IJSRD - International Journal for Scientific Research & Development| Vol.

2, Issue 01, 2014 | ISSN (online): 2321-0613

Optimization of Machining Parameters in CNC Turning Using Firefly

Algorithm
Dr. S. Bharathi Raja1 C. Prem Kumar2 G. Gokul Rajasekar3 R. P. Dinesh Prasad4
1
Principal
1,2,3,4
Department Of Mechanical Engineering
1
Indra Ganesan College of Engineering, Tiruchirapalli, Tamil Nadu. 2, 3,4Parisutham Institute of Technology
& Science, Thanjavur.
Abstract now a days machining is done through various
automated machines. One of the widely used machines is
CNC. Even though the automated machines are used, the
quality of the work is determined by parameters used. The
cutting parameters of the CNC machine determine the
productivity, surface finish, machining time and other
qualities of the product. This study involves the optimization
of those cutting parameters. In order to get optimized CNC
parameters for a specific tool-work combination, Firefly
Algorithm (FA) is used to compute the best parameters
based on the experiments conducted on a CNC Turning
center. The three cutting parameters are cutting speed (V),
feed (f), and depth of cut (d). The practical constraints have
been considered during both practical and experimental
approaches. The result reveals that FA is very well suited in
solving parameters selection problems.

cutting force and power consumption models are dependent

on speed, feed and depth of cut. C.Y.Nian, W.H.Yang and
Y.S.Tarng [2] employed orthogonal array, multi-response
signal-to-noise ratio and ANOVA to study the performance
characteristics in turning operations. Experimental results
were provided to illustrate the effectiveness in using
Taguchi method. The author concluded that the tool life,
cutting force and surface roughness could be improved
simultaneously using the proposed approach instead of
engineering judgment.
C.F.Cheung and W.B.Lee [3]
established and evaluated a model-based simulation system
for analyzing the surface roughness in ultra-precision
diamond turning. The analysis includes the effects of
process parameters, tool geometry and relative vibration
between the tool and work piece. It was found that the
system can accurately predict the surface roughness under
various cutting conditions

I. INTRODUCTION

B.Y.Lee, Y.S.Tarng and H.R.Lii [4] described the

use of polynomial network to construct the machining
database in turning operation. The relationships between the
cutting parameters and the tool life, surface roughness and
cutting force were accurately correlated by a self-organizing
adaptive modeling technique. Experimental results showed
that the machining database has a high accuracy in the
prediction of cutting performance in turning operation. J
Paulo Davim and C.A.Conceicao Antonio [5] proposed GA
to select optimum cutting conditions in turning and drilling
aluminium matrix composites. The obtained results show
that machining was perfectly compatible with the cutting
conditions for cutting time of industrial interest and in
agreement with the optimal machining parameters.
Numerical and experimental models based on GA are a
matter of scientific interest and large industrial applications.
M.Y.Noordin, V.C.Venkatesh, S Sharif, S Elting and A
Abdullah [6] presented the findings of an experimental
investigation of the effect of feed rate, Side Cutting Edge
Angle (SCEA) and cutting speed on the surface roughness
and tangential force in turning steel. ANOVA revealed that
feed is the most significant factor influencing the surface
roughness followed by SCEA, while cutting speed provided
secondary contribution to the tangential force. The quadratic
model developed using RSM was reasonably accurate and
can be used for prediction of performance measures.
Wassilla Bouzid [7] developed empirical models for tool
life, surface roughness and cutting force for calculating
optimum cutting conditions in tuning to achieve maximum
production rate. The coefficients of these models were
determined based on the experiments. The author explained
the relation of feed to the roughness, which depends on
cutting speed and finally concluded that the proposed

Computer Numerical Control (CNC) is one of the most

commonly used machines. The machining is done by
entering the coded information and cutting parameters.
Selecting the optimized parameters to attain efficient
machining is very difficult. It has been proved that proper
selection of parameters decides the productivity rate and the
cost of production per component. The implementation of
CNC needs high initial investment, so it is must have better
usage of the machine to make a profit. The selection of
parameters must be within the range which are taken from
the cutting tool catalogue or hand books. After getting the
range for cutting parameters, the optimization of parameters
within the range is done through optimizing techniques. The
various optimization techniques available are Genetic
Algorithm (GA), Ant Colony Optimization (ACO), and
Particle Swarm Optimization (PSO). In this research work,
we have used recently developed algorithms called Firefly
Algorithm (FA) for the optimization problem. Our ultimate
aim is minimizing machining time and attaining better
surface finish. Here in this work, we fixed surface finish of a
product and we tried to reduce the machining time in order
to increase the production rate. The three parameters were
analyzed in various ways to determine the effect of one over
the other two.
II. LITERATURE REVIEW
M.S.Chua, M.Rahman, Y.S.Wang and H.T.Loh [1]
developed relations between the tool life, cutting force,
power consumption and the cutting conditions using
multiple regression analysis through factorial design of
experiments. Based on the analysis, it was found that the
tool life model is dependent on depth of cut while the

