Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and

Applications (IJERA) ISSN: 2248-9622 www.ijera.com

Vol. 2, Issue 5, September- October 2012, pp.1594-1602

Multi-Response Optimization By Experimental Investigation Of

Machining Parameters In CNC Turning By Taguchi Based Grey
Relational Analysis
Pankaj Sharma*, Kamaljeet Bhambri**
*(Department of Mechanical Engg., M.M. University, Mullana, Ambala (Haryana) - India)
** (Department of Mechanical Engg., S.U.S.C.E.T., Tangori, Mohali (Punjab) - India)

Abstract: As we know in actual machining there are

The AISI H13, a chromium based hot work tool many factors which affect the surface roughness and
Steel has a wide variety of applications in material removal rates i.e. cutting conditions, tool
aluminum casting and extrusion dies, forging variables and work piece variables. Cutting
dies, hot nut tools, hot header dies, extrusion conditions includes speed, feed and depth of cut. The
mandrels, plastic molds, cores, die holder blocks, tool variables includes tool material, nose radius,
hot press dies and specially hot work punches etc. rake angle, cutting edge geometry, tool vibration,
In this study, the optimization of two response tool overhang, tool point angle etc. The work piece
parameters (Surface roughness and Material variables include hardness and mechanical
Removal Rate) by three machining parameters properties of the material. It is very difficult to take
(cutting speed, feed rate and depth of cut) is all the parameters that control the response
investigated in high speed turning of H13 in dry parameters for a particular manufacturing process. In
conditions. Taguchi’s L’18 orthogonal array and a turning operation, it is very difficult to select the
analysis of variance (ANOVA) are used for cutting parameters to achieve the high surface finish
individual optimization. The simultaneous with optimal material removal rate. This study
optimization is done by Grey Relational Analysis would help the operator to select the cutting
approach. The different levels of all machining parameters.
parameters are used and experiments are done The hot work tool steels have the ability to
on HMT STALLION-100 HS CNC lathe resist softening under hot working conditions and
machine. The optimum condition for combined after numerous exposures to elevated operating
effects was found V5-F2-D3 and the optimal temperatures. Hot work material used for the present
value of the surface roughness (Ra) comes out to study is H13 steel. The nominal chemical
be 1.0828 (μm) and of MRR is 554.0.4 (mm³/sec). compositions are {Chromium (5%), Vanadium (1%),
The optimum results are also verified with the Molybdenum (1.5%), Carbon (0.4%), Maganese
help of confirmation experiments. (0.35%)}. This H13 tool steel is suitable for forging
dies, forging die inserts, hot gripper dies, hot nut
Keywords: CNC Turning, Optimization, ANOVA, tools, hot header dies, brass forging and pressing
MRR, Surface Roughness, Taguchi Method, Grey dies, aluminum base dies, aluminum casting and
Relational Analysis, GRG. extrusion dies, zinc die casting dies, extrusion
mandrels, plastic molds, cores, die holder blocks, hot
1. Introduction: press dies and hot work punches etc.
Quality plays a major role in today‟s This paper is about experimentally
manufacturing market. From Customer‟s viewpoint investigating and optimizing the machining
quality is very important because the quality of parameters for Material Removal Rate (MRR) and
products affects the degree of satisfaction of the Surface Roughness in CNC turning by Taguchi
consumers during usage of the product. It also method and grey relational analysis approach..
improves the goodwill of the company. Taguchi‟s orthogonal arrays are highly fractional
High speed turning operation is done on designs, used to estimate main effects using very
CNC lathe. The quality of the surface plays a very few experimental runs. These designs are not only
important role in the performance of dry turning applicable for two level factorial experiments, but
because a good quality turned surface surely also can investigate main effects when factors have
improves fatigue strength, corrosion resistance and more than two levels. Designs are also available to
creep life. Surface roughness also affects on some investigate main effects for some mixed level
functional attributes of parts, such as, contact experiments where the factors included do not have
causing surface friction, wearing, light reflection, the same number of levels. For example, a four-level
ability of distributing and also holding a lubricant, full factorial design with five factors requires 1024
load bearing capacity, coating and resisting fatigue. runs while the Taguchi orthogonal array reduces the
required number of runs to 16 only.

