Vol. 38, No.

4 (2016) 454-462

Tribology in Industry

RESEARCH
www.tribology.fink.rs

Multi-objective Optimization of Process

Performances when Cutting Carbon Steel with
Abrasive Water Jet

M. Radovanović a
a University of Niš, Faculty of Mechanical Engineering, Aleksandra Medvedeva 14, 18000 Niš, Serbia.

Keywords: ABSTRACT
Аbrasive water jet cutting
Multi-objective optimization of process performances (perpendicularity
Мulti-objective optimization
deviation, surface roughness and productivity) when cutting carbon
Perpendicularity deviation surface
steel EN S235 with abrasive water jet is presented in this paper. Cutting
Roughness
factors (abrasive flow rate, traverse rate and standoff distance) were
Productivity
determined when perpendicularity deviation and surface roughness are
minimal and productivity is maximal. Multi-objective genetic algorithm
Corresponding author: (MOGA) was used for the determination set of nondominated optimal
Miroslav Radovanović points, known as Pareto front.
University of Niš, © 2016 Published by Faculty of Engineering
Faculty of Mechanical Engineering,
Aleksandra Medvedeva 14,
18000 Niš, Serbia.
E-mail: mirado@masfak.ni.ac.rs

1. INTRODUCTION versatility and has been proven to be an

effective technology for machining various
There are some machining technologies for engineering materials and wide range of
contour cutting metal or non metal sheets. thickness. Тhickness which can be cut are 100
Flame cutting, plasma cutting, laser cutting, wire mm for stainless steel and 120 mm for
EDM and water jet cutting are more applicable aluminum, for water pressure of 400 MPa. With
machining technologies for contour cutting AWJ it is possible to cut random contours in
sheets. Abrasive water jet (AWJ) cutting is a thick, middle and thin sheets, such as very fine
modern nonconventional machining technology tabs and filigree structures. Tolerances of 0.1
that enables contour cutting a wide range of mm can be realized in metal cutting. [1]
materials and thickness. AWJ cutting has
advantages over the other contour cutting Process of AWJ cutting is based on material
technologies such as: no thermal distortion, no removal from the workpiece by erosion. High
heat affected zone, small cutting force, minimum speed water jet stream accelerates abrasive
stress on the workpiece, high flexibility, high particles and those particles erode the material.

454
 M. Radovanović, Tribology in Industry Vol. 38, No. 4 (2016) 454-462

In AWJ cutting high pressure pump via carbide. Water is pressed out of the orifice in
accumulator and high pressure tubing directs form of very thin jet at a high speed. Speed of
the pressurized water to the cutting head. Types water jet can calculate using equation:
of pumps and levels of pressure are classified by
water jet industry as: 2p . (1)
v wj  
w
 Low pressure, p<10000 psi (p<690 bar,
p<69 MPa) where are: vwj-speed of water jet in m/s, -
 Medium pressure, 10000<p<15000 psi correction factor (0.9-0.98), p-water pressure in
(690<p<1030 bar, 69<p<103 MPa) Pa, w-density of water in kg/m3.

