Journal of Materials Processing Technology 225 (2015) 151–161

Contents lists available at ScienceDirect

Journal of Materials Processing Technology

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

A photoelasticity approach for characterization of defects in

microwave drilling of soda lime glass
Nitin Kumar Lautre a,∗ , Apurbba Kumar Sharma a , Pradeep Kumar a , Shantanu Das b
a
Department of Mechanical and Industrial Engineering, Indian Institute of Technology Roorkee, Roorkee, Uttarakhand 247 667, India
b
Reactor Control Division, Bhabha Atomic Research Center, Mumbai 400 085, India

a r t i c l e i n f o a b s t r a c t

Article history: Need for drilling micro holes has been on rise in many miniature applications including biotechnology.
Received 17 October 2014 Micro machining of such features is difficult to realize, particularly on difficult-to-machine hard and
Received in revised form 20 May 2015 brittle materials like soda lime glass. Defect characterization with precision in micro-nano scale is even
Accepted 26 May 2015
more challenging. Conventional approaches appear inadequate in such cases while fabrication as well
Available online 28 May 2015
as in characterization. In the present work, holes of 900 ␮m diameter were drilled on soda lime glass
using a novel thermal-based approach called ‘microwave drilling’. It uses the phenomenon of thermal
Keywords:
ablation with plasma heat created by the applied microwave energy through a tool. The energy was
Glass drilling
Microwave
applied in the range of 90–900 W at 2.45 GHz in a multi-mode applicator. The glass specimens were
Concentrator subjected to high localized heat, which also caused some defects like cracking and deformation due to
HAZ melting in the drilling zone. A photoelasticity approach was employed to characterize these defects.
Cracks A setup was developed using a polarizer and a CCD camera to obtain the birefringence patterns. The
Photoelasticity patterns were analyzed to assess the defects. The microwave drilling process was also simulated and
the stress-temperature relationships were studied. Simulation results substantiated the experimental
observations. Details are discussed with evidences.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction
Microwave drilling is a relatively new approach to drill micro
holes in different materials. An approach to microwave drilling
was proposed and studied by Jerby et al. (2000, 2002, 2004, 2005,
2009) to get established. In this process, Al-Shamma et al. (2001),
Shayeganrad and Mashhadi (2009), Jerby et al. (2009) used an
electromagnetic wave at 2.45 GHz to create a hot plasma ball at
the tip of a tool. Meir and Jerby (2012a) recently highlighted the
role of drill tool to interact with microwave and generate plasma,
also at the same time maintain the intensity of plasma as per the
power input. Microwaves are continuously attracted toward hot
Abbreviations: ASTM, American society for testing and materials; B, brewsters;
CCD, charge coupled device; FEM, finite element analysis; GaP, gallium phosphate;
GHz, giga Hertz; HAZ, heat affected zone; HRA, rockwell hardness on scale-A; LED,
light emitting diode; LMH, localized microwave heating; Na2 O, sodium oxide; R,
retardation; RGB, red–green–blue; SBR, stress birefringence; SCA, spectral content
analysis; SiO2 , silicon oxide; VGA, video graphic array.
∗ Corresponding author at: IIT Roorkee IIT Roorkee Roorkee 247667, India.
Tel.: +918909737466.
E-mail addresses: nfl 123@rediffmail.com (N.K. Lautre), akshafme@gmail.com
(A.K. Sharma), kumarfme@gmail.com (P. Kumar), shantanu@barc.gov.in (S. Das).
plasma due to the availability of free electrons around the plasma.
Jacob et al. (1995) emphasized on the need of microwave mate-
rial interaction and promotes the microwave specific effect over
the normal dielectric material heating. Whereas, Meir and Jerby
(2012b) showed the applications of localized microwave heating
(LMH) of dielectric material in conjunction with the importance of
concentrated plasma at the tool tip to thermally ablate the target
material, to create a cavity, identified as drilled hole. Conventional
drilling, on the other hand, by its virtue has some limitations in
drilling micro holes. Various difficulties experienced in conven-
tional drilling include more mechanical movement (rotary and
linear), challenge in balancing torque and thrust, use of complex
bits, tool workpiece friction, delamination of workpiece, burr for-
mation, development of micro cracks, strength requirement of
the drill bit at that scale, rigid clamping requirement of tool and
workpiece, complex tool preparation, etc. On the other hand, laser
drilling especially, the ultra short pulsed laser drilling, apart from
water jet drilling, is by far the most popular process for drilling
of micro holes in glass specimen. Recently, an attempt was made
to drill micro holes of diameter 0.4 mm to 1.2 mm through laser
drilling by Wang et al. (2014). Laser drilling is sophisticated, in
general, induces high energy thermal stresses, causes hole taper,
has limited aspect ratio and costly to operate at the micro regime.

