Thinning Process of Recast Layer in Hole Drilling and Trimming by EDM

Abstract

In order to reduce the recast layer thickness of film cooling holes on turbine blade, a novel thinning process of electrical discharge machining (EDM) is investigated. To avoid the contradiction between holes drilling speed and recast layer of nickel-based superalloy, the EDM holes process consists of two steps, the drilling process for high machining speed and the trimming process for thin recast layer. To measure and evaluate the uneven and inconsistent recast layer thickness of film cooling holes, a mesh partition and multi-point averaged measurement by optical microscope is discussed. Experimental investigation is conducted to determine the EDM drilling and trimming parameters in machining Inconel 738 with 6mm thickness. By changing the discharge polarity in the trimming process, the recast layer thickness is obviously reduced. Over the range of parameters tested, the recast layer thickness can be reduced from 10.472μm to 2.416μm, and the entire process including drilling and trimming could be completed within 20 s. The experimental results show that the hybrid process containing drilling and trimming by EDM is feasible for film cooling holes with thin recast layer thickness.

Full text

Procedia CIRP 42 ( 2016 ) 575 – 579
Available online at www.sciencedirect.com
2212-8271 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of 18th CIRP Conference on Electro Physical and Chemical Machining (ISEM XVIII)
doi: 10.1016/j.procir.2016.02.262
ScienceDirect
18th CIRP Conference on Electro Physical and Chemical Machining (ISEM XVIII)
Thinning process of recast layer in hole drilling and trimming by EDM
Li CJ a, b, Li Y a, b,*, Tong H a, b, Zhao L a, b
a Department of Mechanical Engineering, Beijing Key Lab of Precision / Ultra-precision Manufacturing Equipments and Control,
Tsinghua University, Beijing 100084, China;
b State Key Laboratory of Tribology, Tsinghua University, Beijing 100084, China;
* Corresponding author. Tel.: +86-010-62772212; fax: +86-010-62772212. E-mail address: liyong@mail.tsinghua.edu.cn
Abstract
In order to reduce the recast layer thickness of film cooling holes on turbine blade, a novel thinning process of electrical discharge machining
(EDM) is investigated. To avoid the contradiction between holes drilling speed and recast layer of nickel-based superalloy, the EDM holes process
consists of two steps, the drilling process for high machining speed and the trimming process for thin recast layer. To measure and evaluate the
uneven and inconsistent recast layer thickness of film cooling holes, a mesh partition and multi-point averaged measurement by optical
microscope is discussed. Experimental investigation is conducted to determine the EDM drilling and trimming parameters in machining Inconel
738 with 6 mm thickness. By changing the discharge polarity in the trimming process, the recast layer thickness is obviously reduced. Over the
range of parameters tested, the recast layer thickness can be reduced from 10.472 μm to 2.416 μm, and the entire process including drilling and
trimming could be completed within 20 s. The experimental results show that the hybrid process containing drilling and trimming by EDM is
feasible for film cooling holes with thin recast layer thickness.
© 2016 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the organizing committee of 18th CIRP Conference on Electro Physical and Chemical Machining (ISEM
XVIII).
Keywords: EDM drilling, Drilling and trimming, Recast layer, Film cooling holes;
1. Introduction
Film cooling holes are widely used in the advanced aircraft
engine to obtain extremely high turbine firing temperatures and
subsequent high efficiencies [1], and their fabrication requires
high machining speed and accuracy, as well as good surface
quality. A turbine blade may have several hundreds of film
cooling holes with diameter Φ0.20.8 mm [2, 3]. The film
cooling holes of nickel-based superalloys are difficult to
process by traditional machining. Electrical discharge
machining (EDM) is one of the efficient machining processes
for film cooling holes [3-5].
