OPTIMIZATION
OF BURR SIZE, SURFACE ROUGHNESS AND CIRCULARITY DEVIATION DURING DRILLING
OF AL 6061 USING TAGUCHI DESIGN METHOD AND ARTIFICIAL NEURAL NETWORK
Reddy Sreenivasulu
Department of Mechanical Engineering
R.V.R & J.C College of Engineering (A),
Guntur,Andhra Pradesh, India.
E-mail: rslu1431@gmail.com
Submission: 19/06/2014
Revision: 06/08/2014
Accept: 13/08/2014
ABSTRACT
This paper presents the influence of cutting parameters like cutting
speed, feed rate, drill diameter, point angle and clearance angle on the burr
size, surface roughness and circularity deviation of Al 6061 during drilling on CNC vertical machining center. A plan
of experiments based on Taguchi technique has been used to acquire the data. An
orthogonal array, signal to noise (S/N) ratio and analysis of variance (ANOVA)
are employed to investigate machining characteristics of Al 6061 using HSS twist drill bits of
variable tool geometry and maintain constant helix angle of 45 degrees.
Confirmation tests have been carried out to predict the optimal setting of process parameters to validate the used approach,
obtained the values of 0.2618mm, 0.1821mm, 3.7451µm, 0.0676mm for burr height,
burr thickness, surface roughness and circularity deviation
respectively. Finally, artificial neural network has been applied to compare
the predicted values with the experimental values, good agreement was shown
between the predictive model results and the experimental measurements.
Keywords: Al 6061 Alloy, Drilling, Taguchi
Design method, S/N ratio, ANOVA,
Artificial Neural Network.
1. INTRODUCTION
Burr is plastically deformed
material, obtained on the part edge during cutting or punching. Burr reduces
the precision of products and causes additional cost of de burring, the burrs
strongly influence the product quality and assembly. This additional procedure
results in the high cost of the edge finishing of precision parts. Since de
burring process is not well automated, the productivity of production systems
is often reduced.
Understanding the drilling burr
formation and its dominant parameters is essential for reducing burr. To
provide cost effectiveness in production and especially in machining
operations, there is a continuous necessity to decrease tooling costs. The most
well- automated methods used to decrease tooling costs are various applications
of more resistant cutting tool materials, heat treatment methods, cutting
fluids used, speed and feed rates, and the development of coated inserts on
cutting tools.
The surface roughness and roundness
error are influenced by several factors which include cutting tool geometry,
speed, and feed, structure of the work piece and the rigidity of the machine
tool. These parameters affecting the surface roughness and whole qualities
(roundness, cylindricality and whole diameter) can be optimized in various ways
such as Taguchi and multiple regression methods.
Therefore, a number of researchers
have been focused on an appropriate prediction of surface roughness and
roundness error. Taguchi method has been widely used in engineering analysis
and is a powerful tool to design a high quality system. By applying the Taguchi
technique, time for experimental investigations can be reduced, as it is
effective in the investigation of the influence of multiple factors on
performance as well as to study the significance of individual factors to
determine which factor has more influence, which one is less influence.
Yang and Chen (2001) used the
Taguchi parameter design in order to identify optimum surface roughness performance
on an aluminum material with cutting parameters of depth of cut, cutting speed,
feed rate and tool diameter. It reveals that tool diameter is not a significant
cutting factor influence on the surface roughness.
Davim and Reis (2003) presented an
approach using the Taguchi method and ANOVA to establish a correlation between
cutting speed and feed rate with the de lamination in a fiber reinforced
composite. A statistical analysis of whole quality was performed by Furness, Wu
and Ulsoy (1996) they found that feed rate and cutting speed have a relatively
small effect on the measured whole quality features.
Tsao and Hocheng (2008) performed
the prediction and evaluation of thrust force and surface roughness in drilling
of fiber reinforced composite, the approach used is Taguchi and the neural
network methods. The experimental results reveal that the feed rate and the
drill diameter are the most significant factors affecting the thrust force,
while the feed rate and spindle speed less significant but percentage
contribution is more on the surface roughness.
Nalbant, Gokkaya and Sur (2007)
utilized the Taguchi technique to determine the optimal machining parameters
for surface roughness in turning of AISI 1030 steel with Ti N coated inserts on
cutting tool. Three parameters such as radius of insert on cutting tool, feed
rate, and depth of cut\are optimized for surface roughness.
