EXPERIMENTAL ANALYSIS ON

SURFACE ROUGHNESS OF CNC

END MILLING PROCESS USING
TAGUCHI DESIGN METHOD

PATEL K. P.
B. S. Patel Polytechnic, Ganpat University, Kherva-382711, Mehsana, North Gujarat, India.
[email protected]


Abstract:

Here, we study about, influence of various machining parameters like tool speed, tool feed, depth of cut and
tool diameter. In the present study, experiments are conducted on AL 6351 T6 material with four factors
and five levels and try to find out optimum surface roughness by using taguchi method. This paper attempts
to introduce how Taguchi parameter design could be used in identifying the significant processing
parameters and optimizing the surface roughness of end-milling operations.
In this study, it was observed that, the order of signicance of the main variables is as A1 >B4 >C4 >D1
(tool feed (A), tool speed (B), tool diameter(C), depth of cut (D)).

Keywords: CNC end milling, S/N ratio, Taguchi method.

1. INTRODUCTION

Today, industry needs quality and productivity. The increase of consumer needs for quality metal cutting
products has driven the metal cutting industry to continuous improve quality control of metal cutting process.
The end milling process is one of the most fundamental process of metal removing process.In order to obtain
better surface roughness, the proper setting of cutting parameters is crucial before the process takes place.
Several factors will influence the final surface roughness in a CNC milling operation. The final surface
roughness might be considered as the sum of two independent effects: 1.The ideal surface roughness is a
result of the geometry of tool and feed rate.2.The natural surface roughness is a result of the irregularities in
the cutting operation Factors such as spindle speed, feed rate, tool diameter and depth of cut that control the
cutting operation can be setup in advance. However, factors such as too geometry, tool wear, chip loads and
chip formations, or the material properties of both tool and work piece are Even in the occurrence of chatter
orvibrations of the machine tool, defects in the structure of the work material, wear of tool, or irregularities of
chip formation.


2. EXPERIMENTAL WORK

The study was carried out using a BFW V-30 CNC vertical milling machine equipped with a maximum
spindlespeed 10000 rpm ,feed rate 20 m/min, multiple tool-change capabilities (max number of tools =25)
and with 15 HP spindle horsepower The machine is capable of a three-axis movement (along the x, y,and
zplanes). CNC programs can be developed in the hyper mill software. The work piece material used was AL
6351 T6 in the form of 420mm x 120mm x 20mm plate. The surface roughness measured by Mitutoyo-
surface roughness tester.
Experimental design, setup, and results are presented as follows:
2.1 Select the quality characteristics.
2.2 Select noise factor and control factor.
2.2 Select orthogonal array.
2.4 Conduct the experiments.
2.5 Analyze the result and determine optimum factor-level combination.
2.6 Predict optimum performance.

Patel K. P. / International Journal of Engineering Science and Technology (IJEST)
ISSN : 0975-5462 Vol. 4 No.02 February 2012 540

2.1 SELECT THE QUALITY CHARACTERISTICS

There are three types of quality characteristics in the Taguchi methodology, such as smaller-the-better,
larger-the-better, and nominal-the-best. The goal of this research was toproduce minimum surface roughness
(Ra) in an end-milling operation. Smaller Ra Values represent better or improved surface roughness.
Therefore, a smaller-the-better quality characteristic was implemented and introduced inthis study.

2.2 Select Noise Factor And Control Factor

Literature survey and discuss with some industrial person ,we conclude that tool feed(A), tool speed(B), tool
diameter (C) and depth of cut(D) had significant effect on surface roughness in end milling operation. These
factors are controllable factors. Table1 shows all Taguchi design parameters and levels.
One of the most important considerable attributes of Taguchi parameter design was S/N ratio. It was differ at
the different place.

