Mechanics of Materials 166 (2022) 104244

Contents lists available at ScienceDirect

Mechanics of Materials
journal homepage: www.elsevier.com/locate/mechmat

A lightweight rotationally arranged auxetic structure with excellent energy

absorption performance
Wei-Ming Zhang a, Zhen-Yu Li a, Jin-Shui Yang a, b, *, Li Ma c, Zhuang Lin a, Rüdiger Schmidt d,
Kai-Uwe Schröder d
a
College of Aerospace and Civil Engineering, Harbin Engineering University, Harbin, 150001, PR China
b
Qingdao Innovation and Development Base, Harbin Engineering University, Qingdao, 266400, PR China
c
Center for Composite Materials, Harbin Institute of Technology, Harbin, 150080, PR China
d
Institute of Structural Mechanics and Lightweight Design, RWTH Aachen University, Aachen, D-52062, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: As one of the representatives of advanced lightweight cellular materials, the auxetic structure has a great po­
Auxetic structures tential for energy absorption due to its unique deformation mechanism, which makes a very promising appli­
Negative Poisson’s ratio cation prospect in the fields of aerospace, vehicle and engineering protection. In order to further increase the
Mechanical property
energy absorption capacity of auxetic structures, this paper designed and characterized a novel two-dimensional
Energy absorption
(2D) lightweight rotationally arranged auxetic structure. The elastic properties, plastic collapse stress and spe­
cific energy absorption (SEA) of the novel auxetic structure were systematically investigated through theoretical
calculations, numerical simulations and experimental method. Our results verify that the proposed theoretical
model is accurate and applicable. The rotationally arranged shape and stable triangular design inside unit cells
lead to an excellent energy absorption capacity, which is significantly higher than that of conventional auxetic
structures (star-shaped honeycombs and re-entrant hexagonal honeycombs). Moreover, the SEA of novel struc­
ture has extremum conditions, by which the optimal relative density can be determined. According to different
user requirements, the novel structure can be designed as the best energy absorption type or the most concise and
lightweight type by adjusting geometrical parameters.

1. Introduction auxetic structures have appeared. Cabras et al. (Cabras and Brun, 2016)
proposed a class of 3D auxetic lattice structures with an omnidirectional
Cellular, lattice and composite structures have been extensively Poisson’s ratio arbitrarily close to the stability limit of − 1. A cylindrical
developed and widely applied in various engineering fields (Wadley, auxetic structure was expected to be utilized in springs, bumpers,
2006). Thanks to its porosity, it is light in weight compared to a solid dampers and other similar applications (Wang et al., 2016a). A new
bulk made of its base material, while its stiffness, strength or other design thought for 3D auxetic materials was proposed based on the
mechanical properties may have advantages in terms of per unit weight rotation mechanism of chiral honeycombs developed from the
of material, such as high specific stiffness and strength (Yang et al., tetra-chiral honeycombs with circular loops or square loops (Fu et al.,
2020a, 2020b), excellent toughness as well as sound and vibration 2018). The unique sliding mechanism can design 3D printable auxetic
attenuation (Li et al., 2021). In particular, the mechanical significance in metamaterials which has more stable auxetic behavior and higher
biology is gradually being discovered (Wu et al., 2021). compression resistance compared with the conventional auxetic struc­
As a representative cellular mechanical metamaterial, auxetic ma­ tures (Su et al., 2021). Wang et al., 2016b, 2018a, 2018b conducted a
terials with negative Poisson’s ratio have drawn great attention of series of interesting work. Using an interlocking assembly method, they
scholars over the last 30 years ever since Lakes (1987). Poisson’s ratio ν presented composite 3D auxetic structures made from carbon fiber
is one of the most important indexes to indicate the elastic property of reinforced polymer laminates integrating the high specific stiffness, high
solid structures defined as the negative ratio between transversal and specific strength, and lightweight merits of high-performance fiber
longitudinal strains. In recent years, many novel designs of 2D and 3D reinforced polymer composites into 3D auxetic structures with unique

* Corresponding author. College of Aerospace and Civil Engineering, Harbin Engineering University, Harbin, 150001, PR China.
E-mail address: [email protected] (J.-S. Yang).

https://doi.org/10.1016/j.mechmat.2022.104244
Received 15 December 2021; Received in revised form 26 January 2022; Accepted 26 January 2022
Available online 2 February 2022
0167-6636/© 2022 Elsevier Ltd. All rights reserved.
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

