Documente Academic
Documente Profesional
Documente Cultură
16831695
Review
POLITEHNICA University of Bucharest, Materials Science and Engineering Faculty, 313. Spl. Independentei, Bucharest, RO-
060042 Romania.
(Received on November 22, 2012; accepted on May 28, 2013)
The paper reviews original data obtained by the authors, recently disconnected published, concerning
the specific solidification pattern of low S (<0.05%) irons, with very low Al (<0.005%), melted and super-
heated in acid lined, coreless induction furnaces, and how superheating affects the iron quality with effective
metallurgical treatments for use in these conditions. Solidification undercooling increased with increasing
superheat, associated with significant changes in chemical composition, such as C, Si, Mn, Al and Zr,
involved in the nucleation of graphite. The concept in the present paper sustains a three-stage model for
nucleating flake graphite [(Mn,X)S type nuclei]. There are three important groups of elements [deoxidizer/
Mn, S/inoculating] and three technology stages in electric melt iron [superheat/pre-conditioning base iron/
final inoculation]. Different materials were used for pre-treatment of the iron melt, to control oxidization
levels and/or to promote active graphite nucleation sites, including carbon materials and metallurgical sil-
icon carbide. Special attention was paid to maintain Al and Zr recoveries in the melting furnace for their
effects on the iron structure. A double treatment utilizing strong oxide forming elements, such as Al and
Zr for preconditioning, followed by inoculation decreased eutectic undercooling parameters. This treat-
ment improved graphite characteristics and avoided carbides. For foundry application, it is recommended
to ensure (Mn,X)S compound formation, compatible for nucleating graphite with less eutectic undercool-
ing. Attention is drawn to ensuring a control factor (%Mn) (%S) equals 0.03 0.06, accompanied by
0.0050.010% Al and/or Zr content in inoculated grey irons.
KEY WORDS: grey iron; S; Al; Zr; Ti; electric melting; furnace superheating; preconditioning; inoculation;
graphite nucleation; graphite morphology; carbides; thermal analysis.
these irons.2,3)
One of the main differences between a cupola melted iron
and an electric furnace melted iron is the sulphur content,
as melted. It is well established that there is a preferred grey
iron sulphur content at approx. 0.05 wt.% to assist the inoc-
ulation response. It has been demonstrated that the effective-
ness of inoculants in lower sulphur electric melt iron is infe-
rior to the performance in higher sulphur cupola melted
irons of similar carbon equivalent. To ensure that an inocu-
lant still performs in lower sulphur electric melt iron a
foundry has to choose an appropriate inoculant composition,
and then provide an environment in which it can reliably
minimize undercooling.
Grey iron is still the most common foundry metallic Fig. 1. Foundry metallic materials evolution worldwide.
material, at 45% of total world castings production (Fig. 1).4)
It is especially attractive in the manufacture of vital engine
components, using thin walled castings, because of a range
of excellent properties such as machinability, heat conduc-
tivity and vibration damping capacity, in combination with
reasonable strength.
A specific feature of the automotive industry is ongoing
demand for lighter iron castings, so thin wall grey iron cast-
ings (less than 5 mm wall thickness) are more prevalent.
These castings require mandatory high iron pouring temper-
atures, so the melting furnace superheats to more than
1 550C usually. Even if a cupola melted iron could reach a
tap temperature sufficient to allow a pouring temperature of
1 450C the environment in a cupola shaft for the melting
iron or steel is mainly reducing. In contrast the latest core- Fig. 2. Free energy (G1 723 K) formation range of different com-
less furnaces have an oxidizing atmosphere and the seem- pounds of representative elements in the iron melt.
ingly unlimited availability of kWh allows the melt deck to
reach high molten bath temperatures unthinkable in cupola relationship, a similar lattice registry and relatively high sta-
melting. Superheated electric melt irons with low S and very bility of the compounds. The ability of sulphides to nucleate
low Al, typical for these coreless furnaces solidify at high graphite is not obvious, as graphite has a specific hexagonal
cooling rates in thin wall castings and are particularly prone structure, while sulphides are mainly of the cubic types.6)
to free carbides and/or undercooled graphite formation. To become part of a nucleation substrate, foreign particles
The main objective of the present paper is to explore what must satisfy two conditions: (a) conform to certain relation-
conditions are needed to improve the quality of electric melt ships of lattice mismatch or disregistry; a certain lattice face
grey iron superheated in acid lined, medium frequency core- of the foreign particle (inclusion) substrate has a certain cor-
less induction furnaces, for thin wall iron castings. A review responding relationship to the lattice parameter and orienta-
of original data obtained by the authors, relating to the spe- tion of a graphite crystal lattice; (b) satisfy the interfacial
cific solidification pattern of low sulphur irons, examines energy requirement between a heterogeneous nucleus and
how superheating affects the iron quality and which metal- graphite: only when the foreign particles can be wetted by
lurgical treatments are effective. graphite, can graphite form nuclei on the foreign substrates.5)