1594 | P a g e
 Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 5, September- October 2012, pp.1594-1602

Akhyar, G. et al. (2008) [1] has used the studied with minimum possible number of
application of taguchi method in optimization of experiments. Using Taguchi method, Appropriate
cutting parameters for surface roughness in turning Orthogonal Array has been chosen and experiments
Ti-6%Al-4%V extra low interstitial with various tool have been performed as per the set of experiments
grades coated and uncoated cemented carbide tools designed in the orthogonal array. Signal to Noise
under dry cutting condition and high cutting speed. ratios are also calculated for analyzing the effect of
The analysis of results show that the optimal machining parameters more accurately. The results
combination of parameters are at cutting speed of 75 of the experimentation are analyzed analytically and
m/min, feed rate of 0.15 mm/min, depth of cut of graphically as well. ANOVA is used to determine
0.10 mm and tool grade of KC9225. the percentage contribution of all factors upon each
Chakradhar, D. and Venu Gopal, A. (2011) response individually.
[3] has done the multi objective optimization of
electrochemical machining of EN-31 steel by grey 3. Taguchi Method:
relational analysis. The process parameters Traditional experimental design methods
considered are electrolyte concentration, feed rate are very complicated and difficult to use.
and applied voltage and are optimized with Additionally, these methods also require a large
considerations of multiple performance number of experiments when the number of process
characteristics including material removal rate, over parameters increases. In order to minimize the
cut, cylindricity error and surface roughness. With number of tests required, Taguchi experimental
the help of Analysis of variance (ANOVA) it was design method, a powerful tool for designing high-
observed that feed rate is the significant process quality system, was developed by Taguchi. Taguchi
parameter that affects the ECM robustness. method [1, 11, 18] uses a design of orthogonal
Hassan, K. et al. (2012) [11] has done the arrays to study the entire parameter space with small
experimental investigation of material removal rate number of experiments only. Taguchi recommends
(MRR) in cnc turning of C34000 using taguchi analyzing the mean response for each run in the
method using L‟27 array. When the MRR is array, and he also suggests to analyze variation using
optimized alone the MRR comes out to be 8.91. The an appropriately chosen signal-to-noise ratio (S/N).
optimum levels of process parameters for There are 2 Signal-to-Noise ratios of common
simultaneous optimization of MRR have been interest for optimization of static problems used in
identified. Optimal results were verified through present study as are:
confirmation experiments. It was concluded that
MRR is mainly affected by cutting speed and feed (I) SMALLER-THE-BETTER:
rate.
Jaharah, A.G. et al (2009) [13] has studied
the effect of uncoated carbide tool geometries in
turning AISI 1045. This paper presents the
application of Finite element method (FEM) in
simulating the effect of cutting tool geometries on (II) LARGER-THE-BETTER:
the effective stress and temperature increased in
turning. The tool geometries studied were various
rake (α) and clearance (β) in the different ranges.
The minimum effective stress of 1700MPa is
achieved using rake and clearance angles of 5° and
5° respectively with cutting speed of 300mm/min, Where, ƞ - Signal to Noise (S/N) Ratio,
and feed rate of 0.25mm/rev. Yi - ith observed value of the response,
Ojel, T. et al. (2005) [23] has studied the n - Number of observations in a trial,
effects of cutting edge geometry, work piece y - Average of observed values (responses)
hardness, feed rate and cutting speed on surface s - Variance.
roughness and resultant forces in the finish hard
turning of AISI H13 steel. Cubic Boron Nitrite Regardless of the category of the
inserts with two distinct edge preparations performance characteristics, the higher S/N ratio
(chamfered and honed) and through hardened AISI corresponds to a better performance. Therefore, the
H13 steel bars were used. The honed Edge geometry optimal level of the process parameters is the level
and lower work piece surface hardness resulted in with the highest S/N value. The statistical analysis of
better surface roughness. the data is performed by analysis of variance
(ANOVA) [3, 12] to study the contribution of the
2. Design of Experiment: various factors and interactions and to explore the
Experiments were designed using Taguchi effects of each process on the observed values.
method so that effect of all the parameters could be

1595 | P a g e
 Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 5, September- October 2012, pp.1594-1602