 High pressure, 15000<p<40000 psi Abrasive particles are added to water jet in
(1030<p<2760 bar, 103<p<276 MPa) mixing chamber of the cutting head. There are
 Ultra high pressure, 40000<p<75000 psi two methods to add abrasive particles to water
(2760<p<5170 bar, 276<p<517 MPa) jet: suspension and injection. The suspension
method (direct, indirect or bypass) is used in
 Hyper pressure, p>75000 psi (p>5170 bar, industry only at pressures of up to 70 MPa.
p>517 MPa). Today the injection method is mainly used for
industrial applications with operating pressures
There are two types of high pressure pumps: of up to 400 MPa. The high velocity of the water
direct drive pumps and intensifier pumps. Direct jet creates a Venturi effect or vacuum in the
drive pumps generally are found in industrial mixing chamber located immediately beneath
applications with pressure to 380 MPa. These the orifice. Abrasive particles are metered from
pumps use an electric motor to turn a crankshaft a mini-hopper through a plastic tube down to
that moves three or more pistons that create the the cutting head and are sucked into the water
water pressure. But, the most common pumps in jet stream in the mixing chamber. Abrasive
industrial applications are intensifier pumps. particles are accelerated with high speed water
Intensifier pumps are called intensifiers because jet. Abrasive particles are mixed with the water
they use the concept of pressure intensification jet creating abrasive water jet. Speed of abrasive
to generate the desired water pressure. These water jet can calculate using equation [2,4]:
pumps use hydraulics to apply a certain amount
of oil pressure on one side of a piston of a certain v wj . (2)
v awj  
diameter. On the water side of the pump, the m
1 a
diameter of the piston is much smaller. The mw
difference in the surface area between the
hydraulic side and the water side gives a Where are: vawj-speed of abrasive water jet in
multiplication factor, or intensification, to the m/s, -momentum loss factor (0.65-0.85), vwj -
pressure from the oil side. Most intensifier speed of water jet in m/s, ma-abrasive mass flow
pumps have an intensification ratio of 20 times. rate in kg/s, mw-water mass flow rate in kg/s.
Intensifier pumps generally are found in
pressure applications to 600 MPa. With Abrasive water jet is focused by a focusing tube.
intensifier pump can be achieved hyper Focusing tube directs the abrasive water jet to
pressure. Water jet pumps are specified in either cut the workpiece. Abrasive water jet cuts
horsepower (HP) or kilowatts (kW) to indicate workpiece along the programmed contour using
the size of the electric motor that creates the CNC (computer numerical control) motion
force to pressurize the water. High pressure system of the machine. Scheme of the AWJ
pumps power range is from 15 HP (11 kW) to cutting is shown in Fig. 1. [2,3].
200 HP (150 kW). Most common AWJ machine
tools are with power between 30 HP (22 kW) For evaluation of the AWJ cutting, the greatest
and 100 HP (75 kW). Cutting head consists of influence has a group of geometric characteristics
orifice, mixing chamber and focusing tube. of the workpiece issued in cutting process. The
Orifice is with diameter of 0.15 to 0.35 mm and geometric characteristics are: form accuracy,
is made of sapphire, ruby or diamond. Focusing dimensional accuracy and cut quality. Cut quality
tube is with diameter of 0.54 to 1.1 mm, length relates to form of kerf and cut surface. Terms in
of 50 to 100 mm, and is made of tungsten AWJ cutting, according ISO/TC 44 N 1770 [5], are

455
 M. Radovanović, Tribology in Industry Vol. 38, No. 4 (2016) 454-462

shown in Fig. 2, where are: Ra-surface roughness, Response Surface Methodology (RSM). Two
f-pitch of drag line, g-burr, hf-fine cut, hr- objectives (material removal rate and surface
remaining surface, n-drag line, rk-edge radius, s- roughness) are optimized by using Grey Relation
jet direction, sb-jet affected zone, t-workpiece Analysis (GRA).
thickness, u-perpendicularity or angularity
tolerance. Cut quality limits AWJ cutting Chakravarthy P. and Babu N. in [7] studied
application. The perpendicularity deviation and abrasive water jet cutting of granite. They have
surface roughness of the cut are the most investigated the effect of factors (water
significant characteristics of the cut quality. pressure, traverse rate and abrasive flow rate)
on performance (depth of cut). In order to
determine the Pareto front for the multi-
objective optimization problem with three
objectives (cost of production, rate of production
and consumption of abrasives) they have
applied a new approach based on the principles
of fuzzy logic and genetic algorithm (GA).

Chaitanya M. and Krishna A. in [8] studied laser

cutting of aluminum alloy 7075. They have
investigated the effect of factors (pulse-power,
pulse frequency, pulse width and assist gas
pressure) on performances (surface roughness
and heat affected zone). In order to determine
the Pareto front for the multi-objective
optimization problem with two objectives
(surface roughness and heat affected zone) they
have applied GA based non dominated sorting
algorithm-II (NSGA-II).

Fig. 1. Scheme of AWJ cutting. Soni V. et al. in 9 studied turning of aluminum
using carbide cutting tool. They have
investigated the effect of factors (depth of cut,
feed and cutting speed) on performances
(material removal rate and surface roughness).
In order to determine the Pareto front for the
multi-objective optimization problem with two
objectives (material removal rate and surface
roughness) they have applied multi-objective
genetic algorithm (MOGA).