A characteristic parameter of the lattice relationship is
plane mismatch ( ). The smaller the mismatch of the two
2. Low Sulphur Grey Cast Irons Characteristics
substances , the stronger the nucleation potential between
Only when liquid iron undercools sufficiently (200 them. In general the relationship has the strongest nucleating
230C), can the smaller sizes of micro cluster (C6)n exist as ability when is < 6%, medium nucleating ability when =
stable homogeneous nuclei for graphite particles. Under 6 12% and weak nucleating ability at > 12%.5,7,8) Mis-
normal conditions, such high undercooling is very difficult match (%) between a certain lattice face of some substanc-
to achieve; therefore, the nucleation of graphite is mainly by es and the (0001) face of graphite is shown in Fig. 3, accord-
heterogeneous nucleation.5) ing to paper,5) which reviewed the literature data.913)
Various compounds can be formed in the iron melt, A possible ranking of nuclei of graphite is as follows:
depending upon the melting and pouring practices. Thermo- graphite (highest-least energy required), silicates, oxides,
dynamically, the free energy of formation (G1 723 K) (Fig. 2), sulphides, carbides, nitrides and austenite (lowest).14,15) MnS
of oxides/silicates and sulphides allow more stable particles appears to be a medium potency candidate for graphite
to form in commercial iron melts, than with nitrides and car- nucleation, while other sulphides should be more able to
bides, respectively. Silicate particles are apparently favour- nucleate graphite. The silicates of traditional inoculating
able nucleation sites of graphite in grey irons. This is due to elements (Ca, Sr, Ba) also have a good ability to act as
their hexagonal crystallographic structure, suitable wetting nuclei of graphite. It was also found that the elements of Ca,
Ce, La, Pr, Nd, Sr, can be present in MnS nuclei to form ic misfit with graphite.
complex sulphides (Mn,X)S, which provide a good lattice Three groups of elements were found to be important to
match with graphite and become the most effective nuclei sustain graphite nucleation in industrial grey irons: (1)
among any of the heterogeneous nuclei. For example, it was strong deoxidizing elements (for example Si, Al, Zr), to pro-
found that although the lattice parameter disregistry of MnS/ mote early forming, very small micro-inclusions, which will
graphite may be rather large (12.1%), SrMnS as a modified then act as nucleation sites for later forming complex
particle has a better fit for graphite nucleation (5.2%).13) (Mn,X)S compounds; (2) Mn and S to enable a MnS type
A three stage model for the nucleation of graphite in both sulphide formation; (3) inoculating elements (for example
un-inoculated and inoculated grey irons has been proposed Ca, Sr, Ba, La, Ce), which act in the first stage and/or in the
(Fig. 4):6) (a) Small oxide-based sites (0.13 m, usually second stage of graphite formation, to improve the potential
less than 2.0 m) are formed in the melt; (b) complex of (Mn,X)S compounds to nucleate graphite. The chemistry
(Mn,X)S compounds (110 m, usually less than 5.0 m), maps showed that the core (nucleus) and body (shell) parts
where X = Fe, Si, Al, Zr, Ti, Ca, Sr, P, nucleate at these of the inclusions are distinct in terms of chemical composi-
micro-inclusions; and (c) graphite nucleates on the sides of tion (Fig. 5).16)
the (Mn,X)S compounds because of the low crystallograph- Mn and S were found as the basic components of all
micro-inclusions found in commercial irons, regardless of
their position in the iron structure, iron chemistry, or iron
melt treatment. In all of the above conditions manganese
sulphides were a complex rather than the simple type. In
inoculated irons the (Mn,X)S compounds are more complex,
they appear at a lower Mn/S ratio and with higher compati-
bility for graphite nucleation, especially when one or more
Fig. 5. Typical example of chemical analysis along cross-section of complex (Mn,X)S compound associated with graphite
[La,Ca,AlFeSi inoculated iron] [a-Compo; b-Mn, S;c-O, Al, Si; d-Ca, La].
Fig. 6. Eutectic undercooling degree T1, T2 (a) and chill characteristics (b) of un-inoculated (UI) and Ca - inoculated
irons, at different sulphur levels in the base iron.
er. Any significant analysis correction after melting a steel withstand high temperatures (high purity SiO2 materials,
charge needs power for stirring, especially if recarburizing optimum crystalline structure, active agents addition). One
with cheaper amorphous carbon raisers. Post-furnace metal- application uses a SiC addition in a SiO2 B2O3 mass lead-
lurgical treatment, or numerous transfers and long holding ing to improved lining life (20100%), especially with high
time all require much higher tap temperatures. rust levels in the metallic charge or even when employing
The new generation of coreless induction furnaces are oxygen in a furnace process.27,28)
very capable of achieving high superheat temperatures Since superheating is necessary and has become more
above 1 550C, due to high performance electrical equip- possible in modern foundry melt shops, it is important to
ment with high specific power rating above 250 kW/t. Also investigate the effect on iron quality and possible conse-
quality improvements allow the acid refractory lining to quences in a casting. An extensive research programme,
Fig. 8. The evolution of the liquid iron solution (a) and typical cooling curves and resulted structure of irons (b).
Fig. 9. Influence of the superheating temperature and time on the iron chemistry in acid lined induction furnace melting.
