WELDING RESEARCH Effect Of Martensite Start And Finish .
Payares-Asprino Supp Nov 08 corr:Layout 110/9/0810:49 AMPage 279WELDING RESEARCHEffect of Martensite Start and FinishTemperature on Residual StressDevelopment in Structural Steel WeldsThe experimental electrodes with lower Cr-Ni contents were found capable ofpromoting compressive residual stresses in weldsBY M. C. PAYARES-ASPRINO, H. KATSUMOTO, AND S. LIUABSTRACT. Martensite start and finishtemperatures are very important in structural steel welding because they controlthe residual stresses in a weld. Tensileresidual stresses amplify the effect of applied tensile stress. On the other hand,compressive residual stresses are algebraically added to the applied tensilestresses to result in a lower net stress levelexperienced by a weld, thus inhibitingcrack initiation and increasing the fatiguelife of the welded component.The residual stress state, i.e., whethercompressive or tensile, and its magnitudewill depend on the expansion that accompanies the austenite-to-martensite transformation and the thermal shrinkage dueto cooling. High martensite start temperature and low martensite finish temperature will both minimize the effect of transformation-induced compressive stressgeneration. To obtain a full martensiticstructure in a weld metal within an optimalrange of temperatures will depend mainlyon the filler metal composition. A newtype of welding wire capable of inducing alocal compressive residual stress state bymeans of controlled martensitic transformation at relatively low temperatures hasbeen studied.In this study, several low-transformation-temperature welding (LTTW) wireshave been developed and investigated todetermine the martensite start and finishtemperatures of the welds. Also studiedwas the effect of the martensite start andfinish temperatures on microstructuraldevelopment and hardness in single- andmulti-pass weldments.IntroductionIt is well known that high tensile residual stresses near a weld decrease the faM. C. PAYARES-ASPRINO is with UniversidadSimón Bolivar, Caracas, Venezuela. H. KATSUMOTO is with Sumitomo Metal Industries,Ltd., Kashima-City, Ibaraki, Japan. S. LIU is withColorado School of Mines, Golden, Colo.tigue performance of the weld becausethese initial stresses, when superimposedon the applied stresses, elevate the overallmean stress (Refs. 1–4). Several procedures have been developed to relieve thetensile residual stresses in welded joints.Postweld heat treatment (PWHT) andshot peening are two common methodsused to improve fatigue properties (Ref.3). Another way to reduce or eliminate undesirable residual stresses in welded partsis to modify the welding process itself. Forexample, low heat input and small weldpool are known to reduce residual stress.Physical and mechanical properties suchas heat capacity, density, thermal expansion coefficient, and strength of the basemetal and filler metal contribute to themagnitude of the residual stresses generated in a weld (Refs. 4–6).It has long been recognized that phasetransformations in steels can radically affect the development of residual stresses.For example, Jones and Alberry (Ref. 7)showed how transformation temperaturesinfluence the evolution of stress as a constrained sample cools from the austenitestate. It is significant that their experiments showed that the residual stress atambient temperature is smaller when thetransformation temperature is reduced.Ohta et al. (Ref. 8) designed a weldingalloy containing 10% Cr and 10% Ni, withan exceptionally low austenite-to-martensite transformation temperature, TMs. InKEYWORDSMartensitic TransformationMartensite Start TemperatureLow Transformation TemperatureWelding ElectrodesLTTW ElectrodesWeld Metal PhaseTransformationsCompressive Residual StressDilatometric MeasurementsConsumable Developmentthis alloy, martensitic transformation in anunconstrained specimen starts at about180 C and ends right at ambient temperature. By contrast, normal steel welding alloys have transformation temperaturesaround 400 –500 C. As illustrated in Fig.1A, the net strain on cooling betweenTMs and ambient temperature is contraction in the case of the conventional alloy,whereas there is a net expansion for thenew welding wire. As such, local tensileresidual stress results in the conventionalwire and compressive residual stress forthe low-Ms alloy at ambient temperature— Fig. 1B.When fatigue tests were conducted onwelded sections, the structures joinedusing the low-Ms alloy weld metal exhibited much higher fatigue strength (Ref. 8).This improvement of approximately 20%is attributed to the compressive residualstress, which reduces the effective stressrange that the structure experiences during fatigue testing (Ref. 8). The achievement is based entirely on the fact that thereduction of the transformation temperature allows the expansion originated frommartensite transformation to compensatefor the accumulated thermal contractionstrains. The improved results and the substantial benefits are expected to bring radical changes in fatigue design philosophiesfor structural components. This effect hasrecently been confirmed by Eckerlid et al.