ISIJ International, Voi. 32 (1 992), No. 5, pp.

563-574
Revie w
State of the Art of Superalloy Production for Aerospace and Other
Application Using VIMIVARor VIM/ESR

Alok CHOUDHURY
Metallurglcal Department, LEYBOLD GmbH,Rueckinger StraBe 12, 6455 Erlensee, Germany.
DURFERRIT
(Received on December9. 1997.• accepted in final form on February 28. 1992)

Vacuuminduction melting is indispensable in the manufacture of Ni- and Co-basedsuperalloys and other
sophisticated alloys because of their reactivity with atmospheric oxygen and nitrogen. The paper describes
the technology of melting and refining in a vacuuminduction furnace, the programmablefurnace control
and metallurgical results. The paper also describes subsequent remelting processes like VARand ESRwhich
makeit possible to meet the very high quality requirements for aerospace applications.
KEYWORDS:
VIM; VIDP; superalloys; vacuumarc remelting; electroslag remelting; ingot defects.

1. Introduction
less As these elements are of pri-
high concentrations.
mary importance for the properties of a material their
Superalloys can be defined as iron-, nickel- or concentration must be kept within a narrow analytical
cobalt-based alloys having high strength and high range in order to assure that the properties are re-
resistance to oxidation properties at temperatures above producible. Accordingly, during melting of these alloys
650'C. The structure of superalloys is basically an every care must be taken to avoid any undesired
austenitic FCCy-phase with a number of secondary reactions, such as with air, increasing loss of these
phases like carbides, MC, M23C6. M6Cand gamma- elements, deterioration in oxide and nitride cleanliness.
prime (y'). In order to achieve a micro-structure allowing The melting of superalloys should therefore be carried
high strength and resistance to oxidation amongothef out under vacuumor in an inertgas atmosphere.
properties at high working temperature the chemical
composition of superalloys is rather complex and
2. Metallurgy of VacuumInduction Melting (VIM)
includes a variety of alloying elements of different
concentrations. Table I shows the main alloying ad- VIM is the
versatile me]ting process for the
most
ditions and their structural effect in different types of production of speciality alloys.2 ~ 6) This melting
process
superalloys. I ) allows the best control over the entire alloy chemistry
The superalloys also contain several other reactive including beneficial and harmful trace elements. Apart
elements like Ti. A1, Nb, Ta, Zr, B and Hf in more or from this, the of exact composition
reproducibility
control from heat to heat is completely assured. Table 2
Table l.
commonranges of main alloying additions and shows, for example, the excellent chemistry control of
their effects in superanoys.
IN-71 8.7,8) Special features of vacuuminduction melting
Rangein wto/o can be summarizedas follows:
Element FeNi- and Eff ects
-
fiexibility due to small batch sizes
Co-base
Ni-base faster change of melt
- programme
Cr 525 l 9-3 O Oxidation and hot corrosion resis-
hardening
tance; carbides; solution Table 2. Chemical composition control of IN-71 8based on
Mo,
Al
W O- 12
0-6
C~I l
0~.5
Carbides; solution hardening
Precipitation hardening; oxidation
lOO heats,

resistance Specific Final Analytical Analytical acurracy

Ti 0-6 ~4 hardening; carbides af- Element frequenc in "lo*
Precipitation range
in
range
in
y
Co 0-20 fects amount of precipitate o/o olo in "/o* 1971 data 1980 data
Ni 0-22 Stabilizes austenite; forms harden-
ing precipitates C 0.02~.08 0.01~0.05 95 ~0.006 ~0.03
0.003
Cb 0-5 0~ Carbides; solution hardening; pre- Ti 0.80-1 . 15 0.90- I IO
,
97 + 0.05 +
cipitation hardening (FeNi- and Ni- Cb 4. 75-5 .50 5.05--5 .40 99 ~ O. i 2 + 0.08
base) Al 0.30~.70 0.5(}-•O 60 9i -~
OV•~ + o02
Ta O- 12 o9 Carbides; solution hardening; oxi-
dation resistance * Analytical frequency e,g. 950/0 mean*, t.hq.t the c.h_ emical com-
position of 95"/o of the melts are within tllc final ['ange.

563 1992 ISIJ

 ISIJ International, Vol. 32 (1992), No. 5
MELTiNG
PRIMARY SECONDARY
MELTING FINAL PROCESSING

e REFINE REMELT
FOR:
CASTPRODUCTS
e PURIFY CONTROL
e MACROSTRUCTURAL
e CHEMCONTROL RALCONTROL
e MICROSTRUCTU e EQUIAXED (+ HIP)
e DS-POLYCRYSTALLINE
e DS-EUTECTICS
ELT/INVESTMENT
R~M CAST e DS-SINGLECRYSTAL
e DUO-STRUCTURE
l l (DIFFUSIONBOND)

l I ATOMIZE
SHAPES
NEAR-NET

I + HIP
e pOWDER
REMELT/ATOMIZE + HIP + FORGE
, pOWDER
+ EXTRUDE
e POWDER -
FORGE
ISOTHERMAL
VAR e LASER-SPRAY

ESR PRODUCTS
WROUGHT
e VIRGINCHARGE
o SCRAP (REVERT)
EXTRUDE
e BILLET. BARSHEET
[EXTRUDE.FORGE.ROLL
-
e SCRAP (PROCESSED) e
VADBR FORGE TOSHApE]
[EF/AOD HYDROMET] ROLL

Fig• 1. Processing routes for products cast from VIM-ingots or electrodes.

