AUSTEMPERED DUCTILE IRON (ADI) ALTERNATIVE MATERIALFOR HIGH-PERFORMANCE APPLICATIONS
G. ArtolaIK4-Azterlan, Durango, Bizkaia, Spain
Tecnun (University of Navarra), Donostia, Gipuzkoa, Spain
I. Gallastegi and J. IzagaIK4-Azterlan, Durango, Bizkaia, Spain
M. BarrenaAdilan Group, Iurreta, Bizkaia, Spain
A. RimmerADI Treatments Ltd., West Bromwich, UK
Copyright � 2016 American Foundry Society
DOI 10.1007/s40962-016-0085-8
Abstract
Austempered ductile iron (ADI) grades are standardized,
and the requirements of current international standards (EN
1564-12/ASTM A897-15) are given terms of conventional
mechanical properties, such as hardness and tensile
strength. Nevertheless, these properties do not show the real
potential of the ADI grades. In order to promote the use of
ADI parts in place of other materials, this work proposes a
comparison between GJS-900-8 and GJS-1200-3 grades,
both in terms of conventional and advanced mechanical
properties, employing stress intensity factors and critical
CTODs (Crack Tip Opening Displacement). This study is
completed with mechanical fatigue testing, so that it can be
shown that the service life of ADI parts is comparable to that
given by other heavier and more expensive options.
Keywords: austempered ductile iron, ADI, fracture
mechanics, fatigue, alternative materials
Introduction
The critical defect size for the transition between plastic
yielding (plain stress) and brittle fracture (plain strain) is
proportional to the square of the toughness to yield strength
ratio (KIC/Rp0.2)2. This ratio tends to be lower for high-
strength materials in comparison with low-strength mate-
rials. High-strength materials can thus show brittle
behavior in the presence of smaller defects, and both
fracture mechanics and fatigue become more relevant.
This fact must be taken into account when alternative,
higher-strength materials, are employed to substitute any
mechanical part by a lightweight solution. In this situation,
conventional mechanical testing by itself is not anymore
adequate for structural integrity calculations. It must be
combined with toughness measurements in order to prop-
erly assess the mechanical design of the part.
Austempered ductile iron (ADI) is an excellent alternative for
this type of substitution, and in this paper, fracture and fatigue
behaviors of two representative grades, GJS-900-8 and GJS-
1200-3 according to EN 1564-2012, are investigated.
Fracture response of ADI is frequently associated with
V-Notch Charpy impact testing,1 as it is considered a
useful brittleness-related ‘‘go/no-go’’ check for quality
control in certain ductile iron grades. Nevertheless, notched
bar impact testing does not properly describe the toughness
of cast irons2,3 and absorbed energy is not useful for design
purposes. Thus, testing methods that are specific to fracture
mechanics are compulsory in this case.
International Journal of Metalcasting/Volume 11, Issue 1, 2017 131
Experimental Procedure
All the materials involved in the experimental procedure
were processed in industrial facilities and under production
conditions. The cast specimens for the study were obtained
from 22 one-inch thickness type Y2 keel-blocks from the
same heat. It was first checked that the as-cast materials
were capable of reaching GJS-900-8 and GJS-1200-3
grades before proceeding to the austempering heat
treatment.
After confirming that the keel blocks were in good condi-
tion to manufacture ADI, they were machined to obtain
two sets of specimen pre-shapes, with a minimum
machining clearance of 2 mm.
Each set contained the following pre-shapes:
• One Ø 20 9 140 mm cylinder for a tensile testing
specimen.
• Two 12.5 9 12.5 9 165 mm prisms for impact
testing specimens.
• Three 15 9 30 9 160 mm prisms for fracture
toughness testing specimens.
• Four Ø 25 9 180 mm cylinders for fatigue testing
specimens.
Austempering of both sets was performed under controlled
atmosphere with a carbon potential of 0.8 during austeni-
tization and inert gas protection during the transfer to the
isothermal bath. The samples were treated together with
production parts in two different batches, which were
selected according to the heat treaters expertise in order to
achieve each targeted grade.
After heat treatment, pre-shapes were finish machined to
standard testing specimens and it was verified that the two
sets corresponded to GJS-900-8 and GJS-1200-3 grades,
respectively.
Once it was confirmed that the materials were as required,
fracture and fatigue specimens were machined. B(E) spec-
imens with a 12.5 9 25 mm section and a straight notch
were employed for toughness testing.
