See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/369398677

Utilisation of fuzzy logic and genetic algorithm to seek optimal corrugated die
design for CGP of AZ31 magnesium alloy

Article in Advances in Materials and Processing Technologies · March 2023

DOI: 10.1080/2374068X.2023.2192135

CITATION READS

1 37

3 authors:

Muni Tanuja Anantha Tanya Buddi

CVRS College of Engineering Gokaraju Rangaraju Institute of Engineering & Technology
4 PUBLICATIONS 8 CITATIONS 50 PUBLICATIONS 312 CITATIONS

SEE PROFILE SEE PROFILE

Nageswara Rao
Koneru Lakshmaiah Education Foundation
60 PUBLICATIONS 409 CITATIONS

SEE PROFILE

All content following this page was uploaded by Muni Tanuja Anantha on 09 April 2023.

The user has requested enhancement of the downloaded file.

 Advances in Materials and Processing Technologies

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tmpt20

Utilisation of fuzzy logic and genetic algorithm

to seek optimal corrugated die design for CGP of
AZ31 magnesium alloy

Muni Tanuja Anantha, Tanya Buddi & Nageswara Rao Boggarapu

To cite this article: Muni Tanuja Anantha, Tanya Buddi & Nageswara Rao Boggarapu (2023):
Utilisation of fuzzy logic and genetic algorithm to seek optimal corrugated die design for
CGP of AZ31 magnesium alloy, Advances in Materials and Processing Technologies, DOI:
10.1080/2374068X.2023.2192135

To link to this article: https://doi.org/10.1080/2374068X.2023.2192135

Published online: 20 Mar 2023.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tmpt20
ADVANCES IN MATERIALS AND PROCESSING TECHNOLOGIES
https://doi.org/10.1080/2374068X.2023.2192135

Utilisation of fuzzy logic and genetic algorithm to seek

optimal corrugated die design for CGP of AZ31 magnesium
alloy
a,b
Muni Tanuja Anantha , Tanya Buddic and Nageswara Rao Boggarapua
a
Department of Mechanical Engineering, Koneru Lakshmaiah Education Foundation, Green Fields, Guntur,
India; bDepartment of Mechanical Engineering, ANURAG University, Hyderabad, India; cDepartment of
Mechanical Engineering, GRIET, Bachupally, Hyderabad, India

ABSTRACT ARTICLE HISTORY

Magnesium alloys are the first choice among other lightweight Accepted 9 March 2023
structural metals like aluminium, titanium and beryllium due to
KEYWORDS
low density and excellent corrosion resistance. However, low slip AZ31 Mg alloy; fuzzy logic
and cold plastic processing ability limits its applications. model; grooved die; genetic
Constrained groove pressing (CGP), one among the SPD techni­ algorithm; groove pressing;
ques, is well suited for improving the material properties. H13 steel
Corrugated dies are designed to investigate the deformation beha­
viour of AZ31 Mg alloy samples. To minimise the number of numer­
ical simulations, Taguchi’s L9 orthogonal array is selected for the
grooved die dimensions (viz. groove angle, groove width and coef­
ficient of friction). Elasto-plastic finite element analysis is performed
by implementing a multi-criterion-based genetic algorithm optimi­
sation tool for obtaining the optimal die geometry. Reduction in
total deformation, increase in the equivalent stress and elastic strain
of AZ31 sheet were possible with 50° groove angle, 3 mm groove
width and 0.22 friction coefficient. Mamdani-based fuzzy logic soft
computing tool is implemented for the model to examine the
deformation behaviour of AZ31 Mg alloy. 7.86 % deviation is
observed in the fuzzy logic model predictions in comparison with
simulation results. The suggested die design is well suitable for
multiple CGP passes, to improve the grain refinement in the pro­
cessed sheet.

1. Introduction
Demand of lightweight materials for the technology advancement of electric vehicles,
drones and aerospace structures is increasing day-by-day due to cost effectiveness in
power consumption and fuel economy [1]. Magnesium alloys attracted significantly
among low-density metals family owing to abundant availability, high specific strength
and high strength to weight ratio. Good amount of research carried out to improve the
mechanical properties. Due to hexagonal closed packed (HCP) crystal structure, magne­
sium alloys exhibited poor formability and low active slip system constraining to
structural applications [2,3]. Constrained groove pressing (CGP), friction stir processing

CONTACT Muni Tanuja Anantha [email protected] Research Scholar, Department of Mechanical

Engineering, Koneru Lakshmaiah Education Foundation, Green Fields, Vaddeswaram, Guntur, Andhra Pradesh 522502, India
© 2023 Informa UK Limited, trading as Taylor & Francis Group
2 M. T. ANANTHA ET AL.