EDM will generate the recast layer which affects the service
life of turbine blades [5, 6]. Recently, there are three effective
thinning processes of recast layer. One is the hybrid process of
EDM and electrochemical machining (ECM), of which the
pulse power generator and working fluid in the same station
need to be switched, to thin the recast layer through the
ECM[7,8]. But this hybrid process will cause large size
variation of inlet and outlet, and introduce switching time in the
process. Secondly, some subsequent treatment processes are
used after the EDM, such as abrasive flow machining or
chemical polishing [9], which can reduce the recast layer
thickness. But the process has low efficiency. Thirdly, the
parameters of EDM process need to be optimized to reduce the
recast layer [10-12]. In 2004 and 2009, Lee H et al. have
reported that the discharge gasification rate increased when
discharge thermal effect was more concentrated using the high
discharge current[13,14]. This method can reduce the recast
layer, but the recast layer thinning effect was limited.
In 2014, Klocke F et al. have developed the five cutting
process of wire EDM with different discharge pulse types and
electrode materials to reduce the recast layer [15]. The recast
layer can be controlled within 1μm by matching the cutting
offset and discharge energy in the trimming process. The five
cutting process also reduces the processing efficiency.
The EDM drilling is different from the wire EDM, due to the
difficulty to set the offset in the drilling process. In this paper,
taking account of the processing efficiency and recast layer, the
thinning process of drilling and trimming by EDM was
© 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of 18th CIRP Conference on Electro Physical and Chemical Machining
(ISEM XVIII)

576 Li Cj et al. / Procedia CIRP 42 ( 2016 ) 575 – 579
investigated. The base holes were drilled with the optimized
parameters that were obtained by the orthogonal experiment. In
the trimming process, the recast layer thickness was thinned by
changing the discharge polarity and open circuit voltage.
2. Showing and measure of recast layer
2.1. Showing of recast layer
The metallographic structure of recast layer is different from
the bulk metal because it is generated by the rapid quenching of
the partially molten alloy. Therefore, it is easy to distinguish
them via showing the metallographic structure. There are
different physical or chemical properties of the phase separation
and boundary between the recast layer and bulk. By utilizing
these differences, various color and light intensities might be
rendered by the different parts under optical microscope.
Furthermore, the bulk alloy and recast layer can be
distinguished.
There are four steps in the observation of metallographic
structure of workpiece. The workpiece with processed structure
was inlayed with epoxy resin material firstly. The target surface
of the workpiece was coarsely ground, finely ground and
polished until a smooth mirror is achieved in this step. Then,
the target surface was chemically etched to display its
microstructure. Finally, the microstructure on the grinding
surface was observed under optical microscope. The equipment
and parameters are shown in Table 1.
Table 1 Observing process recast layer
Process
Device
Parameter
Inlay
Epoxy resin
150ć
Rough grinding
Coarse abrasive paper
200#ǃ400#
Finish grinding
Fine abrasive paper
800#ǃ2000#
Polishing
Polishing cloth
2.5μmǃ0.5μm
Chemical corrosion
AlcoholǃHClǃCuCl
Observe
VHX-600ǃSEM
Recast layer
Epoxy
resin
Bulk
Radial section
Axial section
Epoxy
resin
a) b) c)
Fig. 1 Recast layer picture with optical microscope
Recast layer observed by optical microscope is shown in
Fig.1. Recast layer (arrowed in Fig. 1b) is the brighter part. It is
significantly different from the bulk alloy (surrounding darker
region) and inset material (black) in contrast. Therefore the
thickness of recast layer at a particular point of hole wall could
be measured using optical microscope. Fig. 2 is the SEM
(scanning electron microscopy) image of recast layer. The
obvious boundary between the bulk alloy and recast layer is
observed in the SEM image and micro cracks propagated into
the region of recast layer. According to the results of Fig. 1 and
Fig. 2, it reveals that the thickness of recast layer is not uniform
along both circumferential and axial direction. So the recast
layer measurement remains to be a challenge. The thickness of
recast layer at particular measuring points cannot present the
recast layer thickness precisely.
Bulk
Epoxy r esin
Recast l ayer
Fig. 2 Recast layer picture with SEM
2.2. Measurement of recast layer thickness
Other foregone literatures also suggest that the thickness
distribution of recast layer is non-uniform. So it is difficult to
measure it. In 2009, researchers from Georgia Institute of
Technology in the United States got the thickness distribution
of recast layer[16]. The average thickness of recast layer was
calculated via measuring the layer area and dividing it by
circumference. In 2005, researchers in Jadavpur University in
India measured about 9 measuring points under optical
microscope and then obtained the average thickness[17]. The
results indicated that recast layer thickness along the
circumferential and axial direction are non-uniform and
discontinuous. So it is more appropriate to represent thickness
of recast layer with the average value. Also the average
thickness could be calculated by measuring multiple points.