S–N fatigue curves were built by regression to two load
conditions, defined depending on the nominal Rp0.2 indi-
cated in EN 1564-2012 for each grade being tested. The
average stress rm was kept constant and equal to half the
nominal Rp0.2, and two different stress amplitudes of
magnitude 0.5�Rp0.2 and Rp0.2 were applied.
All tests followed applicable European EN standards,
except for fracture testing and uniaxial fatigue that were
performed after BS7448-1:1991 and ASTM E466-15.
Results and Discussion
As-Cast Condition Verification
The material in as-cast condition fulfilled the requirements
for a proper austempering treatment both in terms of
chemical composition and microstructure, as shown in
Tables 1, 2 and Figures 1, 2.
The amounts of Ni, Mo and Cu fit industrial experience
recommendations to improve hardenability. The nodule
count resulted in over 200 nodules per square millimeter.
The as-cast material was also subjected to conventional
mechanical testing as shown in Table 3, in order to reflect
the improvement that is achieved by means of the
austempering process.
Notched and unnotched impact test specimens were used to
emphasize the influence of the notch.
Austempered Condition Verification
Microstructure and mechanical testing confirmed that the
heat treatment of both specimen sets was successful as
GJS-900-8 and GJS-1200-3 were obtained. Mechanical
testing results in Tables 4 and 5 back-up this statement.
The obtained microstructures shown in Figures 3 and 4 are
composed of the expected ausferritic matrix, with less
austenite and sharper ferrite needles explaining the strength
increase for GJS-1200-3 grade.
It is remarkable how impact testing results do not differ
much between the two grades, despite the significant ten-
sile resistance increase from GJS-900-8 to GJS-1200-3.
It is also noticeable that elongation values are comparable
for both grades, since even the high-strength ADI speci-
mens reached an elongation close to 10 %.
Fracture and Fatigue Characterization
The above-mentioned ductility is also evident in the crack
tip opening displacement (CTOD) tests. As shown in Fig-
ure 5, the ratio Fmax/FQ, where Fmax is the maximum test
load and FQ is the point where the curve deviates from
elastic response, is greater than 1.1.
Table 1. Chemical Composition (%)
C Si Mn Mg Ni Mo Cu
3.69 2.28 0.19 0.037 2.45 0.20 0.79
132 International Journal of Metalcasting/Volume 11, Issue 1, 2017
This situation corresponds to elastic–plastic fracture
mechanics, and thus, the studied ADI specimens should not
be described as brittle, from a linear elastic fracture
mechanics point of view. KIC cannot be reported and both
KQ and dc are used instead in Table 6.
The values in the extra column for (KQ/Rp0.2)2 are pro-
portional to the critical defect size that would mean a
change from ductile failure to fracture mechanics driven
failure. The immediate conclusion is that despite GJS-
Table 2. As-Cast Microstructural Features of theMaterial
Nodularity(%)
Nodule count (nod./mm2) Ferrite/pearlite ratio
[90 200 20/80
Figure 1. Unetched as-cast microstructure.
Figure 2. Etched as-cast microstructure.
Table 3. As-Cast Mechanical Properties
Hardness (HB10/3000 W) 273
Rp0.2 (MPa) Rm (MPa) A (%)
Tensile testing 625 852 4.7
KV (J) Unnotched (J)
Impact testing 5 5 5 39 34 37
Table 4. Mechanical Properties of the GJS-900-8 Samples
Hardness
(HB10/3000 W)
314
Rp0.2 (MPa) Rm (MPa) A (%)
Tensile testing 622 963 10.4
KV (J) Unnotched (J)
Impact testing 8 9 9 105 99 106
Table 5. Mechanical Properties of the GJS-1200-3 Samples
Hardness
(HB10/3000 W)
397
Rp0.2 (MPa) Rm (MPa) A (%)
Tensile testing 1035 1260 9.8
KV (J) Unnotched (J)
Impact testing 7 7 8 104 102 94
Figure 3. Microstructure of the GJS-900-8 set.
International Journal of Metalcasting/Volume 11, Issue 1, 2017 133
1200-3 samples offer a yield strength 66 % higher than the
GJS-900-8 samples, the increase in toughness is only 27 %
and the threshold defect size for failure mechanism tran-
sition is reduced 40 %.