All rights reserved by www.ijsrd.com

56

Optimization Of Machining Parameters In CNC Turning Using Firefly Algorithm

(IJSRD/Vol. 2/Issue 01/2014/014)

method was capable of selecting the appropriate conditions.

Muthukrishnan and J Paulo Davim [8] studied the surface
roughness of composite bars under different cutting
conditions. Experimental data were tested with ANOVA and
ANN techniques. ANOVA revealed that feed rate has
highest influence (51%) on surface roughness followed by
depth of cut (30%) and cutting speed (12%). The results of
ANN model showed close matching between the model
output and the directly measured surface roughness. S
Bharathi Raja and N Baskar [9] examined SA, GA, and PSO
in three different mathematical models such as single pass
turning operation, multi-pass turning operation and grinding
operation. The authors found that PSO outperformed the
other optimization techniques in all the cases. Bharathi Raja
and Baskar [10] adopted PSO for optimize machining
parameters to get desired surface roughness in minimum
possible machining time. The authors claimed that average
deviation and accuracy rate of predicted surface roughness
to the actual surface roughness value is found to be 0.05
microns and 85% respectively. The average deviation and
accuracy rate of the predicted machining time to the actual
machining time is found to be 2 s and 96% respectively.
S.Bharathi Raja, C.V.Srinivas Pramod, K.Vamshee Krishna,
Arvind Ragunathan and Somalaraju Vinesh [11]
implemented Firefly Algorithm (FA) to explore optimum
Electric Discharge Machine (EDM) parameters to minimize
machining time for the desired surface finish. It is observed
that the proposed FA could predict the set objective function
considerably close to the experimental values. Bharathi Raja
and Baskar [12] conducted face milling experiments on
brass material to study the effect of cutting parameters on
machining time and surface roughness. PSO has been used
in the developed mathematical model for optimization.
Confirmatory experiments reveal that PSO could predict
optimum cutting parameters very close to the experimental
values. Xin-She Yang [13] developed and described a new
meta-heuristic optimization technique named as Firefly
Algorithm (FA). The results of the developed FA depicts
that it is very much suitable for any engineering
optimization as it outperformed the proven and most
efficient Particle Swarm Optimization (PSO) technique.

Fig. 2: CNC LL20T L5

The experiments are conducted for roughing, semifinishing and finishing operations and their corresponding
machining time are observed. Factorial experimentation is
adopted in selecting the combination of parameters, in
which all the parameters are tested at least once. Table 1-3
shows the machining time for roughing, semi-finishing and
finishing operations and their corresponding given cutting
parameters.
Table. 1: Experimental results of machining time for the
given parameters of roughing operation.

(mm/min)

Depth
of cut
(mm)

Machini
ng time
(sec)

900

0.16

0.8

316

900

0.18

0.8

281

900

0.22

0.8

233

900

0.24

0.8

203

1100

0.16

0.8

259

1100

0.18

0.8

232

1100

0.22

0.8

191

1100

0.24

0.8

174

III. PILOT EXPERIMENT

1500

0.16

0.8

196

In pilot experiments, LL-20-T CNC is used to machine SS

316 material using three different tools for three different
machining operations. Fig. 3.1 and Fig. 3.2 shows the raw
material and CNC machine used for pilot experiments. The
cutting parameters are selected based on the operating range
of the machine.

10

1500

0.18

0.8

174

11

1500

0.22

0.8

146

12

1500

0.24

0.8

133

13

1800

0.16

0.8

164

14

1800

0.18

0.8

147

15

1800

0.22

0.8

122

16

1800

0.24

0.8

113

Experiment
number

Speed

Feed

(rpm)

Table. 2: Experimental results of machining time for the

given parameters of semi-finishing operation.

