CHINA FOUNDRY Vol.7 No.

Critical cooling rate on carbide

precipitation during quenching of
austenitic manganese steel
*Youn-Soo Ham1, Jeong-Tae Kim2, Si-Young Kwak2, Jeong-Kil Choi2 and Woo-Young Yoon3
(1. Daeyang R-Tech. Co., Ltd, 290-28, Daehwa-Dong, Daeduk-Gu, Daejeon, Korea; 2. e-Design Center, KITECH, 7-47, Songdo-Dong,
Yeonsu-Gu, Incheon 406-800, Korea; 3. Department of Materials Science and Engineering, Korea University, Anam-Dong, Seongbuk-Gu,
Seoul 136-701, Korea)

Abstract: Critical cooling rate to avoid carbide precipitation during quenching of austenitic manganese steel was
investigated by means of optical microscopy, image analyzer and numerical analysis. An efficient heat treatment
analysis program including temperature-dependent material properties was developed for the prediction of
cooling rate and probability of carbide precipitation during quenching by finite difference method. Time-dependent
heat transfer coefficient was adopted to achieve more precise results. Area ratio of carbide precipitation was
measured by image analyzer to determine the critical point of carbide precipitation. Temperature-dependent
critical cooling rate at that point was calculated by the developed numerical program. Finally, the probability of
carbide precipitation on the whole area of specimen can be predicted by the proposed numerical program and the
numerical result of a specimen was compared with the experimental result.

Key words: austenitic manganese steel; carbide precipitation; quenching; critical cooling rate
CLC number: TG142.72/156.31 Document code: A Article ID: 1672-6421(2010)02-178-05

A ustenitic manganese steel (0.9%-1.4%C, 1%-15%Mn),

so-called Hadfield steel, has good wear resistance, high
toughness and ductility because of its high work-hardening
capacity. This material is widely used in equipment for handling
and processing earthen materials such as rock crushers,
grinding mills and so on. However it is generally considered
too brittle for normal use under as-cast condition. And carbide
precipitation at the grain boundary induces low toughness
and low elongation during casting. So solution treatment and
quenching process should be applied after casting process in
order to obtain the homogeneous austenitic structure[1-3,6]. If the
cooling rate during quenching period is relatively low in heavy
section, the opportunity of carbide precipitation is increased
*Youn-Soo Ham
Male, born in 1956, Ph.D candidate in metallurgical engineering at Korea
University. He earned a B.S. in metallurgical engineering from Korea
Sungkyunkwan University in 1984, and a M.S. in metallurgical engineering
from Korea Chungnam National University in 1992. After graduating, he
gained enormous practical experience during almost 14 years works
in fields of R&D of vehicle parts, casting of abrasion-resistant steel and
stainless steel. With that experience, he has been, since Jan. 2008,
the CEO of Daeyang R-Tech. Co., Ltd, specialized in the research and
production of abrasion-resistant steels and stainless steels.
Corresponding author: Jeong-Tae Kim, naljeby@kitech.re.kr.
Male, born in 1967, Ph.D candidate in metallurgical engineering at Yonsei
University, principal researcher of Korea Institute of Industrial Technology
(KITECH), Director of KITECH e-Design Center. His research mainly
focuses on R&D of heat treatment and its computer simulation.
Received: 2010-02-21; Accepted: 2010-03-28
during heat treatment. In this paper, we studied on critical
cooling rate to avoid carbide precipitation during quenching
of austenitic manganese steel by experimental and numerical
methods. Temperature history was measured by thermo
couples during quenching and microstructure was investigated
by optical microscope. And numerical analysis program was
developed which can predict the temperature distributions and
probability of carbide precipitation during quenching.
1 Experimental methods
A specimen(150 mm 150 mm 250 mm cubic shaped) was
prepared by sand casting of 1.2%C-0.6%Si-13%Mn-1.8%Cr
steel (ASTM A128 grade C) which can represent the thickness
of general crusher part [3,5]. Figure 1 shows the schematic
drawing of specimen and K-type thermocouple was used at
3 points along center line to measure temperature history
during quenching. In order to verify the accuracy of thermal
analysis, calculated temperature history is compared with
experimental ones. In general, full solution of carbide requires
a solution-treating temperature that exceeds Acm by about 30-
50 and soaking for 1-2 hours per 25 mm thickness. For
solution treatment, specimen was heated up to 1,050 for 21
hours and quenched in water without agitation[1,3]. During the
quenching, temperature of specimen was measured in 3 point
by acquisition system and temperature of quenching medium
(water) is also measured. For microstructure investigation, the

178
 Overseas Foundry
May 2010

specimen was cut into 15 small pieces by wire-cutting machine

(a)
to prevent the work hardening or phase transformation by heat
effect during machining. Microstructure is observed by optical
microscope after polishing and etching with 3% nital solution.
In order to determine the critical point of carbide precipitation,
measurement of carbide fraction is also performed by image
analyzer in each smaller area (5 mm 5 mm).

