Int J Adv Manuf Technol (2012) 58:81–91

DOI 10.1007/s00170-011-3392-5

ORIGINAL ARTICLE

An optimization method of the machining parameters

in high-speed machining of stainless steel using coated
carbide tool for best surface finish
Ahmad Hamdan & Ahmed A. D. Sarhan & Mohd Hamdi

Received: 30 November 2010 / Accepted: 16 May 2011 / Published online: 1 June 2011
# Springer-Verlag London Limited 2011

Abstract High-speed machining (HSM) has emerged as a most significant parameters affecting the cutting forces and
key technology in rapid tooling and manufacturing appli- surface roughness. For such application, several machining
cations. Compared with traditional machining, the cutting parameters are considered to be significantly affecting
speed, feed rate has been great progress, and the cutting cutting forces and surface roughness. These parameters
mechanism is not the same. HSM with coated carbide include the lubrication modes, feed rate, cutting speed, and
cutting tools used in high-speed, high temperature situa- depth of cut. Finally, conformation tests were carried out to
tions and cutting more efficient and provided a lower investigate the improvement of the optimization. The result
surface roughness. However, the demand for high quality showed a reduction of 25.5% in the cutting forces and
focuses extensive attention to the analysis and prediction of 41.3% improvement on the surface roughness performance.
surface roughness and cutting force as the level of surface
roughness and the cutting force partially determine the Keywords Taguchi . Optimization . High speed .
quality of the cutting process. This paper presents an Machining . Lubricant . Cutting . Force . Roughness
optimization method of the machining parameters in high-
speed machining of stainless steel using coated carbide tool
to achieve minimum cutting forces and better surface 1 Introduction
roughness. Taguchi optimization method is the most
effective method to optimize the machining parameters, in High-speed machining (HSM) is gaining popularity in
which a response variable can be identified. The standard industry in recent years due to the capability in improving
orthogonal array of L9 (34) was employed in this research machining performance, reducing cost while achieving
work and the results were analyzed for the optimization reduced lead times, and higher productivity [1]. However,
process using signal to noise (S/N) ratio response analysis the demand for high quality focuses attention on the surface
and Pareto analysis of variance (ANOVA) to identify the condition and the quality of the product, especially the
roughness of the machined surface because of its effect on
product appearance, function, and reliability. In addition, a
A. Hamdan : A. A. D. Sarhan (*) : M. Hamdi
good quality machined surface significantly improves
Department of Engineering Design and Manufacture,
University of Malaya, fatigue strength, corrosion resistance, and creep life.
Faculty of Engineering Building, Besides that, the quality of the machined surface is useful
50603 Kuala Lumpur, Malaysia in diagnosing the stability of the machining process, where
e-mail: [email protected]
a deteriorating surface finish may indicate workpiece
A. A. D. Sarhan material non-homogeneity, progressive tool wear, cutting
e-mail: [email protected] tool chatter, etc. For these reasons, it is important to
A. Hamdan maintain consistent tolerances and surface roughness.
e-mail: [email protected] Surface roughness is defined as a group of irregular waves
M. Hamdi in the surface, measured in micrometers (μm). The
e-mail: [email protected] roughness data obtained by measurement can be manipu-
82 Int J Adv Manuf Technol (2012) 58:81–91

