Available via license: CC BY 4.0
Content may be subject to copyright.
Citation: Chen, Y.; Chen, R.; Yao, Y.;
Min, N.; Li, W.; Diao, A. The Effect of
Multi-Step Tempering and Partition
Heat Treatment on 25Cr2Ni3MoV
Steel’s Cryogenic Strength Properties.
Materials 2024,17, 518. https://
doi.org/10.3390/ma17020518
Academic Editors: Alexander
Yu Churyumov and Thomas
Niendorf
Received: 29 November 2023
Revised: 27 December 2023
Accepted: 15 January 2024
Published: 21 January 2024
Copyright: © 2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
materials
Article
The Effect of Multi-Step Tempering and Partition Heat Treatment
on 25Cr2Ni3MoV Steel’s Cryogenic Strength Properties
Ye Chen 1,2, Ran Chen 3, Yanchen Yao 4, Na Min 5, Wei Li 1,2,* and Anna Diao 4 ,*
1Shanghai Key Laboratory of Material Laser Processing and Modification, Shanghai Jiao Tong University,
Shanghai 200240, China
2
Institute of Advanced Steels and Materials, School of Materials Science and Engineering, Shanghai Jiao Tong
University, Shanghai 200240, China
3
The State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, China
4Shanghai Marine Diesel Engine Research Institute, Shanghai 201108, China
5Key Laboratory for Microstructures, Shanghai University, Shanghai 200444, China
*Correspondence: weilee@sjtu.edu.cn (W.L.); 18916625493@163.com (A.D.); Tel.: +86-021-54745420 (W.L.)
Abstract: In this study, the refinement of two microstructures was controlled in medium carbon
25Cr2Ni3MoV steel via multi-step tempering and partition (MTP) to achieve high cryogenic strength–
ductility combinations. Microstructure evolution, the distribution of stress concentration, and mi-
crocrack formation and propagation during cryogenic Charpy impact testing were investigated.
Compared with their performance in the quenching and tempering states (QT), the MTP steels
showed a significant improvement in yield strength (1300 MPa), total elongation (25%), and impact
toughness (>25 J) at liquid nitrogen temperature (LNT). The strengthening contributions mainly
originated from the high dislocation density and refinement cementite (size: 70 nm) in the martensite
lath (width: 1.5
µ
m) introduced by refined reversed austenite and its latter decomposition. The
instrumented Charpy impact results indicated that cracks nucleated in the primary austenite grain
(PAG) boundary for two steels due to the strain concentration band preferring to appear near PAGs,
while cracks in the QT and MTP samples propagated along the PAGs and high-angle grain boundary
(HAGB), respectively. The crystallized plasticity finite element simulation revealed that the PAG
boundary with cementite precipitates of large size (>200 nm) was less able to dissipate crack propaga-
tion energy than the HAGBs by continuously forming a high strain concentration area, thus leading
to the low-impact toughness of the QT steel.
Keywords: 25Cr2Ni3MoV steel; multi-step tempering and partition; cryogenic mechanical
property; CPFEM
1. Introduction
With the development of exploitation technology and the increasing demand for
clean fuel, high-performance compressors and engines are emerging to adapt to lower-
temperature environments, such as applications in liquefied natural gas fields [
1
–
4
]. It is
necessary for the high-speed shaft steel used in such equipment to possess stricter cryogenic
mechanical properties. Medium carbon Cr-Ni-Mo-V alloy steel, due to its high toughness,
abrasive resistance, and hardness, is widely used to manufacture high-speed shafts [
5
–
7
].
However, its toughness drops as the temperature decreases [
8
], and the ductile–brittle
transition observed is in line with the temperature-dependent moving ability of screw
dislocations in the BCC structure of metals [9].
The method used to improve the toughness of single-phase BCC steel during temper-
ing comprises two aspects: reducing the effective martensite grain size [
10
] and controlling
the shape and size of cementite (decreasing the local stress concentration) [
11
]. During
tempering, martensite microstructures mainly manifest as the low-angle grain boundary
(LAGB), vanishing in ferrite, and the high-angle grain boundary (HAGB), remaining due
Materials 2024,17, 518. https://doi.org/10.3390/ma17020518 https://www.mdpi.com/journal/materials
Materials 2024,17, 518 2 of 16
to the pinning of precipitates [
12
]. Reducing the tempering temperature and shortening the
tempering time can effectively help to maintain the grain size and strength of quenching
martensite. However, controlling the growth of carbide particles requires an element diffu-
sion balance between the matrix and carbide, and the interface can increase the diffusion
rate and accelerate abnormal growth. Although elements with large atomic numbers in
cementite (M
3
C, M = Mo and V) can effectively block the interface migration rate from
the cementite to the matrix by the solute drag effect [
13
], the coarsening phenomenon still
occurs at high-temperature tempering [14], especially cementite spheroidization.
The concentrated distribution of carbide neighboring the grain boundary also has
a negative effect on toughness because it promotes potential microvoid nucleation [
15
]
and stress concentration, according to the simulation results of damask software [
16
]. The
inhomogeneous strain distribution at interfaces is the origin of ductility reduction [
17
].
Therefore, the competition between the controlled recovery of martensite to maintain its
strength and relatively uniform small-sized and round-shaped carbides during tempering
restrict the comprehensive strength/toughness properties of these types of steel. Sachin
Kumar [
18
] utilized cementite containing austenite-stabilizing elements as a nucleation
point for martensite reversion transformation, finally forming reversed austenite containing
bulk cementite after cooling at room temperature. The cryogenic stability of reversed
austenite will decrease when the corresponding total volume fraction or the volume of
the single austenite grain exceeds the critical value due to dilution of the austenite stable
element [
19
]. Recently, G.G. Ribamar [
20
] induced blocky austenite to decompose while
tempering at 500
◦
C to obtain small dispersed carbides and film ferrite. On the other
hand, E. Tkachev [
21
] improved the cryogenic impact toughness by promoting cementite
spheroidization (η-Fe2C (rod-like)→Fe3C (round-like)) after tempering at 500 ◦C.
Dislocation density, grain size, and nanoprecipitation characteristics all contribute
to the strength and toughness of materials. The present study aimed to elucidate the
effect of multi-tempering and partition (MTP) on both the refinement of martensite lath
and the dispersion of carbide, as well as the corresponding influence on the cryogenic
properties of 25Cr2Ni3MoV steel. Moreover, the MTP process was designed to activate
the martensite reversion transformation by nucleating carbides. The decomposition of
thin-film reverted austenite was hypothesized to refine the cementite size along the PAG
boundary. Finally, the strengthening contribution and crack propagation behavior were
addressed based on the multiple mechanisms and strain distribution correlated with the
result of the crystallized plasticity finite element model (CPFEM).
2. Experimental Procedures
The material was provided by the China State Shipbuilding Corporation (CSSC, Shang-
hai, China), and this material is widely used in shafts (Figure 1a). It is manufactured by
combining annealing and multi-pass forging as an initial state. The initial-state microstruc-
ture is single-phase martensite (Figure 1(c2)), and the lath size distribution follows a
Gaussian distribution with a mean value of around 2.1
µ
m (Figure 1(c3)). The shaft, in this
case, is medium carbon 25Cr2Ni3MoV steel, and the corresponding chemical composition
is shown in Table 1.
Table 1. Element content of 25Cr2Ni3MoV steel.
Element (wt.%) C Ni Cr Mo Mn Si V Fe
Content 0.25 2.8–3.1 1.4–1.6 0.51 0.23 0.21 0.13
Balanced
It is well-known that the starting temperature (Ac1) and completion temperature (Ac3)
for austenitization can be extracted from the tuning spot in the thermal expansion curve
(Figure 1(c4)). According to the results above, two heat treatment processes were designed
for the shaft. (i) QT: The shaft was austenitized at 900
◦
C for 4 h, quenched into water, and
then tempered for 10 h at 600
◦
C. (ii) MTP: Firstly, the shaft was re-austenitized at 900
◦
C
Materials 2024,17, 518 3 of 16
for 1 h and quenched into water; secondly, the shaft was multi-tempered and partitioned
by S1, S2, and S3 from 500 ◦C to 700 ◦C for 5 h in total.
Materials 2024, 17, x FOR PEER REVIEW 3 of 16
Figure 1b shows the positions (where R is the radius of the shaft) from which the
tensile and impact test samples were extracted and the corresponding three dimensions
of the tested samples. Dog bone-shaped specimens with gauge dimensions of diameters
of 5 mm, lengths of 25 mm, and V-notch samples of 55 mm × 10 mm × 10 mm were cut
from the prepared steels along the axial direction using an electron discharge machine
(APL250, Yokohama, Japan).