4. Experimental Plan and details: which design parameter significantly affects the
In this study, three machining parameters surface roughness. Based on the ANOVA, the
were selected with different levels as given in Table relative importance of the machining parameters
1. The experimental design was according to an with respect to material removal rate and surface
L‟18 array based on Taguchi method. A set of roughness was investigated to determine the
experiments designed and conducted to investigate optimum combination of the machining parameters.
the relation between the process parameters and The Material Removal Rate calculations
response factor. Minitab 16 software is used for and experimental results of the surface roughness for
optimization and graphical analysis of obtained data. turning of H13 with different turning parameters are
shown in Table 3.
Table 1: Turning parameters and levels
Table 3: Design of experiment and calculations
Levels CS F DOC
(m/sec) (mm/rev) (mm) S. Weight Weight Machining Means of
No. before after Time MRR
1 150 0.1 0.5
turning turning (sec.) (mm³/sec)
2 180 0.2 1
(gm) (gm)
3 210 0.3 1.5
1 454.00 434.70 7.3 304.99
4 240
2 434.70 421.62 3.9 424.54
5 270
3 421.62 414.97 2.76 334.66
6 300
4 456.00 437.05 6.1 319.12
5 437.05 423.98 3.33 496.83
The work material selected for this
6 423.98 417.98 2.38 393.24
experiment is H13 of Ø 32 mm, length 70 mm in the
present study. The turning operation is performed in 7 455.00 436.5 5.36 535.71
3 steps of 17 mm length each over the total length of 8 436.5 424.1 2.93 436.90
varying depth of cut. The chemical composition of 9 424.1 417.9 2.12 370.19
H13 sample can be seen in Tables 2. 10 460.00 441.25 4.75 499.67
11 441.25 428.63 2.63 403.48
Table 2: Chemical composition of H13 sample 12 428.63 422.51 1.92 607.40
Ele- C Cr V Mo SI S P 13 450.00 431.51 4.28 767.37
ments 14 431.51 418.84 2.09 546.85
% 0.36 4.9 0.9 1.4 1.04 0.03 0.03 15 418.84 412.6 1.76 448.79
16 455.00 436.18 3.9 610.84
The turning tests were carried out to 17 436.18 423.51 2.2 460.02
determine the Material Removal Rate and Surface 18 423.51 417.55 1.64 728.99
Roughness under various turning parameters. A
HMT STALLION-100 HS CNC lathe machine is Table 4: Design of experiment and calculations
used for experimentation. Ex. CS F D MRR Ra
Roughness is measured using stylus type No. (m/ ( mm/ (mm (mm³/ (μm)
surface roughness tester „Surftest SJ-201‟ made of min) rev) ) sec)
Mitutoyo, Japan. The cut-off length ( ) was chosen 1 150 0.1 0.5 304.99 0.54
as 0.8 for each roughness measurement. An average 2 150 0.2 1.0 424.54 0.96
of 5 measurements of the surface roughness was 3 150 0.3 1.5 334.66 1.29
taken to use in the multi-criteria optimization. 4 180 0.1 0.5 319.12 0.60
The Material Removal Rate, MRR (mm³/ 5 180 0.2 1.0 496.83 0.96
min) was calculated using formula: 6 180 0.3 1.5 393.24 1.09
7 210 0.1 1.0 535.71 0.48
8 210 0.2 1.5 436.90 1.10
9 210 0.3 0.5 370.19 2.19
Where, Wi = Initial weight of work piece in gm 10 240 0.1 1.5 499.67 1.08
Wf= Final weight of work piece in gm 11 240 0.2 0.5 403.48 1.03
t = Machining time in seconds 12 240 0.3 1.0 607.40 2.13
ρs= Density of mild steel 13 270 0.1 1.0 767.37 1.02
= (7.λ x 10ˉ³ gm/mm³). 14 270 0.2 1.5 546.85 1.06
15 270 0.3 0.5 448.79 2.06
5. Experimentation and Calculations: 16 300 0.1 1.5 610.84 0.90
In high speed turning operation, surface 17 300 0.2 0.5 460.02 1.02
roughness is an important criterion. The purpose of
18 300 0.3 1.0 728.99 2.14
the analysis of variance (ANOVA) is to investigate

1596 | P a g e
 Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 5, September- October 2012, pp.1594-1602

Table 5: ANOVA Table for means of MRR 6. Grey Relational Analysis:

V D SS MS F P C In the Grey relational analysis [3, 25] the
F quality characteristics are first normalized, ranging
C 5 14745 2949 25.07 0.001 49.91 from zero to one. This process is known as Grey
S 0 0 *** Relational Generation. Then the Grey Relational
F 2 6078 3039 2.58 0.136 2.06* Coefficient based on normalized experimental data
D 2 13247 6623 56.31 0.002 44.84 is calculated to represent the correlation between the
8 9 ** desired and the actual experimental data. Then
E 8 9410 1176 3.19 overall Grey Relational Grade (GRG) is determined
T 17 29541 100 by averaging the Grey relational coefficient
6 corresponding to selected responses.
The overall performance characteristic of
Table 6: Response Table for MRR the multiple response process depends on the
Level CS F DOC calculated GRG. This Grey relational approach
1 354.7 506.3 384.4 converts a multiple response process optimization
2 403.1 461.4 593.5 problem into a single response optimization
3 447.6 480.5 470.4 problem. The optimal parametric combination is
4 503.5 then evaluated, which would result in the highest
5 587.7 Grey relational grade. The optimal factor setting for
6 maximizing the overall Grey relational grade can be
600.0
performed using the Taguchi method.
Delta 245.2 44.8 209.0
In Grey relational generation, the
Rank 1 3 2
normalized MRR should follow the larger-the-better
(LB) criterion, which can be expressed as:
Table 7: ANOVA Table for means of Surface
Roughness xj(k) = yi(k) – min yi(k)
V D SS MS F P C max yi(k) – min yi(k)
F
C 5 0.8333 0.166 2.21 0.153 16.20 The normalized Ra should follow the smaller-the-
S 66 ** better (SB) criterion which can be expressed as:
F 2 3.5817 1.790 23.74 0.001 69.61
87 *** xj(k) = max yi(k) – yi(k)
D 2 0.1265 0.063 0.84 0.467 2.46* max yi(k) – min yi(k)
27
E 8 0.6036 0.075 11.73 where, xi(k) and xj(k) are the value after Grey
45 Relational Generation for LB and SB criteria. Max
T 17 5.1452 100 yi(k) is the largest value of yi(k) for kth response and
min yi(k) is the minimum value of yi(k) for the kth
Table 8: Response Table for Surface Roughness response.
Level CS F DOC The Grey relational coefficient ξi(k) can be
1 0.9300 0.7700 1.2400 calculated as:
2 0.8833 1.0217 1.2817 ξi(k) = Δmin – Ψ Δmax
3 1.2567 1.8167 1.0867 Δ oi(k) + Ψ Δmax
4 1.4133
5 1.3800 And Δ oi = ║ xo(k) – xi(k) ║
6 1.3533
Delta 0.5300 1.0467 0.1950 Where Δ oi is the difference between absolute value
Rank 2 1 3 xo(k) and xi(k). Ψ is the distinguishing coefficient 0
≤ Ψ ≥ 1. Δ min and Δ max are the minimum and
Where, maximum value among the Δ oi for corresponding
V-Variable, CS-Cutting Speed, kth response.
F-Feed Rate, D-Depth Of Cut,
SR-Surface Roughness, E-Error, Now the Grey Relational Grade (GRG) can be
T-Total, DF-Degree of Freedom, calculated as :
SS-Sum of Squares, MS-Mean of Squares,
F-a statistical parameter, P-Percentage, γi =
C-% Contribution. Where n is the number of process responses.
Here *** & ** represents most significant and
significant parameters and * as less significant.

