materials

Article
Freckle Defect Formation near the Casting Interfaces
of Directionally Solidified Superalloys
Jianping Hong 1,2, *, Dexin Ma 2 , Jun Wang 1,3, *, Fu Wang 2 , Baode Sun 1,3 , Anping Dong 1 ,
Fei Li 1 and Andreas Bührig-Polaczek 2
1 Shanghai Key Lab of Advanced High-Temperature Materials and Precision Forming,
Shanghai Jiao Tong University, Shanghai 200240, China; bdsun@sjtu.edu.cn (B.S.);
apdong@sjtu.edu.cn (A.D.); lifei74@sjtu.edu.cn (F.L.)
2 Foundry Institute, RWTH Aachen University, Aachen 52056, Germany; d.ma@gi.rwth-aachen.de (D.M.);
f.wang@gi.rwth-aachen.de (F.W.); sekretariat@gi.rwth-aachen.de (A.B.-P.)
3 State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, China
* Correspondence: jianping.hong@hotmail.com (J.H.); junwang@sjtu.edu.cn (J.W.);
Tel.: +86-21-5474-2962 (J.H.); +86-21-5474-5387 (J.W.)

Academic Editor: Mark T. Whittaker

Received: 31 August 2016; Accepted: 9 November 2016; Published: 16 November 2016

Abstract: Freckle defects usually appear on the surface of castings and industrial ingots during the
directional solidification process and most of them are located near the interface between the shell
mold and superalloys. Ceramic cores create more interfaces in the directionally solidified (DS) and
single crystal (SX) hollow turbine blades. In order to investigate the location of freckle occurrence
in superalloys, superalloy CM247 LC was directionally solidified in an industrial-sized Bridgman
furnace. Instead of ceramic cores, Alumina tubes were used inside of the casting specimens. It was
found that freckles occur not only on the casting external surfaces, but also appear near the internal
interfaces between the ceramic core and superalloys. Meanwhile, the size, initial position, and area of
freckle were investigated in various diameters of the specimens. The initial position of the freckle
chain reduces when the diameter of the rods increase. Freckle area follows a linear relationship in
various diameters and the average freckle fraction is 1.1% of cross sectional area of casting specimens.
The flow of liquid metal near the interfaces was stronger than that in the interdendritic region in the
mushy zone, and explained why freckle tends to occur on the outer or inner surfaces of castings.
This new phenomenon suggests that freckles are more likely to occur on the outer or inner surfaces
of the hollow turbine blades.

Keywords: freckles; convection resistance; interface; directional solidification

1. Introduction
Superalloys offer excellent high temperature tensile strength, stress rupture and creep properties,
fatigue strength, oxidation and corrosion resistance, and micro-structural stability at elevated
temperatures 600 ◦ C or above [1–6]. Nickel-based superalloys are the most complex and the most
widely used in high temperature applications. However, because of the factors of alloy chemistry,
casting conditions including solidification parameters, and casting geometry, freckles, hot cracking,
low angle grain boundaries, and other solidification defects were found in directional solidified (DS)
and single crystal (SX) superalloy castings. These solidification defects have a negative influence on
high temperature mechanical properties, and reduce the life of aero engine and gas turbine land-based
power generation [7–9].
Freckles are macroscopic channel segregation defects which usually appear as a long trail of
equiaxed grains with a composition shift consistent with alloy segregation in a wide variety of

Materials 2016, 9, 929; doi:10.3390/ma9110929 www.mdpi.com/journal/materials

Materials 2016, 9, 929 2 of 17

industrial castings, such as vacuum arc remelting (VAR) and electro-slag remelting (ESR) superalloy
billets, nickel-based superalloys, or specialty steel, that occur during solidification. They are
presently one of the main defects encountered in the advanced casting technology of superalloys [10].
Macro-segregation such as freckles cannot be removed by post-processing, thermo-mechanical
treatments, or plastic processing [11,12]. Since the 1960s, when they were linked to the failure of
several military engines, freckles are considered to be unacceptable defects in industrial aerospace
castings. Freckle defects are highly undesirable in critical applications because of their deleterious
effect on mechanical performance and because they cause considerable economic loss.
It is generally agreed that freckles are the product of specific fluid flow patterns, known as
thermosolutal convection, originating in the interdendritic liquid during solidification. This flow is
driven by a density inversion occurring in the mushy zone as a result of interdendritic segregation.
In a previous study, freckle occurrence was found to be dependent on three factors: alloy chemistry,
casting conditions, and casting size [10]. Much research has been focused on alloy chemistry and
solidification conditions. However, previous empirical knowledge and empirically determined casting
parameters seem to be of little help in finding the new appropriate solidification conditions. The local
thermal gradient (G), local solidification rate (R), and local solidification time (LST) have already been
investigated. Currently, the general acceptant and most complete predictive criteria is the Rayleigh
criterion, which combines two of the three factors influencing freckle formation, alloy chemistry and
casting conditions, but not casting geometry. Since freckling results from the breakdown of a metastable
equilibrium state (heavier liquid atop a lighter one), it is assumed that freckle initiation will always
produce fully grown freckles.
In the present study, a variety of single crystal superalloy castings with various geometries such
as specimens with multi-interfacial features were created in an industrial Bridgman furnace with
a lower withdrawal rate, which is prone to freckle formation. The main goal of this research is freckle
formation on the external surface of rods with a variety of diameters (cross sectional area) and freckle
formation on both the external and internal surfaces of castings with multi-interfaces. Morphology
and composition analysis were investigated with freckle and freckle free (Matrix) regions. Finally,
the formation mechanism of freckle near the interface is discussed.

2. Materials and Methods

2.1. Selection of Superalloys

Superalloy CM247 LC was selected for the investigation of freckle formation in this work. CM247
LC is widely used as DS and SX parts in the highest temperature parts of aero-engines. They are
working under high temperature and high-speed rotating conditions. Hundreds of pieces of single
crystal turbine blades are assembled in a turbofan engine on airplanes. Of course, they should be the
least amount of defects as possible. Unfortunately, in past research work, freckles which appeared
during directional solidification were observed on the Ni-based DS and SX turbine blades. Superalloy
CM247 LC is prone to freckle formation due to the content of elements such as Rhenium, Tungsten,
Titanium, and so on, which is a kind of deleterious macro-segregation defect.
CM247 LC is a polycrystalline cast nickel base superalloy. It is commonly produced using
directional solidification techniques to improve creep rupture strength. It has exceptional high
temperature strength, corrosion, and oxidation resistance. This alloy can also be obtained with
hafnium additions to control the grain boundary structure; in turn, this addition prevents cracking
and improves ductility during processing. CM247 LC is commonly used for blade rings and high
pressure turbine blades. CM247 LC is a chemically modified superalloy derived in late 1970s from
the MAR-M247 composition, specifically designed for DS blade and vane applications. The nominal
composition of CM247 LC is shown in Table 1 [13]. Figure 1 [14] shows the SEM micrograph of the
as-cast microstructure of CM247 LC superalloy, and γ/γ0 eutectic, MC carbide, γ matrix, and γ0 particle
were detected.
Materials 2016, 9, 929 3 of 17
Materials 2016, 9, 929 3 of 17

Materials 2016, 9, 929 3 of 17

Figure 1. SEM micrograph showing the as-cast microstructure of the CM 247 LC superalloy.
Figure 1. SEMTable
micrograph showing the as-cast microstructure of the CM 247 LC superalloy.
1. Nominal composition of the investigated superalloy CM247 LC (in wt %).

Alloy Cr Co Mo W Ta Al Ti Hf B C Zr Ni
Table 1. Nominal
CM247 LC composition
8.1 9.2 0.5 of9.5the3.2
investigated superalloy
5.6 0.7 1.4 CM247
0.015 0.07 0.015 LC
Bal.(in wt %).