Fig. 10. Influence of superheating temperature on eutectic undercooling Tm (a) and chill tendency (b) [W2 and W3-
wedge samples, ASTM A367].
resenting amorphous carbon materials (9598.7 wt.% fixed had an intermediate behaviour, closer to graphitized petro-
carbon-FC, 0.62.5 wt.%S, 0.51.5 wt.%Ash); b) Graphitized leum coke in these experimental conditions.
Petroleum Coke (GPC) as a crystalline Synthetic Graphite Resistor coke (MCSX), previously used as a furnace insu-
(9999.7 wt.%FC, 0.060.3 wt.%S, 0.10.7 wt.%Ash); c) lator in high temperature furnaces for graphitizing elec-
By-Product Partially Graphitized Packing Coke, as granular trodes was generally found to have an important inoculating
Petroleum Coke (GPCR) (97.497.9 wt.%FC, 0.050.10 influence, giving the lowest chill tendency with less under-
wt.%S, 2.02.4 wt.%Ash) or Metallurgical Coke (GMCR) cooled graphite (type D + E) (Fig. 11). A homogeneous
(8387 wt.%FC, 0.190.34 wt.%S, 11.516.7 wt.%Ash) ini- structure is also a desirable characteristic with the recar-
tially used to surround electrodes in electric graphitizing fur- burised irons using this type of carbon material.
naces; d) Resistor Coke, as a by-product from graphite man- In different experimental conditions, at a higher iron tem-
ufacture, where calcined petroleum coke was covered with perature (1 4801 500C) and longer carburizing time (7
sand to act as an insulator for the furnace wall. It was selected min), with a higher recarburiser addition (0.9 wt.%), both
as a high purity product (MCSX), at more than 95 wt.%FC, graphitized petroleum coke products (granular synthetic
less than 0.2 wt.%S and 23%wt.SiC; e) Natural Crystalline graphite-GPC and packing coke as by-product-GPCR)
Graphite, a high purity mined product (96.9 wt.%FC). appear to be more effective than calcined petroleum coke
The experimental procedure employed consisted of hold- (CPC). Graphitized packing metallurgical coke (GMCR)
ing the melt (3.03.25%CE, 2.52.9 wt.%C, 1.61.85 performance depends on its purity, but could be more effi-
wt.%Si, 0.0160.022 wt.%S) in a 10 kg graphite crucible cient than amorphous calcined petroleum coke (Fig. 12).
induction furnace at constant temperature for about five The results confirmed that graphitized petroleum coke
minutes followed by deslagging. The selected carbon mate- products, including recovered (by-product) materials, gener-
rial was then added (0.6 wt.%) to the melt while power was ally represent higher carbon recovery and inoculation effect
applied to the furnace to maintain the same temperature in grey irons (lower chill tendency). Medium sulphur and
range (1 4501 470C). After a fixed holding time (5 min- high carbon content amorphous calcined petroleum coke
utes), mechanical stirring was applied for 2 minutes. Finally, could be a preferred cheaper carbon raiser for low sulphur
0.70.8 wt.%FeSi75 was added to the furnace before tap- grey iron, considering inoculation effect, due to it also rais-
ping. Chill wedge (W3 , ASTM A367) and forced chill (4C, ing the sulphur level. Recovered granular carbon materials,
ASTM A367) tests were used to evaluate the chill tendency previously used as conducting media are potential carbon
of the recarburised cast irons. raisers with noticeable inoculation effects in grey iron, espe-
The highest carbon recovery level was attained with cially if they were originally petroleum coke.
graphitized petroleum coke products (GPC, MCSX). Natural
graphite (NG) had the lowest range of carbon dissolution 4.1.2. Metallurgical Silicon Carbide
results in molten iron, while calcined petroleum coke (CPC) Silicon carbide is usually produced in special graphite
electric resistance furnaces from high quality raw materials,
including high purity quartzite (>99.5 wt.%SiO2), active
petroleum coke (<0.5 wt.%Ash, <1.5 wt.%S) and other
additions. Two groups of SiC-products are considered: a) -
type at the highest purity (>98 wt.%SiC), surrounding
graphite electric resistance (>1 800C), usually for abra-
sives, heating elements, refractories type applications; b) -
type, at lower purity (>80 wt.%SiC), is the next layer
(1 4001 800C) in a section of the electric furnace, usually
known as Metallurgical SiC.
Generally, granular metallurgical silicon carbide products
include 8098 wt.%SiC, 0.18 wt.%Cfree (graphite), 0.01
1.0 wt.%Sifree, 0.510 wt.%SiO2, 0.020.5 wt.%Al, <0.15
Fig. 11. Chill tendency (TC) and undercooled graphite amount of
wt.%S, <1 000 ppmN, < 200 ppmH, and very low residual
recarburised irons, depending on final carbon equivalent
(CE) and carbon material type [calcined (CPC) and graph-
elements. Metallurgical silicon carbide is a unique C and Si
itized (GPC) petroleum coke; MCSX-resistor coke; NG- bearing material, with 3035 wt.%Ctotal and 5565 wt.%Sitotal,
natural graphite]. for a Si:C = 2:1 ratio.
Fig. 12. Influence of carbon material type on C- recovery [Rc, %], specific sulphur increase [S/C 103, %] and chill
tendency as wedge test (TCW, mm) and forced chill test (TCC, mm) of test irons.
