Materials Characterization 132 (2017) 53–68

Contents lists available at ScienceDirect

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

Microstructure and mechanical properties of surface and subsurface layers T

in broached and shot-peened Inconel-718 gas turbine disc fir-trees
Ahmad Chamanfara,⁎, Hossein Monajatia, Alex Rosenbauma, Mohammad Jahazia, Ali Bonakdarb,
Eric Morinb
a
Département de Génie Mécanique, École de Technologie Supérieure, 1100 rue Notre-Dame Ouest, Montréal, Québec H3C 1K3, Canada
b
Siemens Canada Limited, 9545 Côte-de-Liesse, Dorval, Québec H9P 1A5, Canada

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

Keywords: Metallurgical and mechanical characterization of surface and subsurface regions in broached and shot-peened
Broaching fir-trees in an industrial gas turbine disc made of Inconel-718 were carried out. High resolution scanning electron
Shot-peening microscopy (SEM) equipped with energy-dispersive X-ray spectrometry (EDS), electron backscatter diffraction
Inconel-718 (EBSD), X-ray diffraction, optical microscopy, and microhardness instruments were employed for qualitative and
Surface/subsurface microstructure
quantitative assessment of alterations at surface and subsurface levels. Five specific locations along the broached
Misorientation
Residual stresses
and shot-peened path were selected and thoroughly examined. Original metallography methods were developed
to clearly and reliably reveal microstructure constituents. Special emphasis was placed on the generated defects
in view of the manufacturer's quality indices, formation mechanisms of defects, and their potential impact on the
service capability of the disc. Also, advanced analysis of the EBSD data allowed assessment of the deformed layer
thickness as well as the misorientation angle and grain size variations from the broached and shot-peened
surface towards the bulk parent material (PM). Furthermore, through successive material removal by electro-
polishing, measurement of residual stresses as a function of depth from the surface was performed by the sin2Ψ
method. The obtained results are analyzed in terms of impact of the processing conditions on the evolution of
microstructure, microhardness, and residual stresses. The findings are also related to the geometrical location in
the disc.

1. Introduction Discs and blades are generally assembled mechanically using fir-tree
arrangements and broaching is one of the widely-used processes for
Inconel-718 is a wrought Ni-Fe-Cr superalloy which is widely used manufacturing of the complex fir-tree profiles. This is due to the high
in the hot sections of power generation and aerospace gas turbine en- productivity of this process as a result of very high material removal
gines as disc material with service temperatures up to 650 °C because of rates as well as its high potential to achieve the required surface qua-
its room and high temperature mechanical properties and oxidation/ lities and accuracies [7]. To produce the fir-tree arrangement, the
corrosion resistance [1,2]. The nominal chemical composition of this broached tool is drawn in a single or multiple passes through the disc to
alloy in wt% is 52.50 Ni-18.50 Fe-19.00 Cr-5.10 Nb-3.00 Mo-0.50 Al- remove material by axial cutting [8]. The broaching operation is
0.08 C-1.01 Ti [3,4]. The microstructure of Inconel-718 consists of a usually carried out in three main steps: roughing, semi-finishing, and
face-centered cubic (FCC) Ni γ matrix containing solid solution finishing [9]. In each step, one or multiple tools with various teeth with
strengtheners, i.e., Fe, Cr, Mo, Al and Ti [5]. The alloy is primarily increasing pitches are engaged to remove the material at different
strengthened at room and elevated temperatures by uniformly dis- depths [10]. The standard broaching tool for nickel-based superalloys is
persed Ni3Nb body-centered tetragonal (BCT) γ″ precipitates which are made of high-speed-steel (HSS) [11]. From material aspect, it is ex-
coherent with the γ matrix [5]. Also, Ni3(Al, Ti) FCC γ' precipitates, pected that during broaching, the surface and subsurface areas ex-
Ni3Nb orthorhombic plate-like δ phase, and MC carbides (TiC and NbC) perience plastic deformation under severe strain and strain rates which
supply additional strength for the alloy [6]. Moreover, Al and Cr form could result in considerable temperature gradients at specific points.
the protective impermeable Al2O3 and Cr2O3 oxide films in most at- Plastic deformation during machining occurs due to: (i) shearing of the
mospheres and provide the corrosion resistance for the alloy [2]. work piece material and (ii) friction at the tool-work piece interface

⁎
Corresponding author.
E-mail address: [email protected] (A. Chamanfar).

http://dx.doi.org/10.1016/j.matchar.2017.08.002
Received 3 April 2017; Received in revised form 30 July 2017; Accepted 1 August 2017
Available online 03 August 2017
1044-5803/ © 2017 Elsevier Inc. All rights reserved.
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

[12]. Consequently, under such thermomechanical conditions sig-

nificant changes in the microstructure of the surface and subsurface
layers can occur which will significantly affect the service capability of
the disc. In addition, broaching process may induce tensile residual
stresses at surface and near-surface regions. For instance, Afazov et al.
[13] reported that for a work piece with a u-notch generated by
broaching, the tensile residual stresses are up to 1000 MPa at the
broached surface. The presence of such tensile residual stresses reduces
drastically the fatigue, creep, and corrosion resistance of the component
[14]. Therefore, the fir-trees are shot-peened after the broaching op-
eration to eliminate the surface and near-surface tensile residual
stresses and replace them with compressive residual stresses [15].
However, a review of the literature indicated that the impact of the
broaching operation on the surface and subsurface integrity of Inconel-
718 has been rarely assessed in detail in particular for components
with industrial sizes. In general, compared to other machining pro-
cesses, little has been published on broaching in spite of its wide in-
dustrial application. Recently, Chen et al. have published data on de-
fects produced mostly for one point in the fir-tree profile in the
Fig. 1. An as-received section of the Inconel-718 disc with broached and shot-peened fir-trees.
broached Inconel-718 200 × 200 × 50 mm3 coupons [16], white
layer characterization [17], and effect of broaching and heat treatment
on the bending fatigue behavior of Inconel 718 [18]. In particular, 2. Experimental Material and Procedures
they observed deformed layer, white layer with nano-sized grains
(20–50 nm), and tensile residual stresses at surface and subsurface An as-received section of the broached and shot-peened stage 4
layers [16,17]. For other conventional machining methods such as Inconel-718 low pressure gas turbine disc was used for this study
turning, previous studies on Inconel-718 and RR1000® nickel-based (Fig. 1). The part was supplied by Siemens Canada Limited, the man-
superalloy have indicated surface cracking, formation of white layer ufacturer of the gas turbines. The broaching tools were made of HSS
consisted of nanostructured material with the grain size in the and were dressed after broaching of every three discs. Therefore, it can
50–150 nm range as well as development of surface/subsurface tensile be indicated that sharp broach tools were used in this study. During
residual stresses with values close to the yield strength of the alloy broaching the disc was submerged in an oil-based coolant (Microcool
[5,19–21]. The mechanisms for the grain refinement in the white 153). The broaching and shot-peening process parameters are provided
layers in turned Inconel-718 [21] and broached Inconel-718 coupons in Tables 1 and 2, respectively. The as-broached disc was not available
[17] is related to dynamic recrystallization (DRX) occurrence as a to perform an independent study on the sole impact of the broaching
result of: (i) grain subdivision as a consequence of sever plastic de- process on the surface and subsurface alterations.
formation [21] and (ii) adiabatic shear localization where the gov- To assess any microstructural changes that may have occurred by
erning metallurgical development is rotational DRX due to mechani- the broaching and shot-peening processes, metallographic sections
cally-driven subgrain rotations [17]. However, these mechanisms need were prepared (Fig. 2). In particular, the part was sectioned into 12
further clarification. samples and samples 7, 8, 9, 10, and 12 were considered for surface and
The present research inscribes in this context and its main objective subsurface examinations. Samples 7 and 10 were from the middle
is to develop a better understanding of the impact of the broaching and section of the second and third rows, respectively. Sample 9 was from
shot-peening operations on surface and subsurface quality, micro- the beginning of the third row whereas samples 8 and 12 were from the
structure evolution, and residual stress patterns in industrial Inconel- end of the second and third rows, respectively. The investigated surface
718 gas turbine disc fir-trees. This is very important in the assessment of for each sample is also marked by an arrow in Fig. 2.
the disc service capability. In particular this study is different from It is noteworthy that special attention was paid to preserve the sharp
those of Chen et al. [16–18] in: (i) examining of an industrial-size disc edge finish during metallography sample preparation as the surface and
instead of subsize coupons, (ii) using a sharp tool rather than a semi- near-surface areas were of particular interests. To this end, the samples
worn tool, and (iii) through examination of at least five locations in the were hot mounted using a special conductive graphite powder to have
fir-tree profile in place of limited number of points. Therefore, this the maximum edge retention and avoid edge rounding at the broached
study is more representative for industrial circumstances in terms of and shot-peened surface. Also, due to sensitivity of the surface and
work piece geometry, process parameters, and thermomechanical subsurface regions, automated grinding and polishing procedures were
conditions. However, the results of this study were compared with used to prepare the samples for metallographic examinations. After
those of Chen et al. [16–18] in this contribution. Also, occurrence of conventional grinding with SiC papers, the samples were polished on a
any grain refinement at surface/subsurface layers and related operative medium nap cloth for 2 min while using 1 μm diamond suspension
mechanisms need a through and precise investigation which is con- polishing solution.
ducted in this study by using EBSD and advanced analysis of the ob-
tained results. Also, any improvement in the surface and subsurface Table 1
integrity requires a well understanding of the mechanisms of defect The broaching process parameters.
formation during these two processes. This aspect is also addressed in
this work. To achieve the aforementioned objectives, the surface and Cutting speed Coolant
subsurface regions of broached and shot-peened fir-trees were ex-
Surface feet per minute Meters per minute (m/
amined by advanced characterization tools and the obtained results (SFM) min)
were documented and related to the process parameters and geome-
trical locations in the disc. The generated data will be used for opti- Pass 1 9 2.80 Microcool 153
Pass 2 9 2.80 Microcool 153
mization of broaching and shot-peening operations. Also, they serve as
Pass 3 9 2.80 Microcool 153
key inputs for the development of material-based finite element models Pass 4 5.5 1.72 Microcool 153
predicting the fatigue life of the disc.

