Você está na página 1de 8

Measurement 46 (2013) 25772584

Contents lists available at SciVerse ScienceDirect

Measurement
journal homepage: www.elsevier.com/locate/measurement

Impact of parameters on the process response:


A Taguchi orthogonal analysis for laser engraving
S eka Kasman
_
Turkey
Dokuz Eylul University, Izmir Vocational School, Buca, Izmir,

a r t i c l e

i n f o

Article history:
Received 6 June 2012
Received in revised form 4 January 2013
Accepted 17 April 2013
Available online 2 May 2013
Keywords:
Laser engraving
Surface roughness
Milling depth
Taguchi method

a b s t r a c t
Compared to conventional methods, laser engraving is the most effective technique in the
machining of hard materials that have a complex geometry. Therefore, laser based machining is widely used in many industries like mold making, and the manufacture of automotive, electronics and biomedical parts. The present study investigates the machinability of
hard metal produced with powder metallurgy and puts forward a new approach relating to
the laser engraving of P/M metals. The main objective of this study is to determine the
impact of laser engraving process on Vanadis 10. For this purpose, three process parameters namely effective scan speed, frequency, and laser effective power were correlated
with the surface roughness (Ra) and engraving depth (D). The Taguchi and linear regression
were used in the analysis. The experiments were performed in accordance with an L9
orthogonal array. Based on the S/N ratio for Ra and D, the optimal condition was found
as SS3F2P1 for Ra and SS1F2P3 for D. It was found that scan speed has a statistically signicant
effect on both Ra and D. Furthermore, a mathematical model for both Ra and D was established and estimated using linear regression. The model was also tested using different
experimental conditions than existing ones. The results obtained from the new experimental conditions show that the predicted models could explain the process.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
Machining of high strength material is the most important issue in the manufacturing industry [1]. During the
processing of such materials, the expectation is for high
surface quality, reduced tool wear, as well as tool-based
problems that cause increased machining cost and surface
roughness [2]. A high quality surface is desired in mold and
parts [3] which are used in automotive, electronics and
biomedical industries. Beside this, the machining of complex geometries with a high surface quality is greatly affected by working the material and forming the tool.
Therefore, the machining strategy for a specic material
is the most important criterion [3].

Tel.: +90 2323012579.


E-mail address: seka.kasman@deu.edu.tr
0263-2241/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.measurement.2013.04.022

In recent years, many milling methods have been developed to machine small, complex geometry parts. Conventional milling methods generally cause some tool-based
problems on the machined surface [4]. In some cases, the
geometry to be machined is not achieved due to the geometrical limitations of the cutting tool. Therefore, thermal
machining methods are the best alternative when considering cutting tool based problems. In the metal machining
industry, electrical discharge machining (EDM) and laser
beam machining (LBM) are among the few thermal
machining methods, and they are commonly used to remove material at the interaction region that is created by
the vaporization mechanism. It is well known that lasers
are used in many production methods such as cutting, drilling, scribing and welding. In recent years, lasers have been
also used for deep engraving applications. Laser based
machining processes are non-contact [5] and use a high
intensity beam as the cutting tool. Therefore, problems

