Materials Characterization 207 (2024) 113559

Contents lists available at ScienceDirect

Materials Characterization
journal homepage: www.elsevier.com/locate/matchar

Grain refinement mechanism in chip and the machined subsurface during

high-speed machining of Inconel 718 alloy
Ailing Shu, Junxue Ren, Jinhua Zhou *, Zongyuan Wang
Key Laboratory of High Performance Manufacturing for Aero Engine (Northwestern Polytechnical University), Ministry of Industry and Information Technology, Xi’an,
People’s Republic of China

A R T I C L E I N F O A B S T R A C T

Keywords: Microstructural defects on the machined surface of Inconel 718 alloy can be a significant source of fatigue cracks,
Inconel 718 alloy potentially leading to failures of the structural components of aerospace engines. This work focused on the grain
Dislocation density refinement mechanism induced by the dynamic recrystallization (DRX) behavior in both the chip and machined
Nanocrystallite
subsurface during high-speed machining (HSM) of Inconel 718 alloy. Multiscale metallurgical observations were
Machined subsurface
High-speed machining
conducted using electron backscatter diffraction (EBSD), transmission electron microscopy (TEM), and preces­
sion electron diffraction (PED) to characterize the crystal structures, including the grain size, grain boundary
distribution, dislocation density, and grain orientation distribution. Finite element (FE) simulations were per­
formed to obtain the strain, strain rate, and cutting temperature to analyze the microstructure evolution
mechanism under the effect of severe deformation conditions in HSM. The results indicate that grain refinement
in the chip primarily occurs in the shear band and secondary deformation zone. The primary cause of the
variation in the final grain size is attributed to the distinct thermomechanical loading conditions experienced by
the material. The microstructure observed in the machined subsurface reveals a significant gradient distribution,
characterized by the presence of an amorphous structure, nanocrystallites, refined grains, and elongated grains
along the depth direction. The equiaxed ultrafine nanocrystals observed in the machined near-surface result from
both continuous DRX (cDRX) and discontinuous DRX (dDRX) mechanisms. In the deeper zone, the presence of
numerous subgrains and a high geometrically necessary dislocation (GND) density reaching up to 1016/m2
within the grains indicate that the refined grains in this zone result from the cDRX mechanism. The variation in
DRX mechanisms observed in the machined subsurface are closely related to the gradient distribution of the
thermomechanical load generated in HSM.

1. Introduction safety-critical components when considering a low-cycle fatigue life.

Therefore, a better understanding of the underlying mechanisms of
Inconel 718 alloy is extensively used in the blisks and casings of microstructural evolution in machined surface is not only crucial for
aerospace engines owing to its high fatigue resistance and excellent theoretical insights but also for the development of engineering appli­
thermal stability in severe working environment [1,2]. However, the cations and technologies.
excellent properties also make Inconel 718 alloy a difficult-to-cut ma­ Liao et al. [12] indicated that the microstructure evolution was
terial owing to the rapid tool wear, high cutting forces, and high tem­ primarily driven by mechanical-dominant, thermal-dominant,
perature in machining operation [3]. In high-speed machining (HSM), chemical-dominant, and combined loading mechanisms, which depen­
the microstructure of the machined surface usually changes [4,5], ded on the nature of load induced by machining the workpiece. The first
including phase transformation, dynamic recrystallization (DRX), gen­ three attribute the microstructure evolution to a single factor, which is
eration of white layer, and so on [6,7], which in turn modifies the not suitable for HSM process [13]. Pan et al. [14] indicated that the
physical mechanical properties and final fatigue life of structural com­ microstructure evolution in cutting process was determined by the
ponents [8–10]. La et al. [11] suggested that the existence of nano­ loading history of force and thermal load. In HSM, DRX is easier to occur
crystallite on the machined surface may be not beneficial for the most than phase transformation, and grain refinement is its typical

* Corresponding author.
E-mail address: zhoujinhua@mail.nwpu.edu.cn (J. Zhou).

https://doi.org/10.1016/j.matchar.2023.113559
Received 24 May 2023; Received in revised form 22 August 2023; Accepted 9 December 2023
Available online 12 December 2023
1044-5803/© 2023 Elsevier Inc. All rights reserved.
A. Shu et al. Materials Characterization 207 (2024) 113559

Table 1 combination of FE and CA methods was an effective way to fully un­

Orthogonal cutting conditions for Inconel 718 alloy. derstand the microstructure evolution in HSM. Wang et al. [27]
Test Cutting Uncut chip Test Cutting Uncut chip employed FE simulation to investigate the grain size evolution and
speed (V) thickness (t) speed (V) thickness (t) microhardness in chip for cutting Ti6Al4V alloy, showing that HSM was
(m/min) (mm) (m/min) (mm) a reliable approach for generating exceptionally refined grains if proper
1 100 0.025 6 200 0.05 machining parameters were selected. The microstructural evolution in
2 200 0.025 7 300 0.05 HSM is a complex phenomenon that is not yet fully understood, further
3 300 0.025 8 400 0.05 research is deemed necessary. The investigations above shows that the
4 400 0.025 9 100 0.10
5 100 0.05 / / /
combination of FE and microstructure characterization is an effective
way to fully understand this matter.
In this study, the refinement mechanisms in chip and the machined
characteristic. Previously, heat treatment recrystallization theories, subsurface for HSM Inconel 718 alloy were studied by combining
including the grain boundary migration and subgrain merging theory, experimental and simulated methods. First, orthogonal cutting experi­
have been usually used to explain the recrystallization phenomenon of ments were conducted, and an FE model was developed using ABAQUS/
the machined surface [15]. However, the times required to complete the Explicit. Then, the EBSD, TEM, and PED technologies were applied to
grain boundary migration and subgrain merging were 105 and 106 s at characterize the crystal information, such as grain size, dislocation
1000 K, respectively [16]. Clearly, they cannot explain the DRX phe­ density, grain boundary distribution, and subgrain structure. The plastic
nomenon under a high strain rate condition in HSM. Hines et al. [17] strain, strain rate, and cutting temperature in HSM were obtained
proposed a progressive subgrain site difference recrystallization mech­ through FE simulation. Finally, the grain refinement mechanisms in chip
anism for the microstructure evolution in the shear band, indicating the and at different depths of the machined subsurface were analyzed.
mechanical rotation of subgrains was the key to recrystallization. Ren
et al. [18] studied the grain refinement mechanism of machined surface 2. Experiment and simulation
for cutting Inconel 718 alloy, showing that the grain refinement is the
result of dislocation-twin interaction. As evident from the above dis­ 2.1. Orthogonal cutting tests of Inconel 718 alloy
cussion, most of the past investigations have primarily concentrated on
analyzing the machined surface and chip, with limited attention given to High-speed orthogonal cutting experiments were conducted using an
the grain refinement mechanism of the machined surface at different HK63 lathe. Cooling lubrication cutting condition was applied at four
depths. cutting speeds and three uncut chip thicknesses, the cutting parameters
The development of advanced material testing technologies, such as are listed in Table 1. Sharp and coated carbide cutters, whose initial rake
scanning electron microscopy (SEM), transmission electron microscopy angles, relief angles and edge radiuses were measured using an auto­
(TEM), and electron backscatter diffraction (EBSD) have provided matic tool-measuring instrument (IFM-G4) as 12◦ , 6◦ and 20 μm,
effective means of delving deeper for material microstructure evolution respectively, were used in the cutting experiments. A cylinder of an
[19]. Wang et al. [20] observed the twinning behavior of the machined aging-treated Inconel 718 alloy with the dimensions of Φ213 × 300 mm
surface in cutting Ti6Al4V alloy by EBSD and TEM tests. The formation was selected as the workpiece. The chemical compositions of the forging
mechanisms of deformation twinning and its effect on work hardening factories are listed in Table 2. The original microstructure of the unde­
were revealed and analyzed. Pane et al. [21] observed the metallo­ formed region was measured using a scanning electron microscope
graphic and crystal structures of the machined surface for cutting (EDAX Velocity Super); the obtained EBSD images are shown in Fig. 1.
Ti6Al4V alloy, and attributed the low content of β phase to the high As shown in Fig. 1(a), the original microstructure of Inconel 718 alloy
cutting temperature. When the grains are highly deformed or refined to consisted of equiaxed grains and certain annealing twins. Fig. 1(b)
nano-scale, the resolution of EBSD is too low to obtain high-quality maps shows a grain boundary map and Fig. 1(f) shows the distribution of the
[12]. As an emerging TEM-based technique, precession electron misorientation angles. The analysis revealed that high-angle grain
diffraction (PED) can provide ideal resolution (~2 nm), making it of boundaries (HAGBs) with angles greater than 15◦ accounted for 93.06%
great importance to study the severely deformed grains and nano- of the total, out of these, approximately 46.6% (ranging from 55◦ to
crystalline structures in HSM [22,23]. Xu et al. [24] investigated the 63.5◦ ) were twin boundaries (TBs). Fig. 1(d) depicts the inverse pole
formation mechanism of nano-scale crystalline structure in a highly figure, indicating the presence of weak texture in the original micro­
perturbed surface layer by PED test. The high strain rate and the pre­ structure. Fig. 1(e) shows the results from analyzing the average grain
cipitation of second-phase particles were considered to be responsible diameter and grain number. The volume fraction of small grains with
for the formation of nano-scale grains. Finite element (FE) simulation is diameters less than 5 μm is approximately 25% and the calculated
also an effective method for studying the microstructure evolution in average grain size is approximately 13.6 μm.
HSM. Ding et al. [25] simulated the grain-size evolution for cutting Al Fig. 2 presents the actual experimental setup. The workpiece was
6061 T6 and OFHC Cu. The results indicated that a small strain, high pregrooved to prefabricate the cutting width of 4.0 mm for the orthog­
cutting speed, and high cutting temperature contributed to the forma­ onal cutting. The cutting forces were measured using a Kistler force test
tion of a coarser elongated grain structure. Xu et al. [26] investigated the system. The average cutting force was selected from the end of stable
grain refinement mechanism in chip for cutting Ti6Al4V alloy by cutting section and calculated as the result. After each cutting experi­
combining FE and cellular automata (CA) methods, proving that the ment, the cutter was replaced by a new cutter to avoid the effects of tool