(FSP) and accumulative roll bonding (ARB) are being introduced to improve ductility
and formability. These severe plastic deformation (SPD) processes provide ultra-fine
grain structure that improves formability of metals. Plastic deformation processes
employed for modification in structure at local or bulk level [4]. CGP process utilised
for structural bulk deformation [5,6]. This process includes an alternative pressing under
a set of corrugated dies followed by straight dies and rotated with 180° between them that
takes place in number of cumulative passes in a plane strain pure shear conditions [7].
According to hall-pitch relationship, this process results in uniformly distributed grain
size and improves ductility in metals. Simple tool design in CGP process can be
implemented in bulk deformation of large volumes [8,9]. Few investigations made on
the plastic deformation under CGP process at room temperature due to poor formability
of magnesium alloys [10].

1.1. Constrained groove pressing (CGP) process

At the first deformation stage as in Figure 1(a,b), sample is placed between two grooved
dies. At the second step(c), the deformed sample is flattened using two flat dies. In the
third step(d,e), plastic deformation is imposed on the flattened sample using grooved
dies of the first deformation step. In the fourth step(f), the deformed sample is positioned
between two flat dies which are operated simultaneously to flatten the sheet.
Yogesha et al. [11] studied the tensile and fracture behaviour of Al-Mg alloy (AA
5052) using cryo-rolling and post-deformation annealing. Deformed samples
enhanced the strength from 240 MPa to 291 MPa. Muni Tanuja et al. [12] observed
110 HV hardness and 90% thickness reduction in CR samples. Cryo-groove-rolled
specimens yielded the strength of 313 MPa and hardness of 122 HV due to large
density misplacement and small particle size. The post-annealing of sheets resulted in

Figure 1. Schematic representation of CGP method [11].

 ADVANCES IN MATERIALS AND PROCESSING TECHNOLOGIES 3

90% thickness shrinkage, improved ductility and fracture toughness. Sunil et al. [13]
initially used CGP to produce tiny grain size in AZ31 Mg alloy and examined the
surface characteristics, microstructure evolution on materials’ wettability, corrosion
resistance, in-vitro bioactivity, and cell adhesion. Fong et al. [14] examined the effects
of deformation path and temperature on microstructure, texture and mechanical
properties of AZ31B-O Mg alloy sheets developing an orthogonal CGP procedure
(involving four groove pressings, a 90° rotation of the sheet sample around its ND
after each pressing and a final flattening). Liu et al. [15] treated a highly deformed
material below recrystallisation temperature, which regains ductility and toughness
with considerable impact from precipitates within the matrix. As a result, straining
strengthens the ideal environment for further precipitation.
To achieve the refined grain structure, CGP process implemented in the body centred
cubic (BCC) and face Centered Cubic (FCC) crystal structured metals (such as aluminium,
copper base alloys, low carbon steels and nickel alloys). Thangapandian et al. [16] utilised
corrugated dies (V-groove, flat groove, and semi-circular) in repeating corrugation and
straightening (RCS) process, which improved the mechanical properties (tensile strength
and hardness) and uniform grain refinement with pass number. Moradpour et al. [17] used
CGP-cross route technique on Al – Mg alloy sheet to create fine grain structures with
increased strength. The number of passes relates directly to the degree of grain fineness.

1.2. Effect of die geometry on CGP process

CGP process is restricted to ductile materials. Die design is the primary cause of such
restriction. Improper die design causes failure of low ductile materials. Sajadi et al. [18]
investigated the effect of groove angle in the rubber pad-CGP process on commercial
pure aluminium. The larger groove angle (50°) led to fewer pass numbers and uniform
plastic strain distribution. Wang et al. [19] increased the shear intensity with small
groove width and created large pass numbers with low groove angle. Die design para­
meters play a significant role on the fine structure at room and elevated temperatures.
Lee et al. [20] defined the effective strainðεe Þ that depends on the die geometry (viz.,
groove angle and groove width) and friction coefficient between the die and sheet
mathematically as:
tan θ
εe ¼ 2m pffiffi (1)
3
Here, m is the number of passes and θis the groove angle.
Yanjin Guan et al. [21] worked on commercially pure copper sheets and observed
reduction (from 30 μm to 21 μm) in the grain size with CGP deformation, and 0.5 μm size
of sub-grains at Pass 3. Sharp changes of microstructure and mechanical properties observed
with 2 mm groove width. After 2 CGP passes, improved mechanical properties were achieved
for pure Ni sheets with 2 mm thick, 2 mm groove width and 45° groove angle.
Saeidi Googarchin et al. [22] employed ANN and GA techniques for generation of
data on AA5052 sheets with different die geometries. Improved tensile strength and
hardness of AA5052 sheet obtained using 45° groove angle, 1 mm groove width and 3
mm groove height.
4 M. T. ANANTHA ET AL.

