Metall. Res. Technol.
Volume 117, Number 3, 2020
|Number of page(s)||8|
|Published online||01 June 2020|
Freckle formation and prevention in high strength low alloy steel ingots
State Key Laboratory of Advanced Metallurgy, University of Science and Technology Beijing,
100083, PR China
* e-mail: firstname.lastname@example.org
Accepted: 7 May 2020
Freckles considerably limit the development of larger electro-slag remelting (ESR) ingots. To simulate the freckling conditions in ESR ingots, high-strength low alloy (HSLA) steel was solidified with varying solidification front angles. Note that the freckling potential is enhanced and the orientation of freckle channel tends to be in the vertical direction with increase in solidification front angle. This is because the tilted solidification front contributes the available buoyancy and transports light (Si, Mn, Cr)-enriched liquid flows upwards toward the mush zone and then accumulates in the open segregation channel. For freckle formation in experimental HSLA steel ingots, a modified Rayleigh number (Ra) that considers the anisotropy of permeability and solidification front angle was evaluated and a threshold value of Ra that separates the freckled and freckle-free area was determined to be 0.79.
Key words: freckle / segregation / solidification front / modified Rayleigh number
© EDP Sciences, 2020
HSLA steels are used extensively in the production of aircraft undercarriage, power transmission gears, shafts, heavy-duty highway and off-road vehicles, and pressure vessels. This is because of their high strength and toughness and appropriate weldability and atmospheric corrosion resistance [1–3]. To maximize the service life and reliability of large-sized HSLA steel parts, the ESR process has been extensively used as a preferred technology because of its superiority for producing large-sized ingots and improving inner quality [4,5]. However, in large-sized ESR steel ingots, the deep liquid pool with a long mushy zone is prone to decrease the cooling rate and form coarse dendritic structures, which considerably strengthens the inhomogeneity of solutes at a macro scale. Freckles, macrosegregation defects, appear as long black liner trails having a diameter of up to several millimeters. They are indeed a typical reflection of this inhomogeneity initiated by thermo-solutal convection instabilities during solidification . Consequently, they often act as crack initiation sites to reduce the mechanical properties and production lifetime. These ingots with freckles have to be discarded because the freckle size is considerably large; moreover, it is practically impossible to thoroughly eliminate them in the subsequent forging or heat treatment process .
Considerable efforts have been made to study freckle defects and clarify the formation mechanism for improved freckle control, these studies primarily focused on model systems such as NH4Cl aqueous solution [8,9], Al-Cu [10,11], Pb-Sn and Pb-Sb [12,13] binary alloys, particularly Ni-based superalloy ingots [14–18] produced via unidirectional solidification or single crystal solidification. However, for explaining the freckle formation during the ESR process of large-size steel ingots, these studies on freckles in Ni-based superalloy ingots were inappropriate because the solidification conditions are extremely different. Moreover, although several freckles formed in Ni-based alloy ingots during ESR or vacuum arc remelting (VAR) were examined, they always appeared to be downward because of the downward flow [19,20] being different in terms of formation mechanism from freckles in ESR steel ingots that appeared to be upward sloping at the midradius of ingots.
