Optics & Laser Technology 53 (2013) 2232

Contents lists available at SciVerse ScienceDirect

Optics & Laser Technology

journal homepage: www.elsevier.com/locate/optlastec

Characterization of hole circularity and heat affected zone in pulsed

CO2 laser drilling of alumina ceramics
A. Bharatish a, H.N. Narasimha Murthy a,n, B. Anand a, C.D. Madhusoodana b,
G.S. Praveena a, M. Krishna a
a
b

Department of Mechanical Engineering, RV College of Engineering, Bangalore, India

Ceramic Technological Institute, BHEL, Bangalore, India

art ic l e i nf o

a b s t r a c t

Article history:
Received 21 December 2012
Received in revised form
3 April 2013
Accepted 8 April 2013
Available online 18 May 2013

Circularity of drilled hole at the entry and exit, heat affected zone and taper are important attributes
which inuence the quality of a drilled hole in laser drilling. This paper examines the effect of laser
parameters on the quality of drilled holes in Alumina ceramics which are widely used in microelectronic
devices, based on orthogonal array experimentation and response surface methodology. Both entrance
and exit circularities were signicantly inuenced by hole diameter and laser power. Heat affected zone
was inuenced by frequency. Taper was also signicantly inuenced by laser power. Response surface
model predicted nominal entrance circularity at 2.5 kHz, 240 W, 2.5 mm/s, 1 mm hole, exit circularity
and taper at 7.5 kHz, 240 W, 4.5 mm/s, and 1 mm hole. The model predicted lowest heat affected zone at
7.5 kHz, 240 W, 2.5 mm/s, and 1 mm. Multiobjective optimization achieved using both response surface
model and gray relational analysis indicated that all the four quality parameters are optimized at 7.5 kHz,
240 W, 3.85 mm/s and 1 mm.
& 2013 Elsevier Ltd. All rights reserved.

Keywords:
Laser drilling
Hole circularity
Gray relational analysis

1. Introduction
Laser drilling is an inexpensive and controllable alternative to
CNC, punching, wire EDM, broaching or other popular destructive
hole drilling methods. The use of laser drilling in manufacturing
industry can be attributed to high accuracy, repeatability and
reproducibility in a short period of time with minimum wastage
[1]. It is widely applied in drilling of printed circuit boards,
chemical vapor deposited diamond, microelectronic packaging
and super luminescent diodes [27].
Alumina of 99% purity is one of the most widely used ceramic
substrates in ballistic tiles and for thin lm circuits in electronic
switching systems. Laser drilling is a preferred method in drilling
alumina due to its unique properties such as non-contacting
nature, absence of mechanical cutting force and tool wear, exibility and high machining rate [8]. CO2 laser has excellent
reliability characteristics, high average beam power, better efciency and good beam quality [9]. The laser drilled hole quality is
mainly judged by circularity, taper and extent of heat affected zone
[10].
Most of the researchers have adopted Nd:YAG laser to drill
various metals, alloys and ceramics. Kacar et al. [11] investigated

Corresponding author. Tel.: +91 80 671 78106; fax: +91 80 28602148.

E-mail address: hnmdatta@yahoo.com (H.N. Narasimha Murthy).

0030-3992/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.optlastec.2013.04.010

10 mm thick 99% alumina ceramic plates using 600 W Nd:YAG

percussion laser drilling. Effects of laser parameters such as peak
power, pulse energy and pulse duration were examined on average
hole diameter. The authors reported that there was no variation of
entrance hole diameter with respect to pulse duration and peak
power whereas exit hole diameter of the craters showed linear
variation with respect to these parameters. Hanon et al. [12]
investigated the effects of laser parameters such as peak power,
pulse duration, focal position and repetition rate on the alumina
ceramic plaques of 5 mm and 10.5 mm thickness using 600 W Nd:
YAG laser. The authors identied three different layers namely a
thin layer, a resolidied material inside the hole and a recast layer
at the entrance region of the hole. The crater depth increased with
the number of pulses due to insufcient recoil pressure inside the
cavity. Yinghou et al. [13] investigated the effects of laser parameters such as peak power, pulse laser power and pulse repetition
rate on spatter deposition and hole diameter by drilling 4.4 mm
thick alumina ceramic (95%) using 3.5 kW CO2 laser. The authors
reported that the spatter deposition and hole diameter increased
with increase in peak power and the size and temperature of the
melt front signicantly affected both spatter and hole formation.
The authors observed that HAZ was incorporated with three zones
namely recast layer, over grown grains zone and partial melting
zone.
Ng and Li [14] assessed the effects of laser peak power and pulse
width on the repeatability of hole geometry in terms of percentage