Uniaxial tensile with a strain rate of 0.5 mm/min (using an INSTRON 5565 tensile
testing machine) was performed to measure the mechanical properties at room and liquid
nitrogen temperature (−196 °C). The Charpy impact test (Zwick/Roell, 450 J impact tester,
Ulm, Germany) was implemented at temperatures ranging from −196 °C to −140 °C. Rock-
well hardness measurements were acquired by averaging ten measurements at room tem-
perature. The microstructure was characterized using a JSM-7001F scanning electron mi-
croscope (SEM) after mechanical polishing and etching (4% Nital). Electron backscaering
diffraction (EBSD) with an orientation imaging microscope system was employed on the
SEM (Tescan, Mira, Brno, Czech Republic) to investigate the effective grain size and grain
boundary characteristics after electropolishing using an electrolyte consisting of 10 vol.%
perchloric acid and 90 vol.% alcohol. Aztech 2.1 software was used to process the EBSD
data and images. The TEM samples were prepared by a twin jet electro-polisher at 26 V
and −20 °C in an 8% perchloric and 92% alcohol solution and observed by a transmission
electron microscope (JEOL, JEM 2100F, Tokyo, Japan) at 200 kV. X-ray diffraction analysis
(Zeiss Crossbeam 550, Jena, Germany) carried out using an X-ray diffractometer with Cu-
Kα radiation determined the phase components of the specimens. Next, 2θ angles from
40° to 100° were scanned at 40 kV with a step of 0.02°. The phase transformation behavior
was certified by thermal dilatometry (DIL805A, TA Instruments, New Castle, DE, USA)
using the tangent method, and corresponding phase content variation was analyzed using
the saturation magnetization (SM) method through the use of a quantum-designed phys-
ical property measurement system (PPMS-9T (EC-II)).
Table 1. Element content of 25Cr2Ni3MoV steel.
Element (wt.%) C Ni Cr Mo Mn Si V Fe
Content 0.25 2.8–3.1 1.4–1.6 0.51 0.23 0.21 0.13 Balanced
Figure 1. Materials and experimental details for the 25Cr2Ni3MoV steel. (a) The gas compressor
schematic diagram for liquefied natural gas; (b) the high-speed shaft was fabricated through the hot
forging of the corresponding sample preparation of the tensile and impact tests (units in mm); and
(c) the primary microstructure of the experimental shaft after hot forging, including morphology (c1),
phase composition (c2), grain size distribution (c3), and phase transformation point (c4) (Ac1: the
phase transformation temperature from ferrite to austenite; Ac3: the full austenite region).
Figure 1b shows the positions (where R is the radius of the shaft) from which the
tensile and impact test samples were extracted and the corresponding three dimensions of
the tested samples. Dog bone-shaped specimens with gauge dimensions of diameters of
5 mm, lengths of 25 mm, and V-notch samples of 55 mm
×
10 mm
×
10 mm were cut from
the prepared steels along the axial direction using an electron discharge machine (APL250,
Yokohama, Japan).
Uniaxial tensile with a strain rate of 0.5 mm/min (using an INSTRON 5565 tensile
testing machine) was performed to measure the mechanical properties at room and liquid
nitrogen temperature (
−
196
◦
C). The Charpy impact test (Zwick/Roell, 450 J impact tester,
Ulm, Germany) was implemented at temperatures ranging from
−
196
◦
C to
−
140
◦
C.
Rockwell hardness measurements were acquired by averaging ten measurements at room
temperature. The microstructure was characterized using a JSM-7001F scanning electron
microscope (SEM) after mechanical polishing and etching (4% Nital). Electron backscat-
tering diffraction (EBSD) with an orientation imaging microscope system was employed
on the SEM (Tescan, Mira, Brno, Czech Republic) to investigate the effective grain size
and grain boundary characteristics after electropolishing using an electrolyte consisting of
10 vol.% perchloric acid and 90 vol.% alcohol. Aztech 2.1 software was used to process the
EBSD data and images. The TEM samples were prepared by a twin jet electro-polisher at
26 V and
−
20
◦
C in an 8% perchloric and 92% alcohol solution and observed by a trans-
mission electron microscope (JEOL, JEM 2100F, Tokyo, Japan) at 200 kV. X-ray diffraction
analysis (Zeiss Crossbeam 550, Jena, Germany) carried out using an X-ray diffractometer
with Cu-K
α
radiation determined the phase components of the specimens. Next, 2
θ
angles
from 40
◦
to 100
◦
were scanned at 40 kV with a step of 0.02
◦
. The phase transformation
behavior was certified by thermal dilatometry (DIL805A, TA Instruments, New Castle, DE,
Materials 2024,17, 518 4 of 16
USA) using the tangent method, and corresponding phase content variation was analyzed
using the saturation magnetization (SM) method through the use of a quantum-designed
physical property measurement system (PPMS-9T (EC-II)).
3. Simulation Procedures
A crystal plasticity framework was used to simulate the deformation of the martensite
lath grains, corresponding to displacement-controlled uniaxial tension at a quasi-static
strain rate of 10
−3
s
−1
and room temperature (T= 23
◦
C) imposed on the generated
representative volume elements (RVEs).
3.1. Deformation Kinematics
Mechanical response during uniaxial tension was simulated through the use of the
finite element software Abaqus 2017, with a user material (UMAT) subroutine programmed
based on the continuum crystal plasticity theory. The crystal lattice structures of martensite
and cementite were body-centered cubic (BCC) structures and perfectly elastic states,
respectively. The total deformation gradient tensor Fcan be decomposed into a component
F
e
, representing elastic stretching and rigid body rotation, and a plastic component F
p
(
F=Fe·Fp
). The plastic velocity gradient L
p
can be defined as the multiplication between
the time derivative .
FPand inverse matrix F−1
P(Lp=.
FP·F−1
P).
3.2. Slip Deformation
The plastic velocity gradient
Lp
is related to the slipping rate
.
γα
of the corresponding
slip system by: .
FP·F−1
P=∑
α
.
γα·S0,α⊗m0,α(1)
where the sum ranges over all activated slip systems and unit vectors
s0,a
and
m0,a
are the
slip direction and normal to slip plane in the reference configuration. Plastic deformation
comprises the motion of dislocations along slip planes; this study mainly considers the
motion of 24 slip systems comprising both the primary {110}<111> slip system and the
secondary {211}<111> slip system for martensite in the current sample.
The plastic slip rate .
γαfollows the classic flow rule [22]:
.
γα=.
γ0sgn(τα
s)
τα
gα
n
(2)
where nis the rate-sensitivity exponent of slip,
.
γ0
is the reference slip rate, and
τα
is the
resolved shear stress.
gα
is the current slip resistance, whose initial value is usually the
critical resolved shear stress (CRSS) of the αslip system, as follows:
.
gα=∑
β
haβ
.
γβ(3)
where
.
γβ
is the plastic slip rate of the
β
slip system and
hαβ
is the slip hardening matrix,
which satisfies the following equation:
haβ=qhaa =−qh0sech2
h0γa
g∞−τ0
(4)
where
h0
is the initial hardening modulus,
τ0
is the CRSS of the slip system and
g∞
is
its saturation value, and qis the ratio of the latent hardening modulus h
αβ
(
α=β
) to the
self-hardening modulus h
αα
. Parameter qis set to 1.4 [
22
] for a non-coplanar slip system
because the BCC structure has several slip systems, including multiple slip surfaces.
Materials 2024,17, 518 5 of 16
3.3. Polycrystalline Geometric Based on EBSD
The grain shapes and crystallographic orientations of the RVE microstructure were
generated from the EBSD measurements using MTEX 5.10 and Matlab 2022 software. A
two-dimensional geometric model was meshed with a total of 506,944 elements. The mesh
size was 0.07 µm, the same as the EBSD step size.
In this paper, the boundary conditions employed for uniaxial tensile simulation were
consistent and are listed below. The node on the left was constrained, as there was no
displacement in the X direction (u
x
= 0), and the node on the right boundary was subjected
to displacement, corresponding to the max. uniform strain from the experimental result
(inset in Figure 10(b1)).
Based on preliminary research, the martensite in low-carbon steel produced after
high-temperature tempering treatment is similar to ferrite, leading to the parameter of
tempering martensite being closer to ferrite [
23
]. Thus, this study adopts the parameters of
a ferrite matrix as the foundation [
17
]. Due to the higher dislocation density and the refined
cementite in the MTP sample, the kinetics of dislocation slip were slightly modulated [
24
].
The crystal material parameters are primarily included, as shown in Table 2.
Table 2. Parameters for the crystal plasticity of tempering martensite.
