Kinetics of the peritectic phase transformation: In-situ measurements and phase field modeling

An experimental study has been conducted into the role of cooling rate on the kinetics of the peritectic phase transformation in a Fe-C alloy. The interfacial growth velocities of the peritectic phase transformation were measured in situ for cooling rates of 100, 50, and 10 K/min. In-situ observations were obtained using high-temperature laser scanning confocal microscopy (HTLSCM) in a concentric solidification configuration. The experimentally measured interface velocities of the liquid/austenite (L/γ) and austenite/delta-ferrite (γ/δ) interphase boundaries were observed to increase with higher cooling rates. A unique finding of this study was that as the cooling rate increased, there was a transition point where the L/γ interface propagated at a higher velocity than the γ/δ interface, contrary to the findings of previous researchers. Phase field modeling was conducted using a commercial multicomponent, multiphase package. Good correlation was obtained between model predictions and experimental observations in absolute values of interface velocities and the effect of cooling rate. Analysis of the simulated microsegregation in front of the L/γ and γ/δ interfaces as a function of cooling rate revealed the importance of solute pileup. This microsegregation plays a pivotal role in the propagation of interfaces; thus, earlier modeling work in which complete diffusion in the liquid phase was assumed cannot fully describe the rate of propagation of the L/γ and δ/γ interfaces during the course of the peritectic transformation.     

Kinetics of the peritectic phase transformation: In-situ measurements and phase field modeling

An experimental study has been conducted into the role of cooling rate on the kinetics of the peritectic phase transformation in a Fe−C alloy. The interfacial growth velocities of the peritectic phase transformation were measured in situ for cooling rates of 100, 50, and 10 K/min. In-situ observations were obtained using high-temperature laser scanning confocal microscopy (HTLSCM) in a concentric solidification configuration. The experimentally measured interface velocities of the liquid/austenite (L/γ) and austenite/delta-ferrite (γ/δ) interphase boundaries were observed to increase with higher cooling rates. A unique finding of this study was that as the cooling rate increased there was a transition point where the L/γ interface propagated at a higher velocity than the γ/δ interface, contrary to the findings of previous researchers. Phase field modeling was conducted using a commercial multicomponent, multiphase package. Good correlation was obtained between model predictions and experimental observations in absolute values of interface velocities and the effect of cooling rate. Analysis of the simulated microsegregation in front of the L/γ and γ/δ interfaces as a function of cooling rate revealed the importance of solute pileup. This microsegregation plays a pivotal role in the propagation of interfaces; thus, earlier modeling work in which complete diffusion in the liquid phase was assumed cannot fully describe the rate of propagation of the L/γ and δ/γ interfaces during the course of the peritectic transformation.     

Kinetics of peritectic reaction and transformation in Fe-C alloys

In situ dynamic observation of the progress of a peritectic reaction and transformation of Fe-(0.14 pct C)- and Fe-(0.42 pct C)-peritectic Fe-C alloys has been successfully made with a combination of a confocal scanning laser microscope and an infrared image furnace. The peritectic reaction is characterized by the formation of the γ-austenite phase at the junction of the liquid and the grain boundary of δ-ferrite crystals and subsequent propagation of the three-phase point, liquid/γ/δ, along the liquid/δ boundary, whereas the peritectic transformation occurs by the thickening of the intervening γ toward both the liquid side and the δ side. The rates of the peritectic reaction for the two peritectic alloys are found to be much faster than the rate that would be controlled by carbon diffusion, suggesting that either massive transformation to γ or solidification as γ controls the rate. This is also the case for the Fe-0.14%C transformation in the hypoperitectic alloy. However, the rate of the peritectic transformation in the Fe-0.42%C alloy is determined by carbon diffusion. This article is based on a presentation made in the “Geoffrey Belton Memorial Symposium,” held in January 2000, in Sydney, Australia, under the joint sponsorship of ISS and TMS.     

Microscopic modeling of fundamental phase transformations in continuous castings of steel

A model has been developed to describe the microscopic behavior of phase transformation of carbon steels in the range of cooling rate occurring in continuous casting. In the liquid-to solidphase transformation, this model simulates the phenomena of dendrite nucleation and growth during solidification. Both δ- and γ-dendrites are involved. The nucleation and growth model has been established on the basis of published experimental data and previous work. Also, a model of the peritectic transformation of carbon steels has been included. In the solid-to solidphase transformation, the model considers the δ→ γ, γ→ α, and γ→ α + Fe₃C phase transformations. The δ→ γ and γ→ α phase transformations have been modeled by using the Johnson-Mehl equation, also known as the Avrami equation. For the pearlite transformation, a nucleation law, as well as the growth kinetics, has been established. Good agreement has been found between the prediction of the model and the experimental data.     