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

standard deviation in percussion drilling of 2 mm thick stainless steel

sheets using 400 W Nd:YAG laser. The authors observed that the
increase in interaction time between liquid melt and material
resulted in higher degree of erosion. Better repeatability was
achieved when higher peak power and smaller pulse width were
used thereby a cause of hole geometry variation was identied. Ng
and Li [15] investigated the inuence of laser peak power, pulse
width and melt ejection process on the repeatability of holes in terms
of circularity and percentage standard deviation by drilling in 2 mm
thick mild steel sheets using 400 W Nd:YAG laser. The authors
observed greater spatter thickness for longer pulse width and lower
peak power. Mutlu et al. [16] investigated the inuence of laser
wavelength and operation pressure on the crater depth and diameter
in drilling 0.8 mm alumina ceramic plates using Nd:YAG pulsed laser.
The crater depth and diameter increased non-linearly with laser
power because of the plasma shielding effect. Both wavelength and
ambient pressures showed the same characteristics.
Since laser drilling is associated with several parameters
development of a physical model becomes complicated and hence
researchers have developed statistical models for taper [10,1721],
circularity [1,20,23] and HAZ [21,17] to analyze parameters such as
laser peak power, laser power, pulse width, pulse frequency, focus
plane position, number of pulses and assist gas pressure. But, no
comprehensive research has been reported in the area of multiobjective optimization of the identied responses in laser drilling
of alumina ceramic using gray relation analysis and response
surface methodology. Circularity at the hole exit is an equally
important quality parameter which is scarcely addressed so far.
The aim of this research was to investigate the effects of laser
parameters on the quality of drilled holes including circularity at
the exit based on Taguchi's Orthogonal Array technique. The laser
parameters and responses were correlated using Response Surface
Methodology (RSM). The laser parameters for optimizing the
multiple responses were obtained using gray relational analysis.
Laser drilling experiments were carried out using 300 W CO2 laser
and 2 mm thick alumina ceramic plates. Pulse frequency, laser
power, scanning speed and diameter of the hole as the main
factors and taper angle, circularity at entrance and exit and heat
affected zone as responses were studied. Experimental HAZ and
taper were validated using analytical models. The quality of drilled
holes was examined using SEM.

2. Experimental
2.1. Laser drilling using orthogonal array technique
A 300 W CO2 Ron laser was used to assess the quality of
drilled holes in 2 mm thick alumina (99%) plates the properties of
which are presented in Table 1. The technical specications of
laser system are as shown in Table 2. Pulsed laser trepanning
technique was used, in which the laser beam movement relative to
the part causes to cut a contour out of the plate. In this study, the
experimental plan had four controllable variables, namely pulse
frequency, laser power, scanning speed and hole diameter. Based
on the preliminary experiments, the range of pulse frequency,

Table 1
Properties of alumina ceramic used for drilling.
Density (g/cm3)
Thermal expansion coefcient (106 J/K)
Thermal conductivity (W/mK)
Young's modulus (GPa)
Poisson's ratio
Hardness (HV1.0)

3.893.96
6.48.2
30.4
330400
0.260.24
13001700

23

Table 2
Technical specications of laser system.
Suppliers
Machine model
Wavelength
Maximum eld size
Spot diameter
Working distance
Resolution
Beam type
Peak power
Mode of operation

M/S Suresh Indu Laser Pvt. Ltd., Pune

RonSinar SC  30 sealed slab 300 WE CO2 laser
10.6 mm
300 mm  300 mm
370 mm
480 mm
<6 mm
TEM00
750 W
Continuous and pulsed mode

Table 3
L9 Orthogonal array experimental layout.
Expt.
no.

1
2
3
4
5
6
7
8
9

Experimental factors
Frequency, Fr
(kHz)

Laser power,
Lp (W)

Scanning speed, Scp Hole diameter,

(mm/s)
D (mm)

2.5
2.5
2.5
5
5
5
7.5
7.5
7.5

150
230
240
150
230
240
150
230
240

2.5
3.5
4.5
3.5
4.5
2.5
4.5
2.5
3.5

1
1.5
2
2
1
1.5
1.5
2
1

laser power, scanning speed and hole diameter were selected as

2.57.5 kHz, 150240 W, 24 mm/s and 12 mm respectively.
When laser drilling was attempted on alumina ceramic plates by
employing laser parameters greater than 240 W laser power,
7.5 kHz pulse frequency and 4 mm/s scanning speed, the specimens cracked due to their lower thermal shock resistance.
Similarly, laser drilling attempted by employing laser parameters
less than 150 W laser power, 2.5 kHz pulse frequency and 2 mm/s
scanning speed, through hole drilling was not possible due to
insufcient laser energy for the drilling. Thus, the parameters were
selected between these higher and lower limits for the investigation. The diameter of the hole to be drilled was varied using
WELDMARK software interfaced with the laser system. The laser
power was varied in terms of duty cycle. L9 (34) Orthogonal Array
experimental layout selected for the research is shown in Table 3.
Four replicates were used to ensure repeatability of the responses.
The experimental responses along with their mean and standard
deviation are presented in Table 4.

2.2. Scheme for measuring the responses

The circularity of the hole at the entry and the exit was
determined by using Eqs. (1) and (2) respectively as
C ent

Dmin
Dmax

C ext

dmin
dmax

where Dmin and Dmax as shown in Fig. 1a are the minimum and
maximum diameters at the entrance of the hole respectively.
Similarly, dmin and dmax are the minimum and maximum diameters at the exit of the hole respectively.
The thickness of the heat affected zone (Fig. 1b) was measured
using digital image analysis by employing a 10 Mega pixel camera,

24

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

Table 4
Experimental results of laser drilling of alumina ceramic.
Measured responses

Mean (m) and standard deviation (s)

Expt. no.