Controllable factors Level 1 Level 2 Level 3 Level 4

Level 5

A: Tool feed (mm\min) 500 1000 1500 2000 2500
B: Tool speed (rpm) 2000 3000 4000 5000 6000
C: Tool diameter (mm) 6 8 10 12 16
D: Depth of cut (mm) 0.25 0.50 0.60 0.75

0.85


Table 1. Selected Factors And Levels

2.3 Select Orthogonal Array

Literature survey and discuss with some industrial person, we considered four factors and five levels in our
study. As per above parameters we select L25 orthogonal array (OA) in the Taguchi parameter design. The
layout of L25 orthogonal array is shown in Table-2.

Sr.no. Tool feed Tool speed Tool dia.
Depth of
cut
1 500 2000 08 0.25
2 500 3000 10 0.5
3 500 4000 12 0.75
4 500 5000 16 0.85
5 500 6000 06 0.6
6 1000 2000 10 0.75
7 1000 3000 12 0.85
8 1000 4000 16 0.6
9 1000 5000 06 0.25
10 1000 6000 08 0.5
11 1500 2000 12 0.6
12 1500 3000 16 0.25
13 1500 4000 06 0.5
14 1500 5000 08 0.75
15 1500 6000 10 0.85
16 2000 2000 16 0.5
17 2000 3000 06 0.75
18 2000 4000 08 0.85
19 2000 5000 10 0.6
20 2000 6000 12 0.25
21 2500 2000 06 0.85
Patel K. P. / International Journal of Engineering Science and Technology (IJEST)
ISSN : 0975-5462 Vol. 4 No.02 February 2012 541

22 2500 3000 08 0.6

23 2500 4000 10 0.25
24 2500 5000 12 0.5
25 2500 6000 16 075

Table 2. Orthogonal Array L25

Sr.no. Tool feed Tool speed Tool dia.
Depth of
cut
Surface
roughness
S/N ratio
1 500 2000 08 0.25 0.77 2.2701
2 500 3000 10 0.5 1.34 -2.5421
3 500 4000 12 0.75 1.18 -1.4376
4 500 5000 16 0.85 0.9 -0.4376
5 500 6000 06 0.6 1.07 0.5623
6 1000 2000 10 0.75 1.92 1.5623
7 1000 3000 12 0.85 1.23 2.5623
8 1000 4000 16 0.6 1.19 3.5623
9 1000 5000 06 0.25 1.16 4.5623
10 1000 6000 08 0.5 0.62 5.5623
11 1500 2000 12 0.6 0.66 6.5623
12 1500 3000 16 0.25 1.10 7.5623
13 1500 4000 06 0.5 2.39 8.5623
14 1500 5000 08 0.75 1.39 9.5623
15 1500 6000 10 0.85 1.39 10.562
16 2000 2000 16 0.5 8.02 11.562
17 2000 3000 06 0.75 6.10 12.562
18 2000 4000 08 0.85 0.74 13.562
19 2000 5000 10 0.6 1.25 14.562
20 2000 6000 12 0.25 1.06 15.562
21 2500 2000 06 0.85 4.76 16.562
22 2500 3000 08 0.6 6.73 17.562
23 2500 4000 10 0.25 1.06 18.562
24 2500 5000 12 0.5 1.14 19.562
25 2500 6000 16 075 2.24 20.562

Table 3. Results Of L25 Experiment
2.6 Conduct Thee Experiments






















Figure 1. Experimental Set Up
Patel K. P. / International Journal of Engineering Science and Technology (IJEST)
ISSN : 0975-5462 Vol. 4 No.02 February 2012 542

Figure 1 shows the experimental set up and operation.The work piece material used was AL 6351 T6 in the
form of 420mm x 120mm x 20mm plate. The all 25 experiments were run by BFW V-30 CNC vertical
milling machine. Surface roughness data values collected with the help ofMitutoyo-surface roughness tester
after end milling was completed. S/N ratio was calculated based on following Eq. (1).

S/N () =-1ulog
1
n
yi
2 n
=1
(1)
Where, n =number of measurements Ina trial/row, in this case, n=5
yi =the i
th
measured value in a run/row.

