DaWei Induction Heating Machine Co.

,Ltd
E-mail: inductionheating@gpgyjr.com.cn

Induction Heating Process

During the thixoforming process, it is very important to obtain a uniform temperature, which affects the uniformity of
the solid fraction, throughout the billet. Consequently, a heating method that can provide a suitable temperature profile
throughout the billet must be chosen. Other parameters that must be considered include heating time (in order to
minimize the total processing time), the level of control, and temperature consistency. An optimal design of the
induction coil has been identified that best meets these criteria. In a previous study, the theoretical optimal coil design
was verified through the FEM simulation of the induction heating process by using a general purpose finite element
analysis code, ANSYS. So, in this study, the suitability of the coil design was also demonstrated by conducting
induction-heating experiments. The optimal reheating conditions to apply the thixoforming (thixoforging and semisolid die
casting) process were investigated by varying the reheating time, the holding time, the reheating temperatures, the
capacity of the induction heating system, and the size of the adiabatic material. The final holding time was observed
to be the most important factor in obtaining a fine globular microstructure and to prevent coarsening in the three-step
reheating process.

I. INTRODUCTION the billet and to obtain a globular microstructure for thixo-

forging and semisolid die casting processes by using A356
FOR the thixoforming process, both the coexisting sol- (ALTHIX) alloy with d 3 l 5 76 3 90 (mm) and a 60 Hz
idus-liquidus phase and the reheating conditions to obtain frequency for the induction heating system. Moreover, they
the globular microstructure are very important. In a study on demonstrated the suitability of the coil design by performing
the induction heating process of semisolid materials
reheating experiments.
(SSM), Sebus and Henneberger[1] verified that the design The effects of reheating time, billet size, and holding
of the induction coil had an important influence on the
temperature on microstructures for the semisolid die casting
induction heating process. Matsuura and Kitamura[2]
and thixoforging of Al-6 Pct Si-3 Pct Cu-0.3 Pct Mg alloy
suggested that it is very important not to destroy the
(ALTHIX 86S) have not been reported. Therefore, the objec-
crystallization matrix of semisolid metals due to the change
tive of this study is to determine the effects of the reheating
of the size of spherical dendrite grains during reheating. In a
conditions on the globularization of microstructures. The
study on the induction heating of semisolid aluminum
reheating time, the holding temperature, the holding time,
alloys, Midson et al.[3] pro- posed that a coil design was
the adiabatic material size, and the capacity of the induction
necessary for uniform induction heating. Rudnev and co-
heating system were considered as parameters of the globu-
workers[4–7] showed that the required temperature
larization of microstructure. Particular interest is focused on
distribution along the end of the slug depends on the
frequency, the coil and slug geometry (includ- ing the slug- the solutions avoiding the coarsening phenomena. Moreover,
to-coil air gap and coil overhang), the material properties of as the first part of our investigation, the optimal coil design
the slug, its emissivity, coil refractory, power density, and to reduce the temperature gradient of the ALTHIX 86S billet
cycle time. Hirt and co-workers[8,9] and Garat and co- and to obtain the globular microstructure was theoretically
workers[10–13] have both repeatedly emphasized that it is proposed and manufactured. The suitability of the coil design
important to prevent the coarsening for thixoforming, was demonstrated by using reheating experiments.
because, generally, an average grain size below 100 mm in
the reheated state will be sufficient to ensure a homogeneous II. THE NECESSITY OF INDUCTION COIL
material flow of the semisolid alloy and good dimensional DESIGN
stability during die filling. Kang et al.[14] experimentally
proposed the optimal reheating conditions for thixoforging During induction heating, the relationship between time
and the semisolid die casting processes using SSM (A356 and temperature must be controlled exactly to obtain a uni-
alloys) with d 3 l 5 39 3 85 (mm) and d 3 l 5 76 3 60 form temperature distribution over the entire cross-sectional
(mm). Jung and Kang[15] proposed and manufactured an area. Because the initial eutectic temperature (i.e., initial
solid fraction) in the thixoforming process is the key parame-
ter to filling results in the thixoforming process, an accurately
controllable induction heating method must be selected as
DaWei Induction Heating Machine Co.,Ltd the reheating process.
www.gpgyjr.com.cn www.dw-inductionheating.com For the thixoforming process, the reheating of the billet
E-mail: inductionheating@gpgyjr.com.cn in the semisolid state as quickly and homogeneously as
possible is one of the most decisive aspects. From this point