Kurt, Bagci and Kaynak (2009)
applied the Taguchi method in the optimization of process parameters for
surface roughness and roundness error in dry drilling processes. The objective
of this study is to investigate the effects of the drilling parameters on burr
size, surface roughness average and circularity deviation and is to determine
the optimal drilling parameters using the Taguchi design method and compared
the results with neural network.
2. Experimental Procedure:
2.1.
Material
Al 6061 is one of the 6000 series
aluminum alloy used in the aircraft and aerospace components, marine fittings,
bicycle frames, camera lenses, brake components, electrical fittings and
connectors, valves, couplings etc. The composition of Al 6061 is 0.63% Si,
0.096% Cu, 0.091% Zn, o.466% Fe, 0.179% Mn, 0.53% Mg, 0.028% Ti, 0.028% Cr, and
remaining aluminum. The young’s modulus is 80 G pa and hardness 98 BHN. In this
study 600x50x10mm rectangular bar was used.
2.2.
Schematic
machining:
In this study, the experiments were
carried out on a CNC vertical machining center (KENT and ND Co. Ltd, Taiwan make) shown in figure.1 to perform different size of
holes on Al6061 work piece by alter the point and clearance angles on standard
HSS twist drill bits and maintain constant helix angle of 45 degrees.
Furthermore the cutting speed (m/min), the feed rate (mm/rev) and percentage of
cutting fluid mixture ratio are regulated in this experiment.
Figure 1: Drilling
of Aluminium 6061 alloy
2.3.
Measuring
Apparatus
The burr size (thickness and height)
is measured by digital profile projector. After measuring the burr size, the
burr is removed then measured surface roughness and circularity deviation of
drilled hole by a surface analyser of Talysurf 50 (Taylor Hobson Co Ltd) and
coordinate measuring machine respectively.
3. MOTIVATION OF THE PRESENT WORK
3.1.
Methodology
The orthogonal array forms the basis
for the experimental analysis in the Taguchi method. The selection of
orthogonal array is concerned with the total degree of freedom of process
parameters. Total degree of freedom (DOF) associated with five parameters is
equal to 10 (5X2).The degree of freedom for the orthogonal array should be
greater than or at least equal to that of the process parameters.
There by, a L27 orthogonal array
having degree of freedom equal to (27-1) 26 has been considered, which is used to
optimize the cutting parameters for burr size, surface roughness and
circularity deviation using the S/N ratio and ANOVA for machining of Al 6061
alloy and predicted results were nearer to the experimental results.
Although similar to design of
experiment (DOE), the Taguchi design only conducts the balanced (orthogonal)
experimental combinations, which makes the Taguchi design even more effective
than a fractional factorial design. By Taguchi techniques, industries are able
to greatly reduce product development cycle time for design and production,
therefore reducing costs and increasing profit.
Also neural network technique has
been applied to compare the predicted values with the experimental values and
compared the error between experimental values. Finally, confirmation test have
been carried out to compare the predicted values with the experimental values
confirm its effectiveness in the analysis of burr size, surface roughness and
circularity deviation.
3.2.
Experimentation
as per Taguchi Design Method
A plan of experiments based on
Taguchi technique has been used to acquire the data. An orthogonal array,
signal to noise (S/N) ratio and analysis of variance (ANOVA) are employed to
investigate the drilling characteristics of Al 6061 alloy using HSS twist drill
bits. The complete procedure in Taguchi design method can be divided into three
stages: system design, parameter design, and tolerance design.
Of the three design stages, the
second stage – the parameter design – is the most important stage. Taguchi’s
orthogonal array (OA) provides a set of well-balanced experiments (with less
number of experimental runs), and Taguchi’s signal-to-noise ratios (S/N), which
are logarithmic functions of desired output in the optimization process.
Taguchi method uses a statistical measure of performance called signal-to-noise
ratio.
The S/N ratio takes both the mean
and the variability into account. The S/N ratio is the ratio of the mean
(Signal) to the standard deviation (Noise). The ratio depends on the quality
characteristics of the product/process to be optimized. The machining
parameters and their levels are given in Table1. Plan of experiments based on
Taguchi orthogonal array and observed responses shown in Table 2.