isolation of the melt from air contamination: gas Table 3. Typical refractories for VIM-crucibles for melting
- of superalloys.
atmospherecan be selected at will, resulting in low loss
of reactive elements due to oxidation
Max. melt Thermal shock
achievement of very closely reproducible composi- Refractory
temperature
Density
tional tolerances type
('C)
(g/cm 3 ) resistance

control of pressure above the melt: thus control of

- Good
pressure-dependent chemical reactions MgO l 800 2.8

precise temperature control Al203 l 900 3.7 Good

- MgO-spinell 1900 3,8 Poor
excellent homogenization due to inductive stirring
- Al203-spinell l 900 3.7 Relatively good
removal of undesired trace elements with high vapour Zr02 2300 5,4 Poor
-
pressure
dissolved gases like hydrogen and nitrogen
- removal of Table 4. Reduction pressures of various crucible materials
- removal of oxide inclusions
with Fe-melts(0,lo/o C) at 1600 'C.
environment pollution from dust output.
- Iow of
level
Figure I shows the various uses and applications of Crucible material Pco in Pascal
VIM.8) The tap weights range from a few kg to 15t
depending on whether the furnace is designed for pre- CaO 40 O
.

cision casting applications, for casting of forging ingots Zr02 133.3

MgO 533.2
or for the production of electrodes for further remelting. Al203 533.2
Melting in an induction furnace shows somespecific Si02 81313.0
differences as compared with electric arc furnace or
with ladle metallurgical practice. In spite of a relatively
intensive melt circulation the desired slag/metal exchange
induction furnace has a number of important aspects
reactions are limited. Dueto the characteristic movement
relevant to refractory/metal- and refractory/slag-interac-
of the bath the slag is transported towards the crucible
tions. Someof the significant parameters influencing the
wall, where it solidifies and becomesinactive.
performance of the crucible are:
The metallurgical processes such as desulphurization
reaction of metal or slag with the crucible lining
and dephosphorization have, therefore, only a limited -
erosion of the lining material
effect. An additional disadvantage lies in the fact that -
chemical- and physical stability of the refractories.
the rammedcrucible lining undergoes a higher erosion -
due to slag attack. For this reason, the main metallurgi-
Table 3 shows the VIM
crucible refractories com-
cal refining is primarily limited to the purely pressure
monly used today. When selecting the lining refractory
the behavior of the oxides under at elevated
vacuum
dependent reactions, such as C-, O-, N- and H-removal,
temperature must also be considered. The main pre-
and the evaporation of trace elements with high vapour
requisite here is that these oxides remain stable when
pressure as in the case of As, Cu, Pb, Bi, Te, Sband Sn. in contact with the melt under the given operating
2.1. Slag/Metal-Reactions conditions, such as melt composition, pressure and
The refractory lining of the crucible of a vacuum temperature. As the dissociation and reduction of the

C 1992 ISIJ 564

 ISiJ International. Vol. 32 (1992). No. 5
oxides result in the formation of CO, the reduction of these elements on stress rupture properties of certain
pressure of these oxldes is an important indicator for superalloys.14'Is) Most of these harmful trace elements,
their suitability. In Table 4 the reduction pressure of fortunately, have relatively high vapor pressures and can
varlous crucible materials at 1600'C is listed. It is therefore removed by evaporatlon during melting in
apparent that CaOis the most stable crucible lining vacuuminduction furnaces. Figure 3shows, howvarious
material. But CaOis very sensitive to humidity and trace elements behave under vacuum. In NiCr-melt
therefore not suitable for crucible lining in industrial elements such as As, Sn and Sb cannot be removed via
furnaces. It is further evident from the table, that Zr02. the gas phase at a pressure of 7 mbar, whereas the
MgO Al203 are quite suitable for refractory lining
and elements Cu, Pb. Se and Bi are reduced under vacuum
in a
vacuuminduction furnace, with Zr02 the most stable to concentrations well below 50 ppmin a relatively short
of the three.8) Apart from carbon, superalloys generally time. 16)
contain relatively high concentrations of A1 and Ti. The
reduction of the lining material by Al and/or Ti must
thus also be considered. A
recent theoretical investigation 1,o
of reaction kinetics regarded that an alumina crucible
will be most suitable for melting superalloys.9) J (D
o.8

Slag/metal reactions for desulphurization cause CL 0.6

practical difficulties in
vacuum induction furnaces. ~
Desulphurization of the melt with a reactive basic slag ~ 0,4 Pb
in a
vacuuminduction furnace is only carried out in ~5
cQ
0.2
specific cases because of heavy slag attack of the crucible oc Te
Se
lining. A crucible with rammedlining exhibits low Bi
resistance to the refinery slag, which
mayalso contains o 10 20 30
fluxes such as CaF2' Schlatter7) and Darmaralo) achiev-
Impurity Content, ppm
ed good desulphurization results in vacuuminduction (a)
furnaces with steels and high temperature alloys by using
a lime bearing basic slag. Desulphurization primarily
200
occurs via the stable compounds MgS and CaS.
Desulphurizing agents, Iike Ca, Mgand Ce common in Ag
160
Jc:_

the secondary metallurgy of steel, have also been used o

~.-
for desulphurization of superalloys. With the addition Q)
Se
of O.070/0 Ce in well-deoxidized melts of Inconel 901, *= 120 Pb
sulphur-contents of less than 8ppmcan be achieved. In
~~
:D
cc 80
this case the sulphur is removed by the formation of
stable sulphur in CeS2/CaS.11,12) 40
Bi
2.2. Evaporation Reactions
For the production of engine com-
critical aircraft
O IO 20 30 40 50 60 70 80 90
IMPURITrCONTENT,
ppm
ponents the concentration of certain trace elements (b)
such as Ag, Pb, Bi and Se must be kept very low in Fig. 2. Effect ofvarious trace elements onthe 649'C/690 MPa
superalloys,13) Figure
2 shows the deleterious influence stress rupture of Inconel 718.

a) Iron b) 80'/, Ni and 20"/o Cr

1600 'c As
Sn

0,1 0,1
Sb
As
:~s~o ~:~o

zQ o_z
h h
H
0,01 ~
H
o,ol
z
LU zuJ Se
o
z
o zo
o
o Bi o
Cu
o,Ool 0,001
Te
Pb Bi

o 50 1oo 150 o 50 1oo 150

TIME (min) TIME (min)

Fig. 3. Evaporation of trace elements in a VIM-furnace at 7 mbar.