Taper (deg)

Entrance circularity

Exit circularity

HAZ (mm)

Taper (deg)

Entrance circularity

Exit circularity

HAZ (mm)

2.577
2.005
2.577
1.146
1.962
0.043
1.003
0.286
0.688
2.005
0.329
0.730
0.630
2.577
3.062
2.362
0.029
0.645
0.229
1.575
1.260
2.305
0.473
0.201
1.804
2.133
2.348
0.401
0.773
0.917
1.933
0.272
1.275
0.172
0.329
0.100

0.933
0.931
0.965
0.933
0.905
0.865
0.402
0.973
0.9
0.678
0.982
0.955
0.967
0.788
0.83
0.813
0.985
0.991
0.970
0.922
0.941
0.885
0.983
1.000
0.891
0.798
0.856
0.978
0.966
0.955
0.665
0.983
0.913
0.993
0.971
0.998

0.889
0.872
0.855
0.907
0.44
0.852
0.977
0.955
0.934
0.939
0.987
0.836
0.939
0.771
0.840
0.814
0.990
0.947
0.989
0.908
0.936
0.907
0.980
0.979
0.881
0.788
0.860
0.971
0.974
0.964
0.658
0.973
0.902
0.984
0.971
0.984

0.789
0.679
0.774
0.602
0.727
0.899
0.526
0.755
0.779
0.903
0.736
0.688
0.750
0.688
0.492
0.468
0.607
0.373
0.755
0.464
0.349
0.650
0.617
0.631
0.722
0.789
0.698
0.617
0.621
0.478
0.664
0.531
0.339
0.507
0.540
0.559

m
2.076

s
0.676

m
0.941

s
0.016

m
0.881

s
0.022

m
0.711

s
0.087

0.823

0.861

0.786

0.26

0.806

0.25

0.727

0.154

0.938

0.733

0.878

0.138

0.924

0.063

0.777

0.092

2.158

1.060

0.850

0.080

0.841

0.071

0.600

0.141

0.605

0.704

0.967

0.031

0.959

0.039

0.550

0.168

1.060

0.944

0.952

0.051

0.950

0.035

0.562

0.142

1.672

0.876

0.881

0.075

0.875

0.075

0.706

0.071

0.974

0.697

0.892

0.152

0.892

0.156

0.574

0.085

0.333

0.666

0.969

0.039

0.96

0.039

0.486

0.1

Fig. 1. (a) Dmin and Dmax of laser drilled hole at the entrance and (b) scheme of HAZ thickness measurement.

thickness (t) of the specimen taper angle was calculated using

Eq. (4).

using Eq. (3).

HAZ thickness

HAZ circle diameterEntry hole diameter

2

HAZ thickness was considered by averaging the thickness at

eight segments divided at 451 angle apart. For measuring the taper
of laser drilled hole, the entry diameter (Dent) and the exit
diameter (Dext) as shown in Fig. 2 were measured using Toolmakers Microscope of least count 0.001 mm. Considering the

Taper angle tan 1


Dent Dext
2t

The mean values of the experimental responses of the four

replicates was fed to MINITAB software as inputs to build mathematical models and analyze the parameters to obtain nominal

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

circularity at the entrance and the exit, minimum HAZ and

minimum taper.

25

2.5. Response Surface Methodology (RSM) models

The relationship between the laser drilling parameters and the
responses was modeled using RSM. The general rst order RSM
model used to predict the inuence of laser parameters on the
response factor is given by Eq. (5).

2.3. ANOVA of experimental responses

ANOVA was performed on the experimental responses as
shown in Tables 5 and 6, considering pooled sum of errors at 95
% condence level to assess the signicance of the experimental
parameters. The assessment was made using F and p distributions.

2.4. S/N ratio for the experimental responses

S/N ratios were computed in order to measure the quality
characteristic deviating from the desired value. Smaller the better
characteristic was considered for taper angle and HAZ. Nominal
the best was considered for the entrance and the exit circularities.
Normal probability plots of S/N ratios for each of the responses
shown in Fig. 3ad indicate that the drilling is carried out at stable
conditions and hence was not inuenced by any extraneous
factors.

Y i 0 1 X i1 2 X i2 q X iq i i 1; 2N

where yi is the response factor and xij are the values of ith
observation and jth level of the drilling parameters. The terms i
are the regression coefcients. For the modeling the higher order
linear effects are considered and the interactive effects are not
considered. The residual is a measure of the experimental error.
The Response Surface representing the Entrance circularity
(Cent) as a function of laser drilling parameters such as laser
frequency (Fr), laser power (Lp), scanning speed (Scp) and hole
diameter (D) can be represented by Eq. (6)
C ent 0 1 Fr 2 Lp 3 Scp 4 D

Based on the experimental results of laser drilling, the mathematical relationship established for correlating entrance circularity (Cent) and the laser parameters is presented as Eq. (7)
C ent 0:9200:00409Fr
0:00059Lp 0:00174Scp0:0627D

R value for model was 91.69%.

Similarly, RSM models developed for exit circularity, HAZ and
Taper are presented as Eqs. (8)(10)
C ext 0:7910:00059Fr
0:00081Lp 0:0058Scp0:04763D

R value for model was 95.18%.

HAZ 0:7340:02254Fr0:00075Lp 0:0248Scp0:0776D
9
R2 value for model was 94.05%
T 4:3050:05724Fr0:01376Lp 0:1491Scp0:3519D

10

R value for the model was 97.42%

Fig. 2. Taper in laser drilled hole.