(b)

Fig. 1: Schematic drawing of the specimen and position (c)

of thermocouples

2 Numerical analysis
Thermal analysis program was developed for solving 3D
transient heat conduction problem by conventional FDM
scheme[7] and some numerical expression is also proposed for
prediction of carbide precipitation.

2.1 Thermal properties of ASTM A128 grade C

In order to solve the 3D heat conduction problem, accurate
Fig. 2: Physical properties used in the calculation
thermal properties (density, specific heat, thermal conductivity)
of alloy are needed. It is hard to obtain the accurate calculation
results such as solidification time or cooling rate when constant
thermal properties were used. In the case of casting process
analysis, constant thermal properties are sufficient to predict
the final solidification region(shrinkage defect area). But
temperature dependant thermal properties are necessary when
heat treatment analysis is performed because very small error of
temperature or cooling rate can affect microstructure evolution.
Figure 2 shows the thermal properties of ASTM A128 grade C
as a function of temperature used in heat treatment analysis[3].

2.2 Boundary conditions used in analysis

Temperature of water used in this analysis was regarded as
Fig. 3: Temperature of water during quenching
a function of time from quenching experiment and its result
is shown in Fig. 3. There are three stages of heat removal in
general quenching in liquids[4,8] . The first stage of cooling, this stage is very low because the film of vapor is acting as
so called film boiling stage, is characterized by vapor blanket an insulator. The highest cooling rates occur in second stage,
of quenchant around the metal specimen and cooling rate of so called nucleate boiling stage, after the vapor film has been

179
 CHINA FOUNDRY
destroyed. Heat is rapidly removed from metal surface by
vaporization of quenchant. Convective cooling stage is the final
stage of heat removal. When the temperature of metal surface
falls below the boiling point of quenchant, the boiling stops and
heat removal occurs by conduction and convection only, so the
relatively slow cooling rate can be seen in this period[3,9] . Heat
transfer coefficient was regarded as function of time in order to
consider these three different cooling stages.
2.3 Calculation of carbide precipitation probability
Critical cooling rate of carbide precipitation can not be
determined in short temperature range during whole cooling
period. In order to evaluate the critical cooling rate of each
temperature, start position of carbide precipitation (depth from
surface) should be determined by microstructure observation
after the heat treatment. From the temperature history of this
position which is calculated by thermal analysis, critical cooling
curve can be defined. Start temperature of carbide precipitation
in high manganese steels varies with the composition of carbon
contents and in the case of 1.2% C is known as 1,000 [10]. So,
cooling rate comparison range is settled from 1,000 to 250 .
And probability of carbide precipitation can be calculated by
below numerical expressions:
C
Pcarbide = 100
N the maximum value can be obtained at the center of specimen.
Where:
Pcarbide : Probability of carbide precipitation
(a) (b)
(d)
Fig. 4: Optical microstructure obtained at : (a) as-cast
4.5 cm from surface ; (b), (c) and (d) after heat
treatment 1 cm, 4.5 cm, 8.5 cm from surface
180
Vol.7 No.2
N: Iteration No of time step (1,000 to 250 )
T
C: The number of Ccritical > CijkT
Ccritical : Critical cooling rate of temperature T
T
CijkT : Cooling rate of temperature T of element i, j, k
3 Results and discussions
3.1 Microstructure observation
Figure 4 shows the photos of optical microscope. Figure 4(a)
indicates the result of as-cast specimen (4.5 cm apart from
the surface) and large carbides can be detect along grain
boundaries because of low cooling rate during solidification.
Figure 4 (b),(c),(d) show the results from 1 cm, 4.5 cm and
8.5 cm apart from surface after heat treatment, respectively.
Almost homogeneous austenite grains is shown in Fig. 4(b),
fine and coarse carbides can be seen in Fig. 4(c) and (d).
3.2 Determination of starting point of carbide
precipitation
In order to determine the starting point of carbide precipitation,
the fraction of carbides is measured by image analyzer. The
fraction of carbide is measured for 6 times and calculated to
average value in each small area of specimen (5 mm 5 mm).
Figure 5 indicates its results by graph. It can be seen that the
fraction of carbide is rapidly increased at the 4.5 cm depth and
Therefore the position of 4.5 cm depth can be defined as the
critical point of carbide precipitation.
(c)
Fig. 5: Variation of the carbide ratio with distance from surface
 Overseas Foundry
May 2010

3.3 Determination of heat transfer coefficient

From the quenching experiment, start time of nucleate
boiling stage was evaluated to 20th second and start point of
convective cooling stage was 480th second. The convection
heat transfer coefficient was evaluated by trial-and-error
method using comparison between experimental temperature
history and numerical simulation results (Fig. 6). Heat transfer
coefficient increased rapidly from 20 second to 100 second
(early stage of nucleate boiling) and increasing speed slowed
down between 100-200 second range.