2.2. SpecimenCr
Alloy DesignCo Mo W Ta Al Ti Hf B C Zr Ni
CM247
2.2.1.LC 8.1 9.2
Solid Specimens 0.5 9.5 3.2 5.6 0.7 1.4 0.015 0.07 0.015 Bal.
Figure 1. SEM micrograph showing the as-cast microstructure of the CM 247 LC superalloy.
As shown in Table 2, a series of diameters of solid rod specimens were prepared with diameters
2.2. SpecimenofDesign
5 mm, 7 mm, 9 mm,
Table 11 mm,composition
1. Nominal 13 mm, 15 mm,
of theand 19 mm. Figure
investigated 2a shows
superalloy theLC
CM247 3D(indesign
wt %).of a series
of solid rod specimens, and in the bottom of the specimen, a kind of helix (spiral crystal selector) was
Alloy Cr Co Mo W Ta Al Ti Hf B C Zr Ni
used for the single crystal growth, with a height of 65 mm and a 360 degree helix angle. The height
2.2.1. Solid Specimens CM247 LC 8.1 9.2 0.5 9.5 3.2 5.6 0.7 1.4 0.015 0.072 0.015 2 Bal.
of the rod specimens was 150 mm, and the cross sectional areas were 20 mm , 39 mm , 64 mm2, 95
mm2, 133 mm2, 177 mm2 and 283 mm2, respectively. The wax specimens were assembled around a
As shown in Table
2.2. Specimen 2, a series of diameters of solid rod specimens were prepared with
Design diameters
central rod, in a cluster on a wax disk.
of 5 mm, 7 mm, 9 mm, 11 mm, 13 mm, 15 mm, and 19 mm. Figure 2a shows the 3D design of a series
of solid rod2.2.1. Solid Specimens
specimens, Tablein
and 2. the
Solid bottom
specimens of
in diameter (mm) and cross
the specimen, sectional
a kind of area (mm
helix 2).
(spiral crystal selector) was
As shown in Table
used for the single crystal growth, 2, a series of
with No.
Specimen diameters
a height of of
1 65solid
2 mm rod specimens
3 and
4 were
a5 3606 degree
7prepared with
helix angle. diameters
The height of
of 5 mm, 7 mm, 9 mm, 11 Diameter
mm, 13 mm,(mm)15 mm, 5and 719 mm. 9 11Figure13 2a 15
shows 19the
2 3D design 2 of a series
2 , 95 mm2 ,
the rod specimens was 150 mm, Cross
and the
Sectional
cross
area (mmof
sectional
2)
areas
20specimen,
were
39 64 a95kind
20
133of helix
mm
177 (spiral
283
, 39 mm , 64 mm
of solid rod specimens, and in the bottom the crystal selector) was
133 mm2 , 177
usedmm 2 and 283 mm2 , respectively. The wax specimens were assembled around a central
for the single crystal growth, with a height of 65 mm and a 360 degree helix angle. The height
rod, in a cluster
of the on
rod aspecimens
wax disk. was 150 mm, and the cross sectional areas were 20 mm2, 39 mm2, 64 mm2, 95
mm2, 133 mm2, 177 mm2 and 283 mm2, respectively. The wax specimens were assembled around a
central Table
rod, in2.
a cluster in diameter (mm) and cross sectional area (mm2 ).
on a wax disk.
Solid specimens

Table 2. Solid specimens in diameter (mm) and cross sectional area (mm2).
Specimen No. 1 2 3 4 5 6 7
Specimen No. 1 2 3 4 5 6 7
Diameter (mm) 5 7 9 11 13 15 19
Diameter (mm) 5 7 9 11 13 15 19
Cross Sectional area (mm2 ) 20 39 64 95 133 177 283
Cross Sectional
(a) area (mm )2 20 39 64 95 133 177
(b) 283

Figure 2. Schematic diagram of (a) 3D design of a series of solid rod specimens, and (b) possible
freckle formation on the external surface.

2.2.2. Multi-Interface Specimens

With the increasing demand of the aerospace industry, more and more large and complex
superalloy components are required for airplane engines. The complex components, especially
ceramic cores, are applied in the hollow blades that cause multi-interfaces in the casting specimens.

(a) (b)

Figure 2. Schematic diagram of (a) 3D design of a series of solid rod specimens, and (b) possible
Figure 2. Schematic diagram of (a) 3D design of a series of solid rod specimens, and (b) possible freckle
freckle formation on the external surface.
formation on the external surface.
2.2.2. Multi-Interface Specimens

2.2.2. Multi-Interface
With theSpecimens
increasing demand of the aerospace industry, more and more large and complex
superalloy components are required for airplane engines. The complex components, especially
With the increasing
ceramic demand
cores, are applied ofhollow
in the the aerospace industry,
blades that cause more and
multi-interfaces in themore
castinglarge and complex
specimens.
superalloy components are required for airplane engines. The complex components, especially ceramic
cores, are applied in the hollow blades that cause multi-interfaces in the casting specimens.
Materials 2016, 9, 929 4 of 17

In the present work, it was observed that the freckle defects occurred near the interfaces
Materials 2016, 9, 929 4 of 17
of
superalloys and ceramic materials, as shown in Figure 3a. Indeed, this is the first time that the freckle
defects occurredIn thenearpresentthework,
interface
it wasofobserved
the superalloys and ceramic
that the freckle core inside
defects occurred near of
thecasting
interfacesspecimens.
of
superalloys and ceramic materials, as shown in Figure 3a. Indeed, this is the
Furthermore, due to the arrangement of the specimen on the disc in the Bridgman furnace, shown in first time that the freckle
defects occurred near the interface of the superalloys and ceramic core inside of casting specimens.
Figure 3b, the freckle was observed on the shadow side of the specimens, which face the central rod
Furthermore, due to the arrangement of the specimen on the disc in the Bridgman furnace, shown in
and have a lower temperature than the heater side. A schematic diagram of the freckle defects that
Figure 3b, the freckle was observed on the shadow side of the specimens, which face the central rod
occur nearandthe
have multi-interfaces
a lower temperature is shown
than thein Figure
heater 3a;
side. A the dark parts
schematic diagram areofsuperalloys, andthat
the freckle defects the white
parts areoccur
ceramic materials
near the (ceramic
multi-interfaces core inside
is shown in Figureof the casting
3a; the or alumina
dark parts shell mold
are superalloys, outside
and the white of the
casting).parts are ceramic
Because of thematerials
ceramic(ceramic core there
materials, inside ofwasthemore
castingthan
or alumina shell mold
one interface nearoutside of the
the superalloys.
casting).
The ceramic coreBecause
insideofofthetheceramic
castingmaterials,
provided there
an was more than
additional one interface near the superalloys.
interface.
The ceramic core inside of the casting provided an additional interface.
For simplification, ceramic tubes were used to replace the ceramic core inside of the castings.
For simplification, ceramic tubes were used to replace the ceramic core inside of the castings.
CeramicCeramic
tubes withtubesvarious diameters
with various were
diameters weredesigned,
designed, to investigate
to investigate their
their influence
influence of interface
of interface effects effects
to freckling (Figure 3). The size of the ceramic tube and external diameter of the specimen
to freckling (Figure 3). The size of the ceramic tube and external diameter of the specimen are shown are shown
in Table in
3. Table 3.

(a) (b)

Figure 3. Schematic diagram of (a) 3D design of a series of rod specimens with multi-interfacial
Schematic
Figure 3.features diagram
and (b) possibleof (a) 3D
freckle design of
formation onathe
series of rod specimens with multi-interfacial features
multi-interfaces.
and (b) possible freckle formation on the multi-interfaces.
Table 3. Specimens with multi-interfacial features in diameter (mm) and cross sectional area (mm2).
Table 3. Specimens with multi-interfacial
Specimen No. features in
1 diameter
2 3 (mm)
4 5and6cross
7 sectional
8 area (mm2 ).
External diameter (mm) 19 19 19 19 19 19 19 19
Internal
Specimen No. diameter (mm) 1 12 2 34 54 6 58 11 6 12 7 8
Internal cross sectional area (mm2) 0,8 3 13 20 28 50 95 113
External diameter (mm) 19 19 19 19 19 19 19 19
Internal
2.3. Procedure diameter (mm)
of Directional Solidification 1 2 4 5 6 8 11 12
Internal cross sectional area (mm2 ) 0.8 3 13 20 28 50 95 113
During the casting progress, the ceramic shell mold (See Figure 4a) was placed on the copper
chill plate in the Bridgman furnace shown in Figure 4b. The shell mold was preheated, poured with
2.3. Procedure of Directional
the superalloy Solidification
melt (See Figure 4c), and then withdrawn from the heating zone through the baffle
into the cooling zone, shown in Figure 6d. The employed heater and pouring temperatures were 1450
During the casting progress, the ceramic shell mold (See Figure 4a) was placed on the copper chill
°C and 1500 °C, respectively. The single crystal solidification of the components was achieved by a
plate in grain
the Bridgman furnace
selector. A low shown
withdrawal in Figure
velocity of v =4b. The shell
1.0 mm/min mold
was was
applied to preheated,
promote frecklepoured with the
defects.
superalloy melt The solidification experiments with superalloy CM 247 LC were carried out in a Bridgman into
(See Figure 4c), and then withdrawn from the heating zone through the baffle
the cooling zone,
furnace withshown in Figure 4d.
a low temperature The employed
gradient heater
(2.0 to 5.0 K/mm) andand pouring temperatures
a withdrawal were 1450 ◦ C
velocity of 1.0 mm/min.
◦
The temperature of the two annular upper and lower heating elements in the
and 1500 C, respectively. The single crystal solidification of the components was achieved by a grain vacuum Bridgman
selector. furnace
A low are shown in Figure
withdrawal velocity5. Heating
of v =temperature
1.0 mm/min waswas
controlled
applied accurately by twofreckle
to promote annulardefects.
upper
and lower heating elements. The heating rate is 20 K/min in the range of 800 to 1450 °C during the
The solidification experiments with superalloy CM 247 LC were carried out in a Bridgman
heating process, and the shell mold heating temperature was held at 1450 °C when the specimens
furnace with a low temperature
were moved from the hot zone gradient (2.0zone.
to the cold to 5.0After
K/mm) and a withdrawal
mold cooling, the specimens velocity of 1.0 out
were knocked mm/min.
The temperature of the two annular upper and lower heating elements in the vacuum
of the shell mold, and sand blasting was used to clean the surface. The characteristic freckle defects Bridgman
can be
furnace are observed
shown after macro-etching
in Figure 5. Heating with metallographic
temperature was microscope.
controlled The cooling temperature
accurately curve upper
by two annular
and lower of the heatingelements.
heating and the shadow side of the
The heating Ø19
rate ismm casting rod
20 K/min (at arange
in the height of
of 190
800mm from ◦the
to 1450 chill
C during the
plate) during directional solidification process is presented in Figure 6. As shown ◦ in Aune’s research
heating process, and the shell mold heating temperature was held at 1450 C when the specimens were
[15], the solidus and liquidus of alloy CM247 LC are 1282 °C and 1368 °C, respectively. Therefore, the
moved from the hot zone to the cold zone. After mold cooling, the specimens were knocked out of the
shell mold, and sand blasting was used to clean the surface. The characteristic freckle defects can be
observed after macro-etching with metallographic microscope. The cooling temperature curve of the
heating and the shadow side of the Ø19 mm casting rod (at a height of 190 mm from the chill plate)
Materials 2016, 9, 929 5 of 17