http://dx.doi.org/10.1016/j.jmatprotec.2015.05.026
0924-0136/© 2015 Elsevier B.V. All rights reserved.
152 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161
Nomenclature
Symbols
D Distance between two fringes
F Material fringe value
G Distance between grid array
K Stress optic constant
n Order of the fringe
N Fringe order
S Distance between work specimen and image grab-
ber
t Thickness
tan ı Loss tangent
L × B × H Length × breadth × height
r Dielectric constant
 Refractive index
 Stress
 Wavelength
In both the processes, formation of heat affected zone (HAZ) is
however predominant. In conventional drilling, the HAZ was due
to friction between the tool and the workpiece; whereas, in laser
drilling it was due to the generated heat. A better control of HAZ in
laser leads to better quality of micro holes; whereas in conventional
drilling, it leads to cracking in the work material like glass. Unlike
conventional drilling, laser is also a contactless dry drilling, which
induces less mechanical stress as compared to thermal stresses.
Micro holes in glass are required in many applications. These
include light biotechnology, weight optics, about 0.8 mm holes in
solar panels, wafer chucks, bus bars, etc. Glasses are found better
than the metallic materials in terms of their low thermal expansion,
less chemical reactivity, high transparency and high rigidity. How-
ever, fabrication of precise and high quality micro holes in glasses
is a big processing challenge. The microwave drilling is a thermal-
based drilling approach being investigated in the recent times by
few researchers like Jerby and Dikhtiar (2000), Jerby et al. (2004),
George et al. (2012), and Lautre et al. (2014a,b, 2015). The abla-
tion of work material in microwave drilling does take place mainly
because of two stage interactions. These interactions are of the tool
material with the microwave radiation, which produces plasma
and consequently the interaction of the plasma with the workpiece
material. The ability to control these interactions primarily decides
quality of the drilled hole. At the same time, Alhekail (2001) empha-
sized on ensuring that the interactions are performed in an enclosed
atmosphere (applicator). While considering microwave material
interaction, there are basically three different types of materials
– transparent, absorptive and reflective. Upon irradiation, differ-
ent types of interaction take place with characteristically different
type of work material. Fini and Breccia (1999) showed the impor-
tance of microwave–material interaction and a few consequences.
In microwave drilling, the interactions inside the applicator play
a crucial role in performance of the process and quality of the
drilled holes. In most of the cases, the work piece experiences a
high thermal stress during the interaction with the plasma and gets
fractured. The stress analysis due to microwave plasma heat thus
becomes an important aspect to prevent the occurrence of cracks
and hence material failure.
The method of photoelasticity is being effectively used for study
of fracture. The method is used to analyze the surface measurement
after the occurrence of the fracture. A photoelastic effect is observed
when the stressed glass specimen is kept under light emitting diode
(LED) light source (considered  = 0.565 ␮m for calculation) and
screen through a linear polarized lens. The stressed region shows a
distorted rainbow like effect depending upon the residual stresses
induced in it. Khanna et al. (2004) used an open mode stress field
equations to model the isochromatic fringe patterns according to
orthotropic stress-optic law. A different method of separating prin-
cipal and normal stresses is modeled through single boundary pixel.
Ashokan and Ramesh (2009) showed the relationship of isochro-
matic line in terms of the principal stress difference while the
isoclinic line in terms of principal stress direction. A 3D photoe-
lastic method was tried by Aben et al. (2000) for two different
cases of weak birefringence and constant principle axes, to find
the complete stresses. The authors have identified the problems
of implementation and application of the photelasticity method
in determining the whole field stress. A phase shifting method
was used by Ramji and Ramesh (2008), to find the isoclinic lines
through plane polariscope (phase shift between electromagnetic
component in 2D is zero) and isochromatic lines through circular
polariscope (phase shift between electromagnetic component in 2D
is ␲/2). In this work, the possibility of use of monochromatic and
white light was highlighted along with the three methods in white
light photoelasticity. The method explained were the spectral con-
tent analysis (SCA), the RGB (red–green–blue) photoelasticity and
the phase shifting methods.
The use of white light photoelasticity is supported for mini-
mum interface between the isoclinic and isochromatic lines. The
white light is generated for a continuous and discrete spectrum
by incandescent and fluorescent lamps, respectively. The discrete
fringe order (12 fringes) as discovered by Ajovalasit et al. (2007),
as more than the continuous spectrum (6–7 fringes) during all trial
experiments. Later, an attempt was made by Ajovalasit et al. (2012)
on a tempered glass specimen to estimate residual stresses using
RGB photoelasticity approach. A crack tip stress field of thin glass
was measured with charge coupled device (CCD) of pixelated array,
through instantaneous photoelasticity. The stresses were further
determined from the phase map of retardation by Sakaue et al.
(2008). The study has revealed that it was difficult to predict the
direction of the quenching cracks with the approach.
In general, residual stresses are claimed for stationary and equi-
librium to surrounding specimens. Residual stresses are difficult
to predict; during service, they merge with other stresses to cause
detrimental failures. Withers and Bhadeshia (2001) discovered that
the generation of compressive residual stresses is usually due to
rapid cooling. The authors attempted quantitative measurements
of residual stresses on the GaP (Gallium phosphate) crystal using
photoelasticity. The isochromatic lines of zeroth order retardation
were determined for two different wavelengths (0.55 and 1 ␮m).
The results identified residual and principal stress lines along the
pulling direction and were confirmed with piezo-optical approach.
Kotake et al. (1980) also observed the zeroth order of retardation as
very independent of wavelength and its appearance is always at the
same spot. Beghini and Bertini (1998) used a hole drilling method
in lieu of photoelasticity to find residual stresses. In this method,
strain gauges are fixed near the hole to be drilled to measure the
stresses as per ASTM (American society for testing and materi-
als) standard E 837-95. However, photoelasticity based method is
considered more effective to measure residual stresses in the trans-
parent brittle work material like glass, as this method eliminates
the need for additional hole drilling.
In this study, the residual stresses induced in glass during
microwave drilling were analyzed qualitatively using the photoe-
lastic approach. An interaction between a reflective nichrome tool
and transparent glass workpiece was established via hot plasma
generated by microwaves at different input power. The cracking
of glass, while drilling a hole was identified using the fringe pat-
tern obtained through a photoelastic setup. It was attempted to
correlate the results with the distortion on the glass, which are
responsible for the development of cracks and eventual failure of
the work material in drilling process.
 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161 153

Fig. 1. Microwave drilling setup for glass drilling, Inset: zoomed view of the tool-workpiece interaction zone.

2. Microwave drilling setup Table 1

Thermal properties of soda lime glass (source: Pagni, 2002).

