Materials Science and Engineering A 504 (2009) 8189

Contents lists available at ScienceDirect

Materials Science and Engineering A
j our nal homepage: www. el sevi er . com/ l ocat e/ msea
C, Si and Ni as alloying elements to vary carbon equivalent of austenitic
ductile cast iron: Microstructure and mechanical properties
Nabil Fatahalla
a,
, Aly AbuElEzz
b
, Moenes Semeida
b
a
Mechanical Department, Faculty of Engineering, Al Azhar University, Cairo, Egypt
b
National Institute for Standards, Force and Materials Metrology Department (FMMD), El Haram, Giza, Egypt
a r t i c l e i n f o
Article history:
Received 16 September 2008
Received in revised form 14 October 2008
Accepted 15 October 2008
Keywords:
Austenitic ductile iron
Carbon equivalent
Microstructure
Mechanical properties
Alloying elements
a b s t r a c t
Successful casting of three groups of austenitic ductile irons was achieved covering a carbon equivalent
(CE) range from 3.51% to 5.04%. The three groups implied the change of C, Si or Ni contents to control
the CE%. In case of using Ni element to vary CE%, austenitic ductile iron could be obtained starting from
13.5% up to 34.7% Ni. Generally, the microstructure consisted of graphite nodules embedded in austenitic
matrix. Nodule characteristics were affected by the variation of CE%. Nodularity was almost 100% for all
tested specimens. Slight decrease in hardness and tensile strength (
u
) was observed with increasing the
CE%. 0.2% proof stress (
0.2
) showed almost a constant value with increasing CE%. Tensile elongation was
mainly increased with increasing CE% with different degrees owing to the alloying element (C, Si or Ni).
2008 Elsevier B.V. All rights reserved.
1. Introduction
Austenitic ductile cast irons are series of cast irons that con-
tain nickel from 18 up to 36mass%, having been treated with
magnesium to bring about the formation of nodular graphite
[1]. It contains sufcient nickel to produce an austenitic matrix
structure similar to that of austenitic stainless steel. These irons
have tensile strength ranging from 3870 to 5620MPa, elongation
from 4% up to 40% and Brinell hardness ranging from 1110 to
1710MPa [16]. These high nickel alloyed ductile cast irons are
made in a number of different compositions to produce the desired
properties [13,713]. While conventional foundry practices are
used for the production of Ni-resist ductile iron castings, special
precautions, not normally used, must be taken into consideration.
Treating and gating practices, and pouring temperature must be
modied considerably fromthose used in conventional ductile iron
production. For this reason, design engineers and Ni-resist ductile
cast iron producers should reviewproposed casting designs if min-
imum cost and maximum product reliability are to be obtained
[1]. Numerous data have been published about the production,
microstructure and mechanical properties of austempered duc-
tile cast iron (ADI) [17,1419] and conventional ductile iron
[17,2031]. Few information [13,713] do exist for the produc-

Corresponding author. Tel.: +22 24010200; fax: +2 160854008.

E-mail address: nfatahalla@tedata.net.eg (N. Fatahalla).
tion and properties of austenitic ductile iron in a narrow range of
CE%. To ll this gap, the present investigation focused on studying
the effect of CE% in a wide range, for austenitic ductile cast iron,
on microstructure and mechanical properties. C, Si and Ni were,
each solely, used as alloying elements to vary the CE% in the range
3.515.04.
2. Experimental procedure
Three heats (A, B and C) were prepared in a 90kg high frequency
(1000Hz) induction furnace. Charges were low sulphur, low man-
ganese, and lowphosphor pig iron (Sorel metal) and steel scrap (cf.
