Journal of Metals, Materials and Minerals, Vol.21 No.1 pp.

101-105, 2011

Applying Computer Simulation in Improving Heat Treating Condition of
Thin High-Carbon Steel Parts

Thanaporn KORAD*, Mana POLBOON, Niphon CHUMCHERY and John PEARCE

National Metal and Materials Technology Center, Klong Luang,
Pathumthani 12120 Thailand

Abstract

The study was carried out using COSMOSFlowWorksto predict heat treatment conditions for
AISI 1085 high-carbon steel sheet that is used to produce automotive parts that require wear resistance
and stiffness resulting from controlled moderately high hardness levels. To achieve such properties, flat
high-carbon steel parts need to be hardened to produce acicular matrix structures, and the most suitable
heat treatment process to harden thin parts without distortion is austempering. In producing hard and stiff
thin section parts in this company study the austempering process was performed by soaking at 830
o
C
and then quenching in a NaCl salt bath at 335
o
C. In production hardness testing is performed to ensure
that parts have microstructures of lower bainite and martensite instead of upper bainite. To reduce
distortion without any effect on hardness, modified austempering conditions were determined using
temperature prediction from commercial computational analysis software. This enabled a more suitable
production line practice for the production of austempered parts without distortion whilst avoiding
reduced hardness.

Key words: Austempering, Hardening, High carbon steel, Bainite structure

Introduction

Austempering is a heat treatment practice
that enables the production of hard-acicular-
structures in medium-to-high carbon steels
without distortion. Conventional austempering
ensures the formation of fully bainitic or mixed
bainite-martensite microstructures in steel by
controlling the cooling conditions of parts with
respect to their design and section size in relation
to the Time-Temperature-Transformation
characteristics of the steel.

Thin section steel parts for automotive
applications are made from AISI 1085 high-
carbon steel (Table 1) and require suitable
hardening heat treatment to provide sufficient
strength and stiffness according to Tu et al.
(6)
.
Typically, austempering practice helps in
achieving high strength with good ductility and
toughness by evolving a predominantly bainitic
microstructure in the steel according to Saxena et
al.
(4)
. The parts require specified tensile strengths
of above 1,350 MPa, but only hardness testing
can be used to qualify such parts. The specified
hardness range for the steel parts in this study is
420-520 HV. Most of the parts can satisfy this
requirement, but 5-10% of these parts are rejected
due to low hardness of 400-420 HV from
structures of upper bainite possibly caused by low
cooling rate. To ensure sufficient hardness levels
in heat treated parts, production is operated at
high cooling rate to produce hard bainitic
structures; however such rapid cooling can cause
some heat treated parts to be rejected because of
distortion. Austempering is therefore controlled to
produce mixed lower bainitic-martensitic
structures instead of the lower hardness upper
bainite-ausferrite structures according to
Chakraborty et al.
(2)
, but this must be balanced
against preventing distortion.

The first part of austempering practice is
complete austenitisation at sufficiently high
temperature to enable complete solution of all
carbides and to homogenize the austenite. The
steel is then quenched in a salt bath or fluidized
bed to transform the austenite to controlled
amounts of bainite and martensite. In the current
study the production line is continuous consisting
of austenitizing at 830
o
C, quenching into a salt
bath at 335
o
C and then cooling in water bath at
40
o
C.

Figure 1 shows the isothermal
transformation characteristics of a nominal
*
Corresponding author Tel: +66 2564 6500 ext. 4030; Fax: +66 2564 6336; E-mail: thanapk@mtec.or.th
102
Korad, T. et al.
0.8wt% C steel. Above the nose of the curve
(550
o
C) the austenite will transform to relatively
soft pearlite. Below the nose at temperatures
down to around 250
o
C the austenite will
transform to bainite which consists of acicular
ferrite containing very fine precipitated carbides,
and which is harder and stronger than pearlite.
Martensite forms from any residual austenite on
cooling after isothermal treatment in the salt bath;
the amount of residual austenite and hence
martensite depends on the holding time in the salt
bath.

Treatment in the range 450
o
C down to
350
o
C will normally result in upper bainite,
treatment from 350
o
C down to 250
o
C giving
lower bainite. Upper bainite has a lower hardness
than lower bainite. Transformation to different
microstructures, i.e. martensite, lower bainite or
upper bainite, or mixed structures, results in
different volumetric changes giving rise to
varying degrees of tendency towards shape
distortion of heat treated parts.

Table 1. Chemical composition of typical AISI
1085 compared with a random analysis
check on the examined parts.