Bouzakis K. et al. in 10 studied milling of

aluminum alloy 2024 using cemented carbide
Fig. 2. Terms related to kerf and cut surface. face milling cutting tool. They have investigated
the effect of factors (depth of cut, feed rate and
There are some studies of multi-objective cutting speed) on performances (machining cost
optimization of abrasive water jet machining and machining time). In order to determine the
and multi-objective optimization using genetic Pareto front for the multi-objective optimization
algorithms (GA). Aultrin J. and Anand D. in [6] problem, for rough milling with three objectives
studied abrasive water jet cutting of aluminum (production cost, production time and distance
alloy 6061. They have investigated the effect of from maximum power) and for finish milling
factors (water pressure, abrasive flow rate, with three objectives (production cost,
orifice diameter, focusing tube diameter and production time and cutting tool deflection),
standoff distance) on performances (material they have applied multi-objective genetic
removal rate and surface roughness). using algorithm (MOGA).

456
 M. Radovanović, Tribology in Industry Vol. 38, No. 4 (2016) 454-462

2. AWJ CUTTING FACTORS AND  Factors of motion system (precision,

PERFORMANCES accuracy, stiffness, working conditions).
 Factors of process (water pressure,
AWJ cutting is a complex process with many
traverse rate, abrasive flow rate, standoff
factors that determine performances.
distance, impact angle).
Mechanism of erosion depends on the level of
various process factors and is explained by
AWJ cutting performances can be classified into
multiple phenomena. AWJ cutting factors can be
five categories, Fig. 4:
classified into six categories, Fig. 3:
 Process performances (orifice wear,
 Factors of workpiece (material type,
focusing tube wear, temperature, noise,
thickness, chemical structure, hardness,
vibration).
toughness, grain size, tolerances,
roughness).  Quality performances (form deviations,
dimension deviations, cut quality: kerf
 Factors of water system (pump pressure,
depth, kerf width, perpendicularity
water flow rate, water purity, accumulator
deviation of the cut surfaces-kerf taper,
volume, high pressure tubing).
striations, burr, edge radius, jet affected
 Factors of abrasive system (abrasive zone, surface roughness).
material type, abrasive hardness, abrasive
 Productivity performances (machining
particle size, abrasive particle shape,
time, productivity).
abrasive particle size distribution,
abrasives input method).  Economy performances (manufacturing
cost, power consumption, abrasives
 Factors of cutting head (orifice diameter,
consumption, water consumption).
orifice material, focusing tube diameter,
focusing tube length, focusing tube  Ecology performances (noise, pollution,
material). recycling).

Fig. 3. AWJ cutting factors.

Fig. 4. AWJ cutting performances.