2578

S. Kasman / Measurement 46 (2013) 25772584

based on the cutting tool, the difcult to machine geometry and work piece material hardness do not arise [57].
According to the Hnilica et al. [8] ledeburitic steels produced by powder metallurgy have many advantages compared with steels whose chemical composition is
fabricated using a conventional casting route. Arslan
et al. [9] indicated that, compared to conventional, standard steels, these types of alloyed steels show high
mechanical properties like high hardness, toughness and
wear resistance.
High chromium vanadium steels are used in the mold
industry and, in addition, the complex geometries for these
types of materials are performed by a wire electrical
machining method [10]. Vanadis 10 is one of the high
vanadium alloyed powder metallurgy tool steels [11],
and it is characterized by very good dimension stability
and wear resistance [10]. It is also used in the mold industry and cold forging, and for extruder screws, granulator
knives, deep drawing, blanking of electrical sheets and gasket stamping.
A literature review related to laser assisted milling has
shown that most of the studies have focused on determining the heat affected zone (HAZ) and the thickness and formation of the recast layer. These studies were performed
for the modication of the surface. In additions to these
studies, there are studies on milling and engraving of hard
to machine metals.
Orazi et al. [12] studied laser surface micro-manufacturing to predict the material removal rate. They built a
regression model of material removal rate for AA 7075 aluminum alloy. Zhang et al. [13] performed laser microprocessing and post chemical etching on an ultrathin Al2O3
ceramic substrate. To evaluate the process, laser power
density, laser scan speed and pass delay was corralled to
the kerf width, and surface roughness on the kerf sidewall.
To investigate the effect of the parameters, they conducted
an experimental layout using the orthogonal array design.
They concluded that a higher laser power with combined
higher scan speed and frequency produces higher microprocessing efciency, quality and accuracy. Yasa and Kruth
[14] carried out selective laser erosion applications on an
AISI 1045 steel block by using different process parameters
like scan strategy, scan speed, scan spacing, spot size and
laser power to determine the effect of the parameters on
the surface roughness and removed depth. They also investigated the effect of an overlapping ration on the surface
prole. Their results show that frequency, scan speed and
spot size are the signicant factors for overlapping that directly affects surface prole and heat input. Moreover, scan
speed and power determine the heat input and directly affect the amount of removed depth. Samant and Dahotre
[15] studied the machining of structural ceramics using a
different laser type. They investigated the effect of parameters on the heat input, surface roughness and material removal rate, using different applications. Their results show
that the machined geometry formed by a laser beam and
the material removal rate are both a function of the laser
process parameters. Singh and Melkote [16] carried out
experiments to characterize the hybrid laser-assisted
mechanical micromachining process. The micro-scale
grooving experiments were conducted on H13 tool steel

to evaluate the parameters effect on the process responses


(the cutting force, groove depth and surface nish). They
used statistical tools, the Taguchi method and analysis of
variance (ANOVA). Apart from this, the load deection
curve for a cutting tool was simulated using ANSYS. Their
results show that the parameters signicantly affect the
process responses. Dumitru et al. [17] carried out experiments using pulsed laser systems operating in IR and UV,
with ns and fs pulses to determine the impact of laser
parameters like wave length, energy density and pulse
duration on the WCCo material. They analyzed HAZ and
microstructure. Another study relating to laser milling
was carried out by Saklakoglu and Kasman [18]. They
investigated the effect of process parameters such as laser
power, scan speed, frequency and ll spacing on the surface roughness and amount of removed rate (depth) in
multi-beam scan machining conditions. They modeled
and analyzed the process using regression analysis. Their
results show that power and scan speed has a signicant
effect on the surface roughness and depth.
The present study investigates the effects of laser
engraving parameters on the surface roughness and depth.
In the laser based machining techniques, the scan speed
and power have inverse effect on the surface roughness
and machining depth. In other words, while an increase
in scan speed improves the surface roughness, it causes
to decrease in the milling depth. A similar case is shown
in the effect of laser power. The surface roughness and
the machining depth increase with increase in the laser
power. However, larger surface roughness is not desired
in the precision surface machining technology. Therefore,
the strategy of present study was built on separately to
minimize surface roughness and maximize the machining
depth. It is considered that Taguchi method presents
appropriate statistical solutions to reach desired responses.
The present study investigates the machinability of hard
metal produced with powder metallurgy and puts forward
a new approach relating to the laser engraving of P/M
metals.
To perform the experiment, an experimental design
matrix was constituted using the design of the experiments. For this purpose, the Taguchi method was chosen
as a statistical tool to build an experimental layout. Two
main criteria for the engraving experiments were selected
to minimize the surface roughness and to maximize material removal rate (engraving depth). Therefore, the process
performance criteria for this study are minimum surface
roughness and maximum engraving depth at a xed number of beam scan.

2. Laser engraving process


Laser engraving is one of the laser beam milling methods. In this method, a high intensity beam moves over
the surface in accordance with the desired geometry. During the interaction of the laser beam with the material, an
amount of laser energy is absorbed by the material [5,6].
This energy, which is affected by scan speed, laser power,
and beam overlap rate, determines the amount of removed
material. Therefore, the scan speed and the effective power

