Wfc fp108

70� CONGRESO MUNDIAl

DE FUNDICIÓN

Pre-congreso 23 - 24 de Abril 2012, Saltillo, Coahuila

Congreso 25 -27 Abril 2012, Monterrey, Nuevo León

On The

Austempering

Behaviour and

Toughness of

Austempered

Engineering Grade

Ductile Iron Castings S. E. Kisakurek, Professor, Istanbul University, Istanbul, Turkey

A. Ozel, Associate Professor, Sakarya University, Sakarya, Turkey

Y. Yalcin, Associate Professor, Kocetepe University, Afyon, Turkey

A. Turk, Associate Professor, Sakarya University, Sakarya Turkey

H. Akbulut, Professor, Sakarya University, Sakarya, Turkey

S. C. Okumus, Associate Professor, Sakarya University, Sakarya, Turkey ABSTRACT

Austempering behaviour of engineering grade ductile iron castings have been studied by measuring the effects of the

as-cast structure, the chemical composition, nodule count and

the heat treatment parameters, austenitising temperature,

austenitising time, austempering temperature and

austempering time, on the hardness, impact touhness and

microstructure developments in austempering process

Method of factorial experimentation was employed to

identify the relative effects of the heat treatment

parameters on toughness variations, as well as, changes

occurred in nodular characteristics during austempering.

Finally, results of part of the studies on ductile/brittle transition behaviour of austempered GGG40 and GGG80

grade castings were briefly presented.

Keywords: Austempered ducti le iron, austempering, ba inite, hardenabil i ty, austemperabil i ty, austenite,

ba init ic duct i le iron, toughness

INTRODUCTION Over the past three decades intensive efforts have been

expended by the industry and the academe to earn the

Austempered Ductile Iron (ADI) its present status. Most of

the studies of 1980s1-10 and of 90s11-23 were concerned with

understanding the austempering process, paying particular

attention to the effects of as-cast structure, chemical

composition and heat treatment parameters on the structure

and mechanical properties obtainable by austempering

treatment. The concerns of the researchers of the last decade

were mostly ware, surface hardening, corrosion, fatigue

properties, etc., of ADIs.24

Austempering of ductile iron is a two stage process (Figure

1): (i) austenitising the iron in the temperature range 850-950

o C(1562-1742 o F) to change the as-cast matrix to austenite,

(ii) quenching into a salt bath maintained at a temperature in

the range 250-400 o C (392-752 o F) and hold isothermally at

temperature until a satisfactory matrix structure is obtained.

A requirement for the production of ADI with optimum

combination of properties is that, pearlite formation should

be avoided during quenching the iron into austempering

bath after austenitisation, which can be achieved by making

alloying element additions to the base metal compositions.1

Therefore, attentions in the past have been mostly

directed to understand the austempering behaviour of low

alloyed or alloyed ductile irons, in

particular to those containing Ni, Mo and Cu.

This paper is a brief report of some aspects of the

authors’ studies of the austempering behaviour of

engineering grade ductile iron castings produced in

accordance with Standard DIN 1693 specifications, namely

GGG-40, 50, 60, 70 and 80, with special interests in the

effects of the as-cast structure, chemical composition and

heat treatment parameters on the hardness, impact

toughness and microstructural changes occurring during

austempering.

EXPERIMENTAL PROCEDURE

All castings were produced in local commercial foundries,

Irons No.1, 2 and 3 by Ferrodokum A.S., all others by

Doktas A.S., with chemical compositions given in Table 1

in the form of either 25mm Y or 25mm keel blocks. Test

specimens, machined to final dimensions and embedded in

piles of cast

iron chips, austenitised at different temperatures in the

range 800-950 o C(1472-1742 o F) for 10-250 minutes in a

muffle furnace in air atmosphere, then rapidly transferred

into a salt bath mixture of 50/50 KNaNO3, maintained at a

temperature in the range 225-450 o C(437-842 o F), to

austemper isothermally. Austempered samples were

subsequently allowed to cool in air.