Crystal Parameters QT MTP
{112}<111> {110}<111> {112}<111> {110}<111>
G/GPa
C11 231 231 231 231
C12 134 134 134 134
C44 116 116 116 116
h0/MPa 1100 1100 1300 1300
τs/MPa 300 300 320 320
τ0/MPa 250 250 270 270
q 1 1 1 1
n 20 20 20 20
.
γ0/s−10.001 0.001 0.001 0.001
Elastic modulus (C11, C12, and C44), initial hardening modulus (h0), saturation stress (
τs
), yield stress (
τ0
),
hardening constant (q), strain rate sensitivity coefficient (n), and the referring shear strain rate ( .
γ0).
4. Results
4.1. Microstructure Characterization
Figure 2shows the morphologies of the two types of steel observed using the SEM.
The main microstructural constituents are tempering martensite and carbide, where the
dark matrix is tempering martensite and the white particles are carbide. The coarse carbide
particles were mainly located at the prior austenite grain (PAG) boundary (Figure 2(a2)) by
the spheroidization growth; inversely, the transgranular carbide was the ellipse carbide
with a high aspect ratio (6–8). A similar carbide distribution was obtained in the warm-
rolling high-carbon steel [
12
]. On the contrary, the interface of martensite laths in the MTP
sample was sharper; meanwhile, no granular carbide was obtained in the PAG boundary,
and transgranular carbide was dispersed along the inter-lath with a low aspect ratio (1–3)
(Figure 2(b2)).
Both carbides in the two samples were certified as
θ
-cementite (M
3
C where M: ran-
domly dispersed Fe, Cr, Ni, and Mo) through selected area electronic diffraction (SAED) and
scanning transmission electron microscopy/energy-dispersive spectrometry (STEM/EDS),
as shown in Figure 3. The average size of the coarse
θ
-cementite particles along the PAG
boundary is more than 300 nm in the QT specimen, as indicated by the white arrow in
Figure 3a. The transgranular
θ
-cementite in the MTP specimen is only half the size of the
carbide in the QT, with a grain size of 150 nm, as shown in Figure 3c. The fine
θ
-cementite
in the MTP sample has higher Ni content and lower Cr and Mo content (Figure 3b,d). The
coarse cementite often accompanied the enrichment of Cr and Mo due to the enhancement
Materials 2024,17, 518 6 of 16
of mobility in cementite when partitioned at 600
◦
C, according to the local equilibrium (LE)
deduced by Y.X. Wu [14].
Materials 2024, 17, x FOR PEER REVIEW 6 of 16
Figure 2. SEM image of 25Cr2Ni3MoV steel with the QT (a1,a2) and MTP (b1,b2) treatment.
Both carbides in the two samples were certified as θ-cementite (M3C where M: ran-
domly dispersed Fe, Cr, Ni, and Mo) through selected area electronic diffraction (SAED)
and scanning transmission electron microscopy/energy-dispersive spectrometry
(STEM/EDS), as shown in Figure 3. The average size of the coarse θ-cementite particles
along the PAG boundary is more than 300 nm in the QT specimen, as indicated by the
white arrow in Figure 3a. The transgranular θ-cementite in the MTP specimen is only half
the size of the carbide in the QT, with a grain size of 150 nm, as shown in Figure 3c. The
fine θ-cementite in the MTP sample has higher Ni content and lower Cr and Mo content
(Figure 3b,d). The coarse cementite often accompanied the enrichment of Cr and Mo due
to the enhancement of mobility in cementite when partitioned at 600 °C, according to the
local equilibrium (LE) deduced by Y.X. Wu [14].
Figure 2. SEM image of 25Cr2Ni3MoV steel with the QT (a1,a2) and MTP (b1,b2) treatment.
Materials 2024, 17, x FOR PEER REVIEW 6 of 16
Figure 2. SEM image of 25Cr2Ni3MoV steel with the QT (a1,a2) and MTP (b1,b2) treatment.
Both carbides in the two samples were certified as θ-cementite (M3C where M: ran-
domly dispersed Fe, Cr, Ni, and Mo) through selected area electronic diffraction (SAED)
and scanning transmission electron microscopy/energy-dispersive spectrometry
(STEM/EDS), as shown in Figure 3. The average size of the coarse θ-cementite particles
along the PAG boundary is more than 300 nm in the QT specimen, as indicated by the
white arrow in Figure 3a. The transgranular θ-cementite in the MTP specimen is only half
the size of the carbide in the QT, with a grain size of 150 nm, as shown in Figure 3c. The
fine θ-cementite in the MTP sample has higher Ni content and lower Cr and Mo content
(Figure 3b,d). The coarse cementite often accompanied the enrichment of Cr and Mo due
to the enhancement of mobility in cementite when partitioned at 600 °C, according to the
local equilibrium (LE) deduced by Y.X. Wu [14].
Figure 3. Precipitation of cementites in the martensite matrix following QT and MTP treatment.
(a) BF image of the coarse
θ
-cementite in the QT samples; their average size is approximately 300 nm.
(b) The element distribution mapping of coarse
θ
-cementite in the QT samples, with corresponding
local EDS results gained through the use of point analysis. (c) BF image of the refined
θ
-cementite
in the MTP samples; their average width is approximately 150 nm. (d) The element distribution
mapping of the refined
θ
-cementite in the MTP samples, with corresponding local EDS results gained
through the use of point analysis.
Materials 2024,17, 518 7 of 16
The microstructure feature of the two types of steel was determined using the EBSD
method (Figure 4). The average sizes of prior austenite grains (PAGs) and the width of
martensite lath (12
µ
m and 1
µ
m) in the MTP sample are smaller than the QT sample
(20
µ
m and 4
µ
m), according to the results shown in the inverse pole figure (IPF) map
(Figure 4(a1,b1)). The total interface density (ID is the ratio between the whole length of
the interface and total area), attributed to martensite transformation, is 0.17
µ
m
−1
and
0.29
µ
m
−1
for the QT and MTP samples, respectively, and the corresponding proportion of
LAGB attributed to dislocation recovery is 22.6% and 27.5%, as shown in Figure 4(a2,b2). It
should be noted that the value of kernel average misorientation (KAM) is higher neigh-
boring the LAGB (Figure 4(a3,b3)), which is related to misorientation induced by greater
dislocation densities and more pronounced strain concentrations [
25
]. Therefore, the
martensite laths with a greater dislocation density were obtained in the MTP sample.
Materials 2024, 17, x FOR PEER REVIEW 7 of 16
Figure 3. Precipitation of cementites in the martensite matrix following QT and MTP treatment. (a)
BF image of the coarse θ-cementite in the QT samples; their average size is approximately 300 nm.
(b) The element distribution mapping of coarse θ-cementite in the QT samples, with corresponding
local EDS results gained through the use of point analysis. (c) BF image of the refined θ-cementite
in the MTP samples; their average width is approximately 150 nm. (d) The element distribution
mapping of the refined θ-cementite in the MTP samples, with corresponding local EDS results
gained through the use of point analysis.
The microstructure feature of the two types of steel was determined using the EBSD
method (Figure 4). The average sizes of prior austenite grains (PAGs) and the width of
martensite lath (12 µm and 1 µm) in the MTP sample are smaller than the QT sample (20
µm and 4 µm), according to the results shown in the inverse pole figure (IPF) map (Figure
4(a1,b1)). The total interface density (ID is the ratio between the whole length of the inter-
face and total area), aributed to martensite transformation, is 0.17 µm−1 and 0.29 µm−1 for
the QT and MTP samples, respectively, and the corresponding proportion of LAGB at-
tributed to dislocation recovery is 22.6% and 27.5%, as shown in Figure 4(a2,b2). It should
be noted that the value of kernel average misorientation (KAM) is higher neighboring the
LAGB (Figure 4(a3,b3)), which is related to misorientation induced by greater dislocation
densities and more pronounced strain concentrations [25]. Therefore, the martensite laths
with a greater dislocation density were obtained in the MTP sample.
Figure 4. Inverse pole figure (IPF) image with the reconstructed prior austenite grain (PAG) bound-
ary, the KAM image, and the grain boundary of 25Cr2Ni3MoV steel with the QT (a1–a3) and MTP
(b1–b3) samples (LAGB: low-angle grain boundary (2°–15°), HAGB: high-angle grain boundary
(>15°), and ID: interface density (µm−1)).
4.2. Mechanical Properties
Figure 5a shows the engineering stress–strain curves tested at −196 °C. The compre-
hensive mechanics increased with the decreased grains and precipitated phase size due
to grain refinement strengthening and the low-stress concentration. The MTP sample
achieved a highest yield strength of 1336 MPa, an ultimate tensile strength of 1367 MPa,
and an elongation of 29% (Table 3). Simultaneously, the impact toughness of MTP steel
(25.4 J, −196 °C) is higher than the QT sample (7 J, −196 °C) at various temperatures (Figure
Figure 4. Inverse pole figure (IPF) image with the reconstructed prior austenite grain (PAG) boundary,
the KAM image, and the grain boundary of 25Cr2Ni3MoV steel with the QT (a1–a3) and MTP
(b1–b3) samples (LAGB: low-angle grain boundary (2
◦
–15
◦
), HAGB: high-angle grain boundary
(>15◦), and ID: interface density (µm−1)).
