Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Available online at www.sciencedirect.

com

ScienceDirect
Available online at www.sciencedirect.com

ScienceDirect
Procedia Manufacturing 00 (2019) 000–000
www.elsevier.com/locate/procedia

Procedia Manufacturing 47 (2020) 781–787

23rd International Conference on Material Forming (ESAFORM 2020)

Multi-Objective Optimization of Abrasive Water Jet Cutting Using MOGA


Miroslav Radovanovica,*
a
University
University of Nis, Faculty of Mechanical Engineeing, Aleksandra Medvedeva 14, 18000 Nis, Serbia

* Corresponding author. Tel.:+381 64 264 1540; fax:+381 18 588 244. E-mail address: miroslav.radovanovic@masfak.ni.ac.rs

Abstract

Abrasive water jet (AWJ) cutting is an unconventional machining process used to cut different types and thicknesses of materials. Cutting process
is based on material removal by erosion. Cutting tool is an ultra-fast thin jet of water mixed with abrasive particles. High-speed water jet
accelerates the abrasive particles and these particles erode the material. AWJ cutting is a complex process with many factors that determine
performances. The most important factors are: water pressure, water flow rate, orifice diameter, nozzle diameter, abrasive mass flow rate, traverse
speed and standoff distance. The most important performances are performances related to economy, productivity and quality. Optimizing the
AWJ cutting performances is essential in order to obtain better product quality, higher productivity and higher economic efficiency. In process
optimization it is logical to take economy and productivity performances as objectives and quality performances as constraints. For the realization
of production goal single-objective optimization is a good way, but multi-objective optimization is the right way. In this paper, multi-objective
optimization of abrasive water jet cutting carbon steel S235 is studied. Optimization was performed to simultaneously optimize two objectives
(productivity and operating cost) with three factors (traverse speed, abrasive mass flow rate and standoff distance) and three constraints
(perpendicularity tolerance, surface roughness limit and traverse speed for separation cut). The optimization problem was solved using multi-
objective genetic algorithm (MOGA).

© 2020
© 2020Miroslav
The Authors. Published
Radovanovic. by Elsevier
Published Ltd. Ltd.
by Elsevier
Thisisisan
This anopen
openaccess
access article
article under
under the BY-NC-ND
the CC CC BY-NC-ND licenselicense (https://creativecommons.org/licenses/by-nc-nd/4.0/)
https://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer-reviewunder
Peer-review underresponsibility
responsibilityof of
thethe scientific
scientific committee
committee of 23rd
of the the 23rd International
International Conference
Conference on Material
on Material Forming.
Forming.

Keywords: Abrasive water jet cutting; Multi-objective optimization; Multi-objective genetic algorithm (MOGA).

1. Introduction random contours in thick, middle and thin sheets, such as very
fine tabs and filigree structures. Tolerances of 0.1 mm can be
Modern machining technologies are widely used in realized in metal cutting. Characteristics of AWJ cutting are:
manufacturing industries nowadays. Abrasive water jet (AWJ) no thermal distortion, no heat affected zone, small cutting
cutting is one of the most advanced modern machining force, minimum stress on the workpiece, high flexibility and
technologies used in manufacturing industry for material versatility. AWJ cutting as machining technology has
processing. AWJ cutting is an unconventional machining developed in the early 1980s.
technology. Cutting process is based on the erosion of material The first commercial abrasive water jet cutting system was
by abrasive particles accelerated by high speed water jet. AWJ available in 1983. Hard abrasive particles are added to the
cutting is an effective technology for machining various water jet, which makes it possible to cutting practically any
engineering materials and wide range of thickness. AWJ material. Development of AWJ cutting is a result of
cutting is widely used in the machining of materials such as development of high pressure pump. The water jet industry has
titanium, steel, brass, aluminum, stone, inconel and any kind different definitions for differences in pressure levels. High
of glass and composites. Thickness of the worpkiece that can pressure is between 15,000 psi (103 MPa) to 40,000 psi (276
be cut could have values up to 100 mm for stainless steel and MPa). Ultrahigh pressure is between 40,000 psi (276 MPa) to
up to 120 mm for aluminum. With AWJ it is possible to cut 75,000 psi (517 MPa). Hyper pressure is 75,000 psi (517 MPa)

2351-9789 © 2020 The Authors. Published by Elsevier Ltd.