properties. An innovative 3D hierarchical design strategy systematically conventional structures or making a little modification on the basis of
summarized the configuration characteristics of existing cellular struc­ conventional configurations. There are few reports on the configuration
tures and proposed a novel class of mechanical metamaterials based on design focusing on the improvement of the energy absorption of the
topology analysis, which have tailored Poisson’s ratio and Young’s auxetic structure. Herein, a novel 2D auxetic structure named WSH (The
modulus (Yang et al., 2019a). Some scholars designed auxetic lattice windmill-like configuration composed of stars and hexagons) with
structures with high stiffness and controllable anisotropic elastic prop­ adjustable parameters was designed in order to improve the mechanical
erties providing a general prescription for solving the challenge of poor properties and energy absorption performance of auxetic structures. The
stiffness in traditional auxetic materials (Gao et al., 2020). Yang et al. theoretical formulas for in-plane elastic constants and plastic collapse
(Yang and Ma, 2020) proposed a systematic design strategy of axisym­ stress have been deduced. A series of WSHs and traditional auxetic
metric auxetic metamaterials with rotationally symmetric transverse structures (such as star honeycombs and re-entrant hexagonal honey­
deformation mode. 2D double-U auxetic honeycombs were designed combs) have been fabricated by 3D printing technology named Multi Jet
based on traditional double-V honeycombs, with low stress concentra­ Fusion (MJF). The quasi-static compressive tests were conducted on the
tion and manufacturing damage, leading to better load resistance and structures along the in-plane direction. The comparison of theoretical,
energy absorbing capacity (Guo et al., 2020; Yang et al., 2019b). Some experimental and finite element analysis (FEA) results verified the
scholars have applied the auxetic design to other aspects. Combining the applicability and accuracy of the proposed theoretical model. Based on
auxetic structure and the sensor design to apply the auxetic performance the above, the energy absorption characteristics of the WSHs were
to stretchable strain sensors can enhance the sensitivity significantly analyzed and discussed after the comparison of specific energy absorp­
(Jiang et al., 2018). In recent years, several research works have tion (SEA) between WSHs and traditional auxetic structures. Then the
demonstrated the possibility of combining different properties of me­ influence of geometric parameters was discussed based on the theoret­
chanical metamaterials. A metastructure was devised by Wei et al., ical and FE analysis.
2018, 2021 to obtain tailorable thermal expansion and tunable Poisson’s
ratio which provided expectable feasibility for both 2. Theoretical analysis
temperature-sensitive and mechanical-sensitive structures as well as
multi-field coupling driven devices. This chapter will characterize the geometric relationship, elastic
It is worth noting the outstanding performance of auxetic structures constants, and plastic collapse of the structural unit cell.
in energy absorption (Zhang et al., 2020a). As is well known, the cellular The WSH (The windmill-like configuration composed of stars and
structural feature is desirable when it comes to energy absorption, since hexagons) structure is a combination of star-shaped honeycombs and re-
it can provide a long, nearly constant plateau stress stage under entrant hexagonal honeycombs as shown in Fig. 1(a). The centrally
compressive loading (de Waal et al., 2021; Ha et al., 2021a, 2021b; symmetrical arrangement will lead to rotational deformation under
Wang et al., 2021; Xiang et al., 2016; Yang et al., 2021; Zhang et al., compression loads, which make the structural deformation more uni­
2020b). On the basis of these, going a step further, the deformation form. Moreover, the triangles formed by closed connections in structures
characteristics of auxetic structures will lead to a great improvement in will effectively enhance strength and stability.
the plateau stage resulting in a structural enhancement under The geometrical parameters of unit cells are defined in Fig. 1(b). In
compression loads. Therefore, the research on the impact behavior of addition, the following relationship equations of basic parameters
conventional auxetic structures has gradually attracted attention (Hu should be satisfied for the construction of WSH:
et al., 2018; Li et al., 2020). Recent research shows that the relative
R1
strength for rotation-dominated lattices can be better than λ= = 1 − tan θ (1)
R2
bending-dominated structures but generally lower than
stretch-dominated geometries (Seetoh et al., 2021). The gradient design t
method for traditional structures is also applicable to advanced auxetic γ= (2)
R2
structures. The response of graded auxetic structure is less stiff and
Two important geometric lengths (H and K) are expressed by basic
sustains gradual failure in comparison to the response of uniform
parameters:
structures, which is stiffer but sustains abrupt failure at smaller strains
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
(Novak et al., 2020). Ma et al. (Wang et al., 2016c, 2017, 2019a; Zhao
H = R2 λ2 − 2λ + 2 (3)
et al., 2018) researched and analyzed the impact resistance of auxetic
structures introduced and applied it for the jounce bumper. The auxetic √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
K = R2 λ2 − 4λ + 5 (4)
jounce bumper can achieve similar mechanical behavior but with a
smoother load-displacement curve, which is beneficial to the damping Relative density is vitally important for structural design of cellular
performance compared with traditional jounce bumper. Imbalzano materials, which has great influence on the mechanical properties. It is
et al., 2015, 2016, 2018 conducted a series of work on the energy ab­ defined as the density fraction of structure density ρ* to the solid ma­
sorption behavior of auxetic structures under impact or blast loads, and terials density ρs . In this paper, the relative density of 2D WSH structures
many meaningful conclusions were proposed. Recent years, a series of can be calculated as below:
structures containing sandwich panels, novel auxetic honeycombs or
other conventional auxetic configurations were proposed. In (Qi et al., ρ* 4 ( √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅ ) t
ρr = = 4 λ2 − 2λ + 2 + 2 λ2 − 4λ + 5 − λ + 4 + 2 (5)
2013, 2017, 2019, 2020a, 2020b), parametric studies concerning the ρs 25 R2
thickness and geometric parameters were conducted and multi-scale Some WSH configurations with different λ are shown in Fig. 2(a)-(d).
optimizations were performed to improve anti-blast or anti-impact It can be found the decrease of λ led to an increase of the concave angle,
performance. Studies have shown that the structure with a combina­ which caused an increase of the relative density.
tion of star-shaped and concave honeycombs has better impact resis­ The densification strain is a parameter closely related to the energy
tance (Wang et al., 2019b, 2019c). In order to get the better combination absorption of the cellular and lattice structures. Theoretically, densifi­
of geometric parameters, topology optimization methods have also been cation strain would simply be equal to the porosity and the porosity is
applied. Through the topology optimization method, the novel auxetic
equal to 1-ρr. But it’s found in the experiment that the densification
topologies computed with an objective of maximum energy absorption
could be designed and analyzed (Bohara et al., 2021; Yang and Li, 2020). strain is smaller than the porosity (Qiu et al., 2009). Densification strain
At present, as far as the authors know, most of the research on energy can be calculated by an expression with an empirical coefficient:
absorption performance of auxetic structures has focused on

2
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Fig. 1. The WSH auxetic configurations: (a) design of the unit cells; (b) geometrical parameters of unit cells.

Fig. 2. 2D WSH configurations with different λ: (a) λ is equal to 1; (b) λ is equal to 2/3; (c) λ is equal to 1/2; (d) λ is equal to 0.

4 ( √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅ ) t parameter.

