Kajikawa Et Al. - 2013 - EFFECT OF SOLIDIFICATION FRONT ANGLE ON FRECKLE FORMATION IN ALLOY 625
Kajikawa Et Al. - 2013 - EFFECT OF SOLIDIFICATION FRONT ANGLE ON FRECKLE FORMATION IN ALLOY 625
Kajikawa Et Al. - 2013 - EFFECT OF SOLIDIFICATION FRONT ANGLE ON FRECKLE FORMATION IN ALLOY 625
Edited by: Matthew J. M. Krane, Alain Jardy, Rodney L. Williamson, and Joseph J. Beaman
TMS (The Minerals, Metals & Materials Society), 2013
Keywords: Freckle, Channel segregation, solidification, solidification front angle, Ni-base alloy
Experimental literatures reported the effect of solidification front In this study, we investigated the effect of the solidification front
angle on freckle formation. Sakurai et al.2) studied the effect of angle on freckle formation in the case of the downward type
the solidification front angle on freckle formation in A286 alloy segregation alloy.
using a horizontal directional solidification furnace and assessed
the effect quantitatively. The result revealed that lowering Si and Experimental procedure
Ti contents and reducing the liquid pool angle, which
corresponded to shallow liquid pool in an actual ESR ingot, were Alloy 625 was selected for this experiment, because the value of
favorable for reducing "A" segregation (channel type the freckle formation index2) of the alloy is close to that of A286
segregation). that was reported by Sakurai et al. If the value is too small, it may
be impossible to form freckle in the tilted solidification front
Wan-hong Yang et al. 3) conducted similar experiments, however, condition. Also Alloy 625 is a typical downward type freckling
they did not distinguish between upward type and downward type alloy8).
segregation. Also, they did not show the macrostructure of their
tilted solidification experiment ingots and it is difficult to know The horizontal directional solidification furnace used for this
the details of the experimental results. Auburtin et al. 4) showed study is schematically illustrated in Figure 1. The furnace is a
the effect of the solidification front angle on freckle formation. rectangular-shaped resistant heat furnace using a Kanthal Super
However, their freckle is not channel type segregation but a heater. An Al2O3 crucible was used for the experiment. The inner
freckle in dendrite arms. size of the crucible was 55Wx190Lx230Hmm. The chiller was
cooled by compressed air. In order to simulate the titled
In terms of liquid density difference, researchers have identified solidification front condition, specially designed chillers were
two types of freckle based on the liquid density difference used for this experiment. The designs of the chillers are shown in
between bulk liquid and solute enriched liquid. Grugel and Figure 2.
Brush5) used Sn-Pn binary alloy and two alloy compositions, Pb-
45mass%Sn and Pb-75mass%Sn, were selected for investigation. The chillers are composed of silicate sand filled in stainless steel
The 45mass% tin alloy is a less-dense solute alloy. In contrast, the casing. In this series of experiments, the chiller angle of the
75mass% tin alloy is a dense solute alloy. Although the 45mass% original straight shape was 90 degrees. 60, 45, and 30 degree
tin alloy showed freckles in the ingots, there are few freckles in chillers were made in addition to 90 degrees. In order to work the
the 75mass% alloy ingots. slope portion as the major cooling interface, silicate sand on the
straight portion of the top side of the chiller was thicker than that
181
on the slope portion. This thicker sand portion acted as a type of
insulation. 1400
1380 CH1
CH2
Cooling air Thermocouple Chiller side 1360 CH3
Solidified CH4
alloy 1 2 3 4 5 6 1340 CH5
Temperature(℃)
Chiller
1320
Heater
Growth 1300
direction
1280
Growth 1260
direction
Crucible 1240
Height :200mm
Width : 55mm 1220
Length: 70mm
1200
Furnace
0 20 40 60 80 100
Time from teeming start(min)
Figure 1. Schematic of the horizontal directional solidification Figure 3. Thermal history of 60 degree chiller ingot
furnace
1400
1380 CH1
Thermo couple CH2
chiller
1360 CH3
1340 CH4
Temperature(℃)
metal
CH5
1320
1300
1280
1260
1240
30° 45° 60° 90° 1220
Chiller angle 1200
0 20 40 60 80 100
Figure 2. The chillers used for the experiment Time from teeming start(min)
The alloy was melted in an induction furnace and poured into the Figure 4. Thermal history of 30 degree chiller ingot
crucible using a small ladle. Just after pouring, cooling air was
blown into the chiller. The cooling rates were controlled by the In each photograph, left hand side arrays are the chiller side. The
flow rate of the compressed air. Thermocouples were inserted into directional solidified structure was obtained in each condition
the crucible in order to acquire the temperature profile during although the primary crystal structures were not perpendicular to
solidification. The ingots were cut perpendicular to the cooling the chiller face. Yellow dotted lines in both figures indicate
interface. The cut plane was polished and macro etched to confirm freckles observed in the ingots
the existence or non-existence of freckles in the ingots. If freckles
were found in the ingot, the ingot was cut on the vertical cross Chiller face angle, chemical composition, liquid density
section of the freckle spot that is close to the chiller face. The difference, and freckle formation index in each experiment are
cooling rate (K/min) and growth rate R(mm/min) are obtained shown in Table 1. Liquid density data10-13) and Thermo-calc14)
from the thermal history of the ingot. By using those two values, with Ni-data15) were used for the liquid density calculation. There
are some deviations in the chemical composition.