Considering the harmfulness of the freckles, freckle formation is a prevalent topic. Mathematical modelling is a typical method, however, it is tedious for large-sized ingots because the requirement for small mesh size is unreliable sometimes due to the adoption of complex macromodels and uncertainties of input parameters and auxillary models . To simplify freckle prediction, several criteria were proposed. The early prediction approaches were established with the form of GmRn based only on the thermal gradient G and grain growth rate R. Most studies are of Ni-based alloys [8,22,23]. Few studies such as Suzuki [24,25] proposed the similar criterion of GR2.1 for freckle formation in steel ingots. Mehrabian and Flemings  proposed a criterion of VG/CR < −1 based on interdendritic fluid flow velocity V, thermal gradient G, and cooling rate CR; however, this criterion was considered as defining whether a channel will unstably evolve rather than as a criterion for the nucleation of a channel . Furthermore, an important difference is that the criterion form of GmRn and VG/CR does not consider the alloy or steel compositions, a key factor that affects convective fluid instability. For describing the instability phenomenon, the nondimensional Rayleigh number reflecting the ratio of buoyancy forces and the viscous resistance forces is applicable. After being conceived by Rayleigh, multiple forms of Rayleigh number involving density difference have been proposed by researchers [22,28–33]. Compared to the criterion form of GmRn and VG/CR, the Rayleigh number does not only consider thermal conditions but also incorporates the thermal properties related to alloy compositions. An obvious difference among them is selecting the characteristic length. Yang et al.  compared multiple forms of Rayleigh number with varying forms of characteristic length using experimental data of the upward directional solidification of Pb-Sn, Pb-Sb, and Pb-Sn-Sb systems. Furthermore, compared to those considering dendritic arm spacing or depth of mushy zone as characteristic length, they reported the criterion using the ratio of thermal diffusivity Dt and growth rate R as a characteristic length to yield indication of freckles. Ramirez et al.’s study  verified Yang et al.’s conclusion based on Pb-Sn and Ni-based alloy systems. The represented form of Rayleigh number is expressed in equation (1) [18,31–34]. (1) where H = Dt/Rh is the characteristic length scale, is the mean permeability in the vertical direction, (Δρ/ρ0) is the relative liquid density inversion in the mushy zone, Dt is the thermal diffusivity, R is the solidification rate, and ν is the kinematic viscosity of liquid steel.
Another important factor affecting Rayleigh number is permeability. In a segregation-related study, mushy zone is considered as a porous medium with growing solid fraction. The permeability of interdendritic liquid is involved in the Rayleigh number criterion as a critical parameter to describe the ability of liquid to flow in the mushy zone, which is primarily affected by the liquid fraction and dendritic structure. Many studies applied the permeability parameter to evaluate macrosegregation; however, most studies only considered permeability as an isotropic parameter without concern for its anisotropy [13,22,26,31–33]. Consequently, the Rayleigh number cannot yield a good separation to both freckled and freckle-free area. However, the mushy zone has been accepted to be anisotropic. Therefore, the anisotropy of permeability was considered in certain studies by using parallel and perpendicular components of permeability to the primary dendrites [12,14,35–38]. The prediction of freckles using Rayleigh number involving the anisotropy of permeability showed an improvement [12,14].
Although the Rayleigh number as equation (1) considers thermal conditions and thermal properties related to alloy compositions and anisotropic permeability, there are limitations in describing freckles that are frequently found in midradius region in ESR steel ingots rather than in the center where the Rayleigh number is expected to be maximum because of the extended local solidification time. As per Copley et al.’s study , conducted using the model system of an aqueous solution of NH4Cl, this may be related to the orientation and shape of solidification front. Auburtin et al.  reported that the Rayleigh number containing geometry factor was more suitable for prediction freckle formation in superalloys from horizontal directional solidification. However, they only considered the solidification tilt angle of < 35°; the criterion demonstrated a monotonically increasing trend with increasing tilt angle, although the flow driving force decreased at higher tilt angles. Morita et al.  studied the effect of solidification front on the downward freckle formation of 718 alloy via horizontal directional solidification, and then predicted freckle tendency using the Rayleigh number containing the geometry factor. However, the Rayleigh number did not consider permeability. Ramirez et al.  evaluated the effect of solidification tilt angle on the critical Rayleigh number of Ni-based superalloys, but it cannot separate the freckled and freckle-free area well without containing anisotropy of permeability and tilt angle in Rayleigh number. Compared to common freckles in directionally solidified or single crystal-solidified Ni-based alloys with a horizontal front, ESR process shows considerable change in the solidification front because of its “V”-shaped pool; the freckles in ESR steel ingots are always upward sloping. It is important to study the formation and tendency prediction of freckles in ESR steel ingots with varying solidification fronts. However, studies on freckle formation considering geometry factor and prediction of ESR steel ingots have been rarely reported.