εd = 1 − η 4 λ2 − 2λ + 2 + 2 λ2 − 4λ + 5 − λ + 4 + 2 (6)
25 R2 The elastic constants in the analytical model mainly include Young’s
modulus, Poisson’s ratio and shear modulus. The unit cell of 2D WSH
where η is a factor depends on the cell structure, which ranges in η = structures under in-plane uniaxial compression load is shown in Fig. 3
1–3. For honeycombs, Gibson and Ashby suggested η = 1.4. In this (a). A simplified analysis model can be established according to the load
paper, η = 1.4 was taken only in the part of results comparison. Since the state and geometric characteristics of the unit cell as shown in Fig. 3(b),
value of η does not affect the linear relationship between densification the uniformly distributed load is transferred to the adjacent unit cells
strain and relative density, in the other parts of theoretical analysis, η through the triangular connection part between the unit cells. The in­
was not given a specific value, but was regarded as an unknown ternal force of each part can be calculated by the force method as shown

3
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Fig. 3. Theoretical model of WSH configurations: (a) in-plane uniaxial compression; (b) simplified analysis model; (c) force method model; (d) representative struts.

in Fig. 3(c). Two representative struts “strR” and “strT” are taken out for deflections of strR and strT can be expressed as:
calculations according to the force state and central symmetry as shown
R42 ( 2 )
in Fig. 3(d). δr = σ 1 λ − 2λ + 2 (15)
Es t 3
The bending moments MA , MB , MC and reaction forces PA , PB , PC ,
which are applied to the points A, B and C, respectively, can be given
R42 ( 2 )
according to the force equilibrium in the X direction: δt = σ 1 λ − 2λ + 2 (1 − λ) (16)
Es t3
pA = σ1 Lb (7)
The strains of the 2D WSH unit cell in the X and Y direction can be
obtained by the deflection of representative struts:
1
pB = pC = σ1 Lb (8) √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅)̅( )3
2 δ cos θ σ1 ( 2 R2
ε1 = r = λ − 2λ + 2 (17)
L 2Es t
where b is the width of structures in the stretching direction (out-of-
plane direction). The X direction is defined as the loading direction, and √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ( )
2δr sin θ + δt cos θ+ 3σ1 ( )̅ R2 3
the Y direction is orthogonal to the X direction. The representative struts ε2 = = (1 − λ) λ2 − 2λ + 2 (18)
strR and strT will show bending deformations in the XY plane during 2L 4Es t
loading. The bending moments can be given according to the moment The Young’s modulus in the X direction can be expressed as:
equilibrium: ( )
E1* σ 1 ( )− 1 t 3
1 = = 2 λ2 − 2λ + 2 2 (19)
M A = PA L (9) Es ε 1 R2
8
The Poisson’s ratio and shear modulus in the XY plane can be
1
MC = PA L tan θ (10) expressed as:
8
ε2 3
The deflections of strR and strT can be given by force method: ν*12 = − = (λ − 1) (20)
ε1 2
( ) ( )
1 1
δr = δr11 σ1 Lb cos θ + δr12 σ1 L2 b (11) G*12 E1* 2 ( 2 )− 1
( )3
t
2 8 = = λ − 2λ + 2 2
(21)
Es 2(ν12 * + 1)Es 3λ − 1 R2
( ) ( )
1 1
δt = δt11 σ1 Lb sin θ + δt12 σ1 L2 b tan θ (12) The structures perform linear elastic behavior within the elastic
2 8
stage, while the cellular structures will show plastic collapse after the
material reaches the yield point. The SEA (Specific Energy Absorption)
where δij means the ith displacement component due to the force in j
of cellular materials and structures is mainly attributed to the plastic
direction, given by
collapse stage, so the plateau stress level is a key property of cellular
R32 ( 2 )3 structures. The critical points of the plastic collapse stage will be
δr11 = δt11 = λ − 2λ + 2 2 (13)
3Es I determined when the bending moments of struts are equal to the plastic
moment Mp . In the case of a rigid-plastic beam or plate with a rectan­
R22 ( 2 ) gular cross section, the total work done by external stress is equal to the
δr12 = δt12 = − λ − 2λ + 2 (14)
2Es I plastic work done by plastic hinges as expressed by

where Es is the Young’s modulus of the constitute material of the σ *p1 ⋅ 2Lb⋅2Hφ1 cos θ = 28Mp φ1 (22)
structures, I is the second moment of inertia. Accordingly, the

4
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

1 Table 1
Mp = Yc bt2 (23) Properties of parent material Nylon PA12.
4
Parameters Es (MPa) ν ρs (kg/m3) Yc (MPa)
where φ1 is the plastic rotation and Yc is the yield stress of the parent
Value 1358 0.33 940 27
material, the plastic plateau stress of 2D WSH unit cells under quasi-
static loading can be given as:
( )2
σ *p1 28Mp φ1 7 t Table 2
= = (24)
Yc 2Lb⋅2Hφ1 cos θ⋅Yc 8 R2 Specimen labels and geometrical parameters of 2D WSH auxetic configurations.
Specimen labels λ γ t (mm) Size (mm3)
3. Experiment overview N01 0 0.152 1 125 × 125 × 63
N12 1
This part will describe the experiment preparation and process, 2
N23 2
including sample design, material parameters, structure dimensions, 3
and a brief introduction to 3D printing technology and testing machines. N11 1
The 3D printed models were constructed by CAD software AutoCAD
as shown in Fig. 4. The parent material properties were obtained
through the standard material test of Nylon PA12 by the Center for Table 3
Composite Materials as listed in Table 1. The specimen labels and Specimen labels and geometrical parameters of conventional auxetic
geometrical parameters are listed in Table 2 and Table 3, where Nij configurations.
means the geometrical parameter λ of WSH is equal to i/j; S means star- Specimen Length of unit Angle of unit cell t Size (mm3)
shaped honeycombs and H means re-entrant hexagonal honeycombs. labels cell (mm) (degree) (mm)
The specimens were fabricated via a powerful 3D printing technique
S 16.0 36.8 1 125 × 125 ×
named Multi Jet Fusion (MJF) as shown in Fig. 5(a) and structures H 16.3 63.4 63
printed by MJF techniques have been proved to have good appearance,
quality and details. Quasi-static compression tests are adopted to
investigate the performance parameters of 2D WSHs and conventional behavior will be established. The interaction, boundary conditions,
auxetic structures (star-shaped honeycombs and re-entrant hexagonal element parameters and other settings will be introduced. The numeri­
honeycombs). The quasi-static in-plane experiments were conducted in cal simulations were conducted by commercial finite element software
the compression testing machine (Instron 4505 with a 100 kN load cell) package ABAQUS/Standard, with a comparing of compression behavior
consisting of the electronic dial gauge, force sensor, base platform and of 2D WSH auxetic structures and conventional auxetic structures. The
compress head. A series of compression tests of 2D WSHs and conven­ finite element (FE) models and boundary conditions were constructed as
tional auxetic structures were conducted. During tests, the specimens shown in Fig. 6(a)-(b). The 8-node linear hexahedral solid element
were placed between two metal platforms and compressed by the head C3D8R with reduced integration and hourglass control was used to mesh
at the speed of 0.5 mm/min with displacement control. The values of the the geometric models. Mesh convergence tests of simulations are pre­
pressure/force and compression displacement were measured by testing sented as shown in Fig. 6(c), and the global size of solid elements was set
machine while the laser extensometer recorded lateral displacements up as 0.5 mm with 612,360 elements considering the accuracy and
automatically. A digital camera was placed facing the specimens to re­ calculation cost. The structures were placed between two rigid plates,
cord the videos. The equivalent Young’s modulus of the 2D WSHs and the bottom plate was fully fixed and the downward displacement load in
conventional auxetic structures can be derived from the slope of the vertical direction was applied to the top plate with other freedom de­
stress-strain curves at small displacement. The nominal stress and strain grees fixed. General contact was adopted in the model to prevent the
were calculated by original dimensions and displacements of specimens. penetration during loading. The hard contact was assigned to the normal
In quasi-static collapse tests, the maximum displacement of compression behavior and the tangential friction coefficient was 0.15. FE analyses of
was set up as 80 mm. States of the specimens before and after testing are quasi-static compression adopt ideal elasto-plastic constitutive model,
diagrammed in Fig. 5(b). which does not consider strain rate effect and strain hardening. This
means although the material continues to plastically deform, the stress
4. Numerical simulation remains constant from the initial yield until fracture. The resulting er­
rors will be discussed in the section of results comparison.
In this part, finite element models of quasi-static compression

