1RESIDUAL STRESSES IN WELDMENTS IN HIGH-STRENGTH STEELSbyJYE-SUAN HWANGB.S., Chung-Cheng Institute of Technology1970REPUBLIC OF CHINASUBMITTED IN PARTIAL FULFILLMENT OF THEREQUIREMENTS FOR THE DEGREE OFOGFAN-ENGINEER M.:-'~ .'-at theMASSACHUSETTS INSTITUTE OF TECHNOLOGYFebruary, 19767A . L, .Signature of Author ....... v ......... ,B .Department of Ocean ErgineeringFebruary, 1976Certified by...$,,.. .Accepted by........II" I:I·'/. .I . .. . . I B I . . .Thesis SupervisorA .1.... j . . ~. ~. · . ..w .............Chairman, Departmental Committeeon Graduate StudentsArchivesPnws . eNe. HAPR 151976A ,F6 ES2RESIDUAL STRESSES IN WELDMENTS IN HIGH-STRENGTH STEEISbyJYE-SUAN HWANGSubmitted to the Department of Ocean Engineering on February 5, 1976,in partial fulfillment of the requirements for the Degree of/OceanEngineer ? ' ABSTRACTThe systematic results of series experimental studies on thermalstrain changes, temperature histories, deflection changes duringwelding and residual stresses existed in weldments after coolingwere presented. Investigations were conducted on simple weldmentsmade with several types of materials including mild steel, 308stainless steel, T-1 steel, HY-80 steel and a Ti-6A1-2Cb-lTa-0.8Moalloy. Comparisons on experimental data were made with the resultsof one-dimensional computer program that have been developed inDepartment of Ocean Engineering, M.I.T..The one-dimensional analysis appear to be sufficiently accuratefor temperature predictions. Better agreements were obtained instrain changes and deflection changes in GTA welded plates thanthat in GMA welded plates.Residual stresses are about 50 % yield strength of the materialin high-strength steels such as HY-80, T-1 steel and Ti-6211 alloy.3The accuracy of the analysis is limited due to lack of knowledgeof the variation of the material properties with temperature and thatof the heat losses during welding.THESIS SUPERVISOR : Koichi MasubuchiTITLE Professor of Ocean Engineering & Materials Science4ACXNOWLEDGEMENTI wish to thank my country - The Republic of China - forsponsoring my program of studies at M.I.T.To professor Koichi Masubuchi, who patiently advised andguided me as a professor and a friend, I am forever grateful. Also,I would like to thank Mr. Fred Merlis and Mr. Antony Zona for theirkindly assistance and helpful suggestions during experiments.I would like to extend my thanks to Mr. Michio Nishida forhis friendly assistance and suggestions.To my parent,I am grateful for their support and encouragementnot only during this work but also during my entire course at theM.I.T.5TABLE OF CONTENTSPageTitle Page 1Abstract 2Acknowledgement 4Table of Content 5Lists of Figures 7Lists of Tables 11I. INTRODUCTION 13A. General 13B. Development of Analytical Studies 14C. Effects of Metallurgical Transformation on Thermal 15Stresses and Residual Stresses.D. Previous Experimental Studies 16E. The Purpose of Present Study 17II. MATERIAL CHARACTERISTICS 23A. Materials Selection 23B. Temperature Analysis 23C. Effects of Material Properties 26D. Material Properties Used for This Study 26III. EXPERIMENTAL INVESTIGATION 36A. Objectives 36B. Experimental Procedures 37C. Experimental Data Analysis 456PageIV. RESULTS AND CONCLUSIONS 64A. Results Representations 64B. Conclusions 106V. REFERENCES 107APPENDIX A 110APPENDIX B 1187LESTS OF FIGURESFigure No.12345678910111213141516171819Description PageResidual Stresses in 38 inch long Butt Weld 19in Low Carbon Steel and 4340 SteelResidual Stress Distribution( I Section Member ) 20Welding Residual Stresses in High Strength Steel 21Maximum Mechanical Strain Observed at Locations 22One Inch From Weld LineThermal Properties for Steel and Alumium 25Yield Strength of T-1 Steel Vs. Temperature 28Expansion Coeff. of Titanium Alloy Vs. Temperature 31Specimen Preparations for Mild Steel, T-1 and308 Stainless Steel. 48Specimen Preparations for Ti-6Al-2Cb-lTa-0.8MoAlloy 49Strain Gage Arrangement in HY-80 Specimen 50Outline of Experimental Set-up 51Thermocouple Instrument Circuit 52Strain Gage Instrument Circuit 53'Strain Gage Installation for Experiments 44Strain Gages arrangement on Specimens of MildSteel and T-1 Steel for Phase 2 Experiments 54Strain Gages Arrangement on Specimen of HY-80for Phase 2 Experiment 59Strain Gages Arrangement on Specimen ofTi-6Al-2Cb-lTa-0.8Mo 61Strain Cganges for Mild Steel (GTA) atDifferent Points Along the Welded Edge 66Strain Changes for 308 Stainless Steel (GTA)at Different Points along the Welded Edge 67Figure No.20DescriptionStrain Changes for T-1 Steel (GTA) atDifferent Points Along the Welded EdgeMaximum Reached Temperatureof Mild Steel (GMA)Maximum Reached TemperatureMild Steel(GTA)Maximum Reached Temperature308 Stainless Steel (GMA)Maximum Reached Temperature308 Stainless Steel (GTA)Maximum Reached Temperaturefor T-1 Steel (GMA)Maximum Reached TemperatureFor T-1 Steel (GTA)Maximum Reached TemperatureFor HY-80 (BUTT WELD)Maximum Reached Temperaturefor Titanium Alloy2122232425262728293031323334353637StrainStrainStrain(GMA)StrainStrainStrainStrainChangesChangesChangesChangesChangesChangesChangesforforforforforforforDistributionDistributionDistributionDistributionDistributionDistributionDistributionPage6869707172737475DistributionMild Steel (GMA)Mild Steel (GTA)308 Stainless SteelStainless Steel (GTA)T-1 Steel(GMA)T-1 Steel (GTA)HY-80 Steel(BUTT)Temperature History at 0.625" From WeldedEdge of Ti-6Al-2Cb-lTa-0.8MoStress Changes For Titanium Alloy-6211Observed on the Gage at 0.625" Away FromThe Welded Line7677787980818283848Figure No.3839404142434546484950515253DescriptionStress Changes For Titanium Alloy-6211Observed on The Gage at 1.625" Away FromThe Welded Line.Longitudinal Strain And Temperature at1.0" From Welded Edge. ( Mild Steel GMA )Longitudinal Strain And Temperatureat 1.0" From Welded Edge( Mild Steel GTA )Longitudinal And Temperature at 1.0"From Welded Edge ( Stainless Steel GMA )Longitudinal Strain and Temperature at 1.0"Welded Edge ( Stainless Steel GTA )Longitudinal Strain and Temperature at1.0" From Welded Edge ( T-1 Steel GMA )Longitudinal Strain and Temperature at1.0" From Welded Edge ( T-1 Steel GTA )Longitudinal Strain and Temperature at 1.5"From Welded Line ( HY-80 Butt Weld )Deflection Changes at The Center of MlidSteel ( GMA )Deflection Changes at The Center of MildStee:l ( GTA )Deflection Changes at The Center of 308Stainless Steel ( GMA )Deflection Changes at The Center of 308Stainless Steel ( GTA )Deflection Changes at The Center of T-1Steel Plate ( GMA )Deflection Changes at The Center of T-1Steel Plate ( GTA )Deflection Changes at The Center ofTi-6A1-2Cb-lTa-0.8MoResidual Stresses Distribution In Weld-ment of Mild Steel ( GMA )985868788899091929394959697989910010Figure No. Description Page54 Resiciual Stresses Distribution In Weld- 101ment of Mild Steel ( GTA )55 Residual Stresses Distribution In Weld-ment of T-1 Steel After Cooling ( GTA ) 10256 Residual Stresses Distribution In Weld-ment of T-1 Steel (GMA ) 10357 Residual Stresses Distribution In Weldmentof H"Y-80 ( Butt Weld ) 10458 Residual Stresses Distribution In Weld-ment of Titanium Alloy ( GMA ) 10511LIST OF TABLESFigure No.2-12-22-32-42-52-62-72-82-92-103-13-23-33-43-53-6DescriptionChemical Composition of T-1 Alloy SteelMechanical and Physical Properties ofT-1 SteelThe Effect of Elevated Temperature onModulus of ElasticitySpecification Limited of HY-80 ChemicalCompositionMaterial Properties Vs. Temperature( HY--80 )Chemical Composition of Ti-6211 AlloyMaterial Properties Vs Temperature( Ti--6A1-2Cb-Ta-0 .8Mo )Chemical Composition of 308 StainlessSteel308 Type Stainless Steel PhysicalPropertiesComparison of Some Physical PropertiesBetween Steel, Aluminum and Titanium.Welding Procedures and ConditionsResidual Stresses Measurements on GMA WeldedMild SteelResidual Stresses Measurements on GTAWelded Mild SteelResidual Stresses Measurement on GMAWelded Specimen of T-1 SteelResidual