Table 5
ANOVA for entrance circularity and exit circularity.
Factor

Fr
LP
Scp
D
Error
Total
(Error)

Df

2
2
2
2
0
8
(4)

Entrance circularity

Exit circularity

SSq

MSq

Ftab

SSq

MSq

Ftab

0.0007
0.0058
0.0003
0.0059
0
0.0127
(0.0010)

0.0003
0.0029
0.0001
0.0030

1.38
12.06
0.61
12.29

5.26
45.76
2.32
46.64

3
3
3
3

0.0002
0.0100
0.0003
0.0034
0
0.0139
(0.0005)

0.0001
0.0050
0.0002
0.0017

0.67
39.23
1.32
13.53

1.23
71.63
2.41
24.71

3
3
3
3

(0.0002)

(0.0001)

Table 6
ANOVA for HAZ and taper.
Factor

Fr
LP
Scp
D
Error
Total
(Error)

Df

2
2
2
2
0
8
(4)

HAZ

Taper

SSq

MSq

Ftab

SSq

MSq

Ftab

0.0201
0.0088
0.0041
0.0098
0
0.0427
(0.0129)

0.0100
0.0044
0.0021
0.0049

3.10
1.36
0.63
1.51

46.91
20.56
9.62
22.89

3
3
3
3

0.1611
2.7839
0.1603
0.1858
0
3.2912
(0.3215)

0.0806
1.3920
0.0802
0.0929
0

1.002
17.32
0.99
1.15

4.89
84.58
4.87
5.64

3
3
3
3

(0.0032)

(0.0804)

26

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

Fig. 3. S/N ratio for (a) entrance circularity, (b) exit circularity, (c) HAZ and (d) taper angle.

Table 7
Experimental (Ex) verses predicted (Pr) responses.
Expt. no.

1
2
3
4
5
6
7
8
9

Entrance circularity

Exit circularity

HAZ

Taper

Ex

Pr

Error (%)

Ex

Pr

Error (%)

Ex

Pr

Error (%)

Ex

Pr

Error (%)

0.94
0.78
0.87
0.85
0.96
0.95
0.88
0.89
0.96

0.93
0.94
0.91
0.85
0.96
0.94
0.87
0.89
0.96

1.06
20.22
4.55
0.70
0.31
1.05
0.56
0.36
0.72

0.88
0.80
0.92
0.84
0.95
0.95
0.87
0.89
0.96

0.88
0.92
0.91
0.84
0.95
0.93
0.87
0.90
0.96

0.11
15.01
0.75
0.11
0.10
0.10
0.45
0.86
0.28

0.71
0.72
0.77
0.60
0.55
0.56
0.70
0.57
0.48

0.70
0.70
0.76
0.75
0.63
0.61
0.68
0.60
0.54

0.88
2.57
1.62
25.13
16.0
10.26
3.62
6.20
13.02

2.07
0.82
0.93
2.15
0.60
1.05
1.67
0.97
0.33

2.07
1.00
0.89
2.13
0.53
0.87
1.66
1.04
0.40

0.09
21.87
4.80
0.97
21.32
17.75
0.18
6.98
21.38

The adequacy of the model was further analyzed by using R2

values. These values represent the condence level of regression.
The R2 values obtained for the RSM models for all the responses
indicated that the experimental and the predicted values were in
good agreement.
The experimental and predicted responses along with error in
the prediction are shown in Table 7. The errors indicated that the
RSM models can predict the responses with very good accuracy, in
all the cases.

characteristics can be expressed by Eqs. (11) and (12)

X i k

X i k 1

yi kOV
maxfmaxyi kOV; OVminyi kg

11

12

where max yi (k) and min yi (k) are the maximum and minimum
values of measured mean responses respectively. OV is the target
optimum value. The gray relation coefcient is given by

2.6. Gray Relational Analysis (GRA) of experimental responses

i k
GRA was adopted to measure the degree of relationship
amongst multiple performance characteristics on the basis of gray
relational grades. These included taper, entrance and exit circularities and HAZ. Linear normalization of the experimental results
was performed in the range between zero and unity. The normalized results for smaller the better and Nominal the better

maxyi kyi k
maxyi kminyi k

min max
0i k max

13

where 0i k jjx0 kxi kjj is the deviation sequence of the

reference sequence x0 k and compatibility sequence xi k,
min minjjx0 kxi kjj and max maxjjx0 kxi kjj. is the identication coefcient 0; 1, 0.5 is generally used. The average
of the gray relation coefcients is used as gray relation grade given

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

by Eq. (14)
1 n
i
k
nk1 i

14

The gray relation coefcients, gray relation grade and order for
the experimental responses are shown in Table 8.

3. Results and discussion

3.1. Analysis of parametric inuences on entrance circularity
ANOVA presented in Table 5 revealed that the entrance
circularity is inuenced by laser power and hole diameter. Fig. 4
Table 8
Gray relation coefcients, gray relation grades and their order.
Expt. no.

1
2
3
4
5
6
7
8
9

Gray relation coefcients

Grade Order

Entrance circularity

Exit circularity

HAZ

Taper

0.764
0.680
0.739
0.434
0.886
0.845
0.509
0.543
1.000

0.566
0.692
0.740
0.465
0.982
0.911
0.550
0.603
1.000

0.412
0.414
0.333
0.398
0.639
0.585
0.419
0.661
1.000

0.344
0.651
0.601
0.333
0.770
0.557
0.405
0.587
1.000

0.521
0.609
0.603
0.408
0.820
0.724
0.471
0.599
1.000

6.000
7.000
4.000
9.000
2.000
3.000
8.000
5.000
1.000

Fig. 4. Response surface plot of entrance circularity with laser power and hole
diameter.