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 30A, NOVEMBER 1999—2967

of view, the design of the induction coil is very important. If ALTHIX 86S with d 3 1 5 76 3 70 (mm) is assumed
For a real system consisting of coil and billet, the induced to be 11 billets heating for a unit hour to 578 8C, the thermal
heat over the length of the billet is normally not equally capacity (Q) and the production rate (Pr) are calculated by
distributed, and consequently, there is a nonuniform tempera- Stansel’s[17] data (in the case of a temperature rise to 510
ture distribution. Therefore, an important point for the opti- 8C, the thermal capacity and the production rate are Q 5
mization of coil design is to verify the correct relationship 145 kW h/t and Pr 5 0.01 t/h, respectively) and linear
between coil length and billet length. interpolation.
In this study, the optimal coil design of the commercial The minimum heated surface area (As) and the minimum
induction heating system (for semisolid forming: 60 Hz) to heated length (lw) can be determined as follows:
obtain the globular microstructure for variations of SSM
Pt P r 3 Q
and specimen size (diameter and length) was theoretically As 5 5 [3]
proposed and manufactured. The suitability of the coil de- Pa Pa
sign will be demonstrated by using induction heating As
experiments. lw 5 [4]
pd
To determine the coil inner diameter Di and optimal coil
III. COIL DESIGN FOR INDUCTION HEATING length H, recommended air gaps (1/2 (Di 2 d )) for through-
For uniform reheating in this study, the optimal coil length, heating coils and property values to calculate the optimal
H, and a coil inner diameter, Di , of the induction heating coil length are shown in Tables I and II, respectively.
system were designed, as shown in Figure 1. By using linear interpolation with Table I, the coil inner
To consider the main surface power loss in induction diameter Di is calculated, and from the result of Eq. [4], the
heating, the idealized power density (Ps) must be represented optimal coil length H is calculated as follows:
as the actual power density (Pa), which is modified to allow H 5 lw 1 (25 to 75) [5]
for the ratio (d/2d F) of a finite current depth of penetration
(d F) of material and billet diameter (d ). Therefore, from the preceding considerations, the coil
dimensions for the SSM (ALTHIX 86S alloys) with d 3 l
2r a 5 76 3 70 (mm) are proposed in Table III, and the experi-
dF 5 !mv [1]
ments of induction heating are carried out using the
Ps (us 2 uc)idealized designed dimensions.
6.89K (us 2 uc )
Pa 5 5 [2] We demonstrated the suitability of an optimal coil design
us 2 uc d through the finite element modeling simulation of the induc-
uS 2 uC tion heating process by using a general purpose finite ele-
In Eq. [2], 5 k can be obtained from the ment analysis code, ANSYS.*[20] In addition, the results of
(uS 2 uC)idealized
variation curves of temperature in a cylinder with finite *ANSYS is a trademark of Swanson Analysis Systems, Inc., Houston, PA.
current depth of penetration.[16]
In the preceding equation, r a , m, v, k, and us2uc are induction heating CAE based on the optimal coil design for
the resistivity of ALTHIX 86S, the magnetic constant, the variations of SSM and specimen size coincided with those
angular frequency, the thermal conductivity, and the maxi- of the induction heating experiment.[20,21]
mum surface-center temperature difference, respectively.
IV. INDUCTION HEATING EXPERIMENTS
For the thixoforming process, the billet must be reheated
to a semisolid state. The induction heating process is very
important in the forming process of an SSM billet, the pro-
cess not only being necessary to achieve the required SSM
billet state, but also to control the microstructure of the billet.
The SSM used in this study was an ALTHIX 86S alloy
fabricated by electromagnetic stirring by PECHINEY in
France. This SSM is a casting alloy used in the development
of automotive parts. The chemical composition is shown in
Table IV, and the microstructure of the raw material is shown
in Figure 2.
The ALTHIX 86S alloy is machined to d 3 l 5 76 3 70

Table I. Recommended Air Gaps [1/2 (Di 2 d )] for

Through-Heating Coils[16]

Billet Billet Diameter (d, mm)

Frequency Temperature (8C) 0 to 60 60 to 125 125 to 250
50/60 Hz 550 12 12 12
850 12 20 40
Fig. 1—Schematic illustration of induction heating of cylindrical specimen.

2968—VOLUME 30A, NOVEMBER 1999 METALLURGICAL AND MATERIALS TRANSACTIONS A

 Table II. Property Values to Calculate the Optimal Coil Length (Specimen Size: d 3 l 5 76 3 70 mm, f 5 60 Hz,
and k 5 0.58)
Parameters Symbol Unit Values Reference
Maximum surface-center temperature difference us 2 uc K mm 4 —
Current depth of penetration d W/m K 10.7 18
Thermal conductivity k kW/m2 109 19
Idealized power density Ps mVm 22.91 —
Resistivity of ALTHIX 86S ra H/m 0.0639 19
Magnetic constant m rad/s 4p 3 1027 16
Angular frequency v m 120p —
Finite current depth of penetration dF kW/m2 1.64 3 1022 —
Actual power density Pa kW 39.5 —
Thermal power Pt dimensionless 1.33 —
Production rate Pr kW 0.01 t/h 17
Thermal capacity Q m2 133.08 h/t —
Minimum heated surface area As mm 30.67 3 1023 —
Billet diameter d mm 76 —
Minimum heated length lw 141 —

Table III. Designed Dimensions of Induction Heating (mm).[20,21] Thermocouple holes (2-mm diameter) to mea-
Device ( f 5 60 Hz, d 5 10.7 mm,[18] l 5 70 mm, and k sure the temperature are accurately machined at 45 mm from
5 0.58) the surface of the billet and at 10 mm from the lateral of
Minimum the billet. To accurately control the temperature of the SSM,
Billet Coil Inner Heating Optimal K-type CA thermocouples (sheath type) of w 1.6 mm are
Diameter Diameter Length Coil Length inserted into the billet. Thermocouples are calibrated by
(d, mm) (Di , mm) (lw , mm) (H, mm) means of being inserted into water of 100 8C. The accuracy
76 100 141 166 to 216 of thermocouples is approximately 0.2 pct. Data logger TDS-
302 (Tokyo Sokki Kenkyuio Co., Ltd.) was used to receive
the data, and the heating temperature was set to the datum
at thermocouple position (b) in Figure 1.
Table IV. Chemical Compositions of ALTHIX 86S To determine the optimal reheating conditions, the follow-
ing parameters were used for the globularization of the
Si Fe Cu Mn Mg Cr Zn Ti Pb microstructure and a small temperature gradient: the capacity
Minimum (pct) 5.5 — 2.5 — 0.30 — — — — of the induction heating system (Q), the reheating time (ta),
Maximum (pct) 6.5 0.15 3.5 0.03 0.40 — 0.05 0.20 0.03 the holding temperature (Th), the holding time (th), the reheat-
ing step, and the adiabatic material size. The reheating exper-
iments were performed under the conditions in Table V. The
meanings of the symbols used in Table V are the same as
those shown in Figure 3.