As a carbon source, it is characterized by lower S, N and peratures resulted in loss of carbon equivalent and residual
H contents, compared to non-graphitic recarburisers. As a Si (Al, Zr) levels with a transition from mottled iron to white
source, it contributes less Al, compared to FeSi and has a iron solidification reflected by increased eutectic undercool-
lower level of harmful elements compared to recarburisers ing (see Fig. 10). As Al was identified to be a key element
and FeSi. Supplementary, metallurgical silicon carbide is within complex manganese sulphides acting as graphite
inert towards moisture (no storage problems), resistant to nucleation sites in grey irons, experiments were conducted
oxidizing at temperature (even above 1 500C) and it does to control Al in coreless furnace melting with superheating.24)
not melt (hence its use in refractories), but dissociates in a The iron bath was held in an acid lined coreless furnace
ferrous melt. for up to 40 min at 1 5201 540C. Composition history of
In iron foundries, metallurgical SiC has three important the iron melt was tracked, especially for minor elements Al,
roles: a) Silicon and Carbon additive, with low Al, S, N, H; Zr, Cr, N. In a special section aluminium, as a key element
b) Preconditioning/pre-inoculation effect; c) Efficient deoxi- in graphite nucleation at lower eutectic undercooling, was
dizer lowering FeO/MnO, in slag and metal. It is used as a analyzed accurately, especially after aluminium additions to
furnace addition, in granular form for induction furnaces the molten bath.
(8095 wt.%SiC, 1...10 mm grain size, 2...10 kg/ton addi- Typical changes to the residual aluminium and zirconium
tion rate) and as briquettes for cupola charges (35...65 with time, when holding the induction melted irons in con-
wt.%SiC, 5...20 kg Bricks/ton). tact with air in a silica crucible is shown in Fig. 13(a). Even
As a ladle addition (in-stream or injected), especially in at very low zirconium levels in the iron Zr continued to oxi-
ductile iron production, finer sized products are used, typi- dize to a final content less than 0.00050 wt.%Zr. Aluminium
cally <1.0 mm grain size and >97 wt.%SiC (special prod- content decreased drastically for the observed heats, with a
ucts). Due to the need for SiC to dissociate into molten iron, similar pattern throughout, so the final content was up to
rather than dissolve, the process is temperature sensitive and two times lower than the initial content. A final content of
requires prolonged exposure to the molten iron. For these 0.00040.0005 wt.%Zr and 0.00120.0016%Al in the irons
reasons SiC additions, either at lower temperatures or late is insufficient to assist graphite nucleation with low eutectic
in the process, are not widely practiced. undercooling.
Worldwide foundry experience reports benefits from Having observed that the Al content in the iron melted in
using metallurgical silicon carbide: reduced chill; improved acid lined coreless furnace was less than that needed for
machinability; improved and more uniform mechanical good graphite nucleation (< 0.0030 wt.%Al), an experimen-
properties; less scrap, particularly due to hard edges, less tal programme was established to examine this in more
shrinkage porosity, gas porosity, graphite degeneration; detail.24) In total 11 heats, from the same metallic charge,
higher durability of furnace acid lining; increased Mg-yield; were produced and the Al content was recorded after differ-
reduction in dross formation; higher nodule count. ent Al additions and holding times (Figs. 13(b), 13(c)). This
The preconditioning/inoculation effect of metallurgical showed that the Al content after melting and measured after
silicon carbide is generally promoted by the presence of SiC holding for 1520 min at 1 50010C would be within a
molecules (not from melting but after dissolution) and resid- very narrow range of 0.00110.0012 wt.% from 0.0023
ual crystalline carbon, as graphite. This effect benefits the 0.0030 wt.%Al in the metallic charge.
formation of a fine iron structure, lower chill (carbide form- In some of the heats Al was added after an initial holding
ing) tendency and reduction, or even elimination, of the period to obtain Al-levels from 0.0025 to 0.0060 wt.% fol-
inoculant addition. Some experimental programmes claimed lowed by 20 minutes holding at 1 53010C. The aluminium
a possibility to improve the preconditioning/inoculation per- content in the iron melt was then evaluated at different
formance of metallurgical silicon carbide, by optimizing the times. The higher the aluminium quantity added, the higher
chemical composition, especially by increasing the free car- the final aluminium content, but the level of Al decreased
bon content.3941) two - three times while holding at high temperatures for 20
25 min. Independent of the added aluminium, its content in
4.1.3. Aluminium Recovery by Charge and after Melting the final tapped iron was in the range 0.00150.0025 wt.%
with In-furnace Additions Al. The experiments showed that despite the presence of
It was shown (see Fig. 9) that higher superheating tem- 0.0040.006 wt.% Al in the iron early after melting, and any
Fig. 13. Change in Al and Zr content, with time, when holding induction melted cast irons at 1 530C (a) and change in Al
content after Al addition, in acid lined, coreless induction furnace melted irons (b, c) [M- after melting; Al Add. -
Al addition; 1...11- different heats].
subsequent addition of Al, during holding the Al content in 0.013 wt.%Zr, (%Mn) (%S) = 0.040.06 were electric
the final iron would be too low to sustain graphite nucleation. melted.
