j m a t e r r e s t e c h n o l .

2 0 2 0;9(2):1929–1940

Available online at www.sciencedirect.com

www.jmrt.com.br

Original Article

Constitutive analysis and dynamic recrystallization

behavior of as-cast 40CrNiMo alloy steel during
isothermal compression

Wei Wang a , Rui Ma b,∗ , Lianping Li c , Ruixue Zhai b , Shibo Ma a , Huajun Yan a ,
Shuangjie Zhang a , Siyuan Gong c
a Hebei Key Laboratory of Material Near-Net Forming Technology, Hebei Engineering Laboratory of Aviation Lightweight Composite
Materials and Processing Technology, Hebei University of Science and Technology, Shijiazhuang 0500181, PR China
b Key Laboratory of Advanced Forging & Stamping Technology and Science (Yanshan University), Ministry of Education of China,

Qinhuangdao 066004, PR China

c Hebei Construction & Investment Group Co., Ltd, Shijiazhuang 050051, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Casting-forging combination forming process is an advanced manufacturing technique

Received 2 September 2019 which is applicable to manufacture the parts with both complex shape and high perfor-
Accepted 9 December 2019 mance. As a fundamental research of casting-forging combination forming process, as-cast
Available online 7 January 2020 40CrNiMo alloy steel is obtained through vacuum casting by using metal mold. The isother-
mal compression tests of as-cast 40CrNiMo alloy steel are implemented on a Gleeble-3800
Keywords: thermal simulation machine at deformation temperatures of 800, 900, 1000 and 1100 ◦ , with
As-cast 40CrNiMo alloy steel strain rates of 0.001, 0.01, 0.1, 1 and 10 s−1 . The results indicate that the true stress-strain
Isothermal compression test curves present typical dynamic recovery type under low deformation temperature and high
Constitutive analysis strain rate. With the increases of deformation temperature or the decreases of strain rate,
Dynamic recrystallization the true stress-strain curves gradually transform to dynamic recrystallization type. The
Kinetic model Arrhenius-type constitutive equation with Zener–Hollomon parameter is determined for
Kinematic model constitutive analysis. The kinetic model and kinematic model of dynamic recrystallization
are deduced to describe the dynamic recrystallization behavior.
© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

material utilization. This manufacturing technology is appli-

1. Introduction cable to manufacture the parts with both complex shape and
high performance, the parts can be firstly formed to a near
The casting-forging combination forming process is a com-
final shape by casting, and then formed to the final shape by
bination of casting and forging, which has the technological
forging. The microstructure of parts can be transformed from
characteristics of both casting and forging. So it has the
casting structure to forging structure through the dynamic
advantages of simplifying production process and improving
recrystallization mechanism [1–3].
In the development of the casting-forging combination
forming process, Kim et al. [4] proposed tie-rod ends can be
∗
Corresponding author.
manufactured by casting-forging process, an optimal config-
E-mail: [email protected] (R. Ma).
https://doi.org/10.1016/j.jmrt.2019.12.025
2238-7854/© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
1930 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940