Un-notched Charpy impact specimens were prepared

according to ASTM A 327-72 Standard Specification,

Impact and hardness tests were conducted on

Trebel Charpy Impact and Ernst Brinell HB30D2 test

equipments, respectively. Metallographic examinations

were done by an Olympus BHM 313U microscope, x-ray

studies by a Phillips PW3710 diffractometer.

RESULTS AND DISCUSSION

STRUCTURE AND PROPERTIES OF DUCTILE IRONS

IN AS-CAST CONDITON

The matrix structure of the castings were typical of the

respective grades; namely, either ferritic, or ferritic-pearlitic,

or pearlitic-ferritic or pearlitic. Nodule counts were in the

range 106-175mm-2, and average nodule diameters in the

range 25-44 m; nodule shape factor was better than 0.90,

except for iron No.1, which was in the range 0.59-0.80.

EFFECT OF HEAT TREATMENT PARAMETERS ON

THE HARDNESS-AUSTEMPERING TIME AND

IMPACT ENERGY-AUSTEMPERING TIME

RELATIONSHIPS

Ductile irons No. 1(GGG-50), No.2(GGG-60) and

No.3(GGG-60+1.19Ni+0.14Mo) were cast, with as-cast

properties and microstructures as given in Table 2, to

examine their austempering behaviour by hardness and

unnotched impact toughness measurements through the time

frame 30 -120 minutes in austempering time at 250, 300,

350 and 400 o C ( 482, 572, 662, 752 o F) after austenitising

for 30 and 100 minutes at 850 and 900 o C(1562-1652 o F).

The results are given in Figures 2-4 as hardness vs

austempering time and

impact energy vs austempering time plots. As seen in

Figures 2a to 2d, under the heat treatment conditions

specified for this phase of the work, impact energy values

Iron No.1 attained at the beginning of the time frame in

question, which was the 30th minute of austempering time,

were all collected in a narrow energy range, 10-30J. At this

point, on the other hand, in all austempering treatments

employed, the hardness was in the process of fall towards

a minimum or a plateau (Figures 2e to 2h). Although not

presented here, hardness and impact energy variations of

Iron No.2 and of Iron No.3 at this stage of austempering

were in similar transformations.

On a hardness-austempering time plot, point of time where

the hardness falls to minimum is regarded as the time when

the stage 1 reaction of the austempering process, i.e.

transformation of matrix austenite to ferrite and high carbon

austenite(known as the toughening reaction), is completed.6

Therefore, at this point, the retained austenite content of

the structure would be expected to reach at the maximum

level that the treatment condition would allow.6,7 As the

maximum toughness is correlated with the maximum

amount of retained austenite in the matrix,7 the fall of the

hardness to the minimum would also be expected to have

occured in concurrence with the rise of the impact toughness

to the maximum. However, in majority of the hardness and

toughness vs austempering time plots presented in Figure 2

there were time lags in between the events of the fall of the

hardness to the minimum and the reach of the toughness at

the maximum. Suggesting that, retained austenite content of

the structure develops to its maximum level not when the

hardness falls to the minimum but some time later in the

austempering time.

As seen by comparison between the impact energy-

austempering plots in Figures 2a-2d and hardness-

austempering time plots in Figures 2e-2h, under all

austempering treatment conditions Iron No.2 entered the

time frame, 30-120 minute austempering time, with higher

impact toughness than Iron No.1, therefore, with greater

amounts of retained austenite in its structure, as the impact

toughness was related to the retained austenite content of the

structure,7, 23 This also means, Iron 2 arrived at this stage of

the treatment with a greater austempering rate than Iron

No.1. Major differences between Iron No.1 and Iron No.2

were in chemical compositions and nodular characteristics.