lated to determine the roughness parameter. There are many review for cutting temperature prediction and measurement
different roughness parameters in use, but Ra is by far the methods. Additionally, many experimental methods to
most common. Other common parameters include Rz, Rq, measure temperature directly, only a few systems have as
and Rsk. yet been used this temperature as an indicator for machine
Surface roughness is mainly affected by different performance monitoring and for industrial application.
machining parameters that can be setup in advance, such Therefore, design and develop control system to control the
as rotation speed, feed rate, and cut depth. However, it is temperature lead to better surface finish, as machine
also affected by other uncontrol variables such as the performance parameter is required [14]. Thus, the imple-
mechanical properties of the material, the type of the cutter, mentation of cutting fluid, which acts as a lubricant and a
and the vibration produced during the process. The cutting coolant, are very crucial. However, the usage of conventional
speed and feed rate are significant machining parameters flooding application in machining processes is becoming
affecting surface roughness; however, the effect of depth of debatable because machine operators could be exposed to
cut is small [2–8]. The use of higher cutting speed and possible health problems [9, 15]. In addition, the bulk of
lower feed rate produced a better surface finish and this is waste from the conventional flooding technique might be
mainly attributed to the high temperature [8, 9]. However, disposed off irresponsibly thus causing serious damage to the
with higher cutting speed and temperature, special rapid environment. Economically, the cost related to the cutting
tooling is needed to increase abrasion resistance and hence fluid is 7–17% of total production cost which is normally
produced good surface roughness. For rapid tooling higher than that of cutting tool equipments which incurs only
applications in the die and mold industry, it will lead to 2–4% of total cost [11].
great improvements if tool can be used for machining with An alternative has been suggested in which a high
shorter lead times and better surface finish. The achieve- pressure jet of soluble oil was applied directly to the chip-
ment of these objectives by HSM puts it at the priority of tool interface. This has been reported to reduce cutting
rapid tooling and manufacturing technologies. Whereas, temperature and improve tool life due to its ability to
tool life is important for cost purposes, the surface finish is penetrate into the chip–tool interface [9]. Klocke and
consider a direct measure of the quality of product Eisennblatter [16] state that minimal quantity lubrication
produced. Coating tools are traditionally expected to play (MQL) refers to the use of cutting fluids of only a minute
multiple roles such as reducing cutting temperatures and amount—typically of a flow rate of 50–500 ml/h which is
cutting forces and increasing abrasion resistance [10, 11]. It about three to four orders of magnitude lower than the
is clear that the coated cutting tool provided a lower surface amount commonly used in flood cooling condition. In
roughness; however, the benefits will depend on many addition, it is also reported that MQL increases the
factors, including substrate material, tool coating combina- machining processes efficiency by improving its overall
tions, and the thermophysical conditions of both tool and performance compared to dry and conventional flood
workpiece. The application of coated carbide tools has been machining [3, 9, 17, 18]. Besides that, study on machining
found to bring about an expansion of the region of austenitic stainless steels indicated that high-pressure MQL
machining conditions within which reduced wear rates give better performance in drilling and turning [19].
may be experienced [7]. The effect of tool coatings is thus However, the investigations of milling process with MQL
expected to be the broadening of the lowest wear zone on especially in high-speed milling of austenitic stainless steels
the feed-cutting speed wear map. In addition, coatings are are still at an early stage and many areas of research is yet
expected to result in lower wear regions occurring at higher to be explored. This is because of the high ductility of
spindle speeds, promoting the use of HSM. stainless steel that categorizes it to be amongst the difficult
Although, coating tools are playing a key role to reduce the to cut materials [2, 20, 21].
cutting temperatures, it is however still high as the friction Following the review above, this study included lubri-
between cutting tool and workpiece is also high due to cation mode, feed rate, spindle speed, and dept of cut as
extensive contact during machining. The cutting temperature machining parameters (called control factor hear after)
is a key factor, which directly affects cutting tool wear, affecting the surface roughness. The main objective is to
workpiece surface integrity, and machining precision accord- find the best combination of theses parameters in high-
ing to the relative motion between the tool and work piece speed machining of stainless steel using coated carbide tool
[12]. The amount of heat generated varies with the type of to achieve low cutting force and surface roughness. The
material being machined and machining parameters espe- conventional method to achieve that is to use the “trial and
cially cutting speed, which had the most influence on the error” approach. However, due to the large number of
temperature [13]. Several attempts have been made to predict experiments, the “trial and error” approach is very time
the temperatures involved in the process as a function of consuming. Hence, a reliable systematic approach for
many parameters. Da Silva and Wallbank [14] presented a optimizing the machining parameters is thus required.
Int J Adv Manuf Technol (2012) 58:81–91 83