565 =Q 1992 ISIJ

 ISIJ International, Vol. 32 (1992), No. 5
2.3. GasReactions to concentrate near the gas/metal-interface and this limits
The solubility of hydrogen and nitrogen in Ni-base- the transportation rate of nitrogen from the melt into
superalloy melts depends on chemical composition, the gas phase. Figure 4
shows the effect of dissolved
temperature and pressure and is governed by Sievert's oxygen on the rate of nitrogen-removal of Fe-melts.7)
law: For effective nitrogen-removal and to achieve very low
nitrogen content in superalloys like Inconel 718, the
[o/oH]=KJ~!rr2 """""(1) following practice is advantageous:
""""'
selection of charging material with very low initial
[oloN] =K~~:N2
"""-"(2) -
""-"' nitrogen- and sulphur-content
Hydrogenis easily removedduring vacuuminduction -
effective and prolonged boiling of the melt (CO-
melting as this element strictly follows Sievert's law. The formation) also during the period of melt down in
rate of hydrogen removal is considerably higher during order to reduce the dissolved oxygen content by
CO-boiling and stirring of the melt. Hydrogen- C-O-reaction; subsequently, full deoxidation of melt
concentrations of less than I with Al along with Ar-purging at low pressure
ppmare already attained
by the time the charge is completely molten.
Nitrogen solubility in Ni-base-melts principally obeys 2.4. Deoxidation via Gas Phase
Sievert's law, but the presence of various alloying ele- The melt can be deoxidized in the vacuuminduction
ments, such as Cr, V, Ti, A1, Zr, etc., which appreciably furnace through the gas phase according to the following
reduce the activity of nitrogen influences significantly reaction:
its solubility. Becauseof the formation of brittle nitrides
or carbon nitrides in superalloys containing strong [C] + [O] = {CO}gas """"" """""(3)
nitride-forming elements like Ti, the nitrogen content As thereaction product is a gas, the oxygen-content
must be kept to the lowest possible level. The removal ata given carbon-content is directly proportional to
of nitrogen by vacuummelting wastherefore the subject CO-partial-pressure at the top of the melt:
of several investigations. 17~ 19) Simkovich 19) investigated
the influence of composition, pressure and temperature olo[C] x o/o[O]=Kx Pco """"""""'(4)
on nitrogen-removal of Fe- and Ni-base-alloys. The This deoxidation process, the so-called Vacuum-Carbon-
results showed that under similar Deoxidation (VCD)-Process, for Fe- and Ni-melts is of
vacuum treatment
conditions alloys containing up to 15o/o Cr contained distinctiveadvantage for two reasons:
comparable residual nitrogen content as the Cr free the deoxidation product isagas and escapes from
(1)
alloys. This is due to the fact that Cr is not melt without contaminating the melt with inclusions,
a strong
nitride-former and accordingly the rate of nitrogen and
removal from a Cr containing melt under vacuumis (2) the reaction is highly pressure dependent and can
comparable to that from Cr free melts. Theconcentration therefore be controlled by selecting proper pressure
of nitrogen in alloys containing Cr along with strong during vacuuminduction melting.
nitride-forming elements A1, Nband Ti wasmuchhigher. The theoretical equilibrium value of [C] x [O]-
Another important factor for the removal of nitrogen product for Ni-melts is in the range of 10~7 to l0~8 at
under vacuumis the concentration of dissolved oxygen l.33Pa.20,21) For usual carbon content of 0.050/0 in
and sulphur in the melt. Since oxygen and sulphur are various superalloys, the theoretical oxygen content at
surface active elements in Fe- and Ni-melts, they tend equilibrium is in the range of 2ppm. In industrial melts
of Inconel 718 the oxygen content after VCD-treatment
IOO I
at a pressure of .33 Pa is in the range of 10 to 15 ppm.
Thus, in actual industrial melts the C-O-reaction does
SUMMERY
OFLrrERATURE not reach equilibriums owing to the inhibition of
~;~
CO-nucleation. Additionally, a continuous fiow of ox-
~ 75
ygen into the melt takes place due to the dissociation
~ of the refractory lining of the crucible and leakages of
~
z
the vacuumchamberat low pressure. Deoxidation of the
o
:~:~:
melt by the C-O-reaction proceeds in two steps7,8).
50 boiling phase, i.e, formation of CO-gaswithin the melt
-
~ primarily by heterogeneous nucleation with intensive
bath turbulence as a result of the formation of this gas
desorption phase, in which no CO-bubblesform within
~ 25 -
~ the melt, with CO-formation taking place only at the
melt/vacuum-interface.
The greatest reduction of carbon and oxygen takes
place during the first stage, the boiling phase. During
o
0,00 0,05 IO O, 1
O,
5 0,20 0,25 the desorption phase there is only a reduction of
OXYGEN
CONTENTlN wr. % carbon, but not of oxygen, primarily owing to the
Fig. 4. Effectof oxygen on the reaction rate of nitrogen dissociation of the crucible lining. The melt must
removal from liquid iron under vacuum. therefore be deoxidized, finally by a strong deoxidizing

C 1992 ISIJ 566

 ISIJ lnternational, Vol. 32 (1992), No. 5
agent Al and the reaction product A1203 must be
like infiuence of the oxygen content and thus of the oxide
removed during refining of the melt. cleanliness on the rupture life of superalloys.15) An ap-
preciable removal of the oxide-incluslons is achieved
3. Improvements in Oxide Cleanliness if there Is
a llquid reactive slag in contact with the
melt, which is capable of absorbing oxide Inclusions.
Oxide cleanliness plays a decislve role in superalloys
with their extremely high-strength properties at higher Active slags are normally not employed in a vacuum
induction furnace. This meansthat the oxides can only
service temperature. Figure shows the deleterious 5 precipitate at the crucible wall, which naturally, Iimits
the degree of removal of these inclusions.22)
[h] 200 As mentioned, the oxygen-pick up from the dis-
Udimet 500 Cast
- In-100 (PM)
sociation and erosion of the crucible lining is
an
175 - important factor in the final cleanliness of the melt.
Consequently, it is of paramount importance to select
the right ceramic lining. Apart from that, Iong dwelling
150 time of the melt in the vacuum induction furnace
should be avoided.23)
125 UDIMET500 (Cast)
(670'C1172 MPa) 4. Process Technology, Furnace Conception and Auto-
a)
mation
Jo 100
:: The refractory crucible can be rammedor brick lined
'l with suitable ceramic material, preferably with Al2031
oc
s 75 MgO-spinell. After ramming and before sintering the
IN 100 (PM) crucible should be dried at approx. 200'C with an electric