Spatially resolved X-ray diffraction mapping of phase transformations in the heat-affected zone of carbon-manganese steel arc welds

Phase transformations that occur in the heat-affected zone (HAZ) of gas tungsten arc welds in AISI 1005 carbon-manganese steel were investigated using spatially resolved X-ray diffraction (SRXRD) at the Stanford Synchrotron Radiation Laboratory. In situ SRXRD experiments were performed to probe the phases present in the HAZ during welding of cylindrical steel bars. These real-time observations of the phases present in the HAZ were used to construct a phase transformation map that identifies five principal phase regions between the liquid weld pool and the unaffected base metal: (1) α-ferrite that is undergoing annealing, recrystallization, and/or grain growth at subcritical temperatures, (2) partially transformed α-ferrite co-existing with γ-austenite at intercritical temperatures, (3) single-phase γ-austenite at austenitizing temperatures, (4) δ-ferrite at temperatures near the liquidus temperature, and (5) back transformed α-ferrite co-existing with residual austenite at subcritical temperatures behind the weld. The SRXRD experimental results were combined with a heat flow model of the weld to investigate transformation kinetics under both positive and negative temperature gradients in the HAZ. Results show that the transformation from ferrite to austenite on heating requires 3 seconds and 158°C of superheat to attain completion under a heating rate of 102°C/s. The reverse transformation from austenite to ferrite on cooling was shown to require 3.3 seconds at a cooling rate of 45 °C/s to transform the majority of the austenite back to ferrite; however, some residual austenite was observed in the microstructure as far as 17 mm behind the weld.     

In-situ phase mapping and direct observations of phase transformations during arc welding of 1045 steel

In-situ spatially resolved X-ray diffraction (SRXRD) experiments were performed during gas tung-sten arc (GTA) welding of AISI 1045 C-Mn steel. Ferrite (α) and austenite (γ) phases were identified and quantified in the weld heat-affected zone (HAZ) from the real time SRXRD data. The results were compiled with weld temperatures calculated using a coupled thermal fluids model to create a phase map of the HAZ. Kinetics of the α → γ transformation during weld heating and the reverse γ → α transformation during weld cooling were extracted from the map. Superheating as high as 250 °C above the A3 temperature was observed for the α → γ phase transformation to reach completion at locations near the fusion zone (FZ) boundary. The SRXRD experiments revealed that the newly created γ phase exists with two distinct lattice parameters, resulting from the inhomogeneous distribution of carbon and manganese in the starting pearlitic/ferritic microstructure. During cooling, the reverse γ → α phase transformation was shown to depend on the HAZ location. In the fine-grained region of the HAZ, the γ → α transformation begins near the A3 temperature and ends near the A1 temperature. In this region, where the cooling rates are below 40 °C/s, the transformation occurs by nucleation and growth of pearlite. In the coarse-grained region of the HAZ, the γ → α transformation requires 200 °C of undercooling for completion. This high degree of undercooling is caused by the large grains coupled with cooling rates in excess of 50 °C/s that result in a bainitic transformation mechanism.     

Analysis of solute distribution in dendrites of carbon steel with δ/γ transformation during solidification

Solute distribution in dendrites during solidification of carbon steel was analyzed by unidirectional solidification experiments and mathematical analysis. The characteristic of the mathematical analysis is that diffusion of solutes in solid and redistribution of solutes at solid/liquid andδ/γ interfaces are taken into consideration. Based on the observed and calculated results, it was found that phosphorus was redistributed fromγ-phase toδ-phase, and that manganese was slightly redistributed fromδ-phase toγ-phase. Therefore the concentrated region of phosphorus can be separated from that of manganese duringδ/γ transformation in the case of slow cooling. Moreover, it was concluded that rapid diffusion inδ-phase and the redistribution duringδ/γ transformation played an important role in the variation of the interdendritic concentrations of solutes with lower carbon concentration.     