2 The optimization method 3.1 Identifying the OA

The optimization method presented in this study is an The standardized orthogonal array L9 (34) was used in this
experimental process called the Taguchi optimization method. experiment. It consists of nine experiments with four
Taguchi optimization, developed by Dr. Genichi Taguchi, is a different factors and three different experimental condition
set of methodologies in which the inherent variability of levels. The factors and levels are specified in Table 1. The
materials and manufacturing processes has been taken into feeds and depths of cut were selected based on recom-
account at the design stage. In Taguchi optimization, the mendations given by the insert manufacturer’s recommen-
multiple factors can be considered at once. Moreover, it seeks dation shown in Appendix A. Cutting speeds were chosen
nominal design points that are insensitive to variations in three times above the recommended speeds in order to
production and user environments to improve the yield in move the experiment into the “high cutting speed” category
manufacturing and the reliability in performance of a product [1, 21]. The highest value of cutting speed was selected
[22]. Therefore, not only can control factors be considered, after preliminary test. Total damage is occurred to the tool
but also uncontrol factors (noise factors) that cause the insert beyond the 400 m/min. The nine experiments with
functional characteristics of a product to deviate from their the details of the combination of the experimental condition
target values. By using the Taguchi optimization techniques, levels for each control factor (A–D) are shown in Table 2.
industries are able to greatly reduce product development
cycle time for both design and production, therefore reducing 3.2 Experimental details
costs and increasing profit.
The complete procedure in Taguchi optimization method After the orthogonal array has been selected, the second
is consisting of three stages: system design, parameter step in Taguchi optimization method is running the
design, and tolerance design. However, the parameter experiment based on that OA. The experimental setup used
design is the most important stage [22, 23]. It has been in this research is shown in Fig. 1. The machine used was a
widely applied in the USA and Japan with great success for vertical type-machining center (Mitsui Seiki VT3A). The
optimizing industrial/production processes. The most im- spindle has constant position preload bearings with oil-air
portant step in the parameter design is the selection of lubrication and maximum rotational speed of 20,000 min−1.
control factors. As many factors as possible should be The spindle motor power is 15 kW and maximum cutting
included, so that it would be possible to identify non- feed used is 10 m/min. The KORLOY carbide insert with
significant variables at the earliest opportunity. The steps an ISO designation of APXT 11T3PDSR-MM (coated with
included in the parameter design of an experiment are: titanium aluminum nitride) is shown in Fig. 2. The tool was
selecting the orthogonal array (OA) according to the mounted on a 12 mm diameter KORLOY Alpha mill insert
numbers of controllable factors, running experiments based holder with a code of AMS2012S. The insert holder was
on the OA, analyzing data, identifying the optimum firmly mounted on the BT 40 tool holder with a code CTR
condition, and conducting confirmation runs with the 32–150 KURODA, which is suitable for high-speed milling
optimal levels of all the parameters. The details regarding process. The cutter is a single-flute type that has chip
these steps will be described in the following sections. breaker grooves and is suitable for slot milling of ductile
In this research, the Taguchi optimization method is applied material.
to identify the optimal control factors combination that yield To investigate the machining performances of the cutting
the lowest cutting force and surface roughness value (Ra). To process, the slot-milling test was carried out as a case study
achieve this goal, the following items will be investigated: by using the proposed experimental setup. A cutting
process of a rectangular workpiece of stainless steel AISI
& The relationship between the controllable factors (lubri-
304 (85.1 HRB) prepared in 50×50×220 mm is selected as
cation mode, feed rate, spindle speed, and dept of cut) and
the response factor (surface roughness and cutting force)
& What are the optimal conditions of the controllable Table 1 Factors and levels used in experiment
factors for best surface roughness?
Factor Experimental condition levels

1 2 3
3 Design of experiment
A. Cutting speed (m/min) 300 350 400
B. Feed rate (mm/tooth) 0.05 0.1 0.15
The most important stage in the design of experiment lies in
C. Axial depth of cut (mm) 0.5 0.75 1
the selection of control factors and identifying the OA then
D. Lubrication mode Dry Flood MQL
running the experiment based on that particular OA.
84 Int J Adv Manuf Technol (2012) 58:81–91

Table 2 Measured parameters and calculated S/N ratio values (completed orthogonal array)

Tests Level of input Measured parameters Calculated S/N ratio

parameter
Surface roughness, Ra (μm) Resultant cutting force, N

Ai Bi Ci Di First Second Third Average First Second Third Average Surface Cutting
reading reading reading reading reading reading roughness force