Fig. 4. The 3D printed models of 4 configurations of WSH and 2 configurations of conventional auxetic structures.

5
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Fig. 5. Specimens and compression tests: (a) specimens of 2D WSH, star-shaped honeycombs and re-entrant hexagonal honeycombs; (b) states before testing and
after testing.

5. Result and discussion longest state, which means parameter γ is small and strut BD will be
prone to instability. The stability critical load can be determined by use
5.1. Theoretical, experimental and numerical results of the following formula:

nπ2 Es I
This section will show the experimental and numerical results of four σ cr Lb = (26)
WSH configurations and two traditional structures. The results of the­ (L − R1 )2
ory, experiment and FEA will be compared to verify the accuracy of the
where n is a factor reflecting the restraint of strut. While n is equal to 2,
theoretical model and the reliability of the simulation model. Among
the stability critical load can be expressed as
them, N01 will be discussed as a special configuration.
As shown in Fig. 7, the shrinking of all structures in the horizontal π 2 Es 1
( )3
t
direction reflects the auxetic behavior of WSHs and conventional auxetic σ cr = 2
(27)
12 (2 − λ) R2
structures. As the compression continues, the structure as a whole be­
comes unstable and the plastic collapse of the structures causes the cell The initial elastic strain is axial strain of the strut BD, which can be
walls adhere to each other to reach the densification strain. It can be calculated by
clearly seen that the deformation of configuration S and H are very σ1 Lb
uneven. As a result, the materials in some parts are hardly involved in ε1 = (28)
Es A
energy absorption. WSH structures deform uniformly and the triangular
parts cause structures to be more stable and each part of structures can Therefore, we can get the Young’s modulus as
be well involved in energy absorption.
E1* t
It has been defined that γ is equal to t/R2 in Eq. (2), which represents = (29)
Es 2R2
the relationship between cell wall thickness and the unit cell size. The
relationship between performance parameters and geometric parame­ The strain after instability is mainly caused by the bending of strut
ters (λ and γ) is shown in Table 4. It can be seen that the relative density BD as shown in Fig. 8(b)–(c), the strain in X direction can be estimated
and densification strain are positively correlated with the first power of from rotation angle of strut BD after buckling:
γ. Young’s modulus and shear modulus are positively correlated with the L0 − Ld sin β
cubic power of γ. The plateau stress and quadratic power of γ are posi­ ε1 = =1− (30)
L0 β
tively correlated. Poisson’s ratio is basically not affected by γ under the
same λ. Results of specimen N01 obtained by Eqs. (25)-(29) as listed in
It should be noted that N01 (λ equal to 0) is an extremely special Table 5 are closer to experimental and numerical results.
configuration needed to be analyzed separately. Since each side of the As shown in Fig. 9, once a layer of cells is completely crushed after
star-shaped parts in the unit cell is firmly linked to the center of sym­ rotating deformation, the cell walls contact each other. This localized
metry, the position that had a concave angle gap originally is replaced deformation will propagate to another adjacent layer of cells. When the
by a stable triangle. Connection methods of configuration N01 in the first batch of cells is broken, the next batch of cells will repeat. In the
center position are shown in Fig. 8(a). The theoretical connection above process, the stress rises and falls again which leads to the stepped
method of N01 is separated at the center while the actual connection fluctuation in the plateau stage.
method of N01 is joined at the center. The stress-strain curves are shown in Fig. 10(a)-(f), and it can be
The relative density of N01 with joined center can be calculated as noted that the theoretical, numerical and experimental results have
below: good agreement in all configurations of WSH, which can validate the
accuracy of theoretical analysis. The curves well reflect the compression
8 ( √̅̅̅ √̅̅̅ ) t
ρr = 2 2 + 5 +2 (25) characteristics of the cellular structures, the low peak stress and the long
25 R2 plateau stage. Errors in the results are caused by the following reasons:
The significant improvement in stiffness causes a difficulty in plastic due to internal defects of specimens caused by 3D printing process, parts
deformation. In addition, the side length of star-shaped parts reaches the of the specimens’ load-bearing capacities are completely lost during