Numerical modeling of solidification and subsequent transformation of Fe-Cr-Ni alloys

A computational method for the analysis of phase transformation involving solidification was developed with the assumption of thermodynamic equilibria at interfaces. The region of interest was divided into finite segments, and solute diffusion across the segments was computed by the use of the direct finite difference method (FDM). Simultaneously, thermodynamic equilibrium at each interface was updated at every step of the diffusion analysis to determine the location of the interfaces. The temperature decrease and the increment of fraction solid were calculated based on thermal balance, including a heat extraction condition. Solid state transformation from δ to γ phase within each FDM segment was modeled by the use of a Clyne-Kurz (C-K) type analysis with assumptions of complete mixing of solutes in theδ phase and limited back diffusion in theγ phase. The calculation results were compared with welding solidification experiments in the iron-chromium-nickel ternary system. Good agreement was obtained with respect to solute distribution and residual fraction ofδ phase over different compositions and solidification modes of the alloys used.     

Simple model of microsegregation during solidification of steels

A simple analytical model of microsegregation for the solidification of multicomponent steel alloys is presented. This model is based on the Clyne-Kurz model and is extended to take into account the effects of multiple components, a columnar dendrite microstructure, coarsening, and the δ/γ transformation. A new empirical equation to predict secondary dendrite arm spacing as a function of cooling rate and carbon content is presented, based on experimental data measured by several different researchers. The simple microsegregation model is applied to predict phase fractions during solidification, microsegregation of solute elements, and the solidus temperature. The predictions agree well with a range of measured data and the results of a complete finite-difference model. The solidus temperature decreases with either increasing cooling rate or increasing secondary dendrite arm spacing. However, the secondary dendrite arm spacing during solidification decreases with increasing cooling rate. These two opposite effects partly cancel each other, so the solidus temperature does not change much during solidification of a real casting.     

An Investigation of the Massive Transformation from Ferrite to Austenite in Laser-Welded Mo-Bearing Stainless Steels

A series of 31 Mo-bearing stainless steel compositions with Mo contents ranging from 0 to 10 wt pct and exhibiting primary δ-ferrite solidification were analyzed over a range of laser welding conditions to evaluate the effect of composition and cooling rate on the solid-state transformation to γ-austenite. Alloys exhibiting this microstructural development sequence are of particular interest to the welding community because of their reduced susceptibility to solidification cracking and the potential reduction of microsegregation (which can affect corrosion resistance), all while harnessing the high toughness of γ-austenite. Alloys were created using the arc button melting process, and laser welds were prepared on each alloy at constant power and travel speeds ranging from 4.2 to 42 mm/s. The cooling rates of these processes were estimated to range from 10 K (°C)/s for arc buttons to 10⁵ K (°C)/s for the fastest laser welds. No shift in solidification mode from primary δ-ferrite to primary γ-austenite was observed in the range of compositions or welding conditions studied. Metastable microstructural features were observed in many laser weld fusion zones, as well as a massive transformation from δ-ferrite to γ-austenite. Evidence of epitaxial massive growth without nucleation was also found when intercellular γ-austenite was already present from a solidification reaction. The resulting single-phase γ-austenite in both cases exhibited a homogenous distribution of Mo, Cr, Ni, and Fe at nominal levels.     

Fracture and the formation of sigma phase,M23C6, and austenite from delta-ferrite in an AlSl 304L stainless steel

The decomposition of delta-ferrite and its effects on tensile properties and fracture of a hot-rolled AISI 304L stainless steel plate were studied. Magnetic response measurements of annealed specimens showed that the transformation rate of delta-ferrite was highest at 720 °C. Transformation behavior was characterized by light microscopy, transmission electron microscopy, scanning electron microscopy, and energy-dispersive spectroscopy on thin foils. The initial transformation of delta-ferrite (δ) to austenite (γ) and a chromium-rich carbide (M₂₃C₆) occurred by a lamellar eutectoid reaction, δ⇄M₂₃C₆ +γ. The extent of the reaction was limited by the low carbon content of the 304L plate, and the numerous, fine M₂₃C₆ particles of the eutectoid structure provide microvoid nucleation sites in tensile specimens annealed at 720 °C for short times. Sigma phase(σ) formed as a result of a second eutectoid reaction,δσ +γ. Brittle fracture associated with the plate-shaped sigma phase of the second eutectoid structure resulted in a significant decrease in reduction of area (RA) in the transverse tensile specimens. The RA for longitudinal specimens was not affected by the formation of sigma phase. Tensile strengths were little affected by delta-ferrite decomposition products in either longitudinal or transverse orientations. Y. Shen, formerly with the Department of Metallurgical and Materials Engineering, Colorado School of Mines, is deceased.     