Stresses Measurement on GTAWelded Specimen of T-1 SteelResidual Stresses Measurement on ButtWelded Specimen of HY-80Page27272829303132333335405556575860Figure No.3-7DescriptionResidual Stresses Measurement on WeldedSpecimen of Ti-6Al-2Cb-lTa-0.8MO3-8 Summary of Residual Stresses in ThisStudy12Page6263I.INTRODUCTION 13A. GENERALResidual stresses due to welding are thermal stresses remainingwhen the material has cooled to ambient temperature. Thermal stresseshave been investigated both by analytically and experimentally for manyyesrs. In 1936, Boulton and Lance Martin first presented analyticaland experimental results and shown that welding induced plastic deform-ation of material in and near the weld and that the residual stressesresulting after cooling were due to these plastic deformations. Fujita2studied the thermal and residual stresses arising in a center weldedplate by both experimental and theoretical means. His tests indicatedthat the temperature distribution of a plate can be assumed to be thatof an infinite plate if the width of the plate is wider than about teninches. After that a number of research programs have been carried outon residual stresses and distortion in weldments, several reviews andbooks have been written on various subjects related to residual stressesand distortion in weldments.3 '4 Until recently, however, almost allstudies were on residual stresses and distribution after welding iscompletely and only limited studies were made on transient thermalstresses and metal movement.5 This is due to the complexity of theproblem characterized by large temperature changes in small areas nearthe welding arc with its resulting non-elastic deformation, temperaturedependency of the material properties and/or phase transformations, andcomplex boundary conditions resulting from conditions of geometry andmultipass welding. With the advancement of the computer technology,14it is possible to analyze thermal stresses and metal movement withreasonable cost and time. Hence, investigators at M.I.T. and some otherlaboratories in the world are currently developing computer programs onthermal stresses during welding.B. DEVELOPMENT OF ANALYTICAL STUDIESBased upon the earlier work by Prof. Masubuchi et al. of BattelleMemorial Institute, researchers of M.I.T. first developed a one-dimension-al program of thermal stresses during welding.7 In the one-dimensionalprogram, it was assumed that: (1) The longitudinal stress o' is axfunction of the lateral distance from the weld line, y, only and (2)ty and are zero. Then, two-dimensional programs were developed -1 0y xyRecently, efforts were made to improve the two-dimensional analysis tocover butt joints as well as bead-on-plate welds. Complex boundaryconditions must be considered in calculating transient thermal stressesduring welding a butt joint. The portion of the joint ahead of thewelding arc is separated, while the portion of the joint behind thearc is connected. The complicated boundary conditions keep changingas the welding arc travels.In order to analyze thermal stresses with an reasonable accuracy,it is essential to conduct an accurate analysis of heat flow. Hence,an elementary three-dimensional finite element method program has beendeveloped on heat flow during welding at M.I.T. A limited effort hasbeen made to analyze transient heat flow during multipass welding ofheavy plates, it has been found that a general three-dimensional1513analysis of transient heat flow is expremely costly. From the resultsof recently studies at M.I.T. and other laboratories, an approximateanalysis of thermal stresses and residual stresses in heavy weldmentsis possible by using current 2D plane stress programs properly.C. EFFECTS OF METALLUJRGICAL TRANSFORMATION ON THERMAL STRESSES ANDRESIDUAL STRESSESIt has been well established that in a weldment in low-carbonsteel, longitudinal residual stresses in regions in the weld are tensileand almost as high as the yield strength , several investigators havealso found that residual stresses in weldments in some high-strengthsteels are considerably lower than the yield strength of the base metal.Several investigators thought that metallurgical transformation duringcooling could have important effects on these stresses.Recently a study was made at M.I.T. by Toshioka, results of theanalysis are shown in figure 4,5,6 of Reference 11. The resultsclearly indicate that metallurgical transformation has significantinfluence on the distribution of residual stresses in weldments inhigh-strength alloy steels. It must be mentioned that the analysisby Tosioka is somewhat incomplete in handling welding situations andhe used the exxisting two-dimensional plane-stress program. A furtherstudy is needed to assess the effects of metallurgical transformationin residual stresses in weldments.16D. PREVIOUS EXPERIMEITAL STUDIESSeveral investigations have studied experimentally on residualstresses in weldments in high-strength steels. Masubuchi and Martin 4investigated residual stresses in butt welds 5/8 inches by 38 inchesmade in low-carbon steel and in SAE 4340 steel oil quenched and temperedat 500 F. Figure 1 shows distributions of longitudinal and residualstresses along the transverse line. Residual stresses were tensile inareas near the weld and compressive in areas away from the weld. Theimportant finding obtained in this investigation is that residual stressesin the low-carbon steel and SAE 4340 steel weldments were similiardespite the considerable difference in the yield strengths of the basemetals and the weld metals used. The differences were as follows:Low-carbon steel base metal --------------- 35-40 ksiE 6010 weld metal ------------------------- 60 ksiSAE 4340 Q & T' steel ---------------------- 224 ksiE 15016 weld metal ------------------------ 150 ksithe maximum residual stress to mean yield strength ratio in low-carbonsteel is significant higher than that in SAE 4340.Similar results were made by Kikuchi and his research group15on welded plates of 80 kg/mm2 high tensile steels (HT80) and 50 kg/mm 2high tensile steels (SM50), i.e. small ratio of tensile residual stressto yield stress of HT80. See figure 2 Figure 3 summarizes experimental results by Kihara, et al andYada, et al. 17 The ratio of the maximum residual stresses to the yield17strength decreases as the yield strength of steel increases.18,19n a recent study conducted by Kl in , measurements were madeIn a recent study conducted by Kleinof transient thermal strains during welding butt joints in several high-strength steels. Results are summarized in Figure 4 which shows therelationship between the yield strength of the material and residualstrains( measured after welds cooled to room temperature ) at locationsone inch from the weld center line. Strain decreased as the yield streng-th increased.Experimental results obtained so far indicate that residual stressesin weldments in high-strength steels appear to be not as high as theyield strength of othe material.E. THE PURPOSE OF PRESENT STUDYIn studying thermal stresses during welding, it is important tomaintain a good balance between analysis and experiments. A series ofexperiments were conducted at M.I.T. during the last several years onweldments in various materials and thicknesses, and experimental datawere compared with analytical predictions. Although a number of studieshave been made of the residual stresses on weldments in high-strengthsteels, there has been not even a single study which covers threesubjects, (a) experiments on transient strains. (b) experiments onresidual stresses. (c) comparison of experimental data with analyticalpredictions. Hence, a comprehensive study will be made of thermalstresses and residual stresses in simple welds in low-carbon steels18and in high-strength steels in present work.The study in this paper covers the following two tasks:Task 1: Experiments on transient thermal strains in simplewelds and analysis.Task 2: Experiments on residual stresses and analysis.Distance From Weld Center Line, inchDistribution of Longitudinal stresses along theTransverse Line Passing the Weld CenterFigure 1 RESIDUAL STRESSES IN 38-INCH-LONGBUTT WELD IN LOW CARBON STEEL AND 4340 STEEL.