1597 | P a g e
 Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 5, September- October 2012, pp.1594-1602
The higher value of the GRG corresponds to a Where,
relational degree between the Reference Sequence GRGC- Grey Relational Generation Calculation,
xo(k) and the given sequence xi(k). The Reference RSDC- Reference Sequence Definition
Sequence xo(k) represents the best process sequence. Calculation,
Therefore, a higher GRG means that the GRCC- Grey Relational Coefficient Calculation
corresponding parameter combination is closer to the
optimal. The mean response for the GRG and the Table 11: Grey Relational Grade (GRG)
main effect plot of the GRG are very important calculation
because the optimal process condition can be S. No. GRG Rank
evaluated from this plot. Mean S/N Ratio
1 0.468 -6.5950 12
Table 9: S/N Ratio Calculation for MRR and Ra 2 0.522 -5.6465 11
S. No. Mean Values S/N Ratios 3 0.423 -7.4731 14
Xo MRR Ra MRR Ra 4 0.630 -4.0131 4
1 304.99 0.54 49.6857 5.352 5 0.551 -5.1769 7
2 424.54 0.96 52.5584 0.355 6 0.462 -6.7071 13
3 334.66 1.29 50.4921 -2.212 7 0.706 -3.0239 3
4 319.12 0.60 50.0791 4.437 8 0.539 -5.3682 9
5 496.83 0.96 53.9242 0.355 9 0.351 -9.0938 18
6 393.24 1.09 51.8932 -0.749 10 0.526 -5.5802 10
7 535.71 0.48 54.5786 6.375 11 0.600 -4.4369 5
8 436.90 1.10 52.8076 -0.828 12 0.365 -8.7541 17
9 370.19 2.19 51.3685 -6.809 13 0.563 -4.9898 6
10 499.67 1.08 53.9737 -0.668 14 0.798 -1.9599 2
11 403.48 1.03 52.1164 -0.257 15 0.386 -8.2682 15
12 607.40 2.13 55.6695 -6.568 16 0.547 -5.2402 8
13 767.37 1.02 57.7001 -0.172 17 0.801 -1.9273 1
14 546.85 1.06 54.7574 -0.506 18 0.385 -8.2907 16
15 448.79 2.06 53.0409 -6.277
16 610.84 0.90 55.7185 0.915 Table 12: ANOVA Table for means of GRG
17 460.02 1.02 53.2555 -0.172 V D SS MS F P C
18 728.99 2.14 57.2544 -6.608 F
Max. 304.99 0.48 49.6857 -6.809 C 5 0.0293 0.0058 0.50 0.772 9.35**
Min. 767.37 2.19 57.7001 6.375 S 08 62
F 2 0.1860 0.0930 7.86 0.013 59.31
Table 10: Grey Relational Analysis Calculations 55 27 ***
S. GRGC RSDC GRCC D 2 0.0036 0.0018 0.15 0.860 1.16*
N. MR Ra MR Ra MR Ra 35 17
R R R E 8 0.0946 0.0118 30.18
xo 1.000 1.000 1.000 1.000 1.000 1.000 84 36
1 0.064 0.965 0.936 0.035 0.348 0.588 T 17 0.3136 100
2 0.259 0.719 0.741 0.281 0.403 0.640 82
3 0.000 0.526 1.000 0.474 0.333 0.513
4 0.191 0.930 0.809 0.070 0.382 0.877 Table 13: ANOVA Table for S/N Ratio of GRG
5 0.415 0.719 0.585 0.281 0.461 0.640 V D SS MS F P C
6 0.031 0.643 0.969 0.357 0.340 0.583 F
7 0.285 1.000 0.715 0.000 0.412 1.000 C 5 5.821 1.164 0.48 0.780 7.27
8 0.498 0.637 0.502 0.363 0.499 0.579 S 2 **
9 0.141 0.000 0.859 1.000 0.368 0.333 F 2 53.90 26.95 11.19 0.005 67.33
10 0.421 0.649 0.579 0.351 0.463 0.588 1 07 ***
11 0.654 0.678 0.346 0.322 0.591 0.608 D 2 1.058 0.529 0.22 0.808 1.32
12 0.213 0.035 0.787 0.965 0.389 0.341 0 *
13 0.523 0.684 0.477 0.316 0.512 0.613 E 8 19.27 2.409 24.08
14 1.000 0.660 0.000 0.340 1.000 0.595 3 1
15 0.311 0.076 0.689 0.924 0.421 0.351 T 17 80.05 100
16 0.321 0.754 0.679 0.246 0.424 0.670 3
17 0.994 0.684 0.006 0.316 0.988 0.613
18 0.335 0.029 0.665 0.971 0.429 0.340

1598 | P a g e
 Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 5, September- October 2012, pp.1594-1602

Table 14: Response Table for means of GRG Residual Plots for Means
Level CS F DOC Normal Probability Plot Versus Fits
1 0.4710 0.5733 0.5393 99
0.10
2 0.5477 0.6352 0.5153 90
0.05

Residual
Percent
3 0.5320 0.3953 0.5492 50 0.00
4 0.4970
10 -0.05
5 0.5823
1 -0.10
6 0.5777 -0.1 0.0 0.1 0.4 0.5 0.6 0.7 0.8
Residual Fitted Value
Delta 0.1113 0.2398
Rank 2 1 3 Histogram Versus Order
4 0.10