Table 1). Necessary amounts of Si, C and Ni were added to yield a Si-
content 1.635.31mass%, C-content 2.13.5mass%, and Ni-content
4.9934.70mass%. Melts were superheated to 17731823K. Mag-
nesium treatment and inoculation were performed using the
Sandwich Technique [1] for producing ductile cast iron. The fer-
rosilicon alloy containing 10% Mg was used in the spheroidising
treatment. The heats were inoculated with 0.5mass% of the charge
with FeSi alloy (65% Si). The grain size of inoculants ranged from1.5
to 3mm. Pure Ni was melted with raw materials to get austenitic
ductile iron in the as cast condition. Table 2 lists the actual chem-
ical composition of all heats involved in this study. The melt was
poured at a temperature ranging from 1620 to 1640K into two dif-
ferent moulds to produce specimens for both chemical analysis and
tests. A half-inch Y-block sand mould was used (cf. Fig. 1). Carbon
0921-5093/$ see front matter 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.msea.2008.10.019
82 N. Fatahalla et al. / Materials Science and Engineering A 504 (2009) 8189
Table 1
Chemical composition of the raw materials used to produce austenitic ductile cast
iron in the present study.
Raw materials Composition%
C Si Mn S P Ni Fe
Sorel metal 4.0 0.1 0.1 0.02 0.03 0.0 Balance
Steel scrap 0.16 0.15 0.6 0.02 0.03 0.0 Balance
Ferrosilicon 0.0 65.0 0.0 0.0 0.0 0.0 Balance
Carboriser 100.0 0.0 0.0 0.0 0.0 0.0 0.0
Nickel 0.0 0.0 0.0 0.0 0.0 99 Balance
Table 2
Chemical composition of all heats of austenitic ductile cast iron produced in the
present study.
Group symbol Heat no. Composition
C Si Ni Mn Mg
A A1 2.11 2.12 19.77 1.40 0.043
A2 2.31 2.07 19.44 1.40 0.041
A3 2.53 2.11 19.41 1.40 0.045
A4 2.71 2.08 19.70 1.40 0.050
A5 2.95 2.12 19.54 1.40 0.045
A6 3.16 2.14 19.41 1.40 0.053
A7 3.29 2.08 19.52 1.40 0.048
A8 3.42 2.16 20.02 1.40 0.059
B B1 2.50 1.63 21.54 1.34 0.047
B2 2.53 2.17 21.59 1.33 0.040
B3 2.52 2.76 21.90 1.32 0.042
B4 2.56 3.32 21.67 1.33 0.049
B5 2.54 3.89 21.86 1.34 0.051
B6 2.51 4.41 21.87 1.34 0.049
B7 2.53 4.92 21.65 1.33 0.038
B8 2.50 5.31 21.58 1.33 0.036
C C1 2.90 1.86 4.99 1.77 0.045
C2 2.85 1.82 9.09 1.72 0.069
C3 2.79 1.84 13.50 1.48 0.061
C4 2.80 1.85 16.10 1.56 0.065
C5 2.83 1.75 19.80 1.71 0.051
C6 2.78 1.79 23.90 1.60 0.063
C7 2.77 1.85 30.40 1.59 0.067
C8 2.91 1.83 34.70 1.39 0.062
Table 3
Effect of CE% on nodule-characteristics of all groups of austenitic ductile cast iron
produced in the present study.
Group
symbol
Heat no. CE% Nodule count
nodule (mm
2
)
Nodule size
(m)
Nodularity
(%)
A A1 3.51 80 15 80
A2 3.69 125 28 100
A3 3.91 125 25 100
A4 4.1 125 25 100
A5 4.34 70 20 100
A6 4.55 220 25 100
A7 4.67 220 25 100
A8 5.04 220 25 100
B B1 3.86 130 28 100
B2 4 160 25 100
B3 4.16 200 25 100
B4 4.28 200 25 100
B5 4.38 225 22 100
B6 4.46 225 22 100
B7 4.59 225 22 100
B8 4.64 250 20 100
C C1 3.7 125 28 100
C2 3.79 200 10 100
C3 3.9 250 15 100
C4 4 180 15 100
C5 4.15 200 15 100
C6 4.26 250 15 100
C7 4.49 250 15 100
C8 4.8 250 15 100
equivalent was calculated according to the following formula [1]:
C.E. = C% +0.33Si% +0.047Ni% (0.0055Ni% Si%)
Standard microstructure examination procedures for cast irons
were used [3]. Vickers hardness test was performed at room tem-
perature of 298K using Otto Wolpert Werk tester. Squared based
diamond indenter (Angle 136

), with 125kg load and 15s dura-

tion was applied. Tensile tests were carried out according to ASTM
(A370-2002). Specimens were machined to 5mm gauge diameter
and 30mmgauge length. Tests were conducted in Instron universal
testing machine connected to computer to draw the stressstrain
curves and recording the tensile strength (
u
), 0.2 proof stress
Fig. 1. Schematic of a half-inch Y-block. Dimensions in mm.
