Professional Documents
Culture Documents
(55-61)
ISSN No. 2230 - 7680 MIT Publications
INTRODUCTION
Now-a-days, due to the growing demand of superior quality components for its functional aspect, surface roughness of
a machined part plays a significant role in the modern manufacturing process. A good quality machined surface
appreciably improves fatigue strength, corrosion resistance, and creep life. Surface roughness also influences some
functional characteristics of parts, such as, contact causing surface friction, wearing, light reflection, heat transmission,
ability of distributing and holding a lubricant, load bearing capacity, coating etc. Consequently, the desired finish surface
is generally specified and the appropriate cutting parameters are preferred to attain the required quality [1]. In
manufacturing industries, manufacturers focused on the quality and productivity of the product. To increase the
productivity of the product, computer numerically machine tools have been implemented during the past decades.
Surface roughness is one of the most important parameters to determine the quality of product. The mechanism behind
the formation of surface roughness is very dynamic, complicated, and process dependent. Several factors will influence
the final surface roughness in a CNC milling operations such as controllable factors (spindle speed, feed rate and depth
of cut) and uncontrollable factors (tool geometry and material properties of both tool and workiepce). Most of the
machine operator using trial and errormethod to set-up milling machine cutting
conditions [2]. This method is not effective and efficient and the achievement of a desirable value is a repetitive
and empirical process that can be very time consuming. In order to solve the problem, a surface prediction
technique which is termed the multiple regression prediction models to optimize the cutting conditions is
developed. This method can find the best conditions required for the machining independent variables such as
speed, feed and depth of cut that would result in the best machining response.
LITERATURE REVIEW
Surface finish of milled components has massive influence on the quality of the finished product. Surface finish in milling
had been found to be influenced in varying amounts by a number of factors such as feed rate, work hardness, builtup edge, coolent used, cutting speed, depth of cut, cutting time, cutting edge.According to these parameters, a
comprehensive literature survey is carried out as follows. Srikanth and Kamala [3] developed a Real Coded Genetic
Algorithm (RCGA) to locate optimum cutting parameters and explained its advantages over the existing approach of
binary coded genetic algorithm (BCGA). Franic and Joze [4] used Binary Coded GeneticAlgorithm (BCGA) for the
optimization of machining parameters. This genetic algorithm optimizes the machining conditions having an influence on
production cost, time and quality of the final product. David et al. [5] addressed a
55
MIT International Journal of Mechanical Engineering, Vol. 2, No. 1, Jan. 2012, pp. (55-61)
ISSN No. 2230 - 7680 MIT Publications
methodology to calculate surface roughness in a high speedend-milling process and used Artificial Neural Networks
(ANN) and statistical tools to develop different surface roughness predictors. Oktem et al. [6] used response surface
methodology to generate a mathematical model for surface roughness in terms of cutting parameters: Feed, cutting
speed, axial depth of cut, radial depth of cut and machining tolerance. An optimum surface roughness predictive model
using binary coded genetic algorithm (BCGA) is suggested by Suresh et al. [7]. This model gives minimum and
maximum values of surface roughness and their respective optimal machining conditions. Yang and Tarng [8] used
Taguchi method for design optimization on surface quality. An orthogonal array, thesignal-to-noise (S/N) ratio and the
analysis of variance (ANOVA) were used to explore the cutting characteristics. Uros and Franci [9] suggested a
neural network-based technique to optimize the cutting parameters taking into consideration the technological,
economic and organizational limitations. Al-Ahmari [10] proposed empirical models for tool life, surface roughness and
cutting force for turning operations. Two significant techniques used were response surface methodology and Neural
Networks. Huang and Joseph [11] suggested multiple regression model to calculate in- process surface roughness in
turning operation via accelerometer. Hossain et al. [12] developed an artificial neural network algorithm for calculating
the surface finish in end milling of Inconel 718 alloy.Avisekh et al. [13] performed a feasibility study of on-linesupervising
of surface roughness in turning operations using a developed opto-electrical transducer. Regression and neural network
(NN) models were developed to predict surface roughness and compared to actual and on-line measurements. Groover
and Mikell [14] illustrated the effect of three factors, namely, the feed, nose radius, and cutting-edge angles, on surface
roughness. Azouzi et al. [15] suggested an on-line prediction of surface finish and dimensional accuracy in turning using
neural network based sensor fusion. Feng and Hu [16] performed a comparative study of the ideal and actual surface
roughness in finish turning and also developed the fractional factorial experimentation approach for analysing the impact
of turning parameters on the surface quality of turned surfaces and used ANOVA technique to examine the impact of
turning factors and factor interactions on surface roughness. Muammer et al. [17] proposed regression analysis and
neural network- based method to predict the surface roughness and compare it for various cutting conditions in turning.
