A model for the silicon-manganese deoxidation of steel weld metals

From an analytical and theoretical study of flat and out-of-position gas metal arc (GMA) C-Mn steel welds containing varying additions of silicon and manganese, we conclude that the buoyancy effect (flotation obeying Stokes’ law) does not play a significant role in the separation of oxide inclusions during weld metal deoxidation. Consequently, the separation rate of the particles is controlled solely by the fluid flow pattern in the weld pool. A proposed two-step model for the weld metal deoxidation reactions suggests that inclusions formed in the hot, turbulent-flow region of the weld pool are rapidly brought to the upper surface behind the arc because of the high-velocity flow fields set up within the liquid metal. In contrast, those formed in the cooler, less-turbulent flow regions of the weld pool are to a large extent trapped in the weld metal as finely dispersed particles as a result of inadequate melt stirring. The boundary between “hot” and “cold” parts for possible inclusion removal is not well defined, but depends on the applied welding parameters, flux, and shielding gas composition. As a result of the intricate mechanism of inclusion separation, the final weld metal oxygen content depends on complex interactions among the following three main factors: (1) the operational conditions applied, (2) the total amount of silicon and manganese present, and (3) the resulting manganeseto-silicon ratio. The combined effect of the latter two contributions has been included in a new deoxidation parameter, ([pct Si][pct Mn])^−0.25. The small, negative exponent in the deoxidation parameter indicates that control of the weld metal oxygen concentrations through additions of silicon and manganese is limited and that choice of operational conditions in many instances is the primary factor in determining the final degree of deoxidation to be achieved.     

Thermochemistry of binary liquid gold alloys: The systems Au-Ni,Au-Co,Au-Fe,and Au-Mn

In this paper we report enthalpy of mixing data for the liquid alloys of gold with manganese, iron, cobalt, and nickel obtained by a Calvet-type calorimeter at 1378 K. The enthalpies of mixing are compared with Gibbs energies calculated from earlier emf and vapor pressure studies to yield information on the excess entropies of mixing. The limiting enthalpies of solution of the liquid transition metals in liquid gold are compared with values predicted from the semi-empirical model of Miedemaet al. and with earlier data for the same transition metals in liquid copper. The calculated values of the excess entropy of solution in liquid gold are compared with the corresponding values in liquid copper near 1400 K. For Ni, Co, and Fe as solutes we observepositive shifts of 5 to 9 J K⁻¹ mol⁻¹ which are attributed to vibrational entropy terms. For Mn there is a strongnegative shift of about 35 J K⁻¹ mol⁻¹. This shift probably is due to “complex” or “associate” formation between gold and manganese atoms.     

Sulfidation of binary iron alloys containing cerium or lanthanum

Additions of elements of the rare-earth group to steel, mostly in the form of “mischmetal”, are generally made to improve on the deoxidation and desulfurization. Spring¹ was the first to carry out experiments with the addition of mischmetal to certain types of steel as far back as 1918. A comprehensive review of the effects of lanthanum and cerium additions to steel was given by Richerd.²     

Electron Microscopy Studies of Charge Density Wave (CDW) Transitions in Ti58.7Ni37.5Al3.8 Alloy

Electron diffraction and microscopy and electrical resistancevs temperature measurements of a Ti_58.7 Ni_37.5Al_3.8 alloy between room temperature and liquid nitrogen temperature have been carried out. On cooling, the increase in electrical resistance and the appearance of 1/3(110)- and 1/3(111)-type superlattice reflections are interpreted to be due to the formation of three-dimensional CDW’s. The transformation behavior, structural and microstructural changes of the present alloy are similar to those of the “premartensitic” behavior of a Ti₅₀Ni₄₇Fe₃ alloy which undergoes “normal-to-incommensurate” and “incommensurate-to-commensurate” CDW transitions, as reported earlier. The substitution of Al for Ni in TiNi and the nonstoichiometry of the present alloy apparently cause the incommensurate phase to exist over a larger temperature range. Formerly with the University of Illinois     

Deoxidation of Liquid Copper with Reducing O<Subscript>2</Subscript>/CH<Subscript>4</Subscript> Flames

The rate of deoxidation of molten copper during top blowing with various reducing gases has been investigated using thermogravimetry. It was observed that the rate of deoxidation increases with an increasing flow rate of H₂ or CO and that H₂ is a more effective reducing reagent than CO. The rate of deoxidation using methane was measured for O₂/CH₄ ratios from 1.5 to 2.0. As expected, the deoxidation rate decreased with an increasing O₂/CH₄ feed ratio because the flame became less reducing. For all tests, initially there is a linear decrease in mass as oxygen is removed. However, for some experiments, after some time, a sudden acceleration in the rate of mass loss occurs. Using video and X-ray imaging, it was found that this pattern corresponded to gas evolution from within the molten copper. This finding can be explained by the sudden water vapor evolution because the hydrogen dissolved in the copper reacts with the remaining oxygen, and “boiling” takes place, leading to an enhanced stirring of the copper.     