Metallic Al was added to the furnace prior to tap as a pre-
4.2. Complex Treatments of Electric Melt Grey Irons conditioner followed by a ladle addition of 1.0 wt.%Ca
4.2.1. Al, Zr and Ti Preconditioning Comparison FeSi inoculant. 15 mm diameter bars for structure analysis
The effect of strong deoxidizing elements such as Al, Zr and 4C chill test samples (ASTM A367) were cast in furan
or Ti on graphite nucleation in base iron and Sr-inoculated resin moulds at 1 3301 350C. Similarly, Zr was tested as
grey irons has been investigated in a laboratory programme a pre-conditioner (0.03 wt.% Zr addition as FeSiZr alloy in
(SEM analysis, thermal analysis, chill tendency, structure the stream during furnace tap), followed by 0.3 wt.% Sr
parameters).6,18,25,4244) It was found that Al, Zr and Ti have FeSi inoculation in the ladle. Experimental test samples,
different distribution patterns in the core (nucleus) and the non-pre-conditioned and Zr-pre-conditioned iron, un-inocu-
shell (body) of (Mn,X)S compounds, which were found to lated and Sr-inoculated, were poured at 1 3801 400C
act as nucleation sites for graphite flakes, with different (furan no-bake bonded mould): wedge test sample (20 mm
influences on graphite nucleation. In contrast with Ti result- width, 50 mm height, 75 mm length) and round test bar (20
ing in complex shaped (morphology) (Mn,X)S compounds, mm diameter).
like Chinese-script, which appear to cause type D graphite Low residual Al level (0.0010.003 wt.%) results in high-
formation, Al and Zr led to smaller sized, more compact er chill, more undercooled graphite and lower eutectic cell
compounds, enabling type A graphite nucleation (Fig. 14).6) count, in inoculated irons. A 0.0070.010 wt.% Al content in
the melt is important to promote type A graphite nucleation,
4.2.2. Double Treatments (Al/Zr Preconditioning + Inocu- with reduced chill, at a higher eutectic cell count (Figs. 15(a),
lation) of Grey Cast Irons 15(b), 15(c)). In another experimental programme, (Figs.
In some of the experiments, Al and Zr were considered 16(a), 16(b), 16(c)),18) Zr-preconditioning also contributed
individually, for their ability to support complex (Mn,X)S to reduced chill (carbides) and to an increased eutectic cell
compounds formation, with high activity in graphite nucle- count. Al has a vital role mainly in the first stage of graphite
ation of grey irons. Effectiveness in this area allows carbide nucleation, forming very small oxide-based inclusions. Zr
control (chill), limits undercooled graphite (D, E type) for- has a double role in this process: early Zr-oxide based
mation and affects eutectic cell characteristics.18) Hypoeu- micro-inclusions help (Mn,X)S compounds nucleation and
tectic grey irons (3.63.8%CE), at 1.651.85 wt.%Si, 0.45 improve the structure of these sulphides to enhance their
0.6 wt.%Mn, 0.090.1 wt.%S, 0.0010.01 wt.%Al, 0.0005 ability to nucleate flake graphite with less undercooling.42)
Fig. 14. Graphite nuclei and graphite particle morphologies in Ti, Al or Zr preconditioned and SrFeSi inoculated grey
cast iron.
Fig. 15. Chill (a), undercooled graphite amount (b) and eutectic cell count (c) of un and - Al-preconditioned grey cast
irons [UI-un-inoculated; Ca-inoculated irons].
Fig. 16. Chill (a), undercooled graphite amount (b), carbides and eutectic cell count (c) of un and - Zr-preconditioned
grey cast irons [UI-un-inoculated; Sr-inoculated irons].
Fig. 17. Representative solidification thermal analysis parameters [Tm, T1, T3] (a) and chill tendency of base and
treated irons, depending on cooling modulus of wedge samples [W1, W2, W3 ASTM A 367] (b) and eutectic
undercooling Tm (c).
Consequently, not only inoculation but also the pre- bides, with undercooling at the end of solidification, T3,
conditioning (Al or Zr) of the base iron could have a posi- becoming less negative (less desirable for contraction
tive effect on the solidification pattern of cast irons. These defects and intercellular carbides) (Fig. 17(a)).24,30)
elements were incorporated in a complex FeSi based alloy, The greatest improvement was with a double treatment
very efficient for pre-conditioning grey irons for thin wall incorporating Al,Zr,CaFeSi alloy pre-conditioning fol-
castings, at low addition rates.45) lowed by Ca,Ba,AlFeSi inoculation. The effect of this
An experimental programme,24) exposed an iron melt combination was more powerful when compared to the
with 3.94.0%CE (3.43.5 wt.%C, 1.51.7 wt.%Si, 0.7 results of single iron treatments of either inoculation, or
0.8 wt.%Mn, 0.0120.015 wt.%S, 0.0020.005 wt.%Al, especially preconditioning (Fig. 17(a)).
