You are on page 1of 5

Materials Today: Proceedings xxx (xxxx) xxx

Contents lists available at ScienceDirect

Materials Today: Proceedings


journal homepage: www.elsevier.com/locate/matpr

Simulation of three-point bending test on aluminium (Al 1100) sheets


with parameters optimization
Karpagaraj Anbalagan a,⇑, R. Sarala b, S.M. Sivagami b, S. Thamizhmanii c
a
Department of Mechanical Engineering, National Institute of Technology Puducherry, Karaikal 609609, India
b
Department of Mechanical Engineering, Alagappa Chettiar Government College of Engineering and Technology, Karaikudi, Tamilnadu 630003, India
c
Department of Mechanical Engineering, Academy of Maritime Education and Training (AMET), Chennai 603112, India

a r t i c l e i n f o a b s t r a c t

Article history: Aluminium 1100 (Al 1100) is light in weight and has good formability properties. Due to this, Al 1100 can
Available online xxxx be formed into the desired shape with various forming processes. Among the forming process, the three-
point bending test is one of the essential processes. So, to identify its behavior under a three-point bend,
Keywords: this simulation work is needed. The objective of this work is to know the deformation and computational
Three-point bending test time through the simulation work. The three major input parameters of the three-point bending test,
Optimization such as the thickness of the material (1.2,1.6, 2 mm), mesh size (1,3,5 mm), and deformation depth (-
ANOVA
15, 20, 25 mm), are selected for the simulation study. Using the above process parameter L9 orthog-
Simulation
Aluminium sheet
onal array was formed. The Abacus software is used for the simulation work. Analysis Of Variance
(ANOVA) is prepared to find the contribution of each parameter. The regression equation is formed based
on the ANOVA results for future research. Also, the results are checked with the confidence interval to
identify their smoothness. With a 95% confidence interval, results are fit for deformation behavior and
computational time. The mesh size is scored the first rank for deformation and computational time.
The 10% error range does not cross between the simulated and predicted value.
Copyright Ó 2023 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the International Confer-
ence on Future Technologies in Manufacturing, Automation, Design & Energy.

1. Introduction Sehhat et al. investigated the Laser Foil Printing (LFP) for the AI
1100 and developed the operational window with optimized
Aluminium (Al) is a lightweight and abundant metallic element parameters. ANOVA was used to identify the contribution of each
in Earth’s crust. Al proved an ideal material for building heavier- parameter [6]. Response Surface Analysis (RSM) predicts the melt
than-air aircraft [1]. Every kilogram of Al used in a car reduces pool depth. Gau et al. studied the micro forming of the AI 1100
the overall weight of the vehicle by one kilogram. From Al: engine sheet concerning its thickness to grain diameter (T/D). There was
radiators, wheels, bumpers, suspension parts, engine cylinder a strong correlation exists between the T/D ratio. Also, the spring
blocks, etc., are fabricated [2]. In fin stock, heat exchanger fins, back increases once the T/D percentage is small [7]. Few kinds of
spun hollow ware, dials and nameplates, decorative parts, giftware, literature are fully focused on the experimental analysis of the
cooking utensils, rivets and reflectors, and sheet metal work [3]. three-point bending test for aluminum alloys [8,9]. After carefully
Karpagaraj et al. utilized the abacus package for studying the studying the literature, only limited works were published with
three-point bending behavior of Ti-6AL-4 V alloy sheets [4]. The three points bending of AI 1100 sheets. Therefore, this study uses
experimental work results coincide with the simulated work. This Abaqus for a three-point bending test through finite element mod-
clearly showed the suitability of the Abacus package for simulating eling and simulations.
and performing mechanical behaviors. Sahu et al. studied the
strain-hardening effect of Al alloy AA1100 and used the Johnson-
2. Materials and methods
Cook material model for impact analysis [5].
This study performs a three-point bending simulation with thin
⇑ Corresponding author. sheets of AI 1100. The material properties are taken from the liter-
E-mail address: karpagaraj@nitpy.ac.in (K. Anbalagan). ature and used in the abacus software. Using the Taguchi method,

https://doi.org/10.1016/j.matpr.2023.05.068
2214-7853/Copyright Ó 2023 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the International Conference on Future Technologies in Manufacturing, Automation, Design &
Energy.