2579

S. Kasman / Measurement 46 (2013) 25772584

are the most important parameters on both surface roughness and material removal rate. The scan speed determines
the interaction time between the laser beam and material.
Any increase in interaction time causes an increase in the
local temperature. With the increase in temperature, the
surface temperature reaches the vaporization point, and
an amount of material vaporizes from the interaction region. As a result of the interaction process, a crater and recast layer occurs. The interaction time between the laser
beam and material surface determines the crater form
and depth. The crater forms determine the surface
roughness.
3. Experimental details
3.1. Material
The material used in the experimental studies is a powder metallurgy cold work tool steel: Vanadis 10 (ASSAB
Company). The specication and chemical composition of
the material are listed in Table 1. Before experiments, the
material was hardened to 60 HRC.
3.2. Laser engraving experiments and experimental design
A maximum laser power of 50 W Yb: Fiber Laser marking machine at 1064 nm wavelength was used for the surface engraving process. The focal length, which was held
constant during the experiments, is 160 mm. A sample
for the machined surface is given in Fig. 1. During the
engraving experiments, the sample was xed on a plate.
For all engraving experiments, the beam scan times and ll
spacing were kept at 30 times and 0.04 mm, respectively.
The beam scan direction was selected to be in a parallel
direction (0) to the X-axis. Experiments were performed
at standard atmospheric conditions without using any assist gas.
The three main factors that affect the machining process are the scan speed, frequency and effective power.

Preliminary experiments were performed to determine a


range for the parameters. The ranges for scan speed and
power were selected from within the values which give
the best and worst surface roughness and engraving depth.
Since a value lower than 25 W for laser power causes a
reduction in the engraving capability, a range for the
power was selected between 25 W and 50 W. The frequency ranges are 20 KHz and 80 KHz. Values higher than
50 KHz causes a decrease in the engraving capability when
in interaction with the other parameters. Therefore, range
for frequency was selected between 20 KHz and 40 KHz.
The selected factor levels and machining conditions are given in Table 2. The process responses of this study are surface roughness and engraving depth. The responses were
correlated with the process parameters and machining
strategy. Therefore, an experimental plan was constituted
based on the Taguchi L9 orthogonal array.
The surface roughness of the engraved geometry for
each condition was measured by using a Mitutoyo SJ301
stylus type prolometer. Each surface was characterized
by the average surface roughness (Ra) value. The cutoff
length (kc) and the sampling number (N) were selected as
0.8 and 5, respectively, and the travel length was 4 mm.
In total, four different measurements to scan direction
and perpendicular to the scan direction were taken on
the textured surface. The average of those four measurements was used to nd the ultimate Ra value. The amount
of engraving depth (D) was determined by using Mitutoyo
digimatic indicators. In total, three measurements were
performed to determine the D value. The average of the
three measurements was used as the ultimate D value.
3.3. Methodology of statistical analyses
To determine the optimal combination for engraving
parameters, two stages the Taguchi method and analysis
of variance (ANOVA) were used.
Table 2
Laser engraving parameters.
Notation

Parameters

Table 1
The chemical composition of Vanadis 10.
C

Cr

Mn

Mo

Si

Fe

2.90

8.0

9.8

0.5

1.5

0.5

Bal.

SS
F
P

Scan speed (mm/s)


Frequency (kHz)
Power (W)

Fig. 1. Laser beam scan direction (a) and machined surface (b).

Levels
1

200
20
25

500
30
37.5

800
40
50

2580

S. Kasman / Measurement 46 (2013) 25772584

The trend in manufacturing and metal machining


industry is towards reducing cost, time and nding effective machining solutions. In this context, the researchers
focused on machining cost through decreasing time and
achieving high production efciency such as good surface
and high material removal rate. The number of experiments is the particular criteria that directly affect machining cost. Additionally, the number of parameters and their
levels in a process determine the number of experiments.
Increasing the number of parameters for a specic process
leads to an increase in the number of experiments [19]. It is
known that a well dened experimental layout can be a
solution key for a specic investigation. The design of
experiment (DOE) method provides many solution tools
both to decrease the number of experiments and to optimize the process conditions. The Taguchi method is also
one of the solution tools that help in reducing the number
of experiments [20]. The Taguchi optimization technique
consists of following procedures [21]:
Determining the experimental layout.
Calculation of signal to noise (S/N) ratios.
Selection of optimal levels.
Prediction of optimal surface roughness and engraving
depth.
 The conducting of conrmation experiments.
 Performing the ANOVA.