457
 M. Radovanović, Tribology in Industry Vol. 38, No. 4 (2016) 454-462

Some of the most important performances in levels) with center point. Control factors are:
AWJ cutting are perpendicularity deviation, abrasive flow rate (ma), traverse rate (vf), and
surface roughness and productivity. standoff distance (h). Control factors and their
levels are shown in Table 1. [12]
Perpendicularity deviation of the cut surfaces
(kerf taper) is distance between two parallel Table 1. Control factors and levels
straight lines (tangents) between which the cut Levels
surface profile is inscribed, and within the set Code Control factors
-1 0 +1
angle (e.g. 90 in the case of vertical cuts).
A Abrasive flow rate, ma (g/min) 300 500 700
Perpendicularity deviation includes and the
flatness deviations. B Traverse rate, vf (mm/min) 50 100 150
C Standoff distance, h (mm) 1 3 5
Surface roughness is a measure of the
technological quality of the cut. Commonly Three control factors are arranged in design of
arithmetic average roughness (Ra) was used to experiment with 9 runs (8 runs are for base
describe surface roughness. Ra is defined as the design and 1 run for center point). Measured
arithmetic value of the profile from centerline performances are: perpendicularity deviation
along the sampling length. (u) and surface roughness (Ra). Perpendicularity
Productivity is defined as: deviation has measured using optical
microscope Metkon and surface roughness has
Q  tvf . (3)
measured using surface measuring instrument
where are: Q (mm2/min)-productivity, t (mm)- Hommel Tester T500. Design of experiment with
material thickness, vf (mm/min)-traverse rate. factor levels and measured values of
performances, is presented in Table 2.
Productivity in contour cutting of sheets is
defined as cut surface productivity, i.e. surface of Table 2. Design of experiment and results.
the cut in unit of time. The goal in contour Control factors u Ra
cutting is to have maximum cutting length with Run
A B C (mm) (µm)
minimum kerf width.
1 -1 -1 -1 0.07 4.12
2 -1 -1 1 0.11 4.50
3. EXPERIMENTAL RESEARCH 3 -1 1 -1 0.17 5.18
4 -1 1 1 0.14 5.13
Experimental research was planed and realized 5 1 -1 -1 0.20 4.23
in order to define regression equations for 6 1 -1 1 0.21 4,45
performances (perpendicularity deviation and
7 1 1 -1 0.30 4.99
surface roughness) in correlation of factors
(abrasive flow rate, traverse rate and standoff ... ... ... ... ... ...
distance). The equipment used for 9 0 0 0 0.26 4.22
experimentation was abrasive water jet cutting
machine Hydro Jet Eco 0615 with high pressure Analysis of variance (ANOVA) was carried out to
pump of 150 MPa, power of 7.5 kW and water find the relative effect of factors and interactions
flow rate of 2.4 l/min. Cutting head was on performances perpendicularity deviation and
composed of orifice made of sapphire with inner surface roughness. If p<0.05 the factors have a
diameter of 0.35 mm and focusing tube made of significant effect on performance. If p>0.10 the
tungsten carbide with inner diameter of 1.02 factors have not an effect on performance. If
mm and length of 76 mm. Jet impact angle was 0.05<p<0.10 the factors have a moderate effect
90. All experiments were conducted with water on performance. Analysis of variance for
pressure of 150 MPa. Abrasive material was perpendicularity deviation (u) is shown in Table
Garnet mesh 80 (≈ 177 µm). Workpiece material 3. Standard F table value at 95 % confidence
was carbon steel EN S235, thickness of 6.5 mm. level is F0.05,1,3=10.13.

Design of experiment was conducted using full From Table 3 can see that factors: abrasive flow
factorial design 23 (three factors each on two rate, traverse rate and standoff distance have a