Morphology of y’ and y” precipitates and thermal stability of inconel 718 type alloys

The precipitation of the γ^’ (Ll₂) and γ^" (DO₂₂) phases has been studied in four alloys Fe-Ni-Cr-Ti-Al-Nb containing a higher Ti + Al/Nb ratio than that of the INCONEL 718 alloy. For these alloys, the precipitation microstructure varies rapidly with aging temperature and composition. Bct γ^"particles have always been found to precipitate on γ^’ phase. Moreover, by aging three alloys above a critical temperature, a “compact ntorphology” has been observed: cube-shaped γ^’ particles coated on their six faces with a shell of γ^" precipitate. This microstructure has proved to be very stable on prolonged aging. A thermal stability better than that encountered in nominal INCONEL 718 alloy can thus be achieved. The influence of composition and aging temperature on the conditions that bring about this “compact morphology” has been investigated. A minimal Ti + Al/Nb ratio between 0.9 and 1 has been determined, allowing the “compact morphology” to be obtained. This paper is based upon a thesis submitted by R. COZAR in partial fulfillment of the requirements of the degree of Doctor of Philosophy at the University of Nancy.     

Optimization in castings—An overview of relevant computational technologies and future challenges

The manufacture of defect-free components at low cost and high productivity is as important to the casting industry today as it was 30 years ago. In the past, experience was gained either by using a “trial and error” method or by undertaking expensive experiments. Many “dos” and “don’ts” have evolved in the casting process over a period of time. However, the important ones that come to mind are so fundamental that they challenge the “academic mind” to think all over again. The rules proposed by Professor John Campbell^[1] are classic examples. The message is simple: mathematical complexity in computer models needs to go hand in hand with the rules derived from “first principles.” In the field of optimization, a variety of methods have been proposed over a period of years. At the start of optimization study, the foundryman’s first choice is to use simple but well-established methods such as the use of orthogonal arrays for optimal design of process conditions or the famous “inscribed” or Heuvers’ circle method^[2] for optimal feeding design. The computer simulation software has been based on a variety of computational methods ranging from geometric reasoning techniques (the famous Chvorinov rule and its variants)^{[11,13,15,29–31]} to solving complex partial differential equations using one of the numerical methods. Optimization methods based on solving partial differential methods was an active area of research in the mid-1990s.^[6–10,17] This article reviews a variety of optimization methods including—probably for the first time—geometric reasoning methods. The contribution from various computational methodologies is highlighted with particular emphasis on characterizing “objective functions” and “constraints.” The article also raises some of the challenging issues that the optimization community is facing today for solving casting problems and reports on our recent work on linking geometric reasoning techniques with the finite element method (FEM) and other data mining tools to achieve computationally efficient optimal design of casting processes. This article is based on a presentation made in the John Campbell Symposium on Shape Casting, held during the TMS Annual Meeting, February 13–17, 2005, in San Francisco, CA.     

Anisotropic plastic flow in ferritic stainless steels and the “roping” phenomenon

The recent papers of Chao and Takechi, Kato, Sunami and Nakayama are discussed relative to the “roping” morphology and the textures commonly observed in ferritic stainless steels prone to “roping”. The anisotropic plastic flow associated with textures having [110] and [112J rolling directions is reviewed and a transverse plastic buckling mechanism is proposed as being consistent with “roping” morphology and texture combinations. It is proposed that longitudinal bands with a strong (001) [110] (or similar) texture surrounded by material with a (111) [11άcr2] (or similar) texture will undergo plastic buckling under transverse compressive stresses that result from the texture mix and elongation in the rolling direction. The mechanism predicts amaximum ratio of sheet thickness to corrugation width of about 0.43 for buckling with “clamped ends” and about 0.86 for buckling with “hinged ends”. Profile measurements of 430 and 434 stainless steel sheets pulled to 15 pct in the rolling direction show ratios generally under 0.4. Furthermore, the mechanism predicts that aminimum plastic elongation of 0.4 pct is required in the rolling direction to initiate buckling or “roping”. Profile measurements are presented showing that the “roping” corrugations do not develop until 1.5 to 2.0 pct plastic elongation in the rolling direction. Formerly with Allegheny Ludlum Industries, Inc., Brackenridge, Pa.     