6
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Fig. 6. Finite element models and mesh convergence tests: (a) front view of model; (b) side view of model; (c) stress-strain curves of mesh convergence tests under
uniaxial compression at 10 m/s.

compression. But FE models adopt ideal assumption which can still the structures.
maintain a steady stress and not fail completely after materials reach the
yield point. These are reflected in the stress-strain curves. The plateau 5.2. Energy absorption performance and extremum conditions
stress is flatter during compression, while the simulation curve rises
slightly which reflects the plateau enhancement of auxetic structure. This section will show the energy absorption performance of
Furthermore, considering the calculation efficiency, the limited mesh different structures and the advantages of WSH structures compared to
layers in the cell wall direction of FE models result in a slightly higher conventional auxetic structures. In addition, WSH structures have
overall stiffness of the structures. The theoretical calculations of elastic parameter adjustability and this section will calculate the extreme
constants and yield strength are based on the assumptions of thin rods, conditions of specific energy absorption. Then the expression of the best
thin plates and ideal elasto-plastic model, which result in the theoretical relative density for SEA will be given.
calculations being slightly lower in any case than the measured values of The energy absorption capacity of a structure is usually evaluated by
the test pieces and the simulation results, which use solid elements the energy absorption per unit volume or per unit mass (specific energy
(rather than shell elements). The theoretical model over predicts Pois­ absorption), which can be calculated as follows:
son’s ratios at some configurations as a consequence of the thin-walled ∫ εd
assumption. Ev = σ (ε)dε (31)
The maximum error reaches 1.5 times occurring in the configuration 0

S as shown in Fig. 11. Although there are small errors in theoretical, ∫ εd

v* σ(ε)dε
experimental and simulation results, they are basically reflected in the Em = 0
(32)
numerical value. The development trend of structural characteristics m
and shapes of curves are in good agreement, which does not affect the where v* is the volume of structures, and m is the mass of structures.
study of compression behavior and energy absorption characteristics of Furthermore, Ev represents the energy absorption per unit volume and

7
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Fig. 7. Deformation process of experiment and FEA under quasi-static compression: (a) specimen of N01; (b) specimen of N12; (c) specimen of N23; (d) specimen of
N11; (e) specimen of S; (f) specimen of H.

Em represents the energy absorption per unit mass. inside unit cells, which make the structural deformation more uniform
It can be intuitively seen the energy absorption capacity of each and more rigid in the compression process. Furthermore, this charac­
specimen in the quasi-static uniaxial compression test from Fig. 12. The teristic does not cause a big increase in relative density.
configuration N01 with rotating mechanism has excellent performance In addition, N11 is a configuration worthy of attention, with more
and the highest energy absorption. Other configurations of WSH have concise geometric features and smaller relative density. Moreover, its
similar performance, which indicates that the effect of λ on the energy energy absorption capacity is as good as other configurations, all of
absorption is not significant. Nevertheless, the energy absorption of which show that N11 is more practical and conducive to manufacturing.
WSH is significantly higher than that of conventional auxetic structures Interestingly, if we do not consider the deformation due to lightly
(star-shaped honeycombs and re-entrant hexagonal honeycombs), the rotation of unit cells during small deformation stage, the theoretical
SEA has increased by 60%. The best configuration N01 has increased by calculated Poisson’s ratio of N11 is equal to 0 and the overall structure
110%. will only have inconspicuous expansion and shrink deformation.
The excellent energy absorption capacity of WSH is due to the For cellular structures, plateau stress and densification strain have
rotationally arranged structure and the stable triangular structures the greatest influence on energy absorption. The relationship between

8
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Table 4
1
Performance parameters determined by different geometric parameters. ρr |Em (max) = (37)
2η
Dimensionless λ =0 1 2 λ =1
λ = λ =
parameters 2 3 The specific energy absorption curve is shown in Fig. 13(a), from
ρr 2.487γ 2.078γ 1.967γ 1.799γ which can be found that energy absorption reaches the maximum at the
εd 1− 1− 1− 1− extremum condition point. The SEA will decrease as moving to the left
2.487ηγ 2.078ηγ 1.967ηγ 1.799ηγ from the extremum condition point. It’s caused by the low plateau stress
E1 *
√̅̅̅ 3
2γ
√̅̅̅
4 5 3
√̅̅̅̅̅̅
3 10 3 2γ3 making the overall energy absorption smaller, although the structure
has a long plateau stage and a large densification strain. Similarly,
γ γ
Es 5 5
ν12 * 3 3 1 0
− − −
2 ̅̅̅
√ √4̅̅̅ √2̅̅̅̅̅̅
G12 * 2γ3 8 5 3 3 10 3 γ3
−
γ γ Table 5
Es 5 5
σp1 * 7 2 7 2 7 2 7 2 Performance parameters of configuration N01 with a joined center.
γ γ γ γ
Yc 8 8 8 8 Specimen label ρr εd ν12 * E1 * G12 * σcr
Es Es Es

stress and strain can be considered as the rigid-plastic model, which is N01 2.260γ 1 − 2.260ηγ 0 0.5γ 0.25γ 0.206γ3
completely locked after the densification strain. The SEA (Specific En­
ergy Absorption) of 2D WSH auxetic structures can be expressed as
∫ε
v* 0 d σ(ε)dε σ *p1 εd
Em = = (33)
m ρs ρr
Calculating the specific energy absorption of the structure under
quasi-static load with t/R2 as the independent variable yields:
( ) [ ( )2 ( )]
t 7Yc t 1 t
Em = f = − η + (34)
R2 8ρs R2 k R2

From Eq. (5), at each value of λ, ρr is a linear function of t/R2 as

expressed by
t
ρr = k (35)
R2

where k can be given by

4 ( √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅ )
k= 4 λ2 − 2λ + 2 + 2 λ2 − 4λ + 5 − λ + 4 + 2 (36)
25
Then the extremum condition of SEA can be derived:
Fig. 9. The fluctuation process of the stress-strain curve of N01 at the
plateau stage.