4. Statistical analysis and discussion


The laser engraving experiments were implemented to
determine the effect of scan speed, frequency and effective
power on the surface roughness and engraving depth.
Those parameters with multi responses were analyzed
using the Taguchi method. It is well known that, in a surface milling process, better performance is characterized
by the lowest surface roughness at the highest milling
depth. The laser engraving process is also a surface
machining process that offers the lowest surface roughness. Furthermore, a higher material removal rate also corresponds to a better process performance. Therefore, the
statistical studies in the present study were oriented on
the basis of these outputs.
The Taguchi statistical analysis used to determine the
optimal process conditions was carried out following the
steps below.
4.1. Determining the experimental layout
In this study, the scan speed, frequency and effective
power at three different levels were selected as the control
factors. The three levels of each parameter have two degrees of freedom (DOF), and thus total DOF is calculated
as eight. As indicated in the paper by Kumar et al. [22],
the total DOF of a selected orthogonal array should be
greater than the total DOF required for the experiment.
Therefore, the L9 Taguchi orthogonal array design was selected as the experimental layout. Using the L9 experimental layout, a total of nine experiments were carried out to
determine the optimal value with the highest effect for

minimizing surface roughness and maximizing engraving


depth.
4.2. Calculation of signal to noise (S/N) ratio
The Taguchi method uses the quality characteristics of
signal-to-noise (S/N) ratio to determine the optimal combinations for process response. Mainly, there are three S/N
ratios that are used for calculations based on the objective
function for the response: the lower the better, the nominal the better and the larger the better. For the present
study, the lower Ra and higher D values indicate the better
performance. Therefore, the lower the better was used for
the S/NRa and the larger the better was used for the S/ND.
These two S/N ratios are calculated using the following
equations below [1921].
Smaller the better
If the strategy is to minimize the response, the S/N ratio
is found as follows:
n
1X
g S=N 10 log
Y2
n i1 i

!
1

Larger the better


If the strategy is to maximize the response, the S/N ratio
is found as follows:

g S=N 10 log

n
1X
1
n i1 Y 2i

!
2

where Yi is the performance characteristic (process response, Ra or D), and n is the number of experiments.
To determine the optimal combination for Ra and D, the
mean S/N ratio for each level of the parameters is calculated and a graph is produced using the mean S/N ratio.
The greater S/N ratio represents the optimal level of the related parameter in the graph.
Table 3 shows the experimental layout and measurement results for Ra and D, and the S/N ratios calculated
using Eqs. (1) and (2). The mean value of S/NRa and S/ND
for each parameter is listed in Table 4.
4.3. Selection of optimal levels
The main effect of each parameter on the Ra and D is
shown in Figs. 2 and 3. For the Ra (see Fig. 2) , the S/NRa ratio increased with increasing scan speed, decreased with
increasing power, increased between 20 KHz and 30 KHz,
Table 3
Experimental layout using L9 orthogonal array with responses (Ra, D).
No.

SS

Ra

S/NRa

S/ND

1
2
3
4
5
6
7
8
9

200
200
200
500
500
500
800
800
800

20
30
40
20
30
40
20
30
40

25
37.5
50
37.5
50
25
50
25
37.5

19.01
20.36
25.02
10.91
12.04
9.63
8.31
3.55
7.07

25.58
26.18
27.97
20.76
21.61
19.67
18.39
11.00
16.99

256
387
472
192
270
124
143
84
112

48.16
51.75
53.48
45.67
48.63
41.87
43.11
38.49
40.98

2581

S. Kasman / Measurement 46 (2013) 25772584


Table 4
Main effects of S/NRa and S/ND.
Level

S/NRa

S/ND

SS

SS

1
2
3

26.57
20.68
15.46a

21.58
19.60a
21.54

18.75a
21.31
22.66

51.13a
45.39
40.86

45.65
46.29a
45.44

42.84
46.13
48.40a

Rank

Optimal level.

Mean of S/N ratios for Ra (m)

Table 5
Optimal levels for Ra and D.

SS (mm/s)

-15,0

F (KHz)

P (W)

-17,5
-20,0
-22,5
-25,0
-27,5
200

500

800

20

30

40

25,0 37,5 50,0

Fig. 2. Main effects plot (data means) for S/NRa.