1 1 1 1 1 0.29 0.61 1.47 0.79 40.00 35.45 41.45 38.97 0.60 −31.83
2 1 2 2 2 3.24 1.82 2.67 2.58 100.95 80.56 82.98 88.16 −8.43 −38.95
3 1 3 3 3 1.63 2.76 2.17 2.18 101.19 86.34 100.44 95.99 −6.98 −39.67
4 2 1 2 3 0.44 0.62 1.20 0.75 31.06 30.44 52.44 37.98 1.73 −31.90
5 2 2 3 1 1.43 1.69 3.14 2.09 89.11 90.08 94.30 91.16 −6.92 −39.20
6 2 3 1 2 1.84 0.54 2.00 1.46 46.77 60.12 62.60 56.50 −4.10 −35.11
7 3 1 3 2 0.47 1.53 1.27 1.09 32.75 29.12 29.88 30.58 −1.45 −29.71
8 3 2 1 3 1.57 1.95 2.28 1.94 39.94 38.82 42.41 40.39 −5.83 −32.13
9 3 3 2 1 3.55 3.01 2.74 3.10 83.82 90.28 70.28 81.46 −9.88 −38.26

a case study. The tool moves in +X direction to cut a stroke are dry, flooded, and MQL. The ECOCUT SSN 322 neat
of 220 mm and each test was repeated three times in order lubricant oil type with 40.2 cSt at 40°C from FUCHS was
to have accurate readings of the cutting force and surface used in MQL lubrication mode. Whereas, the ECOCOOL
roughness. The cutting forces were measured using a 6210 IT from FUCHS which contains a water mixture of
Kistler three-axis dynamometer (type 9255B). The mea- volumetric concentration 1:10 is use for flood lubrication
sured cutting force signals (X, Y, and Z direction) were mode. This fluid is free from phenol, chlorine, and other
captured and filtered with low-path filters (10 Hz cut off additive. While in the dry cutting-type application, no
frequency). The sampling frequency was set so that 40 lubricant is used. In the MQL type application, the
points per one spindle revolution can be obtained. To experiments were conducted using a thin-pulsed jet
measure the machined workpiece surface roughness Ra, a nozzle that was developed in laboratory and controlled
portable profilometer M1 perthometer (Mahr, Germany) is by a variable speed control drive. The nozzle is attached
used with a cut distance of 5.6 mm. to a portable fixture at the machining center spindle. The
To control the temperature during cutting for better flexible design allowed the injection nozzle to be located
surface finish, the lubrication system is applied. The three at any desired position without interfering with the tool
different types of lubrication modes used in this research or workpiece during the machining process. The diameter

Fig. 1 Experimental setup

Int J Adv Manuf Technol (2012) 58:81–91 85

APXT 11T3PDSR-MM geometry (mm) Table 2. Figure 3 shows an example of measured cutting
l d t r d1 forces in X-, Y-, and Z-axis direction at MQL lubrication
11.3 6.60 3.6 0.5 2.85 mode, 350 m/min cutting speed, 0.05 mm/tooth feed rate,
and 0.75 mm axial depth of cut. While Fig. 4 shows an
example of measured surface roughness at dry lubrication
mode, 300 m/min cutting speed, 0.05 mm/tooth feed rate,
and 0.5 mm depth of cut.

4.2 The analyze of the results

The procedure after the experimental runs is to analyze the

results to optimize the parameters and to identify which
process parameters are statistically significant. Analyzing
Fig. 2 Geometries of KORLOYS carbide insert data was conducted using S/N ratio response analysis and
Pareto ANOVA.

of nozzle orifice is 1 mm and the MQL are set to pulse a 4.2.1 Signal to noise (S/N) response analysis
rate of 400 pulse/min at 20 MPa with delivery rate of
2 ml/min. The direction of applying fluid nozzle in MQL Taguchi used the S/N ratio as the quality characteristic of
system is set opposite to the feed direction. However, in choice to analyze the data [24, 25]. S/N ratio is used as a
the flood type application, the flood fluid was delivered measurable value instead of standard deviation due to the fact
through three nozzles around the tool at the rate of that as the mean decreases, the standard deviation also
7,000 ml/min. decreases and vice versa. The methods for calculating the S/
N ratio are classified into three main categories, depending on
whether the desired quality characteristics are smaller the
4 Results, analysis, and discussion better, larger the better or nominal the better. In the case of
surface roughness, the smaller values are always preferred.
4.1 The experimental results The equation for calculating the S/N ratio for smaller the better
characteristics (in decibel) is as follows [24]:
The slot-milling test was carried out as a case study by
using the proposed experimental setup. The measured S 1 X 2 
¼ 10 log yi ð1Þ
surface roughness and cutting forces are summarized in N n