Thuy et al. [2] utilised equivalent strain and strain inhomogeneity factor to define the
objective function. Hybrid GA implemented to find the optimal process parameters. 1st
pass of CGP on AZ 31 magnesium alloy with a new die geometry (die angle = 40°, groove
width = 2 mm and radius of corner = 0.3 mm) improved the yield strength by 85%.
Reham Reda et al. [23] worked on 1050 and 5052 aluminium alloys subjecting to 3
passes of CGP with intermediate annealing. For Al 1050, UTS, YS and the ratio of YS to
UTS increased with number of passes, whereas reduction in elongation. For Al 5052, UTS
and YS increased after 1st pass, decreased and increased after 2nd and 3rd passes.
These studies indicate the suitability of CGP method for achieving improved mechan­
ical properties of AZ31 Mg alloy with optimal die. There is a need for optimising the die
design for fabrication and using it in the CGP process to understand the deformation
characteristics of the processed AZ31 Mg alloy sheet.

1.3. Objective of the current study

By optimising the design parameters of the set of corrugated dies being using in CGP
process, it is possible to obtain the fine grain size of the processed AZ31 Mg alloy sheet at
room temperature. This will result in high equivalent stress and low deformation for the
processed AZ31 Mg alloy. A finite element analysis is performed to optimise the die
design parameters using genetic algorithm. Fuzzy logic model is implemented to predict
the performance indicators (output responses) for all possible combinations of die design
parameters. The AZ31 Mg alloy sheet is subjected to groove pressing by using a set of
corrugated dies. The simulation results are compared with the predicted values of fuzzy
logic model.

2. Die design
Two sets of dies are needed in the CGP process. One set of die serves for the corrugating
press, while the other set of die for the flattening press. Figure 2 illustrates the overview of
the corrugated die profile. Three parameters of die geometry under consideration (Groove
angle, Groove width and Coefficient of friction) with three levels are shown in Table 1.

Figure 2. An overview of die profile.

 ADVANCES IN MATERIALS AND PROCESSING TECHNOLOGIES 5

Table 1. Input parameters of die geometry with three levels.

Parameters Level 1 Level 2 Level 3
Groove Angle (°) 40 45 50
Groove width (mm) 3 4 5
Coefficient of friction 0.10 0.16 0.22

Table 2. Combination of die geometry as per the Taguchi’s L9 OA (orthogonal

array).
Coefficient of
Test No. Groove Angle (0) Groove Width (mm) Friction
1 40 3 0.10
2 40 4 0.16
3 40 5 0.22
4 45 3 0.16
5 45 4 0.22
6 45 5 0.10
7 50 3 0.22
8 50 4 0.10
9 50 5 0.16

The possible combination of die geometry as per the Taguchi’s design of experiments by
L9 OA (orthogonal array) is shown in Table 2. The die design is initiated with free sketch by
considering the dimension of the blank as 130 � 100 � 35 mm. The groove angle varies
from 40° to 50° and the groove width varies from 3 to 5 mm. Initially the number of
grooves within 130 mm blank length is determined. Each set of dies is modelled and
assembled as represented in Figure 3 using SolidWorks 2020 as per the L9 OA of Table 2.
The material opted for die is H13 steel (having splendid combination of toughness, heat

Groove angle = 40˚ Groove angle = 40˚ Groove angle = 40˚

Groove width = 3mm Groove width = 4mm Groove width = 5mm

Groove angle = 45˚ Groove angle = 45˚ Groove angle = 45˚

Groove width = 3mm Groove width = 4mm Groove width = 5mm

Groove angle = 50˚ Groove angle = 50˚ Groove angle = 50˚

Groove width = 3mm Groove width = 4mm Groove width = 5mm

Figure 3. Die design in SolidWorks as per the combination of die geometry parameters in Table 2.
6 M. T. ANANTHA ET AL.

Table 3. Mechanical properties and chemical composition of the die material H13 steel.
C Mn P S Si Cr Mo V Fe
Chemical Composition (wt %)
0.3889 0.3616 0.0132 0.0045 0.8921 5.3792 0.3616 0.0132 Balance
Tensile Strength Yield strength Hardness Modulus of Elasticity (GPa) Poisson’s ratio
(MPa) (MPa) (HV)
Mechanical Properties
1434 1356 242, 240, 239 44.85 0.35

and fatigue resistance) and widely used in hot and cold working. The material composition
of H13 steel and the properties associated with it are shown in Table 3.