In this study, to simulate the freckling conditions in the ESR process, a series of solidification experiments and modeling calculations in terms of anisotropy of permeability and varying solidification front angles were conducted. Consequently, the effect and mechanism of solidification front angle on freckle formation was clarified. To predict the freckle in HSLA steel ingots, a critical value was also quantitatively evaluated using a modified Rayleigh number.
In this study, a HSLA steel with composition of Fe–0.3C–0.97Si–0.96Mn–0.96Cr–0.009P–0.03S was used as experiment steel. Figure 1 shows the solidification apparatus for the experiment: the rectangular high-purity alumina crucible was purged with nitrogen before the experiment. The crucible was then heated from three sides and cooled from one side using an alumina chill block on one side of crucible with argon flowing in and out of it. After steel was completely melted in a furnace, the liquid metal was then poured into the preheated alumina crucible at 1783 K (1510 °C) (15 K above the equilibrium liquidus temperature of steel, calculated using Thermal-Calc). The desired cooling patterns in the casting were achieved by regulating the cooling argon flow rate and the heating power of furnace. The crucible is composed of two layers of high-purity alumina to prevent cracking during solidification. To simulate the titled solidification front condition, designed chillers were specially used for the experiments. Figure 2 shows the schematic of the chill blocks used for the series of experiments in which the tilted angles of chill blocks were set to 30°, 45°, 60° and 90°, respectively. To make the slope portion of chill block as the major cooling interface, a rectangular cavity filled of enclosed air was designed in the chill as an insulation layer. To monitor the temperature field, thermocouples at intervals of 20 mm were inserted in the ingots. The tips were then arranged to be in a line perpendicular to the slope portion of chill block. In this study, under controlled conditions, all ingots were solidified in an attempt to prolong the local solidification time that would provide favorable conditions for freckling.
Solidified castings are longitudinally sliced to reveal the macrostructure and freckles. The dendritic structure was demonstrated by etching the samples using a 5 wt.% trinitrophenol solution. Furthermore, for composition analysis, one sample was cut from a freckle channel using scanning electron microscopy (SEM) and energy-dispersive spectrometry (EDS).
Schematic view of the chillers used in experiment.
Figures 3a–3d show the typical appearance of formed freckles in steel ingots during solidification under elevated tilted solidification front angles of 30°, 45°, 60°, and 90°. This clearly shows the interface between chill block and steel ingot. The primary dendrites grow along the almost same direction because of thermal gradient from lower left to upper right, which is nearly perpendicular to the cooling interface. The black linear defects, freckles, can be observed in certain steel ingots, which stretch from bottom to top but lean slightly from the vertical direction. Moreover, the deviation angle from vertical direction demonstrated a gradual decreasing trend with increase in tilted solidification front angle. Note that the freckle was nearly along the vertical direction when the tilted solidification front angle increased to 90°. Indeed, the solidification front angle affected freckle formation.
Figure 4 shows the dendrite structure around the freckle region. It can be seen clearly that the primary dendrites grow from lower left to upper right. However, in the freckle channel, the equiaxed dendritic structure is reported, which is considerably different from the matrix around freckles. To further identify the difference between freckle and matrix, an EDS analysis was conducted to compare their compositions. The freckle region was reported to be 1.33% Si, 1.31% Mn, and 1.37% Cr; these proportions are much higher than 1.12% Si, 0.97% Mn, and 1.06% Cr in the matrix. This indicates that freckle formation is closely related to segregation of elements.
To understand the formation rule of freckles and their dependence on solidification conditions, the experimentally determined cooling parameters and the resulting conditions of freckled and freckle-free are summarized in Table 1: G is the thermal gradient, CR is the cooling rate, R is the solidification rate, and θ is the tilted solidification front angle, the last column denotes whether a thermocouple is within the freckled or non-freckled region. According to Table 1, freckling could occur under the condition of smaller solidification rate R under the same tilted angle. Moreover, under similar solidification conditions such as cooling rate of 1.21–1.23 mm/min, ingots with larger tilted angle exhibited freckles (D3), whereas those with smaller tilted angle were freckle-free (A5, B4, and C4). This result shows the important effect of the tilted angle on freckle formation, namely, the formation potential of freckles increases with increase in tilted solidification front angle under the same thermal conditions.