0.0005 wt.%Zr) heated to and tapped at 1 530C, to different Thermal analysis of this complex grey iron treatment
pre-conditioning and/or inoculation treatments. The pre- shows that despite higher superheat to 1 530C and relatively
conditioning treatment was applied in the furnace, just undesirable hypoeutectic iron chemistries, with low levels of
before tap. A small 0.10 wt.% addition of a Al,Zr,CaFeSi S, Al and the (%Mn) (%S) < 0.015 factor, typical for elec-
complex alloy was used as a pre-conditioner.45) Similarly 0.2 tric melting, it is still possible to achieve good levels for the
wt.% of a commercial Ca,Ba,AlFeSi inoculant46) was add- control factors: Tm < 25C, T1 > 10C and T3 < 20C
ed to the ladle during furnace tap. The base (reference) iron, (less negative). The development of the thermal analysis
pre-conditioned iron, inoculated iron and double treated parameters in Quik-cupTM tests, from the base (reference)
(pre-conditioned + inoculated) iron were all monitored iron up to double treated iron, agrees with a favourable
using conventional Quik-cupTM thermal analysis. decrease in chill tendency in the experimental irons, depend-
These experimental irons were also poured at 1 350C ing on the different cooling modulii of the wedge samples
into wedges of the type W1, W2 and W3, in resin sand (CM) (Fig. 17(b)), and eutectic undercooling expression
moulds as specified in the ASTM A367 standard test meth- Tm (Fig. 17(c)).24,30)
ods of chill testing cast iron. Standard W-wedge samples Cooling modulus (CM) is defined as the ratio between
were designed to represent a large range of solidification volume and the total external casting surface and is an
conditions of foundry castings. In this experimental pro- expression of the capacity to transfer a given quantity of
gramme, W1, W2 and W3 samples were selected to represent heat through an existing surface to the mould. At higher
thin to medium section castings, known to be critical for cooling modulus, equivalent to a slower cooling rate, the
chill control in electric melted, hypoeutectic grey irons. irons chill tendency is less, with different iron chill values
The elements Al and Zr can participate in the pre-condi- in Fig. 17(b) agreeing with their counterparts in the thermal
tioning treatment by using a Al,Zr,CaFeSi alloy. This treat- analysis chart (Fig. 17(a)).
ment supports the formation of (Mn,X)S compounds as A case study (automotive foundry) illustrates the efficien-
effective nucleation sites for graphite, and thermal analysis cy of Al and Zr bearing-FeSi pre-conditioning of electric
confirms this benefit: a lower Tm with higher T1, which melt grey irons inoculated with SrFeSi or Ca,ZrFeSi.25,45)
becomes slightly positive, as an undercooling expression for In the initial foundry process, the melting procedure in a 6.5
the beginning of the eutectic reaction, favourable for no car- t medium frequency induction furnace used 50 wt.% Steel
Fig. 18. Thermal analysis parameters (a) and tensile strength (b) of double treated grey cast irons in foundry conditions.
Scrap, 50% Cast Iron Scrap, Graphite, FeSi, FeMn, FeS in Table 1. Three Steps in Electric Iron Melt Processing.
the charge. The thermal regime required a 1 500C tap tem- No. Step Characteristics
perature and pouring within a 1 4001 450C temperature
range. A single treatment employed 0.3 wt.% SrFeSi or *Minimum 1 500C superheat/
Induction furnace
Ca,ZrFeSi inoculation during ladle filling. Typical chemi- *Scrap steel and Iron charge
cal analysis: 3.36 wt.%C, 2.13 wt.%Si, 0.64 wt.%Mn, 0.094 1 Melting *S correction for optimum
wt.%S, (%Mn) (%S) = 0.06, 0.042 wt.%P, 0.097 wt.%Cr, (%Mn) (%S) Factor [0.030.06]
- generally, in a 0.41.2 wt.%Mn/
0.13 wt.%Cu, 0.01 wt.%Ni, 0.011 wt.%Mo, 0.005 wt.%Sn. 0.040.12 wt.%S range
Despite a favourable iron composition and inoculation prac-
*Al, Zr, CaFeSi alloy
tice, carbides were a major problem, occurring in sections Precon-
2 *0.1 wt.% addition before tap (usually)
ditioning
of 5 mm or less. *0.0050.010 wt.%Al/Zr in final iron
An addition of 0.1 wt% Al,Zr,CaFeSi alloy45) to the *Different inoculation techniques
furnace before tap, in constant conditions for melting/ 3
Inocu- *Normal/lower addition rates
inoculation/pouring procedure decreased eutectic under- lation *SrFeSi/CaFeSi/Ca,ZrFeSi/
Ca,BaFeSi alloys
cooling, the graphite structure was improved and the tensile
strength increased (Fig. 18).25) Undercooling refers to the
Tm parameter, while recalescence represents the differ- tice, to eliminate un-desired heredity influence of charge, to
ence between the highest (TER) and lowest (TEU) eutectic dissolve all charge materials and allow removal of suspend-
temperatures, and reflects the amounts of austenite and ed inclusions, slag, etc. This step is recommended especially
graphite that precipitate during the first part of eutectic for lower quality charge, including less known metallic or/
freezing. Too high a recalescence might be harmful espe- and non-metallic materials, possible suppliers of coarse
cially in green sand moulds, as the volume expansion is high residual graphite or micro-inclusions, etc. Very important is
and might increase the size of the mould cavity. The prob- also sulphur correction, in order to obtain a beneficial Mn
ability of micro-shrinkage and porosity is high if the expan- and S relationship in complex (Mn,X)S compounds forma-
sion of graphite or supply of feed metal cannot compensate tion, found as the major graphite nucleation sites in com-
for the increased volume. Decreasing both eutectic under- mercial grey cast irons. It is recommended to use (%Mn)
cooling and recalescence are the most important objectives (%S) control factor, according to Table 1 data.