Mean of S/N ratios for D (m)

SS (mm/s)

F (KHz)

P (W)

52
50
48
46
44
42
40
200

500

800

20

30

40

25,0

37,5

50,0

Fig. 3. Main effects plot (data means) for S/ND.

and then decreased between 30 KHz and 40 KHz. The


greater value of the mean S/NRa for each parameter minimizes the Ra value. As seen in Fig. 2 and from the data given in Table 4, the Ra reaches the minimum at the 800 mm/
s for scan speed, 30 KHz for frequency and 25 W for effective laser power.
A similar procedure was performed to determine the
optimal level of each parameter that maximizes the
engraving depth (D). The main effect of each parameter
on the S/ND is shown in Fig. 3. The results are different from
those obtained for the Ra. According to Fig. 3, the S/ND ratio
decreased with increasing scan speed, increased with
increasing power, and increased to 30 KHz and then decreased to 40 KHz. The optimal combination was determined using the data given in Table 4 and Fig. 3. As
shown in Table 4 and Fig. 3, the optimal combination for
maximum D is 200 mm/s for scan speed, 30 KHz for frequency, and 50 W for laser effective power.

Parameters

Ra

SS (mm/s)
F (kHz)
P (W)

800
30
25

200
30
50

The results of optimal combinations are summarized in


Table 5. It can clearly be seen in the table that scan speed
and power have the same effect on the results. This means
that an increase in scan speed decreases both Ra and D. In
addition, increasing the power increases both Ra and D. The
expectation is to achieve the best surface roughness at a
deeper cavity. However, the inverse effect of scan speed
and power on both Ra and D has hampered this. The reason
for the inverse effect of scan speed on both Ra and D can be
explained by the interaction time and overlapping ratio.
The interaction time decreases with increasing scan speed,
and a decrease in interaction time causes a decline in heat
input to the material. Thus, the amount of vaporized material at the interaction zone decreases. The difference between the optimal conditions for Ra and D is due to the
interaction time between the laser beam and the material.
The pulse overlapping ratio also affects the Ra and D. Pulse
overlapping is dened as the distance between two successive laser beams. The pulse overlapping ratio decreases
with increasing scan speed, which affects the heat input
per unit area [14]. Based on the above, it can be said that
a greater value of scan speed corresponds to a lower Ra value and that a lower value of scan speed corresponds to a
greater D value. The proposed solution to achieve both
lower Ra value and higher D is to increase the beam scan
repeat number at a high scan speed. The effect of power
on both Ra and D can be also explained by the heat input.
A surface exposed to the laser beam has many microcraters. The geometry of these craters is affected by the overlap
ratio and heat input. A larger overlap leads to smooth surface roughness. In addition to the effect of the overlap ratio, the amount of heat input has an effect on the
formation of the crater geometry (width and depth), which
determines the surface topography. A lower heat input
produces a small crater size, and thus the surface topography improves. In a study by Kasman and Saklakoglu [23],
the effect of the overlap ratio and power on surface topography was revealed. It is apparent from SEM images that a
larger overlap and low power produce a smooth surface.
4.4. Prediction of optimal surface roughness and engraving
depth
In order to predict the Ra or D, the rst step is to select
the optimal level that has a greater S/N ratio than the other
levels of the parameters. The second step is to perform the
experiments to predict the Ra or D. As a result of S/N ratio
analysis, the optimal combination was determined for Ra
and D, and the results are presented in Table 5. To predict
the optimal mean value and S/N ratio for Ra and D, the following equations, Eq. (3), were used [24]:

g gm

n
X
i1

gi  gm

2582

S. Kasman / Measurement 46 (2013) 25772584

Using Eq. (3), the S/NRa and S/ND were calculated as


follows:

Table 7
ANOVA results for S/ND.
Parameters

S=NRa SS3 F 2 P1  2  Y Ra
S=NRa 15:46  19:60  18:75  2  20:91
11:99 dB
Ra 6:30 11:98 10:73  2  25:755 3:27 lm

D.F.
Seq. SS
Adj. MS
F
p
P.C (%)

Process parameters
SS (mm/s)

F (KHz)

P (W)

Error

Total

2
159.07
79.53
741.91
0.001
76.69

2
1.17
0.584
5.45
0.155
0.56

2
46.98
23.49
219.1
0.005
22.65

2
0.214
0.107

8
207.42

0.10

S=ND SS1 F 2 P3  2  Y D
S=ND 51:13 46:29 48:40  2  45:79 54:23 dB
D 371:7 247 295  2  226:67 460:36 lm
4.5. Conducting conrmation experiments
After determination of the optimal condition, an experiment was performed in order to determine the difference
between the predicted and measured values. The result of
the conrmation experiment for Ra is 3.55 lm. In addition,
the optimal combination for Ra is one of the experiment
conditions. A similar investigation was performed for D.
The predicted D value taken from the conrmation experiments is 470 lm. It can be seen that the difference between the predicted and measured values for D is almost
10 lm. The source of this difference can be explained by
the frequency. It is clearly seen in Table 4 that the S/ND ratio for each frequency is almost similar, and the difference
between the S/ND ratios is negligibly small.