Materials 2016, 9, 929 5 of 17

during directional
Materials 2016, 9, 929 solidification process is presented in Figure 6. As shown in Aune’s research 5 of [15],
17
thecooling
solidus and liquidus of alloy CM247 LC are 1282 ◦ C and 1368 ◦ C, respectively. Therefore, the
rate of 2016,
Materials the 9,heating
929 side and the shadow side of the Ø19 mm casting rod are 1.8 K/min 5 of 17 and 1.7
cooling
cooling
K/min, raterate
ofof thetheheating
respectively. heatingside
side and
and the
the shadow
shadowsidesideofofthe Ø19
the mm
Ø19 mm casting rodrod
casting are are
1.8 K/min
1.8 K/minand 1.7and
K/min, cooling rate of the heating side and the shadow side of the Ø19 mm casting rod are 1.8 K/min and 1.7
respectively.
1.7 K/min, respectively.
K/min, respectively.

(a) (b) (c) (d)

(a) (b) (c) (d)
(a)Figure 4. The process of directional
(b) solidification in the industrial (c)
vacuum Bridgman furnace: (a) Shell (d)
Figure The
4. 4.
Figure Theprocess
mold; (b) moldof
process ofdirectional
directional
loaded solidification
furnace;in
solidification
in the Bridgman in the
(c)the
melt industrial
industrial vacuum
vacuum
pouring, and Bridgman
Bridgman
(d) directionally furnace:
furnace:
solidification (a)(a) Shell
Shell
Figure
mold; (b) 4. The
mold
and process
loaded
mold of
in
cooling. directional
the Bridgman solidification
furnace; in
(c) the
meltindustrial
pouring, vacuum
and (d)Bridgman furnace:
directionally
mold; (b) mold loaded in the Bridgman furnace; (c) melt pouring, and (d) directionally solidification (a) Shell
solidification
andmold;
mold
and (b)
mold mold loaded in the Bridgman furnace; (c) melt pouring, and (d) directionally solidification
cooling.
cooling.
and mold cooling.

Figure 5. The temperature of the heating system during the solidification and cooling process in the
Bridgman furnace.

Figure 5. The temperature of the heating system during the solidification and cooling process in the
Figure 5. The
Figure temperature
5. The temperatureofofthe
theheating
heatingsystem
system during
during the solidificationand
the solidification andcooling
coolingprocess
processinin
thethe
Bridgman furnace.
Bridgman furnace.
Bridgman furnace.

Figure 6. The temperature of the heating and the shadow side of the Ø19 mm rod during the
solidification and cooling process (at a height of 190 mm from the chill plate).

2.4. Metallographic and Composition Analysis

A series of rods with varying diameters were directionally solidified in the cluster mold in an
industrial Bridgman furnace. After cooling and being knocked out of the shell mold, the rods were

Figure 6. The temperature of the heating and the shadow side of the Ø19 mm rod during the
Figure6. 6.The
Theand
solidification
Figure temperature of the
cooling process
temperature of the heating
(atheating and
a heightand themm
of 190
the shadow side
from the
shadow of
ofthe
chill
side theØ19
Ø19mm
plate). mmrod
rodduring
duringthethe
solidification and cooling process (at a height of 190 mm from the chill plate).
solidification and cooling process (at a height of 190 mm from the chill plate).
2.4. Metallographic and Composition Analysis
2.4. Metallographic and Composition Analysis
A series of rods with varying diameters were directionally solidified in the cluster mold in an
A series
industrial of rodsfurnace.
Bridgman with varying
After diameters were
cooling and directionally
being knocked outsolidified in the
of the shell cluster
mold, themold in an
rods were
industrial Bridgman furnace. After cooling and being knocked out of the shell mold, the rods were
Materials 2016, 9, 929 6 of 17

2.4. Metallographic and Composition Analysis

A series of rods with varying diameters were directionally solidified in the cluster mold in an
industrial Bridgman furnace. After cooling and being knocked out of the shell mold, the rods were
cut offMaterials
from 2016, 9, 929
the casting 6 of 17
cluster and sand blasted to remove any residue on the surface. Macro-etching
examination with a solution of 50 pct H2 O2 and 50 pct HCl, and a micro-etching in a solution of 60 pct
cut off from the casting cluster and sand blasted to remove any residue on the surface. Macro-etching
alcohol, 40 pct HCl, and 2 g Cu2 Cl·2H2 O was employed to determine whether and where freckles
examination with a solution of 50 pct H2O2 and 50 pct HCl, and a micro-etching in a solution of 60 pct
were alcohol,
present 40 in pct
the HCl,
casting
androds.
2 g Cu2Cl·2H2O was employed to determine whether and where freckles
The microstructures of therods.
were present in the casting freckles on the single crystal superalloy components were characterized
by opticalThe microscopy
microstructures of thescanning
(OM) and freckles electron microscopy
on the single crystal (SEM). Thecomponents
superalloy compositions
were in the
matrix and freckleby
characterized regions
opticalof microscopy
the as-cast CM247
(OM) andLC scanning
were determined by energy dispersive
electron microscopy (SEM). TheX-ray
compositions
analysis (EDS). in the matrix and freckle regions of the as-cast CM247 LC were determined by energy
dispersive X-ray analysis (EDS).
3. Results
3. Results
3.1. Freckle Occurrence on the External Surface
3.1. Freckle Occurrence on the External Surface
Figure 7a shows the CM247 LC rods with diameters of 5 mm, 7 mm, 9 mm, 11 mm, 13 mm, and
Figure 7a shows the CM247 LC rods with diameters of 5 mm, 7 mm, 9 mm, 11 mm, 13 mm, and
15 mm after macro-etching. The freckle chains (Figure 7b, in bright and gray colors) on the surface
15 mm after macro-etching. The freckle chains (Figure 7b, in bright and gray colors) on the surface
were parallel to the solidification direction. The freckles occurred on the external surface of all of the
were parallel to the solidification direction. The freckles occurred on the external surface of all of the
rods with crosscross
sectional areas of 20 mm 2 2 2 95 mm2 , 133 mm2 , and 177 mm2 . Freckle
rods with sectional areas of 20 mm, 239
, 39mm
mm2,,64
64 mm
mm2,, 95 mm2, 133 mm2, and 177 mm2. Freckle
chains chains are marked by circles in Figure 7a, and the initial position ofofthe
are marked by circles in Figure 7a, and the initial position the freckles
freckles (the(the height
height fromfrom
the the
chill) chill)
werewere189 mm, 164164
189 mm, mm,mm,172172
mm,
mm,156
156mm,
mm,158 mm,and
158 mm, and155
155mm,
mm, respectively.
respectively. TheThe initial
initial position
position
of freckles in the
of freckles in 15
themm diameter
15 mm diameter rod
rodwas
waslower
lower than theother
than the otherrods
rods with
with smaller
smaller cross
cross sectional
sectional areas.areas.