N. Fatahalla et al. / Materials Science and Engineering A 504 (2009) 8189 83
Fig. 2. As-polished microstructure of austenitic ductile cast iron with different %CE ranging from 3.51 to 5.04. Group A: different C% ranging from 2.11 to 3.42.
and elongation. Tensile tests were performed at roomtemperature
(298K) at a strain rate of 610
5
s
1
up to fracture.
3. Results and discussion
3.1. Production of austenitic ductile cast iron having different CE%
Successful trials have been achieved, in the present investiga-
tion, to obtain austenitic ductile cast iron having different CE%
ranging from3.51 to 5.04. Three groups (A, Band C) were produced;
each one of which contained eight heats of different CE%. The main
variable was CE% as controlled by; (i) C in group A, (ii) Si in group
B and (iii) Ni in group C (cf. Table 3).
3.2. Microstructure of austenitic ductile cast iron
Figs. 24 showthe as-polished microstructure of austenitic duc-
tile cast ironfor all groups; A, B, andC, respectively. Generally, these
photos show dark graphite nodules embedded in a single bright
matrix (austenite). Lownodule count (80 nodule/mm
2
) is observed
in Fig. 2a. This result stems from the low C-content (2.11%) of this
specic heat. Thereafter, the nodule count (125 nodule/mm
2
) was
almost constant for CE% ranging from3.69% to 4.1% (cf. Fig. 2(bd)).
Nodule count reached its lowest value (70 nodule/mm
2
) at a CE%
close to the eutectic composition(cf. Fig. 2e). This result may be due
to formation of secondary graphite at this composition. Secondary
graphite is the miniature graphite particles observed around the
84 N. Fatahalla et al. / Materials Science and Engineering A 504 (2009) 8189
Fig. 3. As-polished microstructure of austenitic ductile cast iron with different %CE ranging from 3.86 to 4.64. Group B with different Si% ranging from 2.50 to 5.31.
graphite nodules. Thereafter, an increase in nodule count (220
nodule/mm
2
) has been observed for alloys of CE% ranging from
4.55 to 5.04 (cf. Fig. 2fh). This result may refer to the relatively
high C-content of alloys in this range (2.953.42%C).
Fig. 3 shows the increase in nodule count due to an increase
in CE%, which may refer to the increase of Si-content [15,7].
Moreover, Fig. 4 shows, also, the increase in nodule count with
increasing CE% which, on the other hand, may refer to increasing
Ni-content [15,7]. Table 3summarises the effect of CE%onnodule-
characteristics of all groups of austenitic ductile cast iron produced
in the present study. Nodule count ranged from70 to 220 for group
A, and from 130 to 250 for group B and nally from 125 to 250
nodule/mm
2
for group C. These results are believed to depend on
the variation of CE% and the alloying element in each case [15]. On
the other hand, nodule size was around 25m for groups A and
B and it was around 15mfor group C. Result for nodule size of
group C may refer to the effect of Ni-content and different chem-
ical composition in these heats [15]. Nodularity was almost 100%
for all heats with one exception for A1 (80%). This is believed to
refer to both, the low C-content and low CE% [15]. Table 3 shows
that the nodule count of groups B and C are generally higher
than those of group A. These results stem from the higher Si-
content in B than in A. The increase in Si-content avoids the
formation of carbides and allows increasing the amount of free
N. Fatahalla et al. / Materials Science and Engineering A 504 (2009) 8189 85
Fig. 4. As-polished microstructure of austenitic ductile cast iron with different %CE ranging from 3.70 to 4.80. Group C with different Ni% ranging from 4.99 to 34.70.
carbon [15]. On the other hand, the nodule size in group C is
smaller thanthat inA (cf. Table3); consequently, thenodulecount
for the former is higher than that for the latter.