(a) constant heat transfer coefficient

(b) Variable heat transfer coefficient

Fig. 6: Heat transfer coefficient used in numerical analysis
Fig. 7: Thermal histories at different locations

Figure 7(a) indicates the results of thermal analysis using

constant heat transfer coefficient (0.018) and Fig. 7 (b) shows
the results using evaluated time-dependent heat transfer
coefficient. Relatively large discrepancies exist in each three
measure points when using constant heat transfer coefficient
but cooling curves show good agreement with experimental
ones over whole quenching range when using time-dependent
heat transfer coefficient. Finally, adequate time-dependent heat
transfer coefficient was evaluated for quenching analysis of
austenitic manganese steels.

3.4 Prediction of carbide precipitation region

Fig. 8: Cooling rates at different locations obtained
Calculated cooling curves for three different points is shown by the calculation
in Fig. 8. Cooling rate of Horizontal_55cm is smaller than
critical cooling rate over whole range and probability of
carbide precipitation is expected to 100%. But cooling curve
of Horizontal_35cm is located in right side of critical cooling
curve therefore probability of this position is expected to
0%. In the case of Vertical_15cm, cooling rate is greater than
critical value over almost whole range except 750-650
and the probability is expected about 15%. Figure 9 is the
comparison result of carbide precipitation region between
image analyzer and calculated probability. From the result
of image analyzer, specimen is divided into three regions:
homogeneous austenite, fine carbide precipitation (transient)
and coarse carbide precipitation. From the comparison, if the (a) Experimental (b) Calculation
probability is less than 20% in some area that zone can be
expected to homogeneous austenite area. Fig. 9: Carbide precipitation region

181
 CHINA FOUNDRY
3.5 Application for real casting product
Developed analysis program was applied to a real cone crusher
casting and its shape is described in Fig. 10 and its material
and heat treatment condition are the same with the specimen
used in experiment.
Figure 11 indicates the probability of carbide precipitation in
cone crusher by numerical analysis. The carbide precipitation
is predicted at the lower part of the cone crusher in wide range;
the predicted region is agreement with the poor life-time part
under the real operation due to lack of toughness.
Fig. 10: 3-D model of the cone for the industrial application
Fig. 11: Carbide precipitation region of the cone crusher
by numerical analysis
182
Vol.7 No.2
4 Conclusions
(1) The critical cooling rate as carbide precipitation
control parameter was determined during quenching using
experimental and numerical methods for austenitic manganese
steels.
(2) A specimen of ASTM A128 grade C steel is quenched
in water bath with no agitation. Carbide precipitation was
observed from depth of 45 mm from surface by image analyzer
and cooling rate curve of this depth is selected as the critical
cooling rate of the carbide precipitation.
(3) Numerical analysis program was developed which can
predict the probability of carbide precipitation and numerical
result using time-dependent heat transfer coefficient shows
good agreement with experiment ones.
References
[1] The Iron and Steel Institute of Japan. Heat Treatments of Steel.
Gijeon publisher, 1993.
[2] Lee S Y, Kim H K and Kim H Y. Heat Treatments of Steel.
Wonchang publisher, 1997.
[3] Subramanyam D K, Swansinger A E and Avery H S. Austenitic
Manganese Steels. In: Metals Handbook, Vol. 4, American
Society of Metals, 1991: 822-840.
[4] Bates C E, Totten G E. and Brennan R L. Quenching of Steel.
In: ASM Handbook, Vol. 4: Heat Treating, ASM International
Materials Park, OH, 1991: 67-120.
[5] J a c k s o n W J a n d H u b b a r d M W. S t e e l m a k i n g f o r
S t e e l f o u n d e r s . S t e e l C a s t i n g s R e s e a r c h a n d Tr a d e
Association, Sheffield , 1979: 106.
[6] Avery H S. Austenitic Manganese Steels. In: Metals Handbook,
Vol. 1, 8th ed., American Society of Metals, 1961: 835
[7] Hong C P. Numerical Analysis of Heat Transfer by FDM, FEM
and BEM. Bando publisher, 1991. (in Korean)
[8] Incropera F P and Dewitt D P. Introduction to Heat Transfer.
2nd ed. John Wiley & Sons, Inc., N. Y., 1990.
[9] Schwalm M and Tensi H M. Heat and Mass in Transfer
Metallurgical Systems. In: Proceedings of Conference
organised by International Centre for Heat and Mass Transfer,
Dubrovnik, Yugoslavia, Sept., 1979: 563-572.
[10] Bain E C, Davenport E S and Waring W S N. The Equilibrium
Diagram of Iron-Manganese-Carbon Alloys of Commercial
Purity. Trans. AIME, 1932, 100: 228-249.