50
(732• 1689 MPa)C heater. The furnace is then charged with material,
preferably low carbon unalloyed steel scrap, and
heated-up by induction to approx. 800'C under air
25 atmosphere. After holding the charge at 800'C for about
2hr, the furnace is closed and evacuated. The charge is
subsequently heated under vacuumto approx. 1700'C
50 150 250 ppm[O] and held for several hours for final sintering. Themelting
OxygenContent in
ppm cycle for superalloys in a VIM-furnace consists of sev-
Fig. 5. Influence of oxygen on rupture life. eral steps of charging, melting down, refining, chem-

Charglng

'~ Temperature
1ooo .1'1'("'~ ~~~~ ~~' . ~,.1"
1soo Installed Power
1430
143 '~;:Tlt'~"~
r'~1r' -. 1' '~'1
1oo Meftlng j ii
1'
"'~'~ . I cb
o"::
9'
power '
" ' J~o
o'~:
120 l II
l ~~ ~~o
r'~' 11
I[
j J~~3le ~~
-oR:
~
(T'
~)
E 10 ;~
!-
o
I Ii ii
i q, ~io
'1:~
1ooo a,
J:: I Ii ii
i
; i
~':t_J
o
~
~
90 o
l
l
lj

li
ij

ii
i ~,, '

s o
,:,
f 81 t
1 l
CL =
~I
o- I' ii
cT~*~~ o)
J,
,Q l I I~l
o
1)
E &~ c
;~ J
~, I l
. i l gl I i!
CQ E:
o
(D
:~
60 r- - _ .
.1

I f~'~ -'- l Ei ' i!

.1 H l ,1
Ii
I' ' x'
cql
'
il
~ 500
l
l
ll

II
ll
TI

~i
I
t
':"
J l TS'l!
'

L1
:rs

Pressure II
,oi
300
30 I
Il

ll
[l

li
'l i jl
j
l II
I r'~1
i iI ii i i
ll ii
l li ,

~~~~~
i'
ll ii !i
l I' E i I .
,
:1
,oo o
Time [h]
o 1 2 3 4
L~IDJneniSi
Ar-purging and/or stirring Alloying
Casting
Temp.-measurement T Fifst Meking down Superheat and Refining and Control
sarn pling
Argon-stirring
S Charge
U:
Fig. 6. Typical melting cycle for Inconel 718 in the VIM-furnace.

567 ~") 1992 ISIJ

 l~lJ International, Vol. 32 (1992), No. 5

(Il~
l l = 7
l __LL~~i~~
~==1l~

~I
f
r ~T
l :1:~~r_,*. :S J l T~~r *-_~.::;:_ I
_J--
-JS
~~~~~,*_~-_~

t s~1- _ J

2-

Fig. 7.
Schematic diagrammeof a VIM-furnace.

istrycorrection and casting. A typical

melting cycle for
Inconel 718 is shownschematically
pressure-tem- in the
perature-time-diagramme of Fig. 6. The basic charging
material is comprised of alloying elements, except the
reactive constituents. Further charging should be car-
ried out under vacuumthrough a bulk charger located
at the top of the furnace. During the melting down- and
refining-period relatively high outgassing of the charging
material and a strong boiling due to C-O-reaction takes
place. It is necessary to adjust carefully the furnace
pressure in order to maintain a well-controlled boil,
which is very important for the removal of dissolved
gases, including oxygen. The boiling intensity decreases
continuously as the oxygen content of the melt is re-
duced. After completion of this refining period the melt
is finally deoxidized and other reactive elements, Al, Fig. 8. VIDP-concept with combined ingot/electrode-casting
Ti, Zr, B, are added where necessary. After thorough and horizonal continuous casting.
mixing and final chemistry check the proper pouring
temperature is adjusted and the melt is poured. VIM-furnaces, the casting of the melt can be carried out
A typical vacuuminduction furnace is shown sche- under vacuumor under inert gas atmosphere. All hy-
matically in Fig. 7. As described earlier, the oxygen con- draulic systems for tilting and the flexible watercooled
tent of superalloy melts at the end of the refining and de- cables
power are outside the vacuumchamber, which
oxidation period is usually in the range of 10 to 15 makesthe equipment safer. During pouring the furnace
ppm
but it has been found that in the casting such as of body with the coil and the crucible is tilted around a
electrodes for further remelting, the oxygen content is vacuum-tight pouring tunnel. The melt is poured into a
higher. This is due to the fact that during casting via a preheated launder, which transfers the liquid metal into
tundish ceramic contamination of the melt from the the casting chamber. The casting chamber can be
eroded tundish material, caused by heavy turbulence designed to incorporate ingot molds, Iadles, continuous
takes place. As the cleanliness of a VAR-ingot depends casting or atomizing equipment. Whereas the VIDP-
on the cleanliness of the electrode, it is of vital impor- furnace itself is comparable to the classical VIM-furnace
tance to use a clean electrode to get a cleaner ingot. in respect to melt treatment under
vacuum, the special
A newdevelopment in this field of vacuuminduction feature of the equipment is the transport of the melt
melting is the VacuumInduction Degassing and Pour- via a relatively long launder. From the metallurgical
ing (VIDP)-furnace. The VIDP-furnace-design has an point of view, the launder is an important part of
independent melting and treatment unit, which allows the VIDP-furnace for the cleaning of the melt from
use of different casting technique modules (Fig. 8). In non-metallic oxide-inclusions. The launder is equipped
comparison with a classical chambertype VIM-furnace with two slag barriers. A ceramic filter just before the
(Fig. 7), the VIDP-furnace does not have outlet can also be inserted, if desired (Fig. 9). The launder
any vacuum
chamber. The furnace body itself is vacuum-tight. is conceived in such
a way, that a quasi laminar fiow
Accordingly, the melt chamber volume and the of the melt is assured. Accordingly, the removal of
chamber-surface of a VIDP-furnace are much smaller oxide-inclusions by flotation during casting is optimized.
than those of a classical VIM-furnace. Similar to Recent results have shownthat all oxides with diametcrs

