IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Optimization of mechanical design problems using advanced

optimization technique
To cite this article: Jigar L. Patel et al 2018 IOP Conf. Ser.: Mater. Sci. Eng. 455 012091

View the article online for updates and enhancements.

This content was downloaded from IP address 185.204.199.82 on 20/12/2018 at 01:15

ICAAMM IOP Publishing
IOP Conf. Series: Materials Science and Engineering 455 (2018) 012091 doi:10.1088/1757-899X/455/1/012091

Optimization of mechanical design problems using advanced

optimization technique

Jigar L. Patel, Parthiv B. Rana, D. I. Lalwani

Mechanical Engineering Department, SVNIT, Surat, Gujarat-395007.

Corresponding author: [email protected]

Abstract. In design of mechanical elements, designers usually consider certain objectives that
are related with cost, time, quality and reliability of product depending on the requireme nts. In
this paper, parametric optimization of spring design problem and pressure vessel design
problem has been carried out using Amended Differential Evolution Algorithm (ADEA).
ADEA is modified DE that can handle constraint and complex problems effectively. In spring
design problem, parameters are optimized to reduce volume of compress ion and in pressure
vessel design problem, parameters are optimized to reduce total cost. The results obtained using
ADEA are compared with the results reported by other researchers. The comparison of results
shows that ADEA provides better results for both spring and pressure vessel design problems.

1. Introduction
For the competitive market, industries always look for solutions that reduces cost of a component
without compromising quality and time to manufacture a component. One of the ways to reduce the
cost of a component is improvement in design. Optimization of design problems consists of certain
objective function, viable solution and optimization technique. The viable solutions are the set of all
designs that are distinguished by all feasible values of the design parameters. Optimization methods
search for the optimum design from all available viable solutions.
Asoptimization of mechanical assembly leads to a complex objective function with many design
parameters, individual components or intermediate assemblies are optimized in practice.
Appropriateselection of design parameters is necessary to reduce cost, improve productivity and quality,
increase reliability, etc. For the complex real life problems, Evolutionary Optimization Techniques are
preferred than Conventional Optimization Techniques (COTs) (Savsani and Savsani, 2014).
In this work, ADEA is used to optimize two design problems, namely, spring design that has three
design parameters (two continuous and one integer) and eight constraints and pressure vessel design
that has four design parameters and four constraints

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
ICAAMM IOP Publishing
IOP Conf. Series: Materials Science and Engineering 455 (2018) 012091 doi:10.1088/1757-899X/455/1/012091

2. Amende d differential evolution algorithm

Unconstraint problem can be solved by using DE algorithm. Rana and Lalwani (2017) made some
modifications to improve the performance of Differential Evolution (DE) and added constrained
handling technique to solve real life complex problems. The modified DE algorithm is named as
ADEA. They made following modifications in DE (Rana and Lalwani, 2017):
1) Initialization of random population
2) Selection of scale factor (F) and crossover rate (CR) based on rank (R)
3) Selection of mutant vector
4) Condition for crossover vector.
5) Constraint handling technique
The detailed description of all modifications is available in Rana and Lalwani (2017).
3. Formulation of design problems
Two design problems viz., spring design problem and pressure vessel design problem are discussed
below with objective function and constraints.

3.1. Spring design problem

Spring design problem involves integer and continuous design variables. The spring design problem,
investigated by He et al., (2004), is taken to optimize design parameters. In this problem, material
used for spring is music wire spring steel ASTM A228 and three design parameters, such as, the wire
diameter (d), the mean coil diameter(D) and number of active coils (N) are optimized to minimize the
volume of compression spring.
The problem with objective function and constraints is given below (He et al., 2004).
2  2 12( 3+2)
Minimize,   =
4

Figure 1: Spring design (He et al., 2004)

Subject to 8
constraints:
 