However, the location and quantities of the measuring points
directly affect the precision measurement. Naturally, larger
quantity and more uniform distribution of the measuring points
could lead to a more precise measurement result as compared
to the actual thickness. There might be a few human factors
under optical microscopy, for the uniformity of sampling
cannot be guaranteed and larger number of measuring points
may intensify the labor.
Taking into account of those measurement conditions, the
microscope grid measuring function and Excel save function
were used in this paper and the mesh partition measuring
method was proposed. Firstly, the location of the measuring
points were determined as the intersection of grid lines and side
wall of holes. The thickness of valid recast layer at each
measuring point was obtained by the microscope. Finally, the
recorded data were exported and analyzed using Microsoft
Excel to take the average value as the recast layer thickness of
holes with EDM.
Mesh partition and multi-point averaged measurement
method for recast layer thickness in this paper is shown as Fig.3.
The intersection point of grid lines and side wall of holes are
regarded as the measuring point, then the recast layer thickness
was measured along the diameter direction. When there is more
than one intersection between a particular grid line and the

577
Li Cj et al. / Procedia CIRP 42 ( 2016 ) 575 – 579
recast layer, only the largest value was included in the statistics.
As illustrated in Fig. 3, the data for point tT1 was taken but the
data for tR1 was discarded.
Bulk
Measuring
point
Recast
layer
Epoxy
resin
Fig. 3 Mesh partition measure method for recast layer thickness
3. Optimization experiment for film cooling holes
machining parameters
3.1. Orthogonal experiment design
Orthogonal experimental design method is a kind of
scientific method of dealing with multiple factors experiment.
The experiments are reasonably arranged using orthogonal
table in the orthogonal experimental design method. By using
the mathematical statistics principle, the experimental results
would be scientifically analyzed. The five processing
parameters as pulse width, pulse interval, open circuit voltage,
peak current and inner flushing pressure, which have important
influence for the thickness of recast layer in this experiment.
The quasi-level orthogonal experimental method is adopted due
to the open circuit voltage has only four levels. The processing
parameter table is shown in Table 2. Peak current, pulse width,
pulse interval and flushing pressure have five levels and the
machining voltage has four levels. The orthogonal experiment
table is a format as L25 (56).
Table2. Processing parameter for film cooling hole with EDM
Pulse
width
(μs)
Pulse
Interval
(μs)
Voltage
(V)
Current
(Range)*
Flushing
pressure
(MPa)
Level 1
8
32
60
2
1.2
Level 2
16
40
90
3
1.8
Level 3
24
48
120
4
2.4
Level 4
32
56
150
5
3.0
Level 5
40
64
90
6
3.6
* Higher current range indicates larger current.
3.2. Analysis of experimental results
In the experiment of EDM, the brass hollow cylindrical
electrode with diameter of 0.5 mm is used as tool electrode. And
the workpiece is Inconel 738 nickel-based superalloy with
thickness of 6 mm. Deionized water is chosen as working fluid
and the experiment was conducted under the positive polarity
machining mode. The revolving rate of electrode was 120 r/min.
Liquid flowed simultaneously inside and outside of the hollow
cylindrical electrode in the working fluid circulating system.
The flushing pressure is adjustable in the range of 0 ~ 4.2 MPa.
In order to observe experimental regularity perfectly, three
holes are machined with each set of parameters for averaging
and 75 valid holes are obtained in the experiment. The
experimental results are shown in Table 3.