54
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

Table 2 For crystallographic orientation mapping by EBSD, finish polishing

The shot-peening process parameters. of the sample surfaces with colloidal silica suspension (OP-S) solution
was performed in a vibratory polisher for 12 h. The EBSD investigations
Shot ID Diameter range Nozzle distance Number of Shot-peening
of shots (mm) from disc surface nozzles time (min) were carried out at an accelerating voltage of 15 kV on a Hitachi
(mm) SU8230 field emission gun (FEG)-SEM equipped with a Bruker detector
and acquisition system. In EBSD imaging, the electron beam scanned
110 0.3–0.6 63.5 2 5
the polished surface of the sample which was tilted 70° to the back-
scattered electron detector. EBSD patterns were detected by a CCD
In this study, optical microscopy, SEM, and EBSD were used to camera, while the background was subtracted by the orientation ima-
identify and characterize the quality of the broached and shot-peened ging microscopy software. The scanning covered areas from the
fir-trees. Special attention was paid to the generated surface and broached and shot-peened surface towards the bulk PM. Also, the step
subsurface (up to an 800 μm depth) anomalies in view of the quality size for the EBSD imaging was 0.3 μm.
indices acceptable by the manufacturer. To this end, the polished Work-hardening at areas affected by broaching and shot-peening
samples were electroetched at 3 V for 10 s using Kalling's reagent (5 g was evaluated by measuring the microhardness in these regions and
CuCl2, 100 mL HCL, and 100 mL ethanol). Also, carbides and δ parti- comparing it with that of the bulk PM. The microhardness measure-
cles were revealed by this etchant. Clemex® image analysis software ments were carried out on the as-polished cross sections with a 1 μm
was used to assess the size of defects. Phase identification was carried surface finish. At least three Vickers microhardness measurements at a
out by EDS in SEM. load of 25 g and a dwell time of 15 s were performed for near-surface
regions and the PM for each sample, and then an average value was
calculated and reported for each region. During these experiments, the
distance between each indent was always greater than three times the
indent diameter. Also, to avoid edge deformation effect, the hardness
was measured at locations 15 μm away from the surface. Furthermore,
to improve the accuracy of the microhardness measurements at such a
low load, the indents were imaged by FEG-SEM and then their sizes
were measured by Clemex® image analysis software.
Measurement of the residual stresses as a function of depth was
carried out with an X-ray diffractometer combined with successive
material removal by electropolishing technique. The X-ray diffraction
machine was operated at 20 kV and 4 mA and Mn Kα radiation
(λ = 2.103 Å) was used to attain the (311) diffraction peak at a dif-
fraction (2θ) angle of 152.6° using a beam size of 1 mm. Sin2ψ mea-
surement technique was used at multiple ψ tilt angles of 0, ± 6.4, and
13.7° with a 2° oscillation. A ψ tilt angle amplitude of 13.7° was selected
due to geometrical limitations. In other words, for the higher tilt angle
amplitudes, the diffractometer arms hit the sample fixture, as can be
observed from Fig. 3. However, this considered ψ tilt angle amplitude
satisfied the minimum requirement for the measurement based on the
ψmin = (180–2θ)/2 equation. Furthermore, this angle should be in-
cluded in the ψ angles since at this angle the (311) planes are parallel to
the sample surface and their residual stress is zero. Thus, the spacing of
the (311) planes under the unstressed lattice (d0) can be determined.
Higher angle than this improves the accuracy of the measurements
provided that there is no geometrical constraint which is not the case in
the current experimental set up.

Fig. 2. (a) Sectioning of the as-received disc into 12 samples and considering samples 7,
8, 9, 10, and 12 for surface and subsurface examinations. The investigated surfaces are
shown by arrows, and (b) section prepared from each location for defect and micro- Fig. 3. The geometrical position of the diffractometer arms with respect to the sample
structure analysis. and its fixture.

55
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

The penetration depth of Mn Kα radiation was about 10 μm taking (a)

into account X-ray absorption by the disc at ψ=0°. The measured values
were corrected thereafter to compensate the effect of this absorption.
The Young's modulus and Poisson's ratio used for the (311) reflection
were 200 GPa and 0.27, respectively. The details of the theoretical basis
for measurement of residual stresses can be found in [22]. Stress mea- MC Carbide
surements in both axial and transverse directions were carried out from
the surface to a depth of 365 μm beneath the surface. To determine the
residual stress profiles, material layers were removed by electropolishing
which was performed at ambient temperature and a voltage of 60 V
δ
using a solution composed of 10 vol% percholoric acid, 5 vol% butoxy MC Carbide
ethanol, and 85 vol% methanol. The time for 10 μm material removal by
electropolishing was about 10 s. It is noteworthy that electropolishing δ
does not alter considerably the level of residual stresses [23].

3. Results and Discussion

3.1. Microstructure of the PM

(b)
Fig. 4 represents the SEM micrograph of the microstructure in the
PM location, i.e. an area unaffected by broaching and shot-peening, in
sample 10 extracted from the disc. Fine equiaxed γ grains with an
average size of 6.3 ± 1.1 μm can be observed. The precipitates are also
visible in Fig. 4 which are displayed at higher magnifications in Fig. 5.
The δ, γ″, γ′, and primary MC type carbide particles with platelet, disc-
shaped, spherical, and blocky morphologies, respectively, can be seen.
The δ particles precipitated mostly at grain boundaries (Fig. 4) although
intragranular δ precipitation can be observed as well. Both γ″ and γ′
have nano-scale size (Fig. 5b) and are coherent with the γ matrix. Large
primary MC type carbides are generally formed during solidification
and are stable in the microstructure up to the melting point of the alloy,
i.e., 1336 °C. The PM exhibited a 473 ± 5 HV microhardness in-
dicating that it is in the fully aged condition. More details about the
evolution of volume fraction and size of the δ, γ″, γ′, and primary MC
type carbide particles with thermomechanical processing (forging and
heat treatment) can be found in Reference [6].
Fig. 5. (a) Medium and (b) high magnification FEG-SEM images of the PM location in
sample 10. The δ, γ″, γ′, and MC type carbides are shown by arrows.
3.2. Analysis of Defects

3.2.1. Defects Along the Cross Section Perpendicular to the Broaching

turbine engines, in view of safety concerns, the quality of the surface
Direction
and subsurface layers is of great importance. This study focuses on the
Since Inconel-718 turbine disc is a critical rotating component in gas
experimental examination of surface and subsurface integrity in
broached and shot-peened Inconel 718. The integrity can be defined as
the topographical, metallurgical, and mechanical characteristics of the
surface and subsurface layers.
A systematic analysis of defects and discontinuities for five samples
extracted from the disc (Fig. 2) was carried out. These samples were
from the beginning, middle, and end sections of the second and third
rows. Some illustrative examples of the observed defects are presented
below. Initially, as a point of reference an SEM image of the broached
and shot-peened surface with no surface and subsurface anomalies is
presented in Fig. 6. Also, the investigated location in all cases is marked
by a yellow point on the inset.