Fig. 1: Raw material

(mm/min)

Depth
of cut
(mm)

Machin
ing
time
(sec)

0.09

0.2

84

Experiment
number

Speed

Feed

(rpm)

1400

All rights reserved by www.ijsrd.com

57

Optimization Of Machining Parameters In CNC Turning Using Firefly Algorithm

(IJSRD/Vol. 2/Issue 01/2014/014)

A. CNC Parameters
Optimization of cutting speed, feed and depth of cut was
selected because requirements in any change in the cutting
parameters do not involve time and cost. Among the three
parameters, two parameters V and f contribute more on
production time. Depth of cut (d) remains same for the
complete machining of single operation, but it varies
accordingly to the rough, semi-finish and finishing
operations. The cutting parameter limits are chosen based on
the operating range of work piece and machine capability.

1400

0.11

0.2

61

1400

0.15

0.2

53

1400

0.17

0.2

48

1600

0.09

0.2

75

1600

0.11

0.2

61

1600

0.15

0.2

47

1600

0.17

0.2

42

2000

0.09

0.2

62

10

2000

0.11

0.2

52

11

2000

0.15

0.2

40

12

2000

0.17

0.2

36

13

2200

0.09

0.2

57

14

2200

0.11

0.2

48

15

2200

0.15

0.2

37

Vmin V Vmax,

(1)

16

2200

0.17

0.2

34

fmin f fmax,

(2)

Table. 3: Experimental results of machining time for the

given parameters of finishing operation.

B. Constraints Considered
There are always some practical constraints that exist in
actual machining operations. Upper and lower limits of V
and f are considered as practical constraints. Equations for
the constraints considered are given in equations (1) and (2).
Here the depth of cut is kept at the desired value in order to
get the required dimension at the end.

C. Empirical relation
The empirical relation between the cutting parameters and
the machining time for considering three machining
operations are generated using MINITAB 16 software.
Empirical relations for roughing, semi-finishing and
finishing operations are given in equation (3), (4) and (5).

(mm/min)

Depth
of cut
(mm)

Machining
time (sec)

1400

0.06

0.2

170

tm= 1060.31 0.55V 3332.49f + 3333.33f + 0.79Vf

(3)

1400

0.08

0.2

96

tm = 261.11 0.07V 1686.50f + 4062.50f + 0.16Vf

(4)

tm = 737.8 0.4V 4846.2f + 12708.3f + 0.9Vf

(5)

Experiment
number

Speed

Feed

(rpm)

1
2
3

1400

0.12

0.2

66

1400

0.14

0.2

57

V. FIREFLY ALGORITHM

1600

0.06

0.2

111

1600

0.08

0.2

85

1600

0.12

0.2

58

1600

0.14

0.2

51

2000

0.06

0.2

91

10

2000

0.08

0.2

61

11

2000

0.12

0.2

49

12

2000

0.14

0.2

43

Firefly algorithm (FA) was developed by Xin-She Yang. FA

is based on the intensity of the light produced by the
fireflies. The light was produced by the mechanism called
bioluminescence. The light intensity is the key to selecting
the prey for sexual contact; it also acts as a shield from
enemies. The fireflies will keep moving until it identifies a
better light from a firefly to meet its objective. This concept
was used to develop in the proposed optimization
techniques. Optimization of input parameters will be iterated
until the best solution is attained. Programming for the
algorithm was done using JAVA.

13

2200

0.06

0.2

83

14

2200

0.08

0.2

64

15

2200

0.12

0.2

45

16

2200

0.14

0.2

40

A. Population Initiation
The value of V and f are calculated randomly using the
equations (6) and (7)
V = Vmin + (Vmax Vmin) rand(),

(6)

f =fmin +(fmax fmin) rand().

IV. FORMULATION OF EMPIRICAL RELATION
Experimental values from the pilot experiments are used to
find the effect of cutting parameters on machining time.
Moreover, an analysis was done to reveal the individual,
combined and interaction effect of the proposed cutting
parameters on the machining time. With the aid of the
analysis, empirical relation was formulated using Design
Expert software.

(7)

B. Distance
xi and xj are the points of presence of two fireflies i and j
respectively. The distance (rij) between these two fireflies
are represented in terms of Cartesian equation given in
equation (8), where k is the kth component of ith firefly and d
is the dimension
rij = ||xi - xj|| =

(8)

All rights reserved by www.ijsrd.com

58

Optimization Of Machining Parameters In CNC Turning Using Firefly Algorithm

(IJSRD/Vol. 2/Issue 01/2014/014)

C. Attractiveness equation
Firefly algorithm is based on the attractiveness of the
fireflies. The attractiveness function is given in equation (9).
In that r is the distance between two fireflies, 0 is the initial
attractiveness at r=0 and is a coefficient which controls the
lights intensity. M is considered as a higher light intensity
and it is vice versa to attractiveness. In this article, m is
taken as 1to solve the problem in worst case condition.
= 0.e-.rm,