C 1992 ISIJ 568

 ISIJ International, Vol, 32 (1992), No. 5
VI DP- Furn ace

Tapphg
Alterholder Filter S,8g Barrier ~A
A
Stag Barrler

Ill
~~
llb IIa I

l
Nozzle ~B Ceramlc launder

Ceramlc launder

I Tapplng Zone
ll
Caimlng Zone
Ill
Castlng Zone

cross-sectlon
A-B
Frg. 9. Schem2ltic of a VIDP-launder.

of more than 15~mhave been removed during casting for fine correction at the end of the fining period.
T'ia
a launder.24)
The use of a programmable logical control system 5. Remelting Process
in combination with process computers for the auto-
Growing demands for superalloy cleanliness and
mation of the vacuum induction melting improves structuralhomogeneity in particular, cannot be met by
melt reproduclbility. With such a system, increasingly
stringent metallurgical requlrements of the future can vacuuminduction melting and casting alone. This has
led to the practice of remelting of already melted and
also be met. Closer analysis tolerances can be achieved
refined materlai (known as electrode) in a watercooled
by the acquisition, storage and appropriate processing of
all the data necessary for the metallurgical
copper crucible. Principally three proccsses, vacuumarc
process. remelting, electroslag remelting and electron beam
The basic concept involves combining the individual remelting, combined with a controlled solidification in a
modules, such as the opened and closed loop controls watercooled crystallizer have been developed and in
of the furnace, computational modules and the
commercial use. The newest development in this respect
alloying- and charge-calculation operation, in order to is the application of a plasma torch for remelting
automate the vacuum induction process to varying Figure 10 shows the various process combina-
purposes.
degrees up to 100 o/o, depending on requirements. Such
tions for the production of superalloys.
systems have been in use for some years now as a
successful means for monitoring and controlling the 5.1. VacuumArc Remelting (VAR)
vacuum system or for recording melt parameters, for The vacuumarc remelting process was the first com-
detection purposes and for the diagnosis of faults. mercial remelting process for superalloys. It was used in
Using charge- and alloying calculations, it is quite an the late 1950s to manufacture materials for the aircraft
easy matter nowadaysto achieve the required chemical industry. The primary feature of vacuumarc melting and
analysis at minimal cost. In the ideal configuration, the remelting is the continuous melting of a consumable
process computer is directly linked to the computer of electrode (manufactured in a vacuuminduction furnace)
the analysis system so that immediately after the analysis by meansof a dc arc under vacuum. The molten material
has been made, the additions can be calculated and solidifies a watercooled copper mold.
in
possibly even automatically weighed and charged to the The a VAR-furnace has remained
basic design of
melt. If the alloy calculation system also features a charge largely unchanged over the years; however, significant
calculation facility, the charge materials and quantities advances have been madein the field of control and
(scrap) can also be optimized at the start of the process. regulation of the process with the object of achieving
The complete calculation system facilitates accurate a fully automatic melting procedure. This, in turn, has
computatlon of the alloying elements starting from the had a decisive posltive influence on the metallurgical
initial quantlties and ending with the adjustment values properties of the products. The manufacture of homo-

569 1992 ISIJ

 ISIJ International, Vol. 32 (1 992). No. 5

Plasma Cold Plasma

l VIM
Crucible Melting Cold Crucible

and Melting

Atomization and
L~~L~_*P/M I~~J LY~~~L ,~, [i] EBCHR PCHA Casting

l VAR
L
Ceaniiness + Structure Structure C!eanliness + Structure

Hot Forming
Pressing
or
HIP
working

Structurai Part

Fig. lO. Various process combinations for superalloys.

geneous ingots with minimal segregation requires care-

ful machining of remelting parameters. Of these, the
melting current density and melt rate have the greater
infiuence on the melting bath geometry and conditions
of solidification.
Theprimary benefits of melting a consumableelectrode
under vacuumare:
- Removal of dissolved gases such as hydrogen and
nitrogen
Minimizing the content of undesirable trace elements
with high vapor pressure
- Improvement of oxide cleanliness
Achievement of directional solidification of the ingot
from bottom to top in order to avoid macrosegregation
and to minimize microsegregation.
Oxide inclusion removal is optimized because of the

relatively short reaction paths during melting of the hot

electrode end and because of a good drop dispersion in
the plasma arc. Oxide removal is achieved by chemical
and physical processes. Less stable oxides or nitrides are
thermally dissociated or are reduced by carbon present
in the alloy and are removedinto the gas phase. However,
in superalloys and in high-alloyed steels the non-metallic
inclusions, e.g, alumina and titanium carbonitride, are
very stable. The removal of these inclusions during
remelting takes place only by flotation. The remaining
inclusions in the solidified ingot are small and evenly
distributed in the cross section. The solidification
structure of an ingot of a given composition is a function
of the local solidification time and the temperature
gradient at the liquid/solid interface. To achieve a
directed dentritic primary structure, a relatively high
temperature gradient at the solidification front must be
maintained during the entire remelting period. Accord- Fig. Il. Top longitudinal section of cast -50"/. HCl/50*/.
ingly, the growth direction of the cellular dentrites ferric~hloride grain etch.
corresponds to the direction of the temperature gradient
or the direction of the heat flow at the momentof The pool depth increases with melting rate. Thus the
solidification at the solidification front. The direction of gradient of the dentrites, with respect to the ingot axis,
the heat flow is always perpendicular to the solidification increases with melting rate. In extreme cases, the growth
front or, in the case of curved interface, perpendicular of the directed dentrites can cometo a stop. The ingot
to the tangent. The growth direction of the dentrites is core then solidifies nondirectionally in equiaxed grains,
a function of the metal pool profile during solidification. which leads to segregation and microshrinkage. Even