458
 M. Radovanović, Tribology in Industry Vol. 38, No. 4 (2016) 454-462

strong (clearly statistically significant) effect on Table 4. Analysis of variance for Ra.
the perpendicularity deviation with contribution Source DF SS MS F p %
of 90.28 %. Traverse rate is the most significant Main effects 3 1.33330 0.44443 4848 0.000 65.39
factor affecting the perpendicularity deviation
A 1 0.06845 0.06845 746 0.000 3.36
with contribution of 51.04 %. Abrasive flow rate
affects with contribution of 34.70 % and B 1 1.26405 1.26405 13789 0.000 61.99
standoff distance affects with contribution of C 1 0.00080 0.00080 8.73 0.060 0.04
4.54 % on the perpendicularity deviation. Some 2-Way
3 0.03225 0.01075 117 0.001 1.58
interactions have significant effect on the Interactions
perpendicularity deviation. Interaction abrasive AB 1 0.02645 0.02645 288 0.000 1.30
flow rate-traverse rate affects with contribution AC 1 0.00080 0.00080 8.73 0.060 0.04
of 1.27 %, abrasive flow rate-standoff distance BC 1 0.00500 0.00500 54.55 0.005 0.24
affects with contribution of 5.67 % and
3-Way
interaction abrasive flow rate-traverse rate- Interaction
1 0.02000 0.02000 218 0.001 0.98
standoff distance affects with contribution of
ABC 1 0.02000 0.02000 218 0.001 0.98
1.90 % on the perpendicularity deviation.
Curvature 1 0.65340 0.65340 7128 0.000 32.04
Table 3. Analysis of variance for u. Error 3 0.00028 0.00009 - - 0.01
Source DF SS MS F p % Total 11 2.03922 - - - 100
Main effects 3 0.07184 0.02394 261.23 0.000 90.28
A 1 0.02761 0.02761 301.23 0.000 34.70
Traverse rate is the most significant factor
affecting the surface roughness with
B 1 0.04061 0.04061 443.05 0.000 51.04
contribution of 61.99 %. Abrasive flow rate
C 1 0.00361 0.00361 39.41 0.008 4.54 affects with contribution of 3.36 %. Standoff
2-Way
3 0.00584 0.00194 21.23 0.016 7.34 distance has moderate effect with contribution
Interactions of 0.04 %. Some interactions have significant
AB 1 0.00101 0.00101 11.05 0.045 1.27 effect on the surface roughness. Interaction
AC 1 0.00451 0.00451 49.23 0.006 5.67 abrasive flow rate-traverse rate affects with
BC 1 0.00031 0.00031 3.41 0.162 0.39 contribution of 1.30 %, traverse rate-standoff
3-Way distance affects with contribution of 0.24 %,
1 0.00151 0.00151 16.50 0.027 1.90 abrasive flow rate-traverse rate-standoff
Interaction
ABC 1 0.00151 0.00151 16.50 0.027 1.90 distance affects with contribution of 0.98 % on
the surface roughness. Interaction abrasive flow
Curvature 1 0.00010 0.00010 1.14 0.365 0.13
rate-standoff distance has moderate effect with
Error 3 0.00027 0.00009 - - 0.34 contribution of 0.04 % on the surface roughness.
Total 11 0.07957 - - - 100 Regression equation for surface roughness, with
DF-degree of freedom, SS-sum of square, MS-mean square, coefficient of determination of R2=99.99 %, is:
F-variance ratio, p–value, %-percent contribution
R a  3.01125  0.0017125m a  0.015325v
Regression equation for perpendicularity  0.1325h  0.00001325m a v  0.000225m a h (5)
deviation, with coefficient of determination of  0.0015vh  0.0000025m a vh
R2=99.65%, is:
u  0.106562  0.000434375m a  0.00283125v f
 0.0090625h  0.0000031875m a v f (4) 4. MULTI-OBJECTIVE OPTIMIZATION
 0.000009375m a h  0.0000006875m a v f h
Procedure of multi-objective optimization has
Analysis of variance for surface roughness (Ra) is four phases. First phase is mathematical
shown in Table 4. Standard F table value at 95 % modelling of performances in correlation of the
confidence level is F0.05,1,3=10.13. factors. Second phase is determining
optimization problem. In second phase
From Table 4 can see that factors: abrasive flow objectives are selected and constraints are
rate and traverse rate have a strong (clearly defined. Third phase is selection of method for
statistically significant) effect on the surface solution of optimization problem. Fourth phase
roughness with contribution of 65.35 %. is solution of optimization problem.

459
 M. Radovanović, Tribology in Industry Vol. 38, No. 4 (2016) 454-462

Mathematical model of the multi-objective Nondominated points, generated by multi-

optimization was defined with goal to minimize objective genetic algorithm solver, have plotted
perpendicularity deviation and surface rougness in form of the Pareto front, Fig. 5.
and maximize productivity. Mathematical model
of the multi-objective optimization is:
 Objective functions:
f1  min u 
f 2  min R a 
f 3  max Q

 Constraints:
ma ,min  ma  ma ,max
vf ,min  vf  vf ,max
h min  h  h max

Mathematical model of the multi-objective

optimization for this study has form:
Fig. 5. Pareto front points.
 Objective functions:
 u  0.106562  0.000434375m a  Listing of the Pareto front points obtained as
 
  0.00283125v f  0.0090625h  outcomes from the optimization process
f1  min  presented in Table 6.
 0.0000031875m a v f  0.000009375m a h 
 