  2
= −  ≤ 0,   = − ≤0
1   13 2 

3  =   − 1 ≤ 0, 4  = 2 − 
 ≤ 0
  = 3.0 − 2 ≤ 0,   = − ≤0
5 1 6  
(
− )
 7  =  + + 1.05  3 + 2 1 −  ≤ 0and8  =  − (
− ) ≤ 0
 
where, 2
4( −1) 4   

1 0.615 1  1 = , = + 1.05  +2
=
 4(
2
−4)
+ ,= , 3   3 1
2 8 3 2  
1

2
ICAAMM IOP Publishing
IOP Conf. Series: Materials Science and Engineering 455 (2018) 012091 doi:10.1088/1757-899X/455/1/012091

Table 1Design parameters and its ranges

Design parameters Range
x1 0.2-1
x2 0.6-3
x3 1.0 -70
Other parameters
The maximum load (
 ) 1000 lb.
The maximum free length
 14 in.
The minimum wire diameter   0.2 in.
The allowable maximum shear stress () 189000 .0 
The maximum outside diameter of the spring (
 ) 3 in
The preload compression force ( ) 300 
The allowable maximum deflection under preload ( ) 6 
The deflection from preload to maximum load positon ( ) 1.25 

Shear modulus () 11.5 × 106 

3.2. Pressure vessel design problem

The pressure vessel design problem (Fig. 3) was investigated by He et al., (2004). In this problem, the
material used for pressure vessel is carbon steel ASME SA 203 grade B andthe design parameters such
as, the shell thickness ( ), the thickness of the head ( )and the inner radius  are optimized
to minimize the total cost.
The problem with objective function and constraints is given below (He et al., 2004).
Minimize,  = 0.6224134 + 1.7781232 + 3.1661124 + 19.84123

Figure 2 Pressure vessel design (He et al., 2004)

Subject to constraints:
1 = 0.01933 − 1 ≤ 0, 2  = 0.009543 −  2 ≤ 0,
4
−  2
3  = 1296000 − 3 24 3 ≤ 0and 4  = 4 − 240 ≤ 0
3

3
ICAAMM IOP Publishing
IOP Conf. Series: Materials Science and Engineering 455 (2018) 012091 doi:10.1088/1757-899X/455/1/012091

Table 2 Design parameters and its ranges

Design parameters Range
1 0.0625-6.1875
2 0.0625-6.1875
3 10-200
4 10-200

4. Results and discussion

The above two problems are solved using ADEA with following parameters:
Population size () = 25, number of iteration () = 1000, number of runs = 5, control parameters i.e.,
 (0.5 − 0.8) and  (0.85 − 0.95). The results are reported by writing the program in MATLAB®.

4.1. Spring design problem

The spring design problem was optimized by many researchers, such as He et al. (2004), Sandgren
(1990), Deb and Goyal (1997) and Lampinen and zelinka(1999) using different optimization
techniques. Comparison of results obtained using ADEA with other researchers’ work is shown in
Table 3.
Table 3 Comparison of results obtained using ADEA with other researchers’ work for spring design
Design PSO NLPA GA DE
ADEA
Variables He et al. (2004) Sandgren (1990) Deb (1997) Lampinen (1999)
x1 0.29162 0.283 0.283 0.283 0.283
x2 1.38421 1.22304101 1.180701 1.226 1.22304101
x3 7 9 10 9 9
Constraints
g1 -1.2584 -1008.8114 -54309 -713.51 -1008.8114
g2 -9.4584 -8.9456 -8.8187 -8.933 -8.9456
g3 -0.09162 -0.083 -0.08298 -0.083 -0.083
g4 -1.6158 -1.777 -1.8193 -1.491 -1.777
g5 -1.7466 -1.3217 -1.1723 -1.337 -1.3217
g6 -5.4643 -5.4643 -5.4643 -5.461 -5.4643
g7 0 0 0 0 0
g8 -4.2997E-5 0 0 -0.009 0
Objecti ve Function
fx 2.6141 2.65856 2.7995 2.665 2.65856
* NLPA-Nonlinear programming algorithm, * PSO- Particle Swarm Optimization, *GA-Genetic Algorithm,
* DE-Differential Evolution
The comparison of results shows that ADEA gives the f(x) value as 2.6141 that is superiorto results
reported by other researchers. ADEA provides 1.67% enhancement in the results compared with best
value reported by Heet. al. (2004).

