International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882

Volume 4, Issue 8, August 2015

AN APPLICATION OF ARTIFICIAL NEURAL NETWORK TO

PREDICT SURFACE ROUGHNESS DURING DRILLING OF AISI 1020
STEEL
Abhinaw Roy & Mr. Kulbhushan Bhagat
Student, Suresh Gyan Vihar University, Jaipur Rajasthan
Assistant Professor, Suresh Gyan Vihar University, Jaipur, Rajasthan

ABSTRACT
Metal machining has been an interesting topic of
research for decades. Early research efforts include
topics, such as, chip formation, heat generation at toolchip and tool-work interfaces, tool wear, tool life,
cutting force analysis, machinability, power
requirements, manufacturing cost and productivity etc.
With more emphasis on quality of surface finish with
high production rate, in past few decades, research
efforts seem to have been diverted towards the study of
surface finish/roughness in metal machining.
The objectives of the present study is the development
of the surface roughness prediction model using
Artificial neural network in terms of feed, cutting
speed and depth of hole during the drilling of AISI
1020 steel and to Investigate the effect of drilling
parameters on surface roughness using 3D plots
between the parameters and surface roughness .
In the present research a multilayer feed-forward
neural network has been employed using MATLAB
software. The input layer has four neurons
corresponding to each of the three input variables
(cutting speed, feed and depth of cut) and one neuron
in the output layer corresponding to output variable
(surface roughness). The number of hidden layers and
neurons in hidden layers has been determined on the
basis of trial and error method. The mean square error
has been selected as performance function. The ANN
training simulation was carried out using batch
gradient descent with variable learning rateprocedure
traingdx. This procedure helps in improving the
performance of back -propagation algorithm by
allowing the learning rate to change during the training
process based on the complexity of the local error
surface, In this research, a portable surface roughness
tester has been used to measure surface roughness
indicators of finished work pieces.
Keywords: Artificial neural
roughness, AISI 1020 steel.

I.

network,

surface

INTRODUCTION

In the manufacturing industries, various machining

processes (turning, milling, drilling etc.) are adopted
for removing the material from the workpiece to obtain
finished product. Among the various metal removing

processes, drilling is the one of most important metal

removing process as compared to other traditional
machining processes. Drilling is use for making the
hole in the workpieces. Hole making is a most
important machining process in manufacturing. During
the drilling, the drill rotates and fed into the work piece
to remove the material in the form of chips that move
along the fluted shank of the drill.Different drilling
tools and hole making methods are used for drilling.
The selection of different tools and methods depends
on the type of workpiece, size of the hole, the quantity
of holes, and the quantity of the holes in given time
periods. Also, drilling is used as an initial process for
many machining operations, such as reaming, tapping
and boring. Drilling process is widely used in the
aerospace, aircraft and automotive industries. Now
days, due to global competitiveness, manufacturing
industries are more concerned about the quality of their
products with high metal removal rate. These focus on
producing high quality products in time at minimum
cost.
The surface roughness is considered to be a measure of
the technological quality of a product. Surface
roughness is the one of the critical performance
parameter that has an appreciable effect on several
mechanical properties of machined parts such as
fatigue behaviour, corrosion resistance, creep life, etc.
It also affects other functional attributes of machined
parts like friction, wear, light reflection, heat
transmission, lubrication, electrical conductivity, etc.
Hence, achieving the desired surface quality is of great
importance for the functional behaviour of the
mechanical parts.
During the drilling of the workpiece, it has long been
recognized that the drilling conditions (drill point
geometry, drill and workpiece materials, drilling
parameters like feed rate and spindle speed) affect the
performance of the operation to a greater extent. These
drilling conditions should be selected to optimize the
economics of drilling operations. So it can be achieved
by empirical modeling of performance as a function of
machining conditions using artificial neural network.
The proposed work will be employed for investigate
the effect of drilling parameters on surface roughness
and development of surface roughness prediction
model using artificial neural network (ANN).

www.ijsret.org

853

International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
Volume 4, Issue 8, August 2015

traingdx. This procedure helps in improving the

performance of back -propagation algorithm by
allowing the learning rate to change during the training
process based on the complexity of the local error
surface.