V. THE RESULTS OF INDUCTION HEATING

AND DISCUSSION
After induction heating, the microstructure of the SSM
must be globular. Moreover, when the SSM is fed from the
induction heating system to the die, the shape of the SSM
must be maintained. Therefore, the capacity of the induction
heating system (Q), the reheating time (ta), the holding tem-
perature (Th), the holding time (th), the reheating step, and
the adiabatic material size were considered as parameters
of the experiment to observe the globularization of the micro-
structure and a small temperature gradient.
The relationship between temperature and solid fraction of
the ALTHIX 86S alloy is given in Figure 4. The temperatures
Fig. 2—Microstructure for raw material of ALTHIX 86S. corresponding to the solid fractions of 50 and 55 pct are
582 8C and 578 8C, respectively.[10] Generally, these solid
fractions of 50 and 55 pct are for semisolid die casting and
(mm). The experiments of induction heating were performed thixoforging, respectively.
by using an induction heating system (frequency: 60 Hz)
with the capacity of 50 kW. As shown in Figure 1, to achieve VI. SOLID FRACTION 55 PCT
uniform heating, the heating coil of the induction heating Reheating experiments 1 through 19 were performed for
system is made by machining to D0 3 H 5 120 3 180 fs 5 55 pct. From experiments 1 through 10, the experiments

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 30A, NOVEMBER 1999—2969

 Table V. Experimental Conditions for Reheating of Semisolid Aluminum Alloy (ALTHIX 86S), Test Specimen Size:
d 3 1 5 76 3 70 (mm)
Holding Adiabatic
Reheating Time Temperature Holding Time Total Material
ta (min) Th (8C) th (min) Time Capacity Size (mm)
Number ta1 ta2 ta3 Th1 Th2 Th3 th1 th2 th3 (min) Q (kW) D3W3L
1 4 3 1 350 567 578 1 3 2 14 8.398 without
2 4 3 1 350 567 578 1 3 2 14 7.480 without
3 4 3 1 350 567 578 1 3 2 14 5.945 without
4 4 3 1 350 567 578 1 3 2 14 3.912 without
5 4 3 1 350 567 578 1 3 2 14 4.096 without
6 4 3 1 350 567 578 1 3 2 14 4.941 without
7 4 3 1 350 567 578 1 3 2 14 5.292 without
8 4 3 1 350 567 578 1 3 2 14 5.544 without
9 4 3 1 350 567 578 1 3 2 14 5.482 without
10 4 3 1 350 567 578 1 3 2 — 5.012 without
11 10 — — 578 — — 2 — — 12 5.544 50 3 50 3 20
12 12 — — 578 — — 2 — — 14 5.544 50 3 50 3 20
13 8 1 — 567 578 — 3 2 — 14 5.544 50 3 50 3 20
14 8 2 — 567 578 — 1 2 — 15 5.544 50 3 50 3 20
15 4 3 1 350 567 578 1 3 1 13 5.544 50 3 50 3 20
16 4 3 1 350 567 578 1 3 2 14 5.544 50 3 50 3 20
17 4 3 1 350 567 578 1 3 3 15 5.544 50 3 50 3 20
18 4 3 2 350 567 578 1 3 2 15 5.544 50 3 50 3 20
19 4 4 2 350 567 578 1 3 2 16 5.544 50 3 50 3 20
20 4 3 1 350 572 582 1 3 2 14 5.544 50 3 50 3 20
21 4 3 1 350 572 582 1 3 2 14 5.231 50 3 50 3 20
22 4 3 1 350 567 582 1 3 2 14 4.941 50 3 50 3 20
23 4 3 1 350 567 582 1 3 2 14 4.806 50 3 50 3 20
24 4 3 1 350 567 582 1 3 2 14 5.376 50 3 50 3 20
25 4 3 1 350 567 582 1 3 2 14 5.448 50 3 50 3 20
26 4 3 1 350 567 582 1 3 2 14 5.128 53 3 53 3 19
27 4 3 1 350 567 582 1 3 4 16 5.128 53 3 53 3 19
28 4 3 1 350 567 582 1 3 3 15 5.128 53 3 53 3 19

Fig. 3—Input data diagram of reheating conditions to obtain the globu-

lar microstructure. Fig. 4—Relationships between temperature and solid fraction for
ALTHIX 86S.

were carried out at the reheating conditions of ta1 5 4 min,

ta2 5 3 min, ta3 5 1 min, th1 5 1 min, th2 5 3 min, th3 5 Therefore, experiments 11 and 12 were performed at one-
2 min, Th1 5 350 8C, Th2 5 567 8C, and Th3 5 578 8C, with step heating conditions of th1 5 2 min and Th1 5 578 8C,
a changing capacity of induction heating system to determine with the capacity (Q) of 5.544 kW. For experiments 11 and
the optimal capacity according to the heating time. In the 12, when the size of adiabatic materials is D 3 W 3 L 5
case of the capacity (Q) of 5.544 kW, the billets reach the 50 3 50 3 20 (mm), ta1 is 10 and 12 minutes, respectively.
set temperature with a small temperature difference. Figure 5 shows the microstructure of the reheated SSM for