(mm/min)

Depth of cut
(mm)

Machining
time
(seconds)

0.20

0.8

168

Speed

Feed

(rpm)
1342

Table. 8: Experimental result of machining time for semifinishing operation

(mm/min)

Depth of cut
(mm)

Machining
time
(seconds)

0.13

0.2

23

Speed

Feed

Where m 1

(rpm)

D. Movement
The movement of the firefly i towards j due to the
attractiveness through the brightness can be given as in
equation (10) and (11)

2180

(9)

inp= [icp(1- )] + [ib] +[(rand() 0.5)]

(10)

Similarly for the other firefly,

jnp = [ jcp(1- )] + [jb] +[(rand() 0.5)]

(11)

Based on the empirical relations formulated by Design

Expert software, optimization of cutting parameters to attain
minimum machining time is executed by the proposed FA.
1000 iterations with population as 100 are considered for
optimization by JAVA programming. After satisfying the
considered constraints, optimized cutting parameters are
obtained for the minimum possible machining time. The
results of optimization are given in table (4), (5) and (6).
Table 4: Optimized parameters and its corresponding
minimized machining time for roughing operation

(mm/min)

Depth of cut
(mm)

Machining
time
(seconds)

0.20

0.8

160

Speed

Feed

(rpm)
1342

Table. 9: Experimental result of machining time for

finishing operation

(mm/min)

Depth of cut
(mm)

Machining
time
(seconds)

0.09

0.2

52

Speed

Feed

(rpm)
1862

VII. RESULTS AND DISCUSSION

Results of pilot experiment have revealed the effect of
cutting parameters on machining time. Fig. 7.1 shows the
contour plot showing the combined effect of cutting speed
and feed on the machining time. As speed and feed increase
machining time decreases in which machining is more
sensitive to feed when compared to speed.

Table 5: Optimized parameters and its corresponding

minimized machining time for semi-finishing operation

(mm/min)

Depth of cut
(mm)

Machining
time
(seconds)

0.13

0.2

21

Speed

Feed

(rpm)
2180

Table 6: Optimized parameters and its corresponding

minimized machining time for finishing operation

(mm/min)

Depth of cut
(mm)

Machining
time
(seconds)

0.09

0.2

49

Speed

Feed

(rpm)
1862

Fig. 3: Contour plot for spindle speed (rpm) Vs feed rate

(mm) in roughing operation.
Fig. 3 shows the surface plot showing the
combined effect of cutting speed and feed on machining
time. The dominating effect of feed on machining time is
very well revealed in the plot.

VI. CONFIRMATORY EXPERIMENTS

Confirmatory experiments are conducted on CNC turning
center to validate the machining time generated by
optimized parameters of FA. The results of confirmatory
experiments are given in table (7), (8) and (9).
Table. 7: Experimental result of machining time for
roughing operation

All rights reserved by www.ijsrd.com

59

Optimization Of Machining Parameters In CNC Turning Using Firefly Algorithm

(IJSRD/Vol. 2/Issue 01/2014/014)

Fig. 4: Surface graph for spindle speed (rpm) Vs feed rate

(mm) in roughing operation.

in feed decrease machining time, but the finish of the

component comparatively low.

Similarly, the combined effect of speed and feed on

machining time in semi-finishing operation is shown in Fig.
5. The effect of feed is comparatively more than cutting
speed in machining time.

Fig. 8: Surface graph for spindle speed (rpm) Vs feed rate

(mm) in finishing operation

Fig. 5: contour graph for spindle speed (rpm) Vs feed rate

(mm) in semi-finishing operation
Fig. 6 shows the surface plot of the combined effect of
speed and feed on machining time in semi-finishing
operation.

From table (4), (5), (6) and table (7), (8),(9), it is

evident that the predicted machining time by FA is closer
agreement with the actual machining time obtained from
confirmatory experiments.
200

160 168

150
100

49

50

21

52

23

0
Roughing

Semi-Finishing

Finishing

Predicted Machining Time

Actual Machining Time
Fig. 9: Comparison of predicted and actual machining time
Fig. 6: Surface graph for spindle speed (rpm) Vs feed rate
(mm) in semi-finishing operation.
Fig. 7 shows the contour plot drawn to depict the combined
effect of speed and feed on machining time.