Photomicrographs for etched specimens of all heats are shown
in Figs. 57. Generally, all photos show graphite nodules embed-
ded in austenitic matrix. Austenitic ductile cast irons must contain
sufcient amount of Ni to produce an austenitic matrix similar
to that of austenitic stainless steel [3]. Literature indicates a CE%
of around 4.3% for standard types of austenitic ductile cast iron
according to ASTM A439. Successful production of austenitic duc-
tile cast iron was achieved in the present study with a wide CE%
range (3.515.04).
Fig. 5a shows a matrix containing iron carbide together with
austenite. Gradual decrease in iron carbide-content and corre-
sponding gradual increase in the soft phase of austenite can be
clearly seen through Fig. 5(bh). The presence of carbides in Fig. 5(a
and b) is believed to stem from the low C-content (2.112.31%C) in
A1 and A2 heats.
Fig. 6, for group B shows wholly austenitic matrix from the
beginning (cf. Fig. 6a) to the end (cf. Fig. 6h). This refers to the
sufcient amount of C and Si-content of this group [15].
In Fig. 7(a and b) pearlite (dark areas) and martensite (bright
areas) can be observed, in small fractions, beside the austenitic
matrix. Appearance of pearlite and martensite is due to insufcient
amount of Ni-content [15]. Literature showed that the sufcient
amount of Ni-content have been necessary to obtain austenitic
matrix and it was reported to have a minimumvalue of 18mass%Ni
[15]. The present study achieved successful production of
86 N. Fatahalla et al. / Materials Science and Engineering A 504 (2009) 8189
Fig. 5. Effect of variation of %CE ranging from 3.51 to 5.04 on the microstructure of austenitic ductile cast iron etched with 0.5 nital. Group A with different C% ranging from
2.11 to 3.42.
austenitic ductile cast iron having wholly austenitic matrix using
only 13.5%Ni. The production cost can, therefore, be reduced using
lesser Ni-content.
3.3. Mechanical properties of austenitic ductile cast iron
3.3.1. Hardness
Fig. 8 shows the inuence of CE% ranging from 3.51 to 5.04 on
hardness Vickers (HV) for the three groups. These curves showthat
HV is slightly decreased or roughly can be considered constant in
the investigated CE% range. It is worthy to mention that heats C1
and C2 are not implied in the austenitic ductile iron category (their
microstructures revealed small amounts of pearlite and marten-
site). The high values of HV for these two specic heats refer to the
existence of pearlite and martensite in their matrices. Although the
present results cover a wider range of CE% compared to literature,
however, it agreed with literature (ASTMA439) in the range of CE%
reported previously.
Table 4 lists the mechanical properties of the austenitic ductile
cast iron reported in the literature. It shows that Brinell hardness
(HB) of standard grades falls within 12102020MPa [1]. However,
the chemical compositions of standard grades were near the eutec-
tic value of 4.3%. The present study covered the CE% range from
3.51% to 5.04%.
3.3.2. Tensile properties of austenitic ductile cast iron
3.3.2.1. Tensile strength(
u
). Fig. 9shows the inuence of CE%rang-
ing from 3.51 to 5.04 on
u
for all heats investigated. The graph
N. Fatahalla et al. / Materials Science and Engineering A 504 (2009) 8189 87
Fig. 6. Effect of variation of %CE ranging from3.86 to 4.64 on the microstructure of austenitic ductile cast iron. Etched with nital 0.5. Group B with different Si% ranging from
2.50 to 5.31.
Table 4
Summary of the properties of austenitic ductile cast iron available in a narrowrange
of CE% ASTM A439 [3].
Grade CE%
0.2
(MPa) u (MPa) Elongation% HB (MPa)
D2 4.44 207 400 8 13902020
D2B 4.44 207 400 7 14802120
D2C 4.37 183 400 20 12101710
D3 4.33 207 379 6 13902020
D3A 4.92 207 379 10 13101930
D4 4.33 NA 414 NA 20202730
D5 4.31 207 379 20 13101850
D5B 4.31 207 379 6 13901930
D5S 3.17 207 449 10 13101930
shows that as CE% increases the
u
decreases slightly for all heats.