In microwave, drilling, micro holes were drilled with a nichrome Property Value
wire (drill bit, 80% nickel and 20% chromium) of 0.7 mm diameter. Coeffcient of thermal expansion 8.7 × 10−6 mm/K
The bit was gripped in a jig, having a dead weight of 15.7 g. Unlike Specific heat capacity at 25 ◦ C 0.921 kJ/kgK
conventional drill bit, the nichrome drill bit has no rotation. In the Thermal conductivity at 25 ◦ C 0.937 W/mK
setup shown in Fig. 1, the only moving part is the tool in the form of Thermal diffusivity at 25 ◦ C 0.436 × 10−6 m2 /s
Coeffcient of thermal stress 0.62 MPa/◦ C
vertical feed as communicated by Lautre et al. (2014a,b). The appli-
Resistence to shock for 6 mm 38 ◦ C(Annealing) and 204 ◦ C
cator (L × B × H: 0.44 × 0.42 × 0.21 m) is of stainless steel to reflect thickness (tempering)
microwave inside the applicator. The magnetron in the applica- Softening, annealing, strain point 715 ◦ C, 548 ◦ C, 511 ◦ C,
tor generates standing microwaves at 2.45 GHz frequency under respectively, (ASTM 338/336)
the variable power of 90–900 W. A pulsed mode power supply was
used in five distinct steps. The duty cycle were adjusted in the order
hardness and density of the glass used are 0.20, 71.6 GPa, 80HRA
of 0.2, 0.25, 0.5, 0.75 and 1 to obtain the power variation. Refractory
and 2500 kg/m3 , respectively. The fringe value at 565 nm wave-
brick was used as a dummy load and a supporting platform for the
length of light was found to be 485.5 kN/m by Sakaue et al. (2008).
glass specimen as illustrated in the figure.
The critically stressed region as developed around the microwave
A 1.2 mm thick soda lime glass (72.6% SiO2 , 14% Na2 O and others)
drilled hole is mostly due to failure by weak tensile stresses. The
plate was used as the workpiece material. The glass plate and the
glass transitional temperature is 560 ◦ C, but to thermally ablate it,
nichrome wire tip (drill bit) was placed inside the applicator in such
a temperature of more than 1040 ◦ C is required. The other ther-
a way that the microwave energy in spatial domain is maximum.
mal properties of the soda lime glass as obtained from its supplier
The spatial position was predetermined through simulation using
as well as from various other sources (Pagni, 2002) are listed in
COMSOL Multiphysics tool (Lautre et al. (2014a,b)). As the tool-
Table 1. Its refractive index is ∼1.5 and the dielectric constant is
workpiece configuration is exposed to microwaves in the manner
7.75 at 20 ◦ C. The time for microwave drilling of glass was kept
as explained above, microwave energy gets concentrated at the tip
within the range of 30–60 s as per data from the pilot experiments.
of the tool owing to low skin depth of the material; the tip gets
The glass specimens were cleaned with acetone and placed
heated up and a localized hot spot is formed. Once, the tempera-
inside the applicator over the refractory bricks at the pre decided
ture exceeds a critical value, a plasma ball is formed at the tool tip
location. The location of the glass inside the applicator was opti-
and a LMH phenomenon begins in the vicinity of the tool tip. The
mized by simulation (COMSOL Multiphysics version 4.4.0.195) for
high temperature in the plasma ball zone causes melting and/or
obtaining the best LMH. The jig and its supporting assembly were
ablation of the work material (in this case soda lime glass) under-
placed vertically over the glass specimen. The tip of the nichrome
neath it. As the tool is gravity fed, the tip advances automatically
tool was kept in point contact with the glass surface.
as the workpiece material gets ablated. As the process progresses,
the work material gets drilled. Further details of the technique have
3. The photoelasticity methodology
been discussed elsewhere (Lautre et al. (2014a,b)). The glass being a
relatively bad conductor of heat, do not conduct away the heat from
The photoelasticity methodology helps to estimate the resid-
the nearby material affected by the LMH; this gives rise to thermal
ual (or frozen-in) stresses induced in glasses during microwave
stresses in the HAZ leading to often cracking of the workpiece.
drilling. Residual stresses in glass may be due to microwave induced
The soda lime glass is microwave transparent with a loss tan-
high temperature gradient or the high difference in compressive
gent (tan ı) of 0.02 and a dielectric constant (r ) of about 7.75 at
and tensile stresses within and around the HAZ. Primarily, residual
20 ◦ C. Beghini and Bertini (1998) found that the glass could with-
stresses are due to the volumetric changes occurred during melt-
stand compressive and tensile stresses of 345 MPa and 6.9 MPa,
ing of limited portion of glass during microwave drilling. It was
respectively. In conventional processes, the fracture strength is
observed from preliminary simulation studies as well as experi-
15% and in laser drilling it is 23% of the tensile stress of the glass.
mental results that the hot plasma melts the nearby region of the
However, most of the glass breaks at 41 MPa during annealing and
glass (workpiece); consequently, the material tends to expand. On
165 MPa during tempering. The Poisson’s ratio, Young’s modulus,
the contrary, the other region of glass, being relatively cold, tends to
154 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161
Fig. 2. Arrangement for observing photoelasticity effect on glass specimen.
resist the expansion of molten glass. In addition, the underlying lay-
ers of glass tend to oppose the tensile stresses induced at the top
surface of the glass. The glass experience stresses in three ways:
(i) tensile stresses due to melting, (ii) compressive stresses from
nearby region and (iii) the compressive stresses in the underlying
region. The induction of tensile stresses and compressive stresses
tend to induce compressive residual stresses in the glass.
Residual stresses are internally stored stresses, which are devel-
oped without the influence of continuously applied external load.
The sudden cracks under thermal loading are basically due to resid-
ual stresses. The photoelasticity (Birefringence or birefractive) is
one of the optical properties well defined for transparent materi-
als like perspex and glass. Under mechanical stress, the optically
isotropic glass changes to stress birefringence (SBR). The mechan-
ically stressed transparent solution gets converted from optically
isotropic to optically anisotropic. Consequently, the glass shows
change in refractive indices due to the change in density around
the stressed specimen. The refractive index () depends on how
fast the light passes through the specimen and its polarization.
In the present work, a linear polarizer was used to stop
the components of light vibrating in different plane. The other
waveforms of the electromagnetic waves like monochromatic,