Table3. Quasi-level orthogonal experimental table and results
Volt a
-ge
(V)
Curr-
ent
(R)
Pulse
width
(μs)
Pulse
interval
(μs)
Flushing
pressure
(MPa˅
Recast
layer
˄μm˅
Process
time
˄s˅
1
60
2
8
32
1.2
5.18
1025
2
60
3
16
40
1.8
6.79
347
3
60
4
24
48
2.4
9.74
77.3
4
60
5
32
56
3.0
10.60
45.7
5
60
6
40
64
3.6
12.74
33.25
6
90
2
16
48
3.0
7.85
46.7
7
90
3
24
56
3.6
11.14
28.3
8
90
4
32
64
1.2
9.92
157
9
90
5
40
32
1.8
16.04
21.3
10
90
6
8
40
2.4
6.29
97
11
120
2
24
64
1.8
9.49
60
12
120
3
32
32
2.4
18.58
59
13
120
4
40
40
3.0
12.62
10.3
14
120
5
8
48
3.6
6.07
61
15
120
6
16
56
1.2
10.58
38
16
150
2
32
40
3.6
14.68
10.7
17
150
3
40
48
1.2
25.80
23.3
18
150
4
8
56
1.8
6.85
97
19
150
5
16
64
2.4
7.96
44.3
20
150
6
24
32
3.0
15.70
11
21
90
2
40
56
2.4
12.28
42
22
90
3
8
64
3.0
6.33
112
23
90
4
16
32
3.6
7.78
28
24
90
5
24
40
1.2
9.81
64.6
25
90
6
32
48
1.8
13.71
21.6
The variance analysis of experimental results is conducted
with Statistical Product and Service Solutions (SPSS) software.
Analysis results are shown in Table 4, in which, smaller the
value of Sig value indicates more influential impact factor on
target. According to the analysis results, the influence of
various factors on recast layer thickness is: pulse width > open
circuit voltage > peak current > pulse interval > flushing
pressure. The influence of various factors on the machining
efficiency is: open circuit voltage > pulse width > flushing
pressure > peak current > pulse interval.
The processing time and recast layer thickness should be
considered in EDM process of film cooling holes. In order to
optimize the processing parameters for shorter processing time
and thinner recast layer, the grey correlation degree analysis
method is adopted in this paper. If the variation trend of two
factors is consistent, namely with higher synchronous variation
degree, the two factors have greater correlation. Nevertheless,
the correlation degree of them is deemed to be small. In the
analysis of experimental results, there is no comparability over
different technology targets due to the different dimensions and
different orders of magnitude. In the grey correlation method,

578 Li Cj et al. / Procedia CIRP 42 ( 2016 ) 575 – 579
the incomparable items would be converted to comparable ones
through the de-unitization process. That is the process of the
greyscale generation. Finally, the grey correlation degree is
confirmed with Distinguished coefficient in the interval of [0,
1]. The levels of grey correlation degree for various processing
parameters are achieved.
Table. 4 Variance analysis of experimental results with SPSS
Recast layer thickness
Machining time
parameter
MS
F
Sig
MS
F
Sig
check
model
26.94
9.76
0.010
30429.36
2.65
0.142
voltage
26.64
9.65
0.016
49211.47
4.29
0.076
current
15.93
5.77
0.041
10188.65
0.89
0.533
pulse
width
77.13
27.9
5
0.001
40035.74
3.49
0.102
pulse
interval
12.21
4.43
0.067
21480.71
1.87
0.254
Flushing
pressure
2.72
0.99
0.492
35925.76
3.13
0.121
The maximum correlation between recast layer thickness
and the processing time under the condition of the variable
processing parameters can optimize the recast layer thickness
and the optimal parameters of machining efficiency. The
optimal processing parameters are open circuit voltage of 90V,
peak current of range 4 (24A), pulse width of 24μs, pulse
interval of 48μs and flushing pressure of 3.6 MPa. According
to the above analysis, the optimal processing parameters are
adopted for verified experiment. The results of the recast layer
thickness of 7.2μm and processing time of 24s are achieved in
the verified experiment.
4. Rough and trim thinning process of recast layer
In order to reduce the recast layer thickness to 5 μm or less,
the thinning process of recast layer was investigated by drilling
and trimming method. The base holes are drilled by the
optimizing parameters achieved in section 3.2. And in the
trimming process, the experiments of changing the machining
polarity and voltage are studied to reduce the recast layer
thickness.