This is an©open
2351-9789 access
2020 The article
Authors. under by
Published theElsevier
CC BY-NC-ND
Ltd. license (https://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer-review
This is an openunder responsibility
access article of the
under the CC scientific
BY-NC-ND committee
license of the 23rd International Conference on Material Forming.
https://creativecommons.org/licenses/by-nc-nd/4.0/)
10.1016/j.promfg.2020.04.241
Peer-review under responsibility of the scientific committee of the 23rd International Conference on Material Forming.
782 Miroslav Radovanovic / Procedia Manufacturing 47 (2020) 781–787
2 Miroslav Radovanovic/ Procedia Manufacturing 00 (2019) 000–000

and over. Pressure has increased every decade, moving to annealing (SA), ant colony optimization (ACO), particle
55,000 psi (379 MPa) by the end of the 1980s and reaching the swarm optimization (PSO) artificial bee colony (ABC), etc.
current standard of 60,000 psi (412 MPa) in the mid 1990s. In have been employed in place of conventional methods such as
general, standard pressure water jet systems run at ultrahigh linear and non-linear programming, etc.
pressure of 55,000 psi (379 MPa) to 60,000 psi (412 MPa) and Some of the advantages of computational approaches such
more advanced systems run at hyper pressure of 94,000 psi as GA in optimizing machining performances may include
(650 MPa) [3,8]. next: preferred when near-optimal conditions instead of the
Optimizing the AWJ cutting performances is essential in exact optimal solution are cost effective and acceptable for
order to obtain better product quality, higher productivity and implementation by the manufacturers, a derivative free
higher economic efficiency. The current trends of research approach for near-optimal points search direction, able to
show soft computing techniques have been used by many handle objective functions of any complexity with both
researchers to optimize the performances of conventional and discrete and continuous variables, an automatic search for the
unconventional machining technologies. non-linear connection between the inputs and outputs, and a
Evolutionary optimization methods that were inspired by fast and simple optimizing technique [11].
nature behavior such as genetic algorithm (GA), simulated There are few researchers who have studied multi-objective
optimization of abrasive water jet cutting. Aultrin and Anand
Nomenclature in [1] have studied multi-objective optimization of abrasive
water jet cutting aluminum alloy 6061. The effect of water
AWJ abrasive water jet pressure, abrasive flow rate, orifice diameter, nozzle diameter
C operating cost per meter of cut and standoff distance on material removal rate and surface
Ca cost of abrasive per hour roughness was investigated using response surface
Ce cost of electrical energy per hour methodology (RSM). Two objectives (material removal rate
Cm cost of maintenance per hour and surface roughness) were optimized using grey relation
Cr cost of wearing parts per hour analysis (GRA). Nair and Kumanan in [17] have studied multi-
Cw cost of water per hour objective optimization of abrasive water jet cutting inconel
Pa abrasive jet power 617. The effect of water pressure, abrasive mass flow rate,
Pe motor power standoff distance and traverse speed on circularity,
Q productivity cylindricity, perpendicularity and parallelism was
Ra surface roughness investigated. Five objectives (material removal rate,
Ram surface roughness limit circularity, cylindricity, perpendicularity and parallelism)
ca unit cost of abrasive were optimized using weighted principal components analysis
cd coefficient of discharge (WPCA). Deepak and Davim in [7] have studied multi-
ce unit cost of electrical energy objective optimization of abrasive water jet cutting fiber
cv coefficient of velocity reinforced polymer (graphite laced GFRP). The effect of
cw unit cost of water traverse speed, water pressure and standoff distance on kerf
do orifice diameter taper and surface roughness was investigated. Two objectives
df nozzle diameter (kerf taper and surface roughness) were optimized using grey
e specific cutting energy relational method. Radovanovic in [20] has studied multi-
h material thickness objective optimization of abrasive water jet cutting carbon
i quality number steel S235. The effect of traverse speed, abrasive mass flow
ma abrasive mass flow rate rate and standoff distance on perpendicularity deviation and
mw water jet mass flow rate surface roughness was investigated. Three objectives
p water pressure (perpendicularity deviation, surface roughness and
q water flow rate productivity) were optimized using multi-objective genetic
qw water jet flow rate algorithm (MOGA). Chakravarthy and Babu in [5] have
 water density studied multi-objective optimization of abrasive water jet
s standoff distance cutting granite. The effect of water pressure, traverse rate and
t machining time needed for a unit of cut surface abrasive flow rate on depth of cut was investigated. Three
u perpendicularity deviation objectives (production cost, production rate and abrasive
um perpendicularity tolerance consumption) were optimized using a new approach based on
vf traverse speed the principles of fuzzy logic and genetic algorithm (GA).
vs traverse speed for separation cut Jagadeesh et al. in [13] have studied multi-objective
vwj velocity of water jet optimization of abrasive water jet cutting carbon fiber
vawj velocity of abrasive water jet reinforced plastic laminates. The effect of thickness, traverse
w kerf width speed and standoff distance on surface roughness, kerf width
 impact angle and kerf taper was investigated. Two objectives (surface
 momentum loss factor roughness and kerf taper) were optimized using Design expert
 cutting energy efficiency coefficient software. Yuvaraj and Kumar in [24] have studied multi-
objective optimization of abrasive water jet cutting inconel
Miroslav Radovanovic / Procedia Manufacturing 47 (2020) 781–787 783
Miroslav Radovanovic/ Procedia Manufacturing 00 (2019) 000–000 3