Table1: Machining parameters and their levels
LEVELS |
FACTORS |
||||
CUTTING
SPEED (rpm) |
FEED RATE (mm/min) |
DRILL DIAMETER (mm) |
POINT
ANGLE (Degrees) |
CLEARANCE
ANGLE (Degrees) |
|
A |
B |
C |
D |
E |
|
1 |
600 |
0.3 |
8 |
118 |
4 |
2 |
800 |
0.5 |
10 |
110 |
6 |
3 |
1000 |
0.6 |
12 |
100 |
8 |
Table 2: Plan of experiments based on Taguchi
orthogonal array and observed responses
RUNS |
TAGUCHI
RESPONSE DESIGN TABLE |
S/N Ratio |
||||||||
cutting
speed (rpm) |
Feed rate
(mm/min) |
drill dia (mm) |
point
angle (deg) |
clearance
angle (deg) |
burr
height (mm) |
burr
thickness (mm) |
surface
roughness (µm) |
circularity
deviation (mm) |
||
A |
B |
C |
D |
E |
R1 |
R2 |
R3 |
R4 |
||
1 |
1 |
1 |
1 |
1 |
1 |
0.268 |
0.178 |
2.39 |
0.058 |
-1.6278 |
2 |
1 |
1 |
1 |
1 |
2 |
0.254 |
0.166 |
1.16 |
0.063 |
4.4320 |
3 |
1 |
1 |
1 |
1 |
3 |
0.248 |
0.161 |
4.5 |
0.152 |
-7.0672 |
4 |
1 |
2 |
2 |
2 |
1 |
0.287 |
0.208 |
1.25 |
0.064 |
3.7360 |
5 |
1 |
2 |
2 |
2 |
2 |
0.258 |
0.168 |
3.36 |
0.048 |
-4.5433 |
6 |
1 |
2 |
2 |
2 |
3 |
0.264 |
0.197 |
3.72 |
0.127 |
-5.4292 |
7 |
1 |
3 |
3 |
3 |
1 |
0.238 |
0.149 |
4.05 |
0.027 |
-6.1495 |
8 |
1 |
3 |
3 |
3 |
2 |
0.347 |
0.241 |
3.45 |
0.036 |
-4.8008 |
9 |
1 |
3 |
3 |
3 |
3 |
0.242 |
0.184 |
2.29 |
0.174 |
-1.2765 |
10 |
2 |
1 |
2 |
3 |
1 |
0.318 |
0.243 |
3.33 |
0.088 |
-4.4935 |
11 |
2 |
1 |
2 |
3 |
2 |
0.222 |
0.159 |
2.25 |
0.109 |
-1.0965 |
12 |
2 |
1 |
2 |
3 |
3 |
0.328 |
0.218 |
1.06 |
0.122 |
4.9026 |
13 |
2 |
2 |
3 |
1 |
1 |
0.228 |
0.156 |
3.26 |
0.019 |
-4.2749 |
14 |
2 |
2 |
3 |
1 |
2 |
0.200 |
0.151 |
3.60 |
0.041 |
-5.1270 |
15 |
2 |
2 |
3 |
1 |
3 |
0.187 |
0.164 |
1.56 |
0.132 |
2.0188 |
16 |
2 |
3 |
1 |
2 |
1 |
0.324 |
0.228 |
3.54 |
0.026 |
-5.0137 |
17 |
2 |
3 |
1 |
2 |
2 |
0.219 |
0.147 |
2.45 |
0.094 |
-1.8190 |
18 |
2 |
3 |
1 |
2 |
3 |
0.244 |
0.220 |
4.38 |
0.085 |
-6.8348 |
19 |
3 |
1 |
3 |
2 |
1 |
0.214 |
0.189 |
2.89 |
0.066 |
-3.2417 |
20 |
3 |
1 |
3 |
2 |
2 |
0.209 |
0.191 |
2.91 |
0.100 |
-3.3032 |
21 |
3 |
1 |
3 |
2 |
3 |
0.264 |
0.233 |
3.41 |
0.107 |
-4.6847 |
22 |
3 |
2 |
1 |
3 |
1 |
0.254 |
0.212 |
3.11 |
0.089 |
-3.8870 |
23 |
3 |
2 |
1 |
3 |
2 |
0.229 |
0.252 |
3.02 |
0.141 |
-3.6437 |
24 |
3 |
2 |
1 |
3 |
3 |
0.196 |
0.163 |
1.65 |
0.182 |
1.5171 |
25 |
3 |
3 |
2 |
1 |
1 |
0.186 |
0.152 |
2.72 |
0.072 |
-2.7075 |
26 |
3 |
3 |
2 |
1 |
2 |
0.223 |
0.169 |
3.46 |
0.111 |
-4.7936 |
27 |
3 |
3 |
2 |
1 |
3 |
0.241 |
0.198 |
3.59 |
0.105 |
-5.1176 |
3.3.