respectively. The p-value in Table 6 shows that, except


for the scan speed, the other two parameters are statistically insignicant at a 5% level, implying that the parameters have insignicant impact on the Ra value. This table
also shows that the scan speed (SS), power (P) and frequency (F) affect the Ra value by 83.84%, 10.66% and
3.47%, respectively.
Similar investigations were performed for D. The results
given in Table 7 indicate that the scan speed and power are
statistically signicant. The other remaining factor (frequency) has a p-value higher than 0.05 is considered insignicant. Moreover, the scan speed (SS), power (P) and
frequency (F) affect the D value by 76.69%, 22.65% and
0.56%, respectively.
5. Development a regression model
To establish a mathematical relationship between
parameters and responses, linear regression analyses were
performed, and the following equations were developed:

Ra 17:17  0:0253  SS 0:0582  F 0:176  P

4.6. Performing of ANOVA

D 173  0:431  SS 1:95  F 5:61  P

The next analysis was performed by using the analysis


of variance (ANOVA). The ANOVA was used to investigate
the statistical signicance of the parameters at 95% condence level and to determine the percentage contribution
of the parameters to the process response. The signicance
of each parameter was tested using probability values (pvalue). When the p-value in the ANOVA table for S/N ratios
is less than 0.05 (95% condence level), it is considered
that the parameters are statistically signicant [25]. In
addition, the percentage contribution expresses the importance of the parameters for the response.
According to the ANOVA data for Ra given in Table 6, the
signicance and importance of the parameters were expressed by the p-value and percentage contribution,

The signicance of each coefcient in the equation and


the regression model were analyzed by ANOVA and tested
by the probability (p) value. Both regression statistics [19]
and ANOVA results for regression models are reported in
Table 8. Results for Ra indicate that, except for frequency,
all coefcients are statistically signicant at a 95% condence level. A similar investigation was performed for D,
and the results show that, apart from the coefcients for
constant value and frequency, the coefcients are statistically signicant at a 95% condence level.
The ANOVA results for the regression equation given in
Table 8 shows that the regression model for both Ra and D
is statistically signicant at a 95% condence level

Table 6
ANOVA results for S/NRa.

Table 8
Regression analysis and ANOVA results.

Parameters

D.F.
Seq. SS
Adj. MS
F
p
P.C (%)

Process parameters

Parameters

SS (mm/s)

F (KHz)

P (W)

Error

Total

2
185.44
92.72
41.51
0.024
83.84

2
7.69
3.85
1.72
0.367
3.47

2
23.59
11.79
5.28
0.159
10.66

2
4.47
2.23

8
221.20

2.02

Constant.
SS (mm/s)
F (KHz)
P (W)
ANOVA
Regression

Ra

Coeff.

Coeff.

17.17
0.0253
0.0582
0.176
SS
375.42

4.30
8.78
0.67
2.55
F
28.02

0.008
0.000
0.530
0.052
p
0.001

173.22
0.431
1.95
5.613
SS
132184

2.20
7.60
1.15
4.12
F
25.34

0.079
0.001
0.304
0.009
p
0.002

2583

S. Kasman / Measurement 46 (2013) 25772584

525

30
Actual value (A.V.)

Test value

Predicted value (P.V.)

375

20

D (m)

Ra (m)

Actual Value

450

25

15
10

300
225
150

75

Experiments number

Actual value (A.V.)

Predicted value (P.V.)

450

D (m)

375
300
225
150
75
0

Experiments number
Fig. 5. Comparing results of actual value and predicted value for D.

Table 9
Verication results for the regression model.
No.

SS

1
2
3
4
5
6

200
200
500
500
800
800

20
30
40
30
20
40

Ra (lm)

50
25
37.5
25
37.5
50

Experiments number

Fig. 4. Comparing results of actual value and predicted value for Ra.

525

D (lm)

T.V.

A.V.

T.V.

A.V.