2.4 Analyze The Result And Determine Optimum Factor-Level Combination

(1) Analysis Of Raw Data And S/N Ratio

Level Tool feed (A) Tool speed(B) Tool dia. (C) D.o.cut (D)
1 1.052 3.226 3.096 1.030
2 1.224 3.300 2.050 2.702
3 1.386 1.312 1.392 2.566
4 3.434 1.168 1.054 1.804
5 3.186 1.276 2.690 2.180
Delta 2.382 2.132 2.042 1.672
Rank 1 2 3 4

Table 4. Average Effect Response Table For Raw Data

Level Tool feed (A) Tool speed (B) Tool dia. (C) D.o.cut (D)
1 -0.2764 -6.2845 -7.7407 -0.1718
2 -1.2224 -7.4870 -2.0766 -5.0359
3 -2.1015 -1.6815 -2.7025 -6.5351
4 -6.7238 -1.2621 -0.2542 -2.9360
5 -7.7523 -1.3614 -5.3024 -3.3976
Delta 7.4759 6.2249 7.4865 6.3633
Rank 2 4 1 3

Table 5. Average Effect Response Table For S/N Ratio

After raw data were collected, average effect response values (Table 4) and S/N response ratios (Table 5),
respectively, were calculated based on Table 2. The calculation of average effect response values and S/N
ratios were based on the following procedure. For example, the average effect for level one of Tool feed was
computed using data from experimental numbers 1-5 of Table 3.The average effect for level two of Toolfeed
was computed using experimental numbers 6-10 of Table 3. The average effect for level three of Tool feed
was computed using experimental numbers 11-15 of Table 3.The average effect for level four of Tool feed
was computed using experimental numbers 16-20 of Table 3.The average effect for level five of Tool feed
was computed using experimental numbers 21-25 of Table 3. Similarly, the average effect of Tool speed and
Tool diameter and depth of cut was computed for all other cutting levels. The S/N ratio is calculated in the
same way. The average effects and S/N ratios for each level of cutting parameters are summarized and
referred to in the average effects response table and S/N ratios response table for surface roughness (Ra), as
shown in Tables 4 and 5.

(2) Analysis Of Variance

Source Degree of
freedom
Sum of square Mean of
square
F value P value
Tool feed 4 26.625 6.656 2.12 0.170
Tool speed 4 24.335 6.084 1.94 0.197
Tool dia. 4 14.642 3.661 1.17 0.393
D.o.cut 4 9.045 2.261 0.72 0.601
error 8 25.092 3.137
Total 24 99.739
Table 6. ANOVA Table For Response Raw Data
Patel K. P. / International Journal of Engineering Science and Technology (IJEST)
ISSN : 0975-5462 Vol. 4 No.02 February 2012 543

The purpose of the analysis ofvariance (ANOVA) is to determinewhich cutting parameters
significantlyaffect the quality characteristic (Ra).Table 6 shows the results of ANOVAanalysis of raw data
for surface roughness.

(3) Determine Optimum Factor-Level Combination


2500 2000 1500 1000 500
3
2
1
6000 5000 4000 3000 2000
16 12 10 8 6
3
2
1
0.85 0.75 0.60 0.50 0.25
Tool feed
M
e
a
n

o
f

M
e
a
n
s
Tool speed
Tool dia D.O.Cut
Main Effects Plot for Means
Data Means

2500 2000 1500 1000 500
0
-2
-4
-6
-8
6000 5000 4000 3000 2000
16 12 10 8 6
0
-2
-4
-6
-8
0.85 0.75 0.60 0.50 0.25
Tool feed
M
e
a
n

o
f

S
N

r
a
t
i
o
s
Tool speed
Tool dia D.O.Cut
Main Effects Plot for SN ratios
Data Means
Signal-to-noise: Smaller is better