Table 15; Response Table for S/N Ratio of GRG 3 0.05

Frequency

Residual
Level CS F DOC 2 0.00

1 -6.572 -4.907 -5.722 1 -0.05

2 -5.299 -4.086 -5.980 -0.10
0
3 -5.829 -8.098 -5.388 -0.10 -0.05 0.00 0.05 0.10 2 4 6 8 10 12 14 16 18
Residual Observation Order
4 -6.257
5 -5.073 Fig 3: Residual plot for means of GRG
6 -5.153
Delta 1.499 4.012 0.592 Interaction Plot for GRG
Rank 2 1 3 Data Means
0.1 0.2 0.3
0.8 cs
Main Effects Plot for Means 150
180
Data Means
0.6 210
cs f cs 240
270
300
0.6 0.4
0.8 f
0.1
0.5
0.2
Mean of Means

0.6 0.3
f
0.4
150 180 210 240 270 300 0.1 0.2 0.3 0.4
doc 0.8 doc
0.5
0.6 1.0
0.6 1.5
doc
0.5
0.4
0.4 150 180 210 240 270 300 0.5 1.0 1.5
0.5 1.0 1.5

Fig 4: Interaction plot for means of GRG

Fig 1: Main effect plot for means of GRG
7. Determining Optimum Condition:
Main Effects Plot for SN ratios Both the response and S/N ratio are used to
Data Means derive the optimum conditions. Since for quality
-4
cs f characteristic, Grey Relational Grade larger the
-5
better approach is desirable, the largest is the ideal
-6
level for a parameter. The S/N ratio is always
highest at the optimum condition. The graphs of
Mean of SN ratios

-7
-8
Figures 2 and 3 are used to determine the optimum
150 180 210 240 270 300 0.1 0.2 0.3
process parameters combination. The optimum
doc combination is therefore V5-F2-D3.
-4
-5
7.1 Predictive Equation and Verification:
-6
The predicted values of GRG, MRR and Ra
-7
at the optimal levels are calculated by using the
-8
relation:
0.5 1.0 1.5
Signal-to-noise: Larger is better

Fig2: Main effect plot for S/N Ratio of GRG

1599 | P a g e
 Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 5, September- October 2012, pp.1594-1602

Where, The purpose of the ANOVA in this study is to

ñ - Predicted response value after optimization, identify the important turning parameters in
- Total mean value of quality characteristic, prediction of Material Removal Rate and Surface
Roughness. Some important results come from
- Mean value of quality characteristic at ANOVA and plots are given as below.
optimum level of each parameter and
o - Number of main machining parameters that Table 16 shown below shows that optimal values of
affect the response parameter. surface roughness and material removal rate that lie
between the optimal ranges.
Applying this relation, predicted values of GRG,
MRR and Ra at the optimum conditions are Table 16: Optimal values of machining and
calculated as: response parameters
CP OV OL POV EOV OR
1. ñGRG = 0.5346 + [ (0.5823 - 0.5346) +
(0.6352 – 0.5346) + (0.5492 - 0.5346) ] CS 270 V5- MRR= MRR= 544.20
= 0.6975 554.04 544.20 <MRR>
F 0.2 F2- 554.04
2. ñMRR = 482.73 + [ (587.7 – 482.73) +
(461.4 – 482.73) + (470.4 – 482.73) ] D 1.5 D3 Ra= Ra= 1.06
= 554.04 mm³/sec 1.0828 1.06 < Ra >
1.0828
3. ñRa = 1.2028 + [ (1.3800 – 1.2028) +
(1.0217 – 1.2028) + (1.0867 – 1.2028 ] Where,
= 1.0828 µm CP-Cutting Parameters
OV-Optimal Values of Parameters
The robustness of this parameter OL-Optimum Levels of Parameters
optimization is verified experimentally. This POV-Predicted Optimum value
requires the confirmation run at the predicted EOV-Experimental Optimum Value
optimum conditions. The experiment is conducted OR-Optimum Range of MRR and Surface
twice at the predicted optimum conditions. Roughness