(Ref. 9), Martinez Diez (Refs. 10, 11), andDarcis et al. (Ref. 24).Low-TransformationTemperature Welding (LTTW)WiresMartensitic Transformation ApproachMartensitic Transformation — Volume Changeand Residual StressThe principal decomposition productsof austenite during cooling are precipitated phases that include carbides and nitrides, or the polymorphic phases of al-WELDING JOURNAL 279 -s
Payares-Asprino Supp Nov 08 corr:Layout 110/8/084:46 PMPage 280WELDING RESEARCHBElongationStress σ (MPa)AFig. 1 — Comparison between the designed low-TMs wire (10% Cr-10% Ni) and conventional steel wire. A — Transformation of weld metal during unconstrained cooling; B — development of stress during constrained cooling (Ref. 8).loyed iron, which includes the lowtemperature ferrite (α) and the diffusionless transformation products, BCT αmartensite, and HCP ε-martensite (Ref.12). The BCT martensite phase can bethought of as a variant of the thermodynamically favored α-ferrite, which wouldhave formed from the austenite uponcooling were it not for the severe limitation of the diffusional processes due to fastcooling (Ref. 13). In the absence of ferriteformation by nucleation and growth,austenite undergoes the much more dynamic martensite transformation, involving short-range atomic rearrangementsover broad interfaces at high velocity. Thediffusionless shear-type martensitic transformation requires considerably greaterdriving force than the diffusion-controlledgrowth of ferrite in austenite, due to mechanical shearing of the austenite lattice.Consequently, martensitic transformationusually requires a considerably greater undercooling below the equilibrium temper-Table 1 — Compositions of the Welds (in wt-%) Produced Using the CSM ExperimentalMetal-Cored 0.271.7013.20.03 0.01 0.01 0.01 0.01 0.01Mo0.040.040.040.35 0.01Martensitic Transformation TemperatureSiSTi0.200.190.160.220.41 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01Table 2 — Transformation Temperatures Measured Using DilatometryWireAc1Ac3Ar3 (TMs, TFs)(a)Ar1 40.0810.4160.570–0.570(a) TMs and TMf are for A1, A6, B5, and C5.TFs and TFf are for ER70S-3(b) The final strain (εf) is defined as the amount of expansion from the TMs of TFs temperature to the roomtemperature. Positive means an expansive strain exists at room temperature, negative means a contractive strainexists.280 -s NOVEMBER 2008, VOL. 87ature, T0, at which the parent and transformed phases are in thermodynamicequilibrium. The relative stability of theaustenite phase is very important in orderto induce martensitic transformation at adesired temperature or stress level. Thestrain due to phase transformation canalter the state of residual stress or strain.It is well known that the martensitic transformation of the carburized surface of asteel component puts the surface undercompression as a result of the expansion atthe surface due to the formation of thelower-density martensite from austenite.Martensite transformation begins atmartensite start temperature, TMs, (Fig.1), which can vary over a wide temperature range, from as high as 500 C to wellbelow room temperature, depending onthe concentration of γ-stabilizing alloyingelements in the steel. Once TMs is reached,martensite begins to form with furthertransformation taking place during cooling until reaching the martensite finishtemperature, TMf — Fig. 1. At this temperature, all the austenite should havetransformed to martensite, but frequentlyin practice, a small portion of the austenite remains untransformed even at lowtemperatures. Large volume fractions ofaustenite can be retained in some highlyalloyed steels, where the TMf temperatureis well below room temperature.To achieve martensitic transformation,it is usually necessary for the steel to becooled rapidly, fast enough to suppress thehigher temperature, diffusion-controlled