( REFERENCE 14 )0r19OUno 040rno-43C 20-40)riED-20-4020Kg/mm2....___ HT 80...... SI 50RESIDUAL STRESS ( Kg/mm2 )20 40 60WEBFigure 2 Residual Stress Distribution ( REFERENCE 15 )( I Section Member )FLANGE503010-10IrI-20I[ ][[ [ I I I...-IIII- sed )120100IcuMI0*rl$440)M£0v,-HMS.xHxscd:rE--... , V1 ase metal, my ( Ksi )Figure 3 Welding Residual Stresses in High StrengthSteel21IAHY-80II501100HY130I I150180KsiI200YIELD STRENGTH OF BASE PLATE ( KRsi )Maximum mechanical strains observed atlocations one inch from weld line( Reference 18 )222500 -%oz0it)H2000 .1500.1000500oFig. 4-23II. MATERIAL CHARACTERISTICSA. MATERIALS SELECTIONIn order to obtain conclusive evidence for current study on thetransformation effects, experiments should be made on several materialswith different metallurgical characteristics. The following is acombination of material used in present study:1. Low-carbon steel.2. T-1 Quenched & Tempered high-strength steel.3. HY-80 Quenched & Tempered high-strength steel.4. 18-8 Stainless steel.5. Titanium alloy Ti-6Al-2Cb-lTa-0.8Mo.Since the results of the temperature calculation are affected bothby the thermal properties of the materials and by the welding conditions,while the results of the stress calculation are determined by the tem-perature distribution and by the mechanical properties of the materials,the mechanical properties which govern both heat flow and stressanalysis may vary considerably between room temperature and meltingpoint of particular material. Hence, both the physical and mechanicaltemperature dependency properties of the material should be consideredfor experimental data analysis.B .TEMPERATURE ANALYSISIn order to calculate thermal stresses during welding, it is firstnecessary to calculate temperature changes. The temperature distributionis calculate from the two-dimensional solution for a line source in an24infinite plate. This solutions;vq -- wT = T+ e 2K K( ) ...... (1 )0 2s O 2Kwhere r = ( w2+ y2)1/2K (z) = the zero order modified Bessel functionof the second kind.w = x - vt = moving coordinate.T = temperature.TO= initial temperature of the plate.s = thermal conductivity.K = = thermal diffusivity.cgc = specific heat.g = density of the material.q = eff x V x I/ h = heat source intensity.eff = arc efficiency.h = plate thickness in inches.V = arc Voltage in Volts.I = welding current in ampares.From the equation shown above, the temperature distribution is ofcourse controlled by the welding conditions and thermal properties.While the problem of thermal stress analysis is more complicated thanthe calculation of temperature distributions. There are two importantways in which the temperature distribution affects the plastic deform-ation, The first of these is obvious as a result of the increased tem-perature, the material near the weld expands. The second effect involvesDENSITYo4t25WATT-SECF- IN3X SPECIFIC HEAT0n~ cJO4O a 0 0-s cl; S S- II( sl)SJLrMO'-4ooO-OO4v-OOo 4-Oo E-co EO E0\0rH:.ratrQ3HQ)$400)4-t4)$4Q)O0$4-.CooOOcQIn0$4bD*HrL4OX;IAIJflOaNOO IY1WIHLour)oC)a0:00)V~26the variation of yield strength with temperature. At elevated temper-atures the yield strength of the material is considerably reduced andlarge plastic deformations occur for the same total strains.C. EFFECTS OF MATERIAL PROPERTIES :Comparing temperature distributions for different materials is muchmore complicated, since the variations in properties with temperature maybe difference for different materials. Figure 5 shows the thermal con-ductivity and the product of density and specific heat for steel and alu-minum ( Reference 20 ), these properties will affect the accuracy of analy-sis like welding conditions do. The problem we have have is lack of suffi-cient data of material properties at elevated temperature that used foranalytical calculations.D. MATERIAL PROPERTIES USED FOR THIS STUDY1. Mild steel: Mild steel is an ABS class B type steel, most of the mater-ial properties were taken from Reference 20.2. T-1 alloy steel: T-1 steel, an ASTM/A517 steel produced by the US steelcorporation, is a low carbon, Q & T alloy steel which exhibits a combinationof high yield strength, excellent tougness, high resistance to impact abra-tion and good weldable. T-1 steel is commonly used for commercial appli-cations including pressure vessels, storage tanks and merchant ships.Compared with HY-80, T-1 steel contains less nickel and is less costly,The chemical compositions and the heat treatment mechanical propertieswere shown in Table 2-1, 2-2 and 2-3, these data were taken from Ref. 21.27TABLE 2-1 Chemical Composition of T-1 alloy SteelC Mn P S Si0.10/0.20 0.60/i.0 0.035 max 0.040 max 0.15/0.35Ni Cr Mo V Cu B0.70/1.00 o.45/0.65 0.40/0.60 0.03/0.08 0.15/0.50 0.002/0.006TABLE 2-2 Mechanical and Physical Properties of T-1 SteelMechanical PropertiesYield strength min. Psi 100,000Tensile strength Psi 115,000/135,000Elongation in 2 in min % 18Reduction of Area % 45Physical PropertiesDensity, Pci 0.2833Electrical Resistivity, microhm-cm 24.0Modulus of elasticityTension, Psi 28 to 30 x 106Compression, Psi 28 to 30 x 106Coefficient of Expension 6.5 x 10 ( -50 to +150 F)7.74 * 10- 6 ( 70- 1300 F )28TABLE 2-3 The effect of elevated temperature on modulusof elasticityTemperature F Modulus of elasticity Psi80 30,300,000200 28,600,000400 27-,400,000600 27,400,000800 25,600,0001000 22,700,0001200 14,600,000Figure 6 shows the short-time Tensile properties of T-1 atelevated temperatures.120100U,.p.p+H4).,>,806040200Which took from USS United State Steel.1600I 1Temperature ( F )FIC. 6 Yield Strength of T-1 steel Vs.Temperature.293. HY-80 Steel:HY-80 is a low carbon alloy steel which acquires strength and tough-ness through quenching and tempering. It is a material suitable for sub-marine pressure hulls which have modest resistance to corrosion and rela-tively stable physical properties in the range of temperature from -30 Fto 120 F, its modulus of elasticity is fixed at about 30 x 106psi and22density at about 0.283 lb/cu.in. The chemical composition of HY-80is shown in Table 2-4.TABLE 2-4 Specification Limited of HY-80 Chemical CompositionC Mn P* S* Si Cu.18 max 0.10/0.40 .025 max .025 max 0.15/0.35 0.25 maxTi Mo Va Ni Cr.02 max 0.20/0.60 0.03 max 2.00/3.25 1.00/1.80* On both Ladle and Check Analysis the per cent of P and Stogether shall not be more than 0.045.The material properties for HY-80 at elevated temperature wereobtained by either from the data of T-1 steel or that of HY-130 inJ. Schrodt's thesis.2 3Table 2-5 shows the temperature dependent properties of HY-80 usedfor analysis.a,E-I.a,Ha,*rlPIr-IHa)ClE0080000000c'-0000O\00000000400000000cUl0900OOOa,-IOOOOna?,OOOOCQ,HOOQ)Ha1)IE-iCQJ0C\2C-No-0cc-ocoNCccoC\2(o0\ONCC\Oncx)0o0CQHontrd0U)bfl0tor.;j0C\lD\ooxc_tfI-oC-0ocCC-xcON0-cccoiCcm-lco4.)Hb.rl.r-izHV-'o 0cCo\4C-V-'C\cxi0c'---')ON'I'-OC~ a,a,.i00oHU o000o0ooooooooo0E PII- - Stt2X9C-40lx20CT-Nnoo0C\tOcc0U)C\l00C'oOD0C-)rd00H *) >.~ ~ ~ ~ ~ c .Elq ri0\-cccooo01\xl~ODC\iElnc\l0Ci4CV-1-OU-'cc.HC'-PxioCda,X30m m -l - r-- - .. . __ .i _ _E l _ct314. Titanium alloy:Titanium metal is finding many uses. It's alloys are of greatsignificance in strength/weight ratio, corrosion, marine and aerospaceapplications. Ti-6Al-2Cb-iTa-lMo is a near alloy type having inter-mediate ultimate strengths from 125-150 Ksi, strengthed principlally byAl and small amounts of carefully formulated beta phase. It's chemicalcomposition is shown as followed:Table 2-6 Chemical Composition of Ti 6211 alloyN C Fe Al Cb Ta Mo.o5 0.1 0.05 5.9 2.2 1.0 0.8Figure 7 shows the thermal expansions of Ti-6A1-4V and Ti-6211which is used for strain compensation of strain gages that mounted inTi-6211 specimen during experimental study.6.05.0m 4.03.0t 2.01.00.0200 400 600 800Temperature ( F )Fig 7 Expansion coff. vs Temperature32oooooCl0oooo0ooos0o00ONOO2)E-CD-COE'-; sl\0COO0C-oJoo,,-rlrd.