15 views

Uploaded by Nhan Tran

The triangular tubes with multi-cell were first studied on the aspects of theoretical prediction and
crashworthiness optimization design under the impact loading. The tubes' profiles were divided into
2-, 3-, T-shapes, 4-, and 6-panel angle elements. The Simplified Super Folding Element theory was
utilized to estimate the energy dissipation of angle elements. Based on the estimation, theoretical
expressions of the mean crushing force were developed for three types of tubes under dynamic loading.
When taking the inertia effects into account, the dynamic enhancement coefficient was also considered.
In the process of multiobjective crashworthiness optimization, Deb and Gupta method was utilized to
find out the knee points from the Pareto solutions space. Finally, the theoretical prediction showed an
excellent coincidence with the numerical optimal results, and also validated the efficiency of the
crashworthiness optimization design method based on surrogate models

- Lecture15_MPC2
- St-Poz-Garst-3-09
- Master of technology
- Numerical Nonlinear Global Optimization in Mathematica
- Trans
- KEC
- 57c82ea6221906e563c5cf8acba19f84
- The Methods of O R on Russian Financial Markets
- Algorithms Process Optimization
- DBA7301
- Civil Engineering Syllabus Revised Upto 8th Semester 2007
- Lecture Summary
- buckling_analysis_laminates.pdf
- 1-s2.0-S0020746209002078-main(2).pdf
- Cost Optimization of Reinforced Concrete Chimney-libre
- Strut
- Buckling
- Optimization
- Steel Design Manual - Sapstl-editado
- vol7no1_7

You are on page 1of 13

Thin-Walled Structures

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

of multi-cell triangular tubes

TrongNhan Tran a,c, Shujuan Hou a,b, Xu Han a,b,n, Wei Tan a,b, NhatTan Nguyen a,d

a

State Key Laboratory of Advanced Design and Manufacturing for Vehicle Body, Hunan University, Changsha 410082, Hunan, PR China

College of Mechanical and Vehicle Engineering, Hunan University, Changsha 410082, Hunan, PR China

c

Faculty of Mechanical Engineering, Industrial University of Ho Chi Minh City, Go Vap District, HCM City, Vietnam

d

Center for Mechanical Engineering, Hanoi University of Industry, Tu Liem District, Ha Noi, Vietnam

b

art ic l e i nf o

a b s t r a c t

Article history:

Received 2 January 2014

Received in revised form

13 March 2014

Accepted 26 March 2014

Available online 17 May 2014

The triangular tubes with multi-cell were rst studied on the aspects of theoretical prediction and

crashworthiness optimization design under the impact loading. The tubes' proles were divided into

2-, 3-, T-shapes, 4-, and 6-panel angle elements. The Simplied Super Folding Element theory was

utilized to estimate the energy dissipation of angle elements. Based on the estimation, theoretical

expressions of the mean crushing force were developed for three types of tubes under dynamic loading.

When taking the inertia effects into account, the dynamic enhancement coefcient was also considered.

In the process of multiobjective crashworthiness optimization, Deb and Gupta method was utilized to

nd out the knee points from the Pareto solutions space. Finally, the theoretical prediction showed an

excellent coincidence with the numerical optimal results, and also validated the efciency of the

crashworthiness optimization design method based on surrogate models.

Crown Copyright & 2014 Published by Elsevier Ltd. All rights reserved.

Keywords:

Crashworthiness

Multiobjective optimization

Triangular tube

Multi-cell

Energy absorption

Impact loading

1. Introduction

Thin-walled extrusions have been extensively applied in vehicle

crashworthiness components to absorb impact energy in the past

three decades. The tests and the theoretical expressions of square

and circular tubes under axial quasi-static and dynamic loading cases

were rst described by Wierzbicki and Abramowicz [1] and Abramowicz and Jones [2]. From then on, DiPaolo et al. [3,4], Guillow et al.

[5], Ullah [6] Zhang and Zhang [7], Alavi Nia and Parsapour [8] also

did many researches on these aspects. Beside square and circular

tubes, several other proles were also studied on their quasi-static or

dynamic responses, such as triangular tubes [912], hexagonal tubes

[13], etc. The structural collapse modes of triangular and square tubes

are different from those of circular tubes. Nevertheless, the crushing

curves of forcedisplacement of triangular and square tubes are

similar to those of circular tubes. The crushing curves of force

displacement of all the proles show that the crushing force rst

reaches an initial peak, then drops down and then uctuates around

a value of the mean crushing force. The extensional deformation has

n

Corresponding author at: College of Mechanical and Vehicle Engineering,

Hunan University, Changsha, Hunan 410082, PR China.

E-mail addresses: shujuanhou@hnu.edu.cn (S. Hou),

hanxu@hnu.edu.cn (X. Han).

http://dx.doi.org/10.1016/j.tws.2014.03.019

0263-8231/Crown Copyright & 2014 Published by Elsevier Ltd. All rights reserved.

more dominant effect on the crushing responses while the quasiinextensional mode occurs normally [14].

According to studies by Wierzbicki and Abramowicz [1], the

number of angle elements on cross-section of tube decided,

to a certain extent, the effectiveness of energy absorption. As

a matter of fact, it is necessary to design thin-walled multi-cell

tubes for weight-efcient energy absorption. Chen and Wierzbicki

[15] examined the axial crushing resistance of single-cell, doublecell and triple-cell hollow tubes, and the respective foam-lled

tubes under the quasi-static axial loading. The Simplied Super

Folding Element (SSFE) theory was applied to simplify SFE theory,

and three extensional triangular elements and three stationary

hinge lines were comprised instead of the kinematically admissible

model of SFE [1]. The average folding wavelength and the theoretical expression of the mean crushing force were deduced by

dividing the cross-sectional tube into distinct panel section and

angle element, assuming that the roles of each panel and of angle

element were at the same level. The work of Chen and Wierzbicki

[15] showed that the multi-cell tube could increase the specic

energy absorption SEA by approximately 15%, compared to the

respective hollow tube. Kim [16] used Chen and Wierzbicki's model

[15] to study multi-cell tubes with four square elements at the

corner. The SEA of new multi-cell tube was reported to increase by

190%, compared to conventional square tube. Zhang et al. [17] also

applied SSFE theory to derive a theoretical expression of the mean

184

loading. In Zhang's work, the cross-section of tube was divided into

three basic angle elements, and the study also came to the

contribution that plastic energy of each element type was dissipated through membrane action. It was assumed from the

theoretical expression that the average wavelength for the dissimilar folds developed at corners. Thereafter, the SSFE theory was also

adopted by Zhang et al. [18] to predict the mean crushing force of 3panel angle element. At the same time, the SFE theory was

extended by Naja and Rais-Rohani [19] to explore the crushing

characteristics of multi-cell tubes with two different types of threepanel elements. A closed form expression of mean crushing force

was also put forward by Naja and Rais-Rohani [19].