4.2. Mechanical Properties
Figure 5a shows the engineering stress–strain curves tested at
−
196
◦
C. The com-
prehensive mechanics increased with the decreased grains and precipitated phase size
due to grain refinement strengthening and the low-stress concentration. The MTP sample
achieved a highest yield strength of 1336 MPa, an ultimate tensile strength of 1367 MPa,
and an elongation of 29% (Table 3). Simultaneously, the impact toughness of MTP steel
(25.4 J,
−
196
◦
C) is higher than the QT sample (7 J,
−
196
◦
C) at various temperatures
(Figure 2b). The Rockwell hardness for the MTP and QT samples are depicted in Figure 5c,
and they show a similar strength change trend. Accordingly, dual matrix and precipi-
tated phase refinement markedly affect the low-temperature strength without sacrificing
impact ductility.
Materials 2024,17, 518 8 of 16
Materials 2024, 17, x FOR PEER REVIEW 8 of 16
2b). The Rockwell hardness for the MTP and QT samples are depicted in Figure 5c, and
they show a similar strength change trend. Accordingly, dual matrix and precipitated
phase refinement markedly affect the low-temperature strength without sacrificing im-
pact ductility.
Table 3. The mechanical properties of the two types of steel.
Sample YS (MPa) TS (MPa) ETF (%) AK (J) H (HRC)
−196 °C −196 °C −196 °C −150 °C −160 °C −170 °C −180 °C −190 °C −196 °C
QT 1226.6 ± 10.1 1283.2 ± 11.4 18.1 ± 1.1 34.1 ± 4.2 26.6 ± 3.3 19.4 ± 2.5 15.6 ± 3.1 9.1 ± 1.7 6.3 ± 1.5 304 ± 1.7
MTP 1332.1 ± 15.3 1367.7 ± 16.1 25.4 ± 0.7 69.3 ± 2.3 62.9 ± 2.7 50.8 ± 2.1 37.9 ± 1.6 30.6 ± 1.7 24.5 ± 1.1 325 ± 2.5
TS, tensile strength; YS, yield strength; ETF, elongation to fracture; H: hardness; AK: impact tough-
ness.
Figure 5. Mechanical properties of 25Cr2Ni3MoV steel: (a) tensile curves at −196 °C; (b) impact
properties at different temperatures; and (c) hardness.
4.3. Deformation Microstructure
Figure 6 shows the EBSD image after the tensile fractured state of the two samples at
−196 °C. The IPF map (Figure 6(a1,b1)) demonstrates that no deformation texture of mar-
tensite was formed in the two steels during tensile deformation at liquid nitrogen temper-
ature (LNT). By comparing the KAM maps (Figure 6(a2,b2)), two results can be obtained:
(i) a strain hardening behavior originating from dislocation pileup and multiplication was
found in both samples and (ii) the average KAM value of the MTP sample is slightly
higher than the QT sample (Figure 6c). To macroscopically estimate the dislocation incre-
ment state, the XRD method was applied to calculate the dislocation density using the
Williamson–Hall (WH) formula [26], as follows:
βi
= 1
Dv+2εSi (5)
ρ = 3
√
2π
〈
ε
〉
12
⁄
Dvb (6)
βi
= βi
0cos(θi)/λ represents the integral breadth of the diffraction peak in the recipro-
cal space, βi
0 is the integral breadth of the diffraction peak, 𝜃 is the Bragg angle, 𝜆 is the
wavelength (1.54056 Å), and Si = 2sin(θi)/λ represents the length of the diffraction vector
for the diffraction peak.
The result in Figure 6d exhibits the XRD paern of undeformed and fractured sam-
ples in the two samples, which includes corresponding diffraction peaks, referring to the
(110), (200), and (211) crystallographic planes of the bcc phase. An obvious diffraction
peak broadening phenomenon appears in both types of steel. According to the diffraction
peak broadening value obtained in the inset, the values of dislocation density in the matrix
were 2.38 × 1014 m−2 (QT) and 3.91 × 1014 m−2 (MTP). The values of dislocation density of
Figure 5. Mechanical properties of 25Cr2Ni3MoV steel: (a) tensile curves at
−
196
◦
C; (b) impact
properties at different temperatures; and (c) hardness.
Table 3. The mechanical properties of the two types of steel.
Sample
YS (MPa) TS (MPa) ETF (%) AK(J) H (HRC)
−196 ◦C−196 ◦C−196 ◦C−150 ◦C−160 ◦C−170 ◦C−180 ◦C−190 ◦C−196 ◦C
QT
1226.6
±
10.1 1283.2
±
11.4
18.1 ±1.1 34.1 ±4.2 26.6 ±3.3
19.4
±
2.5 15.6
±
3.1
9.1 ±1.7 6.3 ±1.5 304 ±1.7
MTP
1332.1
±
15.3 1367.7
±
16.1
25.4 ±0.7 69.3 ±2.3 62.9 ±2.7
50.8
±
2.1 37.9
±
1.6 30.6
±
1.7 24.5
±
1.1
325 ±2.5
TS, tensile strength; YS, yield strength; ETF, elongation to fracture; H: hardness; AK: impact toughness.
4.3. Deformation Microstructure
Figure 6shows the EBSD image after the tensile fractured state of the two samples
at
−
196
◦
C. The IPF map (Figure 6(a1,b1)) demonstrates that no deformation texture of
martensite was formed in the two steels during tensile deformation at liquid nitrogen
temperature (LNT). By comparing the KAM maps (Figure 6(a2,b2)), two results can be
obtained: (i) a strain hardening behavior originating from dislocation pileup and multipli-
cation was found in both samples and (ii) the average KAM value of the MTP sample is
slightly higher than the QT sample (Figure 6c). To macroscopically estimate the dislocation
increment state, the XRD method was applied to calculate the dislocation density using the
Williamson–Hall (WH) formula [26], as follows:
βi=1
Dv
+2εSi(5)
ρ=3√2π.
ε1/2
Dvb(6)
βi=β0
icos(θi)/λ
represents the integral breadth of the diffraction peak in the recipro-
cal space,
β0
i
is the integral breadth of the diffraction peak,
θi
is the Bragg angle,
λ
is the
wavelength (1.54056 Å), and
Si=2 sin(θi)/λ
represents the length of the diffraction vector
for the diffraction peak.
The result in Figure 6d exhibits the XRD pattern of undeformed and fractured samples
in the two samples, which includes corresponding diffraction peaks, referring to the (110),
(200), and (211) crystallographic planes of the bcc phase. An obvious diffraction peak
broadening phenomenon appears in both types of steel. According to the diffraction peak
broadening value obtained in the inset, the values of dislocation density in the matrix were
2.38
×
10
14
m
−2
(QT) and 3.91
×
10
14
m
−2
(MTP). The values of dislocation density of
deformation were 4.87
×
10
14
m
−2
(QT) and 5.92
×
10
14
m
−2
(MTP). The result of the XRD
analysis shows almost the same trend as the value of KAM.
Figure 7demonstrates the fractured morphology and crack propagation path during
the impact test at
−
196
◦
C for the two types of steel. A typical brittleness fracture with
many cleavage river patterns was obtained in the two samples. Interestingly, the local
impact fracture of the MTP samples has a mixed feature combining quasi-cleavage and
Materials 2024,17, 518 9 of 16
dimples along the tear ridge, as shown by the white arrow (Figure 7b), but no dimples
were discovered in the QT fracture (Figure 7a). Although the tortuous propagation paths
of the main crack occurred in both types of steel, the secondary crack propagation in
the QT sample extended along the PAG boundary with a length of less than 120
µ
m
(Figure 7(c1,c2)) compared to the HAGB in the MTP sample with a length of more than
160
µ
m (Figure 7(d1,d2)). Thus, compared to MTP steel, QT steels have a recognizable
intergranular fracture face along the PAG boundary.
Materials 2024, 17, x FOR PEER REVIEW 9 of 16
deformation were 4.87 × 1014 m−2 (QT) and 5.92 × 1014 m−2 (MTP). The result of the XRD
analysis shows almost the same trend as the value of KAM.
