CAST IRON

UDC 669.15-196.5

EFFECT OF THE ATOMIC RADIUS OF ALLOYING ELEMENTS

ON FORMATION OF STRAIN MARTENSITE IN BAINITIC DUCTILE IRON

Cem Akca1 and Nihat G. Kinikoglu1

Translated from Metallovedenie i Termicheskaya Obrabotka Metallov, No. 9, pp. 19 21, September, 2010.

The effect of the atomic radius of the alloying elements (Ni, Cu, and Mo) on the rate of formation of strain-in-

duced martensite is investigated. Since the final aim of the study is production of railway-wheel bainitic iron,

a special device simulating the operation of material in a railway wheel has been constructed. Small test pieces

in the form of railway wheels with different contents of Ni, Cu, and Mo were cast and subjected to isothermal

hardening at 375C for obtaining a structure of upper bainite with high amount of retained austenite, which

made the observation of the transformation easier. A correlation equation relating the atomic radius of the al-

loying element to the rate of the martensitic transformation is suggested.

Key words: bainitic ductile cast iron, plastic deformation, martensitic transformation.

INTRODUCTION

Bainitic isothermal hardening (austempering) is a rela-

tively new process of heat treatment. Its two main advan-

tages are (1 ) the possibility of accurate control of the struc-

ture of the matrix, which ensures a wide spectrum of me-

chanical properties, and (2 ) development of transformation

by a martensitic mechanism, which raises the wear and fa-

tigue resistances of the metal. Hardening in the bainitic range

includes four stages, i.e., austenization at 850 950C, hard-

ening cooling, isothermal hold in a range of 235 450C,

and cooling. Two reactions occur in the isothermal stage of

the process during formation of bainite. In the first reaction

bainitic ferrite () and carbon-enriched austenite (ce ) form

from the austenite (), i.e.,

+ ce

. (1)

The second reaction consists in transformation of the car-

bon-enriched austenite ce into bainitic ferrite also known

as ausferrite, retained austenite r , and carbides [1, 2], i.e.,

ce ( + Fe

3C) +

r. (2)

The interval between the two reactions is called the

window of the process. Optimum mechanical properties

can be attained if the length of the isothermal hold is chosen

in the window of the process [1, 3]. The capacity for bainitic

hardening is the key factor determining the whole of the pro-

cess, because it is connected with the thickness of the casting

or with the volume fraction of the material of the casting that

transforms into bainite without an the eutectoid reaction [4].

This capacity depends primarily on the chemical composi-

tion of the casting. The main alloying elements used for rais-

ing the capacity for bainitic hardening of ductile iron are Ni,

Cu, and Mo [5]. Strain-induced martensitic transformation in

steels has been well known for more than 30 years; in

bainitic iron it has been studied in the recent years [2, 6, 7]. If

a specimen with an austenite-ferrite structure is loaded to a

stress exceeding the yield limit of its phase components, both

the ferrite and the retained austenite are strained in a plastic

manner, which leads to partial transformation of retained

austenite into very hard and brittle martensite M with

tetragonally distorted body-centered lattice [2] according to

the reaction

r+

r+ + M. (3)

As a result, the material is strengthened due to the com-

bined effects of strain hardening and formation of martensite.

Metal Science and Heat Treatment Vol. 52, Nos. 9 10, 2010

420

0026-0673/10/0910-0420 2010 Springer Science + Business Media, Inc.

1Department of Metallurgical and Materials Engineering, Yildiz

Technical University, Istanbul, Turkey (e-mail: akca@yildiz.edu.tr).

The aim of the present work consisted in determining the

effect of alloying elements on the rate of martensitic transfor-

mation in bainitic iron, which develops under the action of

plastic deformation.

METHODS OF STUDY

Material. Test pieces with chemical composition pre-

sented in Table 1 were melted in an induction furnace and

cast in the form of a wheel. Depending on the content of Ni,

Cu, and Mo the test pieces were divided into groups with

high and low contents of the alloying element. The distribu-

tion of the test pieces in the groups is presented in Table 2.

Heat Treatment. For easier observation of the transfor-

mation the parameters of the treatment for bainite were cho-

sen so as to obtain a maximum content of retained austenite.

The heat treatment included austenization in an Ar electric

resistance furnace at 900C for 30 min, an isothermal hold in

a nitrate salt bath at 375C for 55 min, and cooling in still air.

Plastic Deformation. The load applied to the test pieces

was simulated with the help of the SolidWorks Cosmos

Express software so as to ensure plastic deformation at a

stress exceeding the ranges of fluidity and endurance. The

testing parameters were as follows: total load 3135 N, load

on a test piece 1045 N, speed of rotation 1716 rpm, linear

speed 3 kmh, duration of testing (number of turns) 250,000.

The installation for the fatigue tests is presented schemati-

cally in Fig. 1.

RESULTS AND DISCUSSION

The content of retained austenite Aret was determined by

analyzing images with the help of a Leica DMRX micro-

scope and Leica Qwin software. We measured the fraction of

retained austenite before and after deformation and the con-

tent of martensite formed due to the deformation. The micro-

structure of the test pieces before and after plastic deforma-

tion is presented in Fig. 2. The lighter regions correspond to

retained austenite.

Though the surface contains a maximum amount of mar-

tensite, the high content of molybdenum in test piece 3 seems

to be the reason for its lower hardness due to the segregation

of molybdenum over grain boundaries, because this pro-

motes straining inside the bodies of grains. Having measured

the content of retained austenite we derived a parameter of

the relative rate of martensitic transformation (RRMT, %),

i.e.,

RRMT =A A

A

ret ret

ret

1 2

1

, (4)

where Aret

1and A

ret

2are the contents of retained austenite be-

fore and after plastic deformation, respectively, in percent.