Payares-Asprino Supp Nov 08 corr:Layout 110/8/084:48 PMPage 281WELDING RESEARCHFig. 2 — Compositions of the welds plotted on the Schaeffler diagram.Aferrite and pearlite reactions, as well asother intermediate reactions such as theformation of bainite.According to transformation characteristics, alloys can be divided into twoclasses with respect to martensite formation: 1) those that affect the equilibriumaustenite decomposition temperature(T0), and 2) those that affect the necessaryundercooling below T0, i.e., ΔTm T0 –TMs (Ref. 14). T0 is influenced by thechemical composition of the alloy, degreeof order, hydrostatic stress (Ref. 15) (classical thermodynamic factors), and ΔTm,which is influenced by the difficulty ofmartensite nucleation and growth withinthe austenite matrix (kinetic, activationfactor). The factors that affect the ΔTm include external shear stresses and any otherproducts that may affect the resistance ofthe base austenite lattice to mechanicalshear during martensite transformation,e.g., hardening mechanism, point defects,dislocations, and precipitates.Martensite Formation — Influence of AlloyingElementsThe effect of alloying elements on theTMs temperature has been studied byFig. 3 — Cylindrical specimens according to theGleeble standard specimen were extracted fromthe welds. (All dimensions are given in mm.)BCDEFig. 4 — Dilatometric curves of the welds made using the experimental wires and commercial wire.many researchers. Izumiyama et al. (Ref.16) showed the effects of individual alloying of 13 elements. Their results showedthat Al, Ti, V, Nb, and Co effectively increased the TMs, whereas Si, Cu, Cr, Ni,Mn, and C decreased the TMs temperature. However, Liu (Ref. 17) reported different effects for some of the elements. Heshowed that all alloying elements mentioned earlier (Mn, V, Cr, Cu, W, Si), except Al and Co, decreased the TMs temperature. The different observations arenot unexpected since different processingconditions (e.g., austenitizing conditionsand cooling rates), austenite grain size,and impurity content will significantly af-WELDING JOURNAL 281 -s
Payares-Asprino Supp Nov 08 corr:Layout 110/8/084:49 PMPage 282WELDING RESEARCHBAFig. 5 — Microstructures of A1 in single-pass weld bead and Gleeble specimen. A — Single-pass weldbead; B — Gleeble specimen.BAfect the martensite transformation behavior. All these metallurgical factors need tobe carefully considered in order to manage the martensite transformation behavior of an alloy.This paper describes the developmentand characterization of several LTTWconsumables that contained lower combined alloy contents (than the 10% Cr and10% Ni developed by Ohta et al. (Ref. 8))for the management of weld residualstresses and improvement of weld joint fatigue properties. Metal cored electrodeswere manufactured and welds prepared.The welds were analyzed for chemicalcomposition and specimens were extracted for dilatometric analysis for TMsdetermination. The welds were also crosssectioned for metallography and hardnesstesting.Experimental ProceduresChemical Composition of the WeldFig. 6 — Microstructures of A6 in single-pass weld bead and Gleeble specimen. A — Single-pass weldbead; B — Gleeble specimen.ABFig. 7 — Microstructures of B5 in single-pass weld bead and Gleeble specimen. A — Single-pass weldbead; B — Gleeble specimen.The CSM-designed filler metal produced ferrite-martensite and ferriteaustenite microstructure. The composition of the welds produced using the fourmetal-cored welding wires are shown inTable 1.The chromium and the nickel equivalents of each of the welds were calculatedand plotted on the Schaeffler diagram asshown in Fig. 2. A1, A6, and B5 are expected to result in a ferrite–martensite microstructure while C5 is expected to bemartensitic with some retained austenite.For comparison, a commercial solidwire, AWS ER70S-3, was also included inthe research. Two models were used to calculate the martensite start temperaturesof these alloys, the Self-Olson Equation(Refs. 18, 19) and the methodology proposed by Ghosh and Olson (Refs. 20–22).Scanning electron microscopy (SEM) andenergy-dispersive spectroscopy (EDS)were used to examine the microstructureas well as alloying element segregation inthe welds.Dilatometric MeasurementsTable 3 — Experimentally Measured and Calculated Martensite Start Temperatures of WeldMetalWiresExperimentalTMs(C )A1A6B5C5ER70S-3Self-Olson EquationTMs(C )Ghosh and OlsonTMs(C a)(a) Ferrite transformation start temperature (TFs).282 -s NOVEMBER 2008, VOL. 87Dilatometric specimens were extractedfrom single-pass welds deposited on anA36 grade structural steel using the fourexperimental consumables and the commercial ER70S-3 wire. The dilatometricmeasurements were made on a Gleeblethermomechanical simulator (Fig. 3)using 6-mm-diameter and 80-mm-lengthsamples. The small cylinders were heatedat the rate of 10 C/s to 1050 C and held atthat temperature for 3 min, followed byquenching in a helium jet at the coolingrate of 100 C/s.