Table 2
Chemical composition of Inconel 718 alloy.
Element C Cr Mo Nb Ti Al Ni P B

Wt% 0.015~0.060 17.00~21.00 2.80~3.30 5.00~5.50 0.75~1.15 0.30~0.70 50.00~55.00 0.020~0.027 0.007~0.011
Element Fe Mn Si Ta S Mg Cu Ca Co
≤
Wt% Balance
0.35 0.35 0.10 0.0020 0.005 0.30 0.005 1.0
Element Sn Pb Se Bi Ag Te Tl O N
≤
Wt%
0.005 0.0005 0.0003 0.00003 0.0005 0.00005 0.001 0.005 0.10

2
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 1. Original microstructure structure of Inconel 718 alloy. (a) IPF map, (b) Grain boundary map, (c) Inverse pole map, (d) The {100},{110},{111}pole figures, (e)
A histogram of grain diameters, (f) Grain boundary orientation distribution.

wear on the next experimental data. The VB of the flank face of the cutter performed before observing the chip morphology and dimensions using
was measured using the IFM-G4. The average value from five mea­ the IFM-G4. The microstructures of the chips and machined subsurfaces
surements taken at different locations was used as the result. The chips were observed through EBSD, TEM, and PED tests. Considering that the
collected were inlaid into an edge-preserving resin material as speci­ thickness of the chip could be excessively small, electroplating and
mens, Then, grinding and polishing procedures were sequentially cross-section polishing techniques were adopted to prepare the sample

3
A. Shu et al. Materials Characterization 207 (2024) 113559

⎡ ⎛ ⎞⎤
[ ( )]
ε̇ T − Tr
εJC = [d1 + d2 exp(d3 η) ] 1 + d4 ln⎝ ⎠ ⎦ 1 + d5
⎣ (2)
ε̇0 Tm − Tr

Here, di (i = 1,2,3,4,5) are the failure parameters typically obtained

through mechanical experiments with notched and axisymmetric spec­
imens under different strain rates and temperatures. ε̇0 is the reference
strain rate, η represents the stress triaxiality, which can be calculated by
the ratio of the hydrostatic pressure to the equivalent stress. The pa­
rameters of the Johnson-Cook damage model for the Inconel 718 alloy
are listed in Table 4.
During the cutting process, the combination of the high strain rate,
large strain, and high temperature usually leads to the occurrence of
DRX behavior. Based on the theory proposed by Zener and Hollomon
[32], the evolution of DRX volume fraction and grain size with grain
nucleation and growth can be calculated.
The DRX behavior occurs only when the strain reaches a critical
value. For Inconel 718 alloy, the critical strain proposed by Jafarian
et al. [30], who amended the model of Samantaray et al. [33] by adding
Fig. 2. Experimental setup of the orthogonal cutting experiment.
a calibration constant c to match the machining process, was adopted.
The critical strain εc can be defined as follows:
Table 3 εc = a1 Z m1 /c (3)
Parameters of material constitutive model of Inconel 718 alloy [30].
A B C n m ε̇0 Other constants where a1 and m1 are material parameters, c is the calibration constant, Z
(MPa) (MPa) is the physical parameter of the Zener-Hollomon model, which can be
1241 622 0.0134 0.6522 1.3 1 D = 0.6, S = 0, s = 5, r = expressed as
1.0 ( )
Qact
Z = ε̇exp (4)
RT
for EBSD testing. For preparing the sample for EBSD testing, the
machined surface was electroplated and electrolytically polished. In where Qact is the apparent deformation-activation energy. Geng et al.
contrast, a focused ion beam (FIB) technique with the FEI Helois 5 UX [34] reported that the average Qact at different critical strains for Inconel
was employed to prepare the sample for the TEM and PED testing. 718 was 411 kJ/mol. T and R represent the absolute temperature of
deformation and gas constant, respectively.
According to the modified Avrami model [35], the volume fraction of
2.2. FE orthogonal cutting model of microstructure evolution in HSM DRX can be expressed as follows:
[ ( ) ]
ε − εc kd
2.2.1. Constitutive model XDRX = 1 − exp − βd (5)
ε0.5 − εc
To simulate the metal cutting process, it is crucial to use an appro­
priate material constitutive model able to accurately capture the mate­
where βd and kd are the material parameters. ε0.5 represents the strain
rial behaviors under high-temperature, high-strain rate, and large-strain
required for a 50% volume fraction of DRX and can be defined as
conditions [28]. In the present study, a modified Johnson-Cook model
follows:
known as the TANT model [29], was employed. Considering the effects
of the temperature-dependent flow stress softening, the TANT model can ε0.5 = a5 Z m5 (6)
be expressed as follows:
⎡ ⎛ ⎞⎤ Here, a5 and m5 are the material parameters. The sizes of the newly
ε̇
[ (
T− Tr
)m ][ [ (
1
)]s ] formed DRX grains after nucleation and growth can be calculated as
σ =(A+Bεn )⎣1+Cln⎝ ⎠⎦ 1−
ε̇0 Tm − Tr
D+(1− D) tanh
(ε+S) r follows:
ddrx = a8 Z m8 (7)
(1)
In the above, σ is the equivalent flow stress, ε is the equivalent plastic In the above, a8 and m8 are material parameters. The final average
strain, ε̇ is the equivalent strain rate, ε̇0 is the equivalent reference strain grain size is the average of the remaining original grain size and newly
rate, T is the workpiece temperature, Tr is the room temperature, and Tm formed DRX grain size.
is the melting temperature of the workpiece material. A, B, n, C and m d = d0 (1 − XDRX ) + ddrx XDRX (8)
represent the yield strength, strain hardening modulus, strain hardening
coefficient, strain rate sensitivity coefficient, and thermal softening co­ Here, d0 denotes the original grain size. The DRX kinetic model pa­
efficient, respectively. D, S, s, and r are constitutive model coefficients rameters calculated by Geng et al. [34] and Jafarian et al. [30] were
introduced to describe the dynamic behaviors of the material. The model adopted and are listed in Table 5.
parameters are listed in Table 3. To obtain the strain, strain rate, and cutting temperature in HSM and
In the cutting simulation process, a chip separation criterion should predict the grain size in the ABAQUS/Explicit software, the VUMAT user
be applied to simulate the formation of the chip. In the present FE model, subroutine was developed in FORTRAN by implementing the TANT and
the Johnson-Cook damage strain model was used to describe the ma­ DRX kinetics models.
terial damage evolution process. This model simultaneously accounts for
the stress triaxiality, strain rate, and temperature. Based on the Johnson- 2.2.2. FE model for orthogonal cutting
Cook failure criterion, the equivalent plastic strain for the damage An FE method using the Lagrangian formation in the ABAQUS/
initiation as determined by the stress state is assumed as follows: Explicit software was adopted to obtain the deformation process, the