V.

EXPERIMENTAL WORK

The process parameters that were chosen for

experimentation are given as under:
1. Spindle speed (RPM)
2. Feed (mm/rev.)
3. Depth of hole (mm)
MACHINE USED:
The drilling operations have been carried out on a
CNC MILL MT250 Machining Center

Fig 1: Drilling process

II.

OBJECTIVES
Development of the surface roughness
prediction model using Artificial neural
network in terms of feed, cutting speed and
depth of hole during the drilling of AISI 1020
steel.
Investigating the effect of drilling parameters
on surface roughness using 3D plots between
the parameters and surface roughness.
To investigate the most significant parameters
that affects the surface roughness.

III.

ARTIFICIAL NEURAL NETWORK

Artificial neural network (ANN) is a most attractive

branch of artificial intelligence. It is based on the basic
model of the human brain with the capability of
generalization and learning. An ANN has the potential
to handle problems such as modeling, estimating,
prediction, optimization, diagnosis, and adaptive
control in complex nonlinear systems. It can acquire,
store, and utilize knowledge gained from experience.
The capabilities of ANNs in capturing the
mathematical mapping between input variables and
output features are of primary significance for
modeling machining processes.

IV.

PROPOSED
MODELS

NEURAL

Fig 2: CNC MILL MT250 Machining Center used for

drilling
CUTTING INSERTS:
Coated carbide tool performs better than uncoated
carbide tools. On this basis, commercially accessible
K- series, tungsten carbide high speed core drills with
Tialn coating having 10 mm diameter with two flute
has been used for experimentation.

NETWORK

In the present research a multilayer feed-forward

neural network has been employed using MATLAB
software. The input layer has four neurons
corresponding to each of the three input variables
(cutting speed, feed and depth of cut) and one neuron
in the output layer corresponding to output variable
(surface roughness). The number of hidden layers and
neurons in hidden layers has been determined on the
basis of trial and error method. The mean square error
has been selected as performance function. The ANN
training simulation was carried out using batch
gradient descent with variable learning rateprocedure

Fig 3: Tungsten carbide high speed core drills

COOLANT:
Coolant has been used in all the experiments.
SUPERCUT cutting oil by SHELL COMPANY has
been used in the ratio of 20:1 i.e 20 litres of water and
1 litre of cutting oil in it.
S/R MEASUREMENT

www.ijsret.org

854

International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
Volume 4, Issue 8, August 2015

Surface roughness is defined as the finer irregularities

of the surface texture that usually result from the
inherent action of the machining process or material
condition. There are many parameters used related to
surface roughness in literatures. In this research, a
portable surface roughness tester (Model No TR 210
manufactured by Beijing TIME High Technology Ltd.
Beijing City, China) has been used to measure surface
roughness indicators of finished work pieces.

5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

600
300
600
600
300
900
600
300
600
900
900
900
900
600
300

0.01
0.02
0.03
0.02
0.01
0.01
0.03
0.01
0.02
0.02
0.03
0.03
0.03
0.03
0.02

20
10
10
10
15
15
20
20
20
20
10
20
15
15
15

1.021
1.207
1.133
1.011
1.301
1.311
1.527
1.485
1.229
1.782
1.608
2.182
1.811
1.342
1.474

0.25258
0.37232
0.32468
0.24614
0.43283
0.43927
0.57833
0.55129
0.38648
0.74249
0.63047
1.00000
0.76116
0.45923
0.54421

10
20
10
15

1.259
1.972
1.182
1.199

0.40579
0.86481
0.35622
0.36717

Testing data set

20
21
22
23

Fig 4: surface roughness tester

FORMATION OF DESIGN MATRIX:
S.n
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

Spindle
speed
600
300
900
900
600
300
600
600
300
900
600
300
600
900
900
900
900
600
300
900
300
300
600
300
900
600
300

feed
0.01
0.02
0.01
0.02
0.01
0.02
0.03
0.02
0.01
0.01
0.03
0.01
0.02
0.02
0.03
0.03
0.03
0.03
0.02
0.02
0.03
0.01
0.02
0.03
0.01
0.01
0.03