of metallurgical treatments of molten iron. The second step is pre-conditioning the iron with oxide
forming elements (Al and/or Zr), which will have been
reduced to very low levels during induction melting, espe-
5. Recommendations for Foundry Application
cially for low quality charges and/or poor melting practice,
For foundry application, it is recommended to ensure unsuitable inoculation and/or pouring practice. Both Al and
(Mn,X)S compound formation, compatible for nucleating Zr support type A graphite formation with less undercooling
graphite with less eutectic undercooling: smaller/more and fewer free carbides, with zirconium having the highest
compact/higher count particles with a suitable crystallo- effect. In order to minimize the pinholing, especially in green
graphic fit with graphite. In this respect, three technological sand moulds and to obtain an efficient graphitization process,
steps appear to be necessary to produce high performance an Al content of 0.0050.010 wt.% is recommended17,47)
grey iron, especially in electric melting (Table 1). These (Table 1). At this level (Fig. 19)17,47) the surface tension of
affect solidification with less eutectic undercooling and molten iron is high enough48) while hydrogen pick-up from
recalescence; graphite nucleation with less undercooling for green sand mould is limited.49,50)
promoting type A graphite and avoiding type D graphite; The third and final step is inoculation with Ca or Sr or
fewer carbides; higher eutectic cell count. All these factors Ca + Ba FeSi inoculants, to control eutectic undercooling
are critical for thin wall grey iron castings to avoid carbides and recalescence during solidification, responding to the
and/or type D graphite. base iron quality and casting characteristics. According to
The first of these is superheating of the molten iron, the Table 1 recommendations, for a preferred range of Mn
depending on the charge characteristics and melting prac- and S [controlled factor (%Mn) (%S) = 0.030.06] and
ments and approved reference to those results for publica- 26) R. Gundlach: Proc. 112nd AFS Metalcasting Cong., AFS, Schaumburg,
IL, USA, (2008), Paper 08-158.
tion. The authors would like to recognize and thank Michael 27) I. Riposan, M. Chisamera, I. Petrus and S. Babos: Romanian Patent
Barstow (Consultant) for reviewing and editing this paper. RO 110470-B1/1996.
28) I. Riposan, M. Chisamera, M. Turcu, I. Petrus, S. Babos and I. Mattefy:
Rom. Foundry J., (1996), No. 3, 18.
REFERENCES 29) S. Stan, I. Riposan, M. Chisamera and M. Barstow: Materials and
Technologies, Book Series Advanced Materials Research, Vol. 23,
1) A. Andree: Casting Plant Technol., (2011), No. 2, 4. ed. by I. Chicinas and T. Canta, Trans Tech Publications, Stafa-
2) M. Chisamera, S. Stan, I. Riposan, G. Costache and M. Barstow: AFS Zuerich, Switzerland, (2007), 307.
Trans., 116 (2008), 641. 30) I. Riposan, M. Chisamera, S. Stan and M. Barstow: Int. J. Cast Met.
3) I. Riposan, M. Chisamera, S. Stan, G. Costache and M. Barstow: Res., 26 (2013), No. 3, 152.
Proc. 69th World Foundry Con., WFO, Birmingham, UK, (2010), 342. 31) C. R. Loper, Jr., S. L. Liu, S. Shrivani and T. H. Witter: AFS Trans.,
4) 43rd and 44th Census of World Casting Production. Modern Casting, 92 (1984), 323.
(2009), No. 12, 17 and (2010), No.12, 23. 32) A. Jentsch: Proc. of 66th World Foundry Cong. WFO, Birmingham,
5) J. Zhou: China Foundry, 6 (2009), No, 2, 152. (2004), 729.
6) I. Riposan, M. Chisamera, S. Stan, C. Hartung and D. White: Mater. 33) K. Suzuki and N. Kayama: AFS Trans., 90 (1982), 423.
Sci. Technol., 26 (2010), No. 12, 1439. 34) T. Benecke: Giesserei, 68 (1981), No. 12, 344.
7) R. L. Naro and J. F. Wallace: AFS Trans., 77 (1969), 311; AFS Cast 35) T. Benecke, A. T. Ta, G. Kahr, W. D. Schubert and B. Lux: Giesserei,
Met. Res. J., (1970), 131. 74 (1987), No. 10, 301.
8) J. Tartera: J. AFS-ICM, (1982), No. 12, 46. 36) W. D. Schubert, A. T. Ta, G. Kahr, T. Benecke and B. Lux: Proc. of
9) T. Skaland, F. Grong and T. Grong: Metall. Trans., 24A (1993), No. 3rd Int. Symp. on Physical Metallurgy of Cast Iron, The Physical
10, 2321. Metallurgy of Cast Iron, ed. by H. Fredriksson and M. Hillert, Mater.