Dynamic progressive buckling of thin-walled multi-cell tubes

under axial impact loadings was studied by Jensen et al. [20] and

Karagiozova and Jone [21]. Then, the structural dynamic progressive buckling under the axial loading was summarized by Karagiozova and Alves [22] from a phenomenological point of view.

Consequently, the desirable energy-dissipating mechanism was

a stable and progressive folding deformation pattern for the

structural deformation. On the other hand, the global bending on

a structure was an undesirable energy-dissipating mechanism

mode. At the beginning, multi-cell tubes were mostly employed

from the aspects of theoretical researches, such as by Kim [16] and

Naja and Rais-Rohani [19], Nowadays, either FE solutions [23] or

surrogate models [13,2426] developed appeared in the search

eld of multi-cell tubes under the impact loading. However, there

is seldom a combination study of theory, numeric and optimal

method for thin-walled multi-cell tubes.

Above all, the axial crushing of tube types I, II, III was studied

on both theoretical prediction and numerical optimization design

in this paper. Based on the SSFE theory, theoretical expressions of

the mean crushing force for the three types were derived. All the

proles studies in this paper were divided into 2-, 3-, T-shape,

4 and 6-panel angle elements. In order to obtain the optimal

proles under the crashworthiness criterion, dynamic nite

element analysis code ANSYS/LS-DYNA was executed to simulate

tubes and to obtain the numerical results at the design sampling

points. The multiobjective optimization design was utilized to

obtain the optimal congurations. Finally, the theoretical expressions are employed to validate the numerical optimal solutions.

dissipation over the collapse of a fold, the triangular multi-cell

thin-walled tubes were divided into several basic elements: the 2-,

3-, 4- and 6-panel angle element as shown in Fig. 1.

Based on the principle of global equilibrium for shells, the

internal and external energy dissipations are of equal rate E_ ext

E_ int . The external energy work for a complete single fold is equal to

the sum of dissipated bending and membrane energy. That is

1

P m 2H Eb Em

force, the length of the fold, the bending energy and the membrane

energy, and is the effective crushing distance coefcient. The panel

of folding element after deformation is not completely attened as

shown in Fig. 2. Hence, the available crushing displacement is

smaller than 2H. In this study, the value of was taken as 0.7 since it

was found between 0.7 and 0.8 [1].

2.1.1. The bending energy of tube

The SSFE theory was applied to calculate the dissipated energy

in bending of each panel. Only three extensional and compressional triangular elements and three stationary hinge lines were

used in SSFE theory, which was different from SFE theory. In SFE

theory, a model was built with trapezoidal, toroidal, conical and

cylindrical surfaces of moving hinge lines.

In the work of Chen and Wierzbicki [15], the energy dissipation

for bending of each panel Eb was estimated by summing up the

energy dissipation at three stationary hinge lines. Then

4

Efb M 0 i b

i1

panel, b is the sectional breadth and denotes the rotation angle

at the stationary hinge lines.

In this case, the panel was supposed to completely atten after

the deformation of the wavelength 2H. Consequently, the four

rotation angles at three stationary hinge lines were /2 one by

one (as shown in Fig. 2). By applying Eq. (2), the bending energy at

2. Theoretics

2.1. Theoretical prediction of multi-cell triangular tube

The SSFE theory was applied to solve the axial collapse of

triangular multi-cell thin-walled tubes. In the SSFE theory, the wall

thickness was assumed to be constant and the variation of

Fig. 1. Cross-sectional geometry of triangular multi-cell tube and typical angle element. (a) Tube type I (b) Tube type II and (c) Tube type III.

Efb

2 M 0 b

Since the role of structural each panel is similar and the multicell tube is constituted by m panels (as shown in Fig. 1), the energy

dissipation for bending of multi-cell tube is inferred as

2 M 0 mb 2 M 0 B

Etube

b

for the symmetric mode is determined as

2.1.2.1. The membrane energy of 2-panel, 3-panel and T-shape angle

element. The basic folding element (BFE) was created by using the

triangular elements and the stationary hinge lines (Fig. 3) so as to

calculate the membrane energy of right-corner through asymmetric

or symmetric deformation mode in the SSFE theory. Two possible

collapse modes of the asymmetric and symmetric deformation were

supposed in the establishment of BFE. The symmetric and asymmetric

modes came from the extensional and quasi-inextensional modes

respectively. As for the asymmetric mode, the three triangular

elements were developed for each web after the deformation.

However, the symmetric mode had two triangular elements for each

panel after the deformation. Thus for the asymmetric mode during

one wavelength crushing, the energy dissipation in membrane Em of

each panel was evaluated by integrating the area of triangular

elements (shaded areas in Fig. 3a). Then

Z

1

H2

5

Easym

s0 tds s0 tH 2 2M 0

mf

2

t

s

Each panel was assumed to have the similar contribution. For the

asymmetric mode, the dissipated membrane energy of the right

corner element is double of that in one single panel. Then

H2

t

dissipated energy in membrane of each panel was estimated by

integrating the area of triangular elements (shaded areas in Fig. 3b).