27

shows the effect of laser power and hole diameter on entrance

circularity by keeping scanning speed and frequency constant at
3.5 mm/s and 5 kHz respectively. The entrance circularity
increased with increase in laser power. This was mainly because
high energy laser beam does not allow the molten material to
accumulate and it is transported further from the hole. There is
lesser chance of spatter being accumulated and hence entrance
circularity increases. The entrance circularity decreased with
increase in hole diameter. This may be due to the greater interaction time of the liquid melt on the edges of the entry hole which
results in greater erosion disturbing the periphery of the hole. The
same arguments are provided by Ghorieshi et al. [20] and Ng and
Li [15]. Ghorieshi et al. [20] in which it is reported that lower pulse
frequency produced greater circularity since at lower pulse frequency, the pulse-off time becomes longer and the material has
more time to be cooled down and becomes closer to solid state.
This can prevent any agitation and disorder during the material
removal process and results in greater circularity. Ng and Li [15]
reported that entrance circularity becomes sensitive at higher
laser power and it gradually decreases with increase in hole
diameter.
Biswas et al. [23] reported that initially circularity increases
with increase in pulse frequency thereby decreasing beam energy.
Initially when pulse frequency increases, pulse-off (time between
two successive incidents of laser beam) time becomes shorter and
the beam energy generated becomes lower; as a result, the
material gets melted and solidied with lower agitation and
disorder, and higher circularity is formed. After a certain limit,
when pulse frequency increases, the pulse-off time becomes very
short and the molten material does not get time to be solidied,
due to which agitation and disorder take place in the molten
material during material removal process resulting in lower
circularity.
Fig. 5 shows the optimization results for the nominal entrance
circularity based on the mathematical model developed. The
desirability function approach is used for the optimization of
multiple responses. It nds the experimental factors which provide the most desirable response values. For each response, a
desirability function assigns numbers between 0 and 1 to the
possible values of the response. The response value equal to zero
represents a completely undesirable value and the response value
equal to one represents a completely desirable or ideal response
value. The individual desirabilities are combined using the geometric mean, which gives the overall or composite desirability.
Using MINITAB for optimization of laser drilling of alumina
ceramics, nominal entrance circularity was obtained as 0.9858
when frequency, laser power, scanning speed and hole diameter
are set at 2.5 kHz, 29.39%, 2.5 mm/s, 1 mm respectively.

Fig. 5. Optimization results for nominal entrance circularity.

28

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

3.2. Analysis of parametric inuences on exit circularity

diameter are set at optimal parametric setting i.e. 7.5 kHz, 240 W,
4.5 mm/s, 1 mm respectively.

Most of the researchers, in their laser drilling experiments,

have assumed exit holes to be more circular than the entrance
holes and addressed only entrance circularity. But, in actual
practice, due to the multiple reective characteristic of the laser
beam, the beam propagates back after drilling the hole in the
specimen, thereby increasing the thermal energy which disturbs
the circularity at the exit. Hence, exit circularity also becomes an
important quality characteristic to be addressed. ANOVA of Table 5
showed that exit circularity was inuenced by laser power and
hole diameter. Fig. 6 shows the effect of laser power and hole
diameter on exit circularity by keeping scanning speed and
frequency constant at 3.5 mm/s and 5 Hz respectively. The surface
plot shows that exit circularity increases with increase in laser
power. This is because when laser power increases, thermal
energy increases due to which material removal increases in the
entire cross section of the job. This is in agreement with the results
of Biswas et al. [23]. Also exit circularity increases with increase in
hole diameter. This is because when hole diameter increases,
greater laser energy is required to melt the material. The laser
beam energy decreases by the time it reaches the exit side. This
causes lower levels of agitation and disorder thereby increasing
the exit circularity.
Fig. 7 shows the optimization results for the nominal exit
circularity based on RSM. The model predicted exit circularity of
0.9689 when frequency, laser power, scanning speed and hole

ANOVA presented in Table 6 showed that HAZ is signicantly

inuenced by pulse frequency. Fig. 8 shows the effect of frequency
and hole diameter on HAZ by keeping scanning speed and laser
power constant at 3.5 mm/s and 150 W respectively. The surface
plot reects that HAZ thickness decreases with increase in frequency. This is mainly because as the pulse frequency increases,
pulse-off time becomes shorter due to which beam energy
becomes lower and results in lower HAZ. At low pulse frequency,
peak power of the laser beam is greater which results in excessive
removal of material causing greater HAZ and hence increase in
hole diameter. This is in agreement with Kuar et al. [17].
Bandhopadhyay et al. [21] reported that HAZ thickness nearly
remained constant with pulse frequency while drilling 8 mm
thick IN 718 and Ti6Al4V materials. It slightly increased when
greater pulse energy was employed. Fig. 9 shows the optimization
results for minimum HAZ thickness based on RSM. Using MINITAB
for optimization of laser drilling on alumina ceramics, minimum
HAZ thickness was obtained as 0.5233 when frequency, laser
power, scanning speed and hole diameter were set at optimal
parametric setting i.e. 7.5 kHz, 240 W, 2.5 mm/s and 1 mm
respectively.

Fig. 6. Response surface plot of exit circularity with laser power and hole diameter.

Fig. 8. Response surface plot of HAZ with frequency and hole diameter.

3.3. Analysis of parametric inuences on HAZ

Fig. 7. Optimization results for nominal exit circularity.

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

29

Fig. 9. Optimization results for minimum HAZ thickness.

Fig. 10. Measured and theoretical HAZ thickness.