freckle formation index = ・ R1.1 was calculated. Chemical
analysis of the alloy was conducted by an X-ray fluorescence Accordingly, liquid densities vary slightly. However, the
method. deviation is not too large to analyze the condition of freckle
formation. The relationship between freckle formation indexes
Results and chiller angles is shown in Figure 7. In this figure, the freckle
formation criterion proposed by Suzuki and Miyamoto8) was used
Examples of the thermal histories of the ingots are shown in for the vertical axis. The criterion is an empirical value come from
Figures 3 and 4. In the 30 degree chiller experiment, the sequence their experimental results. The equation of the criterion is
of 4CH and 5CH was reversed, i.e., 5CH was solidified faster than expressed as follows:
4CH. This may be because the location of 5CH is close to the face
and the liquid mass was small. However the freckle initiation
α = ε R1.1 (1)
point was far from 5CH; therefore, data was considered usable for
freckle formation condition analysis. where α is a critical freckle formation index, ε is the cooling rate,
and R is the growth rate. This α means that if the actual
The vertical cross section of each ingot is shown in Figures 5-6.
solidification condition exceeds the value, freckle forms. There
Hereafter, the straight chiller is described as a 90 degree chiller.
seems to be a relationship between the chiller face angle and
182
freckle formation index, i.e., if the chiller angle is smaller, the
freckle formation index becomes smaller. This tendency is almost 8.0
the same as the result of upward segregation alloy A286 2). Alloy625
ε・R1.1(K/min・(mm/min)1.1)
Freckle formation index
However, it is noteworthy that the result of 45 and 30 degree A286※
chiller. In this angle, freckling tendency is steeply decreases 6.0 ※Sakurai et al.:
compared to that of A286. This figure suggests that the effect of
Freckle prone
Tetsu-to-Hagane,
the solidification front angle on freckle formation in downward 74 (1988), p. 2286
type segregation alloy is larger than that of the upward type 4.0
segregation alloy.
2.0
0.0
0 30 60 90
chiller
8.0
※Sakurai et al.:
Alloy625
ε・R1.1(K/min・(mm/min)1.1)
Tetsu-to-Hagane,
Freckle formation index
4.0
chiller 2.0
45°
chiller
30°
50mm 0.0
50mm
0 30 60 90
Figure 6. Vertical cross section of 45 and 30 degree chiller ingot
Solidification front angle (deg.)