Open circuit voltage of pulse power generator can be
regulated from 60 V to 200 V in the EDM drilling machine. The
base holes are drilled by the parameters, open circuit voltage 90
V, current peak level 4, discharge pulse width 24 μs, pulse
interval 32 μs and positive polarity machining (the workpiece
is positive). The voltages were 60 V, 120 V, and 150 V
respectively in the trimming process, the discharge pulse width
was 24 μs, pulse interval was 32 μs, and the current was level
4. The workpiece was nickel-based superalloy Inconel 738 with
6mm thickness. The electrode was a hollow brass material with
a diameter of 0.5 mm and the electrode rotating speed of 120
r/min. The working fluid is deionized water. The cooling type
is internal and external flushing, with 3.6 MPa inner flushing
pressure. Three holes are drilled by the same parameters for
improving the experiment result and the recast layer thickness
is measured by the mesh and multi-point averaged
measurement method.
The recast layer effect of changing polarity and open circuit
voltage is shown in Fig.4. Fig. 4 shows the recast layer
thickness. As shown the Fig.4, the first column presents the
base holes recast layer and the others present the recast layer of
60 V, 120 V, 150 V trim respectively. It is obvious that the
recast layer thickness reduces with the voltage increasing. The
recast layer thickness average is only 2.416 μm, reduced rate
76.9% compared with the base holes recast layer 10.472 μm.
Fig. 4 Discharge polarity effect for recast layer
To improve the EDM drilling efficiency of the base holes,
the pulse interval is changed from the optimizing parameter 48
μs to 32 μs. The base holes drilling time is 16 s, but the base
holes recast layer thickness increased from the 7.117 μm to
10.472 μm compared the verified experiment result in section
3.2. The time for trimming the holes is 3 4 s and the
machining time of the entire hole is approximately 20 s.
The recast layer comparison between base hole and
trimming hole is shown in Fig.5. As shown the Fig.5, the recast
layer after trimming is obviously thinner than the base holes.
Fig. 6 shows the SEM image of the recast layer with trimming.
The recast layer is very thin and has gap and micro crack.
The drilling process mainly improves the machining
efficiency and the recast layer is thinned in the trimming
process. In the trimming process, the discharge erosions are
easily removed and the secondary discharge probability is less,
so the recast layer gets reduced. The discharge polarity is
distinct to thin the recast layer. The discharge gap is big and
after changing the discharge polarity, the probability of erosion
sputtering to the workpiece decreases in the trimming process.
The energy of negative polarity machining applied to
workpiece is weaker compared to positive polarity machining,
and the discharge pits become small. Based on the above
reasons, the drilling and trimming can reduce the recast layer
thickness. The experiments showed that the recast layer
thinning process of drilling and trimming by EDM is feasible.
Epoxy
resin
Recast layer
Bulk
10μm
a)
Epoxy
resin
Bulk
Recast layer
10μm
b)
Fig.5 a) Recast of base holes, b) Recast layer of changing machining polarity

579
Li Cj et al. / Procedia CIRP 42 ( 2016 ) 575 – 579
Epoxy
resin
Recast layer
Bulk
Fig. 6 2.4μm thickness of recast layer (5000 × SEM)
Conclusions
In order to reduce the recast layer thickness of the film
cooling holes drilled by EDM, the thinning process of drilling
and trimming was investigated. The orthogonal experiment
was researched to optimize the discharge machining
parameters. The base holes were EDM-drilled with the
optimized parameters and the discharge polarity was changed
in the trimming process. The experiments showed that the
recast layer thinning process of drilling and trimming by EDM
was feasible.
1) The mesh partition and multi-point averaged
measurement by optical microscope was discussed to improve
the measurement accuracy of recast layer thickness, of which
was found to be non-uniform and discontinuous.
2) The orthogonal experiment of open circuit voltage, peak
current, discharge pulse width, pulse interval and inner flushing
pressure was researched and the discharge machining
parameters were optimized with the SPSS software. The
verified experiment of Inconel 738 with 6 mm thickness
showed that the recast layer thickness can be thinned to 7.117
μm and the EDM drilling time was 24 s using the optimized
parameters.