Fig. 3 An example of cutting

force measurement at MQL
cutting mode, 350 m/min
cutting speed, 0.05 mm/tooth
feed rate, and 0.75 mm
depth of cut
86 Int J Adv Manuf Technol (2012) 58:81–91

Fig. 4 An example of measured surface roughness at dry cutting, 300 m/min cutting speed, 0.05 mm/tooth feed rate, and 0.5 mm depth of cut

Where, yi is the individual measured surface roughness more significant followed by the cutting speed (factor A)
or cutting force in first, second, and third columns in and the axial depth of cut (factor C). The highest cutting
Table 2, and n is the number of the individual measured. In speed (A3, 400 m/min), with the lowest feed rate (B1,
this case, n=3. The calculated S/N ratios are summarized in 0.05 mm/tooth), lowest axial depth of cut (C1, 0.5 mm),
Table 2. The S/N ratio functioned as a performance and the MQL lubrication mode (D3) are determined to be
measurement to develop processes insensitive to noise the best choices for obtaining the lower cutting force.
factors. The degree of predictable performance of a product Therefore, the optimal parameters combination is set as A3
or process in the presence of noise factors could be defined B1 C1 D3.
from S/N ratio values. For each type of the characteristics, For the surface roughness, Fig. 6 shows that the feed rate
the higher the S/N ratio the better is the result. (factor B) is the prominent factor followed by the cutting
Furthermore, the S/N response data of the cutting speed (factor A), the axial depth of cut (factor C), and
force and surface roughness are calculated and summa- cutting mode (factor D). Consequently, the intermediate
rized in Tables 3 and 4, respectively. As for an example of value of cutting speed (A2, 350 m/min), the lowest value of
S/N response calculation, Ai is the summation of all the S/ the feed rate (B1, 0.05 mm/tooth), the lowest value of the
N ratio values of surface roughness or cutting force depth of cut (C1, 0.5 mm) and MQL as lubrication mode is
corresponding to the same level of input parameters (i) suggested as optimal parameters in order to obtain better
under A in Table 2. In this case, (i) is equal to 1, 2, or 3. surface roughness (A2 B1 C1 D3).
Similarly, the S/N ratio response values of the cutting
force and surface roughness are calculated for Bi, Ci, and 4.2.2 Pareto ANOVA: an alternative analysis
Di .
The desired “smaller the better” criteria implies that the Pareto ANOVA is an alternative method used to analyze the
lowest surface roughness would be the ideal result, while data for process optimization. It exhibits the percentage of
the largest S/N ratio response would reflect the best factor influence for each parameter in a very simple way
response which results in the lowest noise. This is the [19, 24]. Pareto ANOVA for resultant cutting force and
criteria employed in this study to determine the optimal surface roughness are constructed in Tables 5 and 6,
machining parameters. The S/N ratio response graphs of the respectively, by using the S/N ratio response data from
resultant cutting force and surface roughness for selecting Tables 3 and 4. The summation of squares of differences
the best combination levels for minimum surface roughness (S) for each control factor is calculated such that, for
are shown in Figs. 5 and 6. example, Sa can be obtained by the following equation:
As seen in Fig. 5, and based on the higher the S/N ratio
the better is the result, the feed rate (factor B) is found to be Sa ¼ ðA1  A2 Þ2 þ ðA1  A3 Þ2 þ ðA2  A3 Þ2 ð2Þ

Table 3 S/N ratio response data

of resultant cutting force Factor and interaction Levels (i) Ai Bi Ci Di

Summation at the level of input parameter 1 −110.45 −93.44 −99.07 −109.30

2 −106.20 −110.28 −109.11 −103.77
3 −100.11 −113.04 −108.58 −103.69
Int J Adv Manuf Technol (2012) 58:81–91 87