Figure 8. The relationship between freckle formation indexes and
solidification front angles
Table 1: Chemical composition of test alloys (mass %)
Angle Chemical composition ⊿ρ
deg C Ni Cr Mo Al Ti Nb Fe g/cm3 Discussion
90 0.055 64.06 20.35 9.00 0.12 0.23 3.56 2.34 0.0115
60 0.046 63.30 21.18 8.86 0.24 0.19 3.60 2.28 0.0134 As described above, the chiller face angle is not equal to the
45 0.049 63.04 21.23 8.96 0.21 0.19 3.59 2.62 0.0118
30 0.025 63.10 21.36 8.93 0.16 0.17 3.57 2.57 0.0132 solidification front angle. The solidification front angle in the
horizontal directional solidification experiment is around 70 – 80
degrees in the case of the straight shape chiller that gives a 90
degree angle. The freckle formation index proposed by Suzuki
and Miyamoto9) is considered to be the index for the solidification
front angle of 70 – 80 degrees. The freckle formation index
indicates good prediction for freckling location in conventional
ingots 16,17), because the solidification front angle in the side mold
183
is in the same range as the solidification front angle in the
horizontal directional solidified ingot. In the previous study8), the
90 degree chiller and the freckle formation index were used to R
evaluate the freckling tendency of commercial alloys. Our unified
evaluation was successful using the same experimental condition
and the same index for each alloy. The freckle formation index
g・cosθ g・sinθ
worked well for the evaluation.
g
θ
Figure 9. A vector diagram of gravity induced force at the tilted
solidification front
u p c
u u 2u 1 L g (3)
Figure10. Vertical cross section of 45 and 30 degree chiller A286 t 0 cL
ingot
cL
The macro structure of A286 alloy horizontal directional u cL D 2cL (4)
t
solidified ingots using tilted chillers is shown in Figures 9, 10. All
freckle streaks shown here grow upward and the directions are where u is flow rate, D is Diffusion coefficient, p is pressure, is
almost the same regardless of those solidification front angles. dynamic viscosity, g is gravity force, is liquid density, and CL is
Meanwhile, in contrast to conventional cast ingots, solidification chemical composition of the liquid. In the calculation, the order of
proceeds with a more angled solidification front in the secondary the ratio of dynamic viscosity and diffusion coefficient (Pr
melt process such as ESR and VAR. The experiment aimed to number in solute) was accorded with liquid metal and viscosity
simulate the tilted solidification front. The effect of the was increased to prevent turbulent flow.
solidification front angle is considered using a vector diagram. A
vector diagram of gravity induced force at the tilted solidification Maximum flow rates were compared at the steady state using the
front is shown in Figure 9. model. It may be necessary to consider the solute distribution in
the liquid. The flow rate equation by solutal effect is expressed as
the following formula:
184
D 1 Thermosolutal convection in the mushy zone affects freckle
U2 g cL L (5) formation in an ingot, which may account for the deviation.
cL Although there is some deviation in freckle formation indexes
Where g’ is the component of gravity force and L is characteristic derived from the horizontal directional solidification experiment,
length. The value g’ along with the solidification front is g・sin this method showed the tendency of freckling. Simple criterion
Accordingly, the flow rate would relate to the solidification front such as ・R1.1or G・R2.1 is desirable from an industrial perspective.
angle . Figure 10 shows the fluid flow simulation result. Criterion using , R, G may need further modification to make
more precise estimation for freckle formation at tilted
4.00 104 solidification front.
185
solute-enriched liquid would be easy to diffuse at the tilted 3. Simple modification using sin may not be able to evaluate
solidification front in upward type segregation alloys. This may be the effect of the solidification front angle on freckle
the effect of freckle suppression by the tilted solidification front formation.
for upward type segregation. 4. The effect of the solidification front angle on freckle
formation varies depending on the type of freckle, such as
In contrast, solute-enriched liquid is dense in downward type upward type and downward type.
freckling alloys. In the case of downward type segregation, 5. There are two types of freckle suppression mechanism. The
gravity driven force owing to liquid density difference is in the suppression mechanism in downward type segregation is
direction of the reverse of upward type segregation. Therefore, different from that in upward type segregation.
solute-enriched liquid tends to move toward the bottom of the
ingot. If the solidification front is tilted much (freckle growth References
direction arrow in the figure is stuck in the mushy zone), it is
impossible to penetrate into the high solid fraction dendrite forest. 1) K. Kajikawa, M. Tanaka, S. Suzuki, and H. Yamada:
Therefore freckle streak grows in low solid fraction regions, i.e., LMPC2011, Nancy, France, 2011.