4
A. Shu et al. Materials Characterization 207 (2024) 113559

Table 4 Table 6
Fracture constants of JC fracture model for Inconel 718 alloy [31]. Physical and mechanical properties of Inconel 718 alloy [36].
d1 d2 d3 d4 d5 Physical properties Values Physical properties Values

0.04 0.75 − 1.45 0.04 0.89 Poission’s ratio 0.3 Thermal conductivity (W/m/ C)
◦
14.7
Density (kg/m3) 8240 Melting temperature (◦ C) 1310
Elastic modulus (Gpa) 199.9 Thermal expansion coefficient 11.8e-6
Table 5 Specific heat (J/kg/◦ C) 435 / /
DRX kinetics model parameters for Inconel 718 alloy.
a1 m1 c a5 m5 βd kd 18.92%, 3.5% and 19.38%, respectively. By the comparative analysis
0.00234 0.1293 10 7
0.018063 0.08518 0.693 1.92 above, the strain, strain rate, and cutting temperature as calculated by
the FE model are considered as reliable.

related 2D orthogonal cutting model is shown in Fig. 3. The FE model

3. Grain refinement mechanism in chip and machined
was established using a coupled dynamic temperature-displacement
subsurface
analysis procedure. The workpiece and cutter were modeled using the
CPE4RT element. The cutter was considered as a rigid body and was set
The FE method is better at demonstrating the distribution of the
as a wear tool considering the distinct tool wear for HSM Inconel 718
grain size in different regions of a workpiece at the macroscale. In
alloy. The rake angle, relief angle, and edge radius were same as the
addition, the deformation parameters, which are closely related to the
experimental parameters. The VB of the flank faces were measured to be
DRX behavior, can be obtained by FE simulations. As described in this
within 75–100 μm, in the present analysis, the average value was
section, FE and experimental methods were combined to comprehen­
calculated as the result (90 μm). The bottom and left sides of the
sively reveal the microstructural characteristics and grain refinement
workpiece were fixed to prevent vertical and horizontal movement. The
mechanism induced by the DRX behavior in the chip and at different
cutting speed direction was parallel to the bottom of the workpiece and
depths of the machined subsurface.
the cutting parameters in the simulation were consistent with those in
the cutting experiments. The contact friction condition at the tool-chip
3.1. Microstructure characteristics and grain refinement mechanism of
interface was set based on Coulomb’s friction law. The physical and
chip
mechanical properties of the Inconel 718 alloy used in the FE model are
presented in Table 6.
Fig. 6 shows experimental and simulated microstructure in chip at V
= 100 m/min and t = 0.1 mm. As shown in Fig. 6(b), the grain defor­
2.3. Validation of FE orthogonal cutting model mation in chip mainly occurred in the shear band and secondary
deformation zone, corresponding to the adiabatic shear deformation and
The simulated cutting forces and chip morphologies were compared friction behavior at tool-chip interface, respectively. Fig. 6(a) and Fig. 6
with the experimental results to evaluate the effectiveness of the FE (c) show the measured microstructure at the tool-chip interface zone and
model. The experimental and simulated principal and thrust cutting in the shear band by EBSD test, respectively, in which a step size of 30
forces for a unit cutting width are compared in Fig. 4. It can be sum­ nm was adopted. As shown in Fig. 6(a), nanocrystallites were formed at
marized that the simulated cutting forces show the same trend as those the tool-chip interface zone. The grains at a deeper depth were severely
of experiments. The average prediction error on the principal and thrust deformed and elongated along the cutting direction, certain recrystal­
cutting forces are 6.69%, 7.43%, respectively, which are considered to lized grains were also formed at the grain boundary. In the shear band,
be within an acceptable range. Fig. 5 compared the experimental and the grains are severely elongated and deformed along the shear direc­
simulated chip morphologies at V = 300 m/min,t = 0.025 mm, V = 100 tion, the grain refinement phenomenon is also evident, as shown in
m/min,t = 0.05 mm, V = 400 m/min,t = 0.05 mm, and V = 100 m/min,t Fig. 6(c). Fig. 6(d) and Fig. 6(e) show the distributions of simulated grain
= 0.1 mm. It is evident that the structures of the simulated and exper­ size and DRX volume fraction. Consistent with the experimental results,
imental serrated chips show similar characteristics. In terms of the peak, the simulated DRX behavior in chip also mainly occurs in the shear band
valley, and space, which are commonly used for quantifying serrated and tool-chip interface zone. In the simulated cloud map, the material
chips dimensions and noted in Fig. 5(d), the chips at V = 400 m/min,t = grain size away from these two regions was essentially unchanged,
0.05 mm and V = 100 m/min,t = 0.1 mm were analyzed, and they are which was equal to the average original grain size. From the grain size
listed in Table 7. For V = 400 m/min,t = 0.05 mm, the FE model shows distribution shown in Fig. 6(d), it is evident that the average grain size at
15.43%, 11.26%, and 10.72% prediction error on the peak, valley, and the tool-chip interface zone was smaller than that in the shear band. The
space of the serrated chips. The data for V = 100 m/min,t = 0.1 mm are DRX volume fraction was higher than that in the shear band, as shown in

Fig. 3. Finite element orthogonal cutting model.

5
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 4. Comparison of simulated and experimental cutting forces. (a) Simulated and experimental principal cutting forces, (b) Simulated and experimental thrust
cutting forces.

Fig. 5. Comparison of simulated and experimental chip morphologies. (a) Chip morphologies at V = 300 m/min,t = 0.025 mm, (b) Chip morphologies at V = 100 m/
min,t = 0.05 mm, (c) Chip morphologies at V = 400 m/min,t = 0.05 mm, (d) Chip morphologies at V = 100 m/min,t = 0.1 mm.

Fig. 6(e). According to the DRX kinetics model, the DRX behavior is
Table 7
closely related to the temperature, strain, and strain rate.
Comparison of chip dimensions between simulation and experiment.
The simulated strain, strain rate, and cutting temperature in the
V = 400 m/min t = 0.05 mm V = 100 m/min t = 0.1 mm serrated chip at V = 100 m/min and t = 0.1 mm are displayed in Fig. 7.
Peak Valley Space Peak Valley Space Evidently, the strain, strain rate, and temperature in the tool-chip
(μm) (μm) (μm) (μm) (μm) (μm) interface zone are higher than those in the shear band. When the
Experiment 70.59 56.75 25.42 152.84 102.03 63.57 machining temperature exceeds the recrystallization temperature and
Simulation 81.48 63.14 28.14 123.92 98.46 51.25 lasts for a certain time and the strain reaches the critical strain, the DRX
Difference 15.43% 11.26% 10.72% 18.92% 3.50% 19.38% process can be fully completed. The differences in the deformation
condition lead to the different final average grain sizes induced by DRX
behavior. The higher strain rate enhances the grain nucleation and
inhibit grain growth, resulting in a smaller DRX grain size [26]. The

6
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 6. Experimental and simulated microstructure in chip at V = 100 m/min and t = 0.1 mm.
(a) Experimental microstructure at secondary deformation zone, (b) Experimental microstructure in chip, (c) Experimental microstructure in shear band, (d) Simulate
grain size distribution (μm), (e) Simulate DRX volume fraction distribution.