  0.0000006875m v h 
 a f  Table 6. Pareto front points.
 R a  3.01125  0.0017125m a  0.015325v f  f1 f2 f3 x1 x2 x3
  Index u Ra Q ma vf h
  0.1325h  0.00001325m a v f 
f 2  min   (mm) (m) (mm2/min) (g/min) (mm/min) (mm)
 0.000225m a h  0.0015v f h
  1 0.134 4.120 325.000 300.000 50.000 1.000
  0.0000025m v h 
 a f  2 0.356 5.153 974.877 507.442 149.981 1.069
f3  min  Q  6.5vf  3 0.251 4.563 588.735 339.636 90.113 1.965
4 0.590 5.121 972.039 579.961 149.545 4.711
 Constraints 5 0.526 5.030 891.874 573.326 137.211 4.248
300  m a  700 6 0.353 5.153 974.757 507.461 149.963 1.016
50  v f  150 7 0.469 4.841 716.290 600.770 110.198 4.359
1 h  5 8 0.507 5.014 873.178 585.376 134.335 3.984
9 0.434 5.033 937.266 367.289 144.195 3.377
Table 5. Parameters of the multi-objective genetic
10 0.201 4.431 505.085 325.091 77.705 1.163
algorithm.
11 0.267 4.723 696.890 309.600 107.214 1.426
Population type Double vector
12 0.487 5.101 947.547 580.924 145.776 3.217
Population size 50
Creation function Constraint dependent 13 0.232 4.408 465.341 364.071 71.591 2.582
Selection function: Tournament, 14 0.382 4.913 806.165 482.089 124.025 2.598
Selection
Tournament size: 2 15 0.421 4.980 844.410 561.978 129.909 2.756
Reproduction Crossover fraction: 0.8 16 0.259 4.474 524.641 332.475 80.714 3.201
Mutation function: Constraint 17 0.333 4.865 757.053 501.921 116.470 1.747
Mutation
dependent
Crossover function:
Crosover From Table 6 it is evident that near optimal
Intermediate, Ratio: 1.0
Migration
Direction: Forward, Fraction: factor levels for rough cutting carbon steel EN
0.2, Interval: 20 S235 with abrasive water jet can be selected as:
Multiobjective Pareto front population fraction:
abrasive flow rate of ma=507.442 g/min,
problem settings 0.35
Generations: 100*number of traverse rate of vf=149.981 mm/min and
Stopping criteria standoff distance of h=1.069 mm. For these
variables
Current iteration 116 factor levels, perpendicularity deviation is

460
 M. Radovanović, Tribology in Industry Vol. 38, No. 4 (2016) 454-462

u=0.356 mm, surface roughness is Ra=5.153 m Acknowledgement

and productivity is Q=974.877 mm2/ min.
The paper is a result of the technological project
From Table 6 it is evident that perpendicularity TR35034 which is supported by the Ministry of
deviation u is important performance for finish Education, Science and Technological
cutting carbon steel EN S235 with abrasive Development of the Republic of Serbia.
water jet. Surface roughness is Ra6.3 m. For
example, if tolerance of dimensions is T=0.25
mm (u0.25 mm) and Ra6.3 m, than the near REFERENCES
optimal factor levels can select as: abrasive flow
rate of ma=325.091 g/min, traverse rate of [1] M. Radovanović and M. Madić, ‘Optimizing
vf=77.705 mm/min and standoff distance of factor levels based on cost in abrasive water jet
cutting using GA‘, Methods and Techniques for
h=1.163 mm. For these factor levels,
Industrial Development, University of Maribor,
perpendicularity deviation is u=0.201 mm, Faculty of Mechanical Engineering, Politecnico
surface roughness is Ra=4.431 m and di Torino, Maribor, Slovenia, pp. 17-30, 2015.
productivity is Q=505.085 mm2/ min.
[2] M. Radovanović, ‘Investigation of surface
roughness in abrasive water jet cutting using
Shainin method‘, Tribological Journal BULTRIB,
5. CONCLUSION vol. 5., no. 5, pp. 105-110, 2015.