Fig. 8. Different connection methods at center and deformation mode of configuration N01: (a) four points separated or joined at the center of star; (b) bending
deformation of strut BD after buckling; (c) experimental photos of strut BD bending state.

9
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Fig. 10. Comparison of theoretical, numerical and experimental results of 2D WSH auxetic configurations and conventional auxetic structures: (a) stress-strain
curves of N01; (b) stress-strain curves of N12; (c) stress-strain curves of N23; (d) stress-strain curves of N11; (e) stress-strain curves of S; (f) stress-strain curves
of H.

moving to the right from the extremum condition point will also reduce
2
the SEA. It’s caused by the short plateau stage making the structure ρr |Em (max) = (39)
3η
reach the densification strain quickly, which will tend to end the energy
absorption stage immediately, although the plateau stress becomes
higher. It also conforms to this law for N01 with joined symmetric center 5.3. Effect of geometrical parameters λ and γ
as shown in Fig. 13(b), which can be calculated as follows:
[( ) This chapter will use the parameters λ and γ as variables to analyze
( ) ( )3 ]
t π 2 Es t
2
t the energy absorption behavior of the WSH structures. The influence of
Em = f = − ηk (38)
R2 48ρs k R2 R2 various parameters on structural elastic constants, collapse stress and
SEA will be discussed. The conclusions will provide a basis for the en­
Similarly, the extremum condition of SEA of N01 can be derived: ergy absorption design of lightweight auxetic structures.

10
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

performance parameters of WSHs will be discussed in detail below.

It can be clearly seen from Fig. 15(a) that the increase of λ resulted in
a small increase of Young’s modulus, which indicates that Young’s
modulus of WSH is not sensitive to the change of λ. However, near the
approach of λ to 0, Young’s modulus increases sharply as a result of the
rigidity advantage of the special configuration. Overall, Young’s
modulus of WSH is more sensitive to the change of γ and increased by 10
times with the change of γ from 0.1 to 0.2.
The effect of geometrical parameters on the shear modulus of WSH is
obvious as shown in Fig. 15(b). The distribution of the shear modulus in
the λ direction has a positive and negative critical position, which is
caused by Poisson’s ratio tending to − 1 and the corresponding λ is 1/3.
The relative density decreases with the increase of λ. While λ ap­
proaches 0 due to the change of configuration, a good phenomenon will
be caused that the relative density decreases as shown in Fig. 15(c).
Overall, the influence of γ on the relative density is dominant, which
indicates that the relative density of the structures can be effectively
controlled by the design of the relationship between cell wall thickness
and cell size.
The Poisson ratio of WSH is a geometric property, which is mainly
determined by λ in thin-walled configurations and is extremely insen­
Fig. 11. Poisson’s ratio of different configurations.
sitive to cell wall thickness as shown in Fig. 15(d).
It can be clearly seen from Fig. 15(e), plateau stress is not sensitive to
Fig. 14 shows the numerical simulation and theoretical calculation
the change of λ, the influence of γ is dominant. The substantial increase
results with different geometric parameters, including the change of γ.
in the upper left corner is an excellent performance of structures with λ
The theoretical prediction and the numerical results are in good agree­
equal to 0.
ment. It can be found that the influence of γ on the stress-strain curves is
The energy absorption capacity is extremely sensitive to the geo­
more obvious than that of λ. According to numerical simulations and
metric parameters λ and γ. The improvement of λ leads to a longer
theoretical calculations, the influence of geometric parameters on the

Fig. 12. Comparison of energy absorption curves between WSHs and conventional auxetic structures: (a) Energy absorption per unit volume; (b) Energy absorption
per unit mass (SEA).

Fig. 13. SEA of WSH as a function of t/R2: (a) configurations with λ > 0; (b) configuration N01 with joined center.

11
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Fig. 14. Comparison of theoretical and numerical results with different geometrical parameters.

plateau stage, which results in the improvement of SEA. As a special (2) The energy absorption of 2D WSHs is significantly higher than
case, the change of the deformation mechanism causes a significant that of conventional auxetic structures (star-shaped honeycombs
improvement in the SEA with λ equal to 0. More interestingly, the in­ and re-entrant hexagonal honeycombs), the SEA has increased at
fluence of γ on SEA has an extremum condition. It can be found from least by 60% while the best configuration among them has
Fig. 15(f) that the maximum values of SEA appear at the vertex position increased by 110%. The excellent energy absorption capacity of
of each contour line, and the connecting line of these points determines 2D WSHs benefits from the rotationally arranged structure and
the extremum condition of SEA. the stable triangular structures inside unit cells, which make the
structural deformation more uniform and more rigid in the
6. Conclusions compression process. Furthermore, this characteristic won’t
cause a big increase in relative density.
In this study, a novel 2D auxetic structure with adjustable parameters (3) As a special case, the configuration with λ equal to 0 (N01) has the
was designed in order to improve the mechanical properties and energy best mechanical properties benefitting from the rotating mecha­
absorption performance of auxetic structures. Theoretical derivation, nism. In addition, the configuration with λ equal to 1 (N11) has
numerical simulation and in-plane quasi-static compression tests were more concise geometric features and smaller relative density.
carried out. Some significant conclusions were obtained based on Moreover, its energy absorption capacity is as good as other WSH
analysis results. configurations, all of which show that it is more practical and
conducive to manufacturing.
(1) Theoretical formulas for elastic constants and plastic collapse (4) The SEA function with γ as the independent variable is estab­
stress in-plane of 2D WSH structures have been deduced. The lished based on the rigid-plastic model. The extremum conditions
comparisons of theoretical, experimental and FEA results verify of SEA and the optimal relative density are proposed.
applicability and accuracy of the proposed theoretical model. The
geometric parameters λ and γ are the basic parameters of the 2D Author statement
WSH structures determining and controlling the performance
parameters and energy absorption capacity. The Young’s We declare that all persons who have substantial contributions to the
modulus, relative density and plateau stress are more sensitive to present work are listed in the manuscript. Prof. Jin-Shui Yang is the
the change of γ while the influence of λ on Poisson’s ratio is corresponding author and has made substantial contributions to the
dominant. Shear modulus and specific energy absorption are conception and design of the work; Mr. Wei-Ming Zhang carried out
under the common influence of both λ and γ. theoretical models and designed experiments; Mr. Zhen-Yu Li carried