uration of the cast preform was determined. Wang et al. [5] The adiabatic shear failure mode predominates in the fracture
suggested the casting-forging process can be applied in the behaviour. The features of dislocations and the precipitation
manufacturing of large aluminum flange, it can reduce press of particles are changed in accordance with the variation of
capacity and material costs. A casting-forging forming process the strain rates and the loading temperatures.
of alternator poles was investigated by Chen et al. [6] and Song The processing maps of 4340 steel were constructed by
et al. [7] Their reports indicated that casting-forging forming Aneta et al. [22] for optimization of the hot forging parame-
process can not only shorten process, decrease material and ters. The isothermal compression tests are performed at the
power demand, but also can increase the forming accuracy temperatures ranging from 800 to 1200 ◦ and at the strain rates
and mechanical properties. in the range of 0.01–100 s−1 . It was found that the temperature
A comprehensive analysis of microstructure and mechan- range of 1050–1200 ◦ and strain rate range of 3–57 s−1 shows
ical properties of casting-forging, low-pressure-casting and the best parameters of processing, and these conditions can
rheo-casting of A356 aluminum alloys was investigated by lead to the occurrence of dynamic recrystallization.
Lee et al. [8]. The results proved that tensile properties and However, it should be pointed out that all the above stud-
fracture toughness of the cast-forged alloy are more superior. ies are about as-rolled 40CrNiMo alloy steel. These research
They also researched the effects of eutectic silicon particles results are not applicable to the process analysis of casting-
on mechanical properties of A356 aluminum alloys fabricated forging combination forming. Therefore, the hot deformation
by casting-forging, low-pressure-casting and squeeze-casting. and dynamic recrystallization behaviors of as-cast 40CrNiMo
The cast-forged alloy also has the best mechanical properties alloy steel are investigated. The Arrhenius-type constitu-
[9]. tive equation with Zener–Hollomon parameter is determined
In order to enhance the mechanical properties of Mg alloy for constitutive analysis. The kinetic model and kinematic
components, a double control forming technology combining model of dynamic recrystallization are deduced to describe
die casting and forging was proposed by Jiang et al. [10–13]. the dynamic recrystallization behavior. The research results
The microstructure and mechanical properties of A356 alu- of this study can provide basis for the process analysis and
minum alloy wheels prepared by thixo-forging combined with numerical simulation of casting-forging combination forming
a low superheat casting process were explored by Wang et al. in 40CrNiMo alloy steel.
[14]. Results indicated that the tensile strength and elongation
of thixo-forged wheels are higher. The casting-forging com-
bination forming technology was suggested to manufacture 2. Materials and methods
automobile brake bracket by Zhou et al. [15], and the pro-
cess parameters and mechanical properties are researched. As-cast 40CrNiMo alloy steel is obtained by vacuum cast-
It is observed that the studies of casting-forging combination ing using vacuum induction melting furnace and metal mold
forming process are focused on the process optimization and with the process parameters casting temperature 1550 ◦ , pre-
mechanical property analysis. The reports of the characteris- heating temperature of metal mold 300 ◦ , casting time 5 s.
tics of as-cast material are very few. Cooling method is furnace cooling. The size of casting blank
40CrNiMo (AISI 4340) alloy steel is a medium carbon low is 42 mm in diameter and 235 mm in height. The chemical
alloy steel. Due to a good balance of strength, toughness components of casting blank are detected by using optical
and wear resistance, it is widely used to manufacture auto- emission spectroscopy method (quantometry analysis). The
motive components, power transmission gears, shaft, and result (wt%) is 0.415C–0.64Si–0.78Mn – 0.74Cr–0.82Ni–0.206Mo-
structural parts [16,17]. There have been some research reports 0.0085S–0.021P–(bal.)Fe.
of hot deformation and dynamic recrystallization behaviors Specimens for isothermal compression are cut from cast-
of 40CrNiMo alloy steel. Hot deformation characteristics and ing blank at different positions using wire cut electrical
dynamic recrystallization behavior of 4340 steel were inves- discharge machining. Specimen size is 10 mm in diameter and
tigated by Sajadifar et al. [18,19]. Hot compression tests were 15 mm in height. The cutting positions are shown in Fig. 1.
performed at a temperature range of 900–1200 ◦ , a strain rate The three positions separately located at the surface zone,
range of 0.01–1 s−1 , and a strain of 0.9. The results proved transition zone and core zone. The metallographic structures
that the evolution of dynamic recrystallization (DRX) grain of the three positions are shown in Fig. 2. They are all typi-
structures can be accompanied by a considerable migration cal casting dendritic microstructure. The average grain size of
of grain boundaries. The activation energy obtained in their each position is tested three times in different areas by using
research is 427.2 kJ/mol. Sanrutsadakorn et al. [20] researched composite grid method according to the ASTM standard. The
the initiation of dynamic recrystallization in AISI 4340 steel. average grain size is the average value of the three test results.
The results showed the DRX occurred during hot deformation Position 1 is 240.03 ␮m, Position 2 is 256.45 ␮m, and Position 3
started when the normalized critical stress and strain reached is 268.94 ␮m. The average grain sizes of the three positions has
the values of 0.735 and 0.324, respectively. little difference, which indicate that the microstructure of the
The deformation behaviour of AISI 4340 alloy steel under casting blank is relative uniform.
high strain rates was investigated by Lee et al. [21]. The tests The isothermal compression tests are implemented on
were performed at strain rates ranging from 500 to 3300 s−1 a Gleeble-3800 thermal simulation machine at deformation
and constant temperatures ranging from 25 to 1100 ◦ on a temperatures of 800, 900, 1000 and 1100 ◦ , with strain rates
split Hopkinson bar. The results indicated that the flow stress of 0.001, 0.01, 0.1, 1 and 10 s−1 . These are 16 test samples for
of AISI 4340 alloy steel increases with the increase of strain each position, a total of 48 test samples. Each surface of the
rate, but decreases with the augmentation of temperature. specimen is covered with the high temperature lubricant and
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940 1931

Fig. 1 – Cutting positions.

Fig. 3 – Flow chart of isothermal compression.

tantalum foil to minimize friction. The flow chart of isother-

mal compression is shown in Fig. 3. The specimen is initially under the same deformation conditions, especially under the
heated to 1200 ◦ C at 10 ◦ /s (step )
1 and held for 120 s for com- deformation condition of high deformation temperature and
plete austenitizing (step ). 2 Then, the specimen is cooled low strain rate. Therefore, the influence of initial grain size for
to deformation temperature at 10 ◦ /s (step ) 3 and held for this batch as-cast 40CrNiMo can be negligible.
180 s for temperature homogenization (step ). 4 After that, the The follow-up study is carried out by using the isother-
specimen is compressed under a constant strain rate (step mal compression tests data of Position 1. Fig. 5 shows the
),
5 reduction rate in the height is 60%. After the compres- true stress-strain curves of Position 1 under different deforma-
sion deformation, the specimen is quickly quenched into cold tion conditions. The true stress-strain curves present dynamic
water to keep the microstructure (step ).
6 The quenched spec- recovery type under low deformation temperature and high
imen is sliced along the axial section, and then polished and strain rate. With the increases of deformation temperature
etched with saturation picric for the observation of the prior or the decreases of strain rate, the true stress-strain curves
austenite grain boundaries. gradually transform to dynamic recrystallization type that the
stress-strain curves present unimodal pattern.