u
for heats of group A is higher than its value for heats of groups
B and C. This is believed to stem from the higher C-content in
the former. The higher
u
of heats of group B compared to that
of group C refers to the higher Si-content in the former. Again,
the results of the two heats C1 and C2, although shown on the dia-
gram, however, are not comparable with other points since they
revealed pearlite and martensite in their matrices (not austenitic
ductile iron). Additionally, these two specic heats, C1 and C2, had
low Ni-content (less than 13.5%Ni) (cf. Table 2). The decrease in
u
is related to the amount of iron carbide (hard phase) and austenite
(soft phase) in the matrices. The
u
values obtained in the present
study agree with those reported in the literature in the common
range of CE% [15]. However, the present research covered a wider
range of CE% than that reported in the literature.
88 N. Fatahalla et al. / Materials Science and Engineering A 504 (2009) 8189
Fig. 7. Effect of variation of %CE ranging from 3.70 to 4.80 on the microstructure of austenitic ductile cast iron etched with 0.5 nital. Group C with different Ni% ranging from
4.99 to 34.70.
3.3.3. 0.2% Proof stress (
0.2
)
Fig. 10 shows the variation of
0.2
with the change in CE% of
austenitic ductile cast iron of all groups investigated. The graph
shows slight decrease in
0.2
for groups A and C and slight
increase in
0.2
for group B. It is believed that the increase in
Si-content of group B resulted in corresponding increase in
0.2
.
Fig. 10 also shows that while the trend of
0.2
for groups A and C
is similar
,
but the values of
0.2
of group C are higher than those
for the former. This result may stem from the higher Ni-content of
C group. This higher
0.2
may also refer to the secondary graphite
particles generated (cf. Fig. 4(ah)). It is suggested to clarify this
phenomenon through future research. Emphases of the present
results are given by the literature [15] for the common range of
CE%.
Fig. 8. Variation of hardness HVwith %CE of austenitic ductile cast iron of all groups
(A, B and C). *This graph implies two alloys of pearliticmartensitic DI of group C.
N. Fatahalla et al. / Materials Science and Engineering A 504 (2009) 8189 89
Fig. 9. Variation of ultimate tensile strength with %CE of austenitic ductile cast iron
of all groups (A, B and C). *This graph implies two alloys of pearliticmartensitic DI
of group C.
Fig. 10. Variation of 0.2 proof stress with %CE of austenitic ductile cast iron of all
groups (A, B and C). *This graph implies two alloys of pearliticmartensitic DI of
group C.
Fig. 11. Variation of elongation%with %CE of austenitic ductile cast iron of all groups
(A, B and C). *This graph implies two alloys of pearliticmartensitic DI of group C.
3.3.4. Elongation%
Fig. 11 delineates the effect of CE% on elongation of austenitic
ductile cast iron for all heats produced in the present study. The
elongation of all the heats of groups A and B has the trend of
slight increase withincreasing CE%. Elongationof the heats of group
B is generally higher than that of group A; this maybe due to
higher Si-content which prevents the formation of any carbides
and increases the amount of soft phase (austenite). The increase in
Ni-content in the third group C resulted in a slight decrease in
ductility as can be seen in Fig. 11.
4. Conclusions
The goals, of the present investigation, have been successfully
achieved implying:
(1) Successful production of austenitic ductile cast iron cover-
ing a wide range of carbon equivalent (3.515.04%). This was
achieved using; carbon, silicon or nickel as alloying elements.
The present results are generally in consistence with those
reported in the literature in the common CE% range. However,
thepresent researchlledthegaps that doexist intheliterature.
(2) Successful casting procedure produced austenitic ductile iron
for a heat with only 13.5mass%Ni. Therefore, a promising
cheaper production cost will be available less than that
presently used (more than 18mass%Ni).
(3) The microstructure of the produced austenitic ductile cast iron
consisted of graphite nodules embedded in austenitic matrix.
The nodule characteristics were affected by the change of CE%.