polychromatic, coherency and collimations (parallel, non diver-
gence) were assumed to be unaltered. The lagging of wave due to
different speed under stressed and unstressed condition is referred
as retardation (R). A stress-optic law was used to correlate the
change of refractive index as a function of stress induced in the
specimen. The Brewster stress-optic law is given as:
1 − 2 = K(1 − 2) = R/t
where, 1 = fast velocity refractive index; 2 = slow velocity refrac-
tive index; 1 = first principal (major) stress (MPa); 2 = second
principal (minor) stress (MPa); K = stress optic constant (Brewsters:
B in m2 /N × 10−12 ); R = retardation or phase shift (nm) of fast and
slow light vector; t = optical path length or thickness of the speci-
men (1.2 mm).
The stress-optic constant (Brewsters) = 2.6 B for soda lime and
3.65 B for borosilicate glass as expressed by Bourhis (2008). The
Brewsters constant shows that borosilicate glass will show better
birefringence under stress than soda lime glass. The retardation can
also be found out from the color pattern with the help of classical
Michel–Levy interference color chart. The relationship of specimen
thickness with retardation and birefringence are shown with color
plots in the classical chart.
The arrangement for obtaining the photoelastic effect on the
glass specimen is shown in Fig. 2. It has a light panel assembly of
visible light source as a workpiece supporting base. The light panel
consists of an array of LED light source over which a polarizer is
placed; this is followed by a color filter and another polarizer. These
arrangements are in-built features of the integrated light panel and
hence are not visible in the figure. The white light as shown in
Fig. 2 is, in fact, coming out through this panel. A RGB grid of size
0.3 mm is placed at below the glass specimen for reference of dif-
ferent fringe order. The stressed glass specimen over the panel and
a linear polarizer were positioned in between the specimen and an
image grabber.
A frame grabber was used to record the best frame of polar-
izing effect in terms of its fringe patterns. The distance between
the fringe patterns was calculated from the light screen grid dis-
tance of 0.3 mm. An image grabber camera (VGA-video graphic
array, 640 × 480 resolutions, CCD imaging sensor) was effectively
used at low light intensity to capture the image of a narrow crack
of the glass. The wavelength range of visible light is 0.4 to 0.7 ␮m
(however, in the present study a source with the wavelength nearly
0.565 ␮m was used only for calculation purposes). The visible light
wave was converted to one direction by a polarizing lens. In the
present study, distortion in HAZ of glass was examined. The glass
specimens were mapped on a grid of 0.3 mm through a linearly
polarized lens (Canon make, 58 mm, two-field filter). The cracks
were generated due to residual stresses induced by microwave heat
energy. The extents of crack formation were constrained within the
(1) boundary of HAZ through a solid precursor of thin perspex (Lautre
et al. (2015)). The collection of pixels (or pels) forms the grid, which
was then used to estimate the residual stresses.
The polarizing filter creates a rainbow like effect (Iridescence)
when light passes through a transparent medium. However, the
incandescent bulb used in setup has significant role in reducing the
iridescence. The stressed areas thus are highlighted on the spec-
imen due to this photoelasticity effect. The rainbow effect is not
always caused by birefringence. The effect is due to the stresses
caused by heating and then cooling of glass, to randomly aligning
the carbon, hydrogen and oxygen atoms. The alignment of chain
atoms changes refractive index along the stressed direction. Very
often, annealing of the glass (heat up to glass transition temperature
560 ◦ C) helps to reduce the glass cracking due to residual stresses.
 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161
The light while coming out from the specimen forms interference:
constructive or destructive effect. The occurrence of constructive
and destructive interference is not too significant for the specimen
thickness, (t) > visible . The interference depends upon the phase lag
of stressed and unstressed wavelengths. The isochrome lines show
a constant color due to equal retardation. Isogyres show the dark
lines (isoclinics fringe) of preferred vibrations whereas melatopes
are points with the isogyres, which defines the optical axis. The con-
structive and destructive interference give rise to various fringes,
which are seen clearly if residual stresses are present in the spec-
imen. The order of a fringe, the fringes and stresses are related
as,
n ×  = N;1 − 2 = F × N/t
where, n = order of the fringe; if n = 0,1,2,3 etc it gives larger inten-
sity of light; if n = 0.5, 1.5, 2.5, etc., it gives destructive interference
with no light intensity;  = light wavelength through soda lime
glass (0.565 ␮m); F = material fringe value = /(K × t) = 181.1 MPa;
N = fringe order which shows the cycle required to generate the
retardation. As the fringe order increases, the color intensity
reduces. When n = 0, the zeroth order has black color (isotropic
point); n = 1 is the first order, it varies from yellow–red to
blue–green–yellow–red; n = 2 is the second order, it varies from
blue–green–yellow–red to green–yellow–red and the n = 3 is the
third order which varies with faint red–green–yellow colors. The
LED array source considered was very cheap and uses five different
colors to generate white light.
The photoelastic approach was used primarily to extract infor-
mation of isochromatic and isoclinic from the fringe patterns. The
relation among the various parameters used in capturing the image
with fringes are stated as,
(D × G)
=
(n × S)
where, D is distance between two fringes (mm); G = 0.3 mm is the
distance between the selected grid array; S = 180 mm is the distance
between the work specimen and the image grabber. The variation
in the distance between two fringes versus the fringe number is
shown in Fig. 3. The zeroth fringe is the reference point of the mea-
surement of the fringe. It is usually of black color with no influence
of wavelength. The first fringe order distance is more than the grid
distance of G = 0.3 mm in case of induced stresses which can be
measured from the grabbed image. The distances may vary as per
the stresses developed in glass; but for estimating small residual
stresses the distance results are largely insignificant. The analy-
sis yields good results until the third fringe order, but it can be
Fig. 3. Distance of fringes formed in stressed glass specimen.
155
extended up to seven fringes. A spectral content analysis (SCA) was
used to find the fringe order from a white light. The relative retarda-
tion increases with the increase in the stress values in accordance