3. Results and discussion

3.2. Constitutive analysis
3.1. Hot deformation behavior
The Arrhenius-type constitutive equation is adopted for con-
stitutive analysis. In addition, the influence of the forming
The true stress-strain curves of the three positions at defor-
temperature and the strain rate on the hot deformation behav-
mation temperature of 900 ◦ and 1100 ◦ are shown in Fig. 4.
ior can be described by Zener–Hollomon parameter Z [23,24],
In the figure, Position 1 with solid line, Position 2 with dash
as shown in Eqs. (1)–(3).
line, Position 3 with dash & dot line. The results show that
the hot deformation behavior of the three positions is very
similar. It also proves that the as-cast structure of this casting ε̇ = AF() exp (−Q/RT) (1)
blank is relatively uniform. The true stress-strain curves of the ⎧ n
three positions all show that the true stress increases with the ⎪
⎨
1 ˛ < 0.8
increase of the strain rate, and it decreases with the increase F() = (2)
exp (ˇ) ˛ > 1.2
of deformation temperature. The true stress-strain curves of ⎪
⎩
the three positions present the same type and variation trend [sin h(˛)]n for ⁄3 all ⁄3 
1 1

Fig. 2 – Metallographic structure of casting blank.

1932 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940

Fig. 4 – True stress-strain curves of the three positions.

Fig. 5 – True stress-strain curves of Position 1.

Z = ε̇ exp (Q/RT) (3) Since that is meaningful to predict the maximum load in hot-
working process. The values of peak stress under different
Where ε̇ is the strain rate (s-1 ),  is the flow stress (MPa), deformation conditions are listed in Table 1. Base on the peak
Q is the activation energy (J·mol-1 ), R is the gas constant stress, the material constants of constitutive equation are con-
(R = 8.314 J·mol-1 K-1 ), T is the absolute temperature, and A, n1 , structed through linear regression method. The average slope
˛, n, and ˇ (ˇ = ˛n1 ) are the material constants. The formula of of the linear fitting curves of [ln p − ln ε̇] under different defor-
flow stress can be derived as follow. mation temperatures is accepted as n1 . The average slope of
  2/n
1/2  the linear fitting curves of [p − ln ε̇] under different deforma-
1
 Z 1/n Z tion temperatures is accepted as ˇ, and then ˛ can be solved
 = ln + +1 (4)
˛ A A by ˛ = ˇ/n1 . The average slope of the linear fitting curves of
[ln[sinh(˛p )] − ln ε̇] is accepted as n. The average slope of the
The peak stress, maximum stress in the true stress-strain linear fitting curves of [1000/T − ln[sinh(˛p )]] under differ-
curve, is usually used to construct the constitutive equation. ent strain rates is accepted as Q/(Rn), then Q (kJ·mol-1 ) can
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940 1933

Fig. 6 – Linear fitting curves for constitutive equation.

as-casting 40CrNiMo is 332.1 kJ/mol, that is less than as-rolled

Table 1 – Values of peak stress under different
deformation conditions (MPa). 40CrNiMo (427.2 kJ/mol) [18,19].
The Arrhenius-type constitutive equation of as-cast 40CrN-
Strain rate Deformation temperature (◦ )
iMo alloy steel is shown in Eq. (5). Fig. 6(f) shows the linear
(s−1 )
800 900 1000 1100 fitting curve of the measured value from isothermal com-
pression tests versus the predicted value from constitutive
0.001 100.15 62.044 40.241 24.687
0.01 145 89.793 59.652 41.071
equation. The R square is 0.992, which reveals the predicted
0.1 199.95 131.65 86.124 53.08 data calculated by Eq. (5) has a high accuracy.
1 234.97 170.97 123.63 90.426
10 257.82 209.36 160.52 124.24 ⎧ 
⎪  1/5.147
⎪
⎪ =
1 Z
⎪
⎪ 9.733 × 10−3
ln
5.544 × 10
⎪
⎪
12
be solved. At last, the intercept of the linear fitting curve ⎪
⎨  
1/2 
of [ln[sinh(˛p )] − ln Z] is accepted as ln A, then A can be Z 2/5.147
+ +1 (5)
solved. The linear fitting curves for each material constant are ⎪
⎪ 5.544 × 1012
⎪
⎪
shown in Fig. 6(a)–(e). The values of the material constants ⎪
⎪ 
⎪
⎪
⎩ Z = ε̇ exp 3.321 × 10
5
are: n1 =7.172, ˇ = 6.981 × 10−2 , ˛ = 9.733 × 10−3 , n =5.147, Q
8.314T
=332.1 kJ/mol, and A = 5.544 × 1012 . The activation energy of
1934 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940