Wt% C Si Mn P S
AISI1085 0.80-
0.90
0.15-
0.30
0.70-
1.00
<0.04 <0.05
Sample 0.83 0.18 0.43 0.01 0.006



Figure 1. TTT diagram for AISI 1080 steel (0.75-
0.85%C), quenching from austenitizing
to austempering temperature typically
is within a few seconds
(1)
.


Materials and Experimental Procedures

This study has examined the
microstructures of three austempered thin parts,
namely a hose ring clamp with outer diameter of
2.4 mm and thickness of 0.9 mm., a shim plate
with thickness of 0.6 mm., and a screen plate with
thickness of 1.5 mm. The hose ring clamp was
designed as a stiff part while the shim plate and
screen plate were designed to have a good wear
resistance. All parts need to be hardened since
high hardness is required. During mass production,
distortion of the parts is the main reason for
rejection rather than low hardness. Detail of the
austempering process and the controlled
parameters are shown in Figure 2, and the
geometry of a hose ring clamp (one of the studied
parts) is shown in Figure 3.


Figure 2. Temperature profile in experiment
performed in continuous furnace.



Figure 3. Hose ring clamp, one of three parts
which found distortion in the
production line.

All three part designs were made from
AISI 1085 eutectoid steel using the same heat
treatment time and temperature control in spite of
their differences in geometry. The manufacturer
had previously studied the problem of distortion
and decided to adjust the heat treatment condition
to softer austempering involving a slower
cooling rate. The trial resulted in decreased rejects
103
Applying Computer Simulation in Improving Heat Treating Condition of Thin High-Carbon Steel Parts
from distortion, but increased rejection due to
unacceptable low hardness levels. The other
possible alternative modification considered by
the manufacturer was increasing thickness of the
raw materials, from 0.9 mm., 0.6 mm., and 1.5
mm. to 1.0 mm., 0.7 mm., and 1.6 mm.,
respectively. Increasing thickness was studied
using flat sheet specimens to verify only
microstructure and hardness. However, before
applying increased thickness on the production
line, the operator wished to perform simulation to
study how cooling conditions such as temperature
gradient and cooling rate compare to those
previously experienced, particularly with regard
to problems of distortion and inadequate hardness
levels. With this aim COSMOSFlowWorksis
the commercial software that was applied to
simulate the austempering conditions.

In this study, the simulated austempering
condition was started at 830
o
C and flow velocity
of salt bath was set at 0.001 m/s for the external
flow simulation. The austempered solid bodies
were set following boundary conditions for AISI
1085 steel such as a density of about 7,400 kg/m
3

which was fixed while thermal conductivity and
specific heat were variable between 535 to 600
J /kgK and 15.1 to 26.6 W/mK, respectively. The
software computed using model equations
(6)
for
the austempered bodies, salt bath and the
temperature gradient such as continuity equation
(1), momentum equation (2) and energy equation
(3) as follows:

(1)
(2)
(3)

Results and Discussion

Microstructural examination of parts with
the original thickness rejected because of
distortion showed that parts produced from the
same steel in different thicknesses and
austempered under the same conditions could
have significantly different microstructures. Some
typical microstructures are shown in Figure 4.

The thinnest part, a shim plate having
thickness of 0.6 mm exhibited the finest acicular
structures (Figure 4 (a)) with HV0.1 hardness
levels of between 538 to 587 HV. The part with a
thickness of 0.9, a hose ring, had hardness of
between 537 to 597 HV (Figure 4 (b)) while a
screen plate, the thickest part at 1.5 mm, had
hardness of between 408 to 494 HV (Figure 4 (c)).
Figure 5 is the scatter plots between predicted
cooling rates and measured hardness levels.
Basically testing hardness on thin metal parts has
limitations that effect on the measurement
accuracy according to Li et al.
(3)
. Hence,
trendlines and error bars are also added in the
plots exhibiting that the increase of the average
hardness is a result of increasing the predicted
cooling rates.


(a)


(b)

(c)

Figure 4. Cross section microstructures of the
examined samples from edge to
center of (a) shim plate, (b) hose ring
clamp and (c) screen plate.
Micrographs show uniform acicular
and fine structures along the cut
surfaces (edge to center) ensuring that
the applied soft austempering could
result in hardening heat treatment.

By using COSMOSFlowWorks
software, heat flow during quenching in the salt
bath was analyzed for each of the three parts to
provide data on temperature gradients at different
positions of each part. This allowed the cooling
rates to be drawn for comparison with suitable
cooling curves for austempering practice in the
AISI 1085 TTT diagram. The simulated cooling
curves from the software calculation show that
significant differences exist between the edge and
the center areas of each part as well as at the outer
and inner edges of the holes in the hose ring and
screen plate designs. Figure 6 illustrates different
cooling curves resulting in significantly different
hardening and residual stress, exhibited from

104
Korad, T. et al.
simulation results while raw data from simulation
analyzed at 2.8 seconds after quenching is
displayed in Figure 7. The data reveals
temperature differential of almost 100
o
C within
the part that could give rise to distortion.