Analysis
of the S/N Ratio
In the Taguchi method, the term
‘signal’ represents the desirable value (mean) for the output characteristic
and the term ‘noise’ represents the undesirable value (Standard Deviation) for
the output characteristic. Therefore, the S/N ratio to the mean to the S.D. S/N
ratio used to measure the quality characteristic deviating from the desired value.
The S/N ratio (h) is defined as
h =
-10 log (M.S.D), Where M.S.D is the mean square deviation for the output
characteristic.
Table 2 shows the experimental
results for observed responses. The S/N ratio table for observed responses is
shown in Table 3.
Table3. Signal to Noise Ratios for Smaller is better
Level |
Cutting speed (rpm) A |
Feed rate (mm/min) B |
Drill diameter (mm) C |
Point angle (Deg) D |
Clearance angle (Deg) E |
1 |
-2.52518 |
-1.79783 |
-2.66049 |
-2.10312 |
-2.44130 |
2 |
-2.41537 |
-2.18147 |
-2.17144 |
-3.45934 |
-2.74395 |
3 |
-3.31802 |
-4.27928 |
-3.42665 |
-2.69612 |
-3.07333 |
Delta |
0.90265 |
2.48145 |
1.25522 |
1.35623 |
0.63203 |
Rank |
4 |
1 |
3 |
2 |
5 |
4. RESULTS AND DISCUSSIONS
From main effects plot of S/N ratio
for, the optimum parameters combination for burr height, burr thickness,
surface roughness and circularity deviation are A2B1C2D1E1 corresponding to the
largest values of S/N ratio for all control parameters. From Table 3, it is
observed that feed rate, point angle, drill diameter, cutting speed and
clearance angle has the order of influence on burr size, surface roughness and
circularity deviation during drilling of Al6061 alloy.
Figure 2: Main effects plot for S/N ratio
Figure 3: Interaction plot of burr
height with effect of other parameters
Figure 4: Interaction plot of burr
thickness with effect of other parameters
Figure 5: Interaction plot of
surface roughness with effect of other parameters
Figure 6: Interaction plot of
circularity deviation with effect of other parameters
4.1.
Results
of ANOVA
The purpose of the analysis of
variance (ANOVA) is to investigate which design parameters significantly affect
the quality characteristic.
Table 4 shows the results of ANOVA
for burr height, cutting speed, feed rate, drill diameter, point angle and
clearance angle are the significant cutting parameters for affecting the burr
height.
Table 5 shows the results of ANOVA
for burr thickness, cutting speed, feed rate, drill diameter, point angle and
clearance angle are the significant cutting parameters for affecting the burr
thickness.
Table 6 shows the results of ANOVA
for surface roughness, cutting speed, feed rate, drill diameter, point angle
and clearance angle are the significant cutting parameters for affecting the
surface roughness.
Table 7 shows the results of ANOVA
for circularity deviation, cutting speed, point angle and clearance angle are
the significant cutting parameters for affecting the circularity deviation.