-HizOIOoHd .H)E -H\iC"OCCloN0C'-.-\ocoo\C-ONONCC)mi-,.riC-)P4C-)O-Pa)Cda)H.ra,a)E-a)-rla):.P-('50(:O4IrlIv1-C-Ia)r-I- 1 -33Table 2-7 shows the material properties of Ti-6211 alloy that wereprepared for experimental analysis.5. Stainless steel:308 Type is one of 18-8 stainless steel which representing chromiumcontents from 17 to 20 percent and nickel contents from 8 to 13 percent.It is available in the form of sheets, strip, plates, bars and tubing.308 type stainless steel is a higher alloy steel having higher corrosionresistance and heat resistance, primarily used for welding filler metalto compensate for alloy loss in welding. Table 2-8 shows the chemicalcomposition of 308 type stainless steel.Cr Ni C Mn Si P S17/21 L0/12 0.08 2.0 1.0 0.045 .030Table 2-8 Chemical Composition of 308 typeStainless steel.Table 2-9 shows the physical propertiesof 308 type of stainless24steel.Table 2-9 308 Type physical propertiesMelting point, F 2550-2650Density, lb/c-u.in .287/.29Specific gravity 7.86/7.94Coeff. of Expansion 9.2 x 10_ 6 ( 68-212 )in/in/ F 11.Ox 10 (68-1600 )Thermal conductivity 9.4-12.4 ( Btu/hr-ft- F)Specific Heat( Btu/lb/ F ) 0.1234Table 2-10 shows the comparaison of steel, aluminum and Titaniumin physical properties.The material properties of each specimen that used for input dataof one-dimensional program were shown in appendix B.Table 2-10 Comparison of some physicalbetween steel, aluminum andpropertiesTitanium.Materials Steel Aluminum TitaniumDensity, lb/in3 0.234 0.1 0.163Young's modulusE x 106 psi 30 10 17Yield strengthC x 103 psi 35-150 30-50 40-150ysStrength/WeightQr/Wp x 103 123-150 300-500 250-920Thermal conduc-tivity, 26.2 130 9BT/hr/f t2/ft/FCoeff. Linearthermal exp. 6.8 13 4.70 x 10-6 /FElectricalresistivity 9.7 2.7 4210-6 cmMelting point 2800 1220 3040OFMelting point Fe Al2 032400 3of oxide OF37003536III EXPERIMENTAL INVESTIGATIONA. OBJECTIVESThe objective of the experimental investigation was to obtain a seriesexperimental data of thermal stresses and residual stresses in simple weld-ments in various strength-level materials. The study covers :( 1 ) Experiments on transienl strains.( 2 ) Experiments on residual stresses.( 3 ) Comparisons of experimental data with analytical predictions.Experiments were conducted in two phases. The major objective ofexperiment phase 1 was to measure temperature and strain changes on testspecimens during welding. Temperature changes during welding were measuredby thermocouples, while strains changes were measured by strain gagesmounted on specimen surfaces, The specimens were welded by either bead-on-edge with GMA process or heat-on-edge with GTA process. The majorobjective of experiment phase 2 was to measure the residual stresses exist-ed near the welded line of the specimens. In order to do that, electricresistance strain gages were used, this is one of stress-relaxation tech-niques for residual stresses measurements. one or two inches wide piececontaining the gages were removed from the specimens and the strains weremeasured by viscorder and so the residual stresses were obtained by simplecalculations.37B. EXPERIMENTAL PROCEDURES1. PHASE 1 EXPERIMENTSSpecimen Preparation. There are five different materials uesd inthis study, among them, two pieces of Mild steel, 308 Stainless steeland T-1 steel were prepared for bead-on-edge and heat-on-edge weld studies.Figure 8 shows the specimens which were 48 inches long, 6 inches wideand 1/2 inch thick. Three single-strain gages and thermocouples weremounted on the center of the specimens, two more gages were mounted on12 inches away from the center of the specimen to measure strain changesat different locations along the lengthwise direction. Figure 9 showsthe specimen of Ti-6Al-2Cb-lTa-lMo Titanium alloy, two rosette straingages were mounted. HY-80 specimen which was prepared for butt weld wasshown in Figure 10.Types and characteristics of the gages and thermocouples were shownas follows.STRAIN GAGES:Materials Gage type gage factorT-1 steelT-1 steel BLH TYPE FAE-25-12S6 2.06 + 1 %Mild steel308 Stain-308 Stain- BLH TYPE FAE-25-12S9 2.02 + 1 %less steelHY-80 steel BLH TYPE FAER-18B-12S6 2.00 + 1 %Titanium BLH TYPE FAER-18B-12S5 2.02 + 1 %alloy38THERMOCOUPLE : BLH TYPE GTM-CA ( Chromel/alumel )They were manufactured by BLH Electronics, Inc., Waltham, Massachu-setts.The strain gages were mounted with the BLH EPY-550 cement which iscompatible with all phenolic and polyimide SR-4 strain gages.The specimen area to be gaged must be chemically clean and free ofall grease, oil, scale or oxidation. In order to obtain best results,following steps were recommended :SURFACE PREPATION1. Remove all foreign matter ( oil, paint, oxide, scale, etc.) fromsurface with disc sander, grinder, or mill file, leaving surfacesmooth.2. Clean surface with gauze saturated with Methyl Ethyl Ketone.3. Dip one-inch strip of Silicon carbide paper into metal condi-tioner, lap surface, and remove residue with clean tissue.4. Indicate gage locations by using blade or 4-H pencil.5. Apply metal Conditioner to surface with cotton swab and removewith one stroke of clean tissue.6. Clean fingers with Neutralizer and cotton applicator.7. Apply Neutralizer to surface with cotton swab and remove withone stroke of clean tissue.GAGE PREPARATIONClean the bonding side of the gage with a clean MEK saturated39cotton swab. Avoid recontamination by handling the gage withtweezers following this operation. As gages were installingon the specimens, a liberal coating of the cement was appliedto the gage bonding surface and smooth to a uniform thickness.Position the gage and press out any trapped air bubbles. ThinTeflon film of .ool" to .003" should be placed between the gageand neuprene rubber pressure pad for easy removal after curing.A bond pressure of 15 psi will insure intimate contact betweenthe gage and the specimen during the curing cycle.Instrumentation and Recording : Figure 11 shows the outline ofthe experimental set-up used. Change of temperature and strains duringwelding, a subsequent cooling were recorded on an oscillographic paper.Figure 12 shows the thermocouple circuit, and Figure 13 shows the straingage circuit.Welding Procedures And Conditions : Table 3-1 shows welding pro-cedures and conditions used in phase 1 experiments.Experiments were conducted on three specimens first, on these spe-cimens three strain gages were mounted at different positions along theedge to be welded with the same distance ( 1 inch ) from the edge, tosee if the one-dimensional program is available for analysis.Two pieces of mild steel, stainless steel and T-1 steel were welded,one was heat-on-edge ( i.e. the three specimens just been mentioned ) byusing GTA welding process and the other was bead-on-edge weld by usingGMA welding process. Butt weld was made of HY-80 by GMA process.*d E-44-)·rl I I I I N I) 0o 0 0 \ .'H I \I' : '!CS E- -H I ) I ) a a a 0 0 0 0 0 0 0E744) CC 4) ad Q) H C0) o \0 o 1 c- oN 0 % o0 0 O 0c) n0) CQ 40 0 \ 40 4 0 . 