Table 2 – Values of peak strain under different

deformation conditions.
Strain rate Deformation temperature (◦ )
(s−1 )
900 1000 1100

0.001 0.2728 0.1614 0.1297

0.01 0.3342 0.2366 0.1634
0.1 0.5393 0.3665 0.2554

of the recrystallization is not sufficient in these deformation

conditions. Partial initial casting structure is replaced by forg-
ing structure through incomplete dynamic recrystallization.
It is different from Fig. 8(a) and (b), there are no coarse
grains are found in Fig. 8(c) and 6(d), lots of recrystallized
Fig. 7 – Dynamic recovery and dynamic recrystallization grain are generated under these deformation conditions. Ini-
type stress-strain curves. tial casting structure has almost completely been replaced by
forging structure through complete dynamic recrystallization.
Fig. 8(e) and (f) shows that complete dynamic recrystalliza-
tion has taken place under these deformation conditions, and
3.3. Dynamic recrystallization behavior the recrystallized grains grow obviously. Although that is not
obvious, it still can be observed that the grain size of Fig. 8(f)
During the hot-working process, the effects of work harden-
is slightly larger than that of Fig. 8(e). This is due to the lower
ing and dynamic softening are concomitant. In initial stage
strain rate allowing the recrystallized grains to gain more time
of metal hot-working process, the proliferation and interac-
for growth.
tion of dislocation cause the work hardening, and the effect
Through comprehensive analysis of microstructures and
of work hardening is much greater than dynamic softening,
deformation conditions, it can be found that as-cast 40CrNiMo
the true stress presents linear growth. As the hot defor-
occurs incomplete dynamic recrystallization at low deforma-
mation progresses, the accumulated energy of the material
tion temperature and high strain rate. With the increase of
increases continuously. Driven by the accumulated energy and
deformation temperature or the decrease of strain rate, it
thermal activation energy, the effect of dynamic softening
makes the transition to complete dynamic recrystallization.
(dynamic recovery softening) is enhanced by the combination
The recrystallized grains significantly grow at high deforma-
and recombination of dislocations. The increasing rate of true
tion temperature and low strain rate.
stress slows down. After the stress reaches the critical value
(critical stress c ), if the dislocation density under a low level,
3.4. Kinetic model of DRX
the driving force for the nucleation of dynamic recrystalliza-
tion will be insufficient. In this case, the true stress will remain
Kinetic model of DRX proposed by Sellars is adopted in this
basically unchanged after reaching the peak value (peak stress
study, including peak strain equation and critical strain equa-
p ). This means that only dynamic recovery occurred, and
tion [27,28], as follow.
the true stress-strain curve presents the dynamic recovery
type. Rather, if the dislocation density is high enough for the   m1
εp = a1 Zm1 = a1 ε̇ exp [Q1 /(RT)]
nucleation of dynamic recrystallization, the dynamic recrys- (6)
tallization mechanism will be activated. In general, the effect εc = kεp
of the dynamic softening (mainly dynamic recrystallization
softening) can overcome the effect of the work hardening. In Where εp is the peak strain, εc is the critical strain for DRX, ε̇ is
this case, the true stress will drop suddenly after reaching the strain rate (s-1 ), T is the deformation temperature (K), R is
the peak value, and the true stress-strain curve presents the the gas constant, Q1 is the recrystallization activation energy
dynamic recrystallization type [25,26], as shown in the Fig. 7. (J·mol-1 ), a1 , m1 and k are material constants.
As shown in Fig. 5, when the deformation temperatures Taking the logarithm of both sides of the expression of εp
at 900 ◦ –1100 ◦ , with strain rates of 0.001 s−1 –0.1 s−1 , the true in the Eq. (6) gives
stress-strain curves present the dynamic recrystallization
type. Fig. 8 shows the microstructures of as-cast 40CrNiMo ln εp = ln a1 + m1 ln˙ ε + m1 Q1 /(RT) (7)
alloy steel after isothermal compression. These are some
banded shadows in the metallographic photos which are The peak strains under deformation temperatures at
derived from the dendrite segregation in the casting process 900 ◦ –1100 ◦ with strain rates of 0.001 s−1 –0.1 s−1 are listed
of alloy steel. The dendrite segregation is concentrated in the in Table 2. Based on these data, the material constants are
interdendritic regions and elongated during the compression resolved using linear regression method. The linear fitting
process. The dendrite segregation can be eliminated by sub- curves for each material constant are shown in Fig. 9(a)–(c).
sequent heat treatment. The values are: m1 = 0.1577, a1 = 0.005, Q1 =311.6 kJ/mol.
Fig. 8(a) and (b) shows that recrystallized grains have The expression of εp is shown in Eq. (8). Fig. 9(d) shows the
formed, but there are still some coarse casting grains. Because linear fitting curve of the measured versus the predicted εp .
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940 1935

Fig. 8 – Microstructures of as-cast 40CrNiMo after isothermal compression.