3.2.1.1. Plucking. As the most frequent observed defect, plucking was

present in almost all locations in all of the examined broached surfaces.
Fig. 7 shows an illustrative example of a plucking defect (sample 9;
beginning of third row). Similarly, plucking has been observed in the
turned NiCr20TiAl nickel-based alloy [24]. Plucking is defined as small
holes or pockets in the machined surface generated as a consequence of
Fig. 4. An SEM micrograph of the PM location in sample 10 extracted from the broached mechanical removal of grains or large particles (e.g., primary MC-type
and shot-peened Inconel-718 disc presenting the γ grain structure with an average size of carbides and/or δ phase, some of which are marked on Fig. 7) from the
6.3 ± 1.1 μm. The investigated location is shown by a yellow point on the inset. (For surface. In particular, MC-type carbides are hard and undeformable and
interpretation of the references to color in this figure legend, the reader is referred to the hence, can be completely removed at once during broaching, leaving
web version of this article.)
cavities behind. It has been reported that very large depth of cut, tool

56
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

Plucking
No surface and subsurface

Crack
anomalies

Fig. 6. An SEM micrograph of sample 10 extracted from the broached and shot-peened Fig. 8. Occurrence of cracking due to removal of a δ or carbide particle in sample 9
Inconel-718 disc with no surface and subsurface anomalies. revealed by SEM. Plucking is also visible.

Plucking

MC Carbide Cracking /
tearing

Fig. 7. SEM image of plucking in sample 9. Fig. 9. Cracking due to over work-hardening in sample 12 revealed by SEM.

wear/damage (tool clearances), and/or swarf entrapment contribute to an elevated hardness (Section 3.4) and residual stress (Section 3.5)
generation of plucking [25], and therefore these factors must be which will have deleterious impact on the service life, particularly, the
controlled in order to minimize plucking occurrence. Moreover, fatigue life of the disc [27].
cutting speed and feed rate influence the generation of plucking. For Cracking of large (5–15 μm) primary MC-type carbides at surface
instance, it has been reported [24] that by increasing the cutting speed and subsurface layers was also observed, as respectively presented in
from 60 to 100 m/min and decreasing the feed rate from 0.25 to Figs. 10 and 11. Fig. 11 is a backscattered electron (BSE) SEM image.
0.15 mm/r during turning of Inconel-718, a relatively smooth surface Shearing of carbides by the broach tool cutting face is probably the
with little plucking can be obtained due to reducing the cutting forces. main root cause for cracking of the surface carbides in Fig. 10. Fig. 11
illustrates the cracking of the subsurface carbides located typically
3.2.1.2. Cracking. In Fig. 8, in addition to plucking, cracks are also 30 μm beneath the broached surface. In the same way, Chen et al.
visible. They appear to be initiated from the broached surface and then [16,18] have reported cracking of the surface and subsurface carbides
have grown inward. It is probable that due to mechanical removal of a in the broached coupons extracted from forged Inconel-718 pancakes.
second phase (δ particle or MC-type carbide) at the surface, a hole is They pointed out that the cracking of the surface and subsurface car-
created and from there cracking is initiated and propagates along the bides occurs due to large plastic deformation while the carbides are not
grain boundary beneath the surface. Cracking has interrupted the able to deform to the same amount as the plastically deformed surface
continuity of the broached surface. and subsurface layers [16]. Accordingly, these particles are resistant to
Severe plastic deformation of the surface as a result of inducing high plastic deformation and result in crack initiation to release the strain
strain and strain rates during the broaching process can lead to [28]. However, in literature often the type of the cracked carbides is not
cracking. Inconel-718 possesses one of the highest work-hardening determined. In this study, in order to better identify the type of the
exponent (n = 0.63 [26]) among metallic materials and therefore is cracked carbide, EDS was performed. Based on the results illustrated in
more prone to cracking. In particular, since the material work-hardens Fig. 12, two types of primary large MC-type carbides were identified: (i)
in the roughing pass of the broaching, its capacity to further work- Nb/TiC carbides rich in Nb with round edges appearing in white/grey
harden in the semi-finishing and finishing steps as well as during shot- in BSE SEM images and (ii) Ti/Nb carbonitrides rich in Ti with sharp
peening will be limited as the material has a saturation limit for work- edges appearing in dark in the BSE SEM images. Accordingly, the
hardening. Once over work-hardened, cracking will occur. An example cracked carbides at surface and subsurface locations were mostly
of surface cracking due to over work-hardening is presented in Fig. 9. complex Nb/TiC carbides rich in Nb. These findings are in agreement
Severely plastically deformed material at the broached surface exhibits with those observed by M'Saoubi et al. [29] regarding cracking of NbC/

57
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

noteworthy that according to Fig. 13, the Nb/Ti C carbides and Ti/Nb
carbonitrides exhibited a clustering tendency as well. Such behavior
was also observed in the bulk PM. Carbide clustering adversely affects
MC the LCF life of the disc [30].

3.2.1.3. White Layer and Deformed Layer. The presence of continuous

MC (Fig. 14a and b) and discontinuous (Fig. 14c) narrow bands with a
white structureless appearance were detected on the surface. The white
layer with either a continuous or discontinuous morphology is
metallurgically different from the bulk PM (Fig. 15). Lack of crystal/
grain structure at the optical microscopy scale is the key metallurgical
characteristic of this layer. This is related to its high resistance to
etching and inadequate resolution of the optical microscope [17]. The
presence of highly deformed grains adjacent to the white layer (Fig. 15)
beneath the broached surface is an indication of severe plastic
deformation. With an increase in depth the severity of the plastic
deformation decreases.
Fig. 10. An SEM image showing the cracking of surface MC-type carbide in sample 10. The white layer characterized by an adjacent distorted grain struc-
ture is generated by a shear transformation as a consequence of ex-
cessive distortion of the surface [25]. The thickness of the deformed
layer is in the 2.3 to 6.4 μm range. Also, the white layer in Fig. 15 is 3.5
to 7.6 μm wide. In the same way, Brasche et al. [32] have observed a
continuous white layer across the length of the broached slot in the
development test samples which were 86 mm long and extracted from
Inconel-718 forged pancakes. In order to better characterize the mi-
crostructure and the mechanical properties of this thin white layer,
EBSD and microhardness were used in this study as discussed in
Sections 3.3 and 3.4.
The mechanisms for white layer formation can be summarized as
[17,33]: (i) rapid heating and cooling, (ii) possible surface reaction
with environment, and (iii) plastic deformation resulting in micro-
structure refinement. In particular, machining under abusive conditions
can generate a surface white layer. Formation of white layer associated
with a deformed layer can be due to lack of sufficient coolant supply,
overheating, high feed rate/speed [25], and/or blunt/worn tools [20].
In particular, Inconel-718 maintains its high strength at elevated tem-
peratures. Therefore, intense work-hardening of the alloy, high cutting
Fig. 11. Cracking of subsurface carbides in sample 8 revealed by BSE imaging in SEM.
forces, and high local broaching temperatures will be resulted. Fur-
thermore, broaching induced plastic deformation leads to adiabatic
heating [34]. Accordingly, because of the low thermal conductivity of
TiC primary carbides during turning of Inconel-718. the alloy, at the broach tool cutting face, temperature increases locally
It should be noted that due to the low intensity of the X-rays gen- and along with the temperature gradients, severe (high strain rate)
erated by light elements such as C and N, the EDS system has limitation plastic deformation of the surface and subsurface layers, rapid cooling,
for quantitative analysis of these light elements and the data presented and high cutting forces/pressure, a white layer is formed on the
here should be considered as qualitative and mostly indicative of the broached surface.
presence of these elements rather than their percentage. It is pointed out that the presence of a white layer on the machined
Surface and subsurface cracking of carbides lead to a considerable surface can decrease the fatigue life up to six times due to easy crack
drop in the low cycle fatigue (LCF) life. For instance, Bhowal et al. [30] initiation and propagation in this region as a consequence of its brittle
compared the LCF lives of machined (lathe turned) and polished In- nature and presence of tensile residual stresses in this region
conel-718. In the former, surface and subsurface cracking of carbides [17,35,36]. Therefore, it is essential to determine optimum broaching
occurred whereas in the latter it did not. It was found that the LCF life process parameters to avoid white layer formation. For instance, Klock
of the polished Inconel-718 was up to 15 times that of the lathe turned et al. [37] indicated that by decreasing the cutting speed in broaching
one. Also, it has been reported [31] that surface carbides are the most of Inconel-718 with HSS tool from 10 to 2.5 m/min, the morphology of
important elements in limiting the LCF life, since they crack frequently the white layer changes from a continuous to a discontinuous one and
during machining and become failure initiation sites under LCF loading. its thickness decreases by 50%. They also reported that by using a ce-
Therefore, the presence of any type of crack is not acceptable from a mented carbide broach tool instead of HSS tool the thickness of the
quality assessment point of view. white layer can be reduced. This was related to decreasing the friction
It is probable that in addition to the broaching operation, the shot- at the tool-workpiece interface and associated alteration in the tem-
peening process has also contributed to the fragmentation of the surface perature profile over the broached surface.
and subsurface carbides especially the large (primary) ones. However, it Similarly, in Fig. 16a, an SEM image of a distorted layer is presented
is interesting to mention that Bhowal et al. [30] reported that the LCF for sample 10. In this layer, which is relatively narrow (7 μm wide) and
life of the lathe turned Inconel-718 with damaged/cracked surface and uniform, the plate-like δ particles were severely deformed and frag-
subsurface carbides was recovered by shot-peening. The improvement mented with an elongated and aligned appearance as a result of the
was related to shot-peening induced compressive residual stresses at severe shearing deformation history. This phenomenon was generally
surface and subsurface layers. Therefore, shot-peening can compensate observed before and/or after the peaks in the fir-tree arrangement and
also for the damaged carbides generated from machining. It is can be related to the strain induced by the broaching and/or shot-