Table 4 S/N ratio response data

of surface roughness Factor and interaction Levels (i) Ai Bi Ci Di

Summation at the level of input parameter 1 −14.81 0.88 −9.33 −16.19

2 −9.28 −21.18 −16.58 −13.98
3 −17.16 −20.95 −15.35 −11.07

Sb, Sc, and Sd are similarly calculated. The contribution 4.3 Discussion
ratio for each factor is obtained as the percentage of the
summation of squares of differences for each factor to In this study, two techniques of data analysis on the
the total of summation of squares of differences. The results of the cutting force and surface roughness have
Pareto diagram is plotted organizing the factors in order been used. Both techniques delivered similar results. The
such that the one has higher contribution comes first feed rate is found to be the most prominent factor
followed by the others. The significant factors or influencing the cutting force and surface roughness for
parameters are chosen from the left-hand side of the high-speed milling followed by depth of cut. As can be
Pareto diagram, which cumulatively contribute up to seen from Figs. 5 and 6, the cutting forces increase with
90%. In Table 5, the feed rate (factor B, percentage of increasing feed rate and depth of cut, while the lower
61.90%), axial depth of cut (factor C, percentage of values of the feed rate and depth of cut is recommended
17.55%), and cutting speed (factor A, percentage of for better surface roughness. That is because the combi-
14.87%) give prominent effect on the cutting force. nation of feed rate and depth of cut determines the
However, the lubrication mode (factor D, percentage of undeformed chip section and hence the amount of the
5.68%) was found to be statistically insignificant. The best cutting forces required to remove a specified volume of
combination for lower resultant cutting force is A3 B1 C1 material. By decreasing feed rate and depth of cut, less
D3. This result is found to be similar to the result obtained material has to be cut per tooth per revolution, thus energy
by S/N ratio response analysis. required is lower. Consequently, this would reduce the
In Table 6, the surface roughness data is similarly cutting forces, consequently, preserve material properties
calculated. It is suggested that the feed rate (factor B, against residual stress, change in micro hardness at the
percentage of 80.85%) is the prominent factor affecting the subsurface, and leading to better surface roughness. This
surface roughness followed by cutting speed (factor A, would explain why feed rate and depth of cut play a key
percentage of 8.24%) and axial depth of cut (factor C, role in obtaining lowest cutting force and better surface
percentage of 7.59%). Similarly, the lubrication mode roughness [19, 26].
(factor D, percentage of 3.32%) was found to be statisti- Furthermore, in machining of the ductile material like the
cally insignificant. The recommended optimal parameter steel AISI 304, the soft ferrite phase causes long tool–chip
combination for better surface roughness is A2 B1 C1 D3. contact length on the rake face. Increasing tool–chip contact
This result is also found to be similar to the results obtained length reduces the shear plane angle and this, in turn, increases
from S/N ratio response analysis. shear stress. In addition, cutting forces also increase due to the

-70 5
Cutting speed
Cutting speed
0 Feed rate
-80
Feed rate
Axial depth of cut
S/N ratio (dB)

S/N ratio (dB)

Axial depth of cut -5

-90 Lubrication mode
Lubrication mode
-10
-100
-15

-110
-20

-120 -25
A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3 A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3
Cutting parameter levels Cutting parameter levels

Fig. 5 S/N ratio response graph of resultant cutting force Fig. 6 S/N ratio response graph of surface roughness
88 Int J Adv Manuf Technol (2012) 58:81–91

Table 5 Pareto ANOVA analysis for cutting force

long tool–chip contact length as a result of increased friction implementation of cutting fluid, which acts as a coolant,
forces. Heat is generated as well at several points during is very crucial. In addition, it is not only important as a
machining of ductile materials. Firstly, at the primary deforma- coolant but also during cutting, the friction coefficient is
tion zone due to shear and plastic deformation. Secondly, at the reduced and the lubrication features enhanced, thus
chip–tool interface due to secondary deformation and sliding. reduce the friction more and consequentially improve
Lastly, at the work–tool interfaces due to rubbing [27]. The the surface roughness of the machined workpiece. The
amount of heat generated varies with the type of material MQL lubrication mode is found to be the most efficient
being machined and cutting parameters especially cutting lubrication system to achieve lower cutting force and
speed, which had the most influence on the temperature [13]. better surface roughness as can be seen in Figs. 5 and 6.
Hence, the higher cutting temperature associated with the high The neat cutting oil in MQL has a better lubrication
spindle speed increases softening of the workpiece material property than the soluble oil in conventional flooding and
and then reduces the cutting forces leading to better surface obviously better than dry cutting.
roughness. In addition, at higher spindle speed, the chip will
break away with less material deformation at the immediate
tool tip, which in turn preserves the machined surface 5 Confirmation test
properties leading to better surface roughness. However, it is
believed that the spindle speed should be controlled at an After the optimal levels of all control factors are identified,
optimum value, as the influence of the high temperature, the last step in the Taguchi optimization method is to
would significantly affect the chip formation mode, cutting conduct a confirmation test using the optimal parameters
forces, tool life, and surface roughness [26]. Thus the A2 B1 C1 D3 to validate the recommendation. The
Int J Adv Manuf Technol (2012) 58:81–91 89