Then

Z

H2

Esym

7

mf

t

s

f

Easym

m_r c 2E m 4M 0

185

sym

Esym

m_r c 2E m_f 8M 0

H2

t

and hexagonal, the cross-section proles were formed by 2-panel

angle elements. The crushing forces of hollow tubes were strengthened in turn from triangular tube to hexagonal tube. When the

central angle varies from 301 to 1201, the crushing force of 2-panel

angle elements has an appropriate increase [27]. Eq. (1) shows that

the mean crushing force is a function of bending and membrane

energy. Because the role of each structural panel is assumed to be

similar, the bending energy of angle elements with the same

numbers of panel will not change. Then, the mean crushing force

will vary with respect to the membrane energy. Thus, the membrane

energy of right corner element is bigger than that of 2-panel angle

element with central angle of 601. From Fig. 4a, the membrane

energy of a 2-panel angle, during one wavelength crushing, is

calculated as

panel

Eind:2

Easym

m

m_r c cos 4M 0

H2

cos

t

was calculated by Zhang and Zhang [18]. Then, the membrane

Fig. 3. Basic folding element: (a) asymmetric mode [15] and (b) symmetric (extensional) mode.

Fig. 4. (a) Relationship between the membrane energy of right corner and of 2-panel angle element and (b) 3-panel angle element.

186

is

E3m panel 4M 0

H2

1 2 tan =2

t

10

one additional panel as shown in Fig. 5. For this reason, the

membrane energy of T-shape element was calculated by a sum

between membrane energy of right-corner element for the symmetric mode and membrane energy of one additional panel.

Because the contribution of each panel is assumed to be similar,

the energy dissipated in membrane of T-shape element, during

one wavelength crushing, is more than triple each panel's membrane energy. Then

ETm shape 3Esym

12M 0

mf

H2

t

summing up the membrane energy of 2-panel angle element for

symmetric mode and the membrane energy of two additional

panels.

It is not easy to give a precise calculation for membrane energy

of two additional panels. In this case, it is too complicated for the

SFE theory to be applied here. So, a simplied deformation model

of two additional panels was proposed and the SSFE theory was

utilized to solve this problem. Fig. 7b shows the membrane

triangular elements developed at the corner line. Accordingly,

the membrane energy Em of 4-panel angle element, during one

wavelength crushing, is evaluated by integrating the triangular

areas as

H2

1

1

12

E4m panel 8M 0

t

cos

11

angle element is a symmetric structure and is constituted by one

2-panel angle element and two additional panels. Therefore, the

energy absorption in membrane of a 4-panel angle element was

calculated by a sum of membrane energy of 2-panel angle element

and membrane energy of two additional panels. Due to similar

deformation mode shown in Fig. 6, it is assumed that 2-panel

angle element and the corresponding 2-panel angle element

in the 4-panel angle element are of equal crushing resistance.

Simultaneously, the deformation mode of 2-panel angle element is

a symmetric mode in this case. For this reason, the dissipated

a symmetric structure and being composed of six panels, the

dissipated energy in membrane of a 6-panel angle element was

calculated by the sum of membrane energy absorbed by all

6 panels. Because of the symmetric structure and the similar

contribution of each angle element, the 6-panel angle element was

formed by two 4-panel angle elements (Fig. 8a). As a result, the

membrane energy of 6-panel angle element, during one

wavelength crushing, was estimated by the sum of membrane

energy absorbed by two 4-panel angle elements. This is

H2

1

E6m panel 2E4panel

1

16M

13

0

m

t

cos

Fig. 5. (a) Collapse mode of T-shape element and (b) extensional elements.

Fig. 6. (a) 2-panel angle element and (b) 4-panel angle element.

187

Fig. 7. (a) Collapse mode of 4-panel angle element and (b) extensional elements.

Fig. 8. (a) Collapse mode of 6-panel angle element and (b) extensional elements.

To build a quantitative prediction of crushing resistance of

triangular multi-cell tubes and to nd out how angle element

affects the structure of tube, the theoretical solution of the mean

crushing force of multi-cell tube was introduced. Fig. 1a shows the

three independent 2-panel angle elements and three 4-panel

angle elements form prole of tube type I. Substituting the terms

of Eqs. (4), (9) and (12) into Eq. (1), the general theoretical

equation of the mean crushing force of tube type I was obtained.

This is

Rigid wall

Lumped Mass

v = 10 m/s

L0 = 250 mm

P m I 2H

panel

Etube

3Eind:2

E4m panel

b

m

2 M 0 B 2M 0

H2

1

6 cos 12 1

:

t

cos

14

Substituting Eq. (16) back into Eq. (15), the mean crushing force in

quasi-static loading of tube type I is obtained as

p

Pm I B H

1

B H

6 cos 12 1

F ;

H t

H t

M0

cos

Pm I

15

B F ;

H

)H

s

Bt

F ;

17

where

The half-wavelength was obtained under the stationary condition of the mean crushing force P m =H 0, then

0

F ;

M0 B M0 H

H

t

2

16

F ; 6 cos 12 1

1

:

cos

constituted by three 3-panel angle elements, three T-shape angle

188

of Eqs. (4), (10), (11) and (13) into Eq. (1), the general theoretical

expression of the mean crushing force of tube type II is obtained

as

P m II 2H Etube

3E3m panel ETm shape E6m panel

b

2 M 0 B 2M 0

H2

8

32 12 tan =2

t

cos

18

P m II B H

8

B H

32 12 tan =2

G;

H t

H t

M0

cos

19

the half-wavelength can be obtained as P m =H 0. Then

s

B G;

Bt

)H

0 2

20

t

G;

H

Substituting Eq. (20) back into Eq. (19), the equation of the mean

crushing force for tube type II under the quasi-static loading is

obtained as

p

M0 B M0 H

G;

P m II

21

2

H

t

where

G; 32 12 tan =2

8

:

cos

The structure of the tube type III (Fig. 1c) was formed by three

independent 2-panel angle elements, six 4-panel angle elements,

189

and one 6-panel angle element. Substituting the terms of Eqs. (4),

(9), (12) and (13) into Eq. (1), the general theoretical solution to

the mean crushing force of tube type III is obtained as

P m III 2H Etube

3E2m panel 6E4m panel E6m panel

b

H2

32

32 6 cos

2 M 0 B 2M 0

t

cos

22

P m III B H

32

B H

32 6 cos

Q ;

H t

H t

M0

cos

23

the half-wavelength is expressed as P m =H 0. Then

s

B Q ;

Bt

)H

24

0 2

t

Q ;

H

Substituting Eq. (24) into Eq. (23), the mean crushing force of tube

type III under quasi-static loading is

p

Q ;

M0 B M0 H

P m III

25

H

t

2

where

Q ; 32 6 cos

32

:

cos

Among all the indicators of crashworthiness optimization

design, the vital analytical objective was the energy-absorption.