Fig. 7. Simulated results of deformation parameters in chip at V = 100 m/min and t = 0.1 mm.
(a) Simulated equivalent strain distribution, (b) Simulated equivalent strain rate distribution, (c) Simulated temperature distribution.

7
A. Shu et al. Materials Characterization 207 (2024) 113559

increase in both the DRX volume fraction and DRX grains eventually distribution. The microstructures in Figs. 10(a) and 11(a) reveal the
leads to the decrease in the average grain size. presence of amorphous structures, nanocrystallite regions, refined
grains, and elongated grains, which are labeled as I, II, III, and IV,
3.2. Grain refinement mechanism of the machined subsurface respectively. In addition, the degree of grain deformation and refine­
ment at V = 400 m/min,t = 0.05 mm exceeds that at V = 400 m/min,t =
3.2.1. Microstructural characteristics of the machined subsurface 0.025 mm. Several specific areas were selected for further analysis to
Figs. 8 and 9 show the microstructure of the machined subsurface show the microstructure evolution, the obtained images are shown in
obtained by the EBSD tests at V = 400 m/min,t = 0.05 mm and V = 400 Figs. 12 and 13.
m/min,t = 0.025 mm, respectively. Figs. 8(a) and 9(a) show the cross- Figs. 12(a)-(c) and 13(a)-(c) are TEM images of the selected areas in
sectional images of the microstructure of machined subsurface with a Zone I in Figs. 10(a) and 11(a), respectively, revealing the amorphous
step size of 0.3 μm. The grains in the machined subsurface are severely structure. Fig. 12(a) shows the amorphous structure at V = 400 m/min,t
elongated along the cutting direction. Moreover, the grains have un­ = 0.05 mm is evident, and its thickness is approximately 100 nm; the
dergone significant refinement. Based on the morphological character­ amorphous area is small at V = 400 m/min,t = 0.025 mm, as shown in
istics of the refined grains, it can be inferred that DRX occurred in HSM, Fig. 13(a). It indicates that a larger uncut chip thickness is conducive to
resulting in the formation of DRX grains. The EBSD images also reveal the formation of an amorphous structure. Figs. 12(b), 12(c), 13(b), and
that the depth of the machined subsurface affected by HSM is about 10 13(c) show that there are significant differences between the amorphous
μm, where significant grain refinement occurred. and crystalline structures. To further investigate the formation mecha­
Considering the effect of the step size on the calculation accuracy, a nism of the amorphous structure, the elemental compositions at the
step size of 30 nm, which is close to the limitation of the EBSD technique, interface were analyzed using EDS mapping in a TEM environment, the
was adopted to further investigate the microstructural characteristics of obtained images of certain elements are shown in Figs. 14 and 15. The
the deformation layer. Finally, the EBSD maps with the sizes of 25 μm × results showed that the contents of certain compositional elements have
10 μm were obtained, as shown in Figs. 8(b) and 9(b), respectively. changed. Fe and Ni in the amorphous region are greatly reduced
Evidently, the microstructure of the machined subsurface exhibits a compared with that in the crystalline structure, while O, C, P, and Ba are
gradient distribution along the depth direction. The deformation of the enriched in the amorphous region. In addition, K, which was not found
machined subsurface can be divided into two zones. According to the in the original material, was observed in both the amorphous and
statistical results of the grain-size distribution in Figs. 8(d) and 9(d), crystalline region. It indicates that the machined surface absorbs ele­
nanocrystallites were formed in the machined near-surface, and a drastic ments from the external environment in HSM. The formation of amor­
crystalline modification may have occurred in this zone. The grains in phous structure is due to the severe plastic deformation and
the deeper zone were severely deformed, elongated along the cutting microstructural alterations. It is not only related to the high heat during
direction, and refined to a certain extent. Certain recrystallized grains cutting process but also to the thermal conductivity of workpiece ma­
were distributed at the grain boundaries. In addition, color differences terial [9]. In HSM, the material undergoes a process of rapid heating and
can be clearly observed within the same grain, where different colors cooling, and the amorphous structure is formed under the effect of
represent different crystal orientations. Therefore, it can be deduced that quenching mechanism. In addition, mechanical effect also plays an
high-density dislocations and various low-angle grain boundaries important role in the formation of amorphous structure. Thus, it can be
(LAGBs) with misorientations in the range of 2◦ -15◦ were present inside concluded that the amorphous structure is the combined result of the
the grains, which are typical features of subgrain structures. Figs. 8(e) elemental mixing, quenching effect, mechanical effects, plastic defor­
and 9(e) show the grain boundary maps. The specimens at V = 400 m/ mation, and microstructural alterations (DRX behavior, phase trans­
min,t = 0.05 mm and V = 400 m/min,t = 0.025 mm contained formation, and so on) [38].
approximately 62.25% and 73.12% LAGBs, respectively, indicating that Figs. 12(d)-(f) and 13(d)-(f) show the nanocrystallites on the
a large number of subgrains had not yet been transformed into complete machined near-surface. Equiaxed ultrafine grains with well-defined
grains. Compared with the original microstructure, it can be concluded boundaries are clearly observed. The corresponding SAED images in
that the HSM process promoted the formation of subgrains. In addition, Figs. 12(d) and 13(d) show almost continuous polycrystalline circular
the specimens contained approximately 7.32% and 9.13% TBs, respec­ distributions, indicating the presence of nano-scale grains. The nano­
tively, they were significantly reduced compared to those of the original crystallites at V = 400 m/min,t = 0.025 mm were slightly larger than
microstructure (46.6% TBs). In HSM, the severe plastic deformation those at V = 400 m/min,t = 0.025 mm, which was consistent with the
refines the twins and original grains into fine subgrains with the increase results of the EBSD tests. Figs. 12(e), 12(f), and 13(e) are the focally
in strain. As reported in [37], a change in the twins is the leading process amplified regions with different multiples. A high density of dislocations
of grain refinement. A comparison of Figs. 1(d), 8(c), and 9(c) shows that and stacking faults at both the interior and boundary of the grains can be
the texture intensity of the Inconel 718 alloy changed after the HSM. The observed. Therefore, it can be deduced that the refined grains in these
texture intensity of the machined surface at V = 400 m/min,t = 0.05 mm zones result from the evolution of a high dislocation density. From the
increased significantly, whereas it changed slightly at V = 400 m/min,t HRTEM image and corresponding SAED image in 13(f), twins can be
= 0.025 mm. This is because the deformation and DRX behavior cause observed.
the grains to rotate and have a certain tendency, resulting in an Figs. 12(g)-(i) and 13(g)-(i) show the elongated and refined grains,
enhanced texture intensity. The high-density dislocation motivation is respectively. With increasing measurement depth, the equiaxed ultra­
most likely generated in terms of the microstructural features and fine nanocrystals gradually transformed into elongated, deformed, and
drastic deformation. TEM has an irreplaceable advantage in the obser­ refined microcrystals. The corresponding SAED images in Figs. 12(g)
vation of crystalline structures. Consequently, the microstructure of the and 13(g) show discontinuous circular distributions, indicating the
machined near-surface was further investigated by TEM test, and the coexistence of nanocrystallites and subgrains. Evidently, the grains were
selected area of the FIB sample is shown in Fig. 9(b). severely deformed and elongated along the cutting direction. As re­
The microstructural characteristics of the machined near-surface at ported by Liang et al. [39], the formation of elongated subgrains occurs
V = 400 m/min, t = 0.05 mm and V = 400 m/min, t = 0.025 mm as between DRX and dynamic recovery (DRV), which is a typical feature of
measured by the Double Cs Corrector Transmission Electron Microscope the ultrahigh shear strain in HSM. In addition, the grain size in the zone
are shown in Figs. 10 and 11, respectively. From the bright-field and at V = 400 m/min,t = 0.025 mm is evidently larger than that at V = 400
dark-field TEM images shown in Figs. 10(a), 11(a), 10(b), and 11(b), it m/min, t = 0.05 mm, and the degree of grain deformation is less severe.
can be observed that both the microstructures of the machined near- Figs. 12(h), 12(i), 13(h) and 13(i) show the microstructures of selected
surface at these two cutting conditions show a significant gradient regions in Figs. 12(g) and 13(g) with different multiples, respectively.