2970—VOLUME 30A, NOVEMBER 1999 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 5—(a) through (c) Microstructure in one-step reheating process of Fig. 6—(a) through (c) Microstructure in two-step reheating process of
semisolid aluminum alloy (experiment 12, 86S, fS 5 55 pct, ta1 5 12 min, semisolid aluminum alloy (experiment 14, 86S, fS 5 55 pct, ta1 5 8 min
Th1 5 578 8C, th1 5 2 min, and Q 5 5.544 kW). ta2 5 2 min, Th1 5 567 8C, Th2 5 578 8C, th1 5 3 min, th2 5 2 min, Q 5
5.544 kW).

experiment 12. The globular microstructure was not obtained

at positions (a), (b), and (c) of Figure 1. Due to the lack of 2 minutes, respectively. Figure 6 shows the microstructure of
sufficient holding time for the separation between solid and reheated SSM for experiment 14 of Table V. The temperature
liquid before and after phase change and for globularization difference of the reheated SSM at the measuring positions
of a small Si primary crystal, one-step reheating is unsuitable is 62 8C. The microstructure of Figure 6 is not globular at
for ALTHIX 86S with d 3 1 5 76 3 70 (mm). positions (a) and (b) of Figure 1, and the globularization is
Experiments 13 and 14 were performed at two-step heat- in progress at the position (c). Due to the lack of sufficient
ing conditions of ta1 5 8 min, th1 5 3 min, th2 5 2 min, Th1 holding time for the separation between solid and liquid
5 567 8C, Th2 5 578 8C, and Q 5 5.544 kW. Moreover, before and after phase change, and for globularization of a
when the size of the adiabatic material is D 3 W 3 L 5 small Si primary crystal, two-step reheating is also unsuitable
50 3 50 3 20 (mm), ta2 (ta2 is the time to reach Th2) is 1 and for ALTHIX 86S with d 3 1 5 76 3 70 (mm).

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 30A, NOVEMBER 1999—2971

 kW. Moreover, when the size of the adiabatic material is D
3 W 3 L 5 50 3 50 3 20 (mm), th3 (th3 is the time to
reach Th3) is 1, 2, and 3 minutes, respectively. In the case
of experiment 17, which obtained the finest globular micro-
structure as shown in Figure 7, the temperature difference
of the reheated SSM at the measuring positions is small.
Figure 8 shows the microstructure of reheated SSM for
experiments 15 through 17. In the case of experiment 15,
globularization was in progress at the thermocouple position
(a) of Figure 1, and a fine globular microstructure was
observed at positions (b) and (c).
In the cases of experiments 16 and 17, a fine globular micro-
structure was obtained at positions (a) through (c) of Figure
1. Comparing experiments 15 through 17, the final reheating
holding time (th3) is 1, 2, and 3 minutes, respectively, and the
temperature difference of reheated SSM at the measuring
positions is small. The microstructure of experiment 15 (per-
Fig. 7—Temperature distributions in three-step reheating process of semi-
formed at the condition of th3 5 1 min) is not globular at posi-
solid alloy ( fS 5 55 pct, ta1 5 4 min, ta2 5 3 min, ta3 5 1 min, Th1 5 350 tion (a) of Figure 1. In the case of experiment 16 performed
8C, Th2 5 567 8C, Th3 5 578 8C, th1 5 1 min, th2 5 3 min, th3 5 3 min, at the condition of th3 5 2 min, the globular microstructure
and Q 5 5.544 kW). was obtained at positions (a) through (c) of Figure 1, but the
degree of the globularization was not improved. The micro-
structure of experiment 17 (performed at the condition of th3
Experiments 15 through 17 of Table V were performed 5 3 min) is finer and more globular than that of experiment
at three-step heating conditions of ta1 5 4 min, ta2 5 3 min, 16 (performed at the condition of th3 5 2 min). Due to the lack
ta3 5 1 min, th1 5 1 min, th2 5 3 min, Th1 5 350 8C, Th2 of sufficient holding time to obtain a fine globular microstruc-
5 567 8C, Th3 5 578 8C, and the capacity of Q 5 5.544 ture by the separation between agglomerated solid before and

Fig. 8—(a) through (c) Microstructure in three-step reheating process of semisolid aluminum alloy (86S, fS 5 55 pct, ta1 5 4 min, ta2 5 3 min, ta3 5 1
min, Th1 5 350 8C, Th2 5 567 8C, Th3 5 578 8C, th1 5 1 min, th2 5 3 min, and Q 5 5.544 kW).