with the characteristics of isochromatic fringes.
4. Results and discussion
Drilling of holes on soda lime glass using microwave energy
could be achieved in the present work; however, cracking of glass
was observed as one of the major defects during the process of irra-
diation and ablation through concentration of the applied energy.
The other typical defect observed during microwave drilling
process is drill tip burning, heat scorch marks on surface, circum-
(2)
ferential pile up of non-ablated glass, debris of tool and glass, heat
affected zone and cracks developed in glass. The cracks developed
on the thin glass due to the thermal stress induced by plasma heat
energy are stochastic. It has a different pattern like linear, curved,
rounded, crossed, spider webbed, multiple or combinational and
random cracks (Lautre et al. (2015)). The variation in the morphol-
ogy of the crack may be due to thermal stresses and glass surface
treatment with chemicals like wax and oil.
4.1. Experimental results
Typical failed glass pieces during the drilling trials with molten
zone and cracks are presented in Fig. 4. Defective pieces were
then taken for characterisation through photoelasticity method.
The isochromatic fringes are found at the crack tip and point of
application of the compressive residual load as shown in Fig. 4(a).
The isotropic point reflects the point of application of compressive
load. The color effect for each fringe (<3 fringes) has the yellow
color as a distinguishable color to decide the number of fringe as
(3) observed by Murphy (2011). The maximum principal stress direc-
tion was found parallel to the direction of the crack. The fringe
order is numbered in Fig. 4(a), the variation in the color and shape
of the fringes show the variation of stresses accordingly. At high
microwave power (900 W); the glass melts around the hole within
few seconds (40 s) as shown in Fig. 4(a). A better variation in the
stress distribution was observed at 600 W power and glass cracking
time of 50 s (Fig. 4(b)). The melting is not uniform due to variation
in temperature and stresses around the hole. The hole circularity
is lost due to the fluctuating nature of heat. The heat flux fluctua-
tion resulted in relatively less stressed region and some with more
stresses.
The tensile stress shows clear formation of the fringes, usu-
ally up to three fringes. The crack propagation under the thermal
156 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161
Fig. 4. Defect formation in glass at (a) 900 W, 40 s (melting) and (b) 600 W, 50 s (variation in tensile stresses).
Fig. 5. Defect formation in glass at (a) 360 W, 50 s (crack propagation under tensile load) and (b) 90 W, 45 s (crack propagation around critically stressed region).
stress shows these fringe patterns along the propagation direction
as shown in Fig. 5(a). It further reconfirms that the residual stresses
also assist in propagation of crack in glass. Even at medium power
of 360 W, the glass cracks within 50 s and shows the fringes along a
linear crack propagation direction. The linearity of crack found ter-
minated within HAZ at the lowest possible power setting of 90 W
within 45 s. Further investigation showed the occurrence of circular
crack as an inhibitor to linear crack as shown in Fig. 5(b). A critically
stressed region around which circular crack occur is easily identi-
fied as an extreme boundary of HAZ, beyond, which compressive
stresses exists. The generation of crack from the hole boundary is
due to variation in stresses near to the hole boundary which gen-
erates crack within HAZ while the circular cracks are around the
hole. Within the HAZ, it is mostly tensile stresses and small portion
of compressive stresses. Circular cracks are found mostly around
the critically stressed region.
The investigation at marginally higher power than the lowest
power level (180 W) reveals similar behavior as observed at 90 W.
However, the occurrence of round crack or linear crack or both were
intensified with much change in the microwave irradiation time.
In addition to crack, the melting of glass around the circumference
(Fig. 6(a)) and non-uniform melting within HAZ (Fig. 6(b)) were
also noticed at 180 W power. It was understood that although the
increasing power level helps in melting the glass, it also generates
cracks within HAZ.
The need of surface treatment of glass surface with wax and oil
precursors were found effective in changing characteristics of the
cracks. A series of liquid and solid materials were used to under-
stand and arrest the crack through glass surface treatment (Lautre
et al. (2015)). The small reduction in the intensity of occurrence
of crack and melting was observed with wax at 180 W. At power
level of 360 W, the role of wax in crack reduction was found sig-
nificant with uniform glass melting around the hole as shown in
Fig. 7(a). The hole generation time with crack is reduced from 45 s
to 30 s. An examination of the profile of a sectioned glass specimen
shows relatively less variation in stresses along the depth of the
glass (Fig. 7(b)). The traces of the spherical plasma ball as a cav-
ity was also observed confirming the role of hot plasma ball (LMH
zone) during microwave drilling of glass. At low power (90 W); the
crack formation was also delayed to 35 s. The small variation, as
observed even away from the blind hole, is due to cutting of glass
for profile projection.
The surface treatment with low viscous material like olive oil
at low power ranging from 90 W to 360 W results in frequent oil
burn patches. At 360 W, the oil burnout is visible on the glass sur-
face (Fig. 8 (a)) and at 90 W, it was observed inside the cavity
(Fig. 8(b)). The behavior of crack is very random with no distinc-
tion between linear and circular cracks at relatively high power.
However, melting of glass was controlled due to the heat energy
balanced by the oil. This highlights the need for further investiga-
tion on suitable surface treatment material to minimize the crack
and melting of glass around the hole. The oil gets burnt inside the
hole, which reduces the plasma intensity. The low viscosity surface
treatment allows the oil to enter into the cavity resulting either
the ceasing of hot plasma and release of black smoke leaving black
patch residue inside the cavity. Moreover, an early investigation
(Fig. 7(b)) suggests less involvement of treatment along the depth
of the glass specimen. The initial defects like occurrence of crack
from hole and around the critically stressed region were totally
transformed to linear crack with offset from the hole as shown in
Fig. 8(b).
A good color intensity of fringes help in better stress pre-
diction, the high color intensity fringe pattern separated by
yellow color resembles less stress. The faded color fringe rep-
 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161 157