The R square is 0.982, which reveals that the expression of εp stress reaches steady state (ss ). From p to ss , the values of
has a high accuracy.  are negative, this is because the dynamic recrystallization
softening plays a dominant role in this stage. Assume that if
3.116 × 105 dynamic recrystallization does not occur, dynamic recovery is
εp = 0.005[ε̇ exp ( )]0.1577 (8) the only softening mechanism,  will maintain a linear decline
RT
after reach the critical stress c until it reaches zero. At this
Critical strain as a criterion for dynamic recrystallization point, the dynamic recovery softening and work hardening
can be obtained by the relation curve of work hardening rate  strengthening reach a balance, stress reaches the saturated
versus stress . The work hardening rate  reflects the change stress s .
rule of the stress change rate with stress. According to the In the work hardening curve, the inflection point from stage
five stage work hardening theory, in the process of metal hot IV to stage V is the initial point of DRX, so the mathematic
deformation, the relation curve of work hardening rate ver- relation between  and  at this point can be express as follow.
sus stress can be divided into (I) easy slip stage, (II) linear
hardening stage, (III) dynamic recovery stage, (IV) large strain
hardening stage, and (IV) dynamic recrystallization softening ∂2 
− =0 (9)
stage [29,30], as shown in Fig. 10. ∂2 
There is an obvious inflection point in the work harden-
ing curve from stage IV to stage V. The stress at this inflection Where, the formula of work hardening rate  is
point is the critical stress c . After the inflection point, hot
deformation enters into the dynamic recrystallization stage.
Work hardening rate  reaches zero twice, the first time is ∂
= (10)
when the stress reaches peak (p ), the second time is when ∂ε
1936 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940

Fig. 9 – Linear fitting curves for equation of εp .

Table 3 – Values of critical strain under different

deformation conditions.
Strain rate Deformation temperature (◦ )
(s−1 )
900 1000 1100

0.001 0.0726 0.0595 0.0456

0.01 0.1649 0.0713 0.0560
0.1 0.2640 0.1572 0.0709

Mirzadeh and Najafizadeh, ln  can be expressed by a cubic

polynomial of ε [31,32].

ln  = A1 ε + A2 ε2 + A3 ε3 + A4 1⁄3 1⁄3 1⁄3 1⁄3 1⁄3 (0 ≤ ε ≤ εp ) (12)

Fig. 10 – Relation curve of work hardening rate versus

So, the critical strain of dynamic recrystallization can be
stress.
expressed as

Eq. (11) can be derived from Eqs. (9) and (10). A2

εc = − (13)
3A3
  ∂   ∂ ln  
∂2  ∂ ∂ ∂ ∂
= = =
∂2  ∂ ∂
∂ε · ∂ε ∂  · ∂ε ∂ ∂ε The cubic polynomial fitting curves of ln  − ε at deforma-

∂
 ∂ ln   ∂
 ∂ ln   ∂2ln 
tion temperature of 1000 ◦ are shown in Fig. 11. The critical
= = = (11) strains under different deformation conditions are listed in
∂
∂ε · ∂ε ∂ε  · ∂ε ∂ε ∂2 ε
Table 3. According to the data in Tables 2 and 3, the material
constant k in Eq. (6) can be ascertained by the linear fitting
According to Eq. (11), the critical strain of DRX can be curve of εp − εc , as shown in Fig. 12, the slope is 0.5348, and
obtained by the ln  − ε curve. Based on the research of the R square is 0.908.
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940 1937

Fig. 13 – Work hardening rate-stress curves.

Fig. 11 – Cubic polynomial fitting curves of ln  − ε.

Table 4 – Values of saturated stress and steady stress

under different deformation conditions (MPa).
Strain rate Deformation temperature (◦ )
(s−1 )
900 1000 1100

s ss s ss s ss

0.001 53.2 62.65 34.1 42.99 22.03 26.8

0.01 75.6 94.5 49.76 64.32 33.79 44.15
0.1 122.2 127.4 76.8 95.08 45.93 58.48

Table 5 – Values of ε0.5 under different deformation

conditions.
Strain rate Deformation temperature (◦ )
(s−1 )
900 1000 1100
Fig. 12 – Linear fitting curve of εp − εc .
0.001 0.3187 0.2206 0.1471
0.01 0.5074 0.3257 0.2435
0.1 0.6698 0.4070 0.3161
So, the kinetic model of dynamic recrystallization of as-cast
40CrNiMo alloy steel can be expressed as
 rate-stress curve. And then XDRX under different deformation
εp = 0.005{ε̇ exp [3.116 × 105 /(RT)]}0.1577
(14) conditions can be calculated.
εc = 0.5348εp The work hardening rate-stress curves at deformation tem-
perature of 1100◦ are shown in Fig. 13. The saturated stress s
3.5. Kinematic model of DRX and steady stress ss under different deformation conditions
are listed in Table 4. Based on the data in the Table 4, the stress
Dynamic recrystallization volume fraction model is an impor- values corresponding to 50% dynamic recrystallization is cal-
tant part of kinematic model. Through the metallographic culated by using Eq. (15). Then the corresponding strain ε0.5
analysis to determine the DRX volume fraction will be a dif- can be determined by the true stress-strain curve. The val-
ficult and tedious process. Therefore, the method presented ues of ε0.5 under different deformation conditions are listed in
by Sellars is adopted, the DRX volume fraction can be ascer- Table 5.
tained by the flow stress-strain curve [33–36]. The DRX volume On the basis of Avrami recrystallization theory, the kine-
fraction XDRX can be expressed as follows. matic model of DRX can be described as follows.