2972—VOLUME 30A, NOVEMBER 1999 METALLURGICAL AND MATERIALS TRANSACTIONS A

after phase change, the reheating conditions of experiment 16 positions is small, and in the case of experiment 18, the
are unsuitable. globular microstructure is obtained at positions (a) through
Experiments 18 and 19 of Table V were performed at the (c) in Figure 1.
conditions of ta1 5 4 min, ta3 5 2 min, th1 5 1 min, th2 5 Compared to experiment 17, which obtained a fine globu-
3 min, th3 5 2 min, Th1 5 350 8C, Th2 5 567 8C, Th3 5 578 lar microstructure, in the case of experiment 18, the tempera-
8C, and the capacity of Q 5 5.544 kW. Moreover, when the ture difference of reheated SSM at the measuring positions
size of the adiabatic material is D 3 W 3 L 5 50 3 50 3 is larger. Experiment 18 shows good globularization, but
20 (mm), ta2 is 3 and 4 minutes, respectively. Figure 9 shows shows cohesion between solid regions at positions (a)
the microstructure of reheated SSM for experiment 18. The through (c) of Figure 1.
temperature difference of reheated SSM at the measuring
VII. SOLID FRACTION 50 PCT
Reheating experiments 20 through 28 were performed for
fs 5 50 pct. From experiments 20 through 26 of Table V,
the experiments are carried out at the reheating conditions
of ta1 5 4 min, ta2 5 3 min, ta3 5 1 min, th1 5 1 min, th2
5 3 min, th3 5 2 min, Th1 5 350 8C, Th2 5 572 8C, and
Th3 5 582 8C, changing the capacity of the induction heating
system to determine the optimal capacity according to the
heating time. The sizes of adiabatic materials are D 3 W
3 L 5 50 3 50 3 20 (mm) from experiments 20 through
25, and D 3 W 3 L 5 53 3 53 3 19 (mm) for experiment
26, respectively. In the case of Q 5 5.128 kW, the billet
reaches the set temperature with a temperature difference
of 62 8C.
The reheating experiments of experiments 27 and 28 of
Table V were performed under the reheating conditions of
experiment 26 with th3 5 4 min and th3 5 3 min (th3 is the
final holding time in a three-step reheating), respectively.
Figures 10 and 11 show the temperature distribution and
microstructure of reheated SSM for experiment 28 of Table
V. The temperature difference of reheated SSM at the mea-
suring positions is small, and in the case of experiment
27, the globular microstructure was coarse at positions (a)
through and (c) of Figure 1. However, in the case of experi-
ment 28, a fine globular microstructure was obtained at
positions (a) through (c). Compared to experiment 28, the
microstructure of experiment 27 becomes coarse by the

Fig. 9—(a) through (c) Microstructure in three-step reheating process of Fig. 10—Temperature distributions in three-step reheating process of semi-
semisolid aluminum alloy (experiment 18, 86S, fS 5 55 pct, ta1 5 4 min, solid aluminum alloy ( fS 5 50 pct, ta1 5 4 min, ta2 5 3 min, ta3 5 1 min
ta2 5 3 min, ta3 5 2 min, Th1 5 350 8C, Th2 5 567 8C, Th3 5 578 8C, th1 Th1 5 350 8C, Th2 5 572 8C, Th3 5 582 8C, th1 5 1 min, th2 5 3 min, th3
5 1 min, th2 5 3 min, th3 5 2 min, and Q 5 5.544 kW). 5 3 min, and Q 5 5.544 kW).

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 30A, NOVEMBER 1999—2973

Fig. 11—(a) through (c) Microstructure in three-step reheating process of semisolid aluminum alloy (86S, fS 5 50 pct, ta1 5 4 min, ta2 5 3 min, ta3 5 1
min, Th1 5 350 8C, Th2 5 572 8C, Th3 5 582 8C, th1 5 1 min, th2 5 3 min, and Q 5 5.128 kW).

growth and coalescence due to longer holding time than that is not a good condition because the size of the solid grain
of experiment 28. is small, but the globularization of the microstructure is not
obtained. As shown in Figure 6, two-step reheating also is
not relevant due to the lack of sufficient holding time to
VIII. COMPARISON BETWEEN SOLID
allow for the separation between solid and liquid before
FRACTION 50 AND 55 PCT
and after phase change and for globularization of a small
Through reheating experiments for fs 5 50 and 55 pct, it Si particle.
was found that the higher the solid fraction, the worse the The holding time of the final step is very important in
accuracy of globularization, but the boundary between the the three-step reheating process. As shown in Figures 8(a)
solid regions and the liquid regions is clear. and (b), if the holding time of the final step is short, the
Figure 12 shows the micrographs magnified to the scale microstructure of the globularization is not obtained or not
of 1000 to observe the eutectic microstructure of experiments accurate due to the lack of sufficient holding time for the
17 and 28, which obtained the finest globular microstructure separation between a solid and a liquid, before and after
for fs 5 55 and 50 pct. Figures 12(a) and (b) show that the phase change, and for globularization by the growth of a
eutectic is melted completely. Therefore, it was found that small Si primary crystal. On the other hand, if the holding
the eutectic must be melted completely at over 572 8C (com-
time of the final step is too long, the microstructure becomes
plete eutectic melting temperature of ALTHIX 86S alloy),
coarse due to the growth and coalescence of solid regions
and the reheating time for the complete eutectic melting is
necessary to obtain a fine globular microstructure. The rise rather than the refining of globular Si primary crystal,[22] as
in temperature does not occur until a sufficient amount of shown in Figure 11(a). Therefore, the optimal holding time
thermal energy to melt the eutectic is provided because so of the final step to obtain a fine globular microstructure
much thermal energy and time are necessary to melt the without agglomeration is 3 minutes.
eutectic.[14,15] Before and after the melting of the eutectic, Figure 13 shows the optimal reheating conditions to obtain
the solid fraction changes rapidly, and a rapid temperature a fine globular microstructure that is available for thixoform-
rise occurs when the eutectic is melted. Due to this tempera- ing according to the alloys and specimen sizes. Compared
ture rise, controlling the reheating temperature is difficult. to the reheating conditions performed by using A356 alloys
Therefore, to homogeneously control the temperature distri- with d 3 l 5 39 3 85 (mm) and d 3 l 5 76 3 60 (mm)
bution and the solid fraction of the SSM, the billet must be by Kang et al.[14] by whom the finest globular microstructure
reheated in three steps. is obtained in one-step reheating conditions (ta1 5 10 min,
It was found that in the case of the ALTHIX 86S alloy th1 5 2 min, Th1 5 5738C, Q 5 3.3 kW) for d 3 l 5 39 3
with d 3 l 5 76 3 70 (mm) (Figure 5), one-step reheating 85 (mm) and in two-step reheating conditions (ta1 5 8 min,