58
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

Fig. 12. (a) and (b) Presence of primary car-

(a) (b) bides in sample 8 revealed by BSE imaging in
SEM, (c) to (e) qualitative EDS analysis for lo-
cations marked on the image shown in b.

(c) (d)

(e)

peening in this particular area. In the same way, a non-uniform dis- 3.2.1.4. Rough Surface. In Fig. 17 the formation of an uneven or rough
torted layer, up to 15 μm thick, containing special alignment of δ phase surface at the broached and shot-peened fir-trees is presented. The alloy
is presented in Fig. 16b. It can be observed that the material is deformed maintains a considerable part of its strength and toughness at high
in a region ahead of the broach tool. In Chen et al. [16] study on temperatures and consequently, its rapid work-hardening tendency
broaching of Inconel-718 coupons, the deformed layer thickness is re- during broaching leads to high cutting forces and formation of a
ported as 45 μm which is significantly higher than that of the current rough surface in the work-hardened layer. Also, local material
study, i.e., 7 to 17 μm. Such a difference can be related to using a semi- removal as a result of vibration or uncontrolled movement of tool,
worn tool in the former and a sharp tool in the current work. A semi- fixture, or disc due to insecure component/fixtures or fixture wear/
worn tool creates a larger contact at the tool-workpiece interface be- damage can lead to formation of a rough surface.
cause of reduced clearance angle on the tool. Therefore, considerable Due to the stress concentration, presence of a rough surface can lead
rubbing friction develops which leads to an increase in the cutting to early fatigue crack initiation and propagation and hence short fatigue
force, plastic deformation, and cutting zone temperature. Consequently, life. For instance, Andrews et al. [38] have shown that for Inconel-718
more plastic deformation and a deeper worked-hardened layer are fatigue crack was initiated from the notch generated intentionally on
generated in the semi-worn tool condition compared to using a sharp the surface by lathe cutting. Also, they indicated that the machined
tool. Consistently, Zhou et al. [12] have also observed deformed layers notch provided a preferential route for growth of the surface fatigue
in turned Inconel-718 and have indicated that when a new tool was crack along the notch root and crack growth was faster in the notched
used, the deformation was concentrated at very thin (10 μm) surface samples compared to the polished ones. However, quantifying the effect
layers while in case of worn tool, the deformed layer extended from the of broaching process parameters on the extent of rough surface for-
surface to subsurface layers and reached to a 50 μm depth. mation is complex and at the present time operator experience as well

59
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

Fig. 13. Clustering of the complex Nb/TiC

(a) (b) carbides and Ti/Nb carbonitrides in sample
10 revealed by SEM.
Nb/TiC Clustering

Nb/TiC Clustering

(c) (d)

(e)

Ti/ Nb(C, N)

as trial and error are used for minimization of rough surface formation. Therefore, it is probable that SiC particles from broach tool and/or shot-
Accordingly, in order to improve the fatigue life, have an accurate di- peening balls became separated from their PM and deposited on the
mension, and easier blade attachment, removal of any remaining rough broached and shot-peened surface. During metallographic sample
surface by manual or automated operations is needed. preparation, these SiC particles and those from grinding paper could
be scattered over the sample surface (Fig. 18a). The existence of the
foreign/nonparent material on the broached/shot-peened surface is not
3.2.1.5. Foreign/nonparent Material. The surface contamination by
acceptable based on the manufacturer requirements. However, to the
foreign/nonparent material was also detected. As illustrated in
authors' knowledge, the level of damage caused by the foreign material
Fig. 18a and b, presence of black spot on the ground and polished
existence to the disc performance has not been quantified and the
cross section of the broached and shot-peened disc (sample 8) can be
preventive measures are still based on trial and error.
observed. The qualitative assessment of these black spots by EDS
(Fig. 18c) revealed them to be SiC particles. These particles could
have three origins: (i) from the broach tool, (ii) from the shot-peening 3.2.2. Defects Along the Broaching Direction
balls, and (iii) could be SiC abrasive particles from the grinding papers. A number of surface defects along the broaching direction were also
The chemical compositions of T1 HSS tool from ASTM A 600-04a and detected by FEG-SEM studies (Fig. 19). Coarse broaching grooves par-
SA2.5 steel ball from Reference [39], used respectively for broaching allel to the cutting direction are visible (Fig. 19a and b). They could be
and shot-peening, are formulated in Table 3. It can be observed that generated by three mechanisms [28]: (i) feed mark ridges (natural
both Si and C are present in the broach tool and shot-peening balls. surface defects), (ii) debris (tool material chips or hard carbide

60
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

(a)
Continuous white layer
Deformed
layer

10 µm

Fig. 15. Presence of a continuous white layer associated with a deformed layer in sample
10 revealed by optical microscopy.
(b)
Continuous white layer

(a)

7 µm

(c)

(b)

15 µm
Fig. 14. SEM images showing the formation of continuous and discontinuous white layers
in (a) sample 9, (b) sample 12, and (c) sample 10.

particles) dragging over the surface, and (iii) ploughing by chips ad- Fig. 16. Presence of the deformed layers in sample 10 for different locations along the fir-
hered to the tool. Broaching grooves were not completely eliminated by tree profile revealed by SEM.
the shot-peening process. During service broaching grooves act as stress
concentration sites. They are preferred sites for fatigue crack initiation
and thus are detrimental for fatigue life. Chen et al. [18] reported that workpiece material to the cutting edge. This also contributes to the
fatigue cracks initiated from the broaching grooves in the broached plastic deformation of the material instead of being cut. Material drag
Inconel-718 coupons. can lead to tearing and cracking of the surface.
Material drag (Fig. 19a) can also be generated as a result of in- Moreover, material side flow created during broaching was detected
adequate cutting ability of the blunt cutting faces as the material is in the vicinity of the broaching grooves (Fig. 19b). Material side flow
plastically deformed rather than being cut as chips [20]. Presence of occurs once the chip thickness is smaller than a critical value. In this
hard abrasive particles such as carbides can lead to intensive wear of way, a considerable portion of the plastically deformed workpiece
the broach tool and thus a blunt tool will be resulted. Also, low thermal material is furrowed to one side aided by the high temperature in the
conductivity of the Inconel-718 results in the increased local tempera- cutting area [16,40]. In addition, surface cavities (plucking) formed due
tures at the broach tool cutting face and thus leads to adhesion of the to mechanical removal of large and hard particles from the surface were

61
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

(a) (a)
Uneven or rough surface

(b)
(b)

Crack

Uneven or rough
surface BSE
Mapping_617 10 m
(c) MAG:2500x HV: 15kV WD: 7.7mm

(c)

Fig. 17. (a) and (b) Rough surface formation in sample 9, and (c) in sample 12 revealed
by SEM.

Fig. 18. Foreign/nonparent material in sample 8 observed by BSE imaging in SEM with
detected. Especially, MC-type carbides are hard and undeformable and related qualitative EDS analysis.

therefore, can be entirely removed at once during broaching, leaving

cavities behind. Also, bombardment by shots during the shot-peening
process resulted in the local plastic deformation of the surface and types of defects or a defect-free surface can be generated. In order to set
formation of a relatively rough surface (Fig. 19b and c). up a reasonable correlation in this regard, characterizing the thermo-
It should be noted that based on the thermomechanical condition mechanical condition (evolution of temperature, strain, and strain rate)
induced by the broaching process for every location along the and its variation with changing the broaching process parameters (e.g.,
broaching path, which could be different from point to point, different cutting speed, cutting feed, lubrication, etc.) is needed by conducting

62
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

Table 3
Chemical composition in wt% for T1 HSS broach tool based on the ASTM A 600-04a and SA2.5 steel ball according to Reference [39].