8
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 8. Microstructure of the machined surface at V = 400 m/min and t = 0.05 mm observed by EBSD test. (a) Cross-sectional image showing the microstructure of
the machined subsurface, (b) Microstructural characteristic of the deformation layer, (c) {100},{110},{111} pole figures of the deformation layer, (d) A histogram of
grain diameters of the deformation layer, (e) Grain boundary orientation distribution of the deformation layer.

9
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 9. Microstructure of the machined surface at V = 400 m/min and t = 0.025 mm observed by EBSD test. (a) Cross-sectional image showing the microstructure of
the machined subsurface, (b) Microstructure characteristic of the deformation layer, (c) {100},{110},{111} pole figures, (d) A histogram of grain diameters, (e) Grain
boundary orientation distribution.

10
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 10. Microstructure of the machined surface at V = 400 m/min, t = 0.05 mm observed by TEM.
(a) Bright-field TEM image, (b) Dark-field TEM image.

Fig. 11. Microstructure of the machined surface at V = 400 m/min, t = 0.025 mm observed by TEM.
(a) Bright-field TEM image, (b) Dark-field TEM image.

High-density dislocation characteristics, such as stacking faults and 3.2.2. Quantitative characterization of nanocrystalline structure
dislocation walls, are observed, These are generally conducive to the The EBSD image in Fig. 8(b) and the TEM morphology in Fig. 10
formation of elongated nanocrystallite with sizes of tens of nanometers demonstrate that a large number of nanocrystallites and subgrains were
[39]. generated in the machined near-surface at V = 400 m/min,t = 0.05 mm.
As the measurement depth increased, the deformation of the grains Although TEM technology can provide high-resolution imaging of a
became less drastic, as shown in Figs. 12(j)-(l) and 13(j)-(l). The corre­ microstructure, it cannot quantitatively characterize the grain orienta­
sponding SAED images in Figs. 12(j) and 13(j) indicate the consistency tion distribution characteristics [40]. Similarly, the resolution of EBSD
of the grain orientation and the existence of grains at the micron level. technology is limited. Overall, the formation mechanism of nano­
Figs. 12(k) and 13(k) are the focally amplified regions showing dislo­ crystallite structures in the machined near-surface of Inconel 718 alloy
cation tangles and lines, which may cause local misorientation inside the has not been well explained [41,38]. Thus, the PED test was applied to
grains. Figs. 12(l) and 13(l) show the HRTEM images and corresponding quantitatively characterize the nanocrystallite structure and reveal its
SAED images. Two kinds of diffraction spots (the bright and less bright formation mechanism [42]. The obtained images with a step size of 5 nm
spots) can be observed evidently in the SAED images, indicating the γ are shown in Figs. 16(a) and 16(b). Consistent with the results of TEM
and γ’ phases in the material of machined surface. and EBSD tests, the grains are significantly refined and elongated. The
From the above analysis, it can be concluded that the microstructure statistical analysis of the grain size in Fig. 16(d) shows that almost all the
distributions of the machined subsurface at V = 400 m/min,t = 0.05 mm grains have sizes smaller than 100 nm, certain grains have sizes smaller
and V = 400 m/min,t = 0.025 mm show similar characteristics. More­ than 10 nm, which is much smaller than the average grain size of the
over, the grain deformation and refinement are more significant at V = original microstructure (13.6 μm). According to Xu et al. [24], a high
400 m/min,t = 0.05 mm than those at V = 400 m/min,t = 0.025 mm. strain rate in HSM is the key factor in the formation of nano-scale grains.
Therefore, the grain refinement mechanism of the machined subsurface The strain, strain rate and temperature would be further analyzed
at V = 400 m/min,t = 0.05 mm is discussed in more detail below. through an FE simulation. Fig. 16(c) shows the distribution of grain
boundaries, the LAGBs accounted for approximately 35.27%, which is

11
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 12. Microstructure at different depths of the machined surface at V = 400 m/min,t = 0.05 mm observed by TEM test. (a)-(c) Selected areas revealing the
amorphous structure, (d)-(f) Selected areas showing the nanocrystallites, (g)-(i) Selected areas showing the elongated and refined grains, (j)-(l) Selected areas
showing the less deformed grains.

significantly lower than the statistical result (62.25% of LAGBs) in Fig. 8 the dislocation motion and large subgrain deformation in HSM. The
(e). This indicates that there are a large number of subgrains that have distribution of dislocation density would be analyzed using the PED test
not yet transformed to HAGBs and complete grains in the deeper zone. results.
The high-density LAGBs in the machined near-surface are the result of

12
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 13. Microstructure at different depths of the machined surface at V = 400 m/min, t = 0.025 mm observed by TEM test. (a)-(c) Selected areas revealing the
amorphous structure, (d)-(f) Selected areas showing the nanocrystallites, (g)-(i) Selected areas showing the elongated and refined grains, (j)-(l) Selected areas
showing the less deformed grains.

13
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 14. EDS mapping images showing the elemental composition difference between amorphous and crystalline structure at V = 400 m/min,t = 0.05 mm obtained
in a TEM environment.

Fig. 15. EDS mapping images showing the elemental composition difference between amorphous and crystalline structure at V = 400 m/min,t = 0.025 mm obtained
in a TEM environment.

3.2.3. Geometrically necessary dislocation density distribution a critical value, new grains begin to form at the boundary of the original
The evolution of the dislocation density is the basis for the evolution grain. They grow along with the difference in the dislocation density
of a material microstructure [43,44]. Severe plastic deformation and between the adjacent grains as the driving force. In contrast to the dDRX
rapid heating can lead to the evolution of the dislocation density in HSM, grains, the cDRX grains nucleate inside the grain. As shown in Fig. 17(b),
this is generally considered to be the driving force for the grain nucle­ when the dislocation density inside the grain reaches a critical value, a
ation and growth during the DRX process [24]. DRX behavior can be local misorientation is generated. With the continuous accumulation of
triggered once the dislocation density inside the grain or at the grain dislocations, LAGBs and subgrains are generated. Subsequently, the
boundary reaches the critical condition for DRX. subgrains rotate and coalesce with neighboring subgrains under the
As reported by Fanfoni and Tomellini [45], DRX behaviors can be influence of continuous dislocation accumulation, resulting in the for­
divided into two main categories: discontinuous DRX (dDRX) and mation of larger subgrains, LAGBs, and eventually HAGBs and complete
continuous DRX (cDRX). Fig. 17(a) presents a schematic of the dDRX grains.
mechanism. When the dislocation density at the grain boundary reaches The geometrically necessary dislocation (GND) density is a key factor

14
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 16. Microstructure of machined near-surface at V = 400 m/min and t = 0.05 mm observed by PED test. (a) PED orientation distribution map, (b) Reliablity map,
(c) Grain boundary orientation distribution of the machined near-surface, (d) Distribution map of grain sizes.

Fig. 17. Schematic diagrams of DRX mechanism caused by dislocation density evolution. (a) dDRX mechanism, (b) cDRX mechanism.