22.06
18.25
13.44
10.66
4.68
8.04

22.52
19.03
12.06
9.76
5.81
8.92

406.67
285.85
246.21
156.55
77.91
167.57

458
278
196
134
109
155

Fig. 7. Test results of regression model for D.

to build a model, notated as actual value (A.V.), were compared, and the results are given as a graph as in Figs. 4 and
5. As shown in Fig. 4, the A.V. value and P.V. for Ra are very
close. A signicant difference was observed for Exp. 5,
which can be ignored. Similar procedures were carried
out for D, and the results are given in Fig. 5. As seen in
Fig. 5, there are no big differences and values are almost
similar. The results showed that the predicted regression
model for the two responses can be used for determining
and estimating the engraving depth in the laser engraving
process.
The models were also tested through experiments with
different experimental conditions used in the establishment of the regression model. Table 9 shows the test condition and its results. The results of the experiments are
presented in Figs. 6 and 7. As seen in Fig. 6, there is a high
correlation between the value taken from the test experiments (actual value) and the test value from the Ra model.
The actual Ra values are very close to the Ra value calculated from the model (predicted test value). This indicates
that the predicted model for Ra explains the process for the
actual laser engraving condition. In order to evaluate the D
model in actual conditions, Fig. 7 shows that, except for
Exp. 3, there is a small difference between the test results
and the actual values. The reason for this difference could
be the statistically insignicant constant value. In general,
the differences in D values are within acceptable limits
when considering the insignicant parameter (frequency).

25
Test value

Actual Value

6. Conclusions

Ra (m)

20
15
10
5
0

Experiments number
Fig. 6. Test results of regression model for Ra.

(a < 0.05). In addition, to verify the regression model (prediction model), the nine experimental conditions were inserted in the model. The data taken from the prediction
model, notated as predicted value (P.V.), and the data used

This study has investigated the effect of process parameters on the performance characteristics of the engraving
process. The Taguchi method was used to determine the
optimal process conditions, and the ANOVA was performed
to evaluate the statistical signicance of each parameter on
the performance characteristics. Based on the analyses, the
following conclusions were drawn:
(1) The scan speed appeared to be the main effective
parameter for the two performance characteristics.
The experimental results showed that increasing
scan speed decreases both Ra and D. To minimize
Ra, the scan speed should be selected at a high level
(800 mm/s), whereas to maximize D, the scan speed
should be selected at a low level (200 mm/s).

2584

S. Kasman / Measurement 46 (2013) 25772584

(2) The optimal level for Ra is where the scan speed is at


800 mm/s, the frequency is at 30 KHz, and power is
at 25 W (SS3F2P1). Optimal levels for D for each
material type were obtained at a scan speed of
200 mm/s, a frequency of 30 KHz, and with power
at 50 W (SS1F2P3).
(3) After determining the optimal level for each parameter, a conrmation experiment was performed for
both Ra and D. The predicted and experimental
results were compared, and the results show that
there is a high correlation between them. To achieve
a good surface nish, a high scan speed at the lower
power level is recommended.
(4) ANOVA was performed using the S/N ratio to investigate the signicance of each parameter and to
determine the percentage contribution to both Ra
and D. The effect of scan speed on both Ra and D is
83.84% and 76.69%, respectively. However, the effect
of frequency has little that determined by the p
value (0.367). The percentage contribution of frequency is 3.47% and 0.56% to the Ra and D,
respectively.
(5) To test the predicted linear regression model, a set of
experiments at different condition were used to predict both Ra and D. Results show that the predicted
results (test results) are very close to the actual
value. This means that Ra explains the model of laser
engraving process for Vanadis 10. Besides this, the
data taken from the test experiments also explain
the model for D.

References
[1] J. Leunda, V. Garca Navas, C. Soriano, C. Sanz, Improvement of laser
deposited high alloyed powder metallurgical tool steel by a posttempering treatment, Physics Procedia 39 (2012) 392400.
[2] H. Ding, R. Ibrahim, K. Cheng, S.J. Chen, Experimental study on
machinability improvement of hardened tool steel using two
dimensional vibration-assisted micro-end-milling, International
Journal of Machine Tools and Manufacture 50 (2010) 11151118.
[3] V.N. Gaitonde, S.R. Karnik, Luis Figueira, J. Paulo Davim, Analysis of
machinability during hard turning of cold work tool steel (Type: AISI
D2), Materials and Manufacturing Processes 24 (2009) 13731382.
[4] G. Poulachon, B.P. Bandyopadhyay, I.S. Jawahir, S. Pheulpin, E.
Seguin, The inuence of the microstructure of hardened tool steel
workpiece on the wear of PCBN cutting tools, International Journal of
Machine Tools and Manufacture 43 (2003) 139144.
[5] A.K. Dubey, V. Yadava, Laser beam machining a review,
International Journal of Machine Tools and Manufacture 48 (2008)
609628.
[6] A. Kaldos, H.J. Pieper, E. Wolf, M. Krause, Laser machining in die
making a modern rapid tooling process, Journal of Materials
Processing Technology 155156 (2004) 18151820.