Bajic et al. [18] modeled machined surface roughness and optimization of cutting parameters in face milling and
scrutinized the influence of cutting parameters on surface roughness in face milling. Sakir et al. [19] worked on the
calculation of surface roughness using artificial neural network in lathe machine operations and examined the effect of
tool geometry on surface roughness on AISI 1040 steel in dry cutting condition at a depth of cut of 0.5 mm. Optimization
of machining parameters not only enhance the utility for machining economics, but also the product quality to a great
extent.
Most of the techniques discussed above are based on trial and error approach to setup machining conditions in order to achieve desired surface quality.
56
Therefore, the purpose of this study is to develop a multiple regression model for the prediction of surface roughness
parameter Ra and to evaluate the prediction ability of the model.
PROBLEM DEFINITION
The goal of present analysis is to develop a technique to predict the surface roughness of a part to be machined and to
avoid trial and error approaches to set-up machining conditions in order to achieve the desired surface roughness. The
objective is to predict surface roughness parameter (R a) under multiple cutting conditions determined by spindle speed,
feed rate and depth of cut. Surface roughness would be measured directly by surface roughness measuring
instruments. Experimental results are expected to show that parameters of spindle speed, feed rate and depth of cut
that could calculate surface roughness (Ra) under different combinations of cutting parameters.
METHODOLOGY
Experiments have been carried out in order to examine the impact of one or more factors of the process parameters
(spindle speed, feed rate and depth of cut) on the surface finish of the machined surface in vertical milling operation.
When an experiment involves two or more factors, the factors can influence the response individually or interactally.
Figure 1 defines the steps involved in the process.
MIT International Journal of Mechanical Engineering, Vol. 2, No. 1, Jan. 2012, pp. (55-61)
ISSN No. 2230 - 7680 MIT Publications
57
Spindle speed: The rate at which the machine spindle rotates. Spindle speed is typically measured in rpm.
Feed: It is the relative velocity at which the cutter is advanced along the workpiece. Feed rate is
measured in mm/ min.
Depth of cut: It is the thickness of the metal that is removed in one cut. It is the perpendicular
distance measured between the machined surface andnon-machined surface of the workpiece.
Ra (surface roughness parameter): This parameter is also known the arithmetic mean roughness
value. It is defined as the average value of the ordinates (y1, y2, y3,.., yn) from the mean line. Ra is
universally recognized and the most used as international parameter of roughness. It is the arithmetic
mean of the departure of the roughness profile from the mean line (Figure 2).
Ra1NYiNi0
Where Ra = the arithmetic average deviation from the mean line.
y = the ordinate of the profile curve.
Figure 3: Machining process on the workpiece in CNC machine
EXECUTION OF EXPERIMENTS
The test specimens are prepared on a CNC vertical milling machine (Figure 3). A groove 10 mm X 1.5 mm is cut
on aluminium workpiece using HSS milling cutter of diameter 10 mm. The dimensions of the workpiece is shown
in Figure 4 and actual workpiece is shown in Figure 5. Surface roughness measurement is done off line with the
usage SJ201 surface roughness tester (Figure 6). A computer numeric control (CNC) program has been written
to perform the grooving operation. The parameters defined in the CNC machine are: Spindle speed (N), feed rate
(F), depth of cut (D).Different levels of cutting parameters are shown in Figure 7. All specimens in this experiment
are machined under dry cutting conditions.
Figure 6: Arrangement of surface roughness tester and workpiece
MIT International Journal of Mechanical Engineering, Vol. 2, No. 1, Jan. 2012, pp. (55-61)
ISSN No. 2230 - 7680 MIT Publications
58
2.The hardness of each work piece is same throughout the length of the work piece.