15
A. Shu et al. Materials Characterization 207 (2024) 113559

for analyzing the grain refinement mechanism [46]. It can be calculated Fig. 18(c). A dislocation concentration can be observed obviously, which
based on the crystal orientation data obtained from the EBSD, TEM and is conducive to the generation of LAGBs [24]. Moreover, it can be seen
PED technologies. The lattice curvature and dislocation density tensor that the GND density reaches 1 × 1017/m2, which is significantly higher
can be derived based on the orientation differences between adjacent than that in the deeper zone. The high GND density is induced by high
points [47]. The EBSD test results shown in Fig. 8(b) were used to strain rate and is a key factor for the final grain refined to nano-scale.
calculate the GND density of the machined subsurface, the calculated
results are mapped in Fig. 18(a). A high GND density can be observed in 3.2.4. Simulated deformation condition at the machined subsurface
the machined near-surface, which will be analyzed in detail later using In HSM, the critical conditions of different recrystallization modes
the test result of PED. In the deeper zone, the GND density at the grain depend on the deformation conditions, such as the temperature, strain,
boundary reaches 1 × 1016/m2, which is much higher than that in the and strain rate [48]. Therefore, the microstructural evolution mecha­
grain interior (1 × 1014/m2). In addition, the GND density in a certain nism of the machined subsurface can be analyzed by establishing the
area inside the grain also reaches 1 × 1016/m2, demonstrating that the relationships between the simulated deformation parameters distribu­
grain refinement in this zone is mainly derived from cDRX behavior. In tion and experimental microstructure distribution. Fig. 19 shows the
this process, the GNDs provide continuous power for cDRX behavior. simulated plastic strain and cutting temperature distributions of the
Regarding the distribution of GND density in the machined near-surface, machined surface. It can be seen that the plastic strain and cutting
the test crystal orientation data obtained from PED, as shown in Fig. 18 temperature exhibited significant gradient distribution characteristics
(b), was utilized for the calculations, whose results are presented in along the depth direction. After the thermal and mechanical load are

Fig. 18. Distribution of GND density in the machined subsurface at V = 400 m/min and t = 0.05 mm. (a) Calculated GND density in the machined subsurface by the
test result of EBSD, (b) Test result of PED, (c) Calculated GND density in the machined near-surface by the test result of PED.

16
A. Shu et al. Materials Characterization 207 (2024) 113559

Fig. 19. Simulated results of deformation parameters of the machined surface at V = 400 m/min and t = 0.05 mm. (a) Simulated equivalent strain distribution, (b)
Simulated temperature distribution.

loaded, the plastic strain and cutting temperature began to decrease. At which lacks a large number of grain boundaries required for dDRX grain
the machined surface, the maximum strain reached 4.43 and the cutting nucleation. cDRX behavior is activated under drastic thermoplastic
temperature was up to 1113 ◦ C. In general, an extremely high cutting deformation conditions, and an increase in the grain boundary propor­
speed (400 m/min) leads to an extremely high strain rate and change in tion promotes cDRX grain nucleation. Thus, it can be concluded that the
the cutting temperature. In HSM, an extremely high strain rate is dDRX grains in the nanocrystallites layer nucleate at the cDRX grain
considered to be the reason for the large number of dislocations and boundaries and the nanocrystallites result from both cDRX and dDRX
ultrafine nanocrystallites, and the rapid temperature increase has a behaviors. The existence of high-density dislocations and a large number
significant effect on the grain deformation. Considering the strong of LAGBs and subgrains in the refined and deformed grain layers indi­
dependence of the DRX behavior on the strain, strain rate, and tem­ cate that the refined grains in this zone are the result of cDRX behavior.
perature conditions, different DRX mechanisms may be activated in
different depth regions of the machined subsurface. 3.2.6. Discussion of grain refinement mechanism
From the above analysis, it can be concluded that in HSM, the
3.2.5. Characteristics of DRX at different depths machined subsurface is subjected to a gradient distribution of the
Fig. 20 shows the bright-field and dark-field TEM images of selected thermomechanical load, resulting in a gradient distribution in the
areas in the nanocrystallite layer. Certain equiaxed ultrafine nano­ microstructure [49]. The original grains in the machined near-surface
crystallites can be found around the grain boundaries of larger grains, region are refined into nanocrystallites. In this section, the grain
whcih is a typical feature of the dDRX behavior. Based on the analysis in refinement mechanisms at different depths of the machined subsurface
Section 3.2.3, the dDRX grains can only nucleate at the grain bound­ are discussed.
aries. However, the grain size of the original microstructure structure As the cutting process began, the original grains in the machined
(13.6 μm) are much larger than that of the nanocrystallites (6–100 nm), subsurface elongated and deformed along the cutting direction,

Fig. 20. cDRX grains in nanocrystallite layer. (a) Bright-field TEM image, (b) Dark-field TEM image.

17
A. Shu et al. Materials Characterization 207 (2024) 113559

resulting in numerous dislocations. The cDRX behavior was activated reached a high level, indicating a cDRX behavior. The distribu­
under the drastic thermoplastic deformation conditions. With a contin­ tion of the GND density indicates that a high GND density is
uous increase in the plastic strain, the elongated and deformed grains in induced by a high strain rate. In HSM, the high GND density
the nanocrystallite layer were fragmented into fine subgrain structures, provides continuous power for DRX.
resulting in a large number of LAGBs and a decrease in TBs. Owing to the (5) In HSM, the machined subsurface is subjected to a gradient dis­
lower strain, strain rate, and cutting temperature in the deeper zone than tribution of the thermomechanical load, resulting in a gradient
those in the machined near-surface, the dislocation density and speed of distribution in the microstructure. The nanocrystallites are the
grain refinement were also lower. Thus, only a large number of dislo­ result of both cDRX and dDRX behaviors, whereas the refined
cations are generated in the deeper zone. grains in the deeper zone are the result of cDRX behavior. The
With further deformation, the size of the subgrains continuously different DRX mechanisms depend on the deformation conditions
decreased. The refined subgrains that reached the critical size could only at different depths of the machined surface.
adapt to deformation by rotation, resulting in a continuous increase in
the local misorientation inside the grain. The key factor for subgrains to Author contribution
become the core of recrystallized grains is the formation of HAGBs [18].
When the LAGBs began to transform into HAGBs, complete cDRX grains Ailing Shu: methodology, investigation, formal analysis, and writing
were formed in the nanocrystalline layer. At this point, subgrains began draft. Junxue Ren: methodology, review, and editing. Jinhua Zhou:
to form in the deeper zone. project administration, methodology, experiment, formal analysis, re­
The newly formed cDRX grains in the nanocrystalline layer gener­ view, and editing. Zongyuan Wang: experiment and review.
ated a large number of grain boundaries and high-density dislocations,
providing favorable conditions for dDRX grain nucleation. Thus, the
dDRX grains begin to nucleate at the grain boundaries of cDRX grains. At Declaration of Competing Interest
this instant, the subgrains in the deeper zone were gradually trans­
formed into complete cDRX grains. The authors declare that they have no known competing financial
Finally, DRV and the growth of dDRX grains occurred in the cooling interests or personal relationships that could have appeared to influence
stage after cutting, these processes occurred mainly through the grain the work reported in this paper.
boundary migration mechanism [50] and were closely related to the
temperature. Eventually, completely dDRX grains were formed. In the Data availability
deeper zone, the dDRX grains could not be generated owing to the low
temperature and strain; thus, only cDRX grains were formed. Data will be made available on request.