Hence, in order to estimate the energy absorption of structural

unit mass m, specic energy absorption (SEA) was formulated as

SEA

EA

m

26

absorption. In Eq. (26), the total strain energy during crushing is

Table 1

Design matrix of three types of tube for crashworthiness.

n

Fig. 11. The crushing foredisplacement curve of (a) tube I, (b) tube II and

(c) tube III.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

1

1.3

1.6

1.9

2.2

1

1.3

1.6

1.9

2.2

1

1.3

1.6

1.9

2.2

1

1.3

1.6

1.9

2.2

1

1.3

1.6

1.9

2.2

80

80

80

80

80

85

85

85

85

85

90

90

90

90

90

95

95

95

95

95

100

100

100

100

100

Tube type II

SEA

PCF (kN) SEA

(kJ/kg)

(kJ/kg)

(kJ/kg)

PCF (kN)

18.262 39.362

20.226 52.274

22.774

65.287

23.854 77.82

24.732 88.936

17.635 42.151

19.491 55.941

21.079 69.78

22.231 82.904

23.891 94.376

16.689 48.183

18.63

63.406

19.883 80.385

21.425 97.058

22.916 111.899

16.092 50.984

17.793 67.207

19.279 85.237

20.865 102.754

22.276 117.81

15.569 50.155

17.083 66.443

19.26

83.029

20.628 98.954

21.053 113.035

49.572

65.965

82.467

98.125

111.611

53.133

70.384

88.192

104.633

118.899

60.979

69.095

99.84

120.552

138.568

65.978

83.194

105.95

128.907

148.295

63.148

83.857

104.85

124.628

142.093

52.592

69.599

85.801

101.135

114.541

56.038

74.202

92.462

109.979

125.572

64.544

85.478

107.641

128.512

146.48

68.196

90.665

114.112

135.856

153.825

66.649

87.852

110.016

130.951

157.552

17.356

19.876

21.86

22.003

23.107

16.842

19.272

20.924

21.602

22.693

16.559

18.9

20.273

20.679

21.524

15.253

16.954

18.229

19.146

19.71

14.398

16.139

17.748

18.55

18.715

23.339

25.227

28.199

29.092

29.644

22.753

24.273

26.388

27.564

28.107

21.338

24.05

25.536

26.272

26.42

20.932

22.371

24.579

25.306

25.948

19.337

20.011

22.979

23.92

24.152

190

Fig. 12. The response surface of (a) peak crushing force and (b) SEA.

Fig. 13. (a) SEA vs structural weight and (b) Pm vs structural weight.

estimated as

Z d

Pxdx

EA

0

27

initial peak crushing force (PCF) of multi-cell thin-walled tube was

used for estimating the impact characteristics. Another crashworthiness indicator is the mean crushing force Pm which is

computed by

Z

E

1 d

Pxdx

28

Pm A

d d 0

where d is the crushing displacement at a specic time.

2.2.1. Response Surface Method (RSM)

A typical surrogate modelling technique was considered appropriate in the multivariate optimization process involving material,

geometrical nonlinearities and contact-impact loading nonlinearities. The primary concept of RSM was applied to the construction

of regression functions for crashworthiness indicators by using the

function values at the design sampling points. The mathematical

Fig. 14. Pareto spaces for multi-objective optimization: (a) tube type I, (b) tube

type II and (c) tube type III.

m

~ x

yx yx

i i

i1

29

imation and the numerical solution denoting y(x). m represents

the total number of basic functions i(x), and i is the unknown

linear polynomial basis function is

1; x1 ; x2 ; ; xn

30

1; x1 ; x2 ; :::; xn ; x21 ; x1 x2 ; :::; x1 xn ; :::; x2n ; x31 ; x21 x2 ; :::; x21 xn ; x1 x22 ; :::;

x1 x2n ; :::; x3n ; x41 ; x31 x2 ; :::; x31 xn ; x21 x22 ; :::; x21 x2n ; :::; x1 x32 ; :::; x1 x3n ; :::; x4n

191

mathematically given as

Maximize x; xL ; xR L R

33

f 2 xR =f 1 xR f 1 x are the left and right bend-angles of x.

31

a better choice for the regression analysis [13,2426]. The quartic

response surface models were consequently adopted in this study.

2.2.2. Multi-objective optimization

With two objectives of SEA and PCF, the multiobjective optimization problem for minimizing PCF and maximizing SEA was

dened by the linear weighted average methods (LWAM) [24].

Then, the mathematical denition for the crashworthiness optimization in terms of the LWAM is given as

8

SEAn

>

Minimize Ft; a wPCFt;a

>

PCFn 1 wSEAt;a

>

>

<

s:t w A 0; 1

32

>

1

r t r 2:2 mm

>

>

>

:

80 r a r 100 mm

where SEAn and PCFn are the given normalizing values for each

cross-sectional prole.

2.2.3. Knee point

In some certain cases, the designer must choose the most

preferred solution (termed as knee point) from optimal solutions

to meet their requirement. Several methods were proposed to

determine a knee point from Pareto set such as Turevsky and

Suresh [28] and Sun et al. [29]. However, if there is a great deviation

among the orders of magnitude of different objectives, these

methods [29] seem to be less effective. A modied multi-objective

evolutionary algorithm suggested by Branke et al. [30] was utilized to

seek the knee regions. Deb and Gupta [31] have recently suggested a

solution to nd a knee point with maximum bend-angle, which is

3.1. Numeric simulation

In this section, the FE model was carried out by ANSYS/LS-DYNA

to simulate the triangular multi-cell thin-walled tubes subjected to

axial dynamic loading with 4, 6 and 9 cells (as shown in Fig. 1). The

side-length a of the cross-sections and the thickness t were chosen

to be design variables, and the design interval is given in Eq. (32).

The total length L0 of all the tubes is 250 mm.