As seen in Table 1, carbon and silicon contents of these two

irons were comparable. Mn and Cu contents of Iron No. 2

were nearly double the amounts of the same in Iron No.1;

0.69wt% Mn and 0.65wt% Cu in Iron No. 2 against 0.43wt%

Mn and 0.34wt% Cu in Iron No.1. Available information

on the role that manganese plays in austempering ductile

irons were limited, however, it is well known that it

segregates to cell boundaries during solidification of the

ductile iron where it delays the transformation of austenite,

and also that, increasing its concentration from 0.07wt% to

0.74wt% results in reductions in impact strength.7 Therefore,

higher austempering rate of Iron No.2 compared to that of

Iron No.1 at the 30. minute of austempering time couldn’t

be accounted for simply by the Mn content difference

between these two irons. As pointed out above, as moved

from Iron No.1 to Iron No.2, not only the Mn content but,

simultaneously, the copper content was also increased. Cu is

one of the popular elements added into ductile irons to avoid

pearlite formation during quenching the castings into the

austempering bath after austenitisation,7 therefore, with the

knowledge available, the austempering rate difference

between Iron No.1 and Iron No.2 couldn’t have been related

to the difference between the chemical compositions of

these irons, either.

As to the austempering rates of Iron No.2 and that of Iron

No.3, comparison between the data of Figure 3 and Figure

4 shows that, of the 16 austempering conditions applied in

the experiments, in 15 of them Iron No.3 entered into the

time frame 30-120 minute in austempering time with greater

toughnesses than Iron No.1, suggesting that, again by

deduction similar to that made above when the impact

energies of Iron No.1 and Iron No.2 were related to the rate,

Iron No.3 completed the first 30 minutes of the isothermal

austempering treatment at a greater rate than Iron No.2.

C, Si, Mn and Cu contents and nodular characteristics of

these irons were very similar to each other, as could be seen

in Table 1 and Table 2. The only major difference between

them appeared to be the additional 1.19wt% Ni and

0.14wt% Mo contents of Iron 3, which was deliberately

created to measure the effect of combined addition of these

elements into the composition. Thus, the apparent rate

difference between Iron No.2 and No.3 could be readily

related to the Ni and Mo contents of Iron No.3.

Figures 2-4 indicate that there were differences in the

maximum impact toughness Irons No.1, No.2 and No.3

achieved by austempering; the toughness of Iron No.2 and

No.3 were both remarkably higher than that of Iron No.1,

which could have arisen from the differences in chemical

compositions and nodule counts. There were also differences

between the maximum toughness values Iron No.2 and Iron

No.3 attained during austempering. As seen in Figures 3 and

4, when austempered after the both were austenitised at 850 o

C(1562 o F) for 30 minutes, the maximum toughness

achieved by Iron No.2 was greater than that of Iron No.3.

However, when austempered after austenitised for 100

minutes at 850 o C(1562 o F) and 30 and 100 minutes at 900 o

C(1652 o F), Iron No.2 was ahead. Indicating that,

1.19wt% Ni and 0.14wt% Mo addition to ductile iron which

also contained some amount of Mn and Cu can benefit from

austempering tretment by significant amount, provided that

appropriate conditions were selected.

The magnitude of the relative effects of heat treatment

parameters on the toughnes variations recorded by

austempering heat treatment could be estimated by factorial

analysis of the experimental results.25 Table 3 shows the

factorial design of experiments and the results of analysis

of the effects of austenitising temperature(factor A),

austenitising time(factor B), austempering temperature(factor

C) and austempering time(factor D), by varying each factor at

two levels, as 800-900 o C(1562-1652 o F), 30-100 minutes,

250-350 o C(482-662 o F) and 30-120 minutes, respectively,

on the maximum impact toughness of Iron No.2.

Results and effects are given in the fifth and sixth columns

of Table 3 show that the most effective factors in

increasing the impact energy of austempered Iron No.2

were austempering temperature and austempering time.

Increasing the austenitising temperature and/or the

austenitising time decreased the toughness. In decreasing

order of effectiveness, austenitising temperature was the

third factor in the queue, increasing the level of this factor

resulted in significant reduction in toughness. The least

effective factor was austenitising time, increase of which

also caused reduction in toughness, but to a lesser extent.

Therefore, the optimum condition to develop an ADI with

maximum toughness was to austemper Iron No.2 at 350 o C

for 120 minutes after austenitising at 850 o C for 30 minutes.