Payares-Asprino Supp Nov 08 corr:Layout 110/8/084:50 PMPage 283WELDING RESEARCHMicrostructural DevelopmentThe specimens were prepared to a mirror-finish using standard metallographictechniques and etched with the KallingNo. 2 reagent (1.5 g CuCl2 33 mL HCl 33 mL ethanol 33 mL H2O) according to ASTM E407 and E340 testing techniques. Photomicrographs were takenusing a LECO Olympus PMG-3 field microscope, coupled to a PaxCAM camera.Area fractions of martensite and ferritewere measured using the point countingtechnique.ABFig. 8 — Microstructures of C5 in single-pass weld bead and Gleeble specimen. A — Single-pass weldbead; B — Gleeble specimen.Microhardness Distribution in the WeldmentsMicrohardness testing was conductedon transverse cross sections taken fromthe welds. Measurements were made on aVickers microhardness scale with a load of300 g. The objective of this study was toevaluate the hardness of single- and multipass welds; both as-solidified and reheated zones were measured. In the caseof multipass welds, measurements weremade in the “No-Reheat” (as-solidified)zone, “Once-Reheated” (reheated weldmetal) zone, and “Twice-Reheated”(overlapping reheated) zones.Samples for microhardness testingwere first etched to identify the weld fusion zone. Hardness measurements weremade across the weld interface at two different distances from the surface of thelast bead of the weldments. Indentationswere made at increments of 0.3175 mm inside the weld metal. However, the increments between successive measurementswere reduced to 0.1588 mm when approaching the reheated zones in order todetect changes in hardness in these areas.Results and AnalysisResults from the dilatometric analysisand microstructural analysis are presented and discussed in the following.Dilatometric AnalysisThe dilatometric curves and themartensite start temperature data are presented. In addition, the experimentally determined values are compared with thepredicted values using the Self-Olsonequations and the Ghosh-Olson methodology.Dilatrometric CurvesFigure 4 shows the dilatometric curvesfor the four designed and conventionalwires obtained using the Gleeble thermomechanical simulator. Martensitic transformation occurred in all four designedwires. The martensite start temperatureBAFig. 9 — Microstructures of ER70S-3. A — Single-pass weld bead; B — Gleeble specimen.Table 4 — Microstructure (in vol-%) of Single-Pass Welds and Gleeble SpecimensA1 WireA6 WireB5 WireC5 WireVolume 9.530.016.624.9Table 5 — Average Hardness Readings of Single-Pass Welds and the Gleeble SpecimensHardnessA1 WireA6 WireB5 WireC5 300386352337337375375400393and the relative strain (compressive ortensile) are recorded on each of the figures. Data from the C5 sample was selected for further description and interpretationofthedilatometricmeasurements. At the beginning of thetest, the percent strain was zero. With increasing temperature, the sample beganto expand as evidenced by the positivestrain. Ferrite-to-austenite transformation (α γ) began at approximately 635 Cand ended at around 720 C, representinga deviation from the equilibrium Ac1 andAc3 temperatures. The negative slopeduring α γ transformation indicatescontraction because of the denser atomicpacking factor of austenite. After theholding temperature of 1050 C, the sample was allowed to cool down at the rate of100 C/s and contraction was observed. At270 C, the slope began to change again indicating that martensite began to form.Up to this point, cooling has amounted toa contraction of about 1.8% (from 1.2 toWELDING JOURNAL 283 -s
Payares-Asprino Supp Nov 08 corr:Layout 110/8/084:51 PMPage 284WELDING RESEARCHFig. 10 — Macrophotograph of weld B5 showingthe three layers and six weld beads.AFig. 11 — Macrophotograph of the multipassweld B5 with lines designating the different weldzones: as-solidified weld metal, once-reheatedzone, and twice-reheated zone.BFig. 12 — Microstructures of as-solidified and once-reheated areas. A — As-solidified fifth pass; B —once-reheated fourth pass.ABFig. 