Fig. 6. Defect formation in glass at (a) 180 W, 45 s (crack propagation and melting) and (b) 180 W, 40 s (HAZ melting).

Fig. 7. Defect in glass at (a) 360 W, 30 s (circular crack, surface treatment with wax) and (b) 90 W, 35 s (stress variation along depth).

Fig. 8. Defect formations in glass at (a) 360 W, 30 s (random crack and burn, surface treatment with oil) and (b) 90 W, 30 s (straight crack with oil burning).

resents the increased value of the stress. For the circular
fringes, the center point highlights the point of application of
the compressive load. The zeroth fringe (n = 0) is independent
of the wavelength and always exists at the same spot, as it
is the point of applicable of active forces. When n = 1, then
N = 0.565 ␮m (from Eq. (2)) to get the difference in two stresses
as 1 − 2 = (181.1 × 565)/1.2 = 0.0852 N/mm2 . On the other hand,
working with the other case, when n = 1, then N = 0.565 ␮m (from
Eq. (1)) and R = 0.575 ␮m for red–blue transition (from Standard
Michel–Levy interference color chart) to get the difference in two
stress as 1 − 2 = 575/(1.2 × 2.6) = 184.29 N/mm2 . The difference
in refractive indices for an order, from Eq. (1), is 1 − 2 = K
(1 − 2) = 2.6 × 184.29 = 479.14 × 10−6 is small and within the
limit of the glass refractive index. However, the refractive indices
from Eq. (2) is = 0.221 × 10−6 (= 2.6 × 0.0852) is very small as if the
refractive index is not changed even after microwave heating. This
clears that the difference in stresses on purely birefringence value
(Eq. (1)) and order basis assumption (Eq. (2)) may give impractical
results. Hence, the stress relation obtained from retracted assump-
tion (Eq. (1)) with fringe order from Eq. (2) was considered for
further analysis. However, the retraction value (R) in Eq. (1) was
obtained for specific fringe order (Eq. (2)) from the Michel–Levy
interference color chart. As a starting case, let us assume the ther-
mal stress for glass with a relationship (Table 1) of coefficient of
thermal stress as  = 0.62 MPa/◦ C and the least shock resistance
annealing temperature as 38 ◦ C, the initial minimum stress value
as 2-init = 23.56 MPa. The variation of maximum principle stress
(1-init) can hence be predicted accordingly.
4.2. Simulation results
In case of the assumption of one of the stresses to 23.56 MPa,
a verification of the results could be obtained through simulation
analysis. The simulations were performed on COMSOL Multiphysics
158 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161

Fig. 9. Meshing (a) Coarse and (b) Finer for the glass specimen with nichrome tool.

simulation tool (version 4.4.0.195) based on Finite element analysis
(FEA) and thermal stress model. The parameters like fringe order,
retardation and others were subjected to variations. However, the
analysis is based on the assumptions matching with experimental
observations. The software tool was used to see the closeness of FEA
results with the experimentally observed photoelasticity results.
The details of the approach, formulation and the procedure have
been presented elsewhere (Lautre et al. (2014a,b)). The meshing
was attempted for different element morphology. A coarser mesh-
ing is shown in Fig. 9(a) and the finer meshing as used in the present
simulation, is shown in Fig. 9(b). It was ensured that the meshing
is beyond the coarse limit, as in coarse meshing the edges, face
or narrow region are too short than the smallest element size of
2.13 mm. Finer meshing was considered with smallest element size
of 0.304 mm, number of vertex elements were 20, number of edge
elements were 408, number of boundary elements were 4296 and
total number of elements were 7131.
The simulation results of von Mises stress variation by Gauss
point evaluation is shown in Fig. 10. The results show the presence
of both the tensile and compressive stresses on the surface and
within the HAZ region. The positive value shows the tensile stress
and compressive stresses are shown as negative value. The con-
tour lines obtained by Sakaue et al. (2008) are similar to the image
obtained by micro-retarder array incorporated in CCD camera. The
macroscopic view shows the contours of compressive load around
the HAZ and the contours tensile load near to the tool. The varia-
tion of these loads (stresses) may lead to circular cracks around the
tool and with or within HAZ as confirmed by Fig. 4(a and b). The
closer view near the tool in the inset (Fig. 10) shows the presence of
compressive stresses in the four quadrants of the tool as A–B–C–D.
The emergence of compressive stresses in quadrant (A–B–C–D)
initiate crack at junction E as the HAZ has tensile stresses near to
the tool. The generated thermal imbalance, even after the tool had
penetrated inside the glass, initiates cracks in glass surface. These

Fig. 10. Variation of stress on the surface during microwave drilling.

 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161 159

Fig. 11. Simulated temperature fields in the tool-surface interaction zone during microwave drilling.