3. Numerical simulations
Finite element method (FEM) adopted for strain distribution in the CGP process.
Elasto-plastic analysis performed to examine the plastic deformation and strain
localisation characteristics of the process after one complete pass [24]. ANSYS
software package was used to simulate the plastic deformation behaviour of speci­
mens after single groove pressing. Two-dimensional (2D) plane strain model is
appropriate for the expected minimal deformation in normal direction. CPE4 ele­
ments employed for modelling and performed simulations for the 2D planar strain
conditions. The work piece dies interface’s coefficient of friction was set at 0.1,
which is within the normal (0.05–0.1) range for cold forming metals. On shing with
4-node isoparametric 2D plane strain element, the optimised model consists of 5496
elements with 26,052 nodes. Dimensions of the specimen specified are: 3 mm
thickness and 80 mm length (width is unity along the plane normal direction in
plane-strain condition). Analytical rigid lines were used to model the dies, while
CPE4 mesh was used to model the specimen. ANSYS simulations performed for all
the combinations of die geometry in Table 2. AZ31 Mg alloy sheets (Table 4 shows
the chemical composition and properties) are used to undergo deformation as per
the combination of parameters in L9 OA of Table 2.
ANSYS – Static structural simulations performed to examine the elasto-plastic
behaviour and changes in mechanical properties of AZ31 sheets during CGP process.
To attain the optimum design, the steps employed are: Defined the 3-D geometry
(length � width � height) of the die (130 � 100 � 35 mm) and AZ31 Mg alloy sheet
(110 � 90 � 3 mm); Attached the geometry (created in SolidWorks 2020) to the work­
bench for performing simulations; Specified the material properties of H13 steel for the
die, whereas the properties of AZ31 Mg alloy for the sheet; CPE4 triangular elements
are selected for modelling, a Four-noded plane strain element. These elements are most
suitable for analysing thin to moderately thick shell structures. Regarding the boundary
conditions, the bottom die is fixed and 40 tons distributed load applied on the upper die
surface at a constant strain rate of 1 mm/s at room temperature. Numerical simulations
performed by clicking on the analysis type ‘structural analysis’. Simulation results of
total deformation, elastic strain and equivalent stress on AZ31 Mg alloy sheet are
summarised in Table 5.
 ADVANCES IN MATERIALS AND PROCESSING TECHNOLOGIES 7

Table 4. Mechanical properties and chemical composition of AZ31 Mg alloy.

Al Zn Mn Zr Si Cu Ca Fe Ni Mg
Chemical Composition (wt %)
2.7201 0.9561 0.2576 0.0010 0.0204 0.0016 0.0026 0.0087 0.0031 95.94
Tensile Strength Yield strength Hardness Modulus of Elasticity (GPa) Poisson’s ratio
(MPa) (MPa) (HV)
Mechanical Properties
238.17 178.54 54, 52, 52 215.7 0.28

Table 5. Simulation results of output responses under different die geometry.

Coefficient of Total deformation Equivalent Equivalent
S. No. Groove Angle (0) Groove Width (mm) Friction (mm) Elastic strain stress (MPa)
1 40 3 0.10 2.748 0.4312 191
2 40 4 0.16 8.166 0.4282 180
3 40 5 0.22 9.505 0.3577 151
4 45 3 0.16 4.338 0.5059 214
5 45 4 0.22 9.938 0.5398 233
6 45 5 0.10 11.130 0.4408 190
7 50 3 0.22 5.212 0.6416 241
8 50 4 0.10 11.150 0.8084 354
9 50 5 0.16 13.110 0.5895 238

4. Optimum die design geometry using GA approach

Genetic algorithm (GA) is well equipped to handle multi-objective optimisation pro­
blems. A universal single-objective GA can be modified to find a variety of distinct non-
dominated solutions in a single run [25–27]. GA’s ability to simultaneously explore
different parts of a solution space allows for the discovery of a wide variety of solutions
for difficult problems with non-convex, discontinuous, and multi-modal solution spaces
[28]. The crossover GA operator may use the structures of successful solutions in
connection to distinct objectives to offer new non-dominated solutions in sections of
the Pareto front [29]. The majority of multi-objective GAs do not require the user to
rank, scale or weigh the objectives. GA is the most commonly used heuristic algorithm,
for resolving multi-objective design and optimisation issues. The results of ANSYS
simulations further processed to trace the optimised set of parameters for effective die
geometry through GA optimisation technique. Three variables are used to define the
geometry of the grooved die (viz., groove width, groove angle and coefficient of friction
between the die and the work-piece), which are designated by A, B and C for simplicity.
To design an effective set of grooved dies, the total deformation should be minimum,
whereas equivalent stress and elastic strain can be maximum.
The optimal geometric values are to be estimated by implementing GA technique in
MATLAB. Regression models for the performance indicators (viz., total deformation,
elastic strain and equivalent stress) in terms of the three variables (A: groove angle, B:
groove width and C: Coefficient of friction) are:
8 M. T. ANANTHA ET AL.