freckle streak grows along the solidification plane. Since 2) T. Sakurai, H. Yamada, K. Kitamura, and T. Takenouchi:
downward type segregation may be difficult to grow upward, the Tetsu-to-Hagane, 74, (1988), 2286-2293
growing direction toward high solid fraction suppresses freckle 3) W-H Yang, J. J. deBarbadillo, K. Morita, T. Suzuki, W.
formation. Chen, and K-M Chang: JOM, (2004), 56-61
4) P. Auburtin, S. L. Cockcroft, and A. Mitchell: Superalloys
1996, TMS, Warrendale PA, (1996) 443-450
Mu
Growth 2477-2483
direction 7) N. Streat and F. Weinberg: Metall. Trans., 5 (1974) 2539-
Mu 2548
sh
y Freckle 8) K. Kajikawa, T. Sato, and H. Yamada: LMPC2009, Santa Fe,
Gravity NM, 2009.
growth
induced Gravity 9) K. Suzuki, T. Miyamoto: Tetsu-to-Hagane, 63, (1977), 45-52
Freckle
driving induced 10) K. Mukai, Z. Liu, and K. C. Mills: Metall. Mater. Trans. 36B,
growth
force driving (2005), 255-262
force 11) D. J. Steinberg: Metall. Trans., 5, (1974), 1341-1343
Solid Solid 12) Z. Li, and K. C. Mills: Metall. Mater. Trans. B, 37B, (2006),
781-790
Figure 14 Schematic of assumed freckle suppression mechanism 13) K. Mukai, Z. Li, and L. Fang: Mater. Trans. JIM, 45, (2004),
by solidification front tilting in downward type freckle 2987-2993
14) B. Sundman, B. Jansson and J.-O. Andersson: CALPHAD, 9
Figure 14 shows the schematic of the assumed freckle suppression (1985) 153-190.
mechanism by solidification tilting in the case of downward type 15) N. Saunders: Ni-database. Thermo Tech Ltd., Guilford, 1995.
segregation. Solute-enriched liquid tends to be trapped by the 16) H. Yamada, T. Sakurai, T. Takenouchi: Tetsu-to-Hagane, 75,
solidified layer before freckle formation in the right side of the (1989), 105-112
solidification condition in Figure 14. This may be the suppression 17) H. Yamada, T. Sakurai, T. Takenouchi: Tetsu-to-Hagane, 75,
mechanism of downward type segregation due to the effect of (1989), 97-104
solidification front tilting. Streat and Weinberg’s experimental 18) T. Sawada and K. Kajikawa: LMPC2011, Nancy, France,
result7) could support this idea. The result showed no freckle was 2011
found in their vertical directional solidification experiment using 19) R. Mehrabian, M. Keane and M.C. Flemings, Metall. Trans.,
dense solute alloy. If the angle of the solidification front becomes 1 (1970), 1209-1220.
much smaller, freckle would not form in Alloy625 because of this 20) M.G. Worster: Ann. Rev. Fluid Mech., 29 (1997), 91-122.
mechanism. 21) H.E. Huppert: J. Fluid Mech., 212 (1990), 209-40.
22) P.W. Emms and A.C. Fowler: J. Fluid Mech., 262 (1994),
Conclusions 111-39.
23) F. Chen, J.W. Lu, and T.L. Yang: J. Fluid Mech., 276 (1994),
How the solidification front angle affects freckle formation in the 163-87
downward type segregation alloy was investigated using the 24) M.H. McCay and T.D. McCay: J. Thermophys. Heat
horizontal directional solidification experiment with tilted surface Transfer, 2 (1988), 197-202.
chillers. The following conclusions can be drawn from this study. 25) J.R. Sarazin, A. Hellawell, and R.S. Steube: Phil. Trans. R.
Soc. London, 345 (1993), 507-44.
1. If the chiller angle decreases, the freckle formation index 26) A.K. Sample and A. Hellawell: Metall. Trans. A, 15A (1984),
becomes smaller. The tilting solidification front suppressed 2163-73.
freckle formation in downward type segregation Alloy625. 27) H. Morikawa, S. Onodera, and Y. Arakida: Tetsu-to-Hagane,
2. The effect of the solidification front angle on freckle 44 (1958) 1259-1265
formation in the downward type segregation alloy is larger
than that of the upward type segregation alloy.
186