Fig. 12. Temperature field on the surface of glass and the nichrome tool.

stresses are induced due to the development of critically stressed
region around the hole. The crack generated from the tool–glass
interface (E) propagates radially out to meet another stress vari-
ation region of the boundary of HAZ as depicted in Figs. 5(b), 6(a
and b), 7(b). The edge of the penetrated tool is marked as F and was
identified as the source of causing the tensile stresses in the glass.
It also validates that the tool tip is maintaining the hot plasma ball
even after the penetration inside the glass specimen. The diame-
ter of spherical ball was assumed to be of nichrome bit diameter
(0.7 mm) for simulation. The microwaves have not affected the bulk
glass volume, due to its characteristic of being microwave transpar-
ent material. The energy of hot plasma ball was also maintained by
the continuous attraction of the microwave even after the pene-
tration, to generate the required heat energy. The crack generating
from the tip of the tool was not observed experimentally as shown
in Figs. 7(b), 8(b). The chances of initiation of the cracks are more
from the top layer of the glass.
It was observed that there is a reasonable agreement between
the simulated and experimental results of the variation of stress
and temperature in terms of initiation of cracks. However, some
of the observations are exceptionally out of control. The induced
tensile stress showed value greater than zero at more than half of
the surface near the drilled hole (Fig. 10). The change of stress value
is small near the hole for compressive stresses (negative stress). The
mismatch in the coefficient of thermal expansion inside and near
the critically stressed region leads to cracking, which are reflected
in the experimental results (Figs. 4–8).
It is thus, clear from the above discussion that, the crack is due
to the variation on glass temperature and corresponding stresses.
However, the tool temperature also plays a crucial role in the for-
mation of the cracks on the glass specimen. The change in surface
temperature as shown in Fig. 11 is between the tool and the glass
specimen. The temperature variation along the y-axis was noted as
A to C and along the x-axis as B to D. The value of the z-axis was
kept at 0.0012 m as the topmost layer of the glass surface. Any of
the coordinated distance from the center of the tool and within the
tool shows relatively lower temperature than that of the glass in
HAZ. The region of the coordinates of the tool shows temperature
in the range of 300 K, which is as good as ambient temperature.
The glass behavior in the HAZ region is totally aggravated by
the simultaneous existence of complex compression and tension
phenomenon of glass. The possible reason of initiation of the crack
may be either due to the tool being relatively cooler than the sur-
face of the melted glass or due to formation of critically stressed
160 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161
Fig. 13. Spatial stress profiles from birefringence and retardation consideration.
region. The simulation results of the temperature variation in HAZ
are shown in Fig. 12. The elevated temperature of glass within
the circular boundary formed by region A–B–C–D is around 491 K.
The nature of the boundary is similar to the experimental results
observed in Figs. 7 and 8. The stress rise in the glass as per its
property is at the rate of 0.62 MPa/ ◦ C. It further gives rise to a ther-
mal stress of about 218 MPa (Table 1). At this, the 2 = 218 MPa is
considered uniformly distributed temperature rise. The major prin-
cipal stress (1) will experience a linear rise, which is more than as
predicted during experimental results as shown in Fig. 13.
The variations are predicted from the birefringence and retarda-
tion consideration expressed in Eq. (1) and plotted in spatial form.
The stresses lead to the induced stresses from the lowest fringe
order 0.5 (Fig. 4) to give a higher induced stress of about 314.45 MPa,
which is sufficient to cause tensile cracking in the glass plate. The
rise of principal stress is not very sharp but it rises vigorously as the
other principle stress (2) is rarely uniform at different fringe order.
The sudden rise of temperature and the stresses or vice versa was
found reasonably sufficient to induce the aggravation of frequent
cracks of different nature during microwave drilling of glass.
5. Conclusion
Photoelasticity methodology is useful in estimating the region
of stresses on the glass. The concept was applied to understand
the residual stress induced in soda lime glass during microwave
drilling. The relationship of induced stresses and cracking of glass
could be effectively established. The observations can further be
used to retune the process for efficient drilling of glass. The stress
intensity around the cracked glass specimen was higher than the
cooler region of the glass. A stress field investigation was carried
out using stress-optic relationship. A problem of low birefringence
in terms of low isochromatic fringes or stress (low retardation)
for the glass was found to be an obstacle; also, a thin glass spec-
imen does not show good intensity fringes. This leads to the use of
the phase analysis method (Retardation), fringe order and standard
chart to analyze the stresses for a reasonable change in refractive
index of glass due to microwave heating. The possible reason of
initiation of the crack is the mismatch in the hot and cool interac-
tion zones. Firstly, the interaction was noticed in the HAZ between
more heat susceptible and less heat susceptible regions. Secondly,
at the extreme boundary of HAZ, where the heat affected glass
experiences tensile stresses and gets surrounded by the compres-
sive stresses of the relatively cool glass surface. Thirdly, it may be
due to the tool being relatively cool than the surface melted glass.