4.2. Pressure vessel design problem

The pressure vessel design problem was optimized by many researchers, such as Rao (2011), He et al.,
(2004), CoelloCoello (2001) and Deb (1997) using different optimization techniques. Comparison of
results obtained using ADEA with other researchers’ work is shown in Table 4.
Table 4Comparison of results obtained using ADEA with other researchers’ work for pressure vessel
design
Design TLBO PSO GA GA
ADEA
Variables Rao (2011) He et al. (2004) Coello (2001) Deb (1997)
x1 0.778168641 NA 0.8125 0.8125 0.9345
x2 0.384649163 NA 0.4375 0.4375 0.5
x3 40.31961872 NA 42.0984456 40.097398 48.329
x4 200 NA 176.6365 176.654047 112.679
Constraints
g1 0 NA 0 -0.00002 -0.00475
g2 -5.55E-17 NA -0.03588083 -0.035891 -0.038941
g3 0 NA 0 -27.886075 -3652.87684
g4 -40 NA -63.36340416 -63.345953 -127.321

4
ICAAMM IOP Publishing
IOP Conf. Series: Materials Science and Engineering 455 (2018) 012091 doi:10.1088/1757-899X/455/1/012091

Objecti ve Function
fx 5885.872354 6059.7143 6059.7143 6059.946341 6410.3811
*NA-Not Available, *TLBO-Teaching Learning Based Optimization, *GA-Genetic Algorithm

The comparison of results shows that ADEA gives the objective function f(x) value as 5885.872354
that is superiorto the results reported by other researchers. ADEA provides 2.87 %enhancement in the
results compared with the best value reported by Rao (2011).

5. Conclusions
Optimization of spring design and pressure vessel design problems have been carried out using ADEA
and the results obtained using ADEA are compared with the results reported by other researchers. The
comparison of results shows that ADEA provides better results than the results reported by other
researchers. In spring design problem, ADEA gives improvement of 1.67 % over PSO, in pressure
vessel design problem, ADEA gives improvement of 2.87 % over TLBO.

References
[1] Coello Coello, C. A. and Mezura Montes, E., 2001. Use of dominance-based tournament
selection to handle constraints in genetic algorithms. In: Intelligent Engineering Systems
through Artificial Neural Networks (ANNIE’2001), Vol. 11, ASME Press, St. Louis,
Missouri, pp. 177–182.
[2] Deb, K., 1997. Geneas: a robust optimal design technique for mechanical component design. In:
Evolutionary Algorithms in Engineering Applications, Ed. D. Dasgupta and Z. Michalewicz,
Springer-Verlag, pp. 497–514.
[3] He et al., 2004. An improved particle swarm optimizer for mechanical design optimization
problems. Engineering Optimization 2004; 36(5):585–605.
[4] Lampinen, J. and Zelinka, I., 1999. Mixed integer-discrete-continuous optimization by
differential evolution. In: Proceedings of the 5th International Conference on Soft
Computing, pp. 71–76.
[5] Rana Parthiv B. and Lalwani D. I., 2017. Optimization of turning process using Amanded
Differential Evolution Algorithm. Engineering Science and Technology, an International
Journal 20: (2017) 1285-1301.
[6] Rao et al., 2011. Teaching learning based optimization: A novel method for constrained
mechanical design optimization problems. Computer-Aided Design 43:303-315 (2011).
[7] Sandgren, E., 1990. Nonlinear integer and discrete programming in mechanical
designoptimization. Journal of Mechanical Design, 112, 223–229.
[8] Savsani Poonam, R. Jhalab L., Savsani Vimal, 2014. Effect of hybridizing Biogeography- based
Optimization (BBO) technique with Artificial Immune Algorithm (AIA) and Ant Colony
Optimization (ACO). Applied Soft Computing: 21. 542–553.