Verifications: 9. Conclusion:
1. For Material Removal Rate (MRR): In this study, the Grey relational based
The calculated value of MRR at the Taguchi method was applied for the multiple
optimum condition (V5-F2-D3) is 554.04 mm³/sec. performance characteristics of turning operations. A
The error in the predicted optimum value (554.04) grey relational analysis of the Material removal rate
and the calculated value (544.20) is only 1.77%. and the surface roughness obtained from the Taguchi
method reduced from the multiple performance
2. For Surface Roughness (Ra): characteristics to a single performance characteristic
The average of two measured values (1.04, which is called the grey relational grade. Therefore,
1.08) of the response at the optimum condition (V5- the optimization of the complicated multiple
F2-D3) is 1.06 m. The error in the predicted performance characteristics of the processes can be
optimum value (1.0828) and experimental value greatly simplified using the Grey relational based
(1.06) is only 2.11%. Taguchi method. It is also shown that the
performance characteristics of the turning
Hence, so good agreement between the operations, such as the material removal rate and the
actual and the predicted results is observed. Since surface roughness are greatly enhanced by using this
the percentage error is less than 5%, it confirms method.
excellent reproducibility of the results. The results
show that using the optimal parameter setting (V5-
F2-D3) a higher material removal rate is achieved
with lower surface roughness.

8. Results:
The effect of three machining parameters
i.e. Cutting speed, feed rate and depth of cut and
their interactions are evaluated using ANOVA and
with the help of MINITAB 16 statistical software.

1600 | P a g e
 Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 5, September- October 2012, pp.1594-1602

References: [11] Hassan, K. et al. (2012) “Experimental

[1] Akhyar, G. et al. (2008), “Application of investigation of Material removal rate in
Taguchi method in the optimization of CNC turning using Taguchi method”
turning parameters for surface roughness”, International Journal of Engineering
International Journal of Science Research and application, vol. 2, no. 2, pp.
Engineering and Technology, vol.1, no.3, 1581-1590
pp. 60-66. [12] Hascalik, A. and Caydas, U. (2008),
[2] Adesta, E.Y.T. et al. (200λ), “Tool wear “Optimization of Turning parameters for
and surface finish investigation in high surface roughness and tool life based on the
speed turning using cermet insert by taguchi method” International Journal of
applying negative rake angles” European advance manufacturing technology, vol. 38,
Journal of Scientific Research, vol. 38, no. pp. 38:896-903.
2, pp. 180-188. [13] Jaharah, A.G. et al. (200λ), “The effect of
[3] Chakradhar, D. and Venu Gopal, A. (2011), uncoated carbide tool geometries in turning
“Multi-objective optimization of AISI 1045 using finite element analysis”
electrochemical machining of en31 steel by European Journal of Scientific research,
grey relational analysis”, International vol. 28, no.2, pp. 271-277.
Journal of Modeling and optimization, vol. [14] Jaharah, A.G. et al. (200λ), “Surface
1, no.2, pp. 113-117. integrity of AISI H13 tool steel in end
[4] David Pye et al. “Chromiun Steel Article milling process” International Journal of
for AISI H13 Hot Work Tool Steel by Pye Mechanical and Materials Engineering
Metallurgical consulting” at (IJMME), vol. 4, no. 1, pp. 88 -92.
http://www.heattreatonline.com/ [15] Kirby, E.D. (2006), “A parameter design
ChromiumSteelArticle.pdf study in a turning operation using the
[5] Datta, R. and Majumdar, A. (2010), taguchi method”, The technology
“Optimization of turning process interface/Fall 2006, pp. 1-14.
parameters using multi-objective [16] Lan, T.S. et al. (2010), “Fuzzy Linguistic
evolutionary algorithm” Institute of optimization on surface roughness for cnc
electrical and electronics engineers (IEEE), turning” Hindawi publishing corporation
vol. 2010, article id 978-1-4244-8126-2/10. Mathematical problems in engineering, vol.
[6] Dhanabalan, S. et al. (2011), “Optimization 10, article id 572506.
of EDM process parameters with multiple [17] Lan, T.S. et al. (2010), “Tool wear
performance characteristics for titanium optimization for general cnc turning using
grades” European Journal of Scientific fuzzy deduction” Scientific research
research, vol. 68, no. 3, pp. 297-305. engineering, vol. 2, pp. 1019-1025.
[7] Dhavamani, C. and Alwarsamy, T. (2011), [18] Lan, T.S. (2011), “Fuzzy parametric
“Reviw on optimization of machining deduction for material removal rate
operation” International Journal of optimization” Journal of mathematics and
academic research, vol. 3, no.3, pp. 4776- statistics, vol.2011, 7-(1), pp.51-56
4785. [19] Lalwani, D.I. et al. (2008), “Experimental
[8] Fnides, B. et al. (2011), “Application of Investigation of cutting parameters
response surface methodology for influencing on cutting forces and Surface
determining cutting force model in turning roughness in finish hard turning of MDN
hardened AISI H11 hot work tool steel” 250 steel”, Journal of Materials processing
Indian Academy of Sciences, vol. 36, part technology, vol. 206, pp. 167-179
1, pp. 109–123. [20] Mahapatra, S.S. et al. (2006), “Parametric
[9] Gonzalez, L.W.H. et al. (2011), “Study of analysis and optimization of cutting
performance of high speed turning of parameters for turning operations based on
AISI1045 using the volumetric dimension taguchi method” Proceedings of the
of cutting tool life” International journal of International Conference on Global
advances in machining and forming Manufacturing and Innovation, - July 27-
operations, vol. 3, no.1, pp.19-32. 29/2006, pp. 1-9.
[10] Ghani, M.U. et al. (2007), “An [21] Modh, N.R. et al. (200λ), “An experimental
investigation of heat partition and tool wear investigation to optimize the process
in hard turning of H13 tool steel with CBN parameters of AISI 52100 steel in hot
cutting tools” International Journal of machining” International Journal of
Advance Manufacturing and Technology, Engineering Research and Applications,
article id 170-007-1282-7. vol. 1, issue 3, pp.483-489.