In addition to the structures, the compositions of freckles and freckle formation under varying solidification front angles and thermal conditions, it can be concluded that solidification front angle and thermal conditions may be important factors to enhance freckle formation. During solidification, lighter elements such as Si, Mn, and Cr are rejected into interdendritic liquid from dendritic arms. Furthermore, parallel concentration and density gradients are formed along the solidification direction in the mushy zone. At the bottom of the mushy zone, a gravitational instability and light (Si, Mn, Cr)-enriched liquid layer initially exists. Once the density gradient increases to a certain extent with solidification, the buoyancy force of the segregated liquid will be capable to overcome the frictional force; consequently, the segregated liquid flows upward. With the progress of the segregated liquid towards the molten pool, the liquid will be heated but its compositions remain almost constant because thermal diffusion is faster than solute diffusion. Therefore, the liquid remains unsaturated under high temperature. Certain solidified dendritic arms in the segregated channel will eventually be eroded and perhaps remelted by the high-temperature and unsaturated segregated liquid. These dendrite fragments from the erosion of initial dendrites have higher density than the surrounding liquid. They will lower the liquid temperature once they sink in the liquid, and certain partially melted fragments may become nucleation sites. However, the heat transfer direction is irregular in the segregation channel and the solute diffusion is extremely slow in the solid; therefore, the segregation channel has to retain its original compositions to solidify with the equiaxed grain structure. The increasing solidification front angle is inclined to contribute all the available buoyancy to transport the light (Si, Mn, Cr)-enriched liquid flows upwards into the mush zone and accumulate along the upper sidewall. The accumulation occurs before channel formation and contributes to the initiation of an open channel. With a larger angle, the more segregated liquid will flow toward the upper sidewall, thus increasing the propensity of segregation channel formation [14,32].
Reynolds number, ratio of buoyancy driving force, and the viscous friction resistance force for the interdendritic liquid flow are identified as an effective approach to evaluate the thermosolutal convection strength in the mushy zone and predict the initiation of freckle channel in castings. As discussed in the introduction, permeability is a critical parameter to describe the ability of the liquid to flow into the mushy zone. Many Rayleigh number criterion only considered it as an isotropic parameter without concern for its anisotropy [13,22,26,31–33]. A typical form of isotropic permeability as expressed in equation (2) [31,32] was always adopted to calculate Ra in the equation (1). (2) where λ1 is the primary dendrite arm spacing and is the average mushy height or mean solid fraction.
In fact, the mushy zone has been considered to be anisotropic. This is especially important for VAR or ESR ingots with a “U” or “V”-shaped molten pool. However, equation (2) is more appropriate for the upward directional solidification but not for ESR solidification process because of the significant effect of the tilted solidification front. Therefore, in this study, as proposed by Fruech and Poirier [12,39], the anisotropy of permeability was involved using perpendicular and parallel components of permeability to the primary dendrites. The parallel and perpendicular components of permeability are expressed as equations (3) and (4) as functions of liquid fraction fL and primary dendrite arm spacing λ1. The primary dendritic arm spacing can be obtained according to the steel compositions and calculated cooling rate CR listed in Table 1 using equation (5) proposed by Bealy and Thomas et al.  based on the fitting data for various low-alloy steels. (3) (4) (5)
For ESR ingots, freckles are always predominantly found at the midradius of ingots rather than in the center where Ra is maximum if adopting this definition as equation (1). This can be attributed to the orientation and shape of solidification front as the review described in the introduction. The result from Table 1 confirms this conclusion. Therefore, the tilted solidification front should be considered to obtain a better prediction on the freckle formation. The solidification front angle is involved as it is incorporated into permeability equation as equation (6) [14,38]. The modified Ra expressed as equation (7) was eventually obtained by considering liquid composition, density, viscosity, isotropy of permeability, solidification front angle, and cooling conditions. (6) (7) where fL is liquid fraction, and other parameters have been specified in the above description.