C 1992 ISIJ 570

 ISIJ International, Vol. 32 (1992), No. 5
in the case of directional dentritic solidification, the fiuctuation.
microsegregation increases wlth dentrite arm spacing. A Freckles and white spots have a muchgreater effect
solidification structure with dentrites almost parallel to on material propertles than tree ring patterns, especially
the Ingot axis (as shownin Fig. Il) yields optimal results. in superalloys. Both defects represent
an important cause
However, this is not always possible. Agood ingot surface of the premature failure of turbine discs in aircraft
requires a minimumenergy input and, accordingly, a engines. Freckles are dark-etching circular or nearly
minimummelting rate. Optimal melt rates and energy circular spots that are generally rich in carbldes or
inputs depend on ingot diameter. This means, that the carbide-forming elements. The formation of freckles is
necessary melting rate for large-dlameter ingots cannot usually a result of a high pool depth and most frequently
be maintained for crystallization parallel to the ingot axis. in a rotating liquid pool. The liquid pool
can be set in
The ingot diameter for several superalloys like Inconel rotation by stray magnetic fields during remelting.
718 or Waspaloy is therefore limited and is less than Freckles can be avoided by maintaining a low pool depth
500mm.Figure 12 showsmelting rates for various steels and by avolding the dlsturbance of magnetic fields
and alloys as a function of ingot diameter. These are through the use of a coaxi_al current supply.
empirical values that were obtained from experience in White spots are typical defects in VAR-ingots. They
operation. Thesemelting rates gave low microsegregation are recognlsable as light-etched spots on a macro-etched
while achieving acceptable surface quality. surface. They are low in alloying elements, e.g. titani-
In spite of directional such
dentritic solidification,
um and niobium in Inconel 718. There are several
defects as tree ring patterns, freckles and white spots can mechanismsthat could account for the formation of
occur in a remelted ingot. This can lead to rejection of white spots25).
the ingot, particularly in the case of superalloys. Tree
- Unmelted
dentrite clusters dropping from the cast
ring patterns can be identified in a macroetched electrode
transverse section as light-etching rings. They usually Particles which falls from the
- crown which forms
represent a negative crystal segregation. Tree ring pat- around the edge of the mold
terns seem to have little effect
on material properties. Particles disintegrating from the solidified edge (shelD
-
They are the result of a wide fluctuation in the remelting of the ingot top.
rate. In modernremelting plants, however, the remelting A11 three of the above mentioned mechanisms,
rate is maintained at the desired value by precise control Individually or combined, can be considered as sources
during operation, so that it exhibits no significant of white spots. This indicates that white spots cannot
completely be avoided during vacuumarc remelting. To
minimize the frequency of occurence of these defects, the
following conditions should be observed during remelt-
12
ing:
the maximumacceptable metal
I~ Io - Use rate permitted
'~E
Fe-base by the ingot macrostructure
,~o)8 - Use a short arc gap to minimize crown formation and
_(D(Ts6 to maximlze arc stability

- Use a homogeneouselectrode free of cavities and

~4 Superalloys (Ni-, Co-base)
cracks.
As mentioned the melt rate is an important
earlier,

2 Inc. 718 factor in of the ingot macrostructure. A

the quality
Waspalo y modernVAR-plant is therefore equipped with a load cell
400 600 800 1OOO system to measure the weight of the electrode of a
Ingot diameter [mm] particular interval of time. The actual values of the melt
Fig. 12. Typical melt rates for different grades. rate are compared by computer with the desired set

Hot Tapping Value

Stlrring Start Value
Meit PhaseValue
Coll Curr
Frequ8ncy Slope 1

Slope 2 Slope 1
Voltage, Slope 3
Slope 3
Pulses
Resist Slope 1 Slope 2
Swing Hold

Slope 2 \
Alternative Sel ection
Melt Rate
Slope 3
Slope 1
Meltlng \ f Slope 3
Current/
Slope 1 !
Pawer Start Fig. 13.
Slope 2 Haid
Control pararneters and setpoint functions.

Time Base Weight Base Time Base

Start-Up Melt Phase Hot Topping

571 (('-')
1992 ISIJ
 IS]J International, Vol. 32 (1992), No. 5
values. Any difference between the measuredmelt rate silica (Si02) are also added, depending on the alloy to
and the desired value is eliminated by the proper be remelted.
accommodationof the power input. Figure 13 showsthe One of the primary advantages of the ESR-process
melt rate and the melting current at start-up, during is the good desulphurization of the metal. The final
steady-state melting and during hottopping. Start-up desulphurization is determined by two reactions. The
and hottopping are usually controlled based on time. first is the metal/slag-reaction, in which sulphur is trans-
The melting phase is controlled based on weight. The ferred from the metal to the slag:
hottopping begins whena preselected residual weight is
reached. A
computer controls the melting parameters,
[S] + (CaO)= (CaS) + [O] ........
..........(5)

which are stored in the form of recipes in the computer. The second reaction is the slaglgas phase-reaction. In
this case, the sulphur absorbed by the slag is removed
5.2. Remelting (ESR)
Electroslag by the oxygen of the gas phase in the form of gaseous
Thetheory behind this process wasknownin the 1930s, sulphur oxide:
but a general breackthrough for this process took more
(CaS) + 3/2 02gas (CaO)+ S02g.~ (6)
than 30 years. Intensive studies carried out in the Soviet = " "" "" "• -