2974—VOLUME 30A, NOVEMBER 1999 METALLURGICAL AND MATERIALS TRANSACTIONS A

 finest globular microstructure is obtained in three-step condi-
tions (ta1 5 4 min, ta2 5 3 min, ta3 5 1 min, th1 5 1 min,
th2 5 3 min, th3 5 2 min, Th1 5 3508C, Th2 5 5758C, Th3
5 5848C, Q 5 8.398 kW, and the size of the adiabatic
material is D 3 W 3 L 5 53 3 53 3 19 (mm)) for fs 5
50 pct and in three-step conditions (ta1 5 4 min, ta2 5 3
min, ta3 5 1 min, th1 5 1 min, th2 5 3 min, th3 5 2 min,
Th1 5 3508C, Th2 5 5708C, Th3 5 5768C, Q 5 12.04 kW,
and the size of the adiabatic material is D 3 W 3 L 5 50
3 50 3 20 (mm)) for the solid fraction of fs 5 55 pct.
In this study using ALTHIX 86S alloys with d 3 l 5 76
3 70 (mm), the finest globular microstructure was obtained
in three-step conditions (ta1 5 4 min, ta2 5 3 min, ta3 5 1
min, th1 5 1 min, th2 5 3 min, th3 5 3 min, Th1 5 3508C,
Th2 5 5678C, Th3 5 5788C, Q 5 5.544 kW, and the size of
the adiabatic material was D 3 W 3 L 5 50 3 50 3 20
(mm)) for fs 5 55 pct and in three-step conditions (ta1 5 4
min, ta2 5 3 min, ta3 5 1 min, th1 5 1 min, th2 5 3 min
th3 5 3 min, Th1 5 3508C, Th2 5 5728C, Th3 5 5828C, Q
5 5.128 kW, and the size of the adiabatic material was D
3 W 3 L 5 53 3 53 3 19 (mm)) for the solid fraction of
fs 5 50 pct.

IX. GRAIN GROWTH MECHANISM OF ALTHIX

86S ALLOY
In the case of eutectic Al-Si alloys, mechanical properties
such as wear resistance, strength, hardness, flowability at
high temperature, and corrosion resistance are excellent due
to the microstructure distributed with Si primary crystals
inside the eutectic matrix. However, a coexisting solidus-
liquidus interval increases with silicon content, and the
Fig. 12—(a) Exp. No. 17, f S 5 55 pct, t a1 5 4 min, t a2 5 3 min, t a3 5 1 microstructure becomes coarse when a strong electromag-
min, Th1 5 350 8C, Th2 5 567 8C, Th3 5 578 8C, th1 5 1 min, th2 5 3 netic stirring force occurs.
min, th3 5 3 min, Q 5 5.544 kW and (b) Exp. No. 28, fS 5 50 pct, ta1 5 Therefore, it is important to prevent the coarsening (over
4 min, ta2 5 3 min, ta3 5 1 min, Th1 5 350 8C, Th2 5 572 8C, Th3 5 582 100 mm) of Si primary crystals for thixoforming, because,
8C, th1 5 1 min, th2 5 3 min, th3 5 3 min, Q 5 5.128 kW. Eutectic
microstructure of semisolid alloy (ALTHIX 86S).
generally, an average grain size below 100 mm in the
reheated state will be sufficient to ensure a homogeneous
material flow of the semisolid alloy and good dimensional
stability during die filling.[9–13,21]
ta2 5 1 min, th1 5 3 min, th2 5 2 min, Th1 5 5758C, Th2 5 The factor commonly used to explain the coarsening phe-
5848C, Q 5 3.00 kW) for d 3 l 5 76 3 60 (mm). nomena of the primary particles is the orientation relation-
According to reported results using A356 alloys with d ship between particles. The experiments were performed by
3 l 5 76 3 90 (mm) by Jung and co-workers,[15,21] the changing the reheating conditions such as solid fraction,