3.Surface roughness values measured within the measuring area are sufficient to represent the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
800
800
800
800
800
800
800
800
800
800
800
800
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
1200
Ra Value
min)
(mm)
Trial 1 Trial 1
Trial 1 Average
200
0.25
2.566 2.783
2.534
2.627
200
200
0.50
0.75
2.566 2.657
2.341 2.512
2.453
2.123
2.558
2.325
300
0.25
2.543 2.842
2.543
2.642
300
0.50
2.518 2.823
2.123
2.488
300
0.75
2.134 2.443
2.223
2.266
400
0.25
2.278 2.876
2.567
2.573
400
0.50
2.546 2.765
2.654
2.655
400
0.75
2.543 2.432
2.761
2.578
500
500
0.25
0.50
2.554 2.654
2.314 2.345
2.654
2.000
2.620
2.345
500
0.75
2.342 2.134
2.145
2.207
200
0.25
2.234 1.999
2.176
2.136
200
0.50
2.456 2.546
2.559
2.520
200
0.75
2.564 2.543
2.543
2.550
300
0.25
1.980 1.988
1.999
1.989
300
0.50
2.001 1.987
1.945
1.977
300
400
0.75
0.25
2.011 2.009
1.897 1.881
2.018
1.883
2.012
1.887
400
0.50
1.892 1.990
1.838
1.906
400
0.75
1.786 1.699
1.700
1.728
500
0.25
1.556 1.456
1.442
1.484
500
0.50
1.446 1.463
1.487
1.465
500
200
0.75
0.25
1.456 1.471
1.432 1.428
1.499
1.415
1.475
1.425
200
0.50
1.400 1.401
1.401
1.400
200
0.75
1.423 1.456
1.500
1.459
300
0.25
1.323 1.387
1.332
1.347
300
0.50
1.399 1.301
1.300
1.333
300
0.75
1.299 1.281
1.272
1.284
400
0.25
1.119 1.187
1.139
1.148
400
0.50
1.101 1.108
1.109
1.106
400
500
0.75
0.25
1.109 1.111
1.098 1.001
1.123
1.006
1.114
1.035
500
0.50
1.101 1.002
1.001
1.034
500
0.75
1.102 1.001
1.002
1.035
MIT International Journal of Mechanical Engineering, Vol. 2, No. 1, Jan. 2012, pp. (55-61)
59
x1x2
S 1
n1
1
x1
n1
1
n2
j1
1
nn
1
xj
n2
n1i1
S2
1n2
,x2
xi
2
2
n1
n2
(xix1)2
(xj x2)
j1
If the calculated value of |t| be < t 0.05, the difference between the sample mean is said to
insignificant at 5% level of significance.
1
2
3
4
5
800
200
800
200
800
200
800
300
800
300
0.25
0.5
0.75
0.25
0.5
Ra Value
(mm)
2.5539025
2.530605
2.5073075
2.5635775
2.534755
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
800
300
800
400
800
400
800
400
800
500
800
500
800
500
1000
200
1000
200
1000
200
1000
300
1000
300
1000
300
1000
400
1000
400
1000
400
1000
500
1000
500
1000
500
1200
200
1200
200
1200
200
1200
300
1200
300
1200
300
1200
400
1200
400
1200
400
1200
500
1200
500
1200
500
0.75
0.25
0.5
0.75
0.25
0.5
0.75
0.25
0.5
0.75
0.25
0.5
0.75
0.25
0.5
0.75
0.25
0.5
0.75
0.25
0.5
0.75
0.25
0.5
0.75
0.25
0.5
0.75
0.25
0.5
0.75
2.5059325
2.5732525
2.538905
2.5045575
2.5829275
2.543055
2.5031825
1.9494025
1.971605
1.9938075
1.7130775
1.729755
1.7464325
1.4767525
1.487905
1.4990575
1.2404275
1.246055
1.2516825
1.3449025
1.412605
1.4803075
0.8625775
0.924755
0.9869325
0.3802525
0.436905
0.4935575
0.1020725
0.050945
0.0011825
The numerical estimates of the effects indicate that the effect of feed is in positive direction. The
positive direction means that the surface finish declined with increasing the cutting feed. This is due to
the increase in distance between the successive cuts made by the tool during the cutting action, as
the cutting feed increases.
MIT International Journal of Mechanical Engineering, Vol. 2, No. 1, Jan. 2012, pp. (55-61)
ISSN No. 2230 - 7680 MIT Publications
The spindle speed has effect in negative direction. This means that increasing the spindle speed
improves the surface finish. It is generally well known that an increase in cutting speed improves
machinability. This may be due to the continuous reduction in the build up edge formation as the
cutting speed increases.
The depth of cut also has negative value, which indicates that increasing the depth of cut improves
the surface finish.
The interaction between the cutting feed and depth of cut significantly affects the surface
roughness. The interaction also suggests that to get a certain surface finish and maximum metal
removal it is preferable to use a high cutting feed associated with depth of cut.
60
It has been observed that increasing the spindle speed improves the surface finish as the cutting
feed decreases. This supports the earlier discussion about the effect of increasing cutting speed on
the surface roughness of the machined workpieces.