3.3.1. Analytical predictions of HAZ

The laser was assumed as a surface heat ux model with
Gaussian distribution of intensity given by Eq. (15) as per Ref. [24].
  2 
2Pa
r0
Ir
15
exp
2
R
R2 tpf
where I(r), , , Pa, tp, f, R, r0 are laser intensity, absorption
coefcient, heat ux efciency, mean power of the laser, pulse
duration, pulse frequency, laser spot radius and radial distance
from the center of the beam respectively.
The radial distance from the center of the beam can be
expressed by Eq. (16)
v
#
u "
R u
2Pa
r 0 p tln
16
2
R2 tpf Ir
The HAZ radius was considered as the limiting value of r0 above
which the laser intensity I(r) is less than the laser intensity (Ircy) at
recrystallization temperature (1550 1C) of alumina ceramic [25]
calculated by Eq. (17)
Ircy H   t

17

Where is the density, t is the thickness of the specimen; H is the

heat liberation rate (kJ/s) during recrystallization process.when
Ir I rcy , r 0 r HAZ .
The absorption coefcient of the alumina material was considered as 6.17  107 [26]. The heat ux efciency was found to be
0.9. The pulse duration and pulse frequency were 0.050.17 ms
and 2.5 kHz7.5 kHz respectively. Theoretical HAZ thickness was
obtained for nine experiments by substituting the values of rHAZ in
Eq. (3). The measured HAZ and theoretical HAZ are presented in
Fig. 10.
As predicted by the theoretical model, HAZ thickness decreased
with increase in frequency because of the lower laser energy
density at recrystallization stage and shorter pulse time leading to

Fig. 11. Response surface plot of taper with laser power and hole diameter.

lower beam energy. Also at lower frequency, absorptivity of the

alumina ceramic is high due to which HAZ thickness increases. The
predicted HAZ is in full agreement with experimental HAZ.
3.4. Analysis of parametric inuences on taper
Minimum value of taper is desired for good quality of laser
drilled holes. ANOVA (Table 6) showed that laser power signicantly inuenced taper. Laser power and taper are negatively
correlated as shown by the surface plot (Fig. 11). As the laser
power increases the diameters at the entrance and at the exit
increase which results in reduced taper. Thermal conductivity of
alumina is low and hence the material removal rate is low at any
given power. The increase in power increases the material removal
rate. Also the taper increased with increase in hole diameter
because of the accumulation of heat energy at higher peak power
and absorptivity leading to greater erosion at the entrance of the
hole. These results are in agreement with that of Ghorieshi et al.
[20] and Kacar et al. [11]. Ghorieshi et al. [20] reported that the
taper decreased with increase in peak power, when the laser beam
impinges the material at higher laser power and hence larger
entrance diameter is produced. At the same time, the beam
produces larger exit diameter because greater energy per pulse
has more effect on the hole exit than that at the entrance. This
results in lower values of taper. Kacar et al. [11] reported that the
taper angle decreased with increase in peak powers. This was
because the diameter at the exit of the hole was greater than that
at the entrance at higher peak powers. Further explanation for
variation of taper with respect to laser power is reported by Tunna
et al. [22] in which it is reported that the melt erosion caused by

30

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

Fig. 12. Optimization results for minimum taper.

Fig. 13. The theoretical and experimental taper.

high vapor pressure above the surface of the material during the
laser irradiation of the surface pushes the melt up and out of the
hole, which affect the geometry of the hole. Yilbas [18] reported
that taper decreased with increase in pulse energy for nickel
whereas it increased for stainless steel.
Fig. 12 shows optimization results for minimum taper based on
RSM developed using MINITAB. Minimum value of taper was
obtained as 0.2551 when frequency, laser power, scanning speed
and hole diameter were set at optimal setting i.e. 7.5 kHz, 240 W,
4.5 mm/s, 1 mm respectively.
Fig. 14. Multi-objective optimization results for experimental responses.

3.4.1. Analytical predictions of taper

The theoretical maximum hole depth is given by Eq. (18) as per
Ref. [13].
HDmax

Ptpf
r2

18

where P is laser peak power, f is the pulse frequency, r is the beam

radius and tp is pulse on time.
Pulse on time tp is given by Eq. (19)
tp

D
f

19

where D is the pulse duty cycle. Since through holes were achieved
in all the experiments, theoretical hole depth was assumed to be
equal to the thickness of the material. combining Eqs. (18) and
(19), taper angle T is expressed as Eq. (20),


DentDextr 2
T tan 1
20
2Ptpf
Pa is the average laser power given by Eq. (21)
Pa P  D

21

Hence, taper can be expressed in terms of beam radius and

average laser power as per Eq. (22)


DentDextr 2
T tan 1
22
2Pa
Considering the beam radius0.28 mm and peak power
intensity 750 W/m2, the taper as per Eq. (22) and experimental
taper are presented in Fig. 13. The taper reached peak levels at
lower duty cycle due to shorter pulse time causing erosion at both
entry and exit sides of the hole. The predicted and experimental
taper are in perfect agreement for the entire range of duty cycle
considered in the experiments.