(4) Slight decrease in hardness, tensile strength, and 0.2% proof
stress with increasing CE% was observed. On the other hand,
a slight increase in ductility was observed with increasing the
CE%.
References
[1] American Foundrymens Society, Ductile Iron Handbook, Inc. Des Plaines, Illi-
nois, (1993).
[2] Metals Handbook on Properties and Selection: Iron and Steel, American Society
for Metals, Metals Park, Ohio USA, vol. 1, 2005.
[3] Metals Handbook on Metallography and Microstructure, American Society for
Metals, Metals Park, Ohio USA, vol. 9, 2004.
[4] H.T. Angus, Cast Iron: Physical and Engineering Properties, 2nd ed., British Cast
Iron Research Association, London, 1978.
[5] N. Fatahalla, H. Hakim, A. AboElEzz, M. Mohamed, Zeitschrift fr Metallkunde
89 (1998) 507513.
[6] H. Bayati, R. Elliott, Materials Science and Technology 11 (1995) 284293.
[7] S. Karsay, Ductile Iron: Production Practices, 2nd printing, American Foundry-
mens Society for Metals, 1979.
[8] D.W. Zeng, C.S. Xie, K.C. Yung, Materials Science and Engineering A333 (2002)
223231.
[9] H. Qin, Y. Zhang, Y. Yougshen, Materials Science and Engineering A393 (2005)
310314.
[10] I. Karaman, H. Sehitoglu, H.J. Maier, Y.I. Chumlyakov, Acta Materialia 49 (2001)
471476.
[11] Y.A. Alzan, A.H.I. Mourad, M. AbouZour, O.A. Abuzeid, Engineering Failure
Analysis 14 (2007) 12941300.
[12] Dsseldorf, Austenitic Cast Ironfor HeavyDutyApplications Technical Report,
Press Release, July 23, 2003.
[13] F. Mampaey, AFS Transactions 02024 (2002) 120.
[14] E. Olivera, D. Rajnovic, S. Zec, L. Sidjanin, M.T. Jovanovic, Materials Characteri-
zation 57 (2006) 211217.
[15] U. Seker, H. Hasirci, Journal of Materials Processing Technology 173 (2006)
260268.
[16] S.K. Putatunda, S. Kesani, R. Tackett, G. Lawes, Materials ScienceandEngineering
A435 (2006) 112122.
[17] Y. Sahin, M. Erdogan, V. Kilicli, Materials Science and Engineering A444 (2007)
3138.
[18] B. Stokes, N. Gao, A.S. Reed, Materials ScienceandEngineeringA445446(2007)
374385.
[19] A. Refaey, N. Fatahalla, Journal of Materials Science 38 (2003) 351362.
[20] K. Aslantas, S. Tasgetiren, Y. Yalcin, Engineering Failure Analysis 11 (2004)
935941.
[21] K. Kocatepe, M. Cerah, M. Erdogan, Materials and Design 28 (2005) 172181.
[22] G. Tokas, M. Tayanc, A. Tokas, Materials Characterization 57 (2006) 290299.
[23] A.N. Damir, A. ElKhatib, G. Nassef, International Journal of Fatigue 29 (2006)
499507.
[24] K. Kocatepe, M. Cerah, M. Erdogan, Journal of Materials Processing Technology
28 (2007) 172181.
[25] U. Seker, I. Ciftci, H. Hasirci, Materials and Design 24 (2003) 4751.
[26] A. AbdAlAal, K.M. Ibrahim, Z. AbdElHamid, Wear 260 (2006) 10701075.
[27] N. Fatahalla, H.A. Hakim, A. AboElezz, M. Mohamed, Journal of Materials Science
31 (1996) 49334937.
[28] R.C. Dommarco, M.E. Sousa, J.A. Sikora, Wear 257 (2004) 11851192.
[29] N. Fatahalla, T. Gomaa, S. Bahi, M. Negm, Zeitschrift fr Metallkunde 89 (1998)
554561.
[30] A. Nofal, N. Fatahalla, M. Morad, Foundry International, Technical Report,
September, 1998, pp. 3136.
[31] N. Fatahalla, S. Bahi, O. Hussein, Journal of Materials Science 31 (1996)
57655772.