Below we present the results of an evaluation of RRMT, i.e.,

Test piece RRMT, %

1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.58

2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36.08

3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45.56

It can be seen that the RRMT depends on the type of the

alloying element that controls bainitic hardening (Cu, Mo, or

Ni, see Table 2). The results of the tests also show the exis-

tence of correlation between the RRMT and the atomic ra-

Effect of the Atomic Radius of Alloying Elements on Formation of Strain Martensite 421

TABLE 1. Chemical Composition of Test Pieces from Bainitic Iron

Test piece

Content of elements, %

C Si Mn P S Mo Ni Cu Mg

1 3.36 2.30 0.240 0.032 0.009 0.007 1.640 0.024 0.053

2 3.52 2.21 0.281 0.030 0.011 0.005 0.018 1.440 0.048

3 3.23 2.19 0.340 0.032 0.006 0.990 0.021 0.104 0.042

TABLE 2. Distribution of Test Pieces with Respect to

the Content of Cu, Mo, and Ni

Test piece Cu Mo Ni

1

2

3

Notations: ) low content of the element; ) high

content.

17

2

6

3

4

5

Fig. 1. Schematic representation of the installation for fatigue tests:

1 ) test piece; 2 ) guides from hardened steel; 3 ) rotating disc with a

load; 4 ) test piece holder; 5 ) immobile disc; 6 ) journal bearings;

7 ) test piece stopper.

dius of the alloying element. We evaluated (in %) the differ-

ence between the atomic radii of the alloying elements and

iron (Table 4). The relation between this difference and the

value of RRMT is plotted in Fig. 3.2 The data are approxi-

mated by a function f (x ) = x + CFe , where x is the differ-

ence between the raii (in %), f (x ) is the RRMT, and CFe is a

constant corresponding to iron. The approximation was made

using the Origin Microcal 6.0 software. The observed de-

crease in the rate of martensitic transformation upon growth

in the difference of the atomic radii is explainable by the fact

that the atoms of the alloying element dissolved in iron by

the principle of substitution of lattice points introduce distor-

tions into the slip planes of the austenite lattice and thus hin-

der the formation of strain-induced martensite.

422 Cem Akca and Nihat G. Kinikoglu

20 m

20 m

20 m

10 m

10 m

10 m

a

b

c

d

e

f

Fig. 2. The microstructure of bainite

cast iron 1 (a, d ) 2 (b, e), 3 (c, f ):

a c) prior to plastic strain; d f ) af-

ter strain.

TABLE 3. Volume Fraction of Retained Austenite and Martensite

Test

piece

Aret

, % M, %

prior to deformation after deformation

1 31.34 17.06 14.28

2 40.30 25.76 14.64

3 42.57 31.68 10.89

2Undoubtedly, the difference in the atomic sizes of alloying ele-

ments can affect the rate of martensitic transformation of retained

austenite. However, this factor is not the only one. All the alloy-

ing elements mentioned influence the stability of austenite by

changing the location of the critical points (including those of the

bainitic and martensitic transformations), the hardenability of the

iron, and the mechanical properties of the austenite and mar-

tensite. Singling out of only one factor of this variety cannot give

a full picture of the effect of the elements (Ed. note).

CONCLUSIONS

1. The rate of transformation of austenite into martensite

in bainitic iron under plastic deformation has been deter-

mined by the method of image analysis. The deformation

was performed in a specially designed device for fatigue test-

ing of test pieces simulating railway wheels.

2. The rate of the transformation of austenite into mar-

tensite upon straining of bainitic iron decreases upon growth

in the difference in the radii of the alloying element (Ni, Cu,

or Mo) and iron.

We are obliged to Assistant O. Nuri Celik for the cast test

pieces supplied and to our colleagues from the Balkan Cen-

ter for Advanced Casting Technologies for their help with the

experiments.

REFERENCES

1. N. Darwish and R. Elliott, Mater. Sci. Technol., 9, 572 585

(1993).

2. J. L. Garin and R. L. Mannheim, J. Mater. Proc. Technol.,

143 144, 347 351 (2003).

3. B. Bosnjak, B. Radulovic, K. Pop-Tonev, and V. Asanovic, ISIJ

Int., 40(12), 1246 1252 (2000).

4. E. Dorazil, High Strength Austempered Ductile Cast Iron, Tech-

nical University, Brno, Czechoslovakia (1991).

5. N. Ahmadabadi, H. M. Ghasemi, and M. Osia, Wear, 231,

293 300 (1999).

6. C. Akca, in: National Heat Treatment Symposium, Istanbul Tech-

nical University, Istanbul, Turkey (2001) (Compact Disc, Pub-

lished in Turkish).

7. C. Akca and N. G. Kinikoglu, in: The 66th World Foundry Con-

gress, Istanbul, 6 9 Sept., 2004 (2004), Vol. 2, pp. 989 1002.

Effect of the Atomic Radius of Alloying Elements on Formation of Strain Martensite 423

r re Fe

rFe, %

50

40

30

20 5 0 5 10 15

RRT, %

R 2 = 0.9765

Fig. 3. Relation between the difference in the atomic radii of the al-

loying elements and iron nrel

= (re

rFe

)rFe

and the relative rate of

martensitic transformation RRMT.

TABLE 4. Difference in the Atomic Radii of the Al-

loying Element and Iron

Element r, nm (re

rFe

)rFe

Ni 0.124 1.58

Cu 0.128 1.58

Mo 0.139 10.31

Fe 0.126 0