Influence of solid/liquid interfaces on the microstructure and stress-rupture life of the single-crystal nickel-base superalloy NASAIR 100

The [001] oriented single crystals of nickel-base superalloy NASAIR 100 with the planar, cellular, coarse-dendritic, and fine-dendritic solid/liquid (S/L) interfaces were prepared, respectively, and their microstructure and stress-rupture behavior at 1050 °C were investigated in both as-cast and solution heat-treated conditions. It was found that in as-cast single crystals of NASAIR 100, microsegregation and γ/γ′ eutectic produced in the solidification process increased, γ′ size decreased, and γ′ shape tended progressively to be cuboidal, with the successive transition of the S/L interface from planar to cellular, then to coarse-dendritic, and finally to fine-dendritic morphology. Furthermore, the solution temperature required to dissolve all as-cast γ′ and most of the γ/γ′ eutectic increased with the aforementioned successive transition of S/L interfaces. The reprecipitated γ′, after solution heat treatment (SHT), was usually fine and cuboidal. However, some W-rich phase was present in the heat-treated dendritic single crystals. Both the planar and the cellular single crystals of NASAIR 100 exhibited no superiority in stress-rupture life, irrespective of the heat-treatment conditions. Instead, the single crystals with dendritic morphology possessed excellent stress-rupture lives, after heat treatment at 1320 °C for 4 hours, followed by air cooling (AC). Perfect γ′ rafts with high-average aspect ratios formed during the stress-rupture tests of dendritic single crystals; in contrast, irregularly coarsening structures appeared in both the planar and cellular single crystals. The microstructure and solution behavior were illustrated in detail. Furthermore, the microstructural factors to affect the high-temperature stress-rupture life of the single crystals of NASAIR 100 were also analyzed.     

Analysis of the <Emphasis Type="Italic">γ</Emphasis> → <Emphasis Type="Italic">α</Emphasis> transformation in a C-Mn steel by phase-field modeling

This article deals with the austenite (γ) decomposition to ferrite (α) during cooling of a 0.10 wt pct C-0.49 wt pct Mn steel. A phase-field model is used to simulate this transformation. The model provides qualitative information on the microstructure that develops on cooling and quantitative data on both the ferrite fraction formed and the carbon concentration profile in the remaining austenite. The initial austenitic microstructure and the ferrite nucleation data, derived by metallographic examination and dilatometry, are set as input data of the model. The interface mobility is used as a fitting parameter to optimize the agreement between the simulated and experimental ferrite-fraction curve derived by dilatometry. A good agreement between the simulated α-γ microstructure and the actual α-pearlite microstructure observed after cooling is obtained. The derived carbon distribution in austenite during transformation provides comprehension of the nature of the transformation with respect to the interface-controlled or diffusion-controlled mode. It is found that, at the initial stage, the transformation is predominantly interface-controlled, but, gradually, a shift toward diffusion control takes place to a degree that depends on cooling rate.     

Nucleation and phase selection in undercooled Fe-Cr-Ni melts: Part II. Containerless solidification experiments

The solidification behavior of undercooled Fe-Cr-Ni melts of different compositions is investigated with respect to the competitive formation of δ-bcc (ferrite) and γ-fcc phase (austenite). Containerless solidification experiments, electromagnetic levitation melting and drop tube experiments of atomized particles, show that δ (bcc) solidification is preferred in the highly undercooled melt even at compositions where δ is metastable. Time-resolved detection of the recalescence events during crystallization at different undercooling levels enable the determination of a critical undercooling for the transition to metastable bcc phase solidifcation in equilibrium fcc-type alloys. Measurements of the growth velocities of stable and metastable phases, as functions of melt undercooling prior to solidification, reveal that phase selection is controlled by nucleation. Phase selection diagrams for solidification processes as functions of alloy composition and melt undercooling are derived from two types of experiments: X-ray phase analysis of quenched samples and in situ observations of the recalescence events of undercooled melts. The experimental results fit well with the theoretical predictions of the metastable phase diagram and the improved nucleation theory presented in an earlier article. In particular, the tendency of metastable δ phase formation in a wide composition range is confirmed.     