Kinetics of deoxidation of liquid copper by graphite particles during submerged injection

Liquid copper can be deoxidized by submerged injection of inert gas in the presence of graphite particles. This paper describes the results of experiments in which approximately 20 kg copper melts were deoxidized by injection of N₂, CO₂, and air in the presence of low sulfur graphite particles. The apparent kinetics of the deoxidation process are first order with respect to the concentration of dissolved oxygen, and the concentration of oxygen in the melt could be reduced to less than 10 ppm by weight after less than 30 minutes of injection. The kinetics are consistent with mixed rate control with both mass transfer and chemical reaction rate affecting the rate of deoxidation. In these experiments, the rate of deoxidation under a graphite covering was slower when particles were injected with the gas stream than when gas was injected alone, although this result may have been influenced by the small size of the melt. Y. W. Chang, formerly Graduate Student with the Department of Civil Engineering, Mechanics, and Metallurgy, University of Illinois at Chicago This paper is based on a presentation made in the T.B. King Memorial Symposium on “Physical Chemistry in Metals Processing” presented at the Annual Meeting of The Metallurgical Society, Denver, CO, February, 1987, under the auspices of the Physical Chemistry Committee and the PTD/ISS.     

Stirring and its effects on aluminum deoxidation in the ASEA-SKF furnace: Part II. Mathematical representation of the turbulent flow field and of tracer dispersion

A mathematical formulation is presented describing fluid flow and tracer dispersion in an ASEA-SKF furnace. The statement of the problem required the simultaneous solution of the turbulent Navier-Stokes equations together with a simplified form of Maxwell’s equations. The resultant partial differential equations were solved numerically using a digital computer. Computed results are presented describing the streamline pattern, the velocity field, the spatial distribution of turbulent energy within the system, together with the rate at which a tracer is dispersed. For a 50 ton furnace, with a coil current of 1300 A the computed linear velocities ranged up to 150 cm/s and the mean values of the eddy diffusivity were of the order of 300 to 500 cm²/s. The computed results were found to be in reasonable agreement with previously reported tracer dispersion measurements. However, there is a disparity between the predicted circulation rates and those deduced from the measurements in an earlier paper using a simple one dimensional model for the interpretation of the results.     

200, 020, and 002 X-ray peaks in tempered Fe-18 Ni-C martensites

Separate 200, 020, and 002 X-ray peaks were recorded for 0.0, 0.4, and 0.8 wt pct carbon (18 pct Ni) martensites after tempering between 25 and 500°C. The carbon bearing martensites studied here have been tempered initially enough to eliminate the “high tetragonality” 002 peak usually recorded for as-quenched martensite and the present results apply to tempered martensite only. The peak maximum is taken to determine the lattice parameter and the peak shape is recorded. At all carbon levels and after all tempering treatments, the “crd parameter is larger than or equal to the “a” or “b”. The relative enlargement is very small (0.08 pct) for the lowest carbon level and for any carbon level after severe tempering (500°C for 15 min). For the two higher carbon alloys tempered at temperatures below 400°C (for 15 min) the “c” parameter is significantly larger than the “a” and “b” and for the 0.4 wt pct C alloy the “b” is significantly smaller than the“a” whereas in the 0.8 pct C alloy the “b” is slightly larger than the “a”. Within experimental error the mean volume of the unit cell does not change during the tempering studied here and is nearly unaffected by the initial carbon content. This indicates that little (at most 0.1 wt pct) carbon is dissolved in tempered martensite. In the low carbon alloy the peaks are symmetric and sharpen symmetrically during tempering. In the higher carbon alloys the peaks are nearly symmetric and sharp after severe tempering. After less severe tempering the 002 peak is asymmetrically broadened toward lower9 values (higher lattice parameters) whereas the 200 and 020 peaks are asymmetrically broadened toward higher 0 values corresponding to lower lattice parameters. This collection of results is tentatively interpreted as being due to strains in martensite due to transformation induced substructure and precipitated carbides.     