Figure 6. The microstructure of 25Cr2Ni3MoV steel after tensile fracture: (a1,b1) IPF image; (a2,b2)
KAM distribution image; (c) KAM values of the QT and MTP samples in undeformed and fractured
states; and (d) the XRD paern of the QT and MTP samples in undeformed and fractured states; the
inset in (d) shows the apparent diffraction peak broadening phenomenon owing to dislocation for-
mation.
Figure 7 demonstrates the fractured morphology and crack propagation path during
the impact test at −196 °C for the two types of steel. A typical brileness fracture with
many cleavage river paerns was obtained in the two samples. Interestingly, the local im-
pact fracture of the MTP samples has a mixed feature combining quasi-cleavage and dim-
ples along the tear ridge, as shown by the white arrow (Figure 7b), but no dimples were
discovered in the QT fracture (Figure 7a). Although the tortuous propagation paths of the
main crack occurred in both types of steel, the secondary crack propagation in the QT
sample extended along the PAG boundary with a length of less than 120 µm (Figure
7(c1,c2)) compared to the HAGB in the MTP sample with a length of more than 160 µm
(Figure 7(d1,d2)). Thus, compared to MTP steel, QT steels have a recognizable intergran-
ular fracture face along the PAG boundary.
Figure 6. The microstructure of 25Cr2Ni3MoV steel after tensile fracture: (a1,b1) IPF image; (a2,b2) KAM
distribution image; (c) KAM values of the QT and MTP samples in undeformed and fractured states;
and (d) the XRD pattern of theQT and MTP samples in undeformed and fractured states; the inset in
(d) shows the apparent diffraction peak broadening phenomenon owing to dislocation formation.
Materials 2024, 17, x FOR PEER REVIEW 10 of 16
Figure 7. Fractography morphology of the QT and MTP samples after the impact test: (a,b) the frac-
ture morphology; (c1,d1) the local profile fractography; and (c2,d2) the crack propagation trace de-
termined by EBSD (the red area is the positions for the SEM in the Charpy samples).
5. Discussion
5.1. Microstructure Evolution during MTP Heat Treatment
The DIL was scheduled to capture the phase transformation behavior during the
MTP process (Figure 8(a1,a2,a3)). No martensite transformation occurs at each step (S1,
S2, and S3), as the temperature falls to room temperature when the cooling rate is 50 °C/s.
Furthermore, the corresponding change in length (ΔL/L) at the isothermal stage of every
step shows a decreasing trend with the extension of isothermal time. SM was measured,
as shown in Figure 8(b1,b2,b3), to analyze phase transformation and the carbide precipi-
tation phenomenon further.
Comparing Figures 1b and 9(a1), the decrease in length in S1 (Figure 8(a1)) with a
stable SM value (Figure 8(b1)) was aributed to the lamellar cementite precipitation
(width: 10 nm; length: 70 nm), which was generated from the interaction between dislo-
cation in the quenching martensite and carbide segment.
The decrease in length in S2 (Figure 8(a2)) with a decreasing SM value (Figure 8(b2))
is related to the occurrence of transformation between different crystalline structures [27],
i.e., a small degree of reversion austenite formation, accompanied by Ni, Mn, and C ele-
ment partition from martensite into austenite. This can be inferred from two perspectives:
(i) the SM value decreasing due to austenite as a non-magnetic phase, whose presence is
often accompanied by magnetic decline [28], and (ii) the retained austenite along the
HAGB after cooling (Figure 9(b2), which is the result of hydrostatic pressure and high
stability [29]. Notably, the martensite reversion transformation happened below Ac1. This
is aributed to the precipitation of cementite with high Ni content (Figure 9(a1,a2)) during
the S1 stage, which not only provided the revision austenite nucleation site but more car-
bon and nickel elements from the dissolution of cementite can also improve the local driv-
ing of reversion martensite transformation and promote the migration of the interface
from austenite to martensite [30].
The decrease in length in S3 (Figure 8(a3)) with an increasing SM value (Figure 8(b3))
was connected to austenite decomposing during tempering [31], with the corresponding
product including ferrite and fine lamellar cementite (Figure 2b) from the austenite
Figure 7. Fractography morphology of the QT and MTP samples after the impact test: (a,b) the
fracture morphology; (c1,d1) the local profile fractography; and (c2,d2) the crack propagation trace
determined by EBSD (the red area is the positions for the SEM in the Charpy samples).
Materials 2024,17, 518 10 of 16
5. Discussion
5.1. Microstructure Evolution during MTP Heat Treatment
The DIL was scheduled to capture the phase transformation behavior during the MTP
process (Figure 8(a1,a2,a3)). No martensite transformation occurs at each step (S1, S2,
and S3), as the temperature falls to room temperature when the cooling rate is 50
◦
C/s.
Furthermore, the corresponding change in length (
∆
L/L) at the isothermal stage of every
step shows a decreasing trend with the extension of isothermal time. SM was measured, as
shown in Figure 8(b1,b2,b3), to analyze phase transformation and the carbide precipitation
phenomenon further.
Materials 2024, 17, x FOR PEER REVIEW 11 of 16
decomposition inherited from the higher Ni content (Figure 3d) rather than spheroidized
cementite with high Cr and Mo content. This can effectively reduce the coarsening of ce-
mentite at the HAGB. A similar phenomenon was observed in high carbon-bearing steel
containing blocky austenite after tempering [20]. The dislocation as a laice defect has a
higher binding energy for the carbon element segment [32], leading to an intragranular
uniform distribution of carbon elements rather than only segregation at the grain bound-
ary. Meanwhile, the kinetics of cementite growth from the martensite during the temper-
ing is directly proportional to the relationship with the local carbon concentration. Con-
sequently, the cementite in the MTP sample was precipitated from martensite, which is
more refined than the QT sample due to the higher dislocation density.
Compared to the QT sample, the MTP process promotes the refinement of martensite
lath. The long tempering time and high tempering temperature led to grain growth due
to the merging of the subgrain boundary, i.e., the vanishing of the LAGB due to disloca-
tion recovery [18]; thus, the lower dislocation density was retained in the martensite lath
of the QT sample (Figure 4(b3)).
Figure 8. Phase transformation behavior of 25Cr2Ni3MoV steel during the isothermal stage in the
MTP process: (a1) the DIL curve showing the length change for S1; (a2) the DIL curve showing the
length change for S2; (a3) the DIL curve showing the length change for S3 (ΔL/L starts from 0%, and
the starting and ending of the isothermal stage are indicated by the red and green circles); and
(b1,b2,b3) the variation in SM during the isothermal stage for S1, S2, and S3.
Figure 8. Phase transformation behavior of 25Cr2Ni3MoV steel during the isothermal stage in the
MTP process: (a1) the DIL curve showing the length change for S1; (a2) the DIL curve showing the
length change for S2; (a3) the DIL curve showing the length change for S3 (
∆
L/L starts from 0%,
and the starting and ending of the isothermal stage are indicated by the red and green circles); and
(b1,b2,b3) the variation in SM during the isothermal stage for S1, S2, and S3.
Comparing Figures 1b and 9(a1), the decrease in length in S1 (Figure 8(a1)) with a
stable SM value (Figure 8(b1)) was attributed to the lamellar cementite precipitation (width:
10 nm; length: 70 nm), which was generated from the interaction between dislocation in
the quenching martensite and carbide segment.
The decrease in length in S2 (Figure 8(a2)) with a decreasing SM value (Figure 8(b2))
is related to the occurrence of transformation between different crystalline structures [
27
],
i.e., a small degree of reversion austenite formation, accompanied by Ni, Mn, and C element
partition from martensite into austenite. This can be inferred from two perspectives: (i) the
SM value decreasing due to austenite as a non-magnetic phase, whose presence is often
accompanied by magnetic decline [
28
], and (ii) the retained austenite along the HAGB after
cooling (Figure 9(b2), which is the result of hydrostatic pressure and high stability [
29
].
Notably, the martensite reversion transformation happened below Ac1. This is attributed
to the precipitation of cementite with high Ni content (Figure 9(a1,a2)) during the S1 stage,
which not only provided the revision austenite nucleation site but more carbon and nickel
Materials 2024,17, 518 11 of 16
elements from the dissolution of cementite can also improve the local driving of reversion
martensite transformation and promote the migration of the interface from austenite to
martensite [30].
Materials 2024, 17, x FOR PEER REVIEW 12 of 16
Figure 9. Microstructures of 25Cr2Ni3MoV steel after the S1 state (a1,a2) and the S2 state (b1,b2):
(a1) SEM image showing the tempering martensite and nano carbide as the inset and (a2) the
bright field (BF) and dark field (DF) image for globular cementite and the corresponding element
content determined using the single-point energy-dispersive spectroscopy (EDS) method. (b1)
SEM image showing the austenite. (b2) EBSD maps of austenite along the high-angle grain bound-
ary.