A critical Rayleigh number value is an indicator suggesting that, freckles can be expected to form if the Rayleigh number exceeds a certain critical value. To determine a reliable critical value indicating freckle formation by evaluating Ra using equation (7) from the experimental results, the following conditions are essential: (1) reliable data of liquid density and thermophysical properties varying with temperature and composition; (2) accurate measurement of solidification parameters such as cooling rate and temperature gradient; (3) strict control of solidification direction; and (4) reasonable evaluation of permeability within mushy zone depending on the dendritic microstructure. The conditions (2–4) can be obtained from the above experiments with designed solidification front angles. For (1), direct measurements are experimentally difficult because of the high melting point of steel; therefore, liquid density and viscosity were estimated as per liquid composition variation during solidification using Thermal-Calc and Java-based software JMatPro, respectively. To estimate liquid density using Thermal-Calc, the difference between density ρ0 at the liquidus temperature (liquid fraction fL = 100%) and density ρT at a temperature T (TS ≤T ≤ TL, TL and TS are the liquidus and solidus temperature, respectively) was represented by Δρ. JMatPro provides extensive information on how the properties of an alloy or each individual phase may change within its specification range, such as the liquid phase in the mushy zone, which is usually beyond the capability of measurement. The calculation basically comprises calculation of amount and constituent of each phase in steel based on thermodynamics and phase transformation kinetics and property calculation of each individual phase using simple pair-wise mixture model equation (8)  based on its constituent from last step and the property calculation of steel using appropriate mixture models. The General Steel module of JMatPro was adopted to calculate the variation of liquid viscosity during solidification. (8) where P is property of phase, is the property of phase in the pure element, is a binary interaction parameter dependent on the value of ν, and xi and xj are the mole fractions of elements i and j in the phase, respectively. Both and are temperature-dependent and it is possible to include ternary or higher order effects.
Figures 5a and 5b show the variation of the liquid density and viscosity for subsequent calculation of Ra, respectively. These variations are the reflections of Si, Mn and Cr segregation during solidification, which have smaller densities than Fe. During solidification, the liquid fraction gradually decreases with temperature and the steel becomes mushy melt containing both solid and liquid phases. Furthermore, in this work, some lightweight elements such as Si, Mn and Cr in steel segregate into the remaining liquid phase, which causes a change in the composition of the remaining liquid phase. The viscosity of the remaining liquid phase eventually decreases under the synergistic effect of both temperature and composition change. Hirai  proposed an empirical equation for alloy viscosity calculation, the equation suggested that the alloy viscosity is proportional to the liquid density. In this work, the segregation of lightweight elements such as Si, Mn and Cr into the liquid phase leads to a decrease in the liquid phase density; therefore, the liquid viscosity is expected to decrease. This conclusion agrees with the calculation result of JMatPro and indicates that the composition change has a dominant effect on liquid phase viscosity.
Ra was calculated based on the abovementioned and calculated data; moreover, the resulting maximum values for each experiment were plotted against solidification front angle and are presented in Figure 6. The negative Ra indicates the negative liquid density difference, which does not influence the result. Freckles formation can be evaluated using absolute value of Ra; the larger absolute value of Ra suggests a higher formation potential of freckles. Figure 6 shows that a clear-cut boundary divides the freckled and freckle-free regions; below the boundary there is no formation of freckles in ingots solidified under the whole range of solidification front angle. Obviously, the boundary value is independent of the solidification front angle. Based on the clear boundary obtained from the experimental results, a critical threshold value of Ra, 0.79, represented by the horizontal dash line in Figure 6, can be easily derived. The derivation of critical threshold value of Ra indicates it is a good predictor for formation of freckles; otherwise, the appearance of overlapping between freckled and freckle-free regions would decrease the prediction resolution of freckles.
