'n ^O M !r pip M ^ N ^ M ^ a I I I i I I c^ 0 00 110 M H M M W) K1 M Q H -^ N M 00 N i ^yM'., N I I ! ^ I I I a. d o^ -, kn W, M 00 a a a d N d N J, M -. o c 0 0 o,n o 00 C CJ I O I O O{ O I ^^' N I M I O^ Ova O $ O O O W O O C C a aq Q H o O O %n N zN v KO o^ ;' + + d H c o^ 0% C2 r h h Wn Vl kn to r ^ r r r n ^, ^ M M M M M M L O O O O O O ORICf':=". P OF POOR Q v, oo v^ 00 v, ^ 00 116 4 N ^+ M M M M M ^f M N '7 ^ N O O O O ^ N N1 It It M n ^D ^f M M M ^f v1 00 00 00 P ^ M M M ^ 1n i I I I I I 00 0^ o Kai M M M v1 r ^ V'1 I i 1 I 1 { 0 ^n r^, kn a ,^ d .. d MO Ol ^ -0 t- wl 0 0 ^o $o 00 00 oc O I OOpI O I O I O I O I Ov'+ aN OCR O^n O^n O O ^O ~ O ^O ^O O C O O 6 C %n %n %n Wn v) d r^ a a ;' :E to to to N V1 h ^ r ^ r ^ ^ M M M M M M O O O O O O 19

Original page 47

V1 .-, N vf N 00 M M ^ M M Q N ^-• d •rn ^ V1 M N M ^ -p O O C ^ O a. 4. CA C M M o0 O N m M d d M M p O^ N O M d d M I i I I m ^p ^D ^O .m ORIGINAL PAGE IS M [^ to OF POOR QUALITY M M M 0p ^ O N N O^ .-, t^ 4A M d h ^D O ^ r ^ M d 'n I i I (^ ^p N h d enM M M d d Men d N OM N ^, N ^ d 00 ^..^ ^ ^.C v ^^ ^ ov0 v ^^ ^.,,i ^ ^ lb... ^ ° ° I I ^D I I ..i N by ^ v v 00 ^ a+ y O^ N N d N v U v ^o lb lb I I .^,, O .- N, 'A ^. -- c M c 00 O O O W u ,1 O OO OO O O O O O O I O C O I E -E I d ^ I N Ea- O p 0 0 o c o0 0 00 0 H ^G I I 0 0 z to V1 to to z C t/1 :3 a z^ ^ M o p4 f^G AG fX Q4 ^ `^ " o E w a OJ ^" a a a 1a C ^ t^ r- ryq M M en M -' O O O O E- 40 a a a. a M[^ M M M O O O O

Original page 48

oo ao O v1 N M CACA 4 M M _ 00 M ^ A "' O, ^ ^: p N d O O N1 N M O^ O '-' w'^' ^ V01 h M M ... ^O d N O ^ 00 ^ N ten M I I I I I p t^ ^O N o? en .-- h M M Q M ^ ^ F et c^ O ^ ^ F be I I I I i V1 I CA dG v N v v v ^ ^ 1n M N 00 N M O O en .-, p ^^ v ^^^ O O C CO CO C v^... ^. ..„ I 0w ON vi,.., $ p O O 4 p O O C W M 00 c'' O O F^cj to M a a z^ aw o a a v 1 N F ^3 co ^^ g g g g 'n +n nZn Zn O O O C O ORlUFAL FADE F3 Of POOR QUALITY ^o o^ ON v 1 aoo N M N M M N N 00 n O; t^ C oo O O t+i p o —y ei ^ ON N en .-^ M M N '^ M r, •--^ O^ O^ t^ eN^ O M en d M M I I I I 1 M ^ .-. 06 N '^ M M M N ^ ^ ^ ^+ cv r1 ^ ^ C4 ^ ^ ^ ,.. ^. ^ 0^0 N 10 %n M pp In to _^ _p p ^O .-.O 0 NO NC OC $ O p p I G I p O O C I O ^ OM ^^ O^ D00 O O p p O O O ^ ^ o M a a4 a a ca cc m cc g g g g g in v! ^n Vi Wi O C O O O O 41

Original page 49

o o co o eq r°o, b -- O O O a^ C O t^ ^O t^ 7 "^ M M M M ea LL 00 00 ^O O M M M M ^ I I I I cv O; 00 N N 00 M In M M r- N N O ^ fn r4 ta E .. C6 'O C 1 I /^ I .^ I ov o^ op I x 00 "= "o 1-1 ^ o a bo o v O N (r! to U vi W O O O a .- C p C C as _ 1 o I I H tn ^ ^ °o o $ °o 0 0 0 Lr 6J uj O O O O Euu O Q. zN u (y E.., F. E.. E-! OD 00 00 O4 a 5 ^ E~ a a a a ^^c g g g g U O O O 0 ARi fir en I ( I N ^ I I I e+i N ^O M ON l^ of e! V1 M N N l^ V V1 ^A M h I i I ^ I 0% N ^/1 M lqt lqt ^, ^ en^ r^ .r %A In 4 op I op I `^ h " v) 1 ..^ ..,^ 6v v ^ b tX h lb O C ^^ $ C O O C 0 8 g C g i I o C TG^ 0o go $ $ $o 0 0 ° ° O O M M M oN fn ON N N E en ^+ C 00 Op ° b d Cd 3 a 3 oG 3 pG a a O ^ ^., ^ u, O O O O O