5.2. Strengthening Contribution
The strengthening contribution in this experimental steel mainly involves intrinsic
strengthening (σINT = σ0 (laice friction stress) + σSS (solid solution strengthening)) [33],
grain boundary strengthening (σGB) [34], precipitation strengthening (σPRE) [35], and dislo-
cation strengthening (σDIS) [26], and the corresponding calculation formula is as follows:
Δ𝜎 =𝐾×(1
√
𝑑
⁄) (7)
Δ𝜎 = 0.538𝐺𝑏
𝑓
𝑑
‾
∙ln 𝑑
‾
2𝑏
⁄ (8)
Δ𝜎 = 𝑀𝛼𝐺𝑏
𝜌 (9)
where K is the strengthening coefficient, d is the average width of martensite lath, f is the
area fraction of cementite, 𝑑
‾ is the average size of cementite, b is the Burgers vector, M is
the average Taylor factor, 𝛼 is the interaction strength between dislocations, and G is the
shear modulus. The σINT is 160 MPa.
From the results in Table 4, it can be seen that the double refinement of the matrix
and precipitated phase promoted strength improvement before yield strength. Mean-
while, dislocation multiplication and pileup provided a working hardening ability during
cryogenic deformation in the MTP sample.
Table 4. Strength contribution for the two types of steel at room temperature.
Sample σINT d σGB 𝒅
‾
𝒇
σPRE 𝝆 σDIS σTotall(cal) σtotal(exp)
MPa μm MPa nm % MPa ×10−14 m−2 MPa MPa MPa
QT 160 4 142 120 7 120 2.38 307 729 746
MTP 160 1.5 168 80 4 138 3.51 380 826 820
Figure 9. Microstructures of 25Cr2Ni3MoV steel after the S1 state (a1,a2) and the S2 state (b1,b2):
(a1) SEM image showing the tempering martensite and nano carbide as the inset and (a2) the bright
field (BF) and dark field (DF) image for globular cementite and the corresponding element content
determined using the single-point energy-dispersive spectroscopy (EDS) method. (b1) SEM image
showing the austenite. (b2) EBSD maps of austenite along the high-angle grain boundary.
The decrease in length in S3 (Figure 8(a3)) with an increasing SM value (Figure 8(b3))
was connected to austenite decomposing during tempering [
31
], with the corresponding
product including ferrite and fine lamellar cementite (Figure 2b) from the austenite de-
composition inherited from the higher Ni content (Figure 3d) rather than spheroidized
cementite with high Cr and Mo content. This can effectively reduce the coarsening of
cementite at the HAGB. A similar phenomenon was observed in high carbon-bearing steel
containing blocky austenite after tempering [
20
]. The dislocation as a lattice defect has a
higher binding energy for the carbon element segment [
32
], leading to an intragranular
uniform distribution of carbon elements rather than only segregation at the grain boundary.
Meanwhile, the kinetics of cementite growth from the martensite during the tempering is
directly proportional to the relationship with the local carbon concentration. Consequently,
the cementite in the MTP sample was precipitated from martensite, which is more refined
than the QT sample due to the higher dislocation density.
Compared to the QT sample, the MTP process promotes the refinement of martensite
lath. The long tempering time and high tempering temperature led to grain growth due to
the merging of the subgrain boundary, i.e., the vanishing of the LAGB due to dislocation
recovery [
18
]; thus, the lower dislocation density was retained in the martensite lath of the
QT sample (Figure 4(b3)).
5.2. Strengthening Contribution
The strengthening contribution in this experimental steel mainly involves intrinsic
strengthening (
σINT
=
σ0
(lattice friction stress) +
σSS
(solid solution strengthening)) [
33
],
Materials 2024,17, 518 12 of 16
grain boundary strengthening (
σGB
) [
34
], precipitation strengthening (
σPRE
) [
35
], and dislo-
cation strengthening (σDIS) [26], and the corresponding calculation formula is as follows:
∆σGB =K×1/√d(7)
∆σPRE =0.538G b f 1
2/¯
d·ln¯
d/2b(8)
∆σDIS =MαGb√ρ(9)
where Kis the strengthening coefficient, dis the average width of martensite lath, fis the
area fraction of cementite,
¯
d
is the average size of cementite, bis the Burgers vector, Mis
the average Taylor factor,
α
is the interaction strength between dislocations, and Gis the
shear modulus. The σINT is 160 MPa.
From the results in Table 4, it can be seen that the double refinement of the matrix and
precipitated phase promoted strength improvement before yield strength. Meanwhile, dis-
location multiplication and pileup provided a working hardening ability during cryogenic
deformation in the MTP sample.
Table 4. Strength contribution for the two types of steel at room temperature.
Sample σINT dσGB −
dfσPRE ρσDIS σTotall(cal) σtotal(exp)
MPa µm MPa nm % MPa ×10−14 m−2MPa MPa MPa
QT 160 4 142 120 7 120 2.38 307 729 746
MTP 160 1.5 168 80 4 138 3.51 380 826 820
5.3. Deformation Compatibility of Tempered Martensite and Precipitate
The microstructural evolutions and associated mechanical behavior of the 25Cr2Ni3MoV
steel during deformation can be depicted using the CPFEM.
Figure 10(a1,a2,b1,b2) show the simulated IPF images of QT and MTP steel, corre-
sponding to the EBSD measurements with the junction of the three PAG boundaries; the
different colors indicate the variations in the crystal orientations among different martensite
variants. Furthermore, the significant difference between QT and MTP steel in terms of
grain size, crystal orientation, and morphology prohibited comprehension.
There is a slight difference between the experimental and simulated true stress–strain
curve, as there is a yield plateau on the experimental tensile curve because of the existence
of plastic, affecting the stability performance at the initial deformation. However, this does
not affect the application of the model at present because the yield strength and subsequent
hardening trends were in good agreement with the experimental curve (Figure 10(c1,c2)).
Figure 10(d1,d2) shows the accumulated plastic shear strain distribution of QT and
MTP steel at 6% true strain. Although the accumulated plastic shear strain degree of the
two samples is similar, the strain concentration band in the QT sample (the area for the
white frame line) is wider than the MTP sample. Meanwhile, cross-distribution occurs at
neighboring PAG boundaries in the QT samples, which can become a crack nucleation
point and aggregate the local deformation concentration, resulting in incompatibility
deformation [
36
]. Conversely, the existence of parallel and dispersive strain concentration
bands in the MTP sample can prove the formation of geometry necessary dislocation (GND)
caused by the high strain gradient, i.e., higher dislocation strength was achieved [37].
The new model was introduced by adding coarse carbide particles near the PAG
boundary based on the QT model (Figure 10(a2)), as shown in Figure 11(a2). Compared
to the no-cementite model (Figure 11(a1)), the high-stress area mainly concentrates on
the interface between the cementite and matrix (Figure 11(b2)). In comparison, the high-
stress area was connected to a curve along the PAG boundary, especially along the tensile
direction, as shown in the area of the white frame in Figure 11(b2); simultaneously, the
Materials 2024,17, 518 13 of 16
formation of the local strain concentration area with the spacing of the precipitated phase
decreasing can be seen in Figure 11(c2).
Materials 2024, 17, x FOR PEER REVIEW 13 of 16
5.3. Deformation Compatibility of Tempered Martensite and Precipitate
The microstructural evolutions and associated mechanical behavior of the
25Cr2Ni3MoV steel during deformation can be depicted using the CPFEM.
Figure 10(a1,a2,b1,b2) show the simulated IPF images of QT and MTP steel, corre-
sponding to the EBSD measurements with the junction of the three PAG boundaries; the
different colors indicate the variations in the crystal orientations among different marten-
site variants. Furthermore, the significant difference between QT and MTP steel in terms
of grain size, crystal orientation, and morphology prohibited comprehension.
There is a slight difference between the experimental and simulated true stress–strain
curve, as there is a yield plateau on the experimental tensile curve because of the existence
of plastic, affecting the stability performance at the initial deformation. However, this does
not affect the application of the model at present because the yield strength and subse-
quent hardening trends were in good agreement with the experimental curve (Figure
10(c1,c2)).
Figure 10(d1,d2) shows the accumulated plastic shear strain distribution of QT and
MTP steel at 6% true strain. Although the accumulated plastic shear strain degree of the
two samples is similar, the strain concentration band in the QT sample (the area for the
white frame line) is wider than the MTP sample. Meanwhile, cross-distribution occurs at
neighboring PAG boundaries in the QT samples, which can become a crack nucleation
point and aggregate the local deformation concentration, resulting in incompatibility de-
formation [36]. Conversely, the existence of parallel and dispersive strain concentration
bands in the MTP sample can prove the formation of geometry necessary dislocation
(GND) caused by the high strain gradient, i.e., higher dislocation strength was achieved
[37].