Figure 2. Response Graph Of Five Cutting Parameters For Mean And S/N Ratio

Figure 2 shows three graphs, each of which contains a curve representingthe mean and a curve representing
theS/N ratio. The values of the graphs arefrom Table 4 and 5. The S/N ratio indicates thedegree of the
predictable performanceof a product or process in the presenceof noise factors. Process parametersettings
with the highest S/N ratioalways yield the optimum quality with minimum variance.The level that hasa
higher value determines the optimumlevel of each factor. For example, in Figure 5, level one for tool
feed(A1=0.02 in) has the highest S/N ratiovalue, which indicated that the machining performance at such
level producedminimum variation of the surfaceroughness. In addition, the lowersurface roughness value
had a bettermachining performance. Furthermore,level two of tool feed (A1 =500 mm/min)has indicated the
optimum situation interms of mean value.Similarly, the level three of cuttingspeed (B4=4000 rpm) and the
level oneof tool dia. (C4=16 mm) and level four of depth of cut (D1=0.25 mm) have alsoindicated the
optimum situation interms of S/N ratio and mean value.


2.5 Predict Optimum Performance

One could predict the optimum surface roughness performance using the following Taguchi equation:

Predicted mean =A1 +B4 +C4 +D1 5 (Y)

=1.052 +1.168 +1.054 +2.566 5 (2.0564)

=-4.44 m.

Similarly, the S/N ratio could be predicted as:

Predicted S/N =pA1 +pB4 +pC4 +pD1 5 (p)

=-0.2764 - 1.2621 - 0.2542 - 0.1718 5 (-3.6152)

=16.1115 Db.

Here, predicted mean Ra =-4.44 m which so that the parameter we used is within the range of specified
cutting conditions (Table 2). A confirmation of the experimental design was necessary in order to verify the
optimum cutting conditions.




Patel K. P. / International Journal of Engineering Science and Technology (IJEST)
ISSN : 0975-5462 Vol. 4 No.02 February 2012 544

3 CONCLUSION

In this study, the analysis of confirmation experiments has shown that Taguchi parameter design can
successfully verify the optimum cutting parameters, which are A1 >B4 >C4 >D1 (tool feed (A), tool
speed(B), tool diameter(C), depth of cut(D) ). The work piece material used was AL 6351 T6. The average
value of surface roughness [Mean (=- 4.44 m) and S/N ratio (=16.1115 Db.)] were calculated and were
found to be within the range.Taguchi parameter design can provide a systematic procedure that can
effectively and efficiently identify the optimum surface roughness in the process control of individual end-
milling machines. It also allows industry to reduce process or product variability and minimize product
defects by using a relatively small number of experimental runs and costs to achieve superior-quality
products. This research only demonstrates how to use Taguchi parameter design for optimizing machining
performance with minimum cost. Further study could consider more factors (e.g. forces, materials, lubricant,
etc.) in the research to see how the factors would affect surface roughness. Also, further study could consider
the outcomes of Taguchi parameter design when it is implemented as a part of management decision-making
processes.

References:

[1] A handbook of a primer on The Taguchi method by: Ranjit.K.roy.
[2] Ab. Rashid M.F.F., Gan S.Y., and Muhammad N.Y. Mathematical Modeling to Predict Surface Roughness in CNC Milling
(2009).
[3] B. C. Routara & A. Bandyopadhyay & P. Sahoo Roughness modeling and optimization in CNC end milling using response surface
method: effect of work piece material variation, 15 March 2008.
[4] Sanjit,Moshat,Saurav Datta, Asish Bandyopadhyay and Pradip Kumar Pal, Optimization of CNC end milling process parameters
using PCA-based Taguchi method,2010.
[5] J ulie Z. Zhang, J oseph C. Chen, E. Daniel Kirby, Surface roughness optimization in an end-milling operation using the Taguchi
design method, 27 September 2006.
[6] Babur Ozcelik, Mahmut Bayramoglu, The statistical modeling of surface roughness in high-speed at end milling, 14 November
2005.


Patel K. P. / International Journal of Engineering Science and Technology (IJEST)
ISSN : 0975-5462 Vol. 4 No.02 February 2012 545