Figure 7. Freckle defects on CM247 LC casting rods with diameters of 5 mm, 7 mm, 9 mm, 11 mm, 13
Figure 7. Freckle defects on CM247 LC casting rods with diameters of 5 mm, 7 mm, 9 mm, 11 mm,
mm, and 15 mm after directional solidification in a vacuum Bridgman furnace with a low withdraw
13 mm, and 15 mm after directional solidification in a vacuum Bridgman furnace with a low withdraw
rate of 1.0 mm/min. (a) freckles location; (b) magnification picture of freckles.
rate of 1.0 mm/min. (a) freckles location; (b) magnification picture of freckles.
The cross sections containing freckle chains on the top of the CM247 LC rods (a height of 210
The cross the
mm from cooling
sections chill) with freckle
containing diameters of 5 mm,
chains 7 mm,
on the top 9ofmm,
the11 mm, 13
CM247 LCmm, 15 (a
rods mm, and 19
height ofmm
210 mm
are marked by a yellow frame in Figure 8a–g under metallurgical microscopy.
from the cooling chill) with diameters of 5 mm, 7 mm, 9 mm, 11 mm, 13 mm, 15 mm, and 19 mm are As shown in Figures
marked7b and
by a8,yellow
the freckle defect
frame in is a linear
Figure trailunder
8a–g of equiaxed grains which
metallurgical have different
microscopy. orientations,
As shown and 7b
in Figures
the freckle area is reflected by different colors (in bright and gray colors), compared with the matrix
and 8, the freckle defect is a linear trail of equiaxed grains which have different orientations, and the
area which has a single crystal structure. It appears as dark and white spots in the cross sections of
freckle area is reflected by different colors (in bright and gray colors), compared with the matrix area
the metallographs (a) to (g) in Figure 8. The area of freckles increases with increasing cross sectional
whicharea
hasofa the
single crystalfrom
specimens structure.
20 mm2Ittoappears
283 mm2as dark and white
. Meanwhile, the γ/γ′spots in the
eutectic cross
phase sectionsbyof the
is depicted
metallographs (a) to (g) in Figure 8. The area of freckles increases with increasing cross
arrows in the freckle area, and it can be seen that the freckle area has a higher fraction of γ/γ′ sectional
eutectic area
of thethan
specimens from 20 mm 2 to 283 mm2 . Meanwhile, the γ/γ0 eutectic phase is depicted by arrows
the matrix.
in the freckle area, and it can be seen that the freckle area has a higher fraction of γ/γ0 eutectic than
the matrix.
Materials 2016, 9, 929 7 of 17
Materials 2016, 9, 929 7 of 17

Figure 8. Freckle observed under metallurgical microscopy on the cross section of CM247 LC casting
Figure 8. Freckle observed under metallurgical microscopy on the cross section of CM247 LC casting
rods (at a height of 210 mm from the cooling chill) with diameters of (a) 5 mm; (b) 7 mm; (c) 9 mm;
rods(d)
(at11a mm;
height of 210 mm from the cooling chill) with diameters of (a) 5 mm; (b) 7 mm; (c) 9 mm;
(e) 13 mm; (f) 15 mm; and (g) 19 mm.
(d) 11 mm; (e) 13 mm; (f) 15 mm; and (g) 19 mm.
Materials 2016, 9, 929 8 of 17

MaterialsOccurrence
3.2. Freckle 2016, 9, 929 on Multi-Interfaces 8 of 17

The CM247
3.2. Freckle
Materials 9,LC
929 casting
2016,Occurrence on specimen with multi-interfaces (Specimen No. 8 with an internal8 diameter
Multi-Interfaces of 17
of 12 mmThe andCM247
an external diameter of 19 mm) is shown in Figure 9. After
LC casting specimen with multi-interfaces (Specimen No. 8 with an internal
removing the shell mold
3.2.
material, Freckle
freckle Occurrence
defects on
were Multi-Interfaces
diameter of 12 mm and an observed on the cross
external diameter of 19 mm) section of the in
is shown specimen. Seen removing
Figure 9. After in Figure the 9b, shell
the black
area between
moldThe the
CM247
material, internal
freckle and external
LC casting
defects specimen
were superalloy
withon
observed is the ceramic
multi-interfaces
the cross tube,
of the meanwhile,
section(Specimen No. 8 Seen
specimen. theinan
with black area
9b, near
internal
Figure
the surface
diameter of the
of 12 external
mm and superalloy
an external is the
diametershell
of 19mold.
mm) The
is internal
shown in and
Figure external
9.
the black area between the internal and external superalloy is the ceramic tube, meanwhile, the black After freckle
removing chains
the shellon the
mold
area
vertical material,
near theof
direction freckle
surface defects were
of the external
the specimen observed on
superalloy
are shown the
in Figure cross
is the section
shell
9a,c, of
mold.were
which the specimen.
The separated
internal and Seen in Figure
external
from 9b,
freckleNo. 8.
Specimen
the black
chains
The freckles on area
occurred between
the vertical the
onlyinternal
notdirection onoftheand
the external
specimen
external superalloy
are
surface shownbutinis the on
Figure
also ceramic
9a,c, tube, meanwhile,
which
the internal were the
separated
interface. Oneblack
fromfreckle
chainarea
wasnear
Specimen the
No.surface
observed 8.onThe offreckles
the the external
surface superalloy
occurred
of the not only
inner is on
the the
superalloy shellexternal
mold.
A, The internal
surface
meanwhile, but andfreckle
also
three external
on thechainsfreckle
internal on the
chains on the
interface. Onevertical
freckle direction
chain wasofobserved
the specimen
on the are shownofinthe
surface Figure
inner9a,c, which were
superalloy separated three
A, meanwhile, from
outer superalloy B were observed at the low height position from the cooling chill, and three freckle
Specimen
freckle chainsNo. on 8. The freckles
the outer occurredB not
superalloy wereonly
observedon theatexternal
the low surface but alsofrom
height position on the
theinternal
cooling
chains converged to one chain at the higher position from the cooling chill. Figure 9b indicates that
interface.
chill, and One
threefreckle
frecklechain wasconverged
chains observed on the surface
to one chain atofthe thehigher
inner superalloy
position from A, meanwhile,
the cooling threechill.
freckles nearchains
freckle
Figure 9b
the inner-interface
on the
indicates outer
that
and outer-interface
superalloy
freckles near theB were observed
inner-interface
occurred
at
and the on
lowthe shadow
height
outer-interface
side,from
position
occurred
which had a lower
cooling
on the shadow
cooling rate.
chill, and three
side, which hadfreckle
a lowerchains
coolingconverged
rate. to one chain at the higher position from the cooling chill.
Figure 9b indicates that freckles near the inner-interface and outer-interface occurred on the shadow
side, which had a lower cooling rate.

Figure
Figure 9. Freckle
9. Freckle chain
chain occurs
occurs onon the:(a)
the: (a)internal
internal surface
surface and
and(b)
(b)cross section
cross of specimen;
section (c) external
of specimen; (c) external
surface of CM247 LC.
surface of CM247 LC.
Figure 9. Freckle chain occurs on the: (a) internal surface and (b) cross section of specimen; (c) external
Figure of
surface 10a shows
CM247 the freckles on the cross sectional surface of the inner superalloy A and Figure
LC.
Figure
10b shows10atheshows theonfreckles
freckles the surfaceon ofthe
thecross
outer sectional
superalloysurface of the
B. The depth andinner
widthsuperalloy
of the freckleA and
Figure
chainFigure
10b on the10a
shows shows
the
surface ofthe
freckles
thefreckles
on the
outer onsurface
the cross
superalloy ofsectional
arethe
1569 surface
outer
µm of the
superalloy
and 833 inner
µm, superalloy
B.respectively.
The A and
depthMoreover,
and Figure
width theof the
10b
freckle shows
depth and
chain the
thefreckles
onwidth of theon
surface ofthe
freckle
thesurface
chain
outeronof the surface
the outer superalloy
superalloy of the
are inner
1569 B. superalloy
µm The
anddepth
833 µm, and
is width
1114 of the
µm and freckle
1236
respectively. µm,
Moreover,
chain on the
respectively. surface
The of
area the
of theouter superalloy
freckles on the are
outer 1569
and µm and
inner 833 µm,
superalloy respectively.
is
the depth and width of the freckle chain on the surface of the inner superalloy is 1114 µm and 1236 1.01 mm 2 Moreover,
and 0.89 mmthe2, µm,
depth and width
respectively. It of the freckle
indicates that thechain
side on
and the surface
area of the of the inner
freckle chainsuperalloy
on the outeris 1114 µm and
superalloy B 1236
is µm,
greater
respectively. The area of the freckles on the outer and inner superalloy is 1.01 mm2 and 0.89 mm2 ,
respectively. Theinner
than that on the area superalloy
of the frecklesA. on the outer and inner superalloy is 1.01 mm2 and 0.89 mm2,
respectively. It indicates that the side and area of the freckle chain on the outer superalloy B is greater
respectively. It indicates that the side and area of the freckle chain on the outer superalloy B is greater
than than
that that
on the inner superalloy A.
on the inner superalloy A.

