Applied Mathematical Modelling 116 (2023) 547–568

Contents lists available at ScienceDirect

Applied Mathematical Modelling

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

Free vibration behaviours of composite sandwich plates with

reentrant honeycomb cores
Weimin Jiang a, Junmeng Zhou d, Jiayi Liu a,b,∗, Mangong Zhang c, Wei Huang a
a
School of Naval Architecture and Ocean Engineering, Huazhong University of Science and Technology, Wuhan 430074, China
b
Hubei Key Laboratory of Naval Architecture and Ocean Engineering Hydrodynamics (HUST), Wuhan 430074, China
c
Wuhan Second Ship Design and Research Institute, Wuhan 430064, China
d
China Ship Development and Design Center, Wuhan 430064, China

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

Article history: The free vibration behaviours of carbon fiber composite sandwich plates with reentrant
Received 17 August 2022 honeycomb cores were explored by experiment, theory and finite element method (FEM)
Revised 20 November 2022
in present study. The composite sandwich plates with reentrant honeycomb cores were
Accepted 2 December 2022
designed and fabricated through hot-press and secondary forming technology. The modal
Available online 5 December 2022
characteristics of composite auxetic honeycomb sandwich plates were tested under sim-
Keywords: ply supported boundary conditions. Furthermore, the influences of relative densities and
Composite dimensions on the natural frequencies of composite auxetic honeycomb sandwich plates
Natural frequencies were studied. The equivalent moduli of the unit cell of reentrant hexagonal core were de-
Sandwich plates duced. The free vibration control equation was derived to gain the natural frequencies of
Free vibration the sandwich plates by Zig-Zag theory. The theoretical results of the natural frequencies
were obtained and compared with the experimental results and finite element calculation
results. The results showed that the theory and FEM could estimate the natural frequencies
of this kind of sandwich plates accurately. The natural frequencies of the sandwich plates
raised as the increasing of relative densities and influenced by dimensions of the core.
© 2022 Elsevier Inc. All rights reserved.

1. Introduction

Composite sandwich structures have an extensively application in aerospace, aircraft, ocean engineering and others ow-
ing to their lightweight characteristics, high specific stiffness, high specific strength, acoustic insulation and multi-functional
potentials [1–6]. In recent years, a wide range of studies are focused on investigating a new fabrication technology and me-
chanical behaviours of these structures. Lou et al. [7] researched the influences of local damage on the modal characteristics
of composite pyramidal truss core sandwich structures by experiment and FEM. And validated numerical analyses were used
to further research the influence of the extent, location and form of the local damage as well as boundary conditions. Yang
et al. [8] studied modal characteristics of the corrugated sandwich cylindrical plates (CSCPs) with and without foam filled
by experiment and numerical. Compared to the CSCPs without foam filled, the damping of the foam filled CSCPs was clearly
raised, but the natural frequencies had no apparent difference. Tu et al. [9] advanced a rectangular element with nine nodes
each of whose has nine degrees of freedom to analyse the bending and vibration behaviours of laminates and sandwich
plates. Sahoo et al. [10] proposed a novel Zig-Zag theory for analyzing the bulking and free vibration behaviours of lami-

∗
Corresponding author at: School of Naval Architecture and Ocean Engineering, Huazhong University of Science and Technology, Wuhan 430074, China.
E-mail address: [email protected] (J. Liu).

https://doi.org/10.1016/j.apm.2022.12.004
0307-904X/© 2022 Elsevier Inc. All rights reserved.
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