In this study, the thin-walled tubes were modelled with the

BelytschkoTsay four-node shell element with the optimal mesh

density of 2.5 2.5 mm. The material AA6060 T4 was modelled

with material model #24 (Mat_Piecewise_Linear_Plasticity) with

mechanical properties: Young's modulus E 68,200 MPa, initial

yield stress sy 80 MPa, ultimate stress su 173 MPa, Poisson's

ration 0.3, and power law exponent n 0.23 [32]. Since the

aluminum was insensitive to the strain rate effect, this effect was

neglected in the nite element analysis. An automatic node to

surface contact between thin-walled tube and rigid wall was

dened to simulate the real contact. Alternatively, an automatic

single surface contact algorithm was utilized for the self-contact

among the shell elements to avoid interpenetration of folding

generated during the axial collapse. In the contact denition,

a friction coefcient of 0.3 among all surfaces was employed. To

generate enough kinetic energy, one end of tube was attached

with a lumped mass of 500 kg whereas another end impacted onto

a rigid wall with an initial velocity of 10 m/s. The schematic of the

computational model is shown in Fig. 9.

All of the tubes were axial symmetric structures. Despite the

same length, same side-length and same thickness, the three tubes

Table 2

Difference of numeric result and theoretical prediction for three tubes.

n

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Tube type I

Tube type II

Num. Pm (kN)

Theo. Pm (kN)

Diff. (%)

Num. Pm (kN)

Theo. Pm (kN)

Diff. (%)

Num. Pm (kN)

Theo. Pm (kN)

Diff. (%)

22.23

34.107

46.497

61.379

77.295

23.173

34.142

47.275

62.465

79.196

23.787

35.095

48.408

64.341

81.044

24.151

35.645

49.676

66.14

82.19

24.819

36.812

51.437

67.423

83.731

23.016

34.028

46.344

59.817

74.337

23.736

35.098

47.809

61.717

76.710

24.435

36.137

49.230

63.561

79.013

25.115

37.146

50.611

65.352

81.250

25.776

38.129

51.956

67.096

83.427

3.54

0.23

0.33

2.54

3.83

2.43

2.80

1.13

1.20

3.14

2.73

2.97

1.70

1.21

2.51

3.99

4.21

1.88

1.19

1.14

3.86

3.58

1.01

0.49

0.36

25.894

41.1

55.331

72.1

91.64

26.75

42.39

58.3

74.137

94.125

27.482

43.184

59.954

76.129

95.039

28.264

43.859

60.278

77.219

95.897

29.109

44.984

60.994

79.107

96.597

27.148

40.143

54.680

70.589

87.738

27.996

41.404

56.406

72.827

90.533

28.820

42.627

58.080

74.998

93.244

29.620

43.816

59.707

77.108

95.879

30.400

44.974

61.291

79.162

98.443

4.84

2.33

1.18

2.10

4.26

4.66

2.33

3.25

1.77

3.82

4.87

1.29

3.13

1.49

1.89

4.80

0.10

0.95

0.14

0.02

4.44

0.02

0.49

0.07

1.91

39.973

60.35

85.345

111.642

139.519

41.129

61.05

87.9

113.899

144.17

42.716

62.715

89.541

117.161

148.092

43.862

63.768

91.482

120.81

151.03

45.268

65.507

93.192

122.495

153.514

41.221

60.954

83.028

107.183

133.223

42.510

62.868

85.648

110.581

137.467

43.760

64.726

88.190

113.878

141.583

44.976

66.531

90.660

117.082

145.584

46.160

68.290

93.066

120.201

149.477

3.12

1.00

2.72

3.99

4.51

3.36

2.98

2.56

2.91

4.65

2.44

3.21

1.51

2.80

4.39

2.54

4.33

0.90

3.09

3.61

1.97

4.25

0.14

1.87

2.63

192

were different in weight. Tube I is the lightest one while tube III is

the heaviest. The axial crushing of multi-cell tubes was presented

with a displacement equal to about 70% of the initial length. Fig. 10

shows the deformation process of three tubes at different times.

Sometimes the exact value of the effective crushing distance on

the crushing forcedisplacement curve was not unique. The

corresponding crushing forcedisplacement curves of three tubes

are also shown in Fig. 11. After reaching the initial peak and before

rising steeply whenever the deformation capacity is exhausted at

the effective crushing distance, the crushing force fell sharply and

then uctuated periodically and around the values of the mean

crushing force in correspondence with the formation, and nally

completed the collapse of folds one by one.

For obtaining the response functions of SEA and PCF, a series of

25 design sampling points (based on a and t) were selected in the

design space to provide sampling design values for FEA and

regression analysis of three types of tubes (Table 1) so as to obtain

the response surface of the SEA and PCF. Fig. 12 shows that the

SEA's and PCF's RS of tube types I, II and III cases behave

monotonically over the design domain. In addition, the curves in

Fig. 13 illustrate the variation of SEA and Pm with changes in

weight. Meanwhile, energy absorptions of the tube types I and III

were better than those of tube type II.

The Pareto sets for these three cross-sectional proles were

obtained by changing the weight coefcient w in Eq. (32), and the

Pareto frontiers are plotted in Fig. 13. In fact, any point on the

Pareto frontier can be an optimum. As a result, some methods

were proposed to determine the best solution (knee point) which

has a large trade-off value in comparison with other Paretooptimal points. In this case, methods of [31,32] were utilized to

determine the knee region and the knee point, respectively. The

results of expression (33) showed that Pareto solutions (Knee

points) for tube types I, II and III were 0.7924, 0.7818 and 0.7773,

respectively. The relative errors (REs) of FE simulation value and

RS approximate value are summarized in Table 3. Therefore the FE

simulation value and RS approximate value at the Knee points

were exactly close to each other. According to the relationship

among the weighted average method and those of [31,32], these

optimal results are plotted in Fig. 14.

The theoretical expressions of the mean crushing force of tube

types I, II and III are proposed in Section 2.1.3. Nevertheless, these

expressions were applied in axial quasi-static loading case in

which the effect of dynamic crushing was not considered. In the

dynamic loading case, dynamic amplication effects, consisting of

inertia and strain rate effects, were considered in the theoretical

equations above. In reality, the aluminum alloy is insensitive to the

strain rate effect that can be neglected. A dynamic enhancing

coefcient was suggested to take the inertia effect into account.