Figure 10. Magnified photos of freckles near the multi-interfaces (cross section of superalloy CM247
LC in Figure 6) on (a) the interface of the inner superalloy A and (b) the interface of the outer
superalloy
Figure B.
10. Magnified photos of freckles near the multi-interfaces (cross section of superalloy CM247
Figure 10. Magnified photos of freckles near the multi-interfaces (cross section of superalloy CM247 LC
LC in Figure 6) on (a) the interface of the inner superalloy A and (b) the interface of the outer
in Figure 6) on (a) the interface of the inner superalloy A and (b) the interface of the outer superalloy B.
superalloy B.
Materials 2016, 9, 929 9 of 17

Materials 2016, 9, 929 9 of 17

The cross sections at a height of 190 mm from the cooling chill, with various diameters of ceramic
tubes, areThe cross sections
shown at a 11a–h
in Figure height of 190 are
and mm asfrom the cooling
follows. chill, with
Freckle various
defects werediameters
found of ceramic
both near the
tubes, are shown in Figure 11a–h and are as follows. Freckle defects were found
inner-interface and outer-interface in the specimens with 11 mm and 12 mm diameters, shown inboth near the inner-
interface
Figure 11 andand outer-interface
Table in thedefects
4. The freckle specimens
werewith 11 mm and
observed on 12
themm diameters,
shadow sideshown
of theinouter-interface
Figure 11
and Table 4. The freckle defects were observed on the shadow side of the outer-interface surface
surface ceramic tubes with diameter of 1 mm, 2 mm, 4 mm, 5 mm, 6 mm, 8 mm, 11 mm, and 12 mm
ceramic tubes with diameter of 1 mm, 2 mm, 4 mm, 5 mm, 6 mm, 8 mm, 11 mm, and 12 mm inside of
inside of the specimens.
the specimens.

Figure 11. Freckles on the specimens with various diameters of ceramic tubes inside: (a) 1 mm;
Figure 11. Freckles on the specimens with various diameters of ceramic tubes inside: (a) 1 mm;
(b) 2 mm; (c) 4 mm; (d) 5 mm; (e) 6 mm; (f) 8 mm; (g) 11 mm; and (h) 12 mm.
(b) 2 mm; (c) 4 mm; (d) 5 mm; (e) 6 mm; (f) 8 mm; (g) 11 mm; and (h) 12 mm.
Materials 2016, 9, 929 10 of 17

Freckle
Table 4.2016,
Materials 9, 929statistics of a series of rod specimens with multi-interfacial features in the diameter.
10 of 17

Table 4. FreckleNo.
Specimen statistics of a series
1 of rod2specimens
3 with multi-interfacial
4 5 features
6 in the 7diameter.8
Specimen
External diameter (mm) No. 19 1 19 2 193 419 5 19 6 197 819 19
Freckle? YES 19YES 19 YES
External diameter (mm) 19 YES
19 19YES 19 YES19 YES
19 YES
Internal diameterFreckle?
(mm) 1 YES 2 YES YES
4 YES5 YES 6 YES YES
8 YES11 12
Freckle? NO
Internal diameter (mm) 1 NO 2 NO4 NO
5 6 NO 8 NO 11 YES
12 YES
Freckle? NO NO NO NO NO NO YES YES
The chemical composition of freckles on the external surface (Inspection zone A1), freckles on the
The chemical composition of freckles on the external surface (Inspection zone A1), freckles on
internal surface (Inspection zone A2), and in the matrix (Inspection zone A3) were investigated by
the internal surface (Inspection zone A2), and in the matrix (Inspection zone A3) were investigated
SEM by
with energy dispersive X-ray analysis (EDS), shown in Figure 9a,b. The results of the chemical
SEM with energy dispersive X-ray analysis (EDS), shown in Figure 9a,b. The results of the
composition
chemicalin zones A1, in
composition A2, andA1,
zones A3A2,
areand
shown in shown
A3 are Figurein
12c.
Figure 12c.

Figure
Figure 12. Composition
12. Composition (in(in
wtwt%)%)ininthe
thematrix
matrix and
and freckle
frecklearea
areaofofas-cast CM
as-cast 247247
CM LC LC
withwith
a 11 amm
11 mm
diameter ceramic tube inside, by SEM with EDS: (a) Inspection zone A1 is the external freckle area,
diameter ceramic tube inside, by SEM with EDS: (a) Inspection zone A1 is the external freckle area,
Inspection zone A3 is the matrix area; (b) Inspection zone A2 is the internal freckle area; (c) EDS
Inspection zone A3 is the matrix area; (b) Inspection zone A2 is the internal freckle area; (c) EDS results.
results.

4. Discussion
4. Discussion

4.1. Influence of Height

4.1. Influence from
of Height Cooling
from Chill
Cooling Chill

Based Based
on theon the present
present experimentsand
experiments andpast
past research
research [16],
[16],freckles
frecklesdodo
notnot
occur on the
occur on bottom of of
the bottom
rods immediately,
rods immediately, butbut start
start from
from a particularheight
a particular height from
from the
thecooling
coolingchill. Metallographic
chill. Metallographicanalysis
analysis
was carried
was carried out out
by by cutting
cutting thesamples
the samples at
at heights
heightsofof
110 mm,
110 150 150
mm, mm,mm,
190 mm,
190 and
mm, 210 mm210
and from
mm thefrom
cooling chill, which is perpendicular to the solidification direction, shown in Figure 13. Freckles were
the cooling chill, which is perpendicular to the solidification direction, shown in Figure 13. Freckles
observed on the cross section a (height of 210 mm) and cross section b (height of 190 mm) on the
were observed on the cross section a (height of 210 mm) and cross section b (height of 190 mm) on
shadow side in the furnace, and no freckle was observed on the section a (height of 150 mm) and
the shadow
section bside in the
(height furnace,
of 110 mm). and no freckle was observed on the section a (height of 150 mm) and
section b (height of 110 mm).
Materials 2016, 9, 929 11 of 17
Materials 2016, 9, 929 11 of 17

Figure 13.Cross
Figure 13. Crosssections
sectionsof CM247 LC at
of CM247 LCdifferent heights
at different from the
heights cooling
from chill (with
the cooling a 5 mm
chill (withdiameter),
a 5 mm
freckles were
diameter), observed
freckles wereabove a height
observed of 150
above mm from
a height of 150the
mm chill
fromplate: (a) height
the chill plate:of
(a)210 mm;of
height (b) height
210 mm;
of
(b)190 mm;of
height (c)190
height
mm;of(c)150 mm;of
height (d)150
height
mm;of(d)110 mm.of 110 mm.
height

Moreover,the
Moreover, theheight
heightofofinitiation
initiationposition
positionofofthe
thefreckle
frecklechain
chainreduces
reduceswhen
whenthethe diameter
diameter ofof rod
rod
increases. As
increases. Asshown
shownin in Figure
Figure 14, 14, the height
the height of freckle
of freckle initiation
initiation positionposition was
was 178.9 mm 178.9
withmm with a
a diameter
diameter of 5 mm, but it decreased to 152.6 mm with a diameter of 19 mm. In the same
of 5 mm, but it decreased to 152.6 mm with a diameter of 19 mm. In the same solidification condition, solidification
condition,
large largespecimens
diameter diameter have
specimens
a higherhave a higher probability
probability and less incubation
and less incubation heightformation.
height of freckle of freckle
formation.
It It isthat
is indicated indicated that theposition
the initiation initiation position
and averageand average
size of thesize of the
freckle freckle
chain chain arewith
are relative relative
rod
with rod diameter in the same solidification
diameter in the same solidification process. process.

Figure 14.
Figure 14. Initiation
Initiation position
position of
of the
the freckle
freckle chain
chain in
in various
various diameters.
diameters.

In Figure 15, the maximum and minimum width of freckle chains on the specimens was
measured using a metalloscope. It indicated that the maximum and minimum width of freckle chain
Materials 2016, 9, 929 12 of 17

In Figure 15, the maximum and minimum width of freckle chains on the specimens was measured
Materials 2016, 9, 929 12 of 17
using a metalloscope. It indicated that the maximum and minimum width of freckle chain increases
when the diameter
increases when the of the rod increases.
diameter of the rodThe maximum
increases. width of the
The maximum widthfreckle
of thechain is 0.94
freckle chainmm when
is 0.94
the mm
diameter
when is the5 diameter
mm, butisit 5rises
mm,to 5.12
but mmtowhen
it rises the when
5.12 mm diameter is 19 mm.
the diameter The
is 19 minimum
mm. width of
The minimum
width ofchain
the freckle the freckle
is 0.40chain is 0.40the
mm when mmdiameter
when theisdiameter is 5itmm,
5 mm, and risesand it rises
to 1.90 mm to when
1.90 mmthewhen the is
diameter
diameter
19 mm. is 19 mm.