Original page 50

REFERENCES i . Maecker, H.: PlasmastrOmungen in Lichtb6gen infolge eigenmagnetischer Kompression. Zeitschrift Mr Physik, Vol. 141, 1955, pp. 198-216, 2. Metcalfe, J. C. and Quigley, M. B. C.: Heat Transfer in Plasma Arc Welding. Welding Journal, Vol. 54, No. 3, March 1975, Research Supplement, pp. 99-5 to 103-S. 3. Petrie, T. W. and Pfender, E.: The Influence of the Cathode Tip on Temperature and Velocity Fields in a Gas-Tungsten Arc. Welding Journal, Vol. 49, No. 12, December 1970, Research Supplement, pp. 588-5 to 596-S. 4. Key, J. F.: Anode/Cathode Geometry and Shielding Gas Interrelationships in GTAW, Welding Journal, Vol. 59, No. 12, December 1980, Research Supplement, pp. 364-S to 370-S. 5. Anonymous: Linde Unveils Inert-Gas Tungsten-Arc Cutting of Aluminum. Welding Journal, Vol. 34, No. 11, pp. 1097-1098. 6. Miller, H. R. and Filipski, S. P.: Automated Plasma Arc Welding for Aerospace and Cryogenic Fabrications. Welding Journal, Vol. 45, No. 6, June 1966, pp, 493.501. 7. Privoinik, L. J. and Miller, H. R.: Evaluation of Plasma Arc Welding for 120 in. Diameter Rocket Motor Cases. Welding Journal, Vol. 45, No. 9, September 1966, pp. 717-725. 8. Cooper, G., Palermo, J., and Browning, J. A.: Recent Developments in Plasma Welding. Welding Journal, Vol. 44, No. 4, April 1965, pp. 268-276. 9. Herbst, H. T.: Production Applications for Inert Gas-Shielded Arc Welding. Welding Journal, Vol. 26, No. 5, May 1947, pp. 410-418. 10. Vilkas, F. P.: Square-Wave Welding. IEEE Transactions on Industry and General Applications, Vol. IGA-3, No. 6, November/December 1967, pp. 525-530. 11. VanCleave, B.: Private Communication 12. Regan, R. E.: Plasma Arc Welding: State-of-Art Survey. Boeing Company Report No. U324- 11012-1, May 1977. Published as Appendix to: McCaw, R. L.: Plasma Arc Welding of High- Perforniance Ship Materials. David W. Taylor Naval Ship Research and Development Center, Report No, DTNSRDC/SME-78/34, May 1979, pp. 19-84. 13. Clover, F. R., Jr.: Welding of the External Tank of the Space Shuttle. Welding Journal, Vol. 54. No. 8, August 1980, pp. 17-26. 14. Nunes, A. C., Jr., Novak, H. L. and Mcllwain, M. C.: Weld Geometry Strength Effect in 2219- T87 Aluminum. NASA Technical Memorandum TM-83404, March 1981. I5. Cibuls, A.: Gases in Cast and Wrought Metals. Twenty-First Autumn Course of the Institution of Metallurgists, Eastbourne, 1967. Printed in Kubaschewski, 0., Cibula, A., and Moore, D. C.: Gases and Metals. American Elsevier Publishing Co., Inc., New York, 1970. 43

Original page 51

1. Ransley, C. E. and Neufeld, H.: The Solubility of Hydrogen in Liquid and Solid Aluminum. Journal of the Institute of Metals, Vol. 74, 1948, pp. 599-620. 17. Grigorenko, G. M. and Lakomskii, V. I.: Solubility of Hydrogen in Aluminum at the Temperatures Characteristic of Arc Welding. Automatic Welding, Vol. 20, No. 12, December 1967, pp. 27-29. 18. Howden, D. G. and Milner, D. R.: Hydrogen Absorption in Arc Melting. British Welding Journal, Vol. 10, No. 6, June 1963, pp. 304-316. 19. Woods, R. A.: Porosity and Hydrogen Absorption in Aluminum Welds. Welding Journal, Vol. 53, No. 3, March 1974, Research Supplement, pp. 97-S to 108-S. 20. McFarlane, W. M.: Interim Report — May through Septemter 1982 — Variable Polarity Plasma Arc (VPPA) 5015 Demonstration. TD 1008 Test Report, November 1982, prepared by Martin Marietta Aerospace, Michoud Division, Advanced Manufacturing Technology, for NASA, George C. M::rshall Space Flight Center. 44

Original page 52

THE VARIABLE POLARITY PLASMA ARC WELDING PROCESS: ITS APPLICATION TO THE SPACE SHUTTLE EXTERNAL TANK — FIRST INTERIM REPORT By A. C. Nunes, Jr., E. O. Bayless, Jr., C. S. Jones, III, P. M. Munafo, A. P. Biddle, and W. A. Wilson The ip. formation in this report has been reviewed for technical content. Review of any information concerning Department of Defense or nuclear energy activities or programs has been made by the MSFC Security Classification Officer. This report, in its entirety, has been determined to be unclassified. 1 , l tr R. 'J. SCHWINGHAMF Director, Materials and *U.S. GOVERNMENT PRINTING OFFICE 1983-848-058/4 45