nates and sandwich structures. Mahi et al. [11] advanced a novel hyperbolic shear deformation theory for the researching of
the bending and free vibration behaviours of various plates. Chen et al. [12] studied the free vibration behaviours of com-
posite sandwich plates with different cores through the classic laminated plate theory, first-order shear deformation theory,
Reddy’s third-order shear deformation theory and a Zig-Zag theory.
The above studies are focused on the structures with positive Poison’s ratio. The auxetic material and structure with
negative Poison’s ratio is a novel material and structure which have extraordinary mechanical properties [13–16]. As a re-
sult, auxetic structures suffered widely attention from the scholars [17–23]. Yang et al. [24] used a finite element with two
dimensions to research the Poisson’s ratio of elastic honeycomb plates. They found that plates had a negative Poisson’s ratio
when the re-entrant angle was appropriate, and the Poisson’s ratio of the plates was influenced by dimensions of the cell.
Zhou et al. [25] studied the dynamic responses of metallic auxetic honeycombs with different cell side length-to thickness
ratios and cell configurations. Compared to traditional honeycombs, auxetic honeycombs had higher plateau stress and spe-
cific energy absorption and smaller deformation. Mukhopadhyay and Adhikari [26] deduced the effective in-plane elastic
properties of auxetic honeycombs with spatial irregularity. The results shown that the effective in-plane elastic properties of
auxetic honeycombs were affected by irregularity. Qi et al. [27] studies impact and close-in blast response of sandwich pan-
els with auxetic honeycomb core experimentally and numerically. The anti-impact performance of auxetic honeycomb-cored
sandwich panels were better than those of traditional honeycomb-cored sandwich panels with same dimensions and relative
densities. Ru et al. [28] researched the load-bearing performance of sandwich structures with auxetic layered honeycomb
cores under quasi-static compression experimentally and numerically. The results showed that this kind of sandwich struc-
tures could increase the shear resistance and overall stability of the structures. Singh et al. [29] researched free vibration
behaviors of different honeycomb sandwich plates included square, auxetic, honeycomb and their hybrids numerically. Hy-
brids were consisted of square-auxetic, auxetic-honeycomb and square-honeycomb. The materials of those objects were PLA
(polylactic acid). Nguyen et al. [30] investigated the nonlinear free and forced vibration of sandwich auxetic cylindrical panel
on visco-Pasternak foundations. The effects of dimensions, visco-Pasternak foundations, initial imperfection, temperature in-
crement, nanotube volume fraction and blast load on the vibration behaviours of the auxetic sandwich cylindrical panel.
Chen et al. [31] researched the vibration and damping properties of carbon fiber composite 3D double-arrow-head auxetic
structures. The negative Poisson’s ratios of auxetic structures were studied based on energy method, and modal strain en-
ergy method were applied to study their vibration and damping properties. Effects of inclined corrugated angles and fiber
orientation on natural frequencies, damping loss factors and Young’s module were analyzed and an evaluation parameter
to characterize the structural bearing capacity and vibration damping properties was improved. Dinh et al. [32] studied the
vibration behaviours of auxetic laminated plate with magneto-electro-elastic face sheets subjected to blast loading. The in-
fluence of dimensions, material parameters, elastic foundations, temperature increment, magnetic and electric potentials on
the vibration behaviours of the auxetic laminated plate was analyzed.
The previous works mainly focused on the mechanical properties of plastic, metal and polymer sandwich structures with
negative Poison’s ratio. However, the study on the composite auxetic honeycomb sandwich plates is rarely. In the present
paper, the free vibration of carbon fiber composite auxetic honeycomb sandwich plates were investigated experimentally,
numerically and theoretically. In Section 2, composite honeycomb sandwich plates were designed and fabricated. The tests
were carried out to reveal the effects of relative densities on the natural frequencies of the sandwich plates. In Section 3,
the equivalent moduli of the reentrant honeycomb core were deduced to obtain the natural frequencies of composite sand-
wich plates by a Zig-Zag theory. The equivalent moduli of the unit cell and modal characteristics of composite honeycomb
sandwich plates were studied using FEM in Section 4. In Section 5, the results of the free vibration behaviors of composite
auxetic honeycomb sandwich plates were analyzed and discussed. Finally, the conclusions were drawn in Section 6.

2. Experiment

2.1. Fabrication

Composite auxetic honeycomb sandwich plates with reentrant honeycomb cores were fabricated by the unidirectional
carbon/epoxy reinforced prepregs using hot-press and secondary forming method. The material parameters are referred to
the reference [33].
Composite auxetic sandwich plates were composed of two face sheets and the middle reentrant honeycomb core, as
shown in Fig. 1(a). The core of the composite sandwich plates consisted two rows and nine columns reentrant honeycomb
unit cell whose schematic was presented in Fig. 1(b). The unit cell is determined by the overall height h, thickness t, platform
length b, inclination angle α and extruded for a depth L2 . The relative density of the unit cell is defined by the ratio of the
volume of the members (the actual area of black bars in Fig. 1(b)) to that of the unit cell (the area of the blue rectangle in
Fig. 1(b)), and it is given as:

2t b sin θ

+1
ρ̄ = h
2b sin θ
h
(1)
h
− cos θ

In this paper, the characterizes of the reentrant honeycomb core were h = 16mm, b = 24mm, L = 300mm and θ = 45°.
The thickness t of the composite auxetic honeycomb sandwich plates made of 4 layers, 8 layers and 12 layers prepregs were

548
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 1. (a) the composite auxetic honeycomb sandwich panels, (b) the unit cell of reentrant honeycomb core.

0.4 mm, 0.8 mm, 1.2 mm, respectively. Correspondingly, the relative densities of the reentrant honeycomb cores made of 4
layers, 8 layers and 12 layers prepregs were 7.29%, 14.57%, 21.86%, respectively.
The composite sandwich plates with three different thicknesses were fabricated as follows which could be seen in Fig. 2.
First, the prepregs were cut to the required size and layered as [0°/90°]s , [0°/90°/0°/90°]s , [0°/90°/0°/90°/0°/90°]s for thick-
ness of 0.4 mm, 0.8 mm, 1.2 mm, respectively. Then the layered prepregs were put into the molds as shown in Fig. 2(a). The
prepregs were also put on the upper and lower surfaces of the molds, and the prepregs were used to form the single core
layer of the composite auxetic honeycomb sandwich plates as shown in Fig. 2(b). In order to simplify the fabrication process
and improve the face-core interface strength of composite auxetic sandwich plates, the face sheet was