Total deformation ðmmÞ ¼ 65:636 þ 0:854A þ 22:294B þ 24:845C 6:14� 10 3 A2

2:0775B2 193:82C2
(2)

Elastic strain ¼ 40:441 3:824A þ 4:232B 1:738C þ 0:0819A2 0:0977B2

þ 7:975C2 (3)

Equivalent stress ðMPaÞ ¼ 24:872 3:824A þ 4:232B 1:738C þ 0:05A2 þ 5:105B2

þ 4:68� 103 C2
(4)
The MATLAB simulation results for 18 trials are in Table 6. The optimal design
parameters to achieve minimum total deformation, high equivalent stress and elastic
strain in the sample are in Trial 8 having the die with groove angle 50° and groove width
of 3 mm.
To locate solutions in the acquired Pareto front that is as diversified as feasible and as
close to the real Pareto optimal solution [30]. To achieve an even distribution along the
Pareto optimum front, the obtained data must be penalised by a loss in fitness [31]. From
numerical simulation results, it is clear that minimum total deformation (first Objective
function) and adequate elastic strain (second objective function) will produce fine grains.
Hence, there is a possibility of a group of optimum solutions, which are called ‘Pareto
frontier or Pareto curve or Pareto front’ in a multi-objective optimisation. The Pareto
curve (see Figure 4) from MATLAB simulations indicate an even distribution of data
between objective 1 and objective 2.
The distribution of total deformation, equivalent strain and von Mises stress with
the die of groove angle : 50° and Groove width : 3 mm is more uniform when
compared with the other simulation trials in Table 6. This optimised design is

Table 6. MATLAB results for different simulation trials.

Total Equivalent
Groove Angle Groove Width Coefficient of deformation Elastic stress
Simulation Trails (0) (mm) Friction (mm) strain (MPa)
1 50.0044 4.7592 0.2192 12.571 0.6023 242.338
2 50.0027 3.2001 0.2174 5.7687 0.7425 240.119
3 50.0063 4.8571 0.2171 12.768 0.5978 236.563
4 50.0035 4.6825 0.2176 12.351 0.6122 245.672
5 50.0035 4.3829 0.2194 11.346 0.6345 254.234
6 50.0033 3.4981 0.2180 8.4387 0.7245 297.677
7 50.0030 3.0523 0.2164 6.4503 0.7434 239.215
8 50.0029 3.0001 0.2199 5.7621 0.7531 245.343
9 50.0019 3.1361 0.2187 7.5630 0.7322 245.670
10 50.0058 4.4414 0.2199 11.594 0.6323 252.344
11 50.0033 4.2602 0.2192 11.348 0.6634 261.967
12 50.0063 5.0000 0.2200 13.565 0.5745 236.379
13 50.0027 3.8606 0.2195 10.788 0.6834 338.299
14 50.0030 3.6993 0.2176 9.0209 0.6998 320.456
15 50.0064 4.5153 0.2195 11.873 0.6279 249.567
16 50.0031 3.5739 0.2192 8.6755 0.7123 305.780
17 50.0039 4.6378 0.2195 12.094 0.6239 248.453
18 50.0048 3.7649 0.2178 9.4656 0.6834 326.881
 ADVANCES IN MATERIALS AND PROCESSING TECHNOLOGIES 9

Figure 4. Pareto curve between objective functions 1 & 2.

recommended for fabrication. Figures 5 to 7 show the total deformation, elastic

strain and equivalent stress for the optimised design under the specified boundary
conditions.
The optimal die design parameters from the results of GA are: Groove angle, A = 50°;
Groove width, B = 3 mm; and Coefficient of Friction, C = 0.22 to attain minimum total
deformation, high equivalent stress and elastic strain of the sample. Numerical validation
of Trial 8 with GA result is as follows.

Figure 5. Total deformation of the sample after 1st pressing.

Figure 6. Elastic strain distribution after 1st pressing.

10 M. T. ANANTHA ET AL.

Figure 7. Equivalent (von Mises) stress distribution after 1st pressing.

From Eq: ð2Þ;

Total Deformation ¼ 65:636 þ 0:854A þ 22:294B þ 24:845C 6:14� 10 3 A2
2:0775B2 193:82C2
¼ 65:936 þ 0:854� 50 þ 22:294� 3 þ 24:845� ð0:22Þ
�
6:14� 10 3 502 2:0775� 32 193:82� ð0:22Þ2
¼ 5:6833 mm; this is in acceptance with the result of GA for Trial 8

From Eq: ð3Þ;

Elastic Strain ¼ 40:441 0:143A þ 0:749B 2:948C þ 0:0189A2 0:0977B2 þ 7:975C2
¼ 40:441 0:143� 50 þ 0:749 � 3 2:948� ð0:22Þ þ 0:01859� 502
0:0977� 32 þ 7:975� ð0:22Þ2
¼ 0:7645 which is in acceptance with the result of GA for Trial 8:

From Eq: ð4Þ;

Equivalen tstress ¼ 24:872 3:824A þ 4:232B 1:738C þ 0:05A2 þ 5:105B2 þ 4:68� 10C3 C2
¼ 24:872 3:824� 50 þ 4:232� 3 1:738� ð0:22Þ þ 0:05� 502 þ 5:105� 32
þ 4:68� 103� ð0:22Þ2
¼ 243:44 MPa; whichis in acceptance with the result of GA for Trial 8:

5. Fuzzy logic predictive model

A multi-valued logic known as fuzzy logic, allows for the establishment of inter­
mediate values between conventional grading scales like true/false, yes/no, high/low,
and so forth [32,33]. This method enables the machine to think more like a person by
expressing assessment ideas like moderately tall or exceptionally swift quantitatively
and analysing them. The data is simulated in MATLAB using Mamdani fuzzy logic
model. The validity of the model depends on its selection [34]. This model uses
defuzzification technique, which is the centroid of area (COA). It offers incredibly
precise prediction and analysis. In order to minimise ANSYS simulation time, a fuzzy
logic predictive model (based on mamdani fuzzy logic) is implemented [35–37]. The
 ADVANCES IN MATERIALS AND PROCESSING TECHNOLOGIES 11

Figure 8. Mamdani based fuzzy logic input and output membership relationship for grooved die
design.

model provides a constructive rule-based interference system between input and

output membership functions. This technique is more appropriate for the die design
having several parameters.
Apart from L9 OA, other combination of input parameters can provide more effective
design. Online simulation for all sets of input parameters is tedious and time consuming.
Therefore a fuzzy-based predictive model is constituted to get the output response for the
intended input geometry. The relation between input and output membership function is
shown in Figure 8 for the grooved die design. A set of rules established between input and
output membership functions based on L9 OA and ANSYS simulations results in Table 7.
Fuzzy model predictions in Table 8 have close resemblance with ANSYS simulation
results.
Figure 9 demonstrates the variation of die parameters (Groove width and
Groove angle) with respect to Total deformation and Elastic strain. It is clear
from the results that groove angle (50°) and groove width (3 mm) leads to
decrease in total deformation and increase in elastic strain which is in accordance
with the simulation results as shown in Table 5. An error chart represented in
Figure 10 depicts an average error of 7.05 % for total deformation, 8.45 % error
for elastic strain and 8.09 % error for equivalent stress i.e. An average permissible
error of 7.86% was achieved in comparison with simulation results, after the
implementation of fuzzy predictive model.

Table 7. Rule editor established between input and output membership functions.
Coeff. of
Groove angle Groove width friction Total deformation Elastic strain Equivalent stress (MPa)
IF L & L & L THEN VL L L
IF L & M & M THEN M VL VL
IF L & H & H THEN H VL VL
IF M & L & M THEN L L VH
IF M & M & H THEN H M M
IF M & H & L THEN H L L
IF H & L & H THEN L M M
IF H & M & L THEN H VH VH
IF H & H & M THEN H M M
VL: VERY LOW; L: LOW; M: MEDIUM; H: HIGH; VH: VERY HIGH
12 M. T. ANANTHA ET AL.

Table 8. Comparison of fuzzy model predictions with ANSYS simulation results.

ANSYS Simulations Fuzzy Predictions
Total Equivalent Equivalent
deformation Elastic stress Total deformation Elastic stress
S. No. (mm) strain (MPa) (mm) strain (MPa)
1 2.748 0.4312 191 2.88 0.471 202
2 8.166 0.4282 180 7.93 0.433 185
3 9.505 0.3577 151 10.3 0.433 185
4 4.338 0.5059 214 4.64 0.495 213
5 9.938 0.5398 233 9.96 0.574 249
6 11.13 0.4408 190 10.3 0.495 213
7 5.212 0.6416 241 5.9 0.574 249
8 11.15 0.8084 354 9.96 0.709 309
9 13.11 0.5895 238 11.9 0.583 253

Figure 9. Fuzzy prediction on total deformation and elastic strain for the specified groove width and
groove angle.

Figure 10. An error chart depicting the deviation of numerical simulations and fuzzy model
predictions.
 ADVANCES IN MATERIALS AND PROCESSING TECHNOLOGIES 13

6. Conclusions
In order to trace the optimal grooved die geometrical parameters, numerical simula­
tions were carried out by adopting a simple and reliable hybrid optimisation techni­
que to enhance the grain refinement in AZ31 Mg alloy through CGP (constrained
groove pressing) process. The observations drawn from the present investigations are
as follows:

● The plastic deformation behaviour of the material has been studied by ANSYS
simulations.
● A fuzzy predictive model was implemented to model an overall groove pressing
operation for a set of input parameters. The aim is to observe the deformation
behaviour of AZ31 magnesium alloy after pressing with a set of optimised grooved
die geometry. The results obtained show a permissible error of 7.86% when com­
pared with simulation results.
● A multi-objective GA optimisation technique is employed to obtain the optimal die
design parameters as 50° groove angle, 3 mm groove width and 0.22 friction
coefficient with the motive of achieving minimum total deformation, high equiva­
lent stress and elastic strain.