Table 6 Pareto ANOVA analyses for surface roughness

measured cutting force and surface roughness were techniques, the S/N ratio response and Pareto ANOVA
reported as the average of three measurements. The response analysis. However, both techniques delivered
cutting force and surface roughness obtained were found similar results. The feed rate is found to be more
to be 22.78 N and 0.44 μm, respectively. The result significant followed by the cutting speed and the depth
shows an improvement of 25.5% and 41.3% compared of cut, while, the lubrication mode was found to be
with the lowest average values obtained during the statistically insignificant. Finally, conformation tests with
experiments of the cutting force and surface roughness the optimal levels of all the parameters were carried out
shown in Table 2. to investigate the improvement of the optimization. The
result showed a reduction of 25.5% in the cutting forces
and 41.3% improvement on the surface roughness
performance. This was accomplished with a relatively
6 Conclusion small number of experimental runs, suggesting that
Taguchi optimization method is an efficient and effective
In this study, the machining parameters in high-speed method for optimizing surface roughness in high-speed
machining of stainless steel were optimized for best surface machining of stainless steel using coated carbide tool.
finish using Taguchi optimization method. With the L9 (34)
orthogonal array, a total of nine experiments are carried out
Acknowledgments This research has been supported by the short-term
with four different factors and three different experimental research grant, University of Malaya. The authors would like to express
condition levels. The results are analyzed in two deferent our sincere appreciation for the continuing support on our research work.
90 Int J Adv Manuf Technol (2012) 58:81–91

Appendix A

Table 7 Recommended cutting conditions (for single-edge type)

Workpiece Grade Cutting diameter

material
∅10, ∅15 ∅16, ∅25 ∅32, ∅63 ∅80, ∅100

V (m/min) fz (mm/tooth) V (m/min) tz (mm/tooth) V (m/min) fz (mmf/tooth) V (m/min) fz (mm/tooth)

Low carbon NCM325 60–120 0.05–0.15 80–180 0.1–0.15 120–250 01–0.15 150–250 0.1–0.15
steels, PC3535
soft steels
High carbon NCM325 50–100 0.05–0.15 80–150 0.1–0.15 100–200 0.1–0.15 100–200 0.1–0.15
steels, PC3535
alloy steels
Stainless NCM335 50–100 0.05–0.15 80–150 0.1–0.15 100–180 0.1–0.15 100–180. 0.1–0.15
steels PC9530
Cast irons NCM320K 80–120 0.08–0.2 100–130 0.15–0.2 120–200 0.15–0.2 120–200 0.15–0.2
PC215K
Aluminum H0l 250–900 0.15–0.3 300–950 0.15–0.3 400–1,000 0.1–0.4 400–1,000 0.1–0.4
alloy
Standard for side cutting, reduce condition 20–30% for slotting