C (min-max) Mn P (max) S (max) Si Cr V W Fe

T1 HSS 0.65–0.80 0.10–0.40 0.03 0.03 0.20–0.40 3.75–4.50 0.90–1.30 17.25–18.75 Balance
SA2.5 steel ball 0.70–1.20 0.35–1.20 0.05 0.05 0.40–1.20 – – – Balance

work-hardening at surface and subsurface layers, the kernel average

(a)
misorientation (KAM) maps were extracted from the EBSD results. In
the KAM map, the average misorientation of a point within a grain
Broaching relative to its neighbors is determined (a value in the 0–5° range) and
Grooves presented by a particular color in the KAM map. In other words, KAM
map represents the variations in local misorientation within the grains.
A higher average misorientation for a point indicates a higher plastic
Dragging deformation and work-hardening at that point.
The KAM map from the surface to the bulk PM is presented in
Rough
Fig. 20 for sample 10. Blue and red colors illustrate the minimum and
Surface maximum amounts of the point-to-point misorientation, respectively. A
noticeable difference in the point-to-point misorientation at the surface
and subsurface layers compared to the unaffected area can be realized.
At the surface and subsurface layers in the KAM maps (shaded green,
yellow, and orange) a high density of low-angle misorientation was
observed. In other words, there is a higher point-to-point misorientation
at the surface and subsurface layers compared to the bulk PM. This can
(b) be related to subjecting the surface and subsurface layers to plastic
Side Flow
Broaching deformation during broaching and shot-peening. In contrast, in the PM
location, a considerable fraction of the microstructure (blue areas)
Groove
displayed a very low KAM (< 1°). In these areas full recovery and re-
Cavity
crystallization have taken place during hot forging, solution heat
treatment, and double aging of the as-received disc which has resulted
Rough in very low point-to-point misorientation. Consequently, due to pre-
Surface sence of the high KAM regions at the surface and subsurface layers
compared to the PM region, the amount of work-hardening in the
former is higher as supported by the microhardness data (Section 3.4).
Since gain size significantly affects the creep and fatigue resistance,
it is important to characterize grain size evolution in the affected zones
and compare it with the PM. An EBSD map presenting the grain
structure from the surface to a 40 μm depth for sample 10 is shown in
Fig. 21. Although, a well-defined grain structure with clear boundaries
is hard to be recognized at the surface, refined discrete areas with a
(c) high level of strain within them as manifested by evolving colors can be
easily observed. These areas are confined by fine-dashed white lines in

Cavity
Rough
Surface

Fig. 19. FEG-SEM images of the surface defects in the broaching direction for sample 12.

some experimental and modeling works and this can be the subject of a
future investigation. In this way, the impact of the process parameters Misorientaion angle
on the microstructural and mechanical properties of the material can be
evaluated even better. 0 5

3.3. EBSD Assessment Fig. 20. The KAM map (low angle misorientation distribution) from the surface to the
bulk PM in sample 10. The normal direction (ND) and transverse direction (TD) are also
shown.
To evaluate the intensity of the plastic deformation and degree of

63
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

Fig. 21. EBSD map from the surface to the bulk PM in

sample 10.

30 μ m

Fig. 21. While the thickness of the layer containing these ultrafine subgrain boundaries and the accompanying gradients demonstrate in-
grains is up to ~5 μm, the layer affected by plastic deformation at creasing misorientation within grains, subgrain rotation, and formation
surface and subsurface layers is up to 11 μm thick. A representative of HAGBs in this layer (Figs. 20–22). Therefore, the grain refinement
deformed layer is illustrated in Fig. 21 by long-dashed white lines. In mechanism can be explained as increasing misorientation within grains
this layer, it can be seen that grains show evolving colors. After this due to plastic deformation, subgrains development and rotation, and
depth the grains' color is solid showing no internal straining. The formation of HAGBs. The plastic deformation rotates the cells/sub-
average grain size in the first layer lies in the 0.6–0.8 μm range which is grains in the favor of those slip systems which have a higher chance for
significantly smaller than that of the core layer, i.e., ~6 μm. Two pos- deformation continuation. Therefore, it can be concluded that the
sible mechanisms for formation of the fine grains at the surface and second proposed mechanism for DRX occurrence during broaching of
subsurface layers can be considered: Inconel-718 is operative as supported by EBSD studies in this work.
(i) Classical DRX for nickel-based superalloys: this means nucleation From Fig. 22, an 11 μm thickness for the layer with an increased
of strain-free grains and growth of these grains. The onset temperature LAGB density indicates that the worked-hardened layer was extended to
for DRX in Inconel-718 is about 950 °C [1]. Klocke et al. [41] recorded a depth of 11 μm. It is noteworthy the surface/subsurface layers char-
surface temperatures in the 700 °C range during broaching of Inconel- acterized in Figs. 20–22 could be indeed a white layer. This is further
718 under relatively low speeds (2.5 to 10 m/min) with HSS tool and supported by the thickness of the white layers, i.e., 3.5 to 7.6 μm, which
using a cooling lubricant. Similarly, based on simulation and tem- is in the same range as the layer with the refined microstructure
perature recording using a 2-color-pyrometer, Vogtel et al. [11] in- (~ 5 μm). In other words, because of the ultrafine grains presence in this
dicated that in broaching of Inconel-718 the temperature remains layer, it appears as a structureless white narrow bands under optical
below 600 °C at the tool cutting face. Therefore, the temperature during microscope (Fig. 15).
broaching remains below the required temperature for DRX occurrence
(950 °C) and most likely this DRX mechanism is not operational during 3.4. Microhardness Evolution
broaching.
(ii) Increasing misorientation within grains, subgrains development and In this study, microhardness measurements were used as well to
rotation, and formation of high angle grain boundaries (HAGBs): In Fig. 22, assess metallurgical changes and work-hardening in the subsurface re-
the low angle grain boundaries (LAGBs), HAGBs, and twin boundaries gion relative to the bulk PM. The average microhardness value for a
are respectively shown by red, yellow, and blue boundaries. The LAGBs subsurface region 15 μm away from the broached and shot-peened
and HAGBs respectively have a misorientation in the 2–10° and > 10° surface is compared to that of the PM in Fig. 23 for sample 10. In this
range. A high density of subgrain boundaries can be observed in an figure the inset represents the measurement locations. The PM location
11 μm thick surface/subsurface layer as a consequence of large plastic is at least 8 mm away from the surface. It can be found that the sub-
deformation in this region induced by broaching and shot-peening. The surface region is harder than the PM. This indicates that the material

Fig. 22. Distributions of LAGBs, HAGBs, and twin boundaries from

HAGB ( >10 ) LAGB (2 < <10 ) Twin boundaries the surface to the bulk PM in sample 10.

64
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

shot-peening induces work-hardening as well. However, the work-

hardening induced by each of these processes affects the service cap-
ability of the component in a different way. For instance, Ghosh and
Kain [44] indicated that generation of a machining-induced work-
hardened region significantly increases the susceptibility to stress-cor-
rosion cracking in 304 L stainless steel. Moreover, reduced ductility in
the broached work-hardened region (white layer) may accelerate fa-
tigue crack initiation and propagation. Accordingly, as a unique cold
working process, shot-peening is performed after broaching to modify
the surface and tailor the surface integrity. In particular, shot-peening
increases the surface hardness significantly and induces compressive
residual stresses as shown in Section 3.5. By this means, the fatigue and
wear resistance event at high temperatures will be enhanced.