Please cite this article as: K. Anbalagan, R. Sarala, S.M. Sivagami et al., Simulation of three-point bending test on aluminium (Al 1100) sheets with param-
eters optimization, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2023.05.068
K. Anbalagan, R. Sarala, S.M. Sivagami et al. Materials Today: Proceedings xxx (xxxx) xxx

Table 1 Boundary conditions are applied to the specimen and the roll-
Input parameters [10–11] and material properties [*,#]. ers. Fig. 1(a-b) shows the assembled view with mesh and con-
S.no Input Parameters straints. The supporter and load pins are placed in position using
1 The thickness of the sheet in mm (A) 1.2 1.6 2 the test constraints.
2 Mesh (global size) mm (B) 1 3 5 The Fig. 1 (b) constrained view has been achieved by keeping
3 Loading pin displacement (depth) mm (C) 15 20 25 the supports will be constrained type symmetry encastre U1 = U
Young’s modulus 70 GPa, Poisson’s ratio 0.33, Density (x1000, kg/m2) 2.71,
2 = U3 = UR1 = UR2 = UR3 = 0. The type loading pin and displace-
*
https://www.metalmensales.com/aluminum-1100-properties-products/. ment load have been constrained as follows U1 = 0, U2 = -20,
#
Design data book by kalaikathir achchagam pg (1.30–1.32). U3 = 0, UR1 = UR2 = UR3 = 0. In the charging step, the U2 = 0 is
modified to check the spring back, which is before the exact value
the L9 orthogonal array is built using the list of parameters as men- of U2 = -20 [4]. Overall mesh is done with a size of 1 mm for the
tioned in Table 1. test specimen, as shown in Fig. 1 (a). Type of mesh used in Alu-
The laptop configuration is mentioned as follows, Intel(R) Core minium sheet C3D8R: An 8-node linear brick, reduced integration,
(TM) i5-9300H CPU @ 2.40 GHz, 8192 MB RAM and NVIDIA hourglass control. For cylindrical support R3D4: A 4-node 3-D
GeForce GTX 1650–3962 MB. All the trials are performed by using bilinear rigid quadrilateral mesh is given [4,15,16].
the same laptop configuration. The ASTM standard of B209 is used
to configure the three-point bend test sample in Abacus software
[4]. Before setting Abacus trials following assumptions are made.
3. Results and discussion

 A central surface of the shell module with a global coordinate


In the results and discussion, the simulation made with the
system is adopted.
Abacus and the prediction made with regression equations is com-
 Stresses in the loaded beam do not reach the limit values. Fail-
pared. The deviations are studied and identified with the error per-
ure criterion the, Tsai-Wu model is followed.
centages. The role of each parameter is ranked based on the
 The load is quasi-static, eliminating the inertia forces. Keeping
ANOVA model.
the supports will be constrained type symmetry encastreU1 =
U2 = U3 = UR1 = UR2 = UR3 = 0, and the type of loading pin dis-
placement load has been constrained U1 = 0, U2 = -20, U3 = 0,
3.1. Simulation work
UR1 = UR2 = UR3 = 0 and then the De-charging step the
U2 = 0 is modified for the checking of the spring back which
Before the simulation, the three-point bending test setup (load
is before the same value of U2 = -20 [4].
pin, Al sheet, and support) is assembled. After creating the step and
interaction modules, the necessary contacts are applied at the
All the simulation, the element type (for Aluminium sheet
point of contact for support and load. The boundary conditions
C3D8R: An 8-node linear brick, reduced integration, hourglass con-
are introduced, and meshing is created based on Table 1. Finally,
trol. for cylindrical support R3D4: A 4-node 3-D bilinear rigid
the job is checked with the data value and simulated. The visual-
quadrilateral) is taken as a constant [12]. Three-point bending test
ization can be seen in Fig. 2 (Specimen under maximum and after
sheets, rollers, and punch with dimensions are created in the aba-
releasing the load).
cus environment. Partition dimensions are mentioned as 5 mm
Using AutoCAD, the specimen after the spring back is calcu-
from the right and left sides. The necessary material properties
lated. Distance is measured at the end of final load to the release.
for AI 1100 are taken from John Cook’s equation through the
Later the spring back angle is measured by finding the difference
stress–strain curve [10,13–15].
between at maximum load to after the release condition.