3) In the thinning process of recast layer, the discharge
polarity was distinct to recast layer, and the recast layer
thickness was reduced with increasing open circuit voltage.
4) The recast layer thickness can be reduced from 10.472
μm to 2.416 μm, and the whole process time including drilling
and trimming was within 20 s with the trimming parameters,
open circuit voltage 150 V and negative discharge model.
Acknowledgements
The research is supported by National Natural Science
Foundation of China (Grant No. 51305230), Autonomous
Scientific Research Project of Tsinghua University (Grant No.
20131089209) and Beijing Higher Education Young Elite
Teacher Project (Grant No. YETP0084).
References
[1] Klocke F, Klink A, Veselovac D,Aspinwall DK,Soo SL, et al. Turbo
machinery component manufacture by application of electrochemical,
electro-physical and photonic processes, CIRP Ann.–Manuf. Technol.
2014; 63: 703–726.
[2] Bogard DG, Thole KA. Gas tu rbine film cooling. Journ al of propulsion and
power. 2006; 22(2): 249-270.
[3] Bunker RS, A review of shaped hole turbine film-cooling technology, J.
Heat Transf. 2005;127: 441–453.
[4] Voisey KT, Clyne TW, Laser drilling of cooling holes through plasma
sprayed thermal barrier, Surface and coatings technology. 2004;176: 296-
306.
[5] Ekmekci B, Residual stresses and white layer in elctric discharge machining
(EDM), Applied surface science. 2007; 253: 9234-9240.
[6] Wang C, Chow H, Yang L, et al. Recast layer removal after electrical
discharge machining via Taguchi analysis: A feasibility study. Journal of
Materials Processing Technology, 2009;209(8): 4134-4140.
[7] Zhang Y, Xu ZY, Zhu D, Xing J. Tube electrode high-speed
electrochemical discharge drilling using low-conductivity salt solution.
International Journal of Machine Tools & Manufacture. 2015;92:10–18.
[8] Wang W, Zhu D, Qu N, Huang S, Fang X, Electrochemical drilling with
vacuum extraction of electrolyte, Journal of Materials Process Technology.
2010;210: 238–244.
[9] Lin YC, Chow HM, Yan BH, Tzeng HJ, Effects of finishing in abrasive
fluid machining on micro-holes fabricated by EDM. Int J Adv Manuf
Technol, 2007;33: 489–497.
[10] Kuppan P, Rajadurai A, Narayanan S. Influence of EDM process
parameters in deep hole drilling of Inconel 718. Int J Adv Manuf Technol.
2008;38:74–84.
[11] Liu HS, Yan BH, Chen CL, et al. Application of micro-EDM combined
with high-frequency dither grinding to micro-hole machining. International
Journal of Machine Tools and Manufacture, 2006, 46(1):80-87.
[12] Lee H T, Tai T Y. Relationship between EDM parameters and surface
crack formation. Journal of Materials Processing Technology, 2003,
142(3):676-683.
[13] Lee H, Hsu F, Tai T. Study of surface integrity using the small area EDM
process with a copper–tungsten electrode. Materials Science and
Engineering: A, 2004, 364(1-2):346-356.
[14] Lee H T, Liu C. Optimizing the EDM hole-drilling strain gage method for
the measurement of residual stress. Journal of Materials Processing
Technology, 2009, 209(15-16):5626-5635.
[15] Klocke F, Welling D, Klink A, Veselovac D, Nöthe T, Perez R. Evaluation
of Advanced Wire-EDM Capabilities for the Manufacture of Fir Tree Slots
in Inconel 718. Procedia CIRP 14, 2014 : 430–435.
[16] Newtona TR, Melkote SN. Investigation of the effect of process
parameters on the formation and characteristics of recast layer in wire-
EDM of Inconel 718 .Materials Science and Engineering A. 2009;(513–
514): 208–215.
[17] Pur A.B. , Bhattacharyya B. Modeling and analysis of white layer depth
in a wire-cut EDM process through response surface methodology. Int J
Adv Manuf Technol. 2005; 25: 301-307.