A new criterion for internal crack formation in continuously cast steels

To estimate the cracking condition in continuously cast steels, a new model for critical fracture stress given from the measured critical strain has been proposed, which can take into account the brittle temperature range and strain rate. The effects of brittle temperature range and strain rate on critical strain for internal crack formation were analyzed. When the brittle temperature range and strain rate were increased, the possibility of internal crack formation increased due to the decreasing critical strain. To describe the thermomechanical property model of the mushy zone between zero strength temperature (ZST) and zero ductility temperature (ZDT), the yield criterion for porous metals, which can take into account δ/γ transformation, was used. Using the fitting equation for the measured critical strain and the microsegregation analysis, the thermomechanical behavior of the mushy zone could be successfully described by the proposed model, which incorporates the effects of microsegregation of solute elements and δ/γ transformation on hot tear during solidification at the given range of steel compositions and strain rates. A cracking criterion based on the difference of deformation energy in the brittle temperature range is proposed to explain the cracking phenomenon of whole carbon range.     

A Comparative Study of the γ/γ′ Eutectic Evolution During the Solidification of Ni-Base Superalloys

The evolution of γ/γ′ eutectic during the solidification of Ni-base superalloys CMSX-10 and CMSX-4 was investigated over a wide range of cooling rates. The microsegregation behavior during solidification was also quantitatively examined to clarify the influence of elemental segregation on the evolution of γ/γ′ eutectic. In the cooling rate ranges investigated (0.9 to 138.4 K/min (0.9 to 138.4 °C/min)), the γ/γ′ eutectic fraction in CMSX-10 was found to be more than 2 times higher than that in CMSX-4 at a given cooling rate. However, the dependence of the γ/γ′ eutectic fraction on the cooling rate in both alloys showed a similar tendency; i.e., the γ/γ′ eutectic fraction increased with increasing the cooling rate and then exhibited a maximum plateau at and above the certain critical cooling rate in both alloys. This critical cooling rate was found to be dependent on the alloy composition and was estimated to be about 12 K/min (12 °C/min) and 25 K/min (25 °C/min) for CMSX-10 and CMSX-4, respectively. The calculated solid compositions based on the modified Scheil model revealed that even a small compositional difference of total γ′ forming elements in the initial composition of the alloy can play a significant role in the as-cast eutectic fraction during the solidification of Ni-base superalloys. The evolution of the γ/γ′ eutectic fraction with respect to the cooling rate could be rationalized by taking into account the effects of back-diffusion in solid and dendrite arm coarsening on decreasing the extent of microsegregation.     

Modeling of the peritectic reaction and macro-segregation in casting of low carbon steel

Macro-microscopic models have been developed to describe the macrosegregation behavior associated with the peritectic reaction of low carbon steel. The macrosegregation model has been established on the basis of previously published work and experimental data. A microscopic model of a three-phase reaction L+δ→γ has been modeled by using Fredriksson’s approach. Four horizontal and unidirectional solidified experimental groups simulating continuous casting have been performed with a low carbon steel containing 0.13 wt pct carbon. The extent of macrosegregation of carbon was determined by wet chemical analysis of millings. It is confirmed, by comparing calculated results with experimental results, that this model successfully predicts the occurrence of macrosegregation. The results indicate that a peritectic reaction which is associated with a high cooling rate generates high thermal contraction and a high tensile strain rate at the peritectic temperature. Therefore, the macrosegregation, particularly at the ingot surface, is very sensitive to the cooling rate, where extremely high positive segregation was observed in the case of a high cooling rate. However, in the case of slow cooling rate, negative segregation was noted. The mechanism of macrosegregation with peritectic reaction is discussed in detail.     

Effect of cold rolling on the precipitation behavior of δ phase in INCONEL 718

Systematic research has been undertaken on the effect of cold rolling on the precipitation kinetics of δ phase in INCONEL 718. Above 910 °C, cold rolling promotes the precipitation of δ phase. Below 910 °C, the precipitation of δ phase is still preceded by the γ″ precipitation in cold-rolled INCONEL 718. Cold rolling promotes not only the precipitation of γ″ phase but also the γ″ → δ transformation. The relationship between the weight percentage of δ phase and aging time follows the Avrami equation. Below 910 °C, as cold rolling reduction and temperature increase, the time exponent (n) decreases, whereas the rate of δ precipitation increases. The apparent activation energy of δ precipitation varies in the range of 1113 to 577 kJ/mol for 25 to 65% cold-rolled INCONEL 718 and decreases as cold rolling reduction increases. Precipitation-time-temperature (PTT) diagrams have been determined for the four cold-rolled INCONEL 718. The noses of the PTT curves are located at about 910 °C. These curves are shifted significantly to longer times as cold rolling reductions decrease.