AWJ cutting is a modern nonconventional [3] M. Radovanović and E. Herghelegiu,

machining technology. Cut quality limits AWJ ‘Perpendicularity deviation and surface
roughness in abrasive water jet cutting of
cutting application. Some of the most important
carbon steel‘, Nonconventioal Technologies
performances in AWJ cutting are Review, no. 2, pp. 39-44, 2016.
perpendicularity deviation, surface roughness
and productivity. To optimize AWJ cutting [4] B. Nedić and J. Baralić, ‘The wear of the focusing
performances it is essential to have knowledge tube and the cut-surface quality‘, Tribology in
Industry, vol. 32, no. 2, pp. 38-43, 2010.
of the perpendicularity deviation and surface
roughness in relation to process factors. When [5] ISO/TC 44 N 1770, Water jet cutting-
AWJ cutting carbon steel EN S235 regression Geometrical product specification and quality,
equations for perpendicularity deviation and 2010.
surface roughness in relation of the abrasive [6] J. Aultrin and D. Anand, ‘Multi-objective
flow rate, traverse rate and standoff distance optimization of abrasive water jet machining of
have defined using experimental research. For aluminum 6061 alloy by Grey Relational
minimize perpendicularity deviation and surface Analysis‘, Journal of Chemical and
rougness and maximize productivity, Pharmaceutical Sciences, vol. 9, no. 1, pp. 410-
mathematical model of the multi-objective 417, 2016.
optimization was defined. Multi-objective [7] P. Chakravarthy and N. Babu, ‘A new approach
genetic algorithm solver was selected as method for selection of optimal process parameters in
for solution of the optimization problem. For abrasive water jet cutting‘, Materials and
rough cutting the near optimal factor levels can Manufacturing Processes, vol. 14, no. 4, pp. 581-
select as: ma=507.442 g/min, vf=149.981 600, 2007.
mm/min and h=1.069 mm. For these factor [8] M. Chaitanya, A. Krishna, Multi-objective
levels perpendicularity deviation is u=0.356 mm, optimization of laser beam cutting process,
surface roughness is Ra=5.153 m and International Journal of Research in Mechanical
productivity is Q=974.877 mm2/min. For finish Engineering & Technology - IJRMET, vol. 3, no. 2,
pp. 279-290, 2013.
cutting, if T=0.25 mm and Ra6.3 m, the near
optimal factor levels can select as: ma=325.091 [9] V. Soni, S. Mandal and B. Singh, ‘Process
g/min, vf=77.705 mm/min and h=1.163 mm. For parameters optimization in turning of
these factor levels perpendicularity deviation is aluminum using a new hybrid approach‘,
u=0.201 mm, surface roughness is Ra=4.431 m International Journal of Innovative Science,
Engineering & Technology - IJISET, vol. 1, no. 3,
and productivity is Q=505.085 mm2/min.
pp. 418-423, 2014.

461
 M. Radovanović, Tribology in Industry Vol. 38, No. 4 (2016) 454-462

[10] K. Bouzakis, R. Paraskevopoulou and G. water jet using Taguchi method‘, Tribological
Giannopoulos, ‘Multi-objective optimization of Journal BULTRIB, vol.4, pp. 47-54, 2014.
cutting conditions in milling using genetic
[14] P. Janković, M. Radovanović, J. Baralić and B.
algorithms‘, in: 3rd International Conference on
Nedić, ‘Prediction model of surface roughness in
Manufacturing Engineering - ICMEN-2008, 1-3
abrasive water jet cutting of aluminium alloy‘,
October, Chalkidiki, Greece, pp. 763-774, 2008.
Journal of the Balkan Tribological Association,
[11] E. Herghelegiu, ‘Contributions to the vol. 19, no 4, pp. 585-595, 2013.
optimization of the water jet working
[15] M. Radovanović, P. Janković and M. Madić,
parameters‘, Dissertation, University Vasile
‘Predictive models of traverse rate in abrasive
Alecsandri of Bacau, Romania, 2011. water jet cutting based on RA and GA‘, Academic
[12] M. Madić, M. Radovanović and B. Nedić, Journal of Manufacturing Engineering, vol. 10,
‘Correlation between Surface Roughness no. 1, pp. 107-112, 2012.
Characteristics in CO2 Laser Cutting of Mild [16] M. Radovanović, ‘Abrasive Waterjet Cutting
Steel‘, Tribology in Industry, vol. 34, no. 4, pp. cost‘, Nonconventional Technologies Review, no.
232-238, 2012. 1, pp. 97-102, 2007.
[13] M. Radovanović, ‘Investigation on surface
roughness of carbon steel machined by abrasive

462