Figure 15. Statistics of the maximum and minimum width of freckle chains with incremental diameter
Figure 15. Statistics of the maximum and minimum width of freckle chains with incremental diameter
of the specimens.
of the specimens.
Specimens were cut-up as cross sections near the initiation position, and the primary dendrite
arm spacing (PDAS,
Specimens λ1) wasasmeasured
were cut-up and calculated
cross sections near the using image
initiation processing
position, andsoftware. It indicated
the primary dendrite
that the primary dendrite arm spacing of incremental diameter (incremental cross
arm spacing (PDAS, λ1 ) was measured and calculated using image processing software. It indicated sectional area)
follows the linear relationship in Figure 16. The dot is the experimental date and
that the primary dendrite arm spacing of incremental diameter (incremental cross sectional area)the red line indicates
linear
follows thefitting
linearofrelationship
the primary dendrite
in Figurearm
16. spacing
The dotas is athe
function of the diameter.
experimental date andThethefitting
red lineequation
indicates
is y = 374.2 + 9.07x. The Rayleigh number developed by Sarrazin and Hellawell could
linear fitting of the primary dendrite arm spacing as a function of the diameter. The fitting equation be written as
follows [17]:
is y = 374.2 + 9.07x. The Rayleigh number developed by Sarrazin and Hellawell could be written as
follows [17]:
dρ (1)
= g dT ××Gλ4
Ra = 
ηDT 1 (1)

Figure
Figure 16. 16. Experimental
Experimental date
date andthe
and thefitting
fittingline
lineof
of the
the primary
primary dendrite
dendritearm
armspacing
spacing(λ(λ
1) at) the height
1 at the height
of the
of the freckle
freckle initiation
initiation positionwith
position withdiameters
diameters ofof 5
5 mm,
mm, 77 mm,
mm,99mm,
mm,1111mm,
mm,1313
mm,
mm, 15 15mm, andand
mm,
19
19 mm. mm.
Materials 2016, 9, 929 13 of 17

Materials 2016,
Materials 2016, 9,
9, 929
929 13 of 17
As the primary dendrite arm spacing (λ1 ) increases, the Rayleigh number (Ra ) 13 of 17
increases.
The tendency theofprimary
As the freckledendrite
primary formation armwould
spacingbe(λpromoted when Ra increases. The (Raaresult in Figure
The 16
As dendrite arm spacing (λ11)) increases,
increases, the Rayleigh
the Rayleigh number
number (R )) increases.
increases. The
agrees with the
tendency of Rayleigh numberwould
freckle formation
formation criterion that freckle
be promoted
promoted chain
when Raa occurs preferentially
increases. The result ininFigure
result in the large
Figure size rods
16 agrees
agrees
tendency of freckle would be when R increases. The 16
because
with it has
the a greater
Rayleigh number
with the Rayleigh number λ 1 , which means
criterion that
that freckle
freckle chain
chain occurshas more space
preferentially and
in theless resistance
large
criterion that freckle chain occurs preferentially in the large size rods size rods from
liquid in
becausethe interdendritic
it has a greater λarea in
1, whichthe mushy
means thatzone.
freckle chain has more space and less
because it has a greater λ1, which means that freckle chain has more space and less resistance from resistance from
liquid in
liquid in the
the interdendritic
interdendritic area
area in
in the
the mushy
mushy zone.
zone.
4.2. Influence of Cross Sectional Area
4.2. Influence
4.2. Influence of
of Cross
Cross Sectional
Sectional Area
Area
The area and area percentage of freckles on the cross sections shown in Figure 8a–f were measured.
As shown The
The area and
in Figure
area and area
17,area percentage
the area of freckles
of freckles
percentage of freckles
on the on on the
topthewere cross mm2 , 0.38
sections
0.26sections
cross shown
shown mmin 2 , 0.67
in Figure
mm
Figure 2 , 0.97
8a–f
8a–f were
weremm2 ,
measured.
2
1.29 measured.
mm , 1.90As As
mm shown
2
shown in Figure
, andin3.39
Figure 17,
2 the
mm17,which area
the area of freckles
of freckles
follows the onon the top
the top
linear were 0.26 mm
were 0.26 with
relationship
2, 0.38 mm2, 0.67 mm2,
mm , the
2 0.38cross
mm ,sectional
2 0.67 mm2,area
0.97
of 200.97
mm mm
mm2 , 39
2, 1.29 mm2, 1.90 mm2, and 3.39 mm2 which follows the linear relationship with the cross
2 , 1.29
mmmm 2 , 64, 1.90
2
mm mm2 , 95 ,mm
2 2
and 3.39
, 133mm 2
mmwhich
2
, 177 mm 2
follows the 284
, and linear
mm 2
relationship
, respectively.with the Thecross
area of
sectional area
sectional area of of 20
20 mm
mm22,, 39
39 mmmm22,, 64
64 mm
mm22,, 95
95 mmmm22,, 133
133 mmmm22,, 177
177 mm
mm22,, and
and 284 284 mm mm22,, respectively.
respectively.
freckles increases when the cross sectional area increases.
The area
The area of of freckles
freckles increases
increases when
when the the cross
cross sectional
sectional area area increases.
increases.
As shown in Figure 16, when the primary dendrite arm spacing (λ1 ) increases, the Rayleigh
As shown in Figure 16, when the primary
As shown in Figure 16, when the primary dendrite arm spacing dendrite arm spacing (λ(λ11)) increases,
increases, the the Rayleigh
Rayleigh
number (R
number (R a ) also
(Raa)) also
increases,
also increases,
which
increases, which
has
which has
a higher
has aa higher
tendency
higher tendency
tendency of
of defect
of defect
formation.
defect formation.
formation. However,
However,
However, there
there
there is
is no
is no
no
number
significant change
significant changein theinareathe percentage
area percentage of freckles
percentage when when
of freckles
freckles the crossthe sectional area increases;
cross sectional
sectional it maintains
area increases;
increases; it
significant change in the area of when the cross area it
a constant level,
maintains a and
constantthe average
level, andvalue
the of the
average area
value percentage
of the area of freckles
percentage was
of 1.1%
freckles of the
was cross
1.1% sectional
of the
maintains a constant level, and the average value of the area percentage of freckles was 1.1% of the
area,cross
shown
cross in Figure
sectional
sectional area,18.
area, shown in
shown in Figure
Figure 18.18.

Figure
Figure
Figure 17. Freckle
17. Freckle
17. Freckle area
area
area and
and fittingcurve
fitting
and fitting curveon
curve onthe
on the top
the top of
top of CM247
CM247 LC
CM247LC rods
LCrods for
rodsfor various
for cross
various
various sectional
cross
cross areas
sectional
sectional areas
areas
from
from 20
20 to
to 284
284
from 20 to 284 mm .mm
2
mm 2..
2

Figure 18.
Figure 18. Freckle
Freckle area
area percentage
percentage and
and fitting
fitting curve
curve on
on the
the top
top of
of CM247
CM247 LC
LC rods
rods for
for various
various cross
cross
Figure 18. Freckle
sectional area20
areas from
from percentage
to 284 mm22and
284 mm Thefitting curve
valueon
is the
1.1%top of this
CM247 LC rods for various cross
sectional areas 20 to .. The average
average value is 1.1% from
from figure.
this figure.
sectional areas from 20 to 284 mm2 . The average value is 1.1% from this figure.
Materials 2016, 9, 929 14 of 17

4.3. Interface Effects of Freckle Formation

In the present work, ceramic cores inside of the shell mold created new interfaces
Materials 2016, 9, 929
in the casted
14 of 17
parts. As the results in Figures 8 and 11 show, freckle defects were observed only on the shadow
4.3. Interface
side which faced Effects of rod,
the center Frecklehowever,
Formation no freckle defects were present on the heating side which
In the present work,
faced the heating ring in this work. The ceramic coresshadow-heating
inside of the shell mold created
effect new interfaces
leads in the casted of primary
to the inclination
parts. As the results in Figures 8 and 11 show, freckle defects were observed only on the shadow side
dendrites during directional solidification. Brewster G. and Dong H.B. indicated that the increased
which faced the center rod, however, no freckle defects were present on the heating side which faced
micro-segregation
the heating atring
theinsurface
this work.isThe
the inclinationeffect
shadow-heating of the primary
leads dendrite
to the inclination stem dendrites
of primary with respect to the
during directional
mold wall [18,19]. As shown solidification.
in Figure Brewster G. and branching
19, extensive Dong H.B. indicated that the increased
of the secondary armsmicro-
on the diverging
segregation at the surface is the inclination of the primary dendrite stem with respect to the mold
side, and the subsequent formation of tertiary arms to become new primary stalks, occurs to fill
wall [18,19]. As shown in Figure 19, extensive branching of the secondary arms on the diverging side,
the open bodyand theofsubsequent
liquid atformationthe groove at arms
of tertiary the tomold
becomewall (near the
new primary left
stalks, side
occurs ofthe
to fill the
openmold wall in
Figure 19). body
Consequently,
of liquid at thedendrite
groove at inclination
the mold wall could
(near thepotentially
left side of thelead
moldtowall
theinformation
Figure 19). of surface
Consequently, dendrite inclination could potentially lead to the formation of surface eutectic.
eutectic. Meanwhile, micro-segregation in the mushy zone caused density inversion near the interface
Meanwhile, micro-segregation in the mushy zone caused density inversion near the interface of the
of the mold mold
wall wall
andand casting
castingon theshadow
on the shadow side.side.