Fig. 1. Assembly model (a) with mesh and (b) constraints.

Fig. 2. Specimen under maximum bending and after bending.

2
K. Anbalagan, R. Sarala, S.M. Sivagami et al. Materials Today: Proceedings xxx (xxxx) xxx

Table 2
Simulated and predicted results.

S.no Thickness (mm) Mesh size (mm) Loading pin displacement (mm) Computational time (min) Deformation (mm)
Simulated Predicted Simulated Predicted
1 1.2 1 15 35 33.385 1.124 1.1785
2 1.2 3 20 13.1 13.9 3.6 3.48
3 1.2 5 25 3.05 3.295 4.2 4.0875
4 1.6 1 20 19.3 20.66 2.2 1.995
5 1.6 3 25 5.3 5.165 3.2 3.158
6 1.6 5 15 7.6 7.755 5.2 5.28
7 2 1 25 10.8 11.335 1.9 1.7315
8 2 3 15 6.4 5.915 6.456 6.287
9 2 5 20 7.8 8.1 3.8 4.01

3.2. Optimization work A comparison of predicted and simulated results is listed in


Table 2. It is found that the error is less than 10% [19,20].
After completing all the cases, the results are listed in Table 2.
The spring-back for one of the nine trials is compared with the lit- Computational time (Min) = 71.0–7.6 *[Thickness (mm)]
erature [4]. Via probability plot, the results obtained from the nine 5.40*[Mesh Size (mm)] + 1.539* [loading pin displacement
trials are plotted and compared with 95% of the confidence inter- (mm)] . . .1.
val. The result points are stuck and close to the line, which also Deformation (mm) = -0.87 + 1.775 *[Thickness (mm)] + 0.919 *
clearly reflects that no deviations occurred in our results [17,18] [Mesh Size (mm)] + 0.0667*[loading pin displacement(mm)]
(Fig. 3). . . .2

Fig. 3. Normal probability plot (a) Computational time (min) and (b) Deformation (mm).

3
K. Anbalagan, R. Sarala, S.M. Sivagami et al. Materials Today: Proceedings xxx (xxxx) xxx

Table 3 put deformation. With delta values of 24.039,15.393,12.701for


Response table for means (Deformation). mesh size, loading pin displacement, and thickness, respectively
Level Thickness (mm) Mesh size(mm) Loading pin Displacement(mm)  Based on the trial and their results, ANOVA ranking has been
1 4.140 1.132 7.398 made, and it has been found that mesh size is the vital param-
2 6.547 5.754 2.369 eter of the whole outcome compared to the other two inputs for
3 16.841 22.907 17.762 output computational time. Delta values of 6.236,5.555,3.820
Delta 12.701 24.039 15.393 for mesh size, loading pin displacement, and thickness,
Rank 3 1 2
respectively.

Table 4
CRediT authorship contribution statement
Response Table for Means (Computational time).
Karpagaraj Anbalagan: Conceptualization, Project administra-
Level Thickness Mesh Size (mm) Loading pin displacement (mm)
(mm)
tion, Resources, Software, Supervision, Writing – original draft. R.
Sarala: . S.M. Sivagami: . S. Thamizhmanii: .
1 10.012 11.721 4.742
2 7.133 6.133 8.300
3 6.193 5.484 10.297 Data availability
Delta 3.820 6.236 5.555
Rank 3 1 2
Data will be made available on request.