1601 | P a g e
 Pankaj Sharma, Kamaljeet Bhambri / International Journal of Engineering Research and
Applications (IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 5, September- October 2012, pp.1594-1602

[22] Natarajan, C. et al. (2011), “Investigation of

cutting parameters of surface roughness for
a non ferrous material using artificial neural
network in cnc turning”, Journal of
mechanical engineering, vol. 3, no. 1, pp. 1-
14.
[23] Ojel, T. et al. (2005), “Effects of cutting
edge geometry, workpiece hardness, feed
rate and cutting speed on surface roughness
and forces in finish turning of hardened
AISI H13 steel” International Journal of
Advance Manufacturing and Technology,
vol. 25, pp. 262-269.
[24] Ojel, T. and Karpat, Y. (2004), “Predictive
modeling of surface roughness and tool
wear in hard turning using regression and
neural networks” International Journal of
Machine Tools and Manufacture, vol. 45,
pp. 467–479.
[25] Pawade, R.S. and Joshi, S.S. (2011),
“Multi-objective optimization of surface
roughness and cutting forces in high speed
turning of Inconel 718 using taguchi grey
relational analysis (TGRA)” International
Journal of Advance Manufacturing
technology, vol. 56, pp. 47-62.
[26] Prasad, M.V.R. et al. (200λ), “Experimental
investigation to study the influence of
process parameters in dry machining”
ARPN Journal of Engineering and Applied
Sciences, vol. 4, no. 3, pp. 91-94.
[27] Qamar, S.Z. et al. (200λ), “Effect of heat
treatment on mechanical properties of H13
steel” 13th international research/expert
conference „Trends in the Development of
Machinery and Associated Technology‟,
TMT2009, Hammamet, Tunisia, 16-21
October, 2009 .
[28] Routara, B.C. et al. (2007), “Use of
desirability function approach for
optimization of multiresponse performance
characteristics of the surface roughness
parameter in cnc turning”, International
conference on Mechanical Engineering,
vol. ICME07-AM-16, pp. 29-31.
[29] Ramanujam, R. et al. (2011), “Optimization
of cutting parameters for turning Al-
SiC(10p) MMC using ANOVA and grey
relational analysis” International Journal of
precision engineering and manufacturing,
vol. 12, no.4, pp. 651-656.
[30] Sahoo, A.K. and Sahoo, B. (2011),
“Surface Roughness model and parametric
optimization in finish turning using coated
carbide insert: Response Surface
Methodology and Taguchi Approach”
International journal of industrial
engineering computations, vol. 2, pp. 819-
830.

1602 | P a g e