13 — Microstructure of twice-reheated areas in weld A1. A — Twice reheated with the fifth and sixthpasses; B — twice reheated with the fourth and sixth passes.284 -s NOVEMBER 2008, VOL. 87–0.6%). Assuming that the welded structure was entirely rigid, the contraction ofaustenite would have resulted in tensileresidual stresses. However, with the formation of martensite and its more openbody-centered tetragonal (BCT) crystalstructure, the contraction reversed to expansion. Finally, at room temperature, thestrainreachedaroundzero.The martensite transformation starttemperature (TMs) is identified as the temperature at which the slope changed frompositive to negative during cooling. Similarly, the martensite transformation finishtemperature (TMf) can be identified as thetemperature when the negative slopeturned to positive. Even though it is notnecessary for all the austenite to decompose into martensite at Mf to control residual stress, the magnitude of the expansionis important since it is responsible for offsetting the residual tensile stress state thatoriginated from thermal contraction.Sample C5 exhibited the lowest TMstemperature and the largest amount of expansion. The expansive strain also remained at room temperature. These results indicate that compressive residualstresses can be induced in the vicinity ofthe weld metal in a welded joint producedusing the C5 welding wire.Martensite transformation initiated insample A1 at 390 C and the amount of expansion was around 0.5% (from –0.3% to0.2%). Nevertheless, the martensite
WELDING RESEARCH WELDING JOURNAL 279-s ABSTRACT. Martensite start and finish temperatures are very important in struc-tural steel welding because they control the residual stresses in a weld. Tensile residual stresses amplify the effect of ap-plied tensile stress. On the other hand, compressive residual stresses are alge-
6.3 Mechanised/automatic welding 114 6.4 TIG spot and plug welding 115 7 MIG welding 116 7.1 Introduction 116 7.2 Process principles 116 7.3 Welding consumables 130 7.4 Welding procedures and techniques 135 7.5 Mechanised and robotic welding 141 7.6 Mechanised electro-gas welding 143 7.7 MIG spot welding 144 8 Other welding processes 147 8.1 .
Martensitic transformation via quenching Martensite: crystal structure and properties The isothermal decomposition of austenite. The TTT (time-temperature transformation) diagram. Formation pearlite and bainite. Decomposition of austenite on continuous cooling (CCT diagram). Formation of martensite and the martensite lines.
seam butt welding resistance butt welding flash butt welding resistance butt welding shielded unshielded other process plasma laser resistance butt welding inert gas welding submerged arc welding atomic hydrogen shielded metal arc welding (coated electrode) esepl w w w . e u r e k a e l e c t r o d e s .
the welding processes most often used in today's industry including plasma arc cutting, oxyfuel gas cutting and welding, Gas Metal Arc Welding (GMAW), Flux-Cored Arc Welding (FCAW), Shielded Metal Arc Welding (SMAW), and Gas Tungsten Arc Welding (GTAW). Flat welding positions and basic joints will be practiced. Pipe and tube welding
3. Classification of Underwater Welding Underwater welding may be divided into two main types: a) Wet welding b) Dry welding Fig. 3.1 Classification of underwater welding 3.1 Wet welding 3.1.1. Wet welding with coated electrode Wet welding is performed at ambient pressure with the welder-diver in the water and no physical barrier
10.6 Braze-welding 460 Exercises 466 11 Joining processes (welding) 467 11.1 Fusion welding 468 11.2 Oxy-acetylene welding 468 11.3 Manual metal-arc welding 490 11.4 Workshop testing of welds 504 11.5 Miscellaneous fusion welding processes 506 11.6 Workholding devices for fusion welding 509 11.7 Resistance welding 515
Welding residual stresses and stress relieve Formation of the weld metal Solidification of the weld metal Phase transformations during welding Basics of welding methods used in welding of zirconium alloys TIC Laser Welding Electron Beam Welding Upset shape welding Resistance Appendix C -R
The API Aboveground Storage Tank Inspector Certification Examination is designed to identify individuals who have satisfied the minimum qualifications specified in API Standard 653, Tank Inspection, Repair, Alteration, and Reconstruction. Questions may be taken from anywhere within each document in this Body of Knowledge (BOK), unless specifically excluded herein. In the event that specific .