4. Multi-objective optimization of laser drilling parameters

Multiresponse optimization was carried out and the results for
minimum taper, nominal entrance and exit circularity and minimum HAZ thickness are presented in Fig. 14. Four responses were
optimized all together in one setting. Each row of the graph
corresponds to a response variable and each column corresponds
to one of the parameters considered during the laser drilling. Each

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

cell of the graph represents the variation of responses with one of

the parameters, keeping other parameters constant. The vertical
line inside the graph represents current parameter settings and a
horizontal dotted line represents the current response values. The
numbers displayed at the top of a column show the current
parameter level settings, high value of that parameter, and low
value of that parameter setting in the experimental design. At the
left side of each row, goal for the response (minimization of taper,
nominal hole circularity at entry and exit, minimum HAZ), predicted response (y) at current parameter settings and individual
desirability scores are given. The goal set, upper limit (UL), lower
limit (LL), weights (Wt), importance of the factors and the
Table 9
Goals set and limits used for the optimization and desirability.
Response

Goal

LL

UL

Wt Importance Solution Desirability

Entrance
circularity
Exit
circularity
HAZ
thickness
Taper

Target

0.850 0.969 1

0.9630

0.8068

Target

0.841 0.960 1

0.9651

0.8330

Minimize 0.547 0.777 1

0.5570

0.9565

Minimize 0.333 2.158 1

0.3515

0.9898

31

solutions obtained for multiple response optimization based on

desirability is presented in Table 9.

5. SEM studies
The laser drilled holes were characterized to determine the
geometrical and metallurgical features and ascertain the inuence
of process parameters. The surface morphology of the laser drilled
holes was examined using scanning electron microscope (Model
Hitachi SU-1500 SEM 15.0 kV). The laser drilled specimens were
sectioned and sputtered with gold to investigate the extent of
taper. The top surfaces of the specimens were gold sputtered to
study the hole circularity, HAZ and microcracks. Fig. 15(a) shows
lowest taper corresponding to highest gray relational grade of 1 i.e.
drilling at 7.5 kHz frequency, 240 W laser power, 3.5 mm/s scanning speed and 1 mm hole diameter. It is observed that the hole
has approached near-cylindricity due to the lower interaction
between the melt and hole walls at the entry side resulting in
lower levels of erosion. This is in agreement with Bandhopadhyay
et al. [21]. Fig. 15(b) shows the SEM image of the taper when the
hole was drilled at 5 kHz frequency, 150 W laser power, 3.5 mm/s
scanning speed and 2 mm hole diameter, corresponding to lowest

Fig. 15. SEM images of laser drilled holes with (a) minimum taper, (b) maximum taper, (c) minimum HAZ, (d) entrance hole and (e) exit hole drilled at 7.5 KHz, 240 W,
3.5 mm/s, 1 mm, (f) entrance hole and (g) exit hole drilled at 5 kHz, 230, 4.5 mm/s, 1 mm, (h) entrance hole and (i) exit hole drilled at 5 kHz, 240 W, 2.5 mm/s, 1.5 mm.

32

A. Bharatish et al. / Optics & Laser Technology 53 (2013) 2232

gray relational grade of 0.408. Highest taper is obtained at lower

laser power because of the faster resolidication at the entry side.
Fig. 15(c) shows HAZ in the SEM images when the hole was
drilled using parameters corresponding to highest gray relational
grade of 1. The extent of HAZ mainly depends on the materials
being drilled and their thicknesses apart from the laser parameters
employed [19]. HAZ was 0.50.7 mm in the 99% alumina ceramics.
A smooth textured zone along with small resolidied particles was
observed at the laser processed region. Fig. 15(d)(i) shows the
entry and exit holes drilled using parameters corresponding to
gray relational order of 1, 2 and 3 in order to investigate the
circularity. It was observed that both entrance and exit circularities
have continuously decreased with decrease in laser power and
increase in formation of microcracks near the edges of the hole.
The quality of the drilled holes is not so high because of the
lower thermal diffusivity of the alumina material which increases
the interaction time of the liquid melt with the material given by
p
Lt 2 kT h
23
where k is the thermal diffusivity, Th is the pulse width and Lt is
the depth of diffusion [15]. This increase in interaction time leads
to higher degree of erosion taking place along the periphery of
the hole.

6. Conclusions
In the present research, CO2 laser percussion drilling of 2 mm
thick alumina ceramic plates was carried out to examine the
effects of laser parameters such as pulse frequency, laser power,
scanning speed and hole diameter on entrance circularity, exit
circularity, HAZ and taper. Based on ANOVA, both entrance and
exit circularities were signicantly inuenced by laser power and
hole diameter. HAZ was mainly inuenced by pulse frequency and
taper was signicantly inuenced by laser power. The entrance
circularity increased with increase in laser power but it decreased
with increase in hole diameter. Exit circularity increased with
increase in both laser power and hole diameter. HAZ thickness
decreased with increase in pulse frequency at constant laser
power. Taper decreased with increase in laser power at constant
pulse frequency. The multi-objective optimization of laser parameters using RSM indicated that high frequency (7.5 KHz), high
laser power (240 W), moderate scanning speed (3.85 mm/s) and
lower hole diameter (1 mm) are preferred to achieve nominal
entrance circularity (0.963) and exit circularity (0.965), minimum
HAZ (0.55 mm) and minimum taper (0.3511) which were in
agreement with Gray Relational Analysis results. Analytically
predicted Taper and HAZ agreed with the experimental results.
SEM studies indicated a smooth textured zone along with small
resolidied particles at the laser processed region.
References
[1] Biswas R, Kuar AS, Sarkar S, Mitra S. A parametric study of pulsed Nd:YAG laser
micro-drilling of gamma titanium aluminide. Optics and Laser Technology
2010;42:2331.