Source: Korloy catalog book version June 2006, page 311

References 10. Sayit E, Aslantas K, Çiçek A (2009) Tool wear mechanism in

interrupted cutting conditions. Mater Manuf Processes 24(4):476–
483, 1532–2475
1. Rama Kotaiah K, Srinivas J, Babu KJ, Srinivas K (2010) 11. Klocke F, Einesblatter G (1997) Dry cutting. Annals CIRP
Prediction of optimal cutting states during inward turning: an 46:519–526
experimental approach. Mater Manuf Processes 25(6):432–441, 12. Ming C, Fanghong S, Haili W, Renwei Y, Zhenghong Q, Shuqiao
1532–2475 Z (2003) Experimental research on the dynamic characteristics of
2. Anthony Xavior M, Adithan M (2009) Determining the influence the cutting temperature in the process of high-speed milling. J
of cutting fluids on tool wear and surface roughness during Mater Process Technol 138:468–471
turning of AISI 304 austenitic stainless steel. J Mater Process 13. Liu CS, Zhao B, Gao GF, Jiao F (2002) Research on the
Technol 209:900–909 characteristics of the cutting force in the vibration cutting of a
3. Dhar NR, Kamruzzaman M, Ahmed M (2006) Effect of particle-reinforced metal matrix composites SiCp/Al. J Mater
minimum quantity lubrication (MQL) on tool wear and surface Process Technol 129:196–199
roughness in turning AISI-4340 steel. J Mater Process Technol 14. Da Silva MB, Wallbank J (1999) Cutting temperature: prediction
172:299–304 and measurement methods—a review. J Mater Process Technol
4. Ezugwu EO, Bonney J, Fadare DA, Sales WF (2005) Machining 88:195–202
of nickel-base, Inconel 718, alloy with ceramic tools under 15. Kundrák J, Mamalis AG, Gyáni K, Markopoulos A (2006)
finishing conditions with various coolant supply pressures. J Environmentally friendly precision machining. Mater Manuf
Mater Process Technol 162–163:609–614 Processes 21(1):29–37, 1532–2475
5. Isik Y (2007) Investigating the machinability of tool steels in 16. Kurniawan D, Yusof NM, Sharif S (2010) Hard machining of
turning operations. Mater Des 28:1417–1424 stainless steel using wiper coated carbide: tool life and surface
6. Teicher U, Künanz K, Ghosh A, Chattopadhyay AB (2008) integrity. Mater Manuf Processes 25(6):370–377, 1532–2475
Performance of diamond and CBN single-layered grinding wheels 17. Paulo Davim J, Sreejith PS, Silva J (2007) Turning of brasses
in grinding titanium. Mater Manuf Processes 23(3):224–227, using minimum quantity of lubricant (MQL) and flooded
1532–2475 lubricant conditions. Mater Manuf Processes 22(1):45–50,
7. Noordin MY, Venkatesh VC, Chan CL, Abdullah A (2001) 1532–2475
Performance evaluation of cemented carbide tools in turning AISI 18. Varadarajan AS, Philip PK, Ramamoorthy B (2002) Investigation
1010 steel. J Mater Process Technol 116:16–21 on hard turning with minimal cutting fluid application (HTMF)
8. Zafer T, Sezgin Y (2004) Investigation of the cutting parameters and its comparison with dry and wet turning. Int J Mach Tools
depending on process sound during turning of AISI 304 austenitic Manuf 42:193–200
stainless steel. Mater Des 25:507–513 19. Ghani JA, Choudhury IA, Hassan HH (2004) Application of
9. Thepsonthi T, Hamdi M, Mitsui K (2009) Investigation into Taguchi method in the optimization of end milling. J Mater
minimal-cutting-fluid application in high-speed milling of hard- Process Technol 145:84–92
ened steel using carbide mills. Int J Mach Tools Manuf 49:156– 20. Abou-El-Hossein KA (2008) Cutting fluid efficiency in end
162 milling of AISI stainless steel. Ind Lubric Tribol 60:115–120
Int J Adv Manuf Technol (2012) 58:81–91 91

21. Gandarias A, de Luis Norberto Lopez L, Aizpitarte X, Lamikiz A (2008) 25. Phadke MS (1989) Quality engineering using robust design.
Study of the performance of the turning and drilling of austenitic stainless Prentice Hall, Englewood Cliffs, NJ
steels using two coolant techniques. Int J Mach Mach Mater 3:1–17 26. Suresh Kumar Reddy N, Venkateswara Rao P (2005) Performance
22. Taguchi G (1990) Introduction to quality engineering. Asian improvement of end milling using graphite as a solid lubricant.
Productivity Organization, Tokyo Mater Manuf Processes 20(4):673–686, 1532–2475
23. Zhang JZ, Chen JC, Kirby ED (2007) Surface roughness 27. Khan MMA, Dhar NR (2006) Performance evaluation of
optimization in an end-milling operation using the Taguchi design minimum quantity lubrication by vegetable oil in terms of cutting
method. J Mater Process Technol 184:233–239 force, cutting zone temperature, tool wear, job dimension and
24. Park SH (1996) Robust design and analysis for quality engineering. surface finish in turning AISI-1060 steel. J Zhejiang University
Chapman & Hall, London 7:1790–1799