It is not easy to nd the accurate value for dynamic enhancing

coefcient, and this coefcient is even a variable for different

geometric parameters as described by Langseth and Hopperstad

[33] and Hanssen et al. [34]. According to these studies, the

coefcient was proposed in the range of 1.31.6 for AA6060 T4

extruded tubes. These coefcients () were simply set to 1.41,

1.3 and 1.45 for tube types I, II and III, respectively. Thus, the

theoretical solution to tube type I is applied as follows:

0:5

P dym:

s0 t 1:5 B0:5

m I I P m I I

p

F ;

2

where

F ; 6 cos 12 1

1

cos

0:5

s0 t 1:5 B0:5

P dym:

m II II P m II II

p

G;

2

where

Fig. 15. Comparison between numerical prediction and theoretical prediction:

(a) tube type I, (b) tube type II and (c) tube type III.

G; 32 12 tan =2

34

8

:

cos

35

hardening, which is calculated [35] as

p

Q ;

0:5

1:5 0:5

s

t

B

P dym:

III

m

III

III

0

m III

2

36

s0

where

Q ; 32 6 cos

32

:

cos

Table 3

Optimal results by using the method of Deb and Gupta (Knee point).

Type of

cross-section

Terms

Optimal design

variables (mm)

SEA

(kJ/kN)

PCF

(kN)

Type I

Approximate value

FE analysis value

RE

Approximate value

FE analysis value

RE

Approximate value

FE analysis value

RE

t 1.23, a 80

19.897

19.728

0.856

25.282

25.717

1.691

25.100

24.800

1.210

49.315

49.088

0.462

63.436

63.056

0.603

63.967

64.267

0.467

Type II

Type III

t 1.25, a 80

t 1.21, a 80

193

sy su

1n

0:106 GPa

37

strength of the material, respectively, and n is the strain hardening

exponent.

Hence, the mean crushing forces were calculated by expressions (34)(36) and these values were used to compare with the

value of mean crushing force of tubes at a displacement of 60% in

each different case. Remarkably, this mean crushing force is

dened as an equivalent constant force with a corresponding

amount of displacement. The deviations among theoretical equations above and numerical predictions for all cases of three types

of tubes are listed in Table 2. For tube types I and II, the differences

of Eqs. (34) and (35) and numeric results were, respectively,

ranging from 3.83% to 4.21% and from 4.26% to 4.84%. With

regard to tube type III, the dissimilarities between Eq. (36) and

numeric results varied in the range of 4.65% to 4.33%. Obviously,

these differences belong to available range (Table 2). Continuously,

Fig. 15 reveals the very close agreement between the theoretical

solutions and the numerical predictions for all cases.

Fig. 16. (a) Deformation result and (b) crushing forcedisplacement curve of tube I.

Fig. 17. (a) Deformation result and (b) crushing forcedisplacement curve of tube II.

194

Fig. 18. (a) Deformation result and (b) crushing forcedisplacement curve of tube III.

Gupta (Table 3), the optimal triangular sections of tube types I, II

and III were considered in this analysis. The deformation results

and crushing forcedisplacement curves of three tubes are presented in Figs. 1618. For the optimal tube type I with 4 cells in

Table 3, the side-length and the wall thickness were of 80 mm and

1.23 mm, respectively (as shown in Fig. 16). The mean crushing

force obtained from FE analysis is 30.43 kN. Visibly, with this

prole, the sum of side-length and of internal web length B is of

352.62 mm. Substituting items into Eq. (34), the theoretical prediction of mean crushing force is

0:5

P dym:

0:106 1:231:5

m I 1:41

352:620:5

6:46

31:335 kN

2 0:7

38

With 6 cells, the optimal tube type II in Table 3 has the sidelength of 80 mm and the wall thickness of 1.25 mm (Fig. 17). The

mean crushing force received from FE analysis was 37.01 kN.

Simultaneously, for this optimal tube, the sum of side-length and

internal web length B is 439.06 mm. To substitute items into

Eq. (35), the theoretical prediction of mean crushing force is

0:5

P dym:

0:106 1:421:5

m II 1:3

439:060:5

7:41

37:864 kN

2 0:7

39

As listed in Table 3, optimal tube type III has 9 cells (Fig. 18).

The width and the wall thickness of this cross-section are

respectively of 80 mm and 1.21 mm. Then, the mean crushing

force in FE analysis was 53.67 kN. As a matter of course, the

parameter of prole of tube III is B 439.376 mm. Replacing items

into Eq. (36), the theoretical prediction of mean crushing force is

0:5

P dym:

0:106 1:361:5

m III 1:45

439:3760:5

10:089

54:903 kN

2 0:7

40

From the results above, Eq. (36) were adopted to calculate the

mean crushing force for three optimal tubes. Subsequently, the

differences between numerical predictions and theoretical solutions for optimal tube types I, II and III were respectively of 2.97%,

2.3% and 2.05%. These differences show that the proposed equations

are appropriate to the numerical predictions. In addition, the stable

and progressive folding deformation patterns that are developed in

all the three types of tube are the desirable energy-dissipating

mechanism.

5. Conclusions

The proles of three types of tubes were divided into the basic

elements: 2-, 3-, T-shape, 4- and 6-panel angle element. Based on

the Simplied Super Folding Element theory, theoretical expressions of the mean crushing force were proposed for the three

types of triangular multi-cell thin-walled tubes under the axial

crushing loading. Numerical simulations of tubes under the axial

dynamic impact loading were also carried out, and a dynamic

enhancement coefcient was introduced to account for the inertia

effects of aluminum alloy AA6060 T4. Numerical results showed

that tube types I and III were better than tube type II in the aspect

of energy absorption. Simultaneously, the stable and progressive

folding deformation patterns appeared for all the three types

of tubes.

The two RS models of PCF and SEA for each tube were constructed. Pareto sets were obtained by using the linear weighted

average methods (LWAM). In this paper, the Pareto solutions of three

types of tubes were identied to seek out the knee points. The

relative errors between RS approximate value and FE analysis value

at the Knee points were obtained and those were also acceptable.

Finally, the theoretical expressions excellently agreed with the

numerical results, and simultaneously validated the efciency of

the crashworthiness optimization design method based on the

surrogate models and the numerical analysis techniques.