Figure 10. Experimental and simulated results for QT and TDP steel at room temperature: (a1,a2)
IPF maps with three PAG boundaries. (b1,b2) The orientation map used for the CPFEM simulation.
(c1,c2) True stress–strain curve, comparing the experimental and simulated observations. (d1,d2)
The accumulated shear strain distribution at a true strain of 6%; note: in the cloud pictures, the strain
ranges are set to 0.0067–0.706, locations with values exceeding the maxima are shown in gray, and
those with values less than the minima are shown in black.
The new model was introduced by adding coarse carbide particles near the PAG
boundary based on the QT model (Figure 10(a2)), as shown in Figure 11(a2). Compared
to the no-cementite model (Figure 11(a1)), the high-stress area mainly concentrates on the
interface between the cementite and matrix (Figure 11(b2)). In comparison, the high-stress
Figure 10. Experimental and simulated results for QT and TDP steel at room temperature: (a1,a2) IPF
maps with three PAG boundaries. (b1,b2) The orientation map used for the CPFEM simulation.
(c1,c2) True stress–strain curve, comparing the experimental and simulated observations. (d1,d2) The
accumulated shear strain distribution at a true strain of 6%; note: in the cloud pictures, the strain
ranges are set to 0.0067–0.706, locations with values exceeding the maxima are shown in gray, and
those with values less than the minima are shown in black.
Materials 2024, 17, x FOR PEER REVIEW 14 of 16
area was connected to a curve along the PAG boundary, especially along the tensile direc-
tion, as shown in the area of the white frame in Figure 11(b2); simultaneously, the for-
mation of the local strain concentration area with the spacing of the precipitated phase
decreasing can be seen in Figure 11(c2).
As is known, the microvoid can easily nucleate between the soft and hard phases due
to deformation incompatibility. In this study, however, the microvoid occurred between
the same phases (tempering martensite lath), as shown in Figure 7(c2) due to the occur-
rence of strain concentration. That is to say, damage nucleation during plastic deformation
was aggravated by the strain concentration between two cementites rather than the stress
concentration induced by dislocation pileup.
On the other hand, the ability of PAG boundaries to dissipate crack propagation en-
ergy was weakened due to the continuous high-stress area and promoted cracking be-
tween two PAGs, compared to other HAGBs. The deformation that occurs during the im-
pact condition is more drastic than that which occurs during the uniaxial tensile condition;
thus, the QT model under the tensile condition can explain the deterioration in fracture
toughness to a certain extent. The crack shows a preference for propagating along the PAG
boundary (parallel to the loading direction) in the QT sample (Figure 7(d2)). The higher
proportion of HAGBs in the MTP sample also provided more effective obstacles to cleav-
age crack propagation, so higher cryogenic impact toughness was obtained.
Figure 11. CPFEM results for QT steel considering the cementite along the PAG boundary: (a1,a2)
phase maps with three PAG boundaries. Note: dark green regions represent the matrix (tempering
martensite), and yellow regions represent cementite. (b1,b2) The stress distribution map with the
ranges is set to 460–3680 MPa. Locations with values exceeding the maxima are shown in gray.
(c1,c2) The strain distribution map.
6. Conclusions
In this study, we evaluated tensile properties and impact toughness at cryogenic tem-
peratures for medium carbon 25Cr2Ni3MoV steel by comparing (i) conventional quench-
ing and tempering (QT) and (ii) multi-step tempering (MTP).
Compared to the QT sample, the MTP sample possesses refined martensite lath and
cementite through the reduction in tempering time and temperature, and corresponding
excellent cryogenic results were obtained for yield strength (1300 MPa), total elongation
(25%), and impact toughness (>25 J) at liquid nitrogen temperature.
Figure 11. CPFEM results for QT steel considering the cementite along the PAG boundary: (a1,a2)
phase maps with three PAG boundaries. Note: dark green regions represent the matrix (tempering
martensite), and yellow regions represent cementite. (b1,b2) The stress distribution map with the
ranges is set to 460–3680 MPa. Locations with values exceeding the maxima are shown in gray. (c1,c2)
The strain distribution map.
Materials 2024,17, 518 14 of 16
As is known, the microvoid can easily nucleate between the soft and hard phases due
to deformation incompatibility. In this study, however, the microvoid occurred between the
same phases (tempering martensite lath), as shown in Figure 7(c2) due to the occurrence
of strain concentration. That is to say, damage nucleation during plastic deformation
was aggravated by the strain concentration between two cementites rather than the stress
concentration induced by dislocation pileup.
On the other hand, the ability of PAG boundaries to dissipate crack propagation energy
was weakened due to the continuous high-stress area and promoted cracking between two
PAGs, compared to other HAGBs. The deformation that occurs during the impact condition
is more drastic than that which occurs during the uniaxial tensile condition; thus, the QT
model under the tensile condition can explain the deterioration in fracture toughness to
a certain extent. The crack shows a preference for propagating along the PAG boundary
(parallel to the loading direction) in the QT sample (Figure 7(d2)). The higher proportion
of HAGBs in the MTP sample also provided more effective obstacles to cleavage crack
propagation, so higher cryogenic impact toughness was obtained.
6. Conclusions
In this study, we evaluated tensile properties and impact toughness at cryogenic tem-
peratures for medium carbon 25Cr2Ni3MoV steel by comparing (i) conventional quenching
and tempering (QT) and (ii) multi-step tempering (MTP).
Compared to the QT sample, the MTP sample possesses refined martensite lath and
cementite through the reduction in tempering time and temperature, and corresponding
excellent cryogenic results were obtained for yield strength (1300 MPa), total elongation
(25%), and impact toughness (>25 J) at liquid nitrogen temperature.
The high interface density and refinement of cementite along the PAG boundary were
obtained through the use of reversion martensite transformation and austenite decomposi-
tion during the MTP process, improving the strength and coordinating the deformation
compatibility by means of dispersive strain concentration bands.
Moreover, coarse cementite along HAGBs will promote local strain concentration and
microvoid nucleation and weak PAG boundaries’ ability to consume crack propagation
energy, which is a fatal threat to ductile metal material.
Author Contributions: Conceptualization, Y.C.; methodology, Y.C., N.M. and W.L.; software, Y.C. and
R.C.; validation, Y.Y. and A.D.; investigation, Y.C. and W.L.; writing—original draft, Y.C.; writing—
review and editing, W.L.; supervision, Y.Y. and A.D. All authors have read and agreed to the published
version of the manuscript.
Funding: The National Natural Science Foundation of China (No. 52071209).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available upon request from the
corresponding author.
Conflicts of Interest: The authors declare no conflicts of interest.
References
1.
Noor Akashah, M.H.; Mohammad Rozali, N.E.; Mahadzir, S.; Liew, P.Y. Utilization of Cold Energy from LNG Regasification
Process: A Review of Current Trends. Processes 2023,11, 517. [CrossRef]
2.
Park, H.; Lee, S.; Jeong, J.; Chang, D. Design of the compressor-assisted LNG fuel gas supply system. Energy 2018,158, 1017–1027.
[CrossRef]
3.
Kumar, S.; Kwon, H.-T.; Choi, K.-H.; Lim, W.; Cho, J.H.; Tak, K.; Moon, I. LNG: An eco-friendly cryogenic fuel for sustainable
development. Appl. Energy 2011,88, 4264–4273. [CrossRef]
4.
Dowson, P.; Bauer, D.; Laney, S. Selection of materials and material related processes for centrifugal compressors and steam
turbines in the oil and petrochemical industry. In Proceedings of the 37th Turbomachinery Symposium; Turbomachinery Laboratories;
Texas A&M University: College Station, TX, USA, 2008; pp. 190–209.
Materials 2024,17, 518 15 of 16
5.
Fadel, A.; Gli
ˆ
si´c, D.; Radovi´c, N.; Drobnjak, D. Influence of Cr, Mn and Mo Addition on Structure and Properties of V Microalloyed
Medium Carbon Steels. J. Mater. Sci. Technol. 2012,28, 1053–1058. [CrossRef]
6.
Hou, N.; Ding, N.; Qu, S.; Guo, W.; Liu, L.; Xu, N.; Tian, L.; Xu, H.; Chen, X.; Zaïri, F.; et al. Failure modes, mechanisms and
causes of shafts in mechanical equipment. Eng. Fail. Anal. 2022,136, 106216. [CrossRef]
7.
Hu, Y.; Wu, S.; Withers, P.J.; Cao, H.; Chen, P.; Zhang, Y.; Shen, Z.; Vojtek, T.; Hutaˇr, P. Corrosion fatigue lifetime assessment of
high-speed railway axle EA4T steel with artificial scratch. Eng. Fract. Mech. 2021,245, 107588. [CrossRef]
8.