Model study of mixing and mass transfer rates of slag-metal in top and bottom blown converters

Water model experiments have been conducted to clarify mixing rates of molten steel and mass transfer rates between slag and metal in LD and Q-BOP furnaces using six different circular tuyere arrangements. Splashing and ‘spitting’ were also examined with a view to finding a quiet bath with minimum mixing time and maximum mass transfer rate. Froude’s similarity criterion was fulfilled to determine gas flow rate and bath depth. Complete mixing time of water determined by tracer technique had been 0.9 second to 1.8 seconds for Q-BOP as compared to 6 seconds to 13 seconds for LD. This shows that the stirring intensity in Q-BOP is remarkably larger than that of LD. A simple relationship τ = 5.9(Q/N) ^−0.49 was obtained with gas flow rateQ and number of tuyereN. This indicates that flow rate of gas per tuyere should be intensified to realize better mixing. Mass transfer coefficient K_Ba for bottom blowing was found to be almost double that for top blowing. Of all the tuyere configurations studied for Q-BOP’s, a half circular tuyere arrangement was found to be the best considering all aspects of mixing, mass transfer, and bath agitation.     

Numerical computations of fluid flow and heat transfer in a gas-stirred liquid bath

The flow and temperature fields due to bottom air injection in a cylindrical vessel containing water were numerically analyzed. The Eulerian approach was used for the formulation of both the continuous and the dispersed phases. The computational domain was extended beyond the undisturbed height of the liquid in the bath to accommodate practical gas injection systems. Turbulence in the liquid phase was modeled using a two-equationk- ε model. Interphase friction and heat transfer coefficients were calculated by using correlations available in the literature. The general-purpose computer program PHOENICS was employed to predict the velocity, vol-ume fraction, and the temperature fields of each phase. Turbulent dispersion of the phases was modeled by introducing a “dispersion Prandtl number.” The predicted flow fields were com-pared with experimental measurements available in the literature. The results are of interest in the design and operation of a wide variety of material processing operations.     

The relation of microstructure and fracture properties of electron beam melted,modified SAE 4620 steels

A method is described for the transmission and scanning electron microscope study of the relationship between the microstructure and the fracture properties of two quenched and tempered, electron beam melted, modified SAE 4620 steels consisting of tempered low carbon martensite. Among all the microstructure constituents considered, the constituentR (randomly oriented, “tempered low carbon martensite, TLCM”) achieved the highest probability for dimple fracture. The thick TLCM laths (designated as the microstructure constituent II) exhibited higher probability of dimple plus quasi-dimple mode of fracture than the thin laths (I). It is concluded that the steel EB1035 derived the high toughness from a) the high concentration of the “high toughness” microstructure constituentsR and II, b) “non-embrittled” prior austenite grain boundaries with 50 pct probability for smooth plus quasi-smooth mode and 50 pct dimple plus quasi-dimple mode of intergranular fracture. In contrast, besides having low content ofR and II, the steel EB1014 displayed “completely embrittled” prior austenite grain boundaries with 100 pct probability for smooth plus quasi-smooth intergranular fracture. The conclusions derived from the microconstituentsR, II and I seemed to reflect the “embrittling” effect of decreased spacings between the pseudo twin related laths and between the lath boundary cementite films, and the “toughening” effect of the randomly oriented laths. Auger spectra obtained from the fracture surface before and after sputtering is analyzed to determine the presence of grain boundary sulfur segregation.     

Effects of frequency and temperature on short fatigue crack growth in aqueous environments

The growth of short fatigue cracks in a NiCrMoV steel forging was examined, under constant applied stress intensity range (ΔK = 31 MPa-m^1/2) in deaerated deionized water and 0.3 M Na₂SO₄ solution, as a function of frequency and temperature. Measurements were also made of the kinetics of electrochemical reactions of bare steel surfaces with the deaerated 0.3 M Na₂SO₄ solution, under free corrosion, to provide for comparison and correlation. Fatigue crack growth rate increased with reductions in frequency and with increases in temperature. The maximum amount of crack growth enhancement by the different environments appeared to be equal, although the crack growth response in deionized water appeared to be consistent with a faster reaction rate. The temperature and frequency dependence for corrosion fatigue crack growth corresponded directly with that for charge transfer between the “bare” and “filmed” metal surfaces under free corrosion. The results showed that shortcrack growth in the aqueous environments is controlled by the rate of electrochemical reactions, and is thermally activated with an apparent activation energy of about 40 kJ/M.     

Experimental evidence for electrochemical nature of the reaction between iron oxide in calcia-silica-alumina slag and carbon in liquid iron

The electrochemical nature of the reaction between iron oxide in calcia-silica-alumina slag and carbon in liquid iron has been studied by measuring the kinetics of the slag-metal reaction. A base slag (48 pct CaO-40 pct SiO₂-12 pct Al₂O₃) containing iron oxide (≤8 wt pct FeO _t ) was reduced by an Fe-C metal bath (∼4 wt pct C) at 1400 °C. The reaction rate was calculated from measurements of the total inlet gas flow rate and the CO concentration in the outlet gas stream. The slag was “internally short circuited” by dipping an iron plate through the slag layer, and this resulted in an increase in the rate of CO evolution. An external circuit was produced by dipping a graphite rod (shielded from the slag) into the metal bath and a steel or molybdenum rod into the slag layer; the open-circuit voltage and short-circuit current were measured when iron oxide was added to the base slag layer. The reaction rate was enhanced by applying a voltage across the slag layer, and an electric arc cathode was employed in some of these “electrolysis” experiments.     