Figure 19. Schematic diagram of the inclination of the primary dendrites with respect to the mold
Figure 19. Schematic diagram of the inclination of the primary dendrites with respect to the mold
wall. Micro-segregation causes density inversion near the interface of the mold wall and casting on
wall. Micro-segregation causes density inversion near the interface of the mold wall and casting on the
the shadow side.
shadow side.
As shown in Figure 11, freckle defects occurred near both the internal and external interfaces if
the height and specimen section size were in the range of freckle tendency. In the previous work [20–
As shown26], in Figure
freckle 11,were
chains freckle defects
observed occurred
inside superalloynearparts,both the internal
especially andtoexternal
in the center mid-radiusinterfaces
of if the
VAR/ESR ingots and superalloy billets. A model sketch of freckle formation and the associated fluid
height and specimen section size were in the range of freckle tendency. In the previous work [20–26],
flow pattern for internal freckles was applied in Figure 20a. Freckle defects were caused by fluid flow
freckle chains werethermosolutal
named observed inside superalloy
convection, originatingparts,
in the especially
interdendriticinregion
the center to mid-radius
of the mushy zone duringof VAR/ESR
ingots and superalloy billets.
solidification. The RayleighA number
modelcan sketch
be used oftofreckle formation
predict freckle formation and
withthe
goodassociated
precision, fluid flow
and combines alloy effects and solidification conditions. In Figure 20a, the
pattern for internal freckles was applied in Figure 20a. Freckle defects were caused by fluid flowfreckle chain is surrounded
by dendrites, so the resistance of fluid flow comes from the surrounding dendrites forest in the mushy
named thermosolutal
zone. This kindconvection, originating
of flow was driven in the
by the density interdendritic
inversion occurring as aregion of the mushy zone during
result of micro-segregation,
solidification. The Rayleigh number can be used to predict freckle formation
as shown in Figure 19. The fluid flow rate in the interdendritic region can be obtained withfromgood precision, and
Darcy’s
law, and the driving force (F 1) of the fluid flow came from the density difference and gravity. The
combines alloy effects and solidification conditions. In Figure 20a, the freckle chain is surrounded by
permeability of the mushy zone is a function of λ1, λ2, and the liquid fraction fL, so the permeability
dendrites, sowhere
the freckles
resistance of fluid flow comes from the surrounding dendrites forest in the mushy
occurred inside of superalloy parts is defined as [27–29]:
zone. This kind of flow was driven by the density inversion occurring as a result of micro-segregation,
= ∗ ∗ = ∗ ∗ ∗ (2)
as shown in Figure 19. The fluid flow rate in the interdendritic region can be obtained from Darcy’s
where is the permeability, λ2 is the second dendrite arm space (SDAS), and is a constant.
law, and the driving force (F1 ) of
Therefore, the Rayleigh
the fluid flow came from the density difference and gravity.
number could be written as:
The permeability of the mushy zone is a function of λ1 , λ2 , and the liquid fraction fL , so the permeability
where freckles occurred inside of superalloy parts is defined as [27–29]:

Π = Π ( f L ) ∗ Π (λ1 ∗ λ2 ) = K p ∗ f La ∗ λ1b ∗ λ2c (2)

where Π is the permeability, λ2 is the second dendrite arm space (SDAS), and K p is a constant.
Therefore, the Rayleigh number could be written as:

ρ ∗ Π ( f L ) Π ( λ1 ∗ λ2 )
Ra = g ∗ ∗ ∗ sinα ∗ cos (∅ + α) (3)
υ ∗ fL R
Materials 2016, 9, 929 15 of 17

As shown in Figure 20a, where ∆F(ρ) is the buoyancy force of the density difference in the mushy
∆mg is2016,
zone, Materials the9,dead
929 15 of 17
weight of the fluid, and f 1 is the resistance of fluid flow from the surrounding
the dendrites forest. Hence, the driving force F1 is defined as:
∆ ∗ ∗
= ∗ ∗ ∗ sin ∗ cos ∅ + (3)
∗F = F (ρ) − mg − f (4)
1 1
As shown in Figure 20a, where ∆F(ρ) is the buoyancy force of the density difference in the mushy
Azone,
new∆mg
sketch of dead
is the freckle formation
weight and associated
of the fluid, and f1 is thefluid flow of
resistance pattern for freckles
fluid flow from thenear the interface
surrounding
the dendrites
was applied forest.
in Figure Hence,
20b. The the driving force
permeability ΠF1 1where
is defined as: occurred near both of the internal and
freckles
external interfaces of superalloy parts is defined
= ∆ as:− ∆mg − (4)
A new sketch of freckle formation and associatedafluidbflowcpattern for freckles near the interface
Π1 = A ∗ K p ∗ f L ∗ λ1 ∗ λ2 (5)
was applied in Figure 20b. The permeability where freckles occurred near both of the internal
and external interfaces of superalloy parts is defined as:
where ∆F(ρ) is the buoyancy force of the density difference in the mushy zone, ∆mg is the dead
= fluid
weight of the fluid, and f 2 is the resistance of ∗ flow
∗ ∗from∗ the surrounding the dendrites forest.
(5) So,
the driving force F in this condition is defined as:
where ∆F(ρ) is the buoyancy force of the density difference in the mushy zone, ∆mg is the dead
2
weight of the fluid, and f2 is the resistance of fluid flow from the surrounding the dendrites forest. So,
2 = F (ρ
the driving force F2 in this condition isFdefined ) − mg − f 2
as: (6)
=∆ − ∆mg − (6)
where A is a constant related to the surrounding dendrite forest, with a value range of 1 > A > 0.
where
Equation (5)Aisisdeveloped
a constant related to the surrounding
from Equation (2) baseddendrite
on the forest, with
location ofafreckle
value range of 1 > It
occurrence. > is
0. clear
Equation (5) is developed from Equation (2) based on the location of freckle occurrence.
that the permeability Π1 is less than Π, therefore, f 2 < f 1 , F2 > F1 , and it has more freckle tendency It is clear that
the permeability < , > , and it has more freckle tendency near
near the interface ratheristhanless than , therefore,
inside the parts. Supported by this work, freckles are observed near
the interface rather than inside the parts. Supported by this work, freckles are observed near the
the interfaces in most situations. The liquid flow easily moved because the flow was not completely
interfaces in most situations. The liquid flow easily moved because the flow was not completely
surrounded by dendrites;
surrounded by dendrites;more
morethan
thanhalf
halfofof the
the space wasinstead
space was insteada relatively
a relatively smooth
smooth surface
surface of of
ceramic materials.
ceramic TheThe
materials. resistance
resistanceofofthe
theceramic
ceramic materials’ surface
materials’ surface is is less
less thanthan in the
in the interdendritic
interdendritic
region, which
region, was was
which completely
completelysurrounded
surroundedby bydendrites. Theflow
dendrites. The flowofof metal
metal liquid
liquid nearnear the interfaces
the interfaces
was stronger thanthan
was stronger in the interdendritic
in the interdendritic region
regionininthe
the mushy zone.
mushy zone.

FigureFigure 20. Schematic

20. Schematic diagram
diagram of of freckleformation
freckle formation (a)
(a) inside
insideofofsuperalloy
superalloyandand
(b) (b)
nearnear
the interfaces
the interfaces
of superalloys and ceramic materials.
of superalloys and ceramic materials.
Materials 2016, 9, 929 16 of 17

5. Conclusions
The results indicated an interesting phenomenon, even though in the invariable casting condition
and superalloy components, most of the freckle chains were observed on the shadow side and near
the interface of the superalloys and ceramic materials. In this work, freckles occurred on the external
surface near the interface of the superalloys and shell mold wall, or occurred on the internal surface near
the interface of the superalloys and the wall of ceramic materials inside of the castings. Furthermore,
the initiation position of freckle formation occurred with some regularity, and the height of the freckle
initiation position from the cooling copper chill reduced from small cross section to large cross section
in a series of rods with varying diameters (from Ø5 mm to 19 mm in diameter). This new phenomenon
provides a reminder that freckles are more likely to occur near the interfaces if ceramic cores are used
for the hollow turbine blade process.
In the present work, the specimens with multi-interfaces were directional solidified. Freckle
formation near the interface of multi-interfaces of the superalloy was investigated. The conclusions are
as follows:

1. Freckle defects do not only occur on the external surface of Ni-based superalloys castings, but
also on the internal surface, near the interface of superalloys and ceramic materials.
2. The height of the initiation position of the freckle chain reduces when the rod diameter increases.
Meanwhile, the size of the freckle chain increases when rod diameter increases.
3. Freckle area on the rod follows a linear relationship with the cross sectional area. However, there
is no significant change of the area percentage of freckles, and the average value is 1.1% of the
cross sectional area.
4. Freckle defects occur on the large size internal superalloy parts (Diameter of 11 mm and 12 mm),
but not on the small sizes (1 to 8 mm).