617. The effect of water pressure, traverse speed, abrasive  Nozzle focuses the abrasive water jet and directs to
mass flow rate and standoff distance on depth of cut, kerf workpiece.
width, kerf taper and surface roughness was investigated. Five  Abrasive particles remove material from the workpiece.
objectives (depth of cut, kerf width, kerf taper, surface High-pressure pump generates water pressure and delivers
roughness and cutting rate) were optimized using technique for continuous water flow. Via accumulator and high pressure
order preference by similarity ideal solution (TOPSIS). tubing the pressurized water directs to the cutting head.
In this paper, multi-objective optimization of abrasive Cutting head consists of orifice, mixing chamber and nozzle.
water jet cutting carbon steel S235 was studied. Optimization In orifice potential energy of water transforms into kinetic
was performed to simultaneously optimize two objectives energy of water jet. Orifice is where the water pressure is
(productivity and operating cost) with three factors (abrasive converted to water velocity. Water is pressed out of the orifice
mass flow rate, traverse speed and standoff distance) and three in form of very thin jet at a high velocity. Abrasive particles
constraints (perpendicularity tolerance, surface roughness are metered from a mini-hopper to the cutting head and added
limit and traverse speed for separation cut). The optimization to water jet in mixing chamber. The high velocity of the water
problem was solved using multi-objective genetic algorithm jet creates a Venturi effect or vacuum in the mixing chamber
(MOGA). located immediately beneath the orifice, so that abrasive
particles are sucked into the water jet. Abrasive particles are
2. Abrasive water jet cutting mixed with the water jet in the nozzle. Abrasive water jet exit
of the nozzle as a mixture of water, abrasive particles and some
Process of abrasive water jet (AWJ) cutting is based on air. Nozzle focuses abrasive water jet and directs to the
material removal from the workpiece by erosion. High velocity workpiece. Кinetic energy of a jet of abrasive particles is used
water jet accelerates abrasive particles and those particles for cutting. Abrasive particles erode material of the workpiece
erode the material. Mechanism of erosion in AWJ cutting is by cutting (small chips) and deformation (small particles). The
explained by multiple phenomena. precision X-Y movement system is used to move the cutting
Scheme of abrasive water jet cutting is shown in Fig. 1. head to create the desired cutting path. Advanced motion
controllers for abrasive water jet systems are computer based
and enable the production of accurate paths. Cutting head
moves along the workpiece at traverse speed [16,21,22].
Traverse speed can be given in the form of equation based on
specific cutting energy or in the form of regression equation
obtained by design of experiment (DoE). Traverse speed based
on specific cutting energy is [9]:

Pa
vf  (1)
eihw

where vf-traverse speed, -cutting energy efficiency


coefficient (   0.9113 v0f .134 ), Pa-abrasive jet power, e-specific
cutting energy, i-quality number, h-material thickness, w-kerf
width.
Specific cutting energy was determined experimentally.
Quality number describes the qualitative level of the surface
roughness, the kerf taper angle and the specific cut surface
characteristics. Quality number has value from 1 to 10 (1 for
separation cut-rough cut strongly tapered, 5 for smooth cut
small tapered, 10 for smooth and square cut).
Abrasive jet power is:

Fig. 1. Scheme of abrasive water jet cutting 1 2


Pa  ma vawj (2)
2
Process of AWJ cutting can be decomposed into next
phases: where Pa-abrasive jet power, ma-abrasive mass flow rate, vawj-
 High-pressure pump generates water pressure and delivers velocity of abrasive water jet.
continuous water flow. Velocity of abrasive water jet is:
 In orifice potential energy of water transforms into kinetic
energy of water jet. 
 In nozzle kinetic energy of the water jet transfers to abrasive vawj  vwj (3)
particles. Abrasive particles are mixed with the water jet 1 ma / mw
creating abrasive water jet.
where vawj-velocity of abrasive water jet, -momentum loss
784 Miroslav Radovanovic / Procedia Manufacturing 47 (2020) 781–787
4 Miroslav Radovanovic/ Procedia Manufacturing 00 (2019) 000–000

factor (=0.65-0.85), vwj-velocity of water jet, ma-abrasive The most important AWJ cutting factors are: water
mass flow rate, mw-water jet mass flow rate. pressure, water jet flow rate, orifice diameter, nozzle diameter,
Velocity of water jet based on Bernoullis equation is: abrasive mass flow rate, traverse speed and standoff distance.
AWJ cutting performances can be classified into next
2p categories:
vwj  cv (4)  Process performances (traverse speed for separation cut,

orifice wear, nozzle wear, temperature, noise, etc.).
 Quality performances (form and dimensional accuracy, cut
where vwj-velocity of water jet, cv-coefficient of velocity quality: kerf depth, kerf width, kerf taper, striations, burr,
(cv=0.85-0.98), p-water pressure, - water density. edge radius, jet affected zone, surface roughness).
Water jet mass flow rate is:  Productivity performances (machining time, productivity).
 Economy performances (machining cost, electric energy
mw  qw (5) consumption, abrasive consumption, water consumption).
 Ecology friendly performances (noise, pollution).
where mw-water jet mass flow rate, -water density, qw-water  Safety performances (hazard, health).
jet flow rate. The most important AWJ cutting performances are quality,
Water jet flow rate is: productivity and economy performances. Cut quality limits
AWJ cutting application. Terms in AWJ cutting, according
d o2 2p ISO/TC 44 N 1770, are shown in Fig. 2, where are: Ra-surface
q w  cd (6) roughness, f-pitch of drag line, g-burr, hf-fine cut, hr-remaining
4  surface, n-drag line, rk-edge radius, sb-jet affected zone, h-
material thickness, u- perpendicularity or angularity tolerance
where qw-water jet flow rate, cd-coefficient of discharge [12].
(cd=0.6-0.8), do-orifice diameter, p-water pressure, -water
density. Kerf width is about 10% larger than the nozzle
diameter. Kerf width is:

w  1.1d f (7)

where w-kerf width, df-nozzle diameter.


Traverse speed based on equations (1), (2), (3), (4), (5) and
(6) is:

1.155
  2cv2 p 
 m 
 eihwq a  Fig. 2. Terms related to kerf and cut surface
v f  0.898  
2.31 (8)
 4 
1  m  Perpendicularity deviation and surface roughness are the
 c d 2 2 p a  most significant characteristics of the cut quality. Deviations
 d o 
from the perpendicularity and surface roughness are usually in
AWJ cutting is a complex process with many factors that the form of regression equations obtained by the design of
determine performances. experiment (DoE).
AWJ cutting factors can be classified into next categories: In this paper selected productivity and economy
 Factors of workpiece (material type, material thickness, performances are machining time needed to produce a unit of
material hardness, material toughness, material grain size, cut surface and operating cost per meter of cut. Quality
tolerances, surface roughness). performances (perpendicularity deviation and surface
 Factors of water system (maximum pressure, maximum roughness) and process performance (traverse speed for
water flow rate, water purity, accumulator volume, high separation cut) have taken as constraints.
pressure tubing).
 Factors of abrasive system (abrasive material type, abrasive 3. Multi-objective optimization
hardness, abrasive particle size, abrasive particle shape,
abrasive particle size distribution, abrasives input method). Optimizing the AWJ cutting performances is essential in
 Factors of cutting head (orifice diameter, orifice material, order to provide a better quality, productivity and economic
nozzle diameter, nozzle length, nozzle material). efficiency.
 Factors of movement system (accuracy, stiffness, working Optimization procedure consists of next phases:
conditions).  Selection of objectives, constraints and bounds.
 Factors of process (water pressure, traverse speed, abrasive  Defining the optimization problem.
mass flow rate, standoff distance, impact angle).  Selection a method to solve the optimization problem.
 Solving the optimization problem.
Miroslav Radovanovic / Procedia Manufacturing 47 (2020) 781–787 785
Miroslav Radovanovic/ Procedia Manufacturing 00 (2019) 000–000 5