Materials science under extreme conditions of pressure and strain rate

Solid-state dynamics experiments at very high pressures and strain rates are becoming possible with high-power laser facilities, albeit over brief intervals of time and spatially small scales. To achieve extreme pressures in the solid state requires that the sample be kept cool, with T _sample<T _melt. To this end, a shockless, plasma-piston “drive” has been developed on the Omega laser, and a staged shock drive was demonstrated on the Nova laser. To characterize the drive, velocity interferometer measurements allow the high pressures of 10 to 200 GPa (0.1 to 2 Mbar) and strain rates of 10⁶ to 10⁸ s⁻¹ to be determined. Solid-state strength in the sample is inferred at these high pressures using the Rayleigh-Taylor (RT) instability as a “diagnostic.” Lattice response and phase can be inferred for single-crystal samples from time-resolved X-ray diffraction. Temperature and compression in polycrystalline samples can be deduced from extended X-ray absorption fine-structure (EXAFS) measurements. Deformation mechanisms and residual melt depth can be identified by examining recovered samples. We will briefly review this new area of laser-based materials-dynamics research, then present a path forward for carrying these solid-state experiments to much higher pressures, P>10³ GPa (10 Mbar), on the National Ignition Facility (NIF) laser at Lawrence Livermore National Laboratory. This article is based on an invited presentation given in the symposium “Dynamic Deformation: Constitutive Modeling, Grain Size, and Other Effects: In Honor of Prof. Ronald W. Armstrong,” March 2–6, 2003, at the 2003 TMS/ASM Annual Meeting, San Diego, California, under the auspices of the TMS/ASM Joint Mechanical Behavior of Materials Committee.     

Deoxidation of molten copper with a rotating graphite cylinder

The kinetics of deoxidation of molten copper by the “vacuum-suction degassing” (VSD) method is investigated. The molten copper is deoxidized by a rotating porous graphite tube immersed in the copper bath. The inside space of the porous graphite tube is evacuated so that the CO gas formed at the graphite-metal interface is removed through the tube wall. The experimental results suggest that the mass transfer of oxygen in the metal phase controls the reaction rate. The kinetic data are arranged with a first-order rate equation. At (ppm O)≥10, the rate constant increases by decreasing the porosity of the graphite and increasing the thickness of the tube wall. This result suggests that the suction of CO gas weakens CO bubble stirring and, thereby, the mass transfer at the tube-melt interface. However, when the rate of CO suction becomes comparable to or larger than the CO gas evolution rate, the effect of CO stirring becomes negligible. This situation appears under the conditions of high porosity and large wall thickness at (ppm O) ≥ 10. At the low oxygen concentration range of (ppm O) ≤ 4, the effect of CO stirring becomes negligible, regardless of the CO suction condition, because of the considerably low CO formation rate. The achievement of deoxidation by the VSD method is evaluated in connection with the final oxygen concentration.     

Plastic flow and fracture of metallic glass

The tensile flow and fracture behavior of three Pdo.₈Sio2-based alloys in the glassy, “microcrystalline,” and fully crystalline condition has been studied. The glassy alloys flow plastically to a total strain of approximately 0.5 pct e, and exhibit proportional limit stresses of approximatelyE x 10~² whereE is Young’s modulus. This plastic flow is accompanied by the formation of shear deformation bands on the specimen surfaces. Fully crystalline alloys are extremely brittle and fracture via intergranular cracking. Fracture surfaces of the amorphous and “microcrystalline” alloys are inclined at 45 deg to the tensile axis and exhibit two morphologically distinct zones. One zone is relatively featureless while the other contains a “river” pattern of local necking protrusions. Detailed comparison of opposing surfaces indicates that fracture is preceded by large local plastic shear which produces the smooth zone while the local necking pattern is produced during rupture. These observations form the basis for the hypothesis that plastic flow in the glassy material occurs via localized strain concentrations and that fracture is initiated by catastrophic, “adiabatic” shear. Formerly Postdoctoral Associate, Yeshiva University, New York, N. Y.