4. Conclusion Acknowledgments

In this study, the microstructural characteristics and grain refine­ This work is supported by the National Natural Science Foundation
ment mechanism in chips and different depths of machined surfaces of China (Grant No. 52075451), the National Science and Technology
were studied for the HSM Inconel 718 alloy. Advanced material testing Major Project (Grant No. J2019-VII-0001-0141), Science Center for Gas
technologies, including EBSD, TEM, and PED, were used to characterize Turbine Project (P2022-B-IV-012-001), the Aeronautical Science Foun­
the crystal structures. The deformation parameters related to the DRX dation of China (Grant No. 2019ZE053008), and the China Postdoctoral
behavior in HSM were obtained via FE simulation. The primary con­ Science Foundation (Grant No. 2020M683551).
clusions are summarized as follows:
References
(1) The DRX behavior in the chip primarily occurred in the shear
[1] E. Hosseini, V.A. Popovich, A review of mechanical properties of additively
band and secondary deformation zone. The degree of grain manufactured Inconel 718, Addit. Manuf. 30 (2019), 100877, https://doi.org/
refinement in the second deformation zone was found to be 10.1016/j.addma.2019.100877.
greater than that in the shear band. This can be attributed to [2] A.D. Bartolomeis, S.T. Newman, I.S. Jawahir, D. Biermann, A. Shokrani, Future
research directions in the machining of Inconel 718, J. Mater. Process. Technol.
differences in the deformation conditions, including the strain,
297 (2021), 117260, https://doi.org/10.1016/j.jmatprotec.2021.117260.
strain rate, and cutting temperature. It is believed that these [3] K. Mahesh, J.T. Philip, S.N. Joshi, B. Kuriachen, Machinability of Inconel 718: a
differences are the main reason for the varying final grain sizes critical review on the impact of cutting temperatures, Mater. Manuf. Process. 36
induced by the DRX behavior. (2021) 753–791, https://doi.org/10.1080/10426914.2020.1843671.
[4] J. Gubicza, L. Farbaniec, G. Csiszár, T. Sadat, H. Couque, G. Dirras, Microstructure
(2) The grains in the machined subsurface were found to be severely and strength of nickel subjected to large plastic deformation at very high strain
deformed and refined. The grain size in the machined near- rate, Mat. Sci. Eng. A. 662 (2016) 9–15, https://doi.org/10.1016/j.
surface region reached the nano-scale. Compared with the orig­ msea.2016.03.046.
[5] B. Wang, Z.Q. Liu, Y.K. Cai, X.C. Luo, H.F. Ma, Q.H. Song, Z.H. Xiong,
inal microstructure, a large number of LAGBs were generated, Advancements in material removal mechanism and surface integrity of high speed
and the TBs were significantly reduced. metal cutting: a review, Int J Mach Tool Manu 166 (2021), 103744, https://doi.
(3) The microstructure of the machined subsurface was gradiently org/10.1016/j.ijmachtools.2021.103744.
[6] B.X. Wang, G.C. Barber, F. Qiu, Q. Zou, H.Y. Yang, A review: phase transformation
distributed along the depth direction. Amorphous structures, and wear mechanisms of single-step and dual-step austempered ductile irons,
nanocrystallites, refined grains, and elongated grains were J. Mater. Res. Technol. 9 (2020) 1054–1069, https://doi.org/10.1016/j.
clearly observed. High-density dislocation characteristics such as jmrt.2019.10.074.
[7] C.Z. Duan, F.Y. Zhang, S.W. Qin, M.J. Wang, Modeling of dynamic recrystallization
stacking faults, dislocation slips, and dislocation walls were also in white layer in dry hard cutting by finite element-cellular automaton method,
observed. As the uncut chip thickness increased from 0.025 mm J. Mech. Sci. Technol. 32 (2018) 4299–4312, https://doi.org/10.1007/s12206-
to 0.05 mm, the amorphous region and degree of grain refine­ 018-0828-y.
[8] A. Rasti, M.H. Sadeghi, S.S. Farshi, An investigation into the effect of surface
ment and deformation significantly increased.
integrity on the fatigue failure of AISI 4340 steel in different drilling strategies,
(4) A concentration of the GND density was clearly observed in the Eng. Fail. Anal. 95 (2019) 66–81, https://doi.org/10.1016/j.
machined near-surface, and the GND density was higher than that engfailanal.2018.08.022.
in the deeper zone. The GND density at the grain boundaries was [9] Q.G. Yin, Z.Q. Liu, B. Wang, Q.H. Song, Y.K. Cai, Recent progress of machinability
and surface integrity for mechanical machining Inconel 718: a review, Int. J. Adv.
significantly higher than that in the grain interiors. In the deeper Manuf. Technol. 109 (2020) 215–245, https://doi.org/10.1007/s00170-020-
zone, the GND density in certain areas inside the grain also 05665-4.