Fig. 13—Input data diagram of optimal reheating conditions to obtain the globular microstructure for variation of alloys and specimen size.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 30A, NOVEMBER 1999—2975

heating time, and reheating holding time. However, no 86S alloy on globular microstructure, grain growth, and the
change of the orientation relationship between particles was prevention of coarsening was researched. The ALTHIX 86S
detected through experiments. alloy contains the composition of Al-6 pct Si-3 pct Cu-0.3
It is seen that the bonding between particles in the reheat- pct Mg and strengthens by age hardening.
ing process does not occur along certain crystallographic In the case of ALTHIX 86S, segregation in which silicon
directions but occurs in disordered directions, and the globu- and copper are the principal alloying elements often occurs
larized particles are coarse due to the coalescence between as a defect in the forming process.[21,27,28] However, a uni-
particles. Therefore, to prevent the coarsening of the parti- form microstructure without segregation can be obtained in
cles, it is important to hold a relevant temperature for a the thixoforming process.
constant time during each step, and in the case of three- In this study, because a fine and completely melted eutec-
step reheating, it is especially important in the final step. tic Si phase of ALTHIX 86S was obtained, it is expected
Induction heating of an aluminum billet over a diameter of that thixoforming parts with good mechanical properties will
30 mm is effective at low frequencies. However, in the case be produced.
of low frequencies, if the holding time at high temperatures
is long, coarsening occurs through Ostwald growth, coales-
X. CONCLUSIONS
cence, and separation processes due to a strong electromag-
netic stirring force.[21,23] To apply the thixoforging and semisolid die casting pro-
In order to prevent the A356 alloy with d 3 1 5 76 3 cesses, the optimal coil design of ALTHIX 86S alloys with
90 (mm) from coarsening during reheating, it should be held d 3 1 5 76 3 70 (mm) was theoretically proposed and
for 2 minutes at 576 8C, and in the case of three-step reheat- manufactured. The suitability of the coil design was demon-
ing, that is the holding time during the final step.[15] A356 strated by conducting induction-heating experiments. Based
and Al2024 alloys with d 3 1 5 60 3 90 (mm) are held on the experiment, the following can be summarized.
for 1 minute at 576 8C and 616 8C, respectively.[20,24] In this
1. For the ALTHIX 86S billet with d 3 1 5 76 3 70 (mm)
study, the ALTHIX 86S alloy with d 3 1 5 76 3 70 (mm) (the frequency of the induction heating system: 60 Hz)
was used and held for 3 minutes at 578 8C. often used in the thixoforming process, the optimal coil
Silicon primary crystals are simultaneously destroyed and design for induction heating was proposed.
coalesced by collisions between themselves, resulting in 2. This study shows that the larger the billet size, the better
polycrystals with low-and high-angle boundaries. Further- the multistep reheating compared with reported results,
more, after the fracture and coalescence processes progress, and the heating time and the capacity of the induction
the interface of Si primary crystals are gradually rounded heating system must be increased. We could see that the
by their collisions.[21,25] The fracture, coalescence, and wear final holding time of 3 minutes is suitable to secure a
of Si primary particles occur compositively, and the initial globular microstructure.
facet Si particle is globularized at the end of induction 3. If the reheating holding time of the final step is too short,
heating. a globular microstructure cannot be obtained and, if the
Research on the globularization phenomena beside a coex- holding time is too long, because the risks of grain coars-
isting solidus-liquidus interval has focused mainly on alloys ening may be increased. A fine globular microstructure
with a nonfaceted interface. In the case of the Al2024 alloy is obtained with a holding time of 3 minutes at the three-
with a nonfaceted interface, the interfacial energy between step reheating process.
solid and liquid does not change for the variation of crystallo-
graphic directions.[21,26]
However, in the ALTHIX 86S alloy with a faceted inter-
face, the flat interface is a simple cubic crystal surrounded
by a dense facet with a lower interfacial energy because
the interfacial energy between solid and liquid changes for
variations of crystallographic directions.[21,25] Therefore, it
is seen that during the solidification of alloys, the globulari-
zation of alloys with a faceted interface such as Si primary
crystals of the ALTHIX 86S alloy can occur due to the REFERENCES
higher entropy of the interfacial energy compared to alloys
with a nonfaceted interface. This is due to the greater differ- 1. R. Sebus and G. Henneberger: Proc. 5th Int. Conf. on Semi-Solid
Processing of Alloys and Composites, Golden, CO, June 1998, A.K.
ence in structure and bonding between the solid and liquid Bhasin, J.J. Moore, K.P. Young, and S. Midson, eds., Colorado School
phases as compared to metals, which exhibit only very small of Mines, Golden, CO, 1998, pp. 481-87.
differences between the two phases. More research on the 2. F. Matsuura and S. Kitamura: Proc. 5th Int. Conf. on Semi-Solid
deformation behavior in the reheating process of eutectic Processing of Alloys and Composites, Golden, CO, June 1998, A.K.
Bhasin, J.J. Moore, K.P. Young, and S. Midson, eds. Colorado School
alloys is necessary. of Mines, Golden, CO, 1998, pp. 489-96.
Currently, research, development, and application for thix- 3. S. Midson, V. Rudnev, and R. Gallik: Proc. 5th Int. Conf. on Semi-
oforming are limited to a few casting alloys such as A356 Solid Processing of Alloys and Composites, Golden, CO, June 1998,
and A357 classified into Al-Si-Mg alloy group. However, A.K. Bhasin, J.J. Moore, K.P. Young, and S. Midson, eds., Colorado
School of Mines, Golden, CO, 1998, pp. 497-504.
it is seen that with the increasingly widespread use of thixo- 4. V.I. Rudnev, R.L. Cook, D.L. Loveless, and M.R. Black: Steel Heat
forming parts, the development of a tailored alloy for a Treatment Handbook, Marcel Dekker Inc., New York, NY, 1997, pp.
specific application has to be chosen. Therefore, in this study, 794-857.
the effect of the induction heating conditions of the ALTHIX 5. V.I. Rudnev and R. L. Cook: Forging Mag., 1995, Winter, pp. 27-30.