Depth
of hole
15
20
10
15
20
10
10
10
15
15
20
20
20
20
10
20
15
15
15
10
20
10
15
10
20
10
15

DEVELOPMENT OF ANN
SURFACE ROUGHNESS:
SN

Spindle
speed

feed

Depth
Of
hole

S/R

1
2
3
4

600
300
900
900

0.01
0.02
0.01
0.02

15
20
10
15

0.982
1.653
1.118
1.498

24
25
26
27

0.03
0.01
0.01
0.03

10
20
10
15

1.54
1.438
0.784
1.774

Training parameters

Values

Network configuration

4-11-1

Number of hidden layer

Maximum number of
epochs to train
Performance goal
Learning rate

NOR
S/R

0.22747
0.65944
0.31502
0.55966

300
900
600
300

0.58670
0.52103
0.10000
0.73734

Neural network design and training:

The simulation for ANN training was conducted using
batch gradient descent with variable learning
rateprocedure traingdx. The trial and error method
was used for values of training parameters for
traingdx and is shown in table.

FOR

Training data sets

0.02
0.03
0.01
0.02

Validation data set

Surface
roughnes
0.982
1.653
1.118
1.498
1.021
1.207
1.133
1.011
1.301
1.311
1.527
1.485
1.229
1.782
1.608
2.182
1.811
1.342
1.474
1.259
1.972
1.182
1.199
1.54
1.438
0.784
1.774

MODEL

900
300
300
600

Ratio to increase learning

rate
Maximum validation
failures
Momentum constant
Minimum performance
gradient
Number of patterns used
for training
Number of patterns used
for testing
Number of patterns used
for validation

TRAINING OF MODEL:

www.ijsret.org

1
10000
0.00001
0.03
1.001
4
0.43
1.00E-04
19
4
4

855

International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
Volume 4, Issue 8, August 2015

VALIDATION OF MODEL:

Fig 5: Correlation between the experimental (Target)

and predicted (Output) surface roughness in training.

Fig 6: Comparison of the experimental and predicted

values of surface roughness in training.

Fig 9: Comparison of the experimental and predicted

values of surface roughness in validation

TESTING OF MODEL:
Performance of the network:

Fig 7: Correlation between the target and output

surface roughness in testing.
Fig 10: Correlation between the experimental (Target)
and predicted (Output) surface roughness for entire
data set
VALUES
S.n
1
2
3
4
5

Figure 8: Comparison of the experimental and

predicted values of surface roughness in testing.
www.ijsret.org

OF

Spindle
speed
600
300
900
900
600

PREDICTED
ROUGHNESS:
feed
0.01
0.02
0.01
0.02
0.01

Depth
of hole
15
20
10
15
20

EXP.
S/R
0.982
1.653
1.118
1.498
1.021

SURFACE
PRD.
S/R
0.926
1.699
1.093
1.503
1.044

856

International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
Volume 4, Issue 8, August 2015
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

300
600
600
300
900
600
300
600
900
900
900
900
600
300
900
300
300
600
300
900
600
300

0.02
0.03
0.02
0.01
0.01
0.03
0.01
0.02
0.02
0.03
0.03
0.03
0.03
0.02
0.02
0.03
0.01
0.02
0.03
0.01
0.01
0.03

10
10
10
15
15
20
20
20
20
10
20
15
15
15
10
20
10
15
10
20
10
15

1.207
1.133
1.011
1.301
1.311
1.527
1.485
1.229
1.782
1.608
2.182
1.811
1.342
1.474
1.259
1.972
1.182
1.199
1.54
1.438
0.784
1.774

1.237
1.077
1.021
1.361
1.295
1.574
1.467
1.272
1.742
1.613
2.156
1.841
1.334
1.403
1.271
1.886
1.116
1.207
1.568
1.480
0.839
1.794

Fig 13: Plot between depth of hole and

surfaceroughness

EFFECT OF DRILLING
PARAMETERS ON S/R:

Fig 14: 3D plot for surface roughness between spindle

speed and feed

Fig 11: Plot between spindle speeds on surface

roughness

Fig 15: 3D plot for surface roughness between spindle

speed and depth of hole.