[7] S.L. Campanelli, A.D. Ludovico, C. Bonserio, P. Cavalluzzi, M.


Cinquepalmi, Experimental analysis of the laser milling process
parameters, Journal of Materials Processing Technology 191 (2007)
220223.
[8] F. Hnilica, J. Cmakal, Peter Jurci, Changes to the fracture behaviour of
the CrV ledeburitic steel Vanadis 6 as a result of pasma nitriding,
Materiali in Tehnologije 38 (5) (2004) 263267.
[9] F.K. Arslan, I. Altnsoy, A. Hatman, M. Ipek, S. Zeytin, C. Bindal,
Characterization of cryogenic heat treated Vanadis 4 PM cold work
tool steel, Vacuum 86 (2011) 370373.
[10] S. Hatami, M.S. Navid, L. Nyborg, Surface preparation of powder
metallurgical tool steels by means of wire electrical discharge
machining, Metallurgical and Materials Transactions A 43A (2012)
32153226.
[11] www.assab.com.tr.
[12] L. Orazi, G. Cuccolini, A. Fortunato, G. Tani, An automated procedure
for material removal rate prediction in laser surface
micromanufacturing, The International Journal of Advanced
Manufacturing Technology 46 (14) (2010) 163171.
[13] F. Zhang, J. Duan, X. Zeng, Y. Cao, UV laser micro-processing and post
chemical etching on ultrathin Al2O3 ceramic substrate, Journal of the
European Ceramic Society 31 (2011) 16311639.
[14] E. Yasa, J.-P. Kruth, Investigation of laser and process parameters for
selective laser erosion, Precision Engineering 34 (2010) 101112.
[15] A.N. Samant, N.B. Dahotre, Laser machining of structural ceramics
a review, Journal of the European Ceramic Society 29 (2009) 969
993.
[16] R. Singh, S.N. Melkote, Characterization of a hybrid laser-assisted
mechanical micromachining (LAMM) process for a difcult-tomachine material, International Journal of Machine Tools and
Manufacture 47 (2007) 11391150.
[17] G. Dumitru, B. Luscher, M. Krack, S. Bruneau, J. Hermann, Y. Gerbig,
Laser processing of hardmetals: physical basics and applications,
International Journal of Refractory Metals and Hard Materials 23
(2005) 278286.
[18] I.E. Saklakoglu, S. Kasman, Investigation of micro-milling process
parameters for surface roughness and milling depth, International
Journal of Advanced Manufacturing Technology 54 (2011) 567578.
[19] N. Mandal, B. Doloi, B. Mondal, Reeta Das, Optimization of ank wear
using ZirconiaToughenedAlumina (ZTA) cutting tool: Taguchi
method and Regression analysis, Measurement 44 (2011) 2149
2155.
_ Asiltrk, H. Akkus, Determining the effect of cutting parameters on
[20] I.
surface roughness in hard turning using the Taguchi method,
Measurement 44 (2011) 16971704.
[21] T. Kvak, G. Samtas, A. iek, Taguchi method based optimisation of
drilling parameters in drilling of AISI 316 steel with PVD monolayer
and multilayer coated HSS drills, Measurement 45 (2012) 1547
1557.
[22] S. Kumar, P. Kumar, H.S. Shan, Optimization of tensile properties of
evaporative pattern casting process through Taguchis method,
Journal of Materials Processing Technology 204 (2008) 5969.
[23] S. Kasman, I.E. Saklakoglu, The effects of laser parameters on the
hardness of engraved surface, in: 14th International Conference on
Advances in Materials & Processing Technology (AMPT 2011), 2011.
[24] Y.C. Lin, A.C. Wang, D.A. Wang, C.C. Chen, Machining performance
and optimizing machining parameters of Al2O3TiC ceramics using
EDM based on the Taguchi method, Materials and Manufacturing
Processes 24 (2009) 667674.
[25] S. Neseli, S. Yaldz, E. Trkes, Optimization of tool geometry
parameters for turning operations based on the response surface
methodology, Measurement 44 (2011) 580587.

Você também pode gostar