EFFECT OF COPPER CONTENT

The effect of Cu content on the development of retained

austenite proportion of the austempered structure was

studied on Irons No.4, No.5 and No.6, cast to meet the

Standard Specifications for Grade-50, Grade-60 and Grade-

60 + 1.27 Cu ductile irons, respectively. As the castings

were poured from a single melt by simply modifying the

copper content of the melt between the taps, carbon, silicon

and manganese contents could be kept constant in all

members of the group. Chemical composition of each

casting was given in Table 1, hardness and structural

characteristics in the as-cast state, as

measured by auto-image analyser, are outlined in Table 4.

All three irons were given equal austempering treatments;

austenitised for 60 minutes at 850 o C(1472 o 2F) and 900 o

C(1562 o F), and then austempered at 350 o C(662 o F) for

10 to 100 minutes. Results are presented in Table 5.

Data shows, when irons were austenitised at 850 o C(1562 o

F), largest fraction of retained austenite in the structure was

obtained within the first 25-50 minutes of the austempering

process, however, when austenitised at 900 o C(1652 o F)

prior to austempering, retained austenite proportion of the

structure rised to its highest level at a later point in

austempering time, in a time frame starting at,

approximately, the 50. minute of the treatment. The increase

with austenitising temperature in the amount of retained

austenite was, on the average, 26vol%. Other than that, there

was no evidence to suggest that retained austenite content

increased with copper content.

EFFECT OF NODULE COUNT AND NODULE SIZE

Irons No.7 and No. 8, cast to meet the Standard DIN 1693

Specificaitons for GGG-50 grade ductile iron, were planned

to investigate the effects of nodular characterisitics, e.g.

nodule count and nodule size, on austempering behaviour.

They were both poured from a single melt: Iron No.8,

however, in addition to the

standard laddle treatment applied to both, was also given in-

mold treatment to modify its nodular characteristics. As seen

in Table 1, only difference in the chemical compositions of

these two irons were in their silicon contents; there was

2.33wt% Si in Iron No.7 and 3.25 wt% Si in Iron No.8,

which stemmed purely from the additional in-mold

treatment applied to Iron 8. Table 6 lists the hardness,

impact toughness and microstructural characteristics, of

the castings, Figure 5 compares the microstructures of both

irons in unetched and etched conditions

Figures 6a and 6b show the variation of impact energies

against time in austempering treatments at 250, 300, 350

and 400 o C(482, 572, 662, 752 o F) of Iron 7 and Iron 8,

respectively, after austenitised at 850 o C(1562 o F) for 60

minutes. A major difference between the austempering

behaviours of these irons was the change of the optimum

temperature to obtain the maximum toughness: In Iron No.7

maximum toughness was achieved at 325 o C(617 o F), in

Iron No.8 at 350 o C(572 o F). Maximum toughness achieved

by Iron No.8(175J) was remarkably higher than that

obtained by Iron 7(125J). There are evidences that, increasing

the silicon content increased the optimum temperature for

maximum ductility, increased fracture toughness by

increasing the retained austenite content of the matrix and

increased the impact strength of ADI.7 Therefore, the

increase in impact toughness, as moved from Iron No.7 to

Iron No.8, was, at least partly, due to the higher silicon

content of the latter. No published data

were available suggesting direct relationship between the

nodule count and the impact toughness, or ductility, of ADIs,

although works on ductile iron castings has shown that

increasing the nodule count effectively improved the tensile

elongation of ductile irons, but room temperature impact

toughness was independent of the nodule count for all matrix

structures .5 Therefore, the nodule count-impact toughness

relationship requires further research to identify the

individual effect of the nodule count.

CHANGES OCCURRED IN NODULAR

CHARACTERISTICS DURING AUSTEMPERING

Significant changes were noted in nodule counts and

average nodule diameters of ductile irons during

austempering heat treatment process. Measurements on the

broken impact test specimens of Irons No. 1, 2, and 3,

suggested that, up to 27% reductions in nodule count and

up to 45 % increases in nodule size were possible. Method

of factorial experimentation was also applied to measure the

relative magnitudes of the effects of the heat treatment

parameters, austenitising temperature(factor A), austenitising

time(factor B), austempering temperature(factor C) and

austempering time(factor D), upon the % variation occurred

in nodule count and in nodule size during austempering of

Iron No.2 with respect to the as-cast state. Results are

presented in Tables 7a and 7b, indicated that, out of the

four heat treatment factors studied, the most effective one in

influencing the variation in the number and size of the

graphite nodules during austempering was the austenitising

temperature. Effectiveness of the factors decreased in the

order: austenitising temperature, austenitising time,

austempering temperature, austempering time.