Acknowledgments

The nancial supports from National Natural Science Foundation of China (Nos. 11232004 and 51175160), New Century Excellent Talents Program in University (NCET-12-0168) and Hunan

Provincial Natural Science Foundation (12JJ7001) are gratefully

acknowledged. Moreover, Joint Center for Intelligent New Energy

Vehicle is also gratefully acknowledged.

References

[1] Wierzbicki T, Abramowicz W. On the crushing mechanics of thin-walled

structures. J Appl Mech 1983;50:72734.

[2] Abramowicz W, Jones N. Dynamic axial crushing of square tubes. Int J Impact

Eng 1984;2:179208.

[3] DiPaolo BP, Monteiro PJM, Gronsky R. Quasi-static axial crush response of a

thin-wall, stainless steel box component. Int J Solids Struct 2004;41:370733.

[4] DiPaolo BP, Tom JG. Effects of ambient temperature on a quasi-static axialcrush conguration response of thin-wall, steel box components. Thin-Walled

Struct 2009;47:98497.

circular aluminum tubes. Int J Mech Sci 2001;43:210323.

[6] Ullah H. Buckling of thin-walled cylindrical shells under axial compression. Int

J Numer Methods Eng 2009;79:133253.

[7] Zhang X, Zhang H. Energy absorption of multi-cell stub columns under axial

compression. Thin-Walled Struct 2013;68:15663.

[8] Alavi Nia A, Parsapour M. An investigation on the energy absorption characteristics of multi-cell square tubes. Thin-Walled Struct 2013;68:2634.

[9] Krolak M, Kowal-Michalska K, Mania R, Swiniarski J. Experimental tests of

stability and load carrying capacity of compressed thin-walled multi-cell

columns of triangular cross-section. Thin-Walled Struct 2007;45:8837.

[10] Liu Y. Collapse behaviour and simplied modeling of triangular crosssection

columns. Indian J Eng Mater Sci 2009;16:718.

[11] Hong W, Jin F, Zhou J, Xia Z, Xu Y, Yang L, et al. Quasi-static axial compression

of triangular steel tubes. Thin-Walled Struct 2013;62:107.

[12] Hong W, Fan H, Xia Z, Jin F, Zhou Q, Fang D. Axial crushing behaviors of multicell tubes with triangular lattices. Int J Impact Eng 2014;63:10617.

[13] Hou S, Li Q, Long S, Yang X, Li W. Design optimization of regular hexagonal

thin-walled columns with crashworthiness criteria. Finite Elem Anal Des

2007;43:55565.

[14] Abramowicz W. Thin-walled structures as impact energy absorbers. ThinWalled Struct 2003;41:91107.

[15] Chen W, Wierzbicki T. Relative merits of single-cell, multi-cell and foam-lled

thin-walled structures in energy absorption. Thin-Walled Struct

2001;39:287306.

[16] Kim H-S. New extruded multi-cell aluminum prole for maximum crash

energy absorption and weight efciency. Thin-Walled Struct 2002;40:31127.

[17] Zhang X, Cheng G, Zhang H. Theoretical prediction and numerical simulation

of

multi-cell

square

thin-walled

structures.

Thin-Walled

Struct

2006;44:118591.

[18] Zhang X, Zhang H. Numerical and theoretical studies on energy absorption of

three-panel angle elements. Int J Impact Eng 2012;46:2340.

[19] Naja, A, Rais-Rohani, M. Inuence of cross-sectional geometry on crush

characteristics of multi-cell prismatic columns. In: Proceedings of the 49th

AIAA/ASME/ASCE/AHS/ASC structures, structural dynamics and materials

conference (2008).

195

behaviour of short to long square aluminum tubes subjected to axial loading.

Int J Impact Eng 2004;30:9731003.

[21] Karagiozova D, Jones N. Dynamic buckling of elasticplastic square tubes

under axial impactII: structural response. Int J Impact Eng 2004;30:16792.

[22] Karagiozova D, Alves M. Dynamic elasticplastic buckling of structural

elements: a review. Appl Mech Rev 2008;61:040803.

[23] Tang Z, Liu S, Zhang Z. Analysis of energy absorption characteristics of

cylindrical multi-cell columns. Thin-Walled Struct 2013;62:7584.

[24] Hou S, Li Q, Long S, Yang X, Li W. Multiobjective optimization of multi-cell

sections for the crashworthiness design. Int J Impact Eng 2008;35:135567.

[25] Hou S, Li Q, Long S, Yang X, Li W. Crashworthiness design for foam lled thinwall structures. Mater Des 2009;30:202432.

[26] Hou S, Han X, Sun G, Long S, Li W, Yang X, et al. Multiobjective optimization

for tapered circular tubes. Thin-Walled Struct 2011;49:85563.

[27] Zhang X, Huh H. Crushing analysis of polygonal columns and angle elements.

Int J Impact Eng 2010;37:44151.

[28] Turevsky I, Suresh K. Efcient generation of pareto-optimal topologies for

compliance optimization. Int J Numer Methods Eng 2011;87:120728.

[29] Sun G, Li G, Zhou S, Li H, Hou S, Li Q. Crashworthiness design of vehicle by

using multiobjective robust optimization. Struct Multidiscip Optim

2011;44:99110.

[30] Branke J, Deb K, Dierolf H, Osswald M. Finding knees in multi-objective

optimization. In: Yao X, Burke E, Lozano J, Smith J, Merelo-Guervs J, Bullinaria J,

Rowe J, Tio P, Kabn A, Schwefel H-P, editors. Parallel problem solving from

naturePPSN VIII. Berlin, Heidelberg: Springer; 2004. p. 72231.

[31] Deb K, Gupta S. Understanding knee points in bicriteria problems and their

implications as preferred solution principles. Eng Optim 2011;43:1175204.

[32] Santosa SP, Wierzbicki T, Hanssen AG, Langseth M. Experimental and

numerical studies of foam-lled sections. Int J Impact Eng 2000;24:50934.

[33] Langseth M, Hopperstad OS. Static and dynamic axial crushing of square thinwalled aluminum extrusions. Int J Impact Eng 1996;18:94968.