In combination with Table 1, a small tilted front angle and high cooling rate are critical factors to suppress the formation of freckles. Moreover, the ingots with different tilted solidification front angles and cooling rates can yield similar Ra values. This is can be attributed to the anisotropy of permeability within the mushy zone. A decreasing solidification front angle indicates an increasing resistance to the upward liquid flow because of the anisotropy of permeability. It is therefore the resistance for the upward flow of solute-rich liquid decreases as the increase of solidification front angle. Furthermore, with a higher cooling rate, a finer dendritic structure will be formed. This indicates the lower permeability within the mushy zone and the lower level of interdendritic segregation. Accordingly, freckle formation can be avoided by combining a small tilted front angle and intensive cooling conditions during the ESR process. The derivation of critical threshold value of Ra provides a basis for determining suitable solidification front and cooling conditions to avoid freckle formation.
Macrostructure of freckles on the longitudinal sections.
Dendrite structure of freckle region.
Parameters from solidification experiments under different solidification front angles.
Variation of (a) liquid density and (b) liquid viscosity during solidification.
Distribution of maximum Ra under varying solidification front angles.
It was experimentally determined that the freckling potential is enhanced and the orientation of freckle channel tends to the vertical direction with increase in the solidification front angle. Therefore, the key factors for preventing the freckle formation in the larger industrial ESR ingots of HSLA are strong cooling and shallow molten pool. A modified Ra that includes the anisotropic nature of permeability and accounts for its dependence on the solidification front angle was proposed. A critical threshold value of Ra, which reflects freckle-free condition, was determined to be 0.79 for the experimental HSLA steel ingot.
- J. Davis, Alloying: Understanding the basics, ASM International, Materials Park, Ohio, USA, 2001 [Google Scholar]
- I. Low, Advances in ceramic matrix composites (Second Edition): The use of ceramic matrix composites for metal cutting applications, Woodhead Publishing, Cambridge, UK, 2018 [Google Scholar]
- K. Best, Aircr. Eng. 58, 14 (1986) [CrossRef] [Google Scholar]
- S. Maity, N. Ballal, R. Kawalla, ISIJ Int. 46, 1361 (2006) [CrossRef] [Google Scholar]
- M. Chatterjee, M. Balasubramanian, K. Gupt, P. Rao, Ironmak. Steelmak. 17, 28 (1990) [Google Scholar]
- J. Moore, N. Shah, Int. Mat. Rev. 28, 338 (1983) [CrossRef] [Google Scholar]
- D. Ma, Q. Wu, A. Buehrig-Polaczek, Metall. Mater. Trans. B 43, 344 (2012) [CrossRef] [Google Scholar]
- S. Copley, A. Giamei, S. Johnson, M. Hornbecker, Metall. Trans. 1, 2193 (1970) [CrossRef] [Google Scholar]
- C. Beckermann, C. Wang, Metall. Mater. Trans. A 27, 2784 (1996) [CrossRef] [Google Scholar]
- R. Mehrabian, M. Keane, M. Flemings, Metall. Trans. 1, 3238 (1970) [Google Scholar]
- C. Beckermann, C. Wang, Metall. Mater. Trans. A 27, 2765 (1996) [CrossRef] [Google Scholar]
- C. Frueh, D. Poirier, S.D. Felicelli, Mat. Sci. Eng. A 328, 245 (2002) [CrossRef] [Google Scholar]