Effect of the as-cast matrix structure on the amount of

changes occurred in nodule count and nodule size, with

respect to the as-cast state, during austempering were also

searched. Irons No.9 and No.10 were prepared to meet the

Standard DIN 1693 specifications for GGG-40 (fully ferritic)

and GGG-80 (fully pearlitic) ductile irons. One surface of

each sample, of dimensions 10mmx10mmx110mm, was

marked,

so that nodule count and nodule size measurements before

and after the treatments could be done on the same surface

of the same specimen. Irons were austenitised at 850 o C

(1562 o C) and then austempered at 350 o C (572 o F) for

different times. Results are presented in Figures 7a and 7b,

as estimated number of graphite nodules per unit volume of

casting vs austempering time, and

% increase in average diameter of volume distribution of

graphite nodules vs austempering time plots. Increase in the

% reduction in nodule count and % increase in average

nodule size, as moved from pearlitic to ferritic iron was

noticable. Changes occurred in nodule counts and nodule size

can well be accounted for, at least qualitatively, by Oswald

Ripening mechanis,26 which is an observed phenomenon in

solid solutions, which describes the change of an

inhomogeneous structure over time. In other words, over

time, small particles, graphite nodules in this case, dissolve,

and redeposit onto larger particles.

DUCTILE/BRITTLE TRANSITON BEHAVIOUR OF

AUSTEMPERED DUCTILE IRONS

In this phase of the studies ductile/brittle transition

behaviour of the whole series of engineering grade ductile

iron castings, were examined. Specimens, following

austenitisation for 100 minutes at 900 o C (1652 o F), were

isothermally austempered at 250, 300, 350 and 400 o C(482,

572, 662, and 752 o F) for 7 to 210 minutes. For below-zero

impact tests, specimens were refrigerated in ethyl alcohol in

a Lab-Plant Refrigerated Immersion Probe, RP-100. For

above-zero tests, specimens were heated in distilled water.

Figures 8a and 8b, showing the variation of impact energies

of Iron No. 11 (GGG40-fully ferritic) and Iron

No.12(GGG80-fully pearlitic) with test temperature in as-

cast and austempered conditions, demonstrate that, both

irons were susceptible to temperature changes of the

environment. As seen in Figure 8a, fully ferritic ductile

iron(Iron No.11) gained no increments in toughness by

austempering. Even after austempered at the optimum

conditions to obtain maximum toughness, its room

temperature impact energywas equal to that of the as-cast

state. With the fall of the test temperature to -40 o C(-40 o F),

the toughness of austempered Iron No.11 fell to 25J, which

was a very low toughness level for a fully ferritic ductile

iron. On the other hand, austempering increased the room

temperature toughness of the pearlitic iron(Iron No.12),

from 5J in as-cast state to 80J. As the temperature of the

media was lowered to -40 o C(-40 o F), toughness of this

pearlitic ductile fell to 40J. A low value, but still higher than

the room temperature toughness of its as-cast state. The

subject will be discussed elsewhere.

CONCLUDING REMARKS

During isothermal austempering, maximum toughness at

temperature is achieved at a later stage than the fall of the

hardness to minimum, suggesting that ADI owes its high

strength and high ductility to its carbide dispersed metal

matrix (ferrite plus high carbon austenite

ADI can achieve high toughness in presence of manganese,

even in amounts up to 0.65wt%, provided that it was added

in combination with copper.

Austempering treatment increases the nodule shape factor to

> 0.90, at the same time nodule count decreases and nodule

size increases.

Nodule count increases the maximum toughness obtainable

from a given austempering treatment.

ADI is susceptible to ductile/brittle transition.