Acknowledgments: The authors are grateful to Sebastian Fischer, Samuel Bogner, Elke Schaberger-Zimmermann,
Elke Breuer, and Dietmar Lembrecht for their assistance with specimen preparation and metallurgical analysis.
The authors gratefully acknowledge the financial support from Deutsche Forschungsgemeinschaft through the
project of Untersuchung der Frecklesbildung in gerichtet erstarrten Superlegierungen (Grant No. BU1072/15-1
and -2) and the Shanghai Science and Technology Committee (Grant No. 14521102902). Jianping Hong would like
to acknowledge the China Scholarship Council (CSC) for supporting his doctoral study at the Foundry Institution
of RWTH Aachen University, Germany.
Author Contributions: Jianping Hong and Dexin Ma. conceived and designed the experiments; Jianping Hong,
Jun Wang, and Fu Wang analyzed the data; Jianping Hong wrote the paper; Fu Wang developed the composites;
Jianping Hong and Fu Wang developed the methodology of the microbiological tests; Jianping Hong performed
the microbiological tests; and Baode Sun, Fei Li, and Andreas Bührig-Polaczek contributed as manuscript advisors.
Each contributor was essential to the production of this work.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Bradley, E.F. Source Book on Materials for Elevated-Temperature Applications; American Society for Metals:
Metals Park, OH, USA, 1969; pp. 275–298.
2. Sabol, G.P.; Stickler, R. Microstructure of nickel-based superalloys. Phys. Status Solidi B 1969, 35, 11–52.
[CrossRef]
3. Jena, A.K.; Chaturvedi, M.C. The role of alloying elements in the design of nickel-base superalloys.
J. Mater. Sci. 1984, 19, 3121–3139. [CrossRef]
4. Bradley, E.F. Superalloys: A Technical Guide; ASM International: Metals Park, OH, USA, 1988.
5. Kotval, P.S. The microstructure of superalloys. Metallography 1969, 1, 251–285. [CrossRef]
6. Sims, C.T. Superalloys: Genesis and Character. In Superalloys II; Sims, C.T., Stoloff, N.S., Hagel, W.C., Eds.;
John Wiley & Sons: Hoboken, NJ, USA, 1987; pp. 3–21.
7. Gao, S.; Liu, L.; Hu, X.; Ge, B.; Zhang, J.; Fu, H. Review of Freckle Defects under Directional Solidification of
Nickel-based Superalloys. J. Mater. Sci. Eng. 2010, 28, 145–151.
Materials 2016, 9, 929 17 of 17

8. Ronan, K. A statistical analysis of variations in hot tear performance and microporosity formation versus
composition in investment cast FSX-414. In Superalloys 2008; The Minerals, Metals & Materials Society:
Pittsburgh, PA, USA, 2008; pp. 157–166.
9. Yang, X.L.; Dong, H.B.; Wang, W.; Lee, P.D. Microscale simulation of stray grain formation in investment
cast turbine blades. Mater. Sci. Eng. A 2004, 386, 129–139. [CrossRef]
10. Auburtin, P.B.L. Determination of the Influence of Growth Front Angle on Freckle Formation in Superalloys.
Ph.D. Thesis, University of British Columbia, Vancouver, BC, Canada, 1998.
11. Auburtin, P.; Wang, T.; Cockcroft, S.L.; Mitchell, A. Freckle formation and freckle criterion in superalloy
castings. Metall. Mater. Trans. B 2000, 31, 801–811. [CrossRef]
12. Fecht, H.-J.; Wunderlich, R.K.; Garandet, J.; Hayashi, M.; Mills, K.C.; Passerone, A.; Quested, P.N.; Ricci, E.;
Schmidt-hohagen, F.; Seetharaman, S. Thermophysical properties of in738Lc, Mm247Lc and Cmsx-4 in the
liquid and high temperature solid phase. Superalloys 2005, 467–476.
13. Harris, K.; Erickson, G.L.; Schwer, R.E. MAR M 247 derivations—CM 247 LC DS alloy CMSX single crystal
alloys properties & performance. In Superalloys 1984; The Minerals, Metals and Materials Society: Pittsburgh,
PA, USA, 1904; pp. 221–230.
14. Huang, H.E.; Koo, C.H. Effect of zirconium on microstructure and mechanical properties of cast fine-grain
CM 247 LC superalloy. Mater. Trans. 2004, 45, 554–561. [CrossRef]
15. Hong, J.; Ma, D.; Wang, J.; Wang, F.; Dong, A.; Sun, B.; Bührig-Polaczek, A. Geometrical effect of freckle
formation on directionally solidified superalloy CM247 LC components. J. Alloys Compd. 2015, 648, 1076–1082.
[CrossRef]
16. Hellawell, A.; Sarazin, J.R.; Steube, R.S. Channel convection in partly solidified systems. Philos. Trans. R. Soc.
B Biol. Sci. 1993, 345, 507–544. [CrossRef]
17. Pollock, T.; Tin, S. Nickel-based superalloys for advanced turbine engines: Chemistry, microstructure and
properties. J. Propuls. Power 2006, 22, 361–374. [CrossRef]
18. Brewster, G.; Dong, H.B.B.; Green, N.R.R.; D’Souza, N. Surface segregation during directional solidification
of Ni-base superalloys. Metall. Mater. Trans. B 2008, 39, 87–93. [CrossRef]
19. Brewster, G.; D’Souza, N.; Ryder, K.S.; Simmonds, S.; Dong, H.B. Mechanism for formation of surface
scale during directional solidification of Ni-base superalloys. Metall. Mater. Trans. A 2012, 43, 1288–1302.
[CrossRef]
20. McDonald, R.J.; Hunt, J.D. Fluid motion through partially solid regions of a casting and its importance in
understanding a type segregation. Trans. Metall. Soc. AIME 1969, 245, 1993–1997.
21. Valdés, J.; King, P.; Liu, X. On the formulation of a freckling criterion for Ni-based superalloy vacuum arc
remelting ingots. Metall. Mater. Trans. A 2010, 41, 2408–2416. [CrossRef]
22. Yu, K.O.; Domingue, J.A.; Maurer, G.E.; Flanders, H.D. Macrosegregation in ESR and VAR processes. JOM
1986, 38, 46–50. [CrossRef]
23. Yu, K.O.; Domingue, J.A. Control of Solidification Structure in VAR and ESR Processed Alloy 718 Ingots.
In Proceedings of the International Symposium on the Metallurgy and Applications of Superalloy 718,
Pittsburgh, PA, USA, 12–14 June 1989.
24. Helms, A.D.; Adasczik, C.B.; Jackman, L.A. Extending the Size Limits of Cast/Wrought Superalloy Ingots.
In Superalloys 1996; The Minerals, Metals & Materials Society: Pittsburgh, PA, USA, 1996; pp. 427–433.
25. Yang, W.; Chang, K.-M.; Chen, W.; Mannan, S.; DeBarbadillo, J. Freckle criteria for the upward directional
solidification of alloys. Metall. Mater. Trans. A 2001, 32, 397–406. [CrossRef]
26. Yang, W.H.; de Barbadillo, J.J.; Morita, K.; Suzuki, T.; Chen, W.; Chang, K.-M. A freckle criterion for the
solidification of superalloys with a tilted solidification front. JOM 2004, 56, 56–61. [CrossRef]
27. Poirier, D.R. Permeability for flow of interdendritic liquid in columnar-dendritic alloys. Metall. Trans. B 1987,
18, 245–255. [CrossRef]
28. Bhat, M.S.; Poirier, D.R.; Heinrich, J.C. Permeability for cross flow through columnar-dendritic alloys.
Metall. Mater. Trans. B 1995, 26, 1049–1056. [CrossRef]
29. Bhat, M.S.; Poirier, D.R.; Heinrich, J.C. A permeability length scale for cross flow through model structures.
Energy 1995, 26, 1091–1092. [CrossRef]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).