s −  1 1 1 ⎧
XDRX = ⁄3 ⁄3 ⁄3 (ε ≥ εc ) (15) ⎪ XDRX = 01⁄3 1⁄3 1⁄3 1⁄3 1⁄3 (0 ≤ ε ≤ εc )
s − ss ⎪
⎪ 
⎨ 

ε − εc kd
XDRX = 1 − exp −ˇd 1
⁄3 1⁄3 1⁄3 1⁄3 1⁄3 (ε ≥ εc ) (16)
Where s is the saturated stress, ss is the steady stress,  is ⎪
⎪ ε0.5
the flow stress.  can be obtained from the true stress-strain ⎪
⎩
curve, s and ss can be obtained from the work hardening ε0.5 = a2 ε̇m2 exp [Q2 /(RT)]
1938 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940

Fig. 14 – Linear fitting curves for ε0.5 .

Fig. 15 – Linear fitting curve of measured value versus Fig. 16 – Linear fitting curves of
predicted value. ln[(ε − εc )/ε0.5 ] − ln[− ln(1 − ln XXRD )].

Taking the logarithm of both sides of the expression of XDRX

Where XDRX is the DRX volume fraction, εc is the critical strain,
in the Eq. (16) gives
ε0.5 is the strain for 50% dynamic recrystallization, ε̇ is the
strain rate (s−1 ), T is the deformation temperature (◦ ), R is ε − ε 
c
the gas constant, ˇd , kd , a2 and m2 are the material constants, ln [− ln (1 − ln XXRD )] = ln ˇd + kd ln (19)
ε0.5
Q2 is the activation energy for 50% dynamic recrystallization
(J·mol−1 ).
Taking the logarithm of both sides of the expression of ε0.5 The relation between XDRX and  is ascertained by Eq. (15),
in the Eq. (16) gives then the relation between XDRX and ε can be ascertained.
Based on the data of XDRX and corresponding ε, the material
constants in Eq. (19) are resolved through the linear regression
ln ε0.5 = ln a2 + m2 ln˙ ε + Q2 /(RT) (17) analysis. Fig. 16 shows the linear fitting curves under the defor-
mation temperature of 1000 ◦ . The average slope of the curves
under the different deformation conditions is 1.88097, and the
Based on the data in the Table 5, the material constants
average intercept is 0.23995. So the values are: kd = 1.88097,
are resolved through the linear regression analysis. The linear
ln ˇd = 0.23995 (ˇd = 1.27119). The expression of XDRX is shown
fitting curves for each material constant are shown in Fig. 14.
in Eq. (20).
The values are: m2 = 0.1534, a2 = 0.0054, Q2 = 50.514 kJ/mol. The
expression of ε0.5 is shown in Eq. (18). Fig. 15 shows the linear  ε − ε 1.88097

fitting curve of the measured versus the predicted ε0.5 . The R XDRX = 1 − exp −1.27119
c 1
⁄3 1⁄3 1⁄3 1⁄3 1⁄3 (ε ≥ εc )
square is 0.979, which reveals that the expression of ε0.5 has a ε0.5
high accuracy. (20)

By combining Eqs. (18) and (20), the kinematic model of

ε0.5 = 0.0054ε̇0.1534 exp [5.0514 × 104 /(RT)] (18) dynamic recrystallization of as-cast 40CrNiMo alloy steel can
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940 1939

under the deformation temperatures at 900 ◦ –1100 ◦ , with

strain rates of 0.001 s−1 –0.1 s−1 . The recrystallization acti-
vation energy is 311.6 kJ/mol. The DRX kinetic model of
as-cast 40CrNiMo alloy steel can be expressed as follow.

 0.1577
εp = 0.005{ε̇ exp[3.116 × 105 /(RT)]}
εc = 0.5348εp

(3) Within the range of deformation conditions where

dynamic recrystallization occurs, as-cast 40CrNiMo occurs
incomplete dynamic recrystallization at low deformation
temperature and high strain rate. With the increase of
deformation temperature or the decrease of strain rate,
it makes the transition to complete dynamic recrystalliza-
tion. The recrystallized grains grow significantly at high
Fig. 17 – Correlation curve. deformation temperature and low strain rate. The DRX
kinematic model of as-cast 40CrNiMo alloy steel can be
expressed as follow.
be expressed as
⎧
⎧ X = 01⁄3 1⁄3 1⁄3 1⁄3 1⁄3 ⎪ XDRX = 0 (0 ≤
ε ≤ εc )
⎪ DRX (0 ≤ ε ≤ εc ) ⎪
⎨  ε − ε 1.88097