REFERENCES

1. Dorazil, E., Barta, B., Munsterova,E., Stransky,L,

Huvar,A., AFS. Int. Cast Metals Research J., 1982,

V.7, No.2, pp.52-60

2. Janowak, J. F., Gundlach, R. B., AFS Transactions,

1983, v.91, pp. 377-381.

3. Voigt, R.C., AFS Transactions, 1983, v.91, pp.253-

262.

4. Janowak, J.F., Morton, P.A., AFS Transactions,

1984, v.92, pp.489-492.

5. Voight, R.C. and Loper, C.R., J..Heat Treating,

1984, No.4, Vol.3, pp.291-309.

6. Moore, D.J., Rouns, T.N., Rundman, K.B., J. Heat

Treat., 1985, V.4 , pp.7-23.

7. P.A.Blackmore, R.A.Harding, J.Heat.Treat., 1984,

v.3, pp.310-325.

8. Moore, D.J., Rouns, T.N., Rundman, K.B., AFS

Transactions, 85-103, pp.705-717.

9. Gilbert, G.N.J., Report No. 1666, BCIRA, July

1986.

10. Rouns, T.N. and Rundman, K.B., AFS

Transactions, 1987, v.95, pp.851-875.

11. Viau, R., Gagne, M., Thibau, R., AFS

Transactions, 87-77, pp.171-178.

12. Rundman, K.B., Moore, D.J., Hayrynen, K.L.,

Dubensky, W.J, Rouns, T.N., J.Heat Treat.,

1988,v.5,pp.79-95.

13. White, P., Report No. 1787, BCIRA, pp.381

14. Grech, M and Young, J.M., Mater. Sci. and Tech.,

1990, v.6, pp.415-421.

15. Krishnaraj, D., Narasimhan, H.N.L., Seshan,S.,

AFS Transactions, 92-100, pp.105-112.

16. Darwish, N and Elliott, R., Mater. Sci. and Tech.,

1993, v.9, pp.572-585.

17. Darwish, N and Elliott, R., Mater. Sci. and Tech.,

1993, v.9, pp.586-602.

18. Darwish, N and Elliott, R., Mater. Sci. and Tech.,

1993, v.9, pp.882-889.

19. Bahmani, M., Elliott, R., Mater. Sci. Tech., 1994,

v.10, pp. 1050-1056.

20. Korichi, J,.Priestner,R., Mater Sci.Tech., 1995,

v.11, pp.901-907.

21. Krishnaraj, D., Seshan,S., AFS Trans., 95-119,

pp.767-776.

22. Hamid Ali, A.S. and Elliott, R., Mater. Sci. and

Tech., 1996, v.12, pp.780-787.

23. Hamid Ali, A.S., Elliott, R., Mater. Sci. Tech.,

1997, v.13, pp.24-30

24. A.A. Nofal, Jekova,, L., Journal of the University

of Chemical Technology and Metallurgy, 2009,

vol.44, pp.213-228

25. Duckworth, W.E., “Statistical Techniques in

Technological Researché, p.60-86, Methuen

and Co., 1968, London.

26. Oriani, R.A., Acta Met., 1964, v.12, p.1399.,

CHEMISTRY EFFECT ON PROPERTIES (SAMPLE-

HEAD2) This is another sample of subhead (HEAD2.) There may be several second level subheadings and within each second level heading, there may be third level subheads (Head3). “Body” text style is used for the main text. You may have several third level subheadings and within each third level subheading, you may have fourth level subheadings (Head4). “Body” text style is used for the main text as shown. Yield Stress (Sample Head3) Predicted Values (Sample Head4) Here is another sub-subheading (Head4). You may have several fourth level subheadings. Use “body” text style. Bullet Styles (Unord list & Ord

list) • An unordered bullet, no numbers (Times 10) • This unordered list has no numbers 1. An ordered bullet list, uses numbers (Times 10) 2. This ordered list contains numbers.

Lists should be flush left. FIGURE

S Figure captions should be a complete sentence and be located below each figure (left justified) in Figure caption format (Figure Arial 9 BF Ital) see example below. The figure should be located in the text after it is cited. Do not use a list of figures at the end of the paper. Do not wrap text around artwork. If artwork spans two columns, see the last page.

www.wfc2012.com