3.5. Residual Stresses

The profile of the residual stresses as a function of the depth below

Fig. 23. Microhardness variation from a point 15 μm away from the broached and shot-
the surface for the central section of the third row measured by X-ray
peened surface to the bulk PM in sample 10. The standard deviations for the measure-
ments are shown as error bars.
diffraction method is presented in Fig. 24. Location and directions of
the residual stress measurements are shown respectively by a yellow
point and arrows on the top image in Fig. 24. It can be seen that, the
had been work-hardened by the broaching and shot-peening processes residual stress profiles in both axial and transverse directions are
as a result of the induced surface and subsurface severe plastic de- compressive. Also, the residual stress profiles are more severe in the
formation (Figs. 9 and 15). Also, a slight scatter in the microhardness of axial direction than in the transverse direction.
the PM due to local variations in the microstructure is visible. The For the axial direction, residual stresses increased in magnitude
average PM microhardness (473 ± 5 HV) is relatively close to the from −638 MPa at surface to the maximum compressive value of
value (430 HV [6]) reported for the fully heat treated (solutionized and − 819 MPa at 35 μm beneath the surface. The latter is ~80% of the PM
double aged) Inconel-718. yield strength, i.e., 1030 MPa [45]. Beyond the 35 μm depth, the
Since the work-hardened region has a narrow thickness (< 30 μm, magnitude of the compressive residual stresses decreased and even-
Fig. 22), obtaining a representative microhardness profile at loads tually approached zero at a depth of 328 μm. The increase in the
higher than 100 g is not practically possible. For instance, for a load of magnitude of the residual stresses in the 0–35 μm interval may be
200 g the indent diameter is ~28 μm. Considering the need for a dis-
tance between each indent to be at least three times the indent dia-
meter, i.e., 84 μm, it is not possible to have a characteristic micro-
hardness profile capable to capture real variation in microhardness. On
Transverse

the other hand, for microhardness measurements at loads ≤ 100 g,

direction

slight measurement inaccuracy in indent diameter (e.g., ± 0.5 μm) will

result in a considerable hardness difference due to small indent size
[42]. Therefore, the data presented in Fig. 23 should be considered as
an indication of a relatively higher hardness in the subsurface region
compared to the bulk PM rather than an indication of the absolute
hardness values. Ideally, nano-indention should be used to map the
variation in the hardness from the area adjacent to the surface towards
the PM and this can be done for a future work. As an example for such a
hardness profile, in the broached Inconel-718 coupons in the fully aged
condition, nanohardness investigations indicated a gradual increase in
the hardness from the PM value (5.7 GPa) at a depth of 50 μm beneath
the surface to 6.3 GPa adjacent to the surface, i.e., 5 μm away from the
surface [16,43].
It is noteworthy that since it is not possible to correctly measure the
hardness at distances few nanometers away from the surface due to
potential edge deformation and lack of edge support, it is expected that
the real degree of work-hardening will be even higher. Therefore, to
better characterize the work-hardened layer and degree of work-hard-
ening, metallography (Section 3.2), EBSD (Section 3.3), and residual
stresses measurement were carried out. The results of the latter are
discussed in the next section.
Although it is difficult to determine how much work-hardening is
induced by broaching compared to shot-peening, based on the results
presented in Section 3.2 (Figs. 9, 15, and 16), broaching definitely in-
duces plastic deformation and thus work-hardening. For instance, as
discussed in Section 3.2, the presence of a white layer is a definite in-
dication of work-hardening during broaching as shot-peening cannot
induce white layer formation. Furthermore, as discussed above, other
researchers [16,43] have reported work-hardening induced by Fig. 24. Residual stress profile for the central section of the third row in the axial and
broaching in the subsurface layers. Also, it is generally accepted that transverse directions.

65
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

related to shot-peening. For the transverse direction, the magnitude of

the residual stresses started to increase from − 243 MPa at the surface
and reached the peak value of − 450 MPa at a depth of 7 μm, which is
in the shot-peened plastically strained layer. From there, a gradual
decay in compressive residual stress was recorded such that the mate-
rial was effectively stress free at a depth of 328 μm.
Therefore, it can be concluded that the depth of the shot-peening
affected zone is up to 35 μm. From this depth to 328 μm, the residual
stresses represent a balance between remnants of residual stresses from
broaching and shot-peening operations. The anisotropy in the residual
stress profile (Fig. 24) is related to the dominant role of the mechanical
effect [28]. This means that the cutting force component during
broaching and the amount of the plastic deformation during shot-pe-
ening are not the same along different directions.
The residual stress measurements at the surface of the valley be-
tween two peaks (Fig. 25) at the center of the third row were also
measured as −756 ± 36 MPa and − 231 ± 47 MPa for the axial and Fig. 26. Comparison of the current study residual stress profile with the as-broached [16]
transverse directions, respectively. Here also, it can be observed that and as shot-peened [46] Inconel-718.
the impact of shot-peening on residual stresses is more pronounced in
the axial direction than in the transverse direction. The magnitude of
the surface residual stresses at the valley in the axial direction stresses increase with cutting speed and broach tool wear. However, in
(− 756 MPa) is larger than the corresponding value (− 638 MPa, his study the quantitative data have not been reported and these trends
Fig. 24) at the peak. In contrast, the magnitude of the surface residual have been shown schematically.
stress at the valley in the transverse direction (− 231 MPa) is smaller Residual stresses in machining are generated by plastic deformation
than the corresponding value (− 243 MPa) at the peak. and local temperature increase [48]. Mechanical plastic deformation
In order to better investigate the individual contribution of mainly creates compressive residual stresses, whereas substantial cut-
broaching and shot-peening on the development of residual stresses in ting heat input typically generates tensile residual stresses. In parti-
Inconel-718, the data available in the literature are compared with the cular, due to low thermal conductivity of Inconel-718, the heat gener-
current results in Fig. 26. In this figure, the residual stresses after ated during broaching is mostly concentrated at the surface. Thus, the
broaching of Inconel-718 under comparable conditions to the current surface and immediate subsurface regions expand while the bulk PM
work and at the same axial direction were extracted from Reference does not. Upon cooling, the former attempt to recover their positions in
[16]. Similarly, the residual stresses after shot-peening of Inconel-718 spite of the bulk PM resistance. Hence, tensile residual stress develops
at comparable shot-peening conditions and at the same axial direction at the surface and subsurface regions due to plastic misfit strain be-
were extracted from Reference [46]. It can be observed that after tween the surface/subsurface and the bulk PM [19]. Therefore, the
broaching, tensile residual stresses develop at the surface and subsur- tensile residual stresses at the surface and subsurface layers after
face regions. In particular, the tensile residual stress increases from broaching (Fig. 26) have thermal origin, while the subsurface com-
~ 500 MPa at the surface to a maximum value of 732 MPa at 9.4 μm pressive residual stresses are primarily generated by the mechanical
beneath the surface. Then, it gradually decreases to compressive values plastic deformation generated by the cutting forces [16].
with a maximum value of − 250 MPa at a depth of 83 μm and finally According to Fig. 26, after shot-peening of Inconel-718, residual
approaches zero at the end. In the same way, it is reported [13] that in a stresses increased in magnitude from − 733 MPa at surface to the
work piece with a u-notch produced by broaching, the tensile residual maximum compressive value of − 1083 MPa at 70 μm beneath the
stress decreased from 1000 MPa at surface to 400 MPa at 125 μm be- surface. Beyond the 70 μm depth, the magnitude of the compressive
neath the surface and gradually approached zero at 430 μm depth. Si- residual stresses decreased and eventually approached zero at a depth
milarly, Taricco [47] has reported tensile residual stresses in the of 300 μm.
broached Inconel-718 disc and has indicated that tensile residual In the current work, compressive residual stresses are present in the
fir-tree region (Fig. 24). This would suggest that the mechanical
straining induced by shot-peening overcomes the tensile residual
stresses induced by broaching and thus a compressive residual stress
state in the surface and subsurface layers is generated. Consistently,
Transverse direction Taricco [47] has pointed out that after shot-peening of the broached
-231 47 MPa Inconel-718 disc, tensile residual stresses shifted to compressive ones
being in the 500 to 800 MPa range at the surface.
Based on the above analysis it can be pointed out that, shot-peening
generates compressive residual stresses at and near the surface by in-
ducing cold plastic deformation. These residual stresses and work-
hardening associated with severe compressive plastic deformation of
the surface and subsurface layers delay fatigue crack initiation and
propagation [49]. It is noteworthy that at typical service temperatures
of Inconel-718 disc, where the component experiences thermo-
mechanical loads, a fraction of the compressive residual stresses in-
duced by shot-peening is relieved at the early stages of the operation;
however, adequate stress remains and thus, the beneficial impact of
shot-peening on service capability is maintained [15]. For instance, it is
observed [47] that by heat treatment of shot-peened Inconel-718 disc at
Fig. 25. Residual stress values in the axial and transverse directions for the central point 482 °C for 16 h, peak residual stress values of about − 1000 MPa were
located in the valley between two peaks for the third row of Fig. 2.
reduced to −400 to − 500 MPa range. Relief of compressive residual

66
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

stresses occurs by the annihilation or reordering of thermally mobile surface and adjacent subsurface layers. Indeed, the useful impact of the
defects (vacancies and dislocations) that exist in the work-hardened shot-peening process in generating compressive residual stresses ex-
regions [50]. tends to 328 μm depth below the surface (Fig. 24) with the maximum
values at the 35 μm locations beneath the surface.
3.6. Correlation Between Process Parameters, Microstructure Evolution,
Microhardness, and Residual Stresses 4. Conclusions