12
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Fig. 15. Dependence of the performance parameters on λ and γ for 2D WSH auxetic structures: (a) Young’s modulus; (b) shear modulus; (c) relative density; (d)
Poisson’s ratio; (e) plateau stress; (f) specific energy absorption.

out experiments and carried out simulations; Prof. Li Ma and Prof. Acknowledgments
Zhuang Lin reviewed and edited this work; Prof. Rüdiger Schmidt and
Prof. Kai-Uwe Schröder revised the manuscript. The present work was supported by the National Natural Science
Foundation of China under Grant No. 12172098, the Outstanding Youth
Declaration of competing interest Project of Heilongjiang Natural Science Foundation under Grant No. YQ
2020A001 and the China Postdoctoral Science Foundation under Grant
The authors declared that they have no conflicts of interest to this No. 2020T130141.
work titled“A lightweight rotationally arranged auxetic structure with
excellent energy absorption performance”. References
We declare that we do not have any commercial or associative in­
terest that represents a conflict of interest in connection with the work Bohara, R.P., Linforth, S., Nguyen, T., Ghazlan, A., Ngo, T., 2021. Novel lightweight
high-energy absorbing auxetic structures guided by topology optimisation. Int. J.
submitted. Mech. Sci., 106793

13
W.-M. Zhang et al. Mechanics of Materials 166 (2022) 104244

Cabras, L., Brun, M., 2016. A class of auxetic three-dimensional lattices. J. Mech. Phys. Wang, Y., Wang, L., Ma, Z-d, Wang, T., 2016a. Parametric analysis of a cylindrical
Solid. 91, 56–72. negative Poisson’s ratio structure. Smart Mater. Struct. 25 (3), 035038.
de Waal, L., Lu, G., Zhang, J., You, Z., 2021. Dynamic behaviour of graded origami Wang, X.-T., Li, X.-W., Ma, L., 2016b. Interlocking assembled 3D auxetic cellular
honeycomb. Int. J. Impact Eng. 157, 103976. structures. Mater. Des. 99, 467–476.
Fu, M., Liu, F., Hu, L., 2018. A novel category of 3D chiral material with negative Wang, Y., Wang, L., Ma, Z-d, Wang, T., 2016c. A negative Poisson’s ratio suspension
Poisson’s ratio. Compos. Sci. Technol. 160, 111–118. jounce bumper. Mater. Des. 103, 90–99.
Gao, Y., Wu, Q., Wei, X., Zhou, Z., Xiong, J., 2020. Composite tree-like re-entrant Wang, Y., Wang, L., Ma, Z-d, Wang, T., 2017. Finite element analysis of a jounce bumper
structure with high stiffness and controllable elastic anisotropy. Int. J. Solid Struct. with negative Poisson’s ratio structure. Proc Inst Mech Eng C J Mech Eng Sci 231
206, 170–182. (23), 4374–4387.
Guo, M.-F., Yang, H., Ma, L., 2020. Design and analysis of 2D double-U auxetic Wang, X.-T., Chen, Y.-L., Ma, L., 2018a. The manufacture and characterization of
honeycombs. Thin-Walled Struct. 155, 106915. composite three-dimensional re-entrant auxetic cellular structures made from carbon
Ha, N.S., Pham, T.M., Chen, W., Hao, H., Lu, G., 2021a. Crashworthiness analysis of bio- fiber reinforced polymer. J. Compos. Mater. 52 (23), 3265–3273.
inspired fractal tree-like multi-cell circular tubes under axial crushing. Thin-Walled Wang, X.-T., Wang, B., Wen, Z.-H., Ma, L., 2018b. Fabrication and mechanical properties
Struct. 169, 108315. of CFRP composite three-dimensional double-arrow-head auxetic structures.
Ha, N.S., Pham, T.M., Hao, H., Lu, G., 2021b. Energy absorption characteristics of bio- Compos. Sci. Technol. 164, 92–102.
inspired hierarchical multi-cell square tubes under axial crushing. Int. J. Mech. Sci. Wang, T., Li, Z., Wang, L., Ma, Z., Hulbert, G.M., 2019a. Dynamic crushing analysis of a
201, 106464. three-dimensional Re-entrant auxetic cellular structure. Materials 12 (3).
Hu, L., Zhou, M.Z., Deng, H.J., 2018. Dynamic crushing response of auxetic honeycombs Wang, H., Lu, Z., Yang, Z., Li, X.J., 2019b. A novel re-entrant auxetic honeycomb with
under large deformation: theoretical analysis and numerical simulation. Thin-Walled enhanced in-plane impact resistance. Compos. Struct. 208, 758–770.
Struct. 131, 373–384. Wang, H., Lu, Z., Yang, Z., Li, X.J., 2019c. In-plane dynamic crushing behaviors of a
Imbalzano, G., Tran, P., Ngo, T.D., Lee, P.V.S., 2015. Three-dimensional modelling of novel auxetic honeycomb with two plateau stress regions. Int. J. Mech. Sci. 151,
auxetic sandwich panels for localised impact resistance. J. Sandw. Struct. Mater. 19 746–759.
(3), 291–316. Wang, H., Long, S., Yao, X., Lu, G., Zhang, X., Han, Q., 2021. Analytical study on the low-
Imbalzano, G., Tran, P., Ngo, T.D., Lee, P.V.S., 2016. A numerical study of auxetic velocity impact penetration of the fully-clamped foam-core composite sandwich
composite panels under blast loadings. Compos. Struct. 135, 339–352. panels. Compos. B Eng. 224, 109214.
Imbalzano, G., Linforth, S., Ngo, T.D., Lee, P.V.S., Tran, P., 2018. Blast resistance of Wei, K., Peng, Y., Qu, Z., Pei, Y., Fang, D., 2018. A cellular metastructure incorporating