40 N o 0 -4--. I -i oV 4 2d [ d - O4k N 4 N uI H n n -4 N -4 N Nu cs u 4) O H CO ) Hoc4 H HO 0 'H4 I 0a I 0sO H 'HO a o . +j o.U >4 IrO4) .4) l ) I I 4 4) I *HOIe U U) c E- E4 U) U) E-I E~~r >2 x -40r.0o01zCa)14a0-4lH~41It is important to note that the wire used for bead-on-edge weldand butt weld of HY-80 was one that used for ordinary steel and not forhigh strength steel. This is due to the lack of appropriated wire, butis it thought that it will not have a big effect on the resulting residualstresses.For Ti-6Al-2Cb-lTa-O..8Mo alloy specimen, bead-on-edge weld was madeby GMA process with matching wire of Ti-6Al-2Cb-lTa-0.8Mo alloy.2. PHASE 2 EXPERIMENTIn order to measure the residual stresses of the weldments that hadbeen made in previous experiments, we used electric resistance straingages. This is one of stress relaxation techniques for measuring resi-dual stresses. It is saed on the fact that strains take place duringunloading are elastic, even when the material has undergone plasticdeformation.Principle of Stress-Relaxation Technique Electric-strain gagesare mounted on the surface of the test specimens, a small picec of metalcontaining the gages is removed from weldments. As the specimen isstrained, the resistance of the gages changes, the magnitude of strainis then determined by measuring the resistance changes, the strain changes,' y' xy that take place during the removal of the piece are thenmeasured., The residual stresses are then calculated by using followingexpressions :x= 1 xy= - CSpecimen Preparationsin this study :.E v2 ( x +V gy) ........(2)E ( y + x) ........ (3)....... .. (4)xyThere are six specimens that are availableSpecimensMild steelMild steelT-1 steelT-1 steelTitaniumalloyHY-80Type of weld Size of specimens48"48"48"48"Bead-on-edgeHeat-on-edgeBead-on-edgeHeat-on-edgeBead-on-edgeButt weldx 6" x 1/2"x 6" x 1/2"x 6" x 1/2"x 6" x 1/2"56" x 7.5" x 1/2"29" x 15" x 1/2"308 Stainless steel specimens weren't used in thisto their smaller width ( 4inches ) that higher residualnot be obtained.this study owingstresses couldTypes and characteristics of the gages were shown as follows :Gage factor width ofpiece removedMild steelsT-1 steelsTitaniumBLH TYPEFAER-18B-12S-62.00 + 1 %42Specimens Gage type2"43BIH TYPEHY-80 steel FAET-18D-12S-6 1.98 + 1 % 1"The strain gages were mounted with EASTMAN 910 ADHESIVE. Curing pro-cess is not necessary for this type of cement, following procedures areuseful for gage installation :GAGE INSTALLATION1. Prepare surface ( Same as that descript in last section )2. Place gage face up on clean surface anf position terminalstrip at end of gage.3. Apply cellophane tape over top of gage and terminal strip.4. Carefully lift assembly from working surface and clean backof gage and terminal with cotton applicator slightly moistenedwith neutralizer.5. Place gage in position on specimen.6. Starting at one end of cellophane tape lift gage assembly,leaving other end of tape attached to specimen.7. Apply thin film of blue 910 catalyst to back of gage andterminal strip and allow to dry.8. Apply EASTMAN 910 ADHESIVE to gage area of specimen.9. Feed gage tape onto surface, holding free end of tape abovesurface with one hand and using ball of tissue in other handto quickly force gage assembly into place with one stroke.10. Within one second press gage firmly into contact with sur-face using thumb or finger. Maintain pressure for approxi-mately 30 seconds.4411. Wait few minutes before removing cellophane tape from top ofgage and terminal.12. Wire gage to terminal with copper jumper wires ( No.34 ) asshown Figure 14.13. Apply gagekote No. 1 or other suitable protective coatings.3 2 11. Strain gage. I /2. Terminal.3. Connectingwire.Figure 14. Strain gage installation for experiments.Residual Stresses Measurement Figure 15, Figure 16 and Figure 17show the strain gage arrangement on specimens of steels ( mild steel,308 stainless steel and T-1 steel ), Titanium alloy and HY-80.As we contact the strain gages through terminals by connecting wiresto the amplifier and the counter, everything is ready for cutting of thesmall piece of the specimens with the strain gages on it. The proceduresin this experiment are recommended as follows :1. Connecting terminals of the strain gages, and let the specimenfree.452. Taking the readings from the counter for each terminal of thegages before cutting.3. Putting the specimen on the plateform of the metal cutter wellin position to be cut and fixed it tightly.4. turn on the switch of the metal cutter and start to cut.5. Let the specimen loose as one side was cut and take the readingsfrom the counter again.6. Position the specimen by putting the other side of the smallpiece well on the line for cutting.7. Fix the specimen and start to cut.8. After 1 or 2 inches wide specimen has already cut, put thison the working table freely.9. Take the readings again after the temperature of the small piecewas cooled to room temperature.The results before cutting, after one side cutting and after bothsides cutting were recorded and were shown in' Table 3-2 to Table 3-7.C. EXPERIMENTAL DATA ANALYSIS1. The principle of the strain gages is based on the physical propertythat a metal strained in tension or in compression will change its re-sistance. Both mechanical strain and thermal strain changed in speci-men may cause the resistance changein strain gages, the mechanical strainmay consist of both elastic and plastic strain, hence, the resistancechange R is made up of the following :46AR = R(0¥) +R( ) +R(IF) + R(T)WhereAR = Total resistance changeAR(01) = Resistance change duein specimen.6 R(ge) = Resistance change dueA R(f ) = Resistance change dueR(T) = Resistance change duegages.to temperature induced thermal strainsto elastic mechanical strain in specimento plastic mechanical strain in specimento thermal resistance characteristics ofSince AR(T) and the difference between the based plate IT and thegageC'T will generate error, a curve of apparent strain versus temperaturewas made by gage manufacturer to compute this error. Hence the gage read-ings recorded during the experiment can be corrected by subtracting theapparent strain which corresponding to the temperature measured at gagelocation. These apparent strains for gages used in present experimentswere shown as follows :Strain gages used on Mild steel, T-1 steel and HY-80 specimens :Gage : EAP = - 107.84 + 2.90 T - 2.58 x 10 -2T2 + 6.79 x 105T 3 - 4.90x 108T4 ..... (5)Rosette :EAP = -77.04 + 2.56T -2.52 x 10-2T2 + 6.72 x 105T 3 - 4.64 x10-8T4 ..... (6)The sample specimen used is 1018 steel.47Strain gages used on 308 stainless steels :Gage : EAP = -158.79 + 4.13T - 3.20 x 10-2T2 + 7.67 x 10-5T 3 -538x 10-8T4 ..... (7)The sample specimen used is 316 stainless steel.StrainRosette :gages used on Ti-6Al-2Cb-lTa-0.8Mo specimen :EAP = -156.08 + 3.37T - 2.18 x o10-2T2 +3.79 x 10-5T3 - 1.26x 10 -8T4 ..... (8)The sample specimen used is Ti-6A1-4V.Since the difference in (T between sample specimen and specimensused in study of steels and stainless steels is small, it is neglectedfor lack of accurate data. For titanium alloy specimen, the differencein T between Ti-6A1-2Cb-lTa-0.8Mo and Ti-6A1-4V was shown in Figure 7.hence the apparent strain correction between room temperature and T Ffor gages should be made as follows:ET 2AP T (Ti-6 21) dT ............ (9)1For T < 400eF :LEAP = 103.92 - 1.87T + 7.25 x 10 - 3T2 - 6.945 10- 6 T 3For T > 400 FA EAP = -60.12 + 0.8T -2.5 x 10 T The true apparent strain becomes :EAP = EAP - EAP .............(10)(11)48a)a)ClE-qtoHrbtD a) 4a) H 0 C) E- Ma) srio aDH C'-4H-co - -4 vC49, s ~kL i-IV'-4\O O· 0o; OCd'-4Ir-r-I0rl(f~~~~ F1I-0 mr3~X 0 o.c: -XIEd o o\ (1I 0 er-a)It kQ, *°14 o C. a) r-I, r-FXIL P- L·IiI I I IL- IIN 614--I L,- L--I e 4 T50_ )tHOLIE-w:k-Etoo0oHH:P XO EzjI I - w l ----.a - t _· - 7 . ;'tclI _ !1Test SpecimenGTA Or GMAeleFIG 11. Outline of Experimental Set-up5152eelco Potentiometelel No. 320POn-OffClose to CalilChromel-Yellow (+)Switch Open toCalibrateType 100CalvanometerHoneywellNo.1508Visicorder2 F ReferenceunctionFIG. 12. THERMOCOUPLE INSTRUMENT CIRCUIT(1)PlateJunctionBoxFIG. 13. STRAIN GAGE INSTRUMENT CIRCUIT53StrainGageDummyResistorHeilandBridgeBalanceNo. 82-16- -2.5"a1" He 0.3"m a- kh~N- 4"- -_ nIoxa)a) P44 XP a)0bD rl-$4H, ,a w- )Co U21 P4to rHrz454Nr.- I-55Table 3-2 Residual Stresses Measurements on GMA weldedMild Steel.56Table 3-3 Residual Stresses Measurement on GTA weldedMild Steel.Readingsbefore cutting, 317503149830464315423150031513325453050830204307763122230384312663039730096One side cutReadings.,adir~s310003162130494316403133631364327213067330224307603130030427313193038929932ReadingsAfter cuttingj _,.308783137330528318543156631215328273058430146307843121830412313263035729930ElementNo.,,123456789101112131415NetReadings.