Table4: Results of ANOVA for burr height
Symbol |
Cutting
Parameters |
DOF |
SS |
MS |
F |
|
A |
Cutting
speed |
2 |
0.00871 |
0.00435 |
36.25 |
significant |
B |
Feed
rate |
2 |
0.00292 |
0.00146 |
12.16 |
significant |
C |
Drill
diameter |
2 |
0.00218 |
0.00109 |
9.08 |
significant |
D |
Point
angle |
2 |
0.00684 |
0.00342 |
28.5 |
significant |
E |
Clearance
angle |
2 |
0.00140 |
0.00070 |
5.83 |
significant |
Error |
|
16 |
0.001926 |
0.00012 |
|
|
Total |
|
26 |
0.023976 |
|
|
|
Significant, F table at
95%confidence level is F0.05, 2, 16 = 3.63, F exp ≥ F
table
Table 5: Results of ANOVA for burr thickness
Symbol |
Cutting
Parameters |
DOF |
SS |
MS |
F |
|
A |
Cutting
speed |
2 |
0.0066 |
0.0033 |
16.75 |
significant |
B |
Feed
rate |
2 |
0.0027 |
0.0013 |
6.598 |
significant |
C |
Drill
diameter |
2 |
0.0029 |
0.0015 |
7.614 |
significant |
D |
Point
angle |
2 |
0.00702 |
0.00351 |
17.766 |
significant |
E |
Clearance
angle |
2 |
0.0053 |
0.0027 |
13.705 |
significant |
Error |
|
16 |
0.00315 |
0.000197 |
|
|
Total |
|
26 |
0.02765 |
|
|
|
Significant, F table at
95%confidence level is F0.05, 2, 16 = 3.63, F exp ≥ F
table
Table 6: Results of ANOVA for surface
roughness
Symbol |
Cutting
Parameters |
DO
F |
SS |
MS |
F |
|
A |
Cutting
speed |
2 |
2.96 |
1.48 |
3.797 |
significant |
B |
Feed
rate |
2 |
4.44 |
2.22 |
5.696 |
significant |
C |
Drill
diameter |
2 |
3.40 |
1.7 |
4.362 |
significant |
D |
Point
angle |
2 |
3.76 |
1.88 |
4.824 |
significant |
E |
Clearance
angle |
2 |
3.43 |
1.715 |
4.4 |
significant |
Error |
|
16 |
6.2353 |
0.3897 |
|
|
Total |
|
26 |
23.3653 |
|
|
|
Significant, F table at
95%confidence level is F0.05, 2, 16 = 3.63, F exp ≥ F
table
Table 7: Results of ANOVA for circularity
deviation
Symbol |
Cutting
Parameters |
DOF |
SS |
MS |
F |
|
A |
Cutting
speed |
2 |
0.00584 |
0.00292 |
3.74 |
significant |
B |
Feed
rate |
2 |
0.00177 |
0.00885 |
1.13 |
Insignificant |
C |
Drill
diameter |
2 |
0.00215 |
0.00107 |
1.37 |
Insignificant |
D |
Point
angle |
2 |
0.00579 |
0.00289 |
3.71 |
significant |
E |
Clearance
angle |
2 |
0.02307 |
0.01153 |
14.78 |
significant |
Error |
|
16 |
0.01248 |
0.00078 |
|
|
Total |
|
26 |
0.0511 |
|
|
|
Significant, F table at
95%confidence level is F0.05, 2, 16 = 3.63, F exp ≥ F
table
Table 8: Optimal values of individual
machining characteristics
Machining
characteristics |
Optimal
combination of parameters |
Significant parameters(at 95% confidence level) |
Predicted
optimum value |
Experimental
value |
Burr height
(R1) |
A1B1C2D1E3 |
A,B,C,D,E |
0.2618 |
0.2622 |
Burr
thickness (R2) |
A3B1C1D1E1 |
A,B,C,D,E |
0.1821 |
0.1843 |
Surface
Roughness (R3) |
A3B3C3D2E3 |
A,B,C,D,E |
3.7451 |
3.7378 |
Circularity
deviation(R4) |
A3B1C1D1E1 |
A,D,E |
0.0676 |
|
Confirmatory experiments were
conducted for burr size, surface roughness and circularity deviation,
corresponding their optimal setting of process parameters to validate the used
approach, obtained the values of 0.2618mm, 0.1821mm, 3.7451µm, 0.0676mm for
burr height, burr thickness, surface roughness and circularity deviation
respectively. Predicted and experimental values of responses are depicted in Table
8.
5. ARTIFICIAL NEURAL NETWORK
Artificial neural systems are that
physical cellular systems which acquire store and utilize experimental
information. Powerful learning algorithm and self-organizing rule allow ANN to self-adapt
as per the requirements in continually varying environment (adaptability
property). The ANN architecture is a multilayer, feed forward back propagation
architecture.