10) Y. Ueda, M. Takita, K. Ishikawa and T. Yamamoto: Imono, 52 Res. Soc. Proc. 34, Pittsburgh, PA, USA, (1985), 119.
(1980), No. 9, 553. 37) M. Chisamera, I. Riposan, S. Stan and C. Diaconu: Proc. of 67th
11) Z. Zhou and B. Liu: Dissertation, Department of Mech. Eng., Tsinghua World Congress, WFO, Birmingham, UK, (2006), Paper 129.
University, (1987). 38) I. Riposan, M. Chisamera, S. Stan and P. Toboc: Materials and
12) C. Wang: Theory and Practice of Cast Iron Inoculation, Tianjin Technologies, Book Series Advanced Materials Research, Vol. 23,
University Press, Tianjin, China, (1991). ed. by I. Chicinas and T. Canta, Trans Tech publications, Stafa-
13) L. Rong De and Y. J. Xiang: AFS Trans., 99 (1991), 707. Zuerich, Switzerland, (2007), 287.
14) C. R. Loper, Jr. and R. B. Gundlach: Proc. AFS Int. Inoculation 39) I. Riposan, M. Chisamera, S. Stan and V. Constantin: Proc. of
Conf., Schaumburg, IL, (1998), Paper 01. ARABCAST-Arab Foundry Symp., EFS, Alexandria, Egypt, (2004),
15) C. R. Loper, Jr.: AFS Trans., 107 (1999), 523. Paper 01.
16) I. Riposan, M. Chisamera, S. Stan, E. Stefan and C. Hartung: Key 40) I. Riposan, M. Chisamera and I. Petrus: Proc. of 60th World Foundry
Engineering Materials-KEM, Science and Processing of Cast Iron, IX, Cong., CIATF, Zurich, Switzerland, (1993), Paper 40.
ed. by A. Nofal, A. and M. Waly, Vol.457, Trans Tech Publications, 41) I. Riposan, V. Pintea and M. Chisamera: Romanian Patents RO
Stafa-Zuerich, Switzerland, (2011), 19. 105.063-A/1991 and RO 101.922-A/1990.
17) I. Riposan, M. Chisamera, S. Stan and T. Skaland: Proc. of AFS 42) I. Riposan, M. Chisamera, S. Stan, P. Toboc, C. Ecob and G.
Conf. on Cast Iron Inoculation, AFS, Schaumburg, IL, USA, (2005), Grasmo: Proc. of 68th World Foundry Cong., WFO, Birmingham,
31. UK, (2008), Paper 53.
18) M. Chisamera, I. Riposan, S. Stan, D. White and G. Grasmo: Int. J. 43) I. Riposan, M. Chisamera, S. Stan, C. Ecob and D. Wilkinson: J.
Cast Met. Res., 21 (2008), No. 14, 39. Mater. Eng. Perform., 18 (2009), No. 1, 83.
19) A. Sommerfeld, B. Bottger and B. Tonn: J. Mater. Sci. Technol., 24 44) I. Riposan, M. Chisamera, S. Stan, P. Toboc, C. Ecob and D. White:
(2008), No. 3, 321. Mater. Sci. Technol., 24 (2008), No. 5, 578.
20) A. Sommerfeld and B. Tonn: Int. J. Cast Met. Res., 21 (2008), No. 45) Preseed Preconditioner Booklet, ELKEM Foundry Products Division,
14, 23. Oslo, Norway, (2007), available at: www.foundry.elkem.com.
21) A. Sommerfeld and B. Tonn: Int. J. Metalcast., 3 (2009), No. 4, 39. 46) Foundrisil 75 Inoculant, ELKEM Foundry Products Division, Oslo,
22) J. Campbell: Metall. Mater. Trans. B, 40 (2009), No. 6, 786. Norway, (2004), available at: www.foundry.elkem.com.
23) M. Chisamera, I. Riposan, S. Stan and T. Skaland: Proc. 64th World 47) I. Riposan, M. Chisamera, S. Stan, C. Gadarautanu and T. Skaland:
Foundry Cong., WFO, Birmingham, UK, (2000), Paper 62. Proc. of 66th World Foundry Cong., WFO, Birmingham, UK, (2004),
24) I. Riposan, M. Chisamera, S. Stan, G. Grasmo, C. Hartung and D. 775.
White: AFS Trans., 117 (2009), 423. 48) B. Hernandez and J. F. Wallace: AFS Trans., 87 (1979), 335.
25) I. Riposan, M. Chisamera, S. Stan, C. Ecob, G. Grasmo and D. 49) J. W. Dawson and L. W. L. Smith: BCIRA J. Res. Dev., 6 (1956), 226.
Wilkinson: Proc. AFS Int. Iron Melting Conf., AFS, Schaumburg, IL, 50) J. F. Wallace and P. F. Wieser: Proc. of AFS Conf., Schaumburg, IL,
USA, (2009), Paper 2. (1989), paper II, A1-A14.