-872-1256431266-29828226-588-42860-40-166II w - I _57Table 3-4 Residual Stresses Measurement on GMA WeldedSpecimen of T-1 Steel.Element Readings One-side Cut Readings NetNo. Before Cutting Readings After Cutting Readings1 30583 29216 29060 -15232 31321 31592 31076 -2453 31600 31822 31744 1444 30848 30920 31420 5725 31030 31448 31411 3816 29410 29238 29134 -2767 29623 30070 30158 5358 29825 30001 30164 3399 29354 29432 29287 -6710 32124 32271 32279 15511 31636 31410 31399 -23712 32706 32821 32749 4313 31088 31180 31093 514 31759 31296 31440 -31915 33054 32480 32480 -57458Table 3-5 Residual Stresses Measurement on GTA WeldedSpecimen of T-1 Steel.L L L o Lo 15" - iCto L -it -Lc ~e .`- 1I" 2"4 A- - -1" 1.5"Lo0 0I )0a)occd OXO.,L_J c59±IT 3/8" --I I IIIl--IIm-r60Table 3- 6 Residual Stresses Measurement on Butt-weldedspecimen of HY-80.Element Readings One-side Cut Readings NetNo. Before Cutting Readings After cutting Readings1 30283 30494 30705 4222 29905 30271 30083 1783 31105 31022 30889 -2164 30436 30718 30774 3385 31540 31849 31745 2056 31056 31108 30851 -2057 30333 29207 28994 -13398 29142 29244 29476 3349 30484 30448 30516 3210 29583 29506 29436 -14711 29560 29596 29679 11912 31070 31175 30907 -16313 29351 29514 29651 30014 29920 30163 29765 -15515 29459 29794 29972 51316 30304 30421 30056 -248'.4 71A t o ~ ~ ~ ~ ~ ~ ~ , _ 4 If Nr( nr h~~~~~~~~~~~~~~~~~F N~~Alf-.1 t5"I -5'1 +I- T u*ul t2-1 t"C -..ouS;'L~~~~- -- Us z . .. _ -CH04)oE.1-C)a)0CD P49 AndoHH r0z -P0)T r.E:u oC w0) )o_ U~~~hlhD0 *HbD +fu UEn rHCdSU4U)r-Hpt4610r-0E-40rlIol-cc-LC~~~~~·~~~~--L·--CL-~~-I- '9 62Table 3-7 Reasidual Stresses Measurement on GMA WeldedSpecimen of Ti-6Al-2Cb-lTa-0.8Mo Alloy.Element Readings One-side Cut Readings NetNo. Before Cutting Readings After Cutting Readings1 30495 29495 28626 -18332 31671 30810 31299 3723 34100 34236 32224 -18764 30958 30945 31184 2265 31824 32255 31436 -3886 30725 30589 30376 -3497 32930 33249 33450 5208 34916 34874 34952 369 37414 37320 37272 -14210 29574 29873 29934 36011 29644 29783 29809 16512 29493 29475 29426 -6713 31764 31964 31976 21214 31412 31459 31514 10115 30019 30005 29924 -9516 30134 30076 29974 19317 30486 30581 30511 2518 29471 29536 29664 -16019 29368 29404 29353 -45020 29326 29038 29211 -11521 26993 26626 26568 -1522 26993 27331 27789 79623 27788 27749 27884 9624 27900 27704 27668 -232632. Residual Stresses wexe obtained by using equation (2) andequation (3) and were summarized in Table 3-8.bVAb6C-0z0\.OnCQt-4Z$:iCOC;C'-0OGDocrtVU)C\iIC-CVN03rd,o1na)a _rd r4U) *rlw- .. _zE t{0ooOIC)IcDC)H00a,<.. ,E 11C\lC0--t0O0C-C\1ONI-q\DHCLI-a,19H\t-qI-tla)^V WI x_. E- 6XC\03OC~lM0\a,Ha)^@ rl0 _IEt (n00_oC\J\0GDI;onCOICO0GD4ooMcoIMOCt0COGDcDCMO\01C\IC"ICV-CVInoH0teC\ClV4t4NI5C--lI-t,rdrlCo.HU'E-1(12.r-a,coHa)WrdU'CH0F-ia,HEacockCOa)CHroPCdUp,-- --IV. RESULTS AND CONCLUSIONSA. RESULTS REPRESENTATIONSIn order to check if one-dimensional computer program is avail-able for analysis in present study, strain gages were mounted alongthe welding edge to measure the strain changes in each specimen atdifferent positions along the welded edge. Figure 18,19 and 20- shownthese strain changes at each points during welding for mild steel,stainless steel and T-1 steel. It is obvious that the history ofstrain changes at each point are similiar, that is, the results pre-dicted by 1-D program may be used for analysis.In order to compare experimental results with that predictedby 1-D program, exact heat input should be determined first. Figure21 to Figure 28 show the maximum reached temperature at differentdistances from the welded edge for various specimens, and the effi-ciency of heat input were determined as follows :Materials Welding Method EfficiencyMild Steel GMA 0.55Mild Steel GTA 0.55Stainless GMA 43SteelStainless GTA 0.51SteelT-1 Steel GMA 0.55T-1 Steel GTA 0.48Ti-6211 GMA 0.35HY-80 Butt Weld 0.696465As thermal effeciency have been determined, we compared the historyof temperature changes and strain changes during welding for both experi-mental results and that of one-dimensional computer program. Figure 29to Figure 45 show the comparisons for both longitudinal strain and temper-ature changes. Deflection changes were also measured by extensiometerat the center bottom of the specimens, and were shown in Figure 46 to52. Since 1-D computer program is developed by assuming the plates wereheated during welding with no deposit metal, hence the comparisons forGTA process plates are similiar but that are poor for GMA process plates.Residual stress distribution predicted by 1-D computer program basedon the assumptions that the aera bounded by tension stresses and that bycompression stress are equal, and the melting point was reached duringwelding so that maximum residual stresses equal to yield strength of theplates could be obtained in the welded line. Residual stresses measuredfor GTA plates are less than that for GMA plates due to less heat inputduring welding. From Table 3-8, the residual stresses measured in lowerstrength materials nearest the welded edge were higher percentage, whilethe residual stresses measured in high strength steels were about 50 %Of yield strength. Residual stress distributions were shown in Figure 53to Figure 58.a0'-I00Co.r5 000a.1Q 0I00011-FIG. 18 Strain Changes for Mild Steel ( GTA ) at Different PointsAlong the Welded Edge.66eli0FIG. 19 Strain Changes for 308 Stainless Steel ( GTA ) at DifferentPoints along the Welded Edge.67OLo0orro4 00..40lI4 2 6FIG. 20 Strain Changes for T-1 Steel ( GTA ) at Different PointsAlong the Welded Edge.68OVT\OO0OI.c)o00$40041-4Id)+r 00~'CJ-P 0U)0-40)riOOp4C)0OD0)HCd4)r-OdOHCd0u 0 4 a3I I I I I I I I_~ ~ ~ ~ ~ ~ ~~ O0ON -o :s:) M,o 4O Po $-4 4( :Ca, aO · 2C)HrX4o,-V-4( HI ) HODZ IaCnSM nOH XVMV ONV¥LSIG69I L _I I j IiaI I I54 ao r.d Hr-I'd p4 e01-- '~ 0 r-i' ES E HH1V 1^ .' 4 Al H0 :N - 0\ .r-Io a, 0-V54 4-)Cd A0 m Qa) p4o 0 (1(1 Cd El (1< Cda) PX: O NVI% Cd E4NN~EU1-) 10u2EI C0 Cd~CV) E( 4ouI ) pga pple WmoJi SBMy eous!aC70: t On (IQ V-4I I - -L I I - I I a F,04 043 HccI- r- · . 0,el 0 4-.· 0 1 -4 cdo iA E4 )I1 o i aI I I ISI I Io00CH43--4 -o 4-'oa0o a 0uc o'c EcU .--0 )oo o 0o *HN Eq ( HONI ) eapa papletm u1mo XRMV aouvesla71000or-o0O~H'H-,I I I I Int.IVr0Ha)0)(I).H.rI4.)00Ia)'0 -P: :3: 3d H0) &d0Vt'IICIII I I I I I I ICNI It ( qouI ) aeu.r peplia oJ AVey s9aous~STa72IIIIOtrN4,O00OoeEo-oo ctN C0 V' 0C)P:H(',00E-40)I3Io_ _I IItrtH4l-4U")0 ICtI11_I I I I0C cu 'PI Id® :>o -Po ,1r3I t 1 I I I IIV-4( HOMI ) eaPa PapTlM o~& AVMY aou 9sl.a73O4Q)0O4OO c:40 00E-400oHo 0~oC CEO Xo cdC\l0.,1OH4-IQ)00Cc40r30HI t I IT1 r _mLzmk oo a3 oo 4' o r-a) * ) £ ,o , laoo ~ ~) cr-(1)-PV) II oq 13I CHI I I I I I ,I I-4I I Io1o.,-4o ' *I0 Cc i0 0o Eo e EC) =ciijjir-N xCi;-I oCQHPL-0( HNI ) 2Ps PPTO'M moa0 keMy aeou.sra74I4)00I>-:V0\\04-0Cd4-)410.1-t'.1-4olli-10O m)CDHds 0o Cd0 HEq 0000I I I I I I I I I I I Io0Oo-4E+4 ,-O oo go ' 0 C) do @o 0'-4Hr Q0( HOMI ) aeu.l PPlTM woaJI XAMyV 9oUB4%S.(75_II76- r qHm8iCVrla)0)CaHr-H.,H0 0 O o U)0,-. IC%it,A4) .-d HEQ 1-pCHod raO p.ci)-p.,1U0U)ror-I+.i00I.,-77-I I I ..HHF-jI-o o o o oo o o oo U) oIHIAC\ZOHO O0 OOI-_(vnII I78to5:,J0-P5.40PixNxa)pa)0)pCHrl0aa)toHcoco.~IF-iCHto0H0OO*--00 0IIIO OO O') 0I79- Nl llIpa)a)CH.Hr-Pa)Hcticov-rlV)HaI03oI-i0er-IO0--zH80HI-00o-0o 0(LEILa:imv(800ooI 4C\l 1 II Ia)I -Ifr B / HCOH0I l II-i 9 II.0) IriEI-0o) o U)o o oFa)a)CHAEH-ta)a)CHprIIIiII'IIIIIIIIl -.