Declaration of Competing Interest


From Table 2 Minimum error percentage for the (Deformation)
trial five is 1.31 %, and the maximum error percentage is identified The authors declare that they have no known competing finan-
for trial four as 9.32 %. The above table (Table 2) depicts that the cial interests or personal relationships that could have appeared
minimum error percentage for the (computational time) of trial to influence the work reported in this paper.
six is 2.04 %, and the maximum error percentage for the time of
trial 3 is 8.03 % [20]. Acknowledgement
The ranking process shows that the mesh size is the most con-
vincing parameter among the three parameters affecting computa- The authors love to thank IIT – Madras and Nit-Trichy for their
tional time and deformation. According to the level, the mesh size support in completing the simulation and optimization work.
at the first level is 1.132 mm, followed by 5.754 mm and
22.039 mm at level 2 and level 3, respectively. The delta value
References
was about 24.039 for the mesh size (Table 3). Since the loading
pin displacement has secured the second position, the maximum [1] K. Anderson, J.G. Kaufman, J. Weritz, (Eds.). (2019). ASM Handbook: Volume 2B
displacement value is at level 3. Properties and Selection of Aluminum Alloys. Materials Park, OH, USA: ASM
International.
From Table 4, the mesh size is the most convincing parameter
[2] W.S. Miller, L. Zhuang, J. Bottema, A. Wittebrood, P. De Smet, A. Haszler, A.J.M.
compared to displacement and thickness for output parameter S. Vieregge, Recent development in aluminium alloys for the automotive
deformation. industry, Mater. Sci. Eng. A 280 (1) (2000) 37–49.
Similarly, for the output parameter computational time, the [3] S. Ghanaraja, H. Rajashekar, K.S. Ravikumar, B.M. Madhusudan, Synthesis and
Study of Microstructure and Mechanical Properties of Cast Al1100 (Mg)-SiC
ranking process shows that the Mesh size is the most convincing Composites, Mater. Today:. Proc. 5 (1) (2018) 2765–2772.
parameter among the three parameters that affect the computa- [4] A. Karpagaraj, N. SivaShanmugam, B. Suresha, S.A. Vendan, (2018). Studies on
tional time and deformation. According to the level, the mesh size Spring Back Effect of TIG Welded Ti-6Al-4V Sheets. In Simulations for Design
and Manufacturing (pp. 147-171). Springer, Singapore.
at the first level is 11.721 mm, followed by 6.113 mm and [5] S. Sahu, D.P. Mondal, M.D. Goel, M.Z. Ansari, Finite element analysis of AA1100
5.484 mm at level 2 and level 3, respectively. elasto-plastic behaviour using Johnson-Cook model, Mater. Today:. Proc. 5 (2)
The delta value is about 6.236 for the mesh size. Since the load- (2018) 5349–5353.
[6] M.H. Sehhat, B. Behdani, C.H. Hung, A. Mahdianikhotbesara, (2021).
ing pin displacement has secured the second position, the maxi- Development of an empirical model on melt pool variation in laser foil
mum displacement value is at level 3. By Table 4, the mesh size printing additive manufacturing process using statistical analysis.
is the most convincing parameter compared to displacement and Metallography, Microstructure, and Analysis, 10(5), 684-691.strength of fsw
samples using taguchi optimization technique.‘‘ J. Crit. Rev. 7 (2020): 179-182.
thickness.
[7] J.T. Gau, C. Principe, F. Yang, (2006, January). An experimental study of the
influence of size effect on spring back of micro sheet forming. In International
4. Conclusion Manufacturing Science and Engineering Conference (Vol. 47624, pp. 313-318).
[8] K. Logesh, V.B. Raja, Investigation of mechanical properties of AA8011/PP/
AA1100 sandwich materials, International Journal of Chem Tech Research 6 (3)
In this study, a three-point bending simulation is performed (2014) 1749–1752.
using Abaqus software in the medium-level configuration laptop. [9] H. Hino, Y. Takayama, H. Kato, H. Watanabe, (September 5-9, 2010)
Deformation Behavior in Three-point Bending of Aluminum Alloy
The input parameters like the thickness of the sheet (mm) [1.2 to Honeycomb Structures. Proceedings of the 12th International Conference on
2], size of the mesh (mm) [1to5], and loading pin displacement Aluminium Alloys, Yokohama, Japan, The Japan Institute of Light Metal 2010,
(mm) [-15 to 25] and their effect are studied through simulation. pp. 625-630,
[10] K. Anderson, J. Weritz, J.G. Kaufman, 1xxx Aluminum Alloy Datasheets.
Properties and Selection of Aluminum Alloys. ASM Handbooks, ASM
 Nine trials are conducted based on the Taguchi method to study International (2019).
the effect of selected parameters. Based on the simulation [11] R. Jain, S.K. Pal, S.B. Singh, Indian Institute of Technology, Kharagpur, India.
Computational Methods and Production Engineering: Research and
results regression equation has been formed and compared
Development, 125 (2017).
with the simulated values, and the results are in the acceptable [12] A. Dylewski, & W. Łogin, (2019, March). Simulation of three-point bending of a
range (10% error). simple composite beam using the commercial FEM software. In AIP
 ANOVA ranking has been made based on the trail of their Conference Proceedings (Vol. 2078, No. 1, p. 020085). AIP Publishing LLC.
[13] K.P.V. Namburi, A.F. Kothasiri, V.S.M. Yerubandi, Modeling and simulation of
results; it has been found that mesh size is the vital parameter Aluminum 1100 alloy in an extrusion process, Mater. Today:. Proc. 23 (2020)
of the whole outcome compared to the other two inputs for out- 518–522.