Union. Germany,United Kingdom, Austria and Japan It is evident that

a saturation of the slag with sulphur
after World War 11 made the use of the electroslag does not take place, therefore, the desulphurization ca-
remelting possible on a production scale.26~28) In pacity of the slag remains intact throughout the entire
contrast to the vacuum arc remelting (VAR), the re- remelting process. With a highly basic slag (CaO/Si02
melting in the ESR-process does not occur by striking greater than 3), more than 800/0 of the sulphur can be
an arc under vacuum. The ingot is builtup in electroslag removed.
remelting in a watercooled mold by melting a consumable As mentioned previously, the ESR-process is usually
electrode immersedin a superheated slag usually under carried out under a normal air atmosphere. Oxidation
normal atmosphere. The heat required is generated by of the metal is unavoidable. Oxygencan be transferred
an electrical current (usually ac) flowing through the Into the metal in several ways:
liquid slag, which provides the e]ectrical resistance. As
- Oxidation of the electrode surface above the slag bath
the slag.temperature rises above the liquidus temperature
- Oxidation of the
slag surface of elements with var-
of the metal, the tip of the electrode melts. iable valences such as iron and
manganese
The molten metal droplets fall through the liquid slag Rust attached to the electrode surface
and are collected in the watercooled mold. During the - Transfer of oxygen due to desulphurization according
formation of the liquid film, the metal is refined and to Eq. (5).
cleaned of contaminants, such as oxide-and sulphide- This results in losses of easily oxidizable elements such
particles. The high degree of superheat of the slag and, as titanium, aluminium and silicon. During remelting
partially, of the metal favors the slag/metal-reaction. of Fe-based alloys this oxidation is usually compensat-
Melting in the form of metal droplets greatly increases ed by a continuous deoxidation of the slag, preferably
the metal/slag-interface surface area. The intensive with aluminium. In case of superalloys such as Inconel
reactions between metal and slag result in a significant 718 or Waspalloy with relatively high titanium and
reduction in sulphur and non-metallic inclusions. The aluminium content the deoxidation of CaF2-CaO-
reduction of non-metallic inclusions is better than with Al203-slag does not help muchto assure the desired
the VAR-process. The remaining inclusions are very narrow analytical range of these elements in the remelted
small and are evenly distributed in the remelted ingot. ingot. The introduction of oxygen into the superalloy
Another feature of the ESR-process, as in vacuumarc melt is primarily caused by the ESR-slag constituents
remelting, is the directional solidification of the ingot FeO, Si02 and Al203.
from bottom to top. Due to the very low FeO- and Si02-content of the
The special feature
of the ESR-process is that a slag the reaction between titanium and these oxides is
continuous transport of liquid metal through the slag of minor Importance, so that the actual oxidation of
takes place. Durlng this transport, the slag and the metal titanium takes place according to the equation:
compositions chang according to the kinetic and
3[Ti] + 2(Al203) = 3(Ti02) + 4[Al] ............(7)
thermodynamic conditions. To perform its intended
function, the slag must have somewell-defined proper- By applying the law of mass action, the equilibrium
ties, for example: condition can be expressed in a simplified mannerby the
Its melting point must be lower than that of the metal. equation:
It must be electrically efficient.

-
Its composition should be such that the desired
Table 5. Al- and Ti-contents after electroslag remelting of
reactions like removal of sulphur and oxides are
Inconel 718.
ensured.
-
It must have suitable viscosity at remelting tempera- Electrode
ESR-ingot
ture. Top Bottom
Slags for the
electroslag remelting are usually
composedof calcium fluoride (CaF2), Iime (CaO) and
Ti in ol.
l .05 l .04 1.04
Al in olo 0.57 0.52 0.55
alumina (A1203). Magnesia (MgO), Titania (Ti02) and

O 1992 ISIJ 572

 ISIJ International, Vol. 32 (1 992), No. 5
[Ti] 3 (Ti02)3 'extensively automated in a manner similar to that
[Al]4~_ K(Al203)2
..........(8)
""' accomplished in the case of VAR-furnaces. The remelting
process takes place under fully automatic control. Table
In order to avoid a Ti-loss in the melt, the Ti02-content
6 shows a comparison between the features offered by
should be adjusted for given Al203-content of the slag the ESR-processand those by the VAR-process.22)
to achieve equilibrium for a given [Ti]/[Al]-ratio of the
superalloy (balanced slag system). Table 5 shows the 6. Summary
content of Ti and Al in an ESR-ingot remelted with a
balanced slag but without any deoxidation of the slag The
increasing demandfor very narrow analytical
during remelting. All superalloys with [Ti]/[Al]-ratio range of the chemical composition, very low concentra-
of
tion of harmful trace elements and g~ses along with
3 or less can be remelted with a balanced slag without
any problems. For higher [Ti]/[Al]-ratio in the alloy,
such as A286, a deoxidation, even of the balanced slag,
wil] be necessary during remelting under air atmo-
sphere.29)
The solidification of an ESR-ingot is governed by the 2
same thermodynamical laws as in the case of VAR- 3
process. The melting rate here is also the essential fac-
tor influencing the pool depth and, accordingly, the 4
ingot structure. In order to achieve a primary structure 5
free from any macrosegregation,
a directed solidification
must be assured. The melt rate depending on ingot
diameter is of similar order of magnitude as in the
VAR-process. Here also the ingot diameter of some
segregation-sensitive super-alloys is limited to max.
500 mm.As in a VAR-ingot, structural defects, e.g, tree
rings, patterns and freckles,
can occur in an ESR-ingot.
The reason for occurrence of these defects and means
for their avoidance are the
same as described in the
VAR-process. It is, of course, important to note that
white spots normally do not occur in an ESR-ingot, as 11

there are no crowns above the solidifying ingot and no 12

shelf formation. The relics of unmelted dentrltes from
the consumableelectrode can, of course, fall down, but
they have to pass through the superheated slag and are
~*~~f**~~;~~i~*',~*~"'
~. s**^** ~^~s
melted. Significant advances have been madein the last
few years in plant design, especially in the area of process 1 RANI Drive System 7 Pivoting Drive
control and coaxial power supply. Figure 14 shows 2 Electrode RAM 8 Electrode
3 XY•Adjustment 9 Mold Assembly
schematically a modern ESR-furnace. Figure 15 shows 4 Load Cell System 10 Coaxial Bus Tube
5 Sliding Contact 11 Base Plate
two modernESR-furnaceseach with two melting stations 6 4 Bus Tubes 12 Multi Contacts
for remelting of superalloys. The ESR-processhas been Fig. 14 Schematic of an ESR-furnacewith a stationary mold.