~~~~~~~~~~~~~~~~~~I HF-1:081I IItoc4'CHr-iOC)i--rJItI4)wzH.rIa,xI--HmH:E0O0-0oU)!Iw )82U)Oc3O0kF)C-rloCHp~I-Cr_O O O O OO O O OI I-OOb,4o·-r~~~ra)C~~~~~lTc0~~P00o~~O ~~hi0 NCdtQa))5HO hI9.xnq.,ejadTra9,880tzi m-4.HH W,r-¢- Zz MZ0 0 0 0 0 0 O0 C 0 0 0 0-n cl40 0 ( ) aJnqeaedwal UItxsoJOsI ''C) O O O O O O OO o O O o O OCA O4 ~-4 %th xt% OwjvxI~ OJOIIAI0. eanvaiedmae89o1:E-4-4Z H:w rH 0)UZ vEnH wH HH 30PL44-> 10E drz CI.H- H--4) )P, rHW 0111 O O O Oo O O OIt t C dl (0o wcO N0SHr-i4)4)o coo to)N-E-4OV.0 0o 0'-IO O O0 0I'-4( !o) aejn jaeduea&9oI0p4EE-4H4ZCrQ IVZ0H _I- I0 wH WH Hns"'UT~x~ssjO~jjo 0 0o 0 Om N '-40 O 0o oo uyvjasoj3lw91(+).H HI !III)p7o:;Poor-4apo>p:HH H°:BoC7r E¢XH WE-¢ $4:E OE1 Hn O wI13:0 o oo ol-aanq~awadrraj,92· 13- -iCH r(E.X-43E-ioH IE--H IE- ZIHD3ZOi-to o 0 0 0 0 0) 00 0 0) 0 0 0N V-4 ~0 V V) 0'- I OeJOJ~iU18 1 L uO J j.SOJ3OWUT-eaqssoaolainqrrviedtual0o 0 0o 0 0O o oO o O( NI )JUTOdI aeqaqD qV UoToalGTJ93wo00Qa).Ha)X4D-rl-Ha))coH.+0a-oH0,0HF(0Cd W& Hr- E-a,a,xF4P:i/0o)OCl0O-( NI ) uoa.,GTJ@94a,,-c;E-i d0 . a,Ha) Oa) HU) Urd Irl I· :: arrd-4aa,,-P.-C)at~rdoQ)aH00ooooIII i95Ha)a(,a)H0oCHoa)a)C-)Eic)coC)tHa)CO,0 o0 o o0i( o o o( NI ) lTua3D 1-8 uo1^DX1JaI*riAr-I0U)HF-I/I\96o///oOoCOo1-( NI ) a,.ua3 e uoT.oa-TJG0 aH HU) rda1)V) -p*rl HH Ud dU)o In I/////a)Hrl1)coHr-cocaCHr3_rlOOci)O0r-O,¥i-ooI' C)0 0O O Oa) &i a)H rl) H AI II0/1/I/( NI ) .aequao IV uocsJgTjaoo\O97ooOoC UU)HE-/0)rlEPHHr)a)U)EH)-Po*rl)00))CdoHU)0H00COOo--0co00o)co0ooooo~o uo XnHE---a)-Ia)a)-pcxiEl4Hoa)a)0C)a)&OcAHa,pxO0OOHFr-O98a)cd--E a)H Ca) Ha) Cd& 0I IE IooCd:&.,H0)ixwiOOO80Vw~!I//IIIS.o0C00o;000 OUHHO00o4( NI ) a@quaS Iv uo osfalJ@G990rQr0E-H0H1a)-PrDo0IdEq0uCMIr-t.rlOH\0a)a)a)E-iHoa)0U)H0I0-00oMild Steel ( GA )A Experiment0 Calculated Value1. 2 4Distance from the welded line5( inch )FIG. 53 RESIDUAL STRESSES DISTRIBUTION IN WELDMENTOF MILD STEEL ( GM )1004o30.-4C,2010Co0)WCD0)5$4'HrqCd010.H0-100101In30'120rUtou 1 0a)c) 0-100 1 2 3 4 5 6Distances from the welded line ( inch )FIG. 54 RESIDUAL STRESSES DISTRIBUTION IN WELDMENTOF MILD STEEL ( GTA )102T-1 Steel ( GTA )--- A Experimento Calculated ValueI ~ t IaA-4--)1 2 3 4 5 61008060a.,w0qa)a,uxtI2cSa,C)wS40200-20FIG. 55 RESIDUAL STRESSES DISTRIBUTION IN T-1STEEL AFTER WELDED ( GTA WELD )1034a Mlut.8C60.-u; 40voEntoH 20HWC0T-1 Steel ( GA )Distance from the welded lineFIG. 56 RESIDUAL STRESSES DISTRIBUTION INmu4030*, 20a)0a,s4 10r1U)-10-200 1. 2 3 4 5 6Distance from the welded line ( IN )FIG. 57 RESIDUAL STRESSES DISTRIBUTION IN WELDMENTOF HY-80 ( BUTT WELD )104105Ti-6Al-2Cb-1Ta-1Mo ( GMA )X ExperimentCalculated Value70 1 2 3 4 5 6Distances from welded line ( inch )FIG. 58 RESIDUAL STRESSES DISTRIBUTION IN WELDMENTOF TITANIUM ALLOY ( GNMA )503o.4 300a)uta)-10Cd106B. CONCLUSIONS :1. In comparing temperatures measured with analytical predictionsby using 1-Dimensional program, good agreements were obtainedif sufficient informations of temperature dependent propertiesof the materials were used.2. In comparing strain changes and deflection changes with onedimensional computer program results, it is found that thecomparisons are similiar for GTA welded plates but not sogood for GMA welded plates.3. In most cases, the analytical predictions for strain changesby using 1D computer program are tensions when arc just passedover, few seconds later the strain changes become compressions.This may be acceptable for lower yield strength steels ( Mildsteel and 308 stainless steel ) but might be not for highstrength steels such as HY-80 and T-1 steel.4. The residual stresses measured in mild steel and 308 stainlesssteel weldments are almost as high as the yield strength ( GMA. )while the residual stresses measured in HY-80, T-1 steel andTi-6Al-2Cb-lTa-0.8Mo alloy weldments would be about 50 % oftheir yield strength.REFERENCES 1071. Boulton, N.S. and Lance Martin, H.E. "Residual Stresses in Arc-Welded Plates" Proc. Inst. Mech. Engs. (London) 133, 295-347 (1936)2. Fujita, Y. "Built-Up Column Strength" Ph. D. Dissertation, Lehigh U.August, 1956.3. Masubuchi, K. "Control of Distortion and Shrinkage in Welding"Welding Research Council Bulletin 149, April, 1970.4. Masubuchi, K. "Residual Stresses and Distortion in Welded AluminumStructures and Their Effects on Service Performance" Welding ResearchCouncil Bulletin 174, July, 1972.5. Hibbit, H.D. and Marcel, P.V. "A Numerical Thermo-Mechanical Modelfor the Welding and Subsequent Loading of a Fabricated Structure"Dept. of Navy, NSRDC Contract No. N00014-67-A-0191-0006, TechnicalReport No.2, May, 1972.6. Masubuchi, K., Simmoms, F.B. and Monroe, R.E., "Analysis of ThermalStresses and Metal movement during Welding", Final report underContract DA-01-021-AMC-14693(z) from Battelle Memorial Institute toRedstone Scientific Information Center, RSIC-820, July, 19687. Andrews, J.B., Arita,M. and Masubuchi,K., "Analysis of thermalstress and metal movement during welding", NASA Contractor reportNASA CR-61351, prepared for the G.C. Marshall space Flight Center,NASA, Decenber,1970.8. Masubuchi,K., and Iwaki, T. ,"Thermo-Elastico-plastic analysis oforthotropic plates by the Finite Element Method", Jounal of thesociety of Naval Architects of Japan, Vol. 130, 195-204 (1971).1089. Muraki, T. and Masubuchi, K., "Analysis of thermal stresses andMetal movement during welding," Phase report under contract No.NAS 8-24365 for the G.C. Marshall space flight center, NASA fromM.I.T., June 1973.10. Masubuchi,K., Andrews, J.B., and Urushihara,A., "Finite ElementAnalysis of thermal elastoplastic deformations in Butt welding,"I.I.W. Document X-692-73, 1973.11. Toshioka, Y., "Deformation of quenched steel bar" at M.I.T., March,1974, unpublished.12. Toshioka, Y., "Effects of material properties on residual stressesand deformation of welded part", at M.I.T., June, 1974, unpublished.13. M. Nishida, "Calculation of welding thermal stress and distortion"at M.I.T. summer, 1974.14. Masubuchi,K. and Martin,D.C., "Investigation of residual stressesby use of hydrogen cracking," Part I and part II, The weldingJournal, Vol. 40, No.12, Research Supplement, pp. 401s-418s, 1966.15. Studies on the characteristics of the residual stress distributionsof high strength steels" by Y.Kikuchi and his group. Jounal ofwelding society in Japan. pp. 24, sep. 1974.16. Iketa, Y. and Kihara,H., "Brittle fracture strength of welded joint"Welding Research Supplement, Welding Journal, Vol.49, p. 106s, 1970.17. Akita,Y and Yada,T., "On brittle fracture initiation characteristicsto welded structures," Journal of society of Naval Architects ofJapan, 117, p. 95,1965.18. Klein,K.M., "Investigation of welding thermal strains in marine steels" Thesis for M.S. degree, M.I.T., May,1971.10919. Klein, K.M. and Masubuchi,K., "Investigation of welding thermalstrains in high strength steels for marine structures," paperpresented at the second international ocean development con-ference held oct. 1972, in Tokyo, Japan.20.J.B. Andrews, M. Arita, and K. Masubuchi " Analysis of ThermalStress and Metal Movement During Welding" MIT. NASA CR-61351Dec. 1970.21. USS T-1 Constructional Alloy Steel Handbook.22. Mckee, A.I., "Recent Submarine Design Practices and Problems,"Trans. SNAME, Vol. 67, 1959, pp. 623-652.23. C. J. Schrodt., " Fracture of High-Strength Welds In High StrengthQ & T Steel" O.E. Thesis MIT. May, 1974.24. "Metal progress Databook" 1974.25. "Military Handbook -- Metal Properties" Aeronautics and AstronauticsDepartment Library.110APPENDIX AExperimental Data Reduction ProgramInput DataU,-JW1-..W,Cz-r1- -JaZAOWuLiWOW a:z cun, zW J 'O W0....* WI~Z Uo -W wI ) W->LL 0cr0COI tL 1 d L la- S 0. L 0 W 4.. s-<1. * :(/ ( Z C O Q Z a OO a J _J o o a WW- O Z a tfaCz z 0 (AU -> .... ........- N C) L D. 0 Z' Q J > C J C Oa _J a: C a:Z 4 < O 4 <4 .. " -- D0 -I -.20 .N000000 0 0cLAn cc a o cc_~- _t (y 4) Z~ \C 0* O n o o o L* 0 *0 0 00 0 0 0 0 CD O 00C PC CL 0 0 CO 0.· % 1 a," CZf* S 0 o m O0000 0 0 e0co o o o oOOq OU OJ N (%J ~ Oq t* 0 * 0 * _('m 0 o*000 * 0* * 0 0o cra- a0 _114eOeO o o O,--,* 0 0 0 r(1 f-(30(30(30a a0 a a0 0 0 C cc cc cc 0c cc o ccc ('o0 0C* , It L- Oo, -0- u,0 ~0 : 0, 0 0 ,- - ,0 0 * 0 0 0 0 0 0n ,* 0 0 * 0 0 * 0 0·* · ·* C) ·O n · o O ) · N o a 0 fo 0000I I I n 4 n-0000000000000000000* 0 * 0 * 0 00 0 0 0 0 0 0 O 0-IL Io 0· Io o* 0In I Lr- -o 0o o* -fl- c* \0- -. r0 \00%',0;* 0 0 0 S c a UI L O N-' O0 00I I I I I IAJUJ WCOCu)~O Ct) 3~JO O O C O O OL I000ol. U " LU I I* 0 * i e ' 0 0O O: O O Ctr t- O~C CL (I* 0 rO C o 0 * * *. T o O' o ' Lo I.