⎪
⎪
⎨  1.88097
XDRX = 1 − exp −1.271187
c
(ε ≥ εc )
ε − εc ⎪
⎪ ε0.5
⎪ XDRX = 1 − exp −1.271187 ⁄3 ⁄3 ⁄3 ⁄3 ⁄3 (ε ≥ εc )
1 1 1 1 1
⎩
⎪
⎪ ε0.5 ε0.5 = 0.0054ε̇0.1534 exp[50513.73/(RT)]
⎩
ε0.5 = 0.0054ε̇0.1534 exp [50513.73/(RT)]
(21)
Conflicts of interest
In order to verify the accuracy of kinematic model, the values
of XDRX are calculated by using Eqs. (15) and (21) respectively
The authors declare no conflicts of interest.
under random deformation conditions. The correlation curve
is shown in Fig. 17. The R square is 0.932, which reveals that
the kinematic model has a high accuracy. Acknowledgements

This work was supported by Key Research and Development

4. Conclusions Program of Hebei Province (Project No. 19251019D), and Youth
Fund Project of Colleges and Universities Science Research
(1) The true stress-strain curves of as-cast 40CrNiMo alloy
Program of Hebei Province (Project No. QN2019112).
steel shows that the true stress increases with the increase
of the strain rate, and it decreases with the increase of
references
deformation temperature. The activation energy of exper-
imental as-casting 40CrNiMo alloy steel is 332.1 kJ/mol,
that is less than as-rolled 40CrNiMo. The Arrhenius type
constitutive equation with Zener–Hollomon parameter for [1] Krüger L, Jentsch E, Brunke L, Keßler A, Wolf G, Lehnert T,
et al. Development of an innovative lightweight piston
as-cast 40CrNiMo alloy steel can be expressed as follow.
through process combination “casting - forging”. Procedia
Manufactur 2019;27:172–6.
⎧ 
⎪  1/5.147 [2] Li T, Wang S, Zhen K, Zhou H. Research progress on casting
⎪
⎪ 1 Z and forging combination forming technology for metal
⎪
⎪ = ln
⎪
⎪ 9.733 × 10 −3
5.544 × 10 12
materials. Mater Rev 2014;28:119–22, in Chinese.
⎪
⎨ 
1/2 
Z
 2/5.147
[3] Wang W, Ma R, Zhao J, Zhai RX. Co-simulation technology of
+ +1 casting-forging combination forming process for grain
⎪
⎪ 5.544 × 1012
⎪
⎪ 
distribution prediction. Mater Rev 2017;31:150–3, in Chinese.
⎪
⎪ [4] Kim HR, Seo MG, Bae WB. A study of the manufacturing of
⎪
⎪ 3.321 × 105
⎩ Z = ε̇ exp tie-rod ends with casting/forging process. J Mater Process
8.314T Technol 2002;125-126:471–6.
[5] Wang SI, Seo MK, Cho JR, Bae WB. A study on the
development of large aluminum flange using the
(2) With the increases of deformation temperature or the
casting/forging process. J Mater Process Technol
decreases of strain rate, the true stress-strain curves of as- 2002;130-131:294–8.
cast 40CrNiMo gradually transform from dynamic recovery [6] Chen HQ, Wang QC, Guo HG. Research on the casting-forging
type to dynamic recrystallization type. The true stress- precision forming process of alternator poles. J Mater
strain curves present the dynamic recrystallization type Process Technol 2002;129:330–2.
1940 j m a t e r r e s t e c h n o l . 2 0 2 0;9(2):1929–1940