Usually objectives are some of the quality, productivity and


economy performances. Ra  Ram (13)
Optimization problem of the multi-objective optimization
was defined with goal to maximize productivity and minimize where Ra-surface roughness, Ram-surface roughness limit.
operating cost. Multi-objective optimizations, where the Perpendicularity deviation and surface roughness are
objectives are generally conflicting, are difficult for solving usually given in the form of regression equations obtained by
[6,18]. The goal in contour cutting is to have maximum cutting design of experiment.
length with minimum kerf width. Productivity in contour  Traverse speed for separation cut
cutting of sheets is defined as cut surface productivity, i.e. cut
surface in unit of time. v f  vs (14)

Q  hv f (9)
where vf-traverse speed, vs-traverse speed for separation cut.
Bounds relate to technological limits for abrasive mass flow
where Q (mm2/min)-productivity, h (mm)-material thickness, rate, traverse speed and standoff distance.
vf (mm/min)-traverse speed. Multi-objective genetic algorithm (MOGA) is used for
Productivity is the objective function that must be solving multi-objective optimization problem. MOGA is
maximized. Inverse function of the productivity is machining considered one of the most popular meta-heuristic approaches
time needed to produce a unit of cut surface. Optimization that are well suited for solving multi-objective optimization
problem is defined with objectives, constraints and bound. problems because it does not require user to prioritize, scale or
Selected objectives are: weight objectives. In multi-objective optimization there is a set
 Machining time needed to produce a unit of cut surface of optimal solutions called Pareto optimal set. Non-
dominated optimal solutions usually were plotted to form of
1 the Pareto front. Based on the set of optimal solutions, one
t (10)
hv f solution must be selected [14].