The following broaching process parameters affect the surface/ Characterization of surface and subsurface integrity, microstructure,
subsurface defect formation, microstructure, microhardness and re- microhardness, and residual stresses in a broached and shot-peened
sidual stress generation: Inconel-718 disc was carried out. From this study, the following con-
(i) Broaching speed: with increasing the broaching speed the tem- clusions can be drawn:
perature in the cutting zone increases which leads to plastic deforma- 1. In spite of broaching under low cutting speeds (≤ 2.8 m/min) and
tion due to material softening. Under these conditions surface/subsur- using a cooling lubricant, a wide range of fatigue life deteriorating
face defects and tensile residual stresses develop. surface integrity issues such as plucking (surface cavities), cracking,
(ii) Use of cooling lubricant during broaching: it decreases the plastically deformed/work-hardened layer (11 μm thick), and white
friction and thus the overall temperature in the cutting zone. In this layer (3.5 to 7.6 μm thick) were formed during broaching of Inconel-
way, the thermal dominant role and thus the tensile residual stresses 718.
can be reduced. 2. Nb/TiC carbides and δ particles generated surface and subsurface
(iii) Broach tool condition: compared to a new broach tool, a worn defects such as grain boundary cracking and plucking as they cannot
or semi-worn tool creates a larger contact at the tool-workpiece inter- deform to the same degree as the γ matrix.
face. Therefore, intense rubbing friction develops which results in in- 3. EBSD misorientation analyses demonstrated a considerably mis-
crease in the cutting force, plastic deformation, and cutting zone tem- oriented subgrain structure in the white layer. Also, EBSD revealed the
perature. Hence, hardening of the surface layer and tensile residual presence of ultrafine grains in the white layer. The mechanism for grain
stresses at the surface develop [28]. refinement in the white layer was identified as DRX occurrence due to
(iv) Cutting action in broaching: in broaching in any time at least increased misorientation within grains as a result of plastic deformation
two cutting faces are in contact with the work piece. So, each cutting induced by broaching/shot-peening, subgrains development and rota-
face can be considered as an independent heat source caused by plastic tion, and formation of HAGBs.
work and friction [16,34,51]. As a result, a high degree of heat buildup 4. Because of the increased density of the low-angle misorientation,
and plastic deformation because of softening occur on the broached grain refinement, and fragmentation of the δ particles at the white layer
surface [16]. (or subsurface layer), the microhardness increased in this region com-
In this study, the broaching speed in the main passes was 2.8 m/ pare to the unaffected PM.
min. In broaching of nickel-based superalloys with HSS tool, the 5. The residual stresses were compressive and extended into the disc
broaching speed is limited to the 2–8 m/min range [16], which is down to a depth of 328 μm below the surface. The maximum com-
considered as a low speed compared to other machining processes. For pressive residual stresses of − 819 MPa and − 450 MPa appeared in the
instance, in turning of Inconel-718, a cutting speed of 100 m/min or near-surface regions in axial and transverse directions, respectively.
more is used. The broaching speed in this study lies in the optimum
range (2 to 4 m/min) in terms of surface quality and tool wear for Acknowledgements
broaching of nickel-based superalloys with HSS tools [52]. Also, as
indicated in Section 2, sharp HSS tools and cooling lubricant were used Support from Natural Science and Engineering Research Council of
during broaching of Inconel-718 in the current study. Canada (NSERC-ENGAGE) and Siemens Canada Limited are gratefully
As discussed in Section 3.3, the surface temperatures during acknowledged. Also, the authors are grateful to Pierre Ladmiral for
broaching of Inconel-718 under relatively low speeds (2.5 to 10 m/min) performing a part of the SEM work.
with HSS tools and using a cooling lubricant remains in the 600 to
700 °C range [11,41]. Therefore, it can be pointed out that even References
broaching at such a low cutting speed (2.8 m/min) under cooling lu-
bricant in this study can increase the temperature at the cutting zone to [1] M. Dehmas, J. Lacaze, N. Niang, B. Viguier, TEM study of high-temperature pre-
~ 700 °C and lead to ploughing and plastic deformation due to material cipitation of delta phase in Inconel 718 alloy, Adv. Mater. Sci. Eng. 2011
(2011) 1–9.
softening. Consequently, broaching of Inconel-718 is associated with [2] C.T. Sims, N.S. Stoloff, W.C. Hagel, Superalloys II, John Wiely & Sons, New York,
surface integrity issues such as generation of surface/subsurface de- 1987.
fects, e.g., white layer (Fig. 15), deformed layer (Figs. 15 and 16) and [3] A. Niang, B. Viguier, J. Lacaze, Some features of anisothermal solid-state trans-
formations in alloy 718, Mater. Charact. 61 (2010) 525–534.
tensile residual stresses (Fig. 26) caused by high temperatures and [4] M.E. Mehl, J.C. Lippold, Effect of δ-phase precipitation on the repair weldability of
plastic deformation at the cutting face. alloy 718, in: E.A. Loria (Ed.), Superalloys 718, 625, 706 and Various Derivatives,
Also, plastic deformation induced by broaching and shot-peening The Minerals, Metals & Materials Society, Pittsburgh, Pennsylvania, 1997, pp.
731–741.
increased the density of the low-angle misorientation (Fig. 20), led to [5] V. Bushlya, J. Zhou, J.-E. Ståhl, Modeling and experimentation on multistage work-
grain refinement (Fig. 21) as a result of DRX occurrence, and frag- hardening mechanism in machining with nose-radiused tools and its influence on
mentation of the δ particles (Fig. 16) due to shear deformation. All of machined subsurface quality and tool wear, Int. J. Adv. Manuf. Technol. 73 (2014)
545–555.
these factors contribute to the increased microhardness at the subsur-
[6] A. Chamanfar, S. Sarrat, M. Jahazi, M. Asadi, A. Weck, A.K. Koul, Microstructural
face layers compared to the unaffected bulk PM (Fig. 23). characteristics of forged and heat treated Inconel-718 discs, Mater. Des. 52 (2013)
Although compressive residual stresses induced by shot-peening are 791–800.
partly neutralized by the broaching tensile residual stresses especially [7] F. Klocke, D. Welling, A. Klink, D. Veselovac, T. Nöthe, R. Perez, Evaluation of
advanced wire-EDM capabilities for the manufacture of fir tree slots in Inconel 718,
at the surface and the immediate subsurface regions (up to 50 μm, Procedia CIRP 14 (2014) 430–435.
Fig. 26), the residual stresses after shot-peening in this study are rather [8] ASM Handbook Machining, 10th ed., vol. 16, ASM International, Metals Park, Ohio,
low when the current results are compared with those of Seddik et al. 1995, pp. 194–211.
[9] H.A. Kishawy, A. Hosseini, B. Moetakef-Imani, V.P. Astakhov, An energy based
[46] in Fig. 26. This indicates a rather milder severity of the shot-pe- analysis of broaching operation: cutting forces and resultant surface integrity, CIRP
ening process in the current study. Ann. Manuf. Technol. 61 (2012) 107–110.
As discussed, shot-peening induces beneficial work-hardening at [10] D.A. Gonçalves, R.B. Schroeter, Modeling and simulation of the geometry and forces