The KQ values obtained in the test plan resemble the KIC
results presented by other authors4–6 and are higher than
those employed in certain studies of ADI application as
alternative to steel.7
Nevertheless, none of the checked references describes the
crack tip ductility attained on 12.5 mm thick proportional
specimens.
Regarding fatigue behavior, the comparison between GJS-
900-8 and GJS-1200-3 samples must be done in terms of
load-carrying capacity. GJS-1200-3 offers a higher abso-
lute fatigue limit, the use of this stronger grade instead of
GJS-900-8 only makes sense if it benefits from higher yield
strength. When stresses are normalized by the yield
strength, it turns out that the S–N curve for GJS-900-8 is
above GSJ-1200-3. Figure 6 reflects this fact.
Figure 7 is a representative micrograph of the fracture
surfaces that were observed in all fatigue tests. The fatigue
cracks grew semi-circumferentially with herringbone
marks pointing to the center of the semi-circumference,
what allows finding the crack nucleation point.
In all cases, fatigue crack nucleation spots were found to
coincide with micro-shrinkage porosity sites located on theFigure 4. Microstructure of the GJS-1200-3 set.
0
5000
10000
15000
20000
25000
0 0,1 0,2 0,3 0,4 0,5 0,6
LOA
D (N
)
COD (mm)
GJS-1200-3GJS-900-8
Figure 5. Average toughness test curves for the studiedmaterials.
Table 6. Average Results of the BS-7448-1 Tests
Grade KQ
(MPa�m1/2)dc(mm)
(KQ/Rp0.2)2
(mm)
GJS-900-8 50.3 0.05 6.5
GJS-1200-3 63.7 0.05 3.9
40%
50%
60%
70%
80%
90%
100%
110%
10000 100000 1000000 10000000
Δσ(%
Rp0
.2)
Number of cycles to failure
GJS-1200-3(σm=425MPa)
GJS-900-8(σm=300MPa)
Figure 6. S–N test curves.
Figure 7. Shrinkage porosity which acted as fatiguenucleation point.
134 International Journal of Metalcasting/Volume 11, Issue 1, 2017
skin of the specimens. Figure 8 shows a detail of the crack
initiation point for the specimen of Figure 7.
This means that the inherent fatigue resistance of a defect-
free ADI is higher than in Figure 6, and therefore, it would
last longer in service.
Conclusions
GJS-900-8 and GJS-1200-3 grades develop plastic crack
tip blunting for the studied thickness. The obtained
toughness results agree with the literature data.
The obtained fatigue curves point to service lives that are
higher than the reference values given in the European
Standard.
Avoiding the exclusive use of yield strength-based failure
criteria in the design is a major concern, since the range of
applications where ADI grades become an alternative
material to current solutions is extended when design cal-
culations take into account fracture and fatigue criteria.
Acknowledgments
The authors most sincerely thank to Furesa S. Coop.,member of ADILAN Group, and to ADI TreatmentsLtd. for their invaluable support in this work.
REFERENCES
1. M.F. Hafiz, Impact propeties and fracture in
austempered SG-cast iron. AFS Tran. 2009, 415–422
(2009)
2. K.R.W. Wallin, Equivalent charpy-V impact criteria
for nodular cast iron. Int. J. Metalcast. 8(2), 81–86
(2014)
3. A. Iglesias, I. Gallastegi, G. Artola, M. Muro et al.,
71st World Foundry Congress (2014)
4. S.K. Putatunda, Development of austempered ductile
cast iron (ADI) with simultaneous high yield strength
and fracture toughness by a novel two-step
austempering process. Mater. Sci. Eng. A 315, 70–80
(2001)
5. H.E. Elsayed, M.M. Megahed, A.A. Sadek, K.M.
Aboulela, Fracture toughness characterization of
austempered ductile iron produced using both
conventional and two-step austempering processes.
Mater. Des. 30, 1866–1877 (2009)
6. A. Basso, J. Sikora, Review on production processes
and mechanical properties of dual phase austempered
ductile iron. Int. J. Metalcast. 200, 7–14 (2012)
7. M. Kuna, M. Springman, M. Madler, P. Hubner, G.
Pusch, Fracture mechanics based design of a railway
wheel made of austempered ductile iron. Eng. Fract.
Mech. 72, 241–253 (2005)
Figure 8. Detail of the shrinkage porosity which acted asfatigue nucleation point in Figure 7.
International Journal of Metalcasting/Volume 11, Issue 1, 2017 135