2976—VOLUME 30A, NOVEMBER 1999 METALLURGICAL AND MATERIALS TRANSACTIONS A

 6. V.S. Nemkov, V.B. Demidovich, V.I. Rudnev, and O. Fishman: Proc. Heat Treatment, Marcel Dekker Inc., New York, NY, 1997, pp.
XIIth Electroheat Congress ELECTROTECH 92, Montreal, 1992, 775-911.
IEEE and EMC Society, Montreal, 1992, pp. 180-88. 19. Metals Handbook: Properties and Selection: Nonferrous Alloys and
7. V.I. Rudnev and D.L. Loveless: Industrial Heating, 1995, Jan., part Special Purpose Materials, 10th ed., ASM INTERNATIONAL, Mate-
1, pp. 29-34. rials Park, OH, 1990, vol. 2, pp. 164-66.
8. G. Wan, T. Witulski, and G. Hirt: Int. Conf. on Aluminium Alloys: 20. C.G. Kang, H.K. Jung, and Y.J. Jung: Proc. 6th Int. Conf. on Technology
New Process Technologies, Marina di Ravenna, June 1993. of Plasticity (ICTP), Nuremberg, Sept. 1999, M. Geiger, ed., VDI and
9. T. Witulski, U. Morjan, I. Niedick, and G. Hirt: Proc. 5th Int. Conf. WGP, Friedrich-Alexander-University, Nuremberg, Germany, 1999,
on Semi-Solid Processing of Alloys and Composites, Golden, CO, June in press.
1998, A.K. Bhasin, J.J. Moore, K.P. Young, and S. Midson, eds., 21. C.G. Kang, Y.H. Kim, and H.K. Jung: “Thixoforming Process Devel-
Colorado School of Mines, Golden, CO, 1998, pp. 353-60. opment of Aluminum Materials,” Technical Report No. NSDM 98K3-
10. M. Garat, S. Blais, C. Pluchon, and W.R. Loue’: Proc. 5th Int. Conf. 0908-01-01-3, Engineering Research Center for Net Shape and Die
on Semi-Solid Processing of Alloys and Composites, Golden, CO, June Manufacturing, Pusan National University, Pusan, 1999.
1998, A.K. Bhasin, J.J. Moore, K.P. Young, and S. Midson, eds., 22. C.G. Kang, H.K. Jung, and N.S. Kim: Int. J. Mech. Sci., 1999, vol.
Colorado School of Mines, Golden, CO, 1998, pp. xvii-xxxi. 41(12), pp. 1423-45.
11. C. Pluchon, W. Loue’ and M. Garat: Proc. Symp. on Forming Technol- 23. I.J. Kim and D.H. Kim: J. Kor. Foundrymen’s Soc., 1997, vol. 17(4),
ogy of Semi-Solid Metals, Pusan, Nov. 1997, C.G. Kang, ed., KSTP pp. 311-18.
and KIMM, Pusan National University, Pusan, 1997, pp. 80-95. 24. C.G. Kang and J.Y. Kang: “The Process Development on Thixoforming
12. W. Loue’ and M. Garat: Proc. Symp. on Forming Technology of Semi- of Aluminium Materials,” Technical Report, No. NSDM 97K3-0908-
Solid Metals, Pusan, Nov. 1997, C.G. Kang, ed., KSTP and KIMM, 01-01-3, Engineering Research Center for Net Shape and Die Manufac-
Pusan National University, Pusan, 1997, pp. 96-104. turing, Pusan National University, Pusan, Feb. 1998.
13. M. Garat: Proc. Symp. on Forming Technology of Semi-Solid Metals, 25. W. Kurz and D.J. Fisher: Fundamentals of Solidification, Trans Tech
Pusan, Nov. 1997, C.G. Kang, ed., KSTP and KIMM, Pusan National Publications, Aedermannsdorf Switzerland, 1984, pp. 34-43.
University, Pusan, 1997, pp. 105-13. 26. G. Hirt, R. Cremer, A. Winkelmann, T. Witulski, and M. Zillgen: J.
14. C.G. Kang, Y.J. Do, and S.S. Kang: J. Kor. Soc. Technol. Plasticity, Mater. Processing Technol., 1994, vol. 45, pp. 359-64.
1998, vol. 7 (3), pp. 215-24. 27. C.G. Kang, K.D. Jung, and H.K. Jung: Proc. 2nd Int. Conf. on Intelli-
15. H.K. Jung and C.G. Kang: J. Kor. Foundrymen’s Soc., 1998, vol. 18 gent Processing and Manufacturing of Materials (IPMM), Honolulu,
(5), pp. 450-61. Hawaii, July 1999, J.A. Meech, M.M. Veiga, M.H. Smith, and S.R.
16. E.J. Davies: Conduction and Induction Heating, Peter Peregrinus Ltd., LeClair, eds., West Coast Reproduction Centers, Vancouver, B.C.,
London, 1990, pp. 100, 164-79, and 202-22. 1999, pp. 593-99.
17. N.R. Stansel: Induction Heating, McGraw-Hill, New York, NY, 1949, 28. S. Kouji: Proc. 175th JSTP Symp. for Semi-Solid Metals Forming,
p. 178. Tokyo, May 1997, M. Kiuchi, ed., JSTP and JSME, Tokyo University
18. V.I. Rudnev, L.C. Raymond, D.L. Loveless, and M.R. Black: Induction of Technology, Tokyo, 1997, pp. 25-33.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 30A, NOVEMBER 1999—2977