18
A. Shu et al. Materials Characterization 207 (2024) 113559

[10] M.C. Hardy, M. Detrois, E.T. McDevitt, C. Argyrakis, V. Saraf, P.D. Jablonski, J. changes, Int. J. Mech. Sci. 88 (2014) 110–121, https://doi.org/10.1016/j.
A. Hawk, R.C. Buckingham, H.S. Kitaguchi, S. Tin, Solving recent challenges ijmecsci.2014.08.007.
forwrought Ni-base superalloys, Metall. Mater. Trans. A 51 (2020) 2626–2650, [31] B. Erice, F. Gálvez, A coupled elastoplastic-damage constitutive model with lode
https://doi.org/10.1007/s11661-020-05773-6. angle dependent failure criterion, Int. J. Solids Struct. 51 (2014) 93–110, https://
[11] A.L. Monaca, J.W. Murray, Z.R. Liao, A. Speidel, J.A. Robles-Linares, D.A. Axinte, doi.org/10.1016/j.ijsolstr.2013.09.015.
M.C. Hardy, A.T. Clare, Surface integrity in metal machining-part II: functional [32] C. Zener, J.H. Hollomon, Effect of strain rate upon plastic flow of steel, J. Appl.
performance, Int J Mach Tool Manu 164 (2021), 103718, https://doi.org/ Phys. 15 (1944) 22–32, https://doi.org/10.1063/1.1707363.
10.1016/j.ijmachtools.2021.103718. [33] D. Samantaray, S. Mandal, M. Jayalakshmi, C.N. Athreya, A.K. Bhaduri, V.
[12] Z.R. Liao, A.L. Monaca, J. Murray, A. Speidel, D. Ushmaev, A. Clare, D. Axinte, S. Sarma, New insights into the relationship between dynamic softening
R. M’Saoubi, Surface integrity in metal machining-Part I: fundamentals of surface phenomena and efficiency of hot working domains of a nitrogen enhanced 316L
characteristics and formation mechanisms, Int J Mach Tool Manu 162 (2021), (N) stainless steel, Mat. Sci. Eng. A-Struct. 598 (2014) 368–375, https://doi.org/
103687, https://doi.org/10.1016/j.ijmachtools.2020.103687. 10.1016/j.msea.2013.12.105.
[13] E. Brinksmeier, S. Reese, A. Klink, L. Langenhorst, T. Lübben, M. Meinke, D. Meyer, [34] P.H. Geng, G.L. Qin, J. Zhou, T.Y. Li, N.S. Ma, Characterization of microstructures
O. Riemer, J. Sölter, Underlying mechanisms for developing process signatures in and hot-compressive behavior of GH4169 superalloy by kinetics analysis and
manufacturing, Nanomanuf. Metrol. 1 (2018) 193–208, https://doi.org/10.1007/ simulation, J. Mater. Process. Technol. 288 (2021) 116879, https://doi.org/
s41871-018-0021-z. 10.1016/j.jmatprotec.2020.116879.
[14] Z.P. Pan, Y.X. Feng, S.Y. Liang, Material microstructure affected machining: a [35] G.Z. Quan, Y.P. Mao, G.S. Li, W.Q. Lv, Y. Wang, J. Zhou, A characterization for the
review, Manuf. Rev. 4 (2017) 5, https://doi.org/10.1051/mfreview/2017004. dynamic recrystallization kinetics of as-extruded 7075 aluminum alloy based on
[15] A. Huang, S.J. Fensin, M.A. Meyers, Strain-rate effects and dynamic behavior of true stress–strain curves, Comput. Mater. Sci. 55 (2012) 65–72, https://doi.org/
high entropy alloys, J. Mater. Res. Technol. 22 (2023) 307–347, https://doi.org/ 10.1016/j.commatsci.2011.11.031.
10.1016/j.jmrt.2022.11.057. [36] N. Chen, Y. Yuan, C. Guo, X.L. Zhang, X.Q. Hao, N. He, Design, optimization and
[16] Y. Guo, C. Saldana, W.D. Compton, S. Chandraseka, Controlling deformation and manufacturing of polycrystalline diamond micro-end-mill for micro-milling of
microstructure on machined surfaces, Acta Mater. 59 (2011) 4538–4547, https:// GH4169, Diam. Relat. Mater. 108 (2020), 107915, https://doi.org/10.1016/j.
doi.org/10.1016/j.actamat.2011.03.076. diamond.2020.107915.
[17] J.A. Hines, K.S. Vecchio, Recrystallization kinetics within adiabatic shear bands, [37] S. M’Guil, W. Wen, S. Ahzi, J.J. Gracio, R.W. Davies, Analysis of shear deformation
Acta Mater. 45 (1997) 635–649, https://doi.org/10.1016/S1359-6454(96)00193- by slip and twinning in low and high/medium stacking fault energy fcc metals
0. using the ϕ-model, Int. J. Plast. 68 (2015) 132–149, https://doi.org/10.1016/j.
[18] X.P. Ren, Z.Q. Liu, Microstructure refinement and work hardening in a machined ijplas.2014.03.020.
surface layer induced by turning Inconel 718 super alloy, Int. J. Miner. Metall. [38] D. Ulutan, T. Ozel, Machining induced surface integrity in titanium and nickel
Mater. 25 (2018) 937–949, https://doi.org/10.1007/s12613-018-1643-2. alloys: a review, Int J Mach Tool Manu 51 (2011) 250–280, https://doi.org/
[19] Q.Q. Wang, Z.Q. Liu, Investigation the effect of strain history on crystallographic 10.1016/j.ijmachtools.2010.11.003.
texture evolution based on the perspective of macro deformation for high speed [39] X.L. Liang, Z.Q. Liu, B. Wang, Dynamic recrystallization characterization in Ti-6Al-
machining Ti-6Al-4V, Mater. Charact. 131 (2017) 331–338, https://doi.org/ 4V machined surface layer with process-microstructure-property correlations,
10.1016/j.matchar.2017.07.017. Appl. Surf. Sci. 530 (2020), 147184, https://doi.org/10.1016/j.
[20] Q.Q. Wang, Z.Q. Liu, Plastic deformation induced nano-scale twins in Ti-6Al-4V apsusc.2020.147184.
machined surface with high speed machining, Mat. Sci. Eng. A. 675 (2016) [40] T.J. Ruggles, T.M. Rampton, A. Khosravani, D.T. Fullwood, The effect of length
271–279, https://doi.org/10.1016/j.msea.2016.08.076. scale on the determination of geometrically necessary dislocations via EBSD
[21] Z.P. Pan, S.Y. Liang, H. Garmestani, D.S. Shih, Prediction of machining-induced continuum dislocation microscopy, Ultramicroscopy. 164 (2016) 1–10, https://
phase transformation and grain growth of Ti-6Al-4V alloy, Int. J. Adv. Manuf. doi.org/10.1016/j.ultramic.2016.03.003.
Technol. 87 (2016) 859–866, https://doi.org/10.1007/s00170-016-8497-4. [41] X.L. Liang, Z.Q. Liu, B. Wang, State-of-the-art of surface integrity induced by tool
[22] A. Saha, S.S. Nia, J.A. Rodríguez, Electron diffraction of 3D molecular crystals, wear effects in machining process of titanium and nickel alloys: a review,
Chem. Rev. 122 (2022) 13883–13914, https://doi.org/10.1021/acs. Measurement. 132 (2019) 150–181, https://doi.org/10.1016/j.
chemrev.1c00879. measurement.2018.09.045.
[23] C. Ophus, Four-dimensional scanning transmission electron microscopy (4D- [42] Y.X. Chne, Y.Q. Yang, Z.Q. Feng, B. Huang, X. Luo, Surface gradient nanostructures
STEM): from scanning nanodiffraction to ptychography and beyond, Microsc. in high speed machined 7055 aluminum alloy, J. Alloys Compd. 726 (2017)
Microanal. 25 (2019) 563–582, https://doi.org/10.1017/S1431927619000497. 367–377, https://doi.org/10.1016/j.jallcom.2017.08.018.
[24] X. Xu, J. Zhang, H.G. Liu, Y. He, W.H. Zhao, Grain refinement mechanism under [43] H.G. Liu, J. Zhang, X. Xu, Y.T. Qi, Z.C. Liu, W.H. Zhao, Effects of dislocation density
high strain-rate deformation in machined surface during high speed machining evolution on mechanical behavior of OFHC copper during high-speed machining,
Ti6Al4V, Mat. Sci. Eng. A. 752 (2019) 167–179, https://doi.org/10.1016/j. Materials. 12 (2019) 2348, https://doi.org/10.3390/ma12152348.
msea.2019.03.011. [44] V. Pandey, J.K. Singh, K. Chattopadhyay, N.C.S. Srinivas, V. Singh, Influence of
[25] H.T. Ding, N.G. Shen, Y.C. Shin, Modeling of grain refinement in aluminum and ultrasonic shot peening on corrosion behavior of 7075 aluminum alloy, J. Alloys
copper subjected to cutting, Comput. Mater. Sci. 50 (2011) 3016–3025, https:// Compd. 723 (2017) 826–840, https://doi.org/10.1016/j.jallcom.2017.06.310.
doi.org/10.1016/j.commatsci.2011.05.020. [45] M. Fanfoni, M. Tomellini, The Johnson-Mehl- Avrami-Kohnogorov model: a brief
[26] X. Xu, J. Zhang, J. Outeiro, B.B. Xu, W.H. Zhao, Multiscale simulation of grain review, Nuovo. Cim. D. 20 (1998) 1171–1182, https://doi.org/10.1007/
refinement induced by dynamic recrystallization of Ti6Al4V alloy during high BF03185527.
speed machining, J. Mater. Process. Technol. 286 (2020), 116834, https://doi.org/ [46] Y.M. Mei, Y.C. Liu, C.X. Liu, C. Li, L.M. Yu, Q.Y. Guo, H.J. Li, Effects of cold rolling
10.1016/j.jmatprotec.2020.116834. on the precipitation kinetics and the morphology evolution of intermediate phases
[27] Q.Q. Wang, Z.Q. Liu, B. Wang, Q.H. Song, Y. Wan, Evolutions of grain size and in Inconel 718 alloy, J. Alloys Compd. 649 (2015) 949–960, https://doi.org/
micro-hardness during chip formation and machined surface generation for Ti-6Al- 10.1016/j.jallcom.2015.07.149.
4V in high-speed machining, Int. J. Adv. Manuf. Technol. 82 (2016) 1725–1736, [47] W. Pantleon, Resolving the geometrically necessary dislocation content by
https://doi.org/10.1007/s00170-015-7508-1. conventional electron backscattering diffraction, Scr. Mater. 58 (2008) 994–997,
[28] G. Chen, J. Caudill, C.Z. Ren, I.S. Jawahir, Numerical modeling of Ti-6Al-4V alloy https://doi.org/10.1016/j.scriptamat.2008.01.050.
orthogonal cutting considering microstructure dependent work hardening and [48] K. Huang, R.E. Logé, A review of dynamic recrystallization phenomena in metallic
energy density-based failure behaviors, J. Manuf. Process. 82 (2022) 750–764, materials, Mater. Design 111 (2016) 548–574, https://doi.org/10.1016/j.
https://doi.org/10.1016/j.jmapro.2022.08.032. matdes.2016.09.012.
[29] T. Ozel, I. Llanos, J. Soriano, P.-J. Arrazola, 3D finite element modelling of chip [49] Y.T. Lv, Z.H. Ding, X.Y. Sun, L. Li, G. Sha, R. Liu, L.Q. Wang, Gradient
formation process for machining Inconel 718: comparison of FE software microstructures and mechanical properties of Ti-6Al-4V/Zn composite prepared by
predictions, Mach. Sci. Technol. 15 (2011) 21–46, https://doi.org/10.1080/ friction stir processing, Materials. 12 (2019) 2795, https://doi.org/10.3390/
10910344.2011.557950. ma12172795.
[30] F. Jafarian, M.I. Ciaran, D. Umbrello, P.J. Arrazola, L. Filice, H. Amirabadi, Finite [50] C.M. Sellars, J.A. Whiteman, Recrystallization and grain growth in hot rolling, Met.
element simulation of machining Inconel 718 alloy including microstructure Sci. 13 (1979) 187–194, https://doi.org/10.1179/msc.1979.13.3-4.187.

19