auxetic and honeycomb sandwich panels: comparisons and parametric designs. coupled negative thermal expansion and negative Poisson’s ratio. Int. J. Solid Struct.
Compos. Struct. 183, 242–261. 150, 255–267.
Jiang, Y., Liu, Z., Matsuhisa, N., Qi, D., Leow, W.R., Yang, H., et al., 2018. Auxetic Wei, K., Xu, W., Ling, B., Yang, X., Fang, D., 2021. Multi-functional cylindrical
mechanical metamaterials to enhance sensitivity of stretchable strain sensors. Adv. metastructures to simultaneously program both thermal expansion and Poisson’s
Mater. 30 (12), 1706589. ratio. Extreme Mech. Lett. 43, 101177.
Lakes, R., 1987. Foam structures with a negative Poisson’s ratio. Science 235 (4792), Wu, Q., Chen, Z., Xiong, J., Wang, Z., Yang, C., 2021. Computational studies of porous
1038–1040. head protection structures for human cranium under impact loading. Acta Mech.
Li, T., Liu, F., Wang, L.J., 2020. Enhancing Indentation and Impact Resistance in Auxetic Solida Sin. 34 (4), 477–493.
Composite Materials, vol. 198, 108229. Xiang, Y., Yu, T., Yang, L., 2016. Comparative analysis of energy absorption capacity of
Li, S., Yang, J.-S., Schmidt, R., Wu, L.-Z., Schröder, K.-U., 2021. Compression and polygonal tubes, multi-cell tubes and honeycombs by utilizing key performance
hysteresis responses of multilayer gradient composite lattice sandwich panels. Mar. indicators. Mater. Des. 89, 689–696.
Struct. 75, 102845. Yang, C., Li, Q.M., 2020. Advanced lattice material with high energy absorption based on
Novak, N., Krstulović-Opara, L., Ren, Z., Vesenjak, M., 2020. Compression and shear topology optimisation. Mech. Mater. 148, 103536.
behaviour of graded chiral auxetic structures. Mech. Mater. 148, 103524. Yang, H., Ma, L., 2020. Design and characterization of axisymmetric auxetic
Qi, C., Yang, S., Yang, L.J., Wei, Z.Y., Lu, Z.H., 2013. Blast resistance and multi-objective metamaterials. Compos. Struct. 249, 112560.
optimization of aluminum foam-cored sandwich panels. Compos. Struct. 105, 45–57. Yang, H., Wang, B., Ma, L., 2019a. Designing hierarchical metamaterials by topology
Qi, C., Remennikov, A., Pei, L.-Z., Yang, S., Yu, Z.-H., Ngo, T.D., 2017. Impact and close- analysis with tailored Poisson’s ratio and Young’s modulus. Compos. Struct. 214,
in blast response of auxetic honeycomb-cored sandwich panels: experimental tests 359–378.
and numerical simulations. Compos. Struct. 180, 161–178. Yang, H., Wang, B., Ma, L., 2019b. Mechanical properties of 3D double-U auxetic
Qi, C., Jiang, F., Yu, C., Yang, S., 2019. In-plane crushing response of tetra-chiral structures. Int. J. Solid Struct. 180–181, 13–29.
honeycombs. Int. J. Impact Eng. 130, 247–265. Yang, J.-S., Chen, S.-Y., Li, S., Pang, Y.-Z., Schmidt, R., Schröder, K.-U., et al., 2020a.
Qi, C., Jiang, F., Remennikov, A., Pei, L.-Z., Liu, J., Wang, J.-S., et al., 2020a. Quasi-static Dynamic responses of hybrid lightweight composite sandwich panels with
crushing behavior of novel re-entrant circular auxetic honeycombs. Compos. B Eng. aluminium pyramidal truss cores. J. Sandw. Struct. Mater. 23 (6), 2176–2195.
197, 108117. Yang, J.-S., Zhang, W.-M., Yang, F., Chen, S.-Y., Schmidt, R., Schröder, K.-U., et al.,
Qi, C., Pei, L.Z., Remennikov, A., Yang, S., Liu, J., Wang, J.S., et al., 2020b. Parametric 2020b. Low velocity impact behavior of carbon fibre composite curved corrugated
study and optimization of the protect system containing a re-entrant hexagon cored sandwich shells. Compos. Struct. 238, 112027.
sandwich panel under blast impact. Compos. Struct. 252, 112711. Yang, H., Lei, H., Lu, G., 2021. Crashworthiness of circular fiber reinforced plastic tubes
Qiu, X.M., Zhang, J., Yu, T.X., 2009. Collapse of periodic planar lattices under uniaxial filled with composite skeletons/aluminum foam under drop-weight impact loading.
compression, part II: dynamic crushing based on finite element simulation. Int. J. Thin-Walled Struct. 160, 107380.
Impact Eng. 36 (10), 1231–1241. Zhang, J., Lu, G., You, Z., 2020a. Large deformation and energy absorption of additively
Seetoh, I.P., Liu, X., Markandan, K., Zhen, L., Lai, C.Q., 2021. Strength and energy manufactured auxetic materials and structures: a review. Compos. B Eng. 201,
absorption characteristics of Ti6Al4V auxetic 3D anti-tetrachiral metamaterials. 108340.
Mech. Mater. 156, 103811. Zhang, X., An, C., Shen, Z., Wu, H., Yang, W., Bai, J., 2020b. Dynamic crushing responses
Su, Y., Xu, X., Shi, J., Huang, G., 2021. A 3D mechanism-driven hexagonal metamaterial: of bio-inspired re-entrant auxetic honeycombs under in-plane impact loading. Mater.
evaluation of auxetic behavior. Int. J. Mech. Sci. 209, 106699. Today Commun. 23, 100918.
Wadley, H.N.G., 2006. Multifunctional periodic cellular metals. Philosoph. Trans. Royal Zhao, X., Gao, Q., Wang, L., Yu, Q., Ma, Z.D., 2018. Dynamic crushing of double-arrowed
Soc. A Math. Phys. Eng. Sci. 364 (1838), 31–68. auxetic structure under impact loading. Mater. Des. 160, 527–537.

14