4. Example of the optimization


where t (min/mm2-machining time needed to produce a unit of
cut surface, h (mm)-material thickness, vf (mm/min)-traverse
An example of multi-objective optimization of abrasive
speed. Machining time needed to produce a unit of cut surface
water jet cutting is presented. Samples of carbon steel S235,
is the objective function that must be minimized.
6.5 mm thick, were cut with a water jet cutting machine Hydro
 Operating cost per meter of cut
Jet Eco 0615. High pressure pump is with maximum pressure
p=150 MPa, engine power Pe=7.5 kW and maximum water
1
C Ce  Cw  Ca  Cr  Cm  flow rate q=2.4 l/min. Cutting head is with orifice diameter
vf do=0.35 mm and nozzle diameter df=1.02 mm. Impact angle is
=90. Abrasive material is a garnet mesh 80. Coefficient of
1 speed is cv=0.95. Coefficient of discharge is cd=0.7.
C ce Pe  cw q  ca ma  Cr  Cm  (11) Momentum loss factor is =0.75. Water density is w=1000
vf kg/m3. Water pressure is p=150 MPa. Specific cutting energy
is e=6975 MJ/m3. Minimum abrasive mass flow rate is
where C (EUR/m)-operating cost per meter of cut, vf (m/h)- ma,min=300 g/min. Maximum abrasive flow rate is ma,max=700
traverse speed, Ce (EUR/h)-cost of electrical energy per hour g/min. Minimum technological traverse rate is vf,min=50
(Ce=cePe), ce (EUR/kWh)-unit cost of electrical energy, P e mm/min. Maximum technological traverse rate is vf,max=150
(kW)-motor power, Cw (EUR/h)-cost of water per hour mm/min. Minimum standoff distance is smin=1 mm. Maximum
(Cw=cwq), cw (EUR/m3)-unit cost of water, q (m3/h)-water flow standoff distance is smin=5 mm. Unit cost of electrical energy
rate, Ca (EUR/h)-cost of abrasive per hour (Ca=cama), ca is ce=0.105 EUR/kWh. Unit cost of water is cw=0.5 EUR/m3.
(EUR/kg)-unit cost of abrasive, ma (kg/h)-abrasive mass flow Unit cost of abrasive is ca=0.7 EUR/kg. Cost of wearing parts
rate, Cr (EUR/h)-cost of wearing parts per hour, Cm (EUR/h)- per hour is Cr=0.858 EUR/h. Cost of maintenance per hour is
cost of maintenance per hour. Cm=1.25 EUR/h. Perpendicularity tolerance is u0.1 mm.
Operating cost per meter of cut is the objective function Surface roughness limit is Ra6.3 m. Regression equations
that must be minimized. for perpendicularity deviation and surface roughness when
Selected constraints are: abrasive water jet cutting carbon steel S235 were determined
 Perpendicularity tolerance by design of experiment (DoE) [10,19].
Perpendicularity deviation is:
u  um (12)
u  0.00025 ma0.28v1f.05 s 0.60 (15)
where: u-perpendicularity deviation, um-perpendicularity
tolerance.
 Surface roughness limit
786 Miroslav Radovanovic / Procedia Manufacturing 47 (2020) 781–787
6 Miroslav Radovanovic/ Procedia Manufacturing 00 (2019) 000–000

where u (mm)-perpendicularity deviation, ma (g/min)-abrasive Table 1. Parameters of the Matlab multiobjective genetic algorithm solver
mass flow rate, vf (mm/min)-traverse speed, s (mm)-standoff Population type Double vector
distance. Population size 50
Surface roughness is: Creation function Constraint dependent
Selection Selection function: Tournament, size: 2
ma0.07 v 0f .29
Ra  0.826 (16) Reproduction Crossover fraction: 0.8
s 0.012 Mutation Mutation function: Constraint dependent
Crosover Crossover function: Intermediate, Ratio: 1.0
where Ra (m)-surface roughness, ma (g/min)-abrasive mass
flow rate, vf (mm/min)-traverse speed, s (mm)-standoff Migration Direction: Forward, Fraction: 0.2, Interval: 20
distance. Equations (15) and (16) are valid for: 300 g/min  ma Multiobjective Pareto front population fraction: 0.35
 700 g/min, 50 mm  vf  150 mm, and 1 mm  s 5 mm. problem settings
Traverse speed for separation cut based on equation for Stopping criteria Generations: 100*number of variables
specific cutting energy (8) is:
Nondominated points generated by the Matlab
m1a.155 multiobjective genetic algorithm solver were plotted to form
vs  0.26 (17) the Pareto front, Fig. 3.
1  0.0003 ma 2.31
Table 2. Part of the Pareto front points
where vs (mm/min)-traverse speed for separation cut, ma Index t (s/cm2) C (EUR/m) vf (mm/min) ma (g/min) s (mm)
(g/min)-abrasive mass flow rate.
1 11.273 3.349 81,857 320.995 1.817
Optimization was performed to simultaneously optimize
2 11.209 3.364 82.326 342.987 1.601
two objectives (machining time needed to produce a unit of cut
surface and operating cost per meter of cut), with three factors 3 10.560 3.447 87.383 359.681 1.761
(abrasive mass flow rate, traverse speed and standoff distance), 4 10.121 2.999 91.173 319.998 1.483
three constraints (perpendicularity tolerance, surface 5 9.488 3.480 97.256 412.842 1.351
roughness limit and traverse speed for separation cut) and 6 9.146 2.651 100.900 311.528 1.382
bounds.
7 8.687 2.716 106.225 341.556 1.278
Mathematical model of the multi-objective optimization
8 8.151 2.352 113.216 309.730 1.205
has form:
 Objective functions 9 7.404 2.091 124.630 301.708 1.014
10 7.266 2.048 127.003 301.020 1.000
922.8
min, t  11 7.275 2.056 126.850 302.000 1.001
vf
12 7.955 2.322 116.002 314.168 1.115
1 13 8.360 2.878 110.382 383.169 1.090
min, C  49.458  0.7ma 
vf 14 8.726 2.617 105.752 324.685 1.129
15 9.051 2.673 101.951 318.596 1.404
 Constraints 16 9.875 2.870 93.445 312.431 1.361