Multilayer perception (MLP) has an
input layer, output layer and hidden layer. Input vector is incident on input
layer and then to hidden layer and subsequently to final layer/output layer via
weighted connections. Each neuron operates by taking the sum of its weighted
inputs and passing the results through a non-linear activation function. A neural network is a machine that is
designed to model the way in which the brain performs a particular task or
function of interest.
To achieve good performance, they
employ a massive interconnection of simple computing cells referred to as
‘Neurons’ or ‘processing units’. Hence a neural network viewed as an adaptive
machine can be defined as a neural network is a massively parallel distributed
processor made up of simple processing units, which has a natural propensity
for storing experimental knowledge and making it available for use.
The experimental observations were
incorporated into the ANN model. A feed forward neural network was developed to
predict burr size, surface roughness and circularity deviation. As in future
without undergoing the machining process able to get good machining data’s and
its very useful ANN model for getting good optimum machining process.
Figure 7: Artificial Neural
Network diagram
Artificial
Neural Network (ANN) Response Graphs:
Figure 8: Response graph of burr height for Experimental Vs ANN
Figure 9: Response graph of burr thickness for
Experimental Vs ANN
Figure 10: Response graph of surface roughness for Experimental
Vs ANN
Figure 11: Response graph of
circularity deviation for Experimental Vs ANN
From
ANN response graphs shown in figure no.8 to11, it is observed that except
experiment no 13 in the case of circularity deviation, remaining experimental
runs obtained less deviation for all responses. Increasing the number of nodes
increases the computational cost and decreases the error. Good agreement was
shown between the predictive model results and the experimental measurements.
6. CONCLUSIONS
The machining characteristics of
Al6061 alloy have been studied. The primary machining characteristics such as
burr size, surface roughness and circularity deviation were studied for
drilling. The results obtained from the experiments as follows.
From S/N Ratio response graph, the
combination of parameters having the values of 800 rpm, 0.3 mm/min ,10mm.118
degrees and 4 degrees obtained for cutting speed, feed rate drill diameter,
point angle and clearance angle respectively for optimizing burr size, surface
roughness and circularity deviation.
From S/N Ratio response table, feed
rate, point angle, drill diameter, cutting speed and clearance angle has the
order of influence on burr size, surface roughness and circularity deviation
during drilling of Al6061 alloy.
From the results of ANOVA for
cutting speed, feed rate, drill diameter, point angle and clearance angle, all
parameters are significant for all responses except circularity deviation, for
this cutting speed, drill geometry are more significant ,
From results of ANN, it is concluded
that experiment No13 obtained relatively more error than remaining. The
deviation between experimental values and prediction values are found in the
range of 3 to 4%. Finally, concluded that increasing the number of nodes
increases the computational cost and decreases the error. Good agreement was
shown between the predictive model results and the experimental measurements.
REFERENCES
DAVIM, J. P.; REIS, P. (2003). Drilling
carbon fiber reinforced plastic manufactured by auto clave experimental and
statistical study. Materials and Design,
v. 24, n. 5, p. 315-324,
FURNESS, R. J.; WU, C. L.; ULSOY, A. G. (1996).
Statistical analysis of the effects of feed, speed, and wear of hole quality in
drilling. Journal of Manufacturing
Science and Engineering, v. 118, n. 3, p. 367-375
KURT, M.; BAGCI, E.; KAYNAK, Y. (2009). Application of
Taguchi methods in the optimization of cutting parameters for surface finish
and hole diameter accuracy in dry drilling processes. International Journal of
Advanced Manufacturing Technology, v.40, n. 5-6, p. 458-469
NALBANT, M.; GOKKAYA, H.; SUR, G. (2007). Application
of Taguchi method in the optimization of cutting parameters for surface
roughness in turning. Materials and
Design, v. 28, n. 4, p. 1379-1385.
TSAO, C. C.; HOCHENG, H. (2008). Evaluation of thrust
force and surface roughness in drilling composite material using Taguchi
analysis and neural network. Journal of
Materials Processing Technology, v. 203, n. 1-3, p. 342-348.
YANG, J. L.; CHEN, J. C. (2001). A systematic approach
for identifying optimum surface roughness performance in end-milling
operations. Journal of Industrial
Technology, v. 17, n. 2, p. 1-8.