- W. Yang, W. Chen, K. Chang, S. Mannan, J. Barbadillo, Metall. Mater. Trans. A 32, 397 (2001) [CrossRef] [Google Scholar]
- P. Auburtin, T. Wang, S.L. Cockcroft, A. Mitchell, Metall. Mater. Trans. B 31, 801 (2000) [CrossRef] [Google Scholar]
- K. Morita, T. Suzuki, T. Taketsuru, D. Evans, W. Yang, The tendency for freckle formation in alloy 718, Superalloys 718, 625, 726 and various derivatives, TMS, Warrendale, PA, 2001 [Google Scholar]
- D. Ma, Q. Wu, A. Buehrig-Polaczek, Metall. Mater. Trans. B 43, 344 (2012) [CrossRef] [Google Scholar]
- J. Hong, D. Ma, J. Wang, F. Wang, B. Sun, A. Dong, F. Li, A. Bührig-Polaczek, Materials 9, 929 (2016) [CrossRef] [Google Scholar]
- S. Gao, L. Liu, X. Hu, B. Ge, J. Zhang, H. Fu, J. Mater. Sci. Eng. 28, 145 (2010) [Google Scholar]
- J. Valdes, P. King, X. Liu, Metall. Mater. Trans. A 41, 2408 (2010) [CrossRef] [Google Scholar]
- Z. Long, X. Liu, W. Yang, K. Chang, E. Barbero, Mat. Sci. Eng. A 386, 254 (2004) [CrossRef] [Google Scholar]
- J. Gu, C. Beckermann, Metall. Mater. Trans. A 30, 1357 (1999) [CrossRef] [Google Scholar]
- T. Pollock, W. Murphy, Metall. Mater. Trans. A 27, 1081 (1996) [CrossRef] [Google Scholar]
- K. Yu, J. Oti, W. Walston, J OM 45, 49 (1993) [Google Scholar]
- K. Suzuki, T. Miyamoto, Trans. Iron Steel Inst. Jpn. 18, 80 (1978) [CrossRef] [Google Scholar]
- K. Suzuki, T. Miyamoto, Trans. Iron Steel Inst. Jpn. 20, 375 (1980) [CrossRef] [Google Scholar]
- R. Mehrabian, M. Keane, M. Flemings, Metall. Trans. 1, 1209 (1970) [CrossRef] [Google Scholar]
- A. Mitchell, Mater. Sci. Eng. A 413–414, 10 (2005) [CrossRef] [Google Scholar]
- M. Worster, J. Fluid Mech. 237, 649 (1992) [Google Scholar]
- J. Sarazin, A. Hellawell, Metall. Mater. Trans. A 19, 1861 (1988) [CrossRef] [Google Scholar]
- S. Tewari, R. Shah, Metall. Mater. Trans. A 27, 1353 (1996) [CrossRef] [Google Scholar]
- J. Ramirez, C. Beckermann, Metall. Mater. Trans. A 34, 1525 (2003) [CrossRef] [Google Scholar]
- C. Beckermann, J. Gu, W. Boettinger, Metall. Mater. Trans. A 31, 2545 (2000) [CrossRef] [Google Scholar]
- M. Rad, P. Kotas, C. Beckermann, Metall. Mater. Trans. A 44, 4266 (2013) [CrossRef] [Google Scholar]
- S. Karagadde, L. Yuan, N. Shevchenko, S. Eckert, P. Lee, Acta Mater. 79, 168 (2014) [Google Scholar]
- D. Poirier, Metall. Mater. Trans. B 18, 245 (1987) [CrossRef] [Google Scholar]
- S. Sinha, T. Sundararajan, V. Garg, Int. J. Heat Mass Tran. 35, 2865 (1992) [CrossRef] [Google Scholar]
- M. Schneider, C. Beckermann, Int. J. Heat Mass Tran. 38, 3455 (1995) [CrossRef] [Google Scholar]
- A. Scheidegger, Pure Appl. Geophys. 33, 111 (1956) [CrossRef] [Google Scholar]
- S. Felicelli, D. Poirier, J. Heinrich, Metall. Mat. Trans. B 29, 847–855 (1998) [CrossRef] [Google Scholar]
- M. El-Bealy, B. Thomas, Metall. Mater. Trans. B 27, 689 (1996) [CrossRef] [Google Scholar]
- Z. Guo, N. Saunders, P. Miodownik, J. Electron. Mater. 37, 23 (2008) [CrossRef] [Google Scholar]
- M. Hirai, ISIJ Int. 33, 251 (1993) [CrossRef] [Google Scholar]
Cite this article as: Wei Yan, Yang Zhang, Weiqing Chen, Jing Li, Freckle formation and prevention in high strength low alloy steel ingots, Metall. Res. Technol. 117, 309 (2020)
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.