[7] Song JL, Hu DJ, Chen HQ, Guo HG. Experimental study and [21] Lee WS, Yeh GW. The plastic deformation behaviour of AISI
numerical simulation of the casting-forging complex near 4340 alloy steel subjected to high temperature and high
net forming of alternator claw-pole. J Shanghai Jiaotong Univ strain rate loading conditions. J Mater Process Technol
2004;38:79–82. 1997;71:224–34.
[8] Lee K, Kwon YN, Lee S. Correlation of microstructure with [22] Aneta ŁS, Janusz K, Tomasz Ś, Jacek G. Optimization of the
mechanical properties and fracture toughness of A356 hot forging parameters for 4340 steel by processing maps. J
aluminum alloys fabricated by low-pressure-casting, Mater Res Technol 2019;8:3281–90.
rheo-casting, and casting–forging processes. Engng Fract [23] Chen X, Liao Q, Niu Y, Jia W, Le Q, Cheng C, et al. A
Mech 2008;75:4200–16. constitutive relation of AZ80 magnesium alloy during hot
[9] Lee K, Kwon YN, Lee S. Effects of eutectic silicon particles on deformation based on Arrhenius and Johnson-Cook model. J
tensile properties and fracture toughness of A356 aluminum Mater Res Technol 2019;8:1859–69.
alloys fabricated by low-pressure-casting, casting-forging, [24] Liu J, Wang X, Liu J, Liu Y, Li H, Wang C. Hot deformation and
and squeeze-casting processes. J Alloys Compd dynamic recrystallization behavior of Cu-3Ti-3Ni-0.5Si alloy.
2008;461:532–41. J Alloys Compd 2019;782:224–34.
[10] Jiang J, Wang Y, Li Y, Qu J, Shan W, Luo S. A double control [25] Yanagida A, Yanagimoto J. A novel approach to determine
forming technology combining die casting and forging for the kinetics for dynamic recrystallization by using the flow
the production of Mg alloy components with enhanced curve. J Mater Process Technol 2004;151:33–8.
properties. J Mater Process Technol 2012;212:1191–9. [26] Mirzadeh H, Parsa MH. Hot deformation and dynamic
[11] Jiang J, Wang Y, Chen G, Liu J, Li Y, Luo S. Comparison of recrystallization of NiTi intermetallic compound. J Alloys
mechanical properties and microstructure of AZ91D alloy Compd 2014;614:56–9.
motorcycle wheels formed by die casting and double control [27] Zhang J, Lu Z, Jia L, Xie H, Tao S, Wei X, et al. Critical
forming. Mater Des 2012;40:541–9. conditions and grain growth model for dynamic
[12] Jiang J, Wang Y, Li Y, Shan W, Luo S. Microstructure and recrystallization of Cu-Ni-Si alloy. Results Phys
mechanical properties of the motorcycle cylinder body of 2019;15:102593.
AM60B magnesium alloy formed by combining die casting [28] Liu S, Li D, Guo S. Critical conditions of dynamic
and forging. Mater Des 2012;37:202–10. recrystallization for B4 Cp /6061Al composite. Rare Metal Mat
[13] Jiang J, Wang Y, Qu J. Effect of process parameters on Eng 2017;46:1815–20.
microstructure and properties of AM50A magnesium alloy [29] Estrin Y, Tóth LS, Molinari A, Bréchet Y. A dislocation-based
parts formed by double control forming. Trans Nonferrous model for all hardening stages in large strain deformation.
Met Soc China 2014;24:321–33. Acta Mater 1998;46:5509–22.
[14] Wang S, Zhou N, Qi W, Zheng K. Microstructure and [30] Gottstein G, Frommert M, Goerdeler M, Schäfer N. Prediction
mechanical properties of A356 aluminum alloy wheels of the critical conditions for dynamic recrystallization in the
prepared by thixo-forging combined with a low superheat austenitic steel 800H. Mat Sci Eng A-Struct
casting process. Trans Nonferrous Met Soc China 2004;387-389:604–8.
2014;24:2214–9. [31] Tian S, Jiang H, Guo W, Zhang G, Zeng S. Hot deformation
[15] Zhou HT, Xu SX, Li WD, Wang SC, Peng Y. A study of and dynamic recrystallization behavior of TiAl-based alloy.
automobile brake bracket formed by casting-forging Intermetallics 2019;112:106521.
integrated forming technology. Mater Des 2015;67: [32] Wang W, Zhao J, Zhai RX, Ma R. Arrhenius-Type constitutive
285–92. model and dynamic recrystallization behavior of 20Cr2Ni4A
[16] Natasha AR, Jaharah AG, Che HCH. Machining-induced grain alloy carburizing steel. Steel Res Int 2017;88:1600196.
refinement of AISI 4340 alloy steel under dry and cryogenic [33] Xu Y, Birnbaum P, Pilz S, Zhuang X, Zhen Zhao, Kräusel V.
conditions. J Mater Res Technol 2019, Investigation of constitutive relationship and dynamic
http://dx.doi.org/10.1016/j.jmrt.2019.07.045. In Press. recrystallization behavior of 22MnB5 during hot
[17] Voorwald HJC, Bonora RG, Oliveira VMCA, Cioffi MOH. deformation. Results Phys 2019;14:102426.
Increasing fatigue resistance of AISI 4340 steel by nitrogen [34] Quan G, Shi R, Zhao J, Liu Q, Xiong W, Qiu H. Modeling of
plasma ion-implantation. Eng Fail Anal 2019;104:490–9. dynamic recrystallization volume fraction evolution for
[18] Sajadifar SV, Ketabchi M, Bemanizadeh B. Dynamic AlCu4SiMg alloy and its application in FEM. Trans
recrystallization behavior and hot deformation Nonferrous Met Soc China 2019;28:1138–51.
characteristics in 4340 steel. Metallurgist 2012;56:310–20. [35] Chen L, Sun W, Lin J, Zhao G, Wang G. Modelling of
[19] Sajadifar SV, Yapici GG, Ketabchi M, Bemanizadeh B. High constitutive relationship, dynamic recrystallization and
temperature deformation behavior of 4340 steel: activation grain size of 40Cr steel during hot deformation process.
energy calculation and modeling of flow response. J Iron Results Phys 2019;12:784–92.
Steel Res Int 2013;20:133–9. [36] Xu Y, Chen C, Zhang X, Dai H, Jia J, Bai Z. Dynamic
[20] Sanrutsadakorn A, Uthaisangsuk V, Suranuntchai S. recrystallization kinetics and microstructure evolution of an
Determination of initiation of dynamic recrystallization in AZ91D magnesium alloy during hot compression. Mater
AISI 4340 steel. Adv Mater Res 2014;893:381–6. Charact 2018;145:39–52.