Fig 12: Plot between feed on surface roughness

Fig: 3D plot for surface roughness between feed and
depth of hole
www.ijsret.org

857

International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
Volume 4, Issue 8, August 2015

VI.

CONCLUSION & FUTURE SCOPE

The objective of the present work is to optimize

drilling parameters for minimum surface roughness.
An attempt has also been made to investigate the
effects of the drilling parameters on surface roughness.
The important conclusions drawn from the present
work are summarized as follows:
1.
The predicted values of surface roughness
using proposed ANN model was found to be in close
agreement with the experimental data.
2.
The results so obtained clearly illustrates that
the as the spindle speed increase the surface roughness
decreases up to a certain value of spindle speed
afterward roughness increases with increase in spindle
speed
3.
The surface roughness increases with increase
in feed.
4.
The surface roughness increases with increases
in depth of hole.
5.
The minimum surface roughness is achieved at
low level of feed, low level of depth of hole and
middle level of spindle speed.

REFERENCESES
[1] Ahamed, A.R., Asokan, P., Aravindan, S. and
Prakash, M.K. (2010). Drilling of hybrid Al-5%SiCp5%B4Cp metal matrix composites, International
Journal of Advanced Manufacturing Technology,
49(9-12), 871-877.
[2] Bahce, E. and Ozel, C. (2013). experimental
investigation of the effect of machining parameters on
the surface roughness and the formation of built up
edge (BUE) in the drilling of Al 5005, Tribology in
Engineering, InTech, DOI: 10.5772/56027.
[3] Bai, B. H., Kumar, G.V. and Babu, K.A.(2014).
Optimization of surface roughness and burr height in
drilling of AL-6061 under MQL condition using DFA,
Mechanica Confab, 3(4), 10-17.
[4] Cicek, A. (2007), Step based geometric and
topological similarity assessment of mechanical parts,
Mathematical and Computational Applications, 12(3),
141-150.
[5] Cicek, A., Kivak, T. and Samtas, G. (2012).
Application of Taguchi method for surface roughness
and roundness error in drilling of AISI 316 Stainless
Steel, Journal of Mechanical Engineering, 58(3), 122129.
[6] Davim J.P. and Reis, P. (2003). Study delamination
in drilling carbon fiber reinforced plastics (CFRP)
using design experiments. University of Aveiro
Campus Santiago, Portugal. 59, 481-487.
[7] Gaitonde V.N. Karnik S.R., Acyutha B.T. and
Siddeswarappa B. (2007). Methodology of Taguchi
optimization for multi-objective drilling problem to
minimize burr size. International J. Advance
Manufacturing Technology, 34, 1-8.

[8] Gul T. and Muratoglu M. (2004). The drilling of

Al/SiCp metal-matrix composites. Part II: wokpiece
surface intergrity. Firat University, Turkey. 64, 14131418.
[9] Hsu, I. and Tsao C.C. (2008). Optimization of
process parameter in drilling composite materials by
Grey-Taguchi method. Tahua Insitute of Technology,
Taiwan.
[10]
Kaneriya,
N.G.
and
Sharma,
G.K.
(2014).Experimental evaluation and optimization of
dry drilling parameters of AISI304 austenitic stainless
steel using different twist drills,5th International &
26th All India Manufacturing Technology, Design and
Research Conference (AIMTDR 2014) December
12th14th, 2014, IIT Guwahati, Assam, India, 1-7.
[11] Karnik S.R., Gaitonde V.N. and Davim J.P.
(2008). Acomparative study of the ANN and RSM
modelling approaches for predicting burr size in
drilling. International J. Advance Manufacturing
Technology, 38, 868-883.
[12] Khan, I., Luthra, D., Verma, M. and Singh,
S.(2012). Selection of optimum drilling parameter in
drilling of commercial acrylic sheet to achieve
minimum hole expansion by using Taguchi approach,
International Journal of Engineering Science and
Technology 4(4), 1256-1261.
[13] Khashaba, U.A., Seif, M.A. and Elhamid, M.A.
(2007). Drilling analysis of chopped composites,
Composites Part A, Applied Science and
Manufacturing, 38(1), 61-70.
[14] Kivak, T., Habali, K.and Seker, U. (2012). The
effect of cutting paramaters on the hole quality and
tool wear during the drilling of Inconel 718, Gazi
University Journal of Science, 25(2),533-540.
[15] Kumar, D., Singh, L.P. and Singh, G.(2012).
Operational modeling for optimizing surface
roughness in mild steel drilling using Taguchi
technique, International Journal of Research in
Management. 2(3) 66-77.
[16] Kurt M., Kaynak Y., Bakr B., KokluU., Atakok
G. and Kutlu L. (2009). Experimental investigation
and Taguchi optimization for the effect of cutting
parameters on the drilling of Al 2024-T4 alloy with
diamond like carbon (DLC) coated Drills, Sixth
International Advanced Technologies Symposium
(IATS09), Karabuk, Turkey.
[17] Kurt, M., Bagci, E. and 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 J.
Advance Manufacturing Technology, 40, 458-469
[18] Lin, S.C. and Chen, I.K.(1999). Drilling of carbon
fiber-reinforced composite material at high speed,
Wear, 194, 156162.
[19] Madiwal,S. (2006). Analysis of surface finish in
milling of composite using neural Networks, Wichita
State University.2006.