The interaction between the depth of cut and spindle speed reveals that increasing the spindle speed and
increasing the depth of cut deteriorates the surface finish.
CONCLUSION
In the present analysis, 36 specimens of Aluminum alloy have been machined in a CNC milling machine with
HSS tool and then a SJ 201 surface roughness tester has been used to determine R a values of all the
specimens.Amultiple regression model has been developed and evaluated by means of significant different
between the predicted Ravalues and the actual Ra values with the help of t-test. The significant conclusions drawn
from the present research are recapitulated as follows:
1.The surface roughness could be efficiently calculated by using spindle speed, feed rate, and depth of
is no significant different between the mean of R avalues of theoretical and experimental data at 5%
level of significance.
4.As the spindle speed increases for lower feed rates, the surface roughness decreases. For higher feed
rates, the surface roughness changes considerably.
5.As the depth of cut influences the surface roughness considerably for a given feed rate, the increase in
feed rate causes the surface roughness to increase. For lower depth of cut, the feed rate increases
with surface roughness.
REFERENCES
[1]Mike S.L., Joseph C., Chen, Li M. (1998). Surface Roughness Prediction for CNC End Milling, Materials and
Processes Quality Control Manufacturing. J. Ind. Technol., 15(1): (1999).
[2]Zhang J.Z., Joseph C.Chen, E.Daniel Kirby, Surface Roughness Optimization in an End-milling Operation using
the Taguchi Design Method, 27 January 2006
[3]Kamala, V., Srikanth T. (2008). A Real Coded Genetic Algorithm for Optimization of Cutting Parameters in Turning.
[6]Oktem H., Erzurumlu T., Kurtaran H. (2005). Application of Response Surface Methodology in the Optimization of
Cutting Conditions for Surface Roughness. J. Mat. Process. Technol., 170: 11-16.
[7]Suresh P.V.S., Venkateswara, R.P., Deshmukh, S.G. (2002). A Genetic Algorithmic Approach for Optimization of
Surface Roughness Prediction Model.Int. J. Mach. Tools Manuf., 42: 675-680.
[8]Yang, W.H., Tarng, Y.S., (1998). Design Optimization of Cutting Parameters for Turning Operations Based on the Taguchi
Method. J. Mater. Process. Technol., 84: 122-129.
[9]Uros, Z., Franci, C. (2003). Optimization of Cutting Conditions During Cutting by using Neural Networks. Robot.
[12]Hossain, M.I., Amin, A.K.M., Patwari, A.U. (2008). Development of an Artificial Neural Network Algorithm for
Predicting the Surface Roughness in End Milling of Inconel 718 alloy. International Conference on Computer and
Communication Engineering, ICCCE (2008), 13-15: 1321- 1324.
[13]Avisekh, B., Evgueni, V., Bordatchev, S., Kumar, C. (2009). On-line monitoring of Surface Roughness in Turning
Operations with Optoelectrical Transducer.Int. J. Manuf. Res., 4(1): 57-73.
[14]Groover, Mikell (1996). Fundamentals of Modern Manufacturing. Prentice Hall, Upper Saddle River, NJ (now
Published by John Wiley & Sons, New York).
[15]Azouzi, R., Guillot, M. (1997). On-line Prediction of Surface Finish and Dimensional Deviation in Turning using
Neural Network based Sensor Fusion. Int. J. Mach. Tool Manuf., 37(9): 1201-1217.
MIT International Journal of Mechanical Engineering, Vol. 2, No. 1, Jan. 2012, pp. (55-61)
ISSN No. 2230 - 7680 MIT Publications
[16]Feng, C. X., Hu ZJ (2001). A Comparative Study of the Ideal and Actual Surface Roughness in Finish Turning.
[17]Muammer, N., Hasan, G., Iahsan, T. (2007). Comparison of Regression and Artificial Neural Network Models for Surface
Roughness Prediction with the Cutting Parameters in CNC Turning. Modelling and Simulation in Engineering, Hindawi
Publishing Corp. New York, NY, United States, 3:2.
[18]Bajic, D., Lela, B., Zivkovic, D. (2008). Modeling of Machined Surface Roughness and Optimization of Cutting
Parameters in Face Milling, 0543-5846.
[19]Sakir, T., Sleyman, N., Ismail, S., Sleyman, Y. (2008). Prediction of Surface Roughness using Artificial Neural Network
in Lathe. International Conference on Computer Systems and Technologies - CompSysTech08.
61