0.00025 m 0.28 1.05 0.60


v s  0.1 17 10.877 3.321 84.841 331.831 1.557
a f
18 11.209 3.364 82.326 324.987 1.601
ma0.07 v 0f .29
0.826  6.3
s 0.012
m1a.155
v f  0.26
1  0.0003 ma 2.31
50  v f  150 ; 300  ma  700 ; 1  s  5

where t (s/cm2)-machining time needed to produce a unit of cut


surface, C (EUR/m)-operating cost per meter of cut, vf
(mm/min)-traverse speed, ma (g/min)-abrasive mass flow rate,
s (mm)-standoff distance. Optimization problem was solved
using Matlab multiobjective genetic algorithm solver for
determination a set of nondominated optimal points, known as
Pareto front. Parameters of the Matlab multiobjective genetic
algorithm solver are presented in Table 1. Fig. 3. Pareto front points
Miroslav Radovanovic / Procedia Manufacturing 47 (2020) 781–787 787
Miroslav Radovanovic/ Procedia Manufacturing 00 (2019) 000–000 7

Part of the Pareto front points obtained as a result of the [2] Akkurt A, Mustafa K, Ulvi S, Fevzi E. Effect of feed rate on surface
optimization process is presented in Table 2. roughness in abrasive water jet cutting applications. J Mater Process Tech
2004;147: 89–396.
Table 2 shows that the following factor levels can be [3] Axinte D, Kong M. An integrated monitoring method to supervise waterjet
selected for abrasive water jet cutting carbon steel S235, 6.5 machining. Cirp Ann-Manuf Techn 2009;58(1):303–306.
mm thick: traverse rate vf=127 mm/min, abrasive flow rate [4] Caydas U, Hascalik A. A study on surface roughness in abrasive waterjet
ma=300 g/min and standoff distance h=1 mm. machining process using artificial neural networks and regression analysis
For these factor levels, machining time needed to produce method. J Mater Process Tech 2008;202:574–582.
[5] Chakravarthy P, Babu N. A new approach for selection of optimal process
a unit of cut surface is t=7.266 s/cm2 (productivity is Q=8.258 parameters in abrasive water jet cutting. Mater Manuf Process 2007;14(4):
cm2/min) and operating cost per meter of cut is C=2.048 81-600.
EUR/m. [6] Chaturvedi V, Singh D. Multi response optimization of process parameters
of abrasive water jet machining for stainless steel AISI 304 using VIKOR
5. Conclusion approach coupled with signal to noise ratio methodology. J Adv Manuf
Syst 2015;14(2):107-121.
[7] Deepak D, Davim P. Multi-response optimization of process parameters in
Abrasive water jet (AWJ) cutting is a complex machining AWJ machining of hybrid GFRP composite by grey relational method.
process with many factors that determine performances. Procedia Manufacturing 2019;35:1211-1221.
Optimizing the AWJ cutting performances is essential in order [8] Gostimirovic M, Pucovsky V, Sekulic M, Rodic D, Pejic V. Evolutionary
to obtain better product quality, higher productivity and higher optimization of jet lag in the abrasive water jet machining. Int J Adv Manuf
Technol 2019; 101(9-12):3131-3141.
economic efficiency. To optimize machining performances [9] Hoogstrate AM. Towards high-definition abrasive waterjet cutting. PhD
optimization problem was defined. Optimization problem thesis, Delft University of Technology; 2000.
consists of two objectives (machining time needed to produce [10] Herghelegiu E. Contributions to the optimization of the water jet working
a unit of cut surface and operating cost per meter of cut), three parameters. PhD thesis, University Vasile Alecsandri of Bacau; 2011.
factors (abrasive mass flow rate, traverse speed and standoff [11] Iqbal A, Udar N, Hussain G. Optimization of abrasive water jet cutting of
ductile materials. J Wuhan Univ Technol 2011; 26(1):88-92.
distance), three constraints (perpendicularity tolerance, [12] ISO/TC 44 N 1770: Water jet cutting-Geometrical product specification
surface roughness limit and traverse speed for separation cut) and quality; 2010.