www.ijsret.org

858

International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
Volume 4, Issue 8, August 2015

[20] Mohan, N.S., Kulkarni, S.M. and Ramachandra,

A. (2007). Delamination analysis in drilling process of
GFRP composite materials. J. Mater. Process.
Technol., 186, 265-271
[21] Navanth, A. and Sharma,K. (2013). A study of
TAGUCHI method based optimization of drilling
parameter in dry drilling of AL 2014 alloy at low
speeds, International Journal of Engineering Sciences
& Emerging Technologies, 6(1) 65-75.
[22] Neseli, S.(2014). Optimization of process
parameters with minimum thrust force and torque in
drilling operation using Taguchi method, Advances in
Mechanical Engineering, 2014, 1-10.
[23] Noorul Haq A., Marimuthu P. and Jeyapaul J.
(2008). Multi response optimizatio of machining
parameters of drilling Al/SiC metal matrix composite
using grey relational analysis in the Taguchi method.
International J. Advance Manufacturing Technology.
37, 250-255.
[24] Nouari M., List G., Girot F. and Gehin D. (2005).
Effect of machining parameters and coating on wear
mechanisms in dry drilling of aluminum alloys.
International Journal of Machine Tools &
Manufacture. 45, 14361442.
[25] Shunmugesh, K., Panneerselvam, K. and Thomas,
J. (2014). Optimising drilling parameters of GFRP by
using grey relational analysis, International Journal of
Research in Engineering and Technology,3(6) 302305.
[26] Siddiquee, A.N., Khan, Z.A., Goel, P. Kumar,
M., Agarwal, G. and Khan, N.Z.(2014). Optimization
of deep drilling process parameters of AISI 321 steel
using Taguchi method, Procedia Materials Science, 6,
1217 1225.
[27] Sundeep, M., Sudhahar, M., Kannan, T. T. M.,
Kumar,V. and Parthipan,N. (2014).Optimization of
drilling parameters on austenitic stainless steel (AISI
316) using TAGUCHI'S methodology, International
Journal of Mechanical Engineering and Robotics
Research, 3(4),388-394.
[28] Tamilselvan and Raguraj (2014). Optimization of
process parameters of drilling in Ti-Tib composites
using Taguchi technique, International Journal on
Mechanical Engineering and Robotics, 2(4), 1-5.
[29]Tosun, N. (2006). Determination of optimum
parameters for multi-performance characteristics in
drilling by using grey relational analysis. International
J. Advance Manufacturing Technology. 28, 450-455.
[30] Tsao C.C. (2006). Taguchi analysis of drilling
quality associated with core drill in drilling of
composite material. International J. Advance
Manufacturing Technology,32, 877-884.
[31] Tyagi, Y., Chaturvedi, V. and Vimal, J. (2012).
Parametric optimization of drilling machining process
using Taguchi design and ANOVA approach, Journal
of Emerging Technology and Advanced Engineering,
2 (7), 339-347.
www.ijsret.org

859