[34] Hanssen AG, Langseth M, Hopperstad OS. Static and dynamic crushing of

circular aluminum extrusions with aluminum foam ller. Int J Impact Eng

2000;24:475507.

[35] Tam LL, Calladine CR. Inertia and strain-rate effects in a simple plate-structure

under impact loading. Int J Impact Eng 1991;11:34977.

- Lecture15_MPC2Uploaded bysch203
- St-Poz-Garst-3-09Uploaded bybajra88
- Master of technologyUploaded byC V CHANDRASHEKARA
- Numerical Nonlinear Global Optimization in MathematicaUploaded byAndres RH
- TransUploaded byutpalit
- KECUploaded bysappy2shail
- 57c82ea6221906e563c5cf8acba19f84Uploaded byMoseley Miranda
- The Methods of O R on Russian Financial MarketsUploaded byRomina Sessarego
- Algorithms Process OptimizationUploaded byKakashi Hatake
- DBA7301Uploaded bySaravana Kumar
- Civil Engineering Syllabus Revised Upto 8th Semester 2007Uploaded bysubhajit29in
- Lecture SummaryUploaded bysamir_ssh7151
- buckling_analysis_laminates.pdfUploaded byPratik D Upadhyay
- 1-s2.0-S0020746209002078-main(2).pdfUploaded byAmit Chavan
- Cost Optimization of Reinforced Concrete Chimney-libreUploaded byJhonatan SQ
- StrutUploaded bycpsankar
- BucklingUploaded byAnonymous wWOWz9UnW
- OptimizationUploaded byRupak Sapkota
- Steel Design Manual - Sapstl-editadoUploaded bywillys8
- vol7no1_7Uploaded bykutancan
- SW Press Frame Opt ExampleUploaded byMX100
- Terjemahan Materi Dari BlodgettUploaded bylistiyonobudi
- CSD365 Lecture 10 DCT Combined Actions Member RestraintsUploaded bygabuchia
- A12_Design_for_buckling-columns_and_platesUploaded byOSCARDELTA
- Assignment ModelUploaded byHamza
- 72072_sindxUploaded byNhân Nguyễn
- 15 CH-4 Columns and StructsUploaded byTooling ganesh
- Modi MethodUploaded byMehran M Zargar
- Optimal-D.G._1Uploaded byAhmed Zidan
- Urban Management Concepts, principles techniques and education.pdfUploaded byIndra Sugi

- Mechanics of Solids-FinalUploaded byalirezakhadir
- 8 Principal Stresses-Mechanics of materialsUploaded byNhan Tran
- Composite CourseUploaded byNhan Tran
- Columns (1)Uploaded byAbhijith Madabhushi
- 6 shearing stressesUploaded bytigin
- Course NotesUploaded byItaPurwitaSundari
- 11 energy methodsUploaded bytigin
- Classical Mech 2nd Ed Goldstein Solutions 2 Problems 00 Reid p70 pIRXUploaded byelke_hk
- strength of materialsUploaded bySaurabh Sharma
- Contents--Mechanics of MaterialsUploaded byNhan Tran
- Op Tim MatlabUploaded byBiguest Bags
- 1_introductionUploaded byapi-3766020
- Mechanics of Materials Chapter 5Uploaded byJonathan Widodo
- Mechanics of Materials Chapter 5Uploaded byJonathan Widodo
- 3 torsionUploaded bytigin
- 9 Beam Deflection-Mechanics of MaterialsUploaded byNhan Tran
- Presentation Optimization 2-LibreUploaded byNhan Tran
- Crashing Analysis and Multiobjective Optimization for Thin-walled Structures With Functionally Graded ThicknessUploaded byNhan Tran
- Mechanic of MaterialsUploaded byArieska Hermastiadi

- Manual Fabricante TechnicolorTC7210Uploaded byMr Lopez
- DATA ASIMILASI UNTUK PEMULAUploaded bySii Olog-olog Plonk
- amuLogUploaded byreikijoni
- Avamar Fundamentals v2Uploaded byNilesh Race
- Test Paper on Linear Equation and PolynomialsUploaded byAshutosh Kataria
- 664 Optimal ControlUploaded byrodenderon
- Exploration Network Chapter9Uploaded bymeme saleh alharbi
- SAP Diagnostics Agent Logical Hostnames Importance of Sapagentconfigxml FileUploaded byamit435
- PaaS-IaaS-Guide-06042013_0Uploaded byzorgan
- Abulfedae Historia Anteislamica ArabiceUploaded byColin Naturman
- Accenture Achieving High Performance the Value of Bench MarkingUploaded byBiswadip Banerjee
- Mysql ConnectorUploaded byanil
- GS33Y05P10-33E_003Uploaded byapisitu
- TAFJ-UpdaterUploaded byAnonymous GOu6xbdRK
- Reinforcement LearningUploaded bymehdichiti
- openSAP_bw4h1_Unit_1_VALPROP_Presentation.pdfUploaded byStanislava Stanoeva
- GAMING HACKINTOSH READY 2.8GHZ Quad i7 6GB 1TB PC MAC -Uploaded byusernamewiser
- JasperReports-Server-Admin-Guide.pdfUploaded byMohamad Amin
- Fuzzy Equivalence ClassesUploaded bybkjr2008
- Unusable Indexes in OracleUploaded byJP Vijaykumar
- Links Cursos Hotmart - Photoshop , Corel, Sistemas, Premiere.Uploaded byVDF
- Unix Lab SyllabusUploaded byJoinKing
- Social Commerce HypesUploaded byPaulMarsden
- Orange Mali_3G Node B Capacity Assessment Ed01Uploaded bySORO YALAMOUSSA
- 94764_The Finite Element Analysis of Trolly Frame of Stacker Reclaimer Running MechanismUploaded byFarhan Al Azhar
- Secure Active Directory Infrastructure Design Giac Enterprises 163Uploaded byMariano Sanchez
- Data ProcessingUploaded byCarl Saber
- Oracle Vm for Sparc 3.6Uploaded bysc425000
- Functional Programming Using Caml LightUploaded byAnyelo Alejandro Garcia Castellanos
- TLD(x) cannot dismount drive x, slot x already is fullUploaded byN@ts