and bounds. Cut quality performances limit productivity and [13] Jagadeesh B, Babu PD, Mohamed MN, Marimuthu P. Experimental
economy performances in AWJ cutting applications. investigation and optimization of abrasive water jet cutting parameters for
Perpendicularity tolerance and surface roughness limit are the improvement of cut quality in carbon fiber reinforced plastic laminates.
J Ind Text 2018; 48(1):178-200.
conditions for good quality of the parts. In order to define the [14] Kovacevic M, Madic M, Radovanovic M, Rancic D. Software prototype
optimization problem it is necessary to know the for solving multi-objective machining optimization problems: Application
perpendicularity deviation, the surface roughness and the in non-conventional machining processes. Expert Syst Appl
traverse speed for separation cut in correlation to the process 2014;41(13):5657-5668.
factors. Equations for perpendicularity deviation and surface [15] Madic M, Markovic D, Radovanovic M. Comparison of meta-heuristic
algorithms for solving machining optimization problems. Facta
roughness were determined by design of experiment (DoE). Univeristatis, Series: Mechanical Engineering 2013;11(1):29-44.
Equation for traverse speed for separation cut was determined [16] Momber AW, Kovacevic R. Principles of abrasive water jet machining.
using specific cutting energy. Multi-objective optimizations, London: Springer; 1998.
where the objectives are generally conflicting, are difficult for [17] Nair A, Kumanan S. Multi-performance optimization of abrasive water
solving. Matlab multiobjective genetic algorithm solver was jet machining of Inconel 617 using WPCA. Mater Manuf Process
2017;32(6): 693-699.
used for solving multi-objective optimization problem. The [18] Qiang Z, Miao X, Wu M, Sawhney R. Optimization of abrasive waterjet
following factor levels were selected for abrasive water jet machining using multi-objective cuckoo search algorithm. Int J Adv
cutting carbon steel S235, 6.5 mm thick: traverse rate vf=127 Manuf Technol 2018;99:1257-1266.
mm/min, abrasive flow rate ma=300 g/min and standoff [19] Radovanovic M, Madic M. Optimizing factor levels based on cost in
distance h=1 mm. For these factor levels, machining time abrasive water jet cutting using GA. In: Cus F, Gecevska V, Chiampo F,
editors. Methods and techniques for industrial development. Maribor:
needed to produce a unit of cut surface is t=7.266 s/cm2 Faculty of Mechanical Engineering; 2015. p. 17-30.
(productivity is Q=8.258 cm2/min) and operating cost per [20] Radovanovic M. Multi-objective optimization of process performances
meter of cut is C=2.048 EUR/m. when cutting carbon steel with abrasive water jet. Tribol. ind.
2016;38(4):454-462.
Acknowledgement [21] Sharma V, Chattopadhyaya S, Hloch S. Multi response optimization of
process parameters based on Taguchi - Fuzzy model for coal cutting by
water jet technology. Int J Adv Manuf Technol 2011;56:1019-1025.
The paper is a result of the technological project TR35034 [22] Srinivasu D, Babu, N. An adaptive control strategy for the abrasive
which is supported by the Ministry of Education, Science and waterjet cutting process with the integration of vision-based monitoring
Technological Development of Republic of Serbia. and a neuro-genetic control strategy. Int J Adv Manuf Technol 2008; 38(5-
6):514-523.
[23] Valicek J, Hloch S, Kozak D. Surface geometric parameters proposal for
References the advanced control of abrasive waterjet technology. Int J Adv Manuf
Technol 2009;41:323-328.
[1] Aultrin J, Anand D. Multi-objective optimization of abrasive water jet [24] Yuvaraj N, Kumar MP. Multiresponse optimization of abrasive water jet
machining of aluminum 6061 alloy by Grey Relational Analysis. JCHPS cutting process parameters using TOPSIS approach. Mater Manuf Process
2016;9(1):410-417. 2015;30(7): 882-889.

You might also like