Fig. 7: on tour graph for spindle speed (rpm) Vs feed rate

(mm) in finishing operation
Fig. 8 shows the surface plot of combined effect of speed
and feed on machining time in finishing operation. Increase

VIII. CONCLUSION
In this research article, pilot experiments are conducted to
study the effect of cutting parameters on machining time for
a continuous finished profile. Further, predicted machining
time based on cutting parameter optimization by Firefly
algorithm was verified by experimental results. The
followings are the observations evolved from the present
research work:
1) The effect of feed on machining time was dominating
more than cutting speed.
2) Firefly algorithm could able to predict the machining
time for the proposed continuous profile in terms of
accuracy as follows:
Roughing: 95.2%
Semi-finishing: 91.3%
Finishing: 94.2%
3) Firefly algorithm has proved to be one of the best tools
for optimization for solving any engineering problems.
4) The present method of selection of cutting parameters
can very well be replaced by the proposed method of
selection of optimized cutting parameters by FA.
5) Since a real time job profile was undertaken, optimized
parameters by FA have increased the production rate

All rights reserved by www.ijsrd.com

60

Optimization Of Machining Parameters In CNC Turning Using Firefly Algorithm

(IJSRD/Vol. 2/Issue 01/2014/014)

when compared to the parameters utilized before

optimization.
6) Time and cost is not a major function in introducing
the proposed method of parameters selection.
7) Use of FA is completely generalized and it can be very
well used for solving any engineering problems.
8) FA can be interfaced with any CNC machine for online
selection of cutting parameters.
A. Scope for future work
In the present work, optimization was performed by
considering three parameters. But, the inclusion of other
machining parameters may enhance the required results.
Moreover, conflicting objectives such as surface roughness,
cutting force, tool life can be included as objectives or for
the prediction of the same.

Optimization of electrical discharge machining

parameters on hardened die steel using Firefly
algorithm, Engg with Comp, doi:10.1007/s00366013-0320-3, 2013.
[12] S.Bharathi Raja and N.Baskar, Computational
solution for multi-objective optimization in CNC
machining operation using Particle Swarm
Optimization technique, J Bio Info Int Control, vol.2,
pp.1-9, 2013.
[13] Xin She Yang, Firefly Algorithms for multimodal
optimization, Lecture Notes in Computer Science,
vol. 5792, pp.169-178, 2009.

REFERENCES
[1] M.S.Chua, M.Rahman, Y.S.Wang and H.T.Loh,
Determination of optimal cutting conditions using
design of experiment and optimization techniques,
Int J Mach Tool Manuf, vol.33, pp.297-305, 1993.
[2] C.Y.Nian, W.H.Yang and Y.S.Tarng, Optimization
of turning operations with multiple performance
characteristics, J Mater Process Technol, vol.95,
pp.90-96, 1999.
[3] C.F.Cheung and W.B.Lee, A theoretical and
experimental investigation of surface roughness
formation in ultra-precision diamond turning. Int J
Mach Tools Manuf, vol.40, pp.979-1002, 2000.
[4] B.Y.Lee, Y.S.Tarng and H.R.Lii, An investigation of
modelling of the machining database in turning
operation, J Mater Process Technol, vol.105, pp.1-6,
2000.
[5] J Paulo Davim and C.A.Conceicao Antonio,
Optimization of cutting conditions in machining of
aluminium matrix composites using a numerical and
experimental model J Mater Process Technol
vol.112, pp.78-82, 2001.
[6] M.Y.Noordin, V.C.Venkatesh, S Sharif, S Elting and
A Abdullah, Application of Response Surface
Methodology in describing the performance of coated
carbide tools when turning AISI 1045 steel, J Mater
Process Technol, vol.145, pp.46-58,2004.
[7] Wassilla Bouzid, Cutting parameter optimization to
minimize production time in high speed turning, J
Mater Process Technol, vol.16, pp.388-395, 2005.
[8] Muthukrishnan and J Paulo Davim, Optimization of
machining parameters of Al/SiC- MMC with
ANOVA and ANN analysis J Mater Process
Technol, vol.209, pp.225-232, 2009.
[9] S Bharathi Raja and N Baskar, Optimization
techniques for machining operations: A retrospective
research based on various mathematical models, Int
J Adv. Manuf Technol, vol.48, pp. 1075-1090, 2010.
[10] S Bharathi Raja and N Baskar, Application of
Particle Swarm Optimization technique for achieving
desired milled surface roughness in minimum
machining time, Exp Sys Appln, vol.39, pp.59825989, 2012.
[11] S.Bharathi Raja,C.V.Srinivas Pramod, K.Vamshee
Krishna, Arvind Ragunathan and Somalaraju Vinesh,

All rights reserved by www.ijsrd.com

61