Krauss, G. Tempering of Lath Martensite in Low and Medium Carbon Steels: Assessment and Challenges. Steel Res. Int. 2017,
88, 1700038. [CrossRef]
9.
Lee, C.; Maresca, F.; Feng, R.; Chou, Y.; Ungar, T.; Widom, M.; An, K.; Poplawsky, J.D.; Chou, Y.-C.; Liaw, P.K. Strength can be
controlled by edge dislocations in refractory high-entropy alloys. Nat. Commun. 2021,12, 5474. [CrossRef]
10.
Morris, J.W., Jr.; Kinney, C.; Pytlewski, K.; Adachi, Y. Microstructure and cleavage in lath martensitic steels. Sci. Technol. Adv.
Mater. 2013,14, 014208. [CrossRef]
11.
Hayakawa, K.; Ogawa, T.; He, L.; Sun, F.; Adachi, Y. Improvement in the Strength–Ductility Balance of Tempered Martensite Steel
by Controlling Cementite Particle Size Distribution. J. Mater. Eng. Perform. 2023, 1–11. [CrossRef]
12. Prasad, C.; Bhuyan, P.; Kaithwas, C.; Saha, R.; Mandal, S. Microstructure engineering by dispersing nano-spheroid cementite in
ultrafine-grained ferrite and its implications on strength-ductility relationship in high carbon steel. Mater. Des. 2018,139, 324–335.
[CrossRef]
13.
Miyamoto, G.; Yokoyama, K.; Furuhara, T. Quantitative analysis of Mo solute drag effect on ferrite and bainite transformations in
Fe-0.4 C-0.5 Mo alloy. Acta Mater. 2019,177, 187–197. [CrossRef]
14.
Wu, Y.; Sun, W.; Gao, X.; Styles, M.; Arlazarov, A.; Hutchinson, C. The effect of alloying elements on cementite coarsening during
martensite tempering. Acta Mater. 2020,183, 418–437. [CrossRef]
15.
Rawal, S.; Gurland, J. Observations on the effect of cementite particles on the fracture toughness of spheroidized carbon steels.
Metall. Trans. A 1977,8, 691–698. [CrossRef]
16.
Umar, M.; Qayyum, F.; Farooq, M.U.; Khan, L.A.; Guk, S.; Prahl, U. Investigating the Effect of Cementite Particle Size and
Distribution on Local Stress and Strain Evolution in Spheroidized Medium Carbon Steels using Crystal Plasticity-Based Numerical
Simulations. Steel Res. Int. 2020,92, 2000407, 1–13. [CrossRef]
17.
Yin, W.; Briffod, F.; Yamazaki, K.; Shiraiwa, T.; Enoki, M. Effect of limited retained austenite on the strength–ductility trade-off in
low-alloyed TRIP steel. Mater. Sci. Eng. A 2022,861, 144337. [CrossRef]
18.
Kumar, S.; Samanta, S.; Singh, S.B. An alternative quenching and partitioning (Q&P) process via spheroidization. Mater. Charact.
2022,191, 112049. [CrossRef]
19.
Zhang, H.; Sun, M.; Liu, Y.; Ma, D.; Xu, B.; Huang, M.; Li, D.; Li, Y. Ultrafine-grained dual-phase maraging steel with high
strength and excellent cryogenic toughness. Acta Mater. 2021,211, 116878. [CrossRef]
20.
Ribamar, G.G.; Escobar, J.D.; Kwiatkowski da Silva, A.; Schell, N.; Ávila, J.A.; Nishikawa, A.S.; Oliveira, J.P.; Goldenstein, H.
Austenite carbon enrichment and decomposition during quenching and tempering of high silicon high carbon bearing steel. Acta
Mater. 2023,247, 118742. [CrossRef]
21.
Tkachev, E.; Borisov, S.; Belyakov, A.; Kniaziuk, T.; Vagina, O.; Gaidar, S.; Kaibyshev, R. Effect of quenching and tempering on
structure and mechanical properties of a low-alloy 0.25C steel. Mater. Sci. Eng. A 2023,868, 144757. [CrossRef]
22.
Xie, M.; Chen, G.; Yu, J.; Wu, Y.; Liu, X.; Yang, J.; Xu, W. A dual-phase crystal plasticity finite-element method for modeling the
uniaxial deformation behaviors of thermally aged SAC305 solder. Int. J. Plast. 2022,155, 103342. [CrossRef]
23.
Peng, J.; Zhou, B.; Li, Z.; Huo, D.; Xiong, J.; Zhang, S. Effect of tempering process on the cryogenic impact toughness of 13Cr4NiMo
martensitic stainless steel. J. Mater. Res. Technol. 2023,23, 5618–5630. [CrossRef]
24.
Feng, X.; Yang, Z.; Zhang, X.; Shen, Y. Mechanical Properties of Fine-Grained Low Alloy Steel Investigated by Crystal Plasticity
Finite Element Method. J. Mater. Eng. Perform. 2023, 1–12. [CrossRef]
25.
Tanaka, Y.; Masumura, T.; Tsuchiyama, T.; Takaki, S. Effect of dislocation distribution on the yield stress in ferritic steel under
identical dislocation density conditions. Scr. Mater. 2020,177, 176–180. [CrossRef]
26. He, Z.; Jia, N.; Wang, H.; Yan, H.; Shen, Y. Synergy effect of multi-strengthening mechanisms in FeMnCoCrN HEA at cryogenic
temperature. J. Mater. Sci. Technol. 2021,86, 158–170. [CrossRef]
27.
Souza Filho, I.R.; Sandim, M.J.R.; Cohen, R.; Nagamine, L.C.C.M.; Sandim, H.R.Z.; Raabe, D. Magnetic properties of a 17.6
Mn-TRIP steel: Study of strain-induced martensite formation, austenite reversion, and athermal
α′
-formation. J. Magn. Magn.
Mater. 2019,473, 109–118. [CrossRef]
28.
Kaputkina, L.M.; Svyazhin, A.G.; Smarygina, I.V.; Kindop, V.E. Light Non-Magnetic Steels Based on the Fe–25 Mn–5 Ni–Al–C
System. Steel Transl. 2020,50, 22–30. [CrossRef]
29.
Nakada, N.; Ishibashi, Y.; Tsuchiyama, T.; Takaki, S. Self-stabilization of untransformed austenite by hydrostatic pressure via
martensitic transformation. Acta Mater. 2016,110, 95–102. [CrossRef]
30.
Xu, Y.; Li, W.; Du, H.; Jiao, H.; Liu, B.; Wu, Y.; Ding, W.; Luo, Y.; Nie, Y.; Min, N.; et al. Tailoring the metastable reversed austenite
from metastable Mn-rich carbides. Acta Mater. 2021,214, 116986. [CrossRef]
31.
Li, J.; Bai, S.; Li, H.; Jiang, X.; Meng, F.; Luo, H. Decomposition in Mn50Ni40In10 melt-spun ribbons and its influence on magnetic
properties. J. Magn. Magn. Mater. 2023,588, 171502. [CrossRef]
Materials 2024,17, 518 16 of 16
32.
Morsdorf, L.; Kashiwar, A.; Kübel, C.; Tasan, C.C. Carbon segregation and cementite precipitation at grain boundaries in
quenched and tempered lath martensite. Mater. Sci. Eng. A 2023,862, 144369. [CrossRef]
33.
Nikitin, I.; Fedoseeva, A.; Kaibyshev, R. Strengthening mechanisms of creep-resistant 12%Cr–3%Co steel with low N and high B
contents. J. Mater. Sci. 2020,55, 7530–7545. [CrossRef]
34.
Dudko, V.; Yuzbekova, D.; Kaibyshev, R. Strengthening Mechanisms in a Medium-Carbon Steel Subjected to Thermo-Mechanical
Processing. Appl. Sci. 2023,13, 9614. [CrossRef]
35. Gladman, T. Precipitation hardening in metals. Mater. Sci. Technol. 1999,15, 30–36. [CrossRef]
36.
Tang, A.; Liu, H.; Chen, R.; Liu, G.; Lai, Q.; Zhong, Y.; Wang, L.; Wang, J.; Lu, Q.; Shen, Y. Mesoscopic origin of damage nucleation
in dual-phase steels. Int. J. Plast. 2021,137, 102920. [CrossRef]
37.
Hansen, L.T.; Fullwood, D.T.; Homer, E.R.; Wagoner, R.H.; Lim, H.; Carroll, J.D.; Zhou, G.; Bong, H.J. An investigation of
geometrically necessary dislocations and back stress in large grained tantalum via EBSD and CPFEM. Mater. Sci. Eng. A 2020,
772, 138704. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