- LL. LL LLO ti N ?I M 4 C_ O OI I 03 0 -- aoI0 I0 D00 z C0 0* C * * WL I | -"O L O O Oo N· I-U n UM * ) a0 0 LL: 0v 0 0a! co cc .- OC f'I r_4 4 4 4 Z 0 -4 A N* 0 0 ) LL aN cc cc dr cr I ·- 0.- cl0 0 00tr * > 0 0 *! 0C N 4 · C , nN K P- r_, oC N r-N t4 U4 L C)00 0 0 0 · 0 00 0o 0 in 00 00· o 4 4 0000Ln o N 0o·* 0 · * ·0 * 00000000r- ~. m-. i0 0 0 00 C LUt C LO o Ln O ur -(4 Ln LI I I* 000 0000 * 000 * 00 0 0 0 0 000o00 o0o000oo00000 00 * * * *0 0 * 00 0 0 0 0 0 0 0 0 0· * O0 ) *0 LO* *o iq , n O0 Ll N t - ·rl r- M* 0 0 0 . *oi n Inr-L a L c; ,Cr 3n ail | t * 0 0OOOOLn * * 0N O c Lro Lr, O' "I I0 0 0000 0 0 0 0 0 0 0 0 a 0 00000000000000000000oo o o ooo oooooooo· · · · · ·· · · O t · · · · ·0 0 00 0 00 0 0 0 0 0 0 0 0r M M M M M M M M M M M D cc W3 r arl r- - fl- r- I r C r P r r r w m ·C r I * * M , C MO r 0 r r r l- l- l- - r- P.- rlc 0 O0 0 0 00 0 00 00 r4 N -_4 -_ _q0 0 0 0 r COa % L ULn D O% Ln- - -4 --* * *~ a_ * *; * CZ* 0o 0 0 0 o r 0 M 0D N M M N - Q r-C NNNNNN.-'.-4. .-*, L* * * 0 * MW LP -mw 1 CC 4 CON%C 1- 0L, _ t _ _ CoJ M) '115*. * * *Bo o o L InCi 0 0a 0 a co* - * - - -00000@* 0 UL LA LA 0 00"O _ Jn aJ (Oa a l N N M -4 -4 -4 -4 -4 MC00NN~')u, * * t no mOU Q tUB * * iL0C O oCO r-- 0a InNU .-* 4-tC 00 000000 C0CO CO O 0 cC) OO O O O O OO O O O O O O O O O O O O* * * * * * ·* un Ln o Ln n o*OO 0 0 C'n U'4 In r- -c CY- - N N N CM * * L* 0 0 0 L U A0 -i UL0C Ln 4Ž 0 M LA -- i a tn* * 0 CI * * CD I I I I I I I· O L OLA U\ L 41 L V;r'n -4 >* B", S)* 0 *0o tn u o o° -4 C 0 0* *00 Bo cLL n CO- I I I I I* tn C0 0 1- U' U ,O I, C 5_- , I t I I* * * 0 0 * *0000000000000000C00 0000O UBal LOOLAVW0'* L *V 0 4 L n LaOu-N 4 C') L 'C C'* * \o *n 0 *n -4C C -4 0 4 U)L -4N I -- N I I00 0 SO*, I a * s s * * s * I0 a .* f) . ra .r 0 cL LA Co *no 4T-m N o* InD aD m C 4 C-Ln 4 4CJI) I I I I 4 IL I I I I Io~~~l III ** U) a Lo * Ln c O4 Ir- C* * -' Ml ? 'O C *- U 4 ;s L I I I I C I I I I* * * 0 0 *: 0 0 00 a N ) 4 LA 1 C N34 m \: a c0 0- U) 0Ž t 00 0a 4 0 %-* L-N O O0 0 0 0 0 L0 C x C0 5 a B B B o 0Ž 0 c4 ' - 4 ' 4- N ( CZ ) 4 4 CD L A LA 'C %C L C 4; ~ C' C ; NIx 0 ·- c4 c c - c N M 3 3 M W CO \I 3? c C M N Uc* 0 f) * * 0 , 0 0 0 0 0S 0 00 LtW 0 C c \0'C ' C \C \C f- r- 5-I- C C .-C!-0CO (, - \J* * * S0000c '~C c r o116U)Ir0.IIJ0-4a0za0Cu!o4I-.-JII--J.- 4 N 0a I I I LA- r- I-000IW L bIj Ii0oo* 0* 000I I11I LL: IO O0N 0G Cc c: O ) N0 0 00 c0 cIIz *AJ *~c, CO G :~, O.-O O O0ONO I) !-i00o ac0 N CC.N 117* - * 0 e * - * * . * * * 0 * ·0000000000 C C C 00 00000 00000000000000000000000*0 C 000* Ln a n n a nO 4 w c I IC CC C efl o C01n 040'-Lar.0'()M P_1 1 # I* * ·4 Lf0L · · ·O C\J NLU) O f - MU C _, _- _, fy CC CoCO *OO a0LA L L aOOO;;L A a' -n C'- c a o'- uC N r u c u soOI -4 ( 6 -4 N' Cu o (n M M N MC - L iLO C C C C C~0 N c ( ,M rn %D4( 3 6 1 C cOSI I i * r4 r4* t O in o N in0, I I \( -c COeL * LA LAC I C a 0 LA LA rL*0 0 ALn o oL)I I I IMrC I IrvosI0 * 0* II Ioe 0~' oC Lr, 0* Irn c aOIn -rnmN *Ln I O a*· ein 0U O L O O 0 I C ( CL C C ' C Co ~ C* * ' L C . 1 r u C. W -, F I I I I I I I I *; *4 *. * * *4 * *; 4r 1 PI. C) n r- m ? oiC r C 0 C Ce e r r eC r C C C C C - Cr eC CC · C C C C C 0 C C 4 Crl , i .r rr (~ f~!,r ~- .* aO aC*0 C C C C e 0 C 0 C4 \ C M 0o * - · · a -· - e\ '- ~O ',eC L. fI 0- ,- -4 M1 0, C C CO N m m N m 4 4 m % f'-* * * **0'TI - -4 .m .- -7 - -- - - -- -- -4 -4 - - - -C C'aD P-4 -C C000 0-- N o r- *N.: \ A I0 'Lo 0 LA*N C Losas1 C o0 IOh1 C-.rI33JwI-Cra-J-J01J Oa]lrI0000!LL L MW0 00* I II I,0 D 0,' a, r'I I 000a ,'mniaO ,n I O' CC L; PtL Lon o? o CQ _U) F _ _118APPENDIX BInput Data For Temperature and StrainCalculations0. 0 0 ro co 0o0 C e LnN(J 0 · · ·*-c ,O- C - C O 00_ It4 C O-4 0 - 4C> Q -Co 0·0 Co -- 0. C . 0 C.ci o- cc _ a a:LA0CDCL* C *C C CL 0 N * *Do * L * N * c 0 C N' !C - 0OLar -oeCeN 0 0M CDOC; C a Z nLC-1LA *YM - 0 (\ 0 -L , 0'Lro f- 't -4 0 N N I -- c0L)Co*aco Ln0J ·e Cr·r I- Orj,-4.o0O I", * r- Lno Ln N · o0· -,-C0 * 0 n0n OL' O Co'· .a ea .2 *,r C C cO *r * * o O-C .c : f c0* I a * L' aOL, AL J * * *-0 Oa .PJr-4 .t 0r,-- r1- cN r- CoC00 zC0(5sRI * flr%.0 %.AO 00 4 -z C -ON MLn* (1_S O ·oo o a L0 05 t (V c*Ln tfooN M -o00 *5 eta * c · *3 o _ o \U C U LAC 01 0 c ~ N * M 0 * .o · nJ e ~ Ca * C,(50-ONN'OCcxAI * r *' LAmCJ *R *o *e0J ' 0 N · e OL G C ·C *O L0 * N Z r- : -o 0t en c 00LA0· 4· OL 0 J* *00OINaO a N CJ00 L 000O-- 0 O O ,C-oALnLn0C~~00 0CL 0t c0 *fM aL a O o* * Na N, * N% tmJ r~- Ca LP UL CCCO CJ * o * -o e~ e CD ·C ..+ O,-OCN N Oc o0 CD0 · eaO0 s, N * 'CON * X- O r- no D LnC) \0 M M * * )wLn-C D *c0 f 00 eO * 0 LnG LT. . 0et 1nX 0000OLna, M 0Ca 0 N N0CLnA o0 eO0O 0 00 Pcc * * * !c o o o 19c t11190Lr,0n.LC,LnC,0CLn0-LAp-C0L000zLAI,.-4¢1t;tNNo 0-4 oo oC .,- ;4 o -O--C o00 00 0 0 aC ·-- .. ·C C-.-.C; -* *oN.-'_oo 00 C 0Ni .Ic-cc -0CD0La-0oo L cl ao C*f r *r aC Lr OCo MCN C . 0Q 0 _O G C _ 00Ln4=o OO " CO eO -- 0 -400,tCD 0CO - -0 0 · C t ·O- O1 *1 . Lr· O -- _- ·O N -- 00U Lr a0 MN 0O i * -00 r --cr a,--f C0 t O 0* rra N N* C -* iC) Z C>-a0 0Nr s r 5o N N _ oa L0N NN * * -0LnUIa00CD0Itco - O~ ac oem *M cr- Do .r- · · *o ei · s 00 ·rF-o,-oNN.-o* 00o ~Lo N0 ·'C 0, -· O Ln r N')C-ION, · · ·c - O n _c - O0 a, *OLn e _L~ OU ,-CC n L-N 0* a'N N 5 Cco -c c 0P · *in t 4 o N a 0LOC)0 In fl NLn0~Ln00 c*O Ln n L·4 N ON ' .c^ * 5 : oaaN-000 n L aN( · C*FI * O tlaJ ~0 aoON * 00c .4 aNOC-NinM c OinL0p·* r_4o Ln-- O ·Lr N 0NonaN N0LA 00 1 0e. -- C 0 5 s (L.C 0 s0'CC-4 fP.JI-L-N U)CV)-Jz4N' t.. I oLLno a o*o * 0 ·-t o c_ cac CCO NLAN s'0O Ot-i st M q O 4 oLn4 * 04 0- c N ao o0OU) ·a * , Ln*s * 00 m nONON s OM ", C- * M CC, M'*-O o- a uno ( s ej . .o0 C -0 , a r-Qo _oN nV 00o1200Ln00LnLnC1a0LO00in5000LnMo00inIn r- N0.>, inJO 1 0 L I) Ln U) c WOJ OZ N-4oe(~ MQj~~ILa5,00Ln0"0.. .0 0C)C,- 0eO CaM0N NOONo % - Lro r-oN* r- 0 L 0 o Cc , a S @ u -e · .N 1- · 0- L,-4 0 %i 0 o-4 = 4 N o 0 o0LO0CLne -0 ·Gi CMO0 .-4 C C.Q ·CC M NI Oc r- o0 LI a0 MO C OO0LAO CD N ·* *40 el~ eP-L-L ·ap a -o 0 * r- 0 rl- c c4 C: 1- a, ol--0LaC00Lno r-a r- 0 C,0( -N * Ln ·N a r- * 4 Ln 0 S0 C c f- ·orr- Ll -ctC, . .M0Oo o N CC0 0'C0 L ( a o 0 t . %'0 · 0 CC O-4 0 a- 0O L. cc M*LAo * *ou *eO-L 0 0O e t'- e! * I' ·I-o.. oOUO', Ir-O00M .· e a l- Ine O In ru'0J * N0 c .(: . r-- . .O 0o - a N f 0Ln Ln o 0o 0 o 0 C -0aOSe0La3 N ao Ln> * *r P*F F * a'*N * 0 CGO0C ·nU ma *a·0 N *sCC L o-4 0N-F CC LP )0 e(N O eN -( o-0fl C C- 00J L.J0LALnO0 aSO. -q 0Cj Q *o 0 0Lna0 0a0 ·0 0% C! * M -\- C-4cc a, '"et L N * 4 ·S oio.-CrOLUJS. 0 2U) ' r~ ("mo CZ .- . 00 c 0,- . 4N N L LAL I *- C C0 0 j I- - -, - ( "1,-- * D %0 _c r4 C0LAO00 ·CD _- oC- 0f, 0aD0CD0 · 0O * O"-L LAO * P-J0 LLUt0 I.-*0L 00 0 I OLn _N A s0 - CC L> 0".(v .c0 .cU) N *0 0 ml .*0 * OL a4 ( * N o-,o *N M o*4 -" a -t L o400'N,.-~ OU-..0t O ,, - -, *_ * C *oeO, · 0 a · -( C C C- 4'r, Cco *a, o *- c N0 0. - ONN, Ca i Cl M0 U'_ X Co * o C Oi N N U, Ln.r- C> i a 1f \C *00oLrn - Nr000)e OO Co'l N0* CO00a0 ·_ OLa .OO,C- 0 0 0)-1210LP00.L00LAC>UE0OV00',10%OLnLOLAo00LPr- *o* M - 0 * CON 0C 0 IfD sC%O * Cj N-4 C@0 ·o(N C0f-tn *0M.,C a' _ %C ,c t L Cr- r ",C M Ln N l* O C-_o c ( , 'to 40. CN tr-LO00C)W,0LALO0.40 *I O00 4.4 0 * So · e LoO O0 (C 0%'0N *_4 o c N LnCWAac OI * e0- e0C C ( .OLALno-0 C 0 0000s000 -0 0OOO*0ooo* OsCy _z L0LA0)c, a* NOLAO nL0 LAO,n CC,N O 0Nq rC0 n -r r OlC U X e N * CLOP 4 O* ·00 N *0o 0j 4O C N f- 0cr hial LLI -OAOCV) 0: O-OI 3:oO¢ I NL' r-- qC C- 4 0u N .N .0 C o c CN a o o<40' 0 N 4 Ln N 0 o -4 .0 -( 0 00 -O N OM r- ( 0·- . C C CO-,'I(~O · oLnA0 ®C)OLn .00000LPA0La00CO(3 CJO U1220c--40%CC00'C0SC0I40000N0UsLA.0LCG,0Ln00 GO _a Co -40N4c~0LA1a0LA0 0\e0010O 0 -I t Cao- 4O oI 00J D4 Z O>-40 C O(: c 5'>- 0 CCLALn ·-.C Or 0_ c0 *-*0

Journal Article

Abstract

Extracted data

Journal Article

Abstract

Extracted data

Related items

Related items