[2] Moorhouse Colin J, Villarreal Francisco, Wendland Jozef J, Baker Howard J, Hall
Denis R, Hand Duncan P. CO2 laser processing of alumina (Al2O3) printed
circuit board substrates. IEEE Transactions on Electronics Packaging Manufacturing 2005;28(3):24957.
[3] Kononenko TV, Konov VI, Garnov SV, Klimentov SM, Dausinger F. Dynamics of
deep short pulse laser drilling: ablative stages and light propagation. Laser
Physics 2001;11(3):34351.
[4] Han Wei. Computational and experimental investigations of laser drilling and
welding for microelectronic packaging. PhD thesis. Worcester: Worcester
Polytechnic Institute; 2004.
[5] Zang Z, Mukai K, Navaretti P, Duelk M, Velez C, Hamamoto K. Thermal
resistance reduction in high power superluminescent diodes by using active
multi-mode interferometer. Applied Physics Letters 2012;100(3):031109.
[6] Zang Z, Minato T, Navaretti P, Hinokuma Y, Duelk M, Velez C, et al. High
power (>110 mW) superluminescent diodes using active multi-mode interferometer. IEEE Photonics Technology Letters 2010;22(10):7213.
[7] Zang Z, Mukai K, Navaretti P, Duelk M, Velez C, Hamamoto K. High power and
stable high coupling efciency (66%) Superluminescent light emitting diodes
by using active multi-mode interferometer. IEICE Transactions on Electronics
2011;E94C(5):8624.
[8] Samant Anoop N, Dahotre Narendra B. Laser machining of structural ceramics
a review. Journal of the European Ceramic Society 2009;29(6):96993.
[9] Dubey Avanish Kumar, Yadav Vinod. Laser beam machininga review.
International Journal of Machine Tools and Manufacture 2008;48(6):60928.
[10] Yilbas BS. Parametric study for laser hole drilling of Inconel 617 alloy. Lasers in
Engineering 2002;12(1):116.
[11] Kacar E, Mutlu M, Akman E, Demir A, Candan L, Canel T, et al. Characterization
of the drilling alumina ceramic using Nd:YAG pulsed laser. Journal of Materials
Processing Technology 2008;209(4):200814.
[12] Hanon MM, Akman E, GencOztoprak B, Gunes M, Taha ZA, Hajim KI, et al.
Experimental and theoretical investigation of the drilling of alumina ceramic
using Nd:YAG pulsed laser. Optics and Laser Technology 2011;44(4):91322.
[13] Yan Yinzhou, Ji Lingfei, Bao Yong, Jiang Yijian. An experimental and numerical
study on laser percussion drilling of thick-section alumina. Journal of Materials Processing Technology 2012;212(6):125770.
[14] Ng GKL, Li L. Repeatability characteristics of laser percussion drilling of
stainless steel sheets. Optics and Lasers in Engineering 2003;39(1):2533.
[15] Ng GKL, Li L. The effect of laser peak power and pulse width on the hole
geometry repeatability in laser percussion drilling. Optics and Laser Technology 2001;33(6):393402.
[16] Mutlu M, Kacar E, Akman E, Akkan CK, Demir P, Demir A. Effects of the laser
wavelength on drilling process of ceramic using Nd-YAG Laser. Journal of Laser
Micro/Nanoengineering 2009;4(2):848.
[17] Kuar AS, Doloi B, Bhattacharyya B. Modeling and analysis of pulsed Nd:YAG
laser machining characteristics during micro-drilling of zirconia (ZrO2). International Journal of Machine Tools and Manufacture 2005;46(12):130110.
[18] Yilbas BS. Parametric study to improve laser hole drilling process. Journal of
Materials Processing Technology 1996;70(1):26473.
[19] Li L, Low DKY, Ghoreishi M. Hole taper characterization and control in laser
percussion drilling. CIRP Annals Manufacturing Technology 2002;51(1):1536.
[20] Ghoreishi M, Low DKY, Li L. Comparative statistical analysis of hole taper and
circularity in laser percussion drilling. International Journal of Machine Tools
and Manufacture 2002;42(9):98595.
[21] Bandyopadhyay S, Sarin Sundar JK, Sundararajan G, Joshi SV. Geometrical
features and metallurgical characteristics of Nd:YAG laser drilled holes in thick
IN718 and Ti6Al4V sheets. Journal of Materials Processing Technology
2002;127(1):8395.
[22] Tunna L, Neill WO, Khan A, Sutcliffe C. Analysis of laser micro drilled holes
through aluminium for micro manufacturing applications. Optics and Lasers in
Engineering 2005;43(9):93750.
[23] Biswas R, Kuar AS, Biswas SK, Mitra S. Characterization of hole circularity in
pulsed Nd:YAG laser micro-drilling of TiNAl2O3 composites. International
Journal of Advanced Manufacturing Technology 2010;51(9):98394.
[24] Dong-Gyu AHN, JUNG Gwang-Won. Inuence of process parameters on
drilling characteristics of Al 1050 sheet with thickness of 0.2 mm using
pulsed Nd:YAG laser. Transactions of Nonferrous Metals Society of China
2009;19(1):s15763.
[25] Damani RJ, Makroczy P. Heat treatment induced phase and microstructural
development in bulk plasma sprayed alumina. Journal of the European
Ceramic Society 2000;20(7):86788.
[26] Yilbas BS, karatas C, Arif AFM, Abdul Aleem BJ. Laser control melting of
alumina surfaces and thermal stress analysis. Optics and Laser Technology
2011;43(4):85865.