67
A. Chamanfar et al. Materials Characterization 132 (2017) 53–68

associated with the helical broaching process, Int. J. Adv. Manuf. Technol. 83 friction welded joints in dissimilar nickel based superalloys, Mater. Sci. Technol. 23
(2016) 205–215. (2007) 1408–1418.
[11] P. Vogtel, F. Klocke, H. Puls, S. Buchkremer, D. Lung, Modelling of process forces in [32] L. Brasche, B. Thompson, W. Hassan, T. Patton, T. Bantel, K. Smith, J. Lively,
broaching Inconel 718, Procedia CIRP 8 (2013) 409–414. J. Umbach, Nondestructive evaluation (NDE) of manufacturing induced anomalies,
[12] J.M. Zhou, V. Bushlya, R.L. Peng, S. Johansson, P. Avdovic, J.-E. Stahl, Effects of Engine Titanium Consortium – Quarterly Report, 2007, pp. 28–46.
tool wear on subsurface deformation of nickel-based superalloy, Procedia [33] B. Griffiths, Mechanisms of white layer generation with reference to machining and
Engineering 19 (2011) 407–413. deformation processes, J. Tribol. 109 (1987) 525–530.
[13] S.M. Afazov, A.A. Becker, T.H. Hyde, Effects of micro-stresses from machining and [34] A. Chamanfar, M. Jahazi, J. Gholipour, P. Wanjara, S. Yue, Evolution of flow stress
shot-peening processes on fatigue life, J. Adv. Manuf. Technol. 51 (2010) 711–722. and microstructure during isothermal compression of Waspaloy, Mater. Sci. Eng. A
[14] B.R. Sridhar, S. Ramachandra, U. Chandrasekar, Residual stress in nickel base super 615 (2014) 497–510.
alloy Udimet 720 for different surface conditions, Int. J. Eng. Sci. Technol. 3 (2011) [35] D.W. Schwach, Y.B. Guo, A fundamental study on the impact of surface integrity by
36–43. hard turning on rolling contact fatigue, Int. J. Fatigue 28 (2006) 1838–1844.
[15] S.-B. Kim, A. Evans, J. Shackleton, G. Bruno, M. Preuss, P.J. Withers, Stress re- [36] A.M. Wusatowska-Sarnek, B. Dubiel, A. Czyrska-Filemonowicz, P.R. Bhowal, N. Ben
laxation of shot-peened Udimet 720Li under solely elevated-temperature exposure Salah, J.E. Klemberg-Sapieha, Microstructural characterization of the white etching
and under isothermal fatigue, Metall. Mater. Trans. A 36A (2005) 3041–3053. layer in nickel-based superalloy, Metall. Mater. Trans. A 42A (2011) 3813–3825.
[16] Z. Chen, R. Lin Peng, J. Moverare, P. Avdovic, J.M. Zhou, S. Johansson, Surface [37] F. Klocke, P. Vogtel, S. Gierlings, D. Lung, D. Veselovac, Broaching of Inconel 718
integrity and structural stability of broached inconel 718 at high temperatures, with cemented carbide, Production Engineering Research and Development 7
Metall. Mater. Trans. A 47A (2016) 3664–3676. (2013) 593–600.
[17] Z. Chen, M.H. Colliander, G. Sundell, R.L. Peng, J. Zhou, S. Johansson, J. Moverare, [38] R.G. Andrews, A.K. Koul, P. Au, Fatigue crack initiation in alloy 718 at 650 °C, in:
Nano-scale characterization of white layer in broached Inconel 718, Mater. Sci. Eng. E.A. Loria (Ed.), Superalloys 718, 625 and Various Derivatives, The Minerals,
A 684 (2017) 373–384. Metals & Materials Society (TMS), Warrendale, PA, 1991, pp. 943–954.
[18] Z. Chen, J.J. Moverare, R.L. Peng, S. Johansson, D. Gustafsson, On the conjoint [39] http://www.grit-shot.com/News/493.html, 2017.
influence of broaching and heat treatment on bending fatigue behavior of Inconel [40] M.C. Shaw, Metal Cutting Principles, Oxford University Press, New York, 2005.
718, Mater. Sci. Eng. A 671 (2016) 158–169. [41] F. Klocke, S. Gierlings, D. Veselovac, Concept for temperature control in broaching
[19] A. Madariaga, J.A. Esnaola, E. Fernandez, P.J. Arrazola, A. Garay, F. Morel, Analysis nickel-based alloys, Key Eng. Mater. 523 (2012) 469–474.
of residual stress and work-hardened profiles on Inconel 718 when face turning [42] G.F. Vander Voort, R. Fowler, Low-load Vickers microindentation, Adv. Mater.
with large-nose radius tools, Int. J. Adv. Manuf. Technol. 71 (2014) 1587–1598. Process. 1 (2012) 28–33.
[20] J. Kwong, D.A. Axinte, P.J. Withers, The sensitivity of Ni-based superalloy to hole [43] Z. Chen, R.L. Peng, P. Avdovic, J. Zhou, J. Moverare, F. Karlsson, S. Johansson,
making operations: influence of process parameters on subsurface damage and re- Effect of thermal exposure on microstructure and nano-hardness of broached
sidual stress, J. Mater. Process. Technol. 209 (2009) 3968–3977. Inconel 718, EuroSuperalloys 2014- 2nd European Symposium on Superalloys and
[21] V. Bushlya, J.M. Zhou, F. Lenrick, P. Avdovic, J.-E. Ståhl, Characterization of white their Applications, EDP Sciences, French Riviera, Giens, France, 2014.
layer generated when turning aged Inconel 718, Procedia Engineering 19 (2011) [44] S. Ghosh, V. Kain, Microstructural changes in AISI 304L stainless steel due to sur-
60–66. face machining: effect on its susceptibility to chloride stress corrosion cracking, J.
[22] A.L. Ortiz, J.W. Tian, J.C. Villegas, L. Shaw, P. Liaw, Interrogation of the micro- Nucl. Mater. 403 (2010) 62–67.
structure and residual stress of a nickel-base alloy subjected to surface severe plastic [45] A. Chamanfar, M. Jahazi, J. Cormier, A review on inertia and linear friction welding
deformation, Acta Mater. 56 (2008) 413–426. of nickel-based superalloys, Metall. Mater. Trans. A 46A (2015) 1639–1669.
[23] D. Ulutan, T. Ozel, Machining induced surface integrity in titanium and nickel al- [46] R. Seddik, M. Seddik, A. Atig, R. Fathallah, Thermo-mechanical relaxation of
loys: a review, Int. J. Mach. Tools Manuf. 51 (2011) 250–280. compressive residual stresses induced by shot peening, Procedia Structural Integrity
[24] B. Zou, M. Chen, C. Huang, Q. An, Study on surface damages caused by turning 2 (2016) 2182–2189.
NiCr20TiAl nickel-based alloy, J. Mater. Process. Technol. 209 (2009) 5802–5809. [47] F. Taricco, Effect of machining and shot peening on the residual stresses of super-
[25] Rolls-Royce plc Document Number CME5043, Section 1, (2010), pp. 1–25. alloy turbine discs, International Gas Turbine and Aeroengine Congress and
[26] J.M. Pereira, B.A. Lerch, Effects of heat treatment on the ballistic impact properties Exposition ASME, Houston, Texas, 1995, pp. 1–6.
of Inconel 718 for jet engine fan containment applications, Int. J. Impact Eng. 25 [48] K. Jacobus, R.E. DeVor, S.G. Kapoor, Machining-induced residual stress: experi-
(2001) 715–733. mentation and modeling, J. Manuf. Sci. Eng. 122 (2000) 20–31.
[27] T. Connolley, M.J. Starink, P.A.S. Reed, Effect of broaching on high-temperature [49] T.E. Strangman, Thermal-mechanical fatigue life model for coated superalloy tur-
fatigue behavior in notched specimens of Inconel 718, Metall. Mater. Trans. A 35A bine components, in: S.D. Antolovich, R.W. Stusrud, R.A. McKay, D.L. Anton,
(2004) 771–783. T. Khan, R.D. Kissinger, D.L. Klarstrom (Eds.), Superalloy 1992, TMS, Warrendale,
[28] A. Thakur, S. Gangopadhyay, State-of-the-art in surface integrity in machining of PA, 1992, pp. 795–804.
nickel-based super alloys, Int. J. Mach. Tools Manuf. 100 (2016) 25–54. [50] W. Cao, M. Khadhraoui, B. Brenier, J.Y. Guédou, L. Castex, Therrnorneehanieal
[29] R. M'Saoubi, T. Larsson, J. Outeiro, Y. Guo, S. Suslov, C. Saldana, S. Chandrasekar, relaxation of residual stress in shot peened nickel base superalloy, Mater. Sci.
Surface integrity analysis of machined Inconel 718 over multiple length scales, CIRP Technol. 10 (1994) 947–954.
Ann. -Manuf. Technol. 61 (2012) 99–102. [51] N.A. Abukhshim, P.T. Mativenga, M.A. Sheikh, Heat generation and temperature
[30] P.R. Bhowal, D. Stolz, A.M. Wusatowska-Sarnek, R. Montero, Surface effects on low prediction in metal cutting: a review and implications for high speed machining,
cycle fatigue behavior in In718 alloy, in: R.C. Reed, K.A. Green, P. Caron, T.P. Gabb, Int. J. Mach. Tools Manuf. 46 (2006) 782–800.
M. Fahrmann, E.S. Huron, S.A. Woodard (Eds.), Superalloy 2008, The Minerals [52] S.P. Mo, D.A. Axinte, T.H. Hyde, N.N.Z. Gindy, An example of selection of the
Metals & Materials Society (TMS), Champion, Pennsylvania, USA, 2008, pp. cutting conditions in broaching of heat-resistant alloys based on cutting forces,
417–427. surface roughness and tool wear, J. Mater. Process. Technol. 160 (2005) 382–389.
[31] H.Y. Li, Z.W. Huang, S. Bray, G. Baxter, P. Bowen, High temperature fatigue of

68