Fig. 15.
Modern ESR-furnace. (Courtesy of Wiggin
Alloys, Hereford/UK)

573
o 1992 ISIJ
 ISIJ International, Vol. 32 (1 992). No. 5
Table 6. Specral features of the Automatic Melt Control (AMC)system
Features VAR ESR
Complete automatic operation of the furnace by preset melt profiles x >

Automatic start-up and hottopping by preset current-time profile X

Automatic start-up and hottopping by preset power-time profile X
Melt rate control maintaining a preset melt rate profile X x
Alternatively selectable current control maintaining a preset profile X
Alternatively selectable power control maintaining a present profile X
Preset profiles for arc gap control (voltage, pulse rate); arc gap is controlled by slave processor board fully X
integrated in AMC
Prdset profiles for immersion depth control (resistance, swing) X
Computer-calculated basic speed of feed based on melt geometry and melt rate X
Preset profiles for stirring coil (current, reversal time) x
Preset profile for operating pressure (optional) x
Monitoring of selectable limits for melt rate and melting power x
Free selection of four melt parameter signals onto four analog outputs X X
Color-plotter software package for four free selectable melt parameter signals X x
Function keys to influence programmeexecution such as hold, run, step-forward, break-off, start X X
hotto pping
Printed melt record containing process data as well as special events; events are printed with elapsed X X
actual time
Video terminal designed for simple dialog for data entry; graphic display available X >

Established melt profiles are stored on hard disk and on floppy disk for backup purposes X X
Integrated diagnostic maintainance programme x X
Host computer interface with own CPU X x

an exceptional cleanliness in sophisticated superalloys 5) W. Zschintzsch: Die neue Giesserei, 39 (1952), 381,
makes the induction melting of these alloys 6) F. Neumann and F. Hegerwaldt: Stahl Eisen, 94 (1974), 53.
vacuum 7) T. Schlatter: J. Met.. 24 (i972), 17.
absolutely necessary. Vacuuminduction melting is a
8) J. W. Pridgeon et a/.: Metall. Treat., (1981), 261.
versatile process and allows independent control of 9) K. Yeumand D. R. Poirier: Proc. 9th Int. Conf. Vacuum
temperature and pressure. A stirring of the melt, either Metallurgy, San Diego, Ca., (1988), 68.
inductive or by Ar-bubbling influences the mass trans- lO) F. N. Darmara: J. Met., 19 (1967), 42.
port to a great extent, which accelerates the desired met- l l) R. E. Bailey et a!.: Proc. 3rd Int. Symp. of Superalloys, Seven
Springs, Pa., (1976), 109.
allurgical reaction favourably. Arelative new devel- l 2) M. F. Rothman: Trans. VacuumMetall. Conf., (1977), 49.
opment of the classical VIM-furnace is the VIDP-
13) R. T. Holt and W. Wallace: Int. Met. Rev. Review203, The Met.
furnace with a long launder. Pouring the melt via laun- Soc., (1976), 1,

der by creating a quasi-laminar flow greatly improves l4) W. B. Kent: J. VacuumSci. Tech., Il (1974), 1038.
the degree of oxide removal. l 5) P. Hupfer: Fachberichte Huttenpraxis Metallverarbeitung,
773.
(1986),
In spite of the large metallurgical potential of VIM-
furnaces the solidification of the ingots or electrodes l 6) P. P. Turillon: Trans. 6th ICVM, (1963), 88.
l 7) G. K. Sigworth Trans. Mel. Soc. CanadianInst. Mel., (1977), 104.
cannot be controlled, which results in unacceptable in- l 8) W. A. Fischer: Arch. Eisenhtittenlves., 33 (1960), l.
homogeneity of the primary structure. Hence a second l 9) A. Simkovich: J. Met., 253 (1966), 504.
step of remelting with controlled directed solidification 20) K. W. Krone: Ph. D. Thesis of the Technical University of Aachen,
Germany,(1969).
ina watercooled copper mold inevitable. Thesecondary
is
21) M. Gard et a/.: Proc. Con. Int. Sur Les Applications Des
remelting of VIM-cast electrodes is carried out either in
Technique Strasbourg, (1967), 177.
a vacuumarc furnace or an ESR-furnace. With proper 22) A. Choudhury: VacuumMetallurgy, ASM-Int.. Metals Park,
controllof the process, especially of the melt rate, Ohio, (1990), 37,
macrosegregation-free ingots can usually be produced 23) W. H. Sutton and E. Manner: Proc. 6th ICVM, San Diego, Ca.,
(1979), 340.
by both means.
24) A. Choudhuryet al.: Paper presented at ATS-Steelmaking Day,
Paris, (1990).
REFERENCES 25) A Mitchell: Proc. Vac. Met. Con. on Special Metals Melting,
Pittsburgh, (1986), 55.
l) M, Donachie, Jr: Superalloys Source Book. American Soc.
J.

Met., Metals Park, Ohio, (1984), 3. 26) W, Holzgruber and E. Pl6ckinger: Stah! Eisen, (1968), 638.
2) C. Le Chevallier Rev. Met., (1980), 791.
et al,:
27) A. Choudhuryet a/.: Stah/ Eisen, (1968), 193. 1
3) H. Burghardt and G, Neuhof: Steelmaking, VEB-Verlag ftir 28) A. Choudhuryet al.: Arch. Eisenhtiltenwes., (1971), 299.
Grundstoffindustrie, Leipzig, (1982), 391. 29) A. Choudhury et al.: Proc, Vac. Met. Con. on Special Metals
Melting, Pittsburgh, (1986), 141.
4) W. Esmarch: Wissenschaftliche Ve,'bffentlichung Siemens, lO
(1931), 173

C 1992 ISIJ 574