530
540
550
560
570
580
590
600
1300 1500 1700 1900 2100
Predicted cooling rate (C/s)
H
a
r
d
n
e
s
s

(
H
V
)

(a) Shim plate, predicted cooling rate is 1,330-2,020
o
C /s.

530
540
550
560
570
580
590
600
190 240 290 340 390 440
Predicted cooling rate (C/s)
H
a
r
d
n
e
s
s

(
H
V
)

(b) Hose ring, predicted cooling rate is 200-434
o
C /s.

400
410
420
430
440
450
460
470
480
490
500
270 320 370 420 470
Predicted cooling rate (C/s)
H
a
r
d
n
e
s
s

(
H
V
)

(c) Screen plate, predicted cooling rate is 285-504
o
C /s.

Figure 5. Plots relationship between predicted
cooling rates only at the steep stage
from austenitizing temperature to
austenite start (723
o
C) and hardness
checks from different positions in (a)
shim plate (b) hose ring clamp and (c)
screen plate.

Cooling was then simulated after
allowing for increased sheet thickness in each part.
It was seen that cooling rate was less variable
when sheet thickness was increased as indicated
by the data in Figure 8. By following the same
analysis steps on each part, the cooling curves
from the simulation for thicker raw material are
illustrated on the TTT diagram (Figure 9)
suggesting that the modified cooling conditions
would be suitable for austempering of AISI 1085
before actual application on the production line.



Figure 6. Plots of cooling curve simulated from
COSMOSFlowWorks analyzing on
sections that possibly cause distortion
happened on hose ring clamp.




Figure 7. Simulated data of temperature change
in hose ring clamp during
austempering, comparing between the
fastest cooling area and the thick
section.

The simulation experiment showed that
the redesign of the parts by increasing thicknesses
could reduce the differences between the fastest
and slowest cooling regions in each part. Equally
it was predicted that this modified cooling would
be suitable for AISI 1085 austempering practice.

Simulation helped the manufacturer to
redesign the parts with increasing thickness. This
105
Applying Computer Simulation in Improving Heat Treating Condition of Thin High-Carbon Steel Parts

was reinforced by trial samples to check
microstructural features and hardness. Bainitic
and martensitic structures in the thicker parts were
similar to those in the previous thinner sheet
designs; the average hardness levels were slightly
lower, but no distortion problems were
experienced in the trial production batch.



Figure 8. Comparing between before and after
increasing thickness aiming to reduce
distortion on screen plate.


Figure 9. Illustration of simulated data plotted to
show temperature profile for
austempering the studied part after the
design improvement.

Conclusions

The computational analysis simulation
technique of heat flow analysis showing
temperature gradient data assists producers in the
redesign of metal parts to meet exacting hardness
and mechanical property requirements without the
dangers of part shape distortion that can result
from excessive temperature differences within a
part during quenching treatments for hardening.

References

1. American Society for Metals. (1977). Atlas of
isothermal transformation cooling
transformation diagrams. : ASM.

2. Chakraborty, J ., Bhattacharjee, D. & Manna, I.
(2008). Austempering of bearing steel for
improved mechanical properties. Scripta
Mater. 59(2): 247-250.

3. Li, H. & Zhou, X. (2007). The Application and
Effects of Thin Sheet Hardness Reference
Materials without Aging Effect. In:
Proceedings of HARDMEKO 2007
Recent Advancement of Theory and
Practice in Hardness Measurement.
November 19-21, Tsukuba, J apan: 117-
119.

4. Saxena, A., Prasad, S.N., Goswami, S.,
Subudhi, J . & Chaudhuri, S.K. (2006).
Influence of austempering parameters on
the microstructure and tensile properties
of a medium carbonmanganese steel.
Mat. Sci. Eng. A-Struct 431(1-2): 53-58.

5. Tu, M.Y., Hsu, C.A., Wang, W.H. & Hsu, Y.F.
(2008). Comparison of microstructure and
mechanical behavior of lower bainite and
tempered martensite in J IS SK5 steel.
Mater. Chem. Phys. 107(2-3): 418-425.

6. Zienkiewicz, O.C. & Tayler, R.L. (2000).
Finite element method (5
th
ed.). (Vol.3
Fluid Dynamics): Butterworth-
Heinemann.