4
K. Anbalagan, R. Sarala, S.M. Sivagami et al. Materials Today: Proceedings xxx (xxxx) xxx

[14] A.S. Khan, A. Pandey, T. Stoughton, Evolution of subsequent yield surfaces and International Journal of Materials and Biomaterials Applications 2 (1) (2012)
elastic constants with finite plastic deformation. Part II: A very high work 1–4.
hardening aluminum alloy (annealed 1100 Al), Int. J. Plast 26 (10) (2010) [18] B. Sutharson, M. Rajendran, R. Sarala, A. Karapagaraj, Experimental
1421–1431. investigation of oven cured jute fiber/glass reinforced polyester composites,
[15] T. Srinivas, A.C. Reddy, Parametric optimization of warm deep drawing process International Journal of Mechanical and Materials Engineering 7 (3) (2013).
of 1100 aluminum alloy: validation through FEA, Int. J. Sci. Eng. Res. 6 (4) [19] A. Karpagaraj, N. Rajesh Kumar, N. Thiyaneshwaran, N. Siva Shanmugam, M.
(2015) 425–433. Cheepu, R. Sarala, Experimental and numerical studies on gas tungsten arc
[16] V. Mishin, I. Shishov, A. Kalinenko, I. Vysotskii, I. Zuiko, S. Malopheyev, R. welding of Ti–6Al–4V tailor-welded blank, J. Braz. Soc. Mech. Sci. Eng. 42 (10)
Kaibyshev, Numerical Simulation of the Thermo-Mechanical Behavior of 6061 (2020) 1–11.
Aluminum Alloy during Friction-Stir Welding, Journal of Manufacturing and [20] D. Srinivas, R.S. Kadadevaramath, B.L. Shankar, P.M. Nagraj, J. Bhaskaran, D.G.
Materials Processing 6 (4) (2022) 68. Mallapur, Optimization of machinability parameters of Al1100-B4C
[17] B. Sutharson, M. Rajendran, A. Karapagaraj, Optimization of natural fiber/glass composites using Taguchi method, Mater. Today:. Proc. 4 (10) (2017)
reinforced polyester hybrid composites laminate using Taguchi methodology, 11305–11313.

You might also like