The probability of the initiation of the crack is more at the top
glass surface, which propagates further along the depth and across
the surface. The cracking could be controlled to some extent with
the help of suitable surface treatment during microwave drilling of
glass. The simulation studies confirmed that the stresses developed
due to the thermal load that lead to the induced stresses to a value,
which is sufficient to cause tensile cracking in the glass plate. The
photoelasticity approach hence helps to understand the variation
in stresses at some stage in microwave drilling, which are mainly
responsible for the occurrence of the cracks.
Acknowledgements
Authors wish to acknowledge the Board of Research in Nuclear
Science (BRNS), Government of India, for the financial assistance
provided for the work through the DAE project no. 2010/36/60-
BRNS/2048 titled Material joining and drilling with microwave.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jmatprotec.2015.
05.026
References
Aben, H., Ainola, L., Anton, J., 2000. Integrated photoelasticity for nondestructive
residual stress measurement in glass. Opt. Lasers Eng. 33, 49–64.
Ajovalasit, A., Petrucci, G., Scafidi, M., 2007. Phase shifting photoelasticity in white
light. Opt. Lasers Eng. 45, 596–611.
Ajovalasit, A., Petrucci, G., Scafidi, M., 2012. RGB photoelasticity applied to the
analysis of membrane residual stress in glass. Meas. Sci. Technol. 23, 25601 (8
pp).
Alhekail, Z.O.I., 2001. Electromagnetic radiation from microwave ovens. J. Radiol.
Prot. 21, 251–258.
Al-Shamma, A.I., Wylie, S.R., Lucas, J., Stuart, R.A., 2001. Design and construction of
a microwave plasma jet system for material processing. Pulsed Power Plasma
Sci. Dig. Tech. Pap. 2, 1308–1311.
Ashokan, K., Ramesh, K., 2009. An adaptive scanning scheme for effective whole
field stress separation in digital photoelasticity. Opt. Lasers Eng. 41, 25–31.
Beghini, M., Bertini, L., 1998. Recent advances in the hole drilling method for
residual stress measurement. J. Mater. Eng. Perform. 7, 163–172.
Bourhis, E.L., 2008. Glass: Mechanics and Technology, second ed. John Wiley &
Sons, ISBN: 978-3-527-61701-2.
Fini, A., Breccia, A., 1999. Chemistry by microwaves. Pure Appl. Chem. 71, 573–579.
George, T.J., Sharma, A.K., Kumar, P., 2012. A feasibility study on microwave drilling
of metallic materials. I-Manager’s J. Mech. Eng. 2, 1–6.
Jacob, J., Chia, L.H.L., Boey, F.Y.C., 1995. Thermal and non-thermal interaction of
microwave radiation with materials. J. Mater. Sci. 30, 5321–5327.
Jerby, E., Dikhtiar, V., 2000. Method and Device for Drilling, Cutting, Nailing and
Joining Solid Non-Conductive Materials using Microwave Radiation. US patent
6, 114–676.
Jerby, E., Dikhtyar, V., Aktushev, O., Grosglick, U., 2002. The microwave drill.
Science 298, 587–589.
Jerby, E., Aktushev, O., Dikhtyar, V., Livshits, P., Anaton, A., Yacoby, T., Flax, A.,
Inberg, A., Armoni, D., 2004. Microwave drill applications for concrete, glass
and silicon. Proceeding of 4th World Congress Microwave and
Radio-frequency Applications, 156–165.
Jerby, E., Aktushev, O., Dikhtyar, V., 2005. Theoretical analysis of the
microwave-drill near-field localized heating effect. J. Appl. Phys. 97, 34909 (7
pp).
Jerby, E., Golts, A., Shamir, Y., Wonde, S., Mitchell, J.B.A., LeGarrec, J.L., Narayanan,
T., Sztucki, M., Ashkenazi, D., Barkay, Z., Eliaz, N., 2009. Nanoparticle plasma
ejected directly from solid copper by localized microwaves. Appl. Phys. Lett.
95, 91501 (3 pp).
 N.K. Lautre et al. / Journal of Materials Processing Technology 225 (2015) 151–161
Khanna, S.K., Ellingsen, M.D., Winter, R.M., 2004. Investigation of fracture in
transparent glass fiber reinforced polymer composites using photoelasticity. J.
Eng. Mater. Technol. 126, 1–7.
Kotake, H., Hirahara, K., Watanabe, M., 1980. Quantitative photoelastic
measurement of residual stress in lec grown gap crystals. J. Cryst. Growth 50,
743–751.
Lautre, N.K., Sharma, A.K., Kumar, P., Das, S., 2014a. Performance of different drill
bits in microwave assisted drilling. Int. J. Mech. Eng. Rob. Res. 1, 22–29.
Lautre, N.K., Sharma, A.K., Kumar, P., Das, S., 2014b. Microwave drilling with litz
wire using a domestic applicator. Bonfring Int. J. Ind. Eng. Manag. Sci. 4,
125–131.
Lautre, N.K., Sharma, A.K., Kumar, P., Das, S., 2015. On crack control strategy in near
field microwave drilling of soda lime glass using precursors. J. Ther. Sci. Engg.
Appl. 7, 041001 (15 pp).
Meir, Y., Jerby, E., 2012a. Localized rapid heating by low-power solid-state
microwave drill. IEEE Trans. Microw. Theory 60, 2665–2672.
Meir, Y., Jerby, E., 2012b. The localized microwave-heating (lmh) paradigm –
theory, experiments, and applications. Proceeding of the 2nd Global Congress
on Microwave Energy Applications, 131–138.
161
Murphy, E.B., 2011. The return of photoelastic stress measurements: utilizing
birefringence to monitor damage and repair in healable materials. J. Mater.
Chem. 21, 1438–1446.
Pagni, P.J., 2002. Thermal glass breakage. Fire Safety Science – proceedings Of The
Seventh International Symposium, 3–22.
Ramji, Ramesh, M., k, 2008. Whole field evaluation of stress components in digital
photoelasticity – issues, implementation and application. Opt. Lasers Eng. 46,
257–271.
Sakaue, K., Yoneyama, S., Kikuta, H., Takashi, M., 2008. Evaluating crack tip stress
field in a thin glass plate under thermal load. Eng. Fract. Mech. 75, 1015–1026.
Shayeganrad, G., Mashhadi, L., 2009. Theoretically and experimentally
investigation of sparking of metal objects inside a microwave oven. PIERS
Proc., 632–641.
Wang, C.J., Chen, G., Luan, D., Zhang, P., 2014. Effect of laser power on the quality of
drilled micro hole using cu50 zr50 amorphous alloys foils. Micromachines 5,
1061–1068.
Withers, P.J., Bhadeshia, H.K.D.H., 2001. Residual stress part 1–measurement
techniques. Mater. Sci. Technol. 17, 355–365.