Disclosure statement
No potential conflict of interest was reported by the authors.

ORCID
Muni Tanuja Anantha http://orcid.org/0000-0001-5808-7661

References
[1] Pouraliakbar H, Jandaghi MR, Heidarzadeh A, et al. Constrained groove pressing,
cold-rolling, and post-deformation isothermal annealing: consequences of their synergy
on material behavior. Mater Chem Phys. 2018;206:85–93.
[2] Thuy PT, Hue DTH, Ngung DM, et al. “A study on microstructure and mechanical
properties of AZ31 magnesium alloy after constrained groove pressing”. 2019 IOP Conf
Ser: Mater Sci Eng. 2019;611:611 012005.
[3] Brandão P, Infante V, Deus AM. Thermo-mechanical modeling of a high pressure turbine
blade of an airplane turbine engine. Procedia Struct Integr. 2016;1:189–196.
[4] Nazari F, Honarpisheh M. Analytical and experimental investigation of deformation in
constrained groove pressing process. Proc Inst Mech Eng Part C J Mech Eng Sci. 2019;233
(11):3751–3759.
[5] Mwita WM, Akinlabi ET. Numerical prediction of tensile yield strength and microhardness
of Ti6Al4V alloy processed by constrained bending and straightening severe plastic
deformation. Mater Res Express. 2019;6:No.10.
[6] Moskvichev E, Kozulin A, Krasnoveikin V, et al., “Numerical simulation of deformation
behavior of aluminum alloy sheets under processing by groove pressing method”, MATEC
Web of Conferences 143, 01011 (2018), –6 pages. 10.1051/matecconf/201814301011
[7] Solhjoet N, Varposhty AR, Mokhtarian H, et al. A Comparative study to evaluate the
efficiency of RCS and CGP processes. Indian J Sci Res. 2014;1(2):563–572.
14 M. T. ANANTHA ET AL.

[8] Mwita WM, Akinlabi ET, Sanusi KO. Performance and prospects of severe plastic deforma­
tion for effective biomedical titanium alloys. J Mod Mater. 2018;5(1):8–23.
[9] Xue KM, Liu M, Yan SL, et al. Synergic improvement of plasticity and strength of Al–Zn–
Mg–Cu alloy by grain refinement and precipitates redistribution using cyclic extrusion
compression. Adv Eng Mater. 2018;20(8):1–9.
[10] Hoseini-Athar MM, Mahmudi R, Prasath Babu R, et al. Microstructural evolution and
superplastic behavior of a fine-grained Mg–Gd alloy processed by constrained groove
pressing. Mater Sci Eng A. 2019;754:390–399.
[11] Zimina M, Bohlen J, Letzig D, et al. The study of microstructure and mechanical properties
of twin- roll cast AZ31 magnesium alloy after constrained groove pressing. IOP Conf Ser:
Mater Sci Eng. 2016;63:012078.
[12] Muni Tanuja A, Kumar A, Nageswara Rao B. Review on the application of CGP to Improve
AZ31 Mg alloy properties. In: Applications of computational methods in manufacturing
and product design, lecture notes in mechanical engineering.Singapore:Springer; 2022.
DOI:10.1007/978-981-19-0296-3_21
[13] Sunil BR, Kumar AA, Sampath Kumar TS, et al. Role of biomineralization on the degrada­
tion of fine grained AZ31 magnesium alloy processed by groove pressing. Mater Sci Eng C.
2013;33(3):1607–1615.
[14] Fong KS, Tan MJ, Ng FL, et al. Microstructure Stability of a fine-grained AZ31 magnesium
alloy processed by constrained groove pressing during isothermal annealing. J Manuf Sci
Eng. 2017 Aug;139(8):081007 (9 pages). DOI:10.1115/1.4036529.
[15] Liu K, Chen X-G. Development of Al–Mn–Mg 3004 alloy for applications at elevated
temperature via dispersoid strengthening. Mater Des. 2015;84(5):340–350.
[16] Thangapandian N, Prabu SB, Padmanabhan KA. Effects of die profile on grain refinement
in Al-Mg alloy processed by repetitive corrugation and straightening. Mater Sci Eng A.
2016;649:229–238.
[17] Moradpour M, Khodabakhshi F, Eskandari H. Microstructure–mechanical property rela­
tionship in an Al – Mg alloy processed by constrained groove pressing-cross route. Mater
Sci Technol. 2018;34(8):1003–1017.
[18] Sajadi A, Djavanroodi F, Borhani M. Experimental and numerical investigation of groove
pressed UFG pure aluminum. Int J Adv Des Manuf Technol. 2012;5(2):1–5.
[19] Wang Z, Guan Y, Wang T, et al. Microstructure and mechanical properties of AZ31
magnesium alloy sheets processed by constrained groove pressing. Mater Sci Eng A.
2019;745:450–459.
[20] Lee JW, Park JJ. Numerical and experimental investigations of constrained groove pressing
and rolling for grain refinement. J Mater Process Technol. 2002;130–131:208–213.
[21] Guan Y, Wang Z, “Numerical and experimental study on constrained groove pressing”,
InTech open, 10.5772/intechopen.68504.
[22] Saeidi Googarchin H, Teimouri B, Hashemi R. Analysis of constrained groove pressing and
constrained groove pressing-cross route process on AA5052 sheet for automotive body
structure applications. J Automob Eng. 2018;1–17:1436–1452. IMechE. DOI:10.1177/
0954407018785734.
[23] Reda R. Constrained groove pressing (CGP): die design, material processing and mechanical
characterization. Fluid Dyn Mater Proc FDMP. 2019;15(3):171–185.
[24] Kurra S, Prakash Regalla S, Pérez-Santiago R. Study on influence of process parameters in
incremental forming using finite element simulations and Taguchi orthogonal array. Adv
Mater Process Technol. 2015;1:201–209.
[25] Kumar S, Hariharan K, Digavalli R. Hybrid optimization of die design in constrained groove
pressing. Mater Manuf Process. 2020;35(6):687–699.
[26] Purohit GN, Sherry AM, Saraswat M. Optimization of function by using a new MATLAB
based genetic algorithm procedure. Int J Comput Appl. 2013;61(15):1–5.
[27] Tsoukalas VD. Optimization of porosity formation in AlSi9Cu3 pressure die castings using
genetic algorithm analysis. Mater Des. 2008;29(10):2027–2033.
 ADVANCES IN MATERIALS AND PROCESSING TECHNOLOGIES 15

[28] Karthikeyan R, Senthil Kumar V, Punitha A, et al. An integrated ANN – GA approach to

maximise the material removal rate and surface roughness of wire cut EDM on titanium
alloy. Adv Mater Process Technol. DOI:10.1080/2374068X.2020.1793267
[29] Alimirzaloo V, Sadeghi MH, Biglari FR. Optimization of the forging of aerofoil blade using
the finite element method and fuzzy-Pareto based genetic algorithm. J Mech Sci Technol.
2012;26(6):1801–1810.
[30] Wahed MA, Gupta AK, Gadi VSR, et al. Parameter optimisation in V-bending process at
elevated temperatures to minimise spring back in Ti-6al-4V alloy. Adv Mater Process
Technol. 6:350–364. DOI:10.1080/2374068X.2020.172865.
[31] Ghorbanhosseini S, Fereshteh-Saniee F. Multi-objective optimization of geometrical para­
meters for constrained groove pressing of aluminium sheet using a neural network and the
genetic algorithm. JCAMECH. 2019 December;50(2):275–281. DOI:10.22059/jcamech.
2018.267948.335.
[32] Bhiradi I, Raju L, Hiremath SS. Adaptive neuro-fuzzy inference system (ANFIS): modelling,
analysis, and optimisation of process parameters in the micro-EDM process. Adv Mater
Process Technol. 2020;6(1):133–145.
[33] Maghzi P, Mohammadi M, Pasandideh SHR, et al. Operating room scheduling optimization
based on a fuzzy uncertainty approach and metaheuristic algorithms. Int J Eng Trans
B Appl. 2022;35(2):258–275.
[34] Azami G, Narm HG. Bandwidth management with congestion control approach and fuzzy
logic. Int J Eng Trans A Basics. 2021;34(4):891–900.
[35] Nozari H, Tavakkoli-Moghaddam R, Ghareman-Nahr J. A neutrosophic fuzzy program­
ming method to solve a multi-depot vehicle routing model under uncertainty during the
COVID-19 pandemic. Int J Eng Trans B Appl. 2022;35(2):360–371.
[36] Srivastava V, Gupta M. Evaluation and prediction of self-healing assessments for AA2014
based hybrid smart composite structures: a novel fuzzy logic approach. Int J Eng Trans
A Basics. 2022;35(10):1841–1857.
[37] Srivastava V, Gupta M. Design and prediction of healing assessment for AA2014-Nitinol
strip-Solderalloy based hybrid self- healing composite structure via Taguchi analysis and
fuzzy logic approach. Mech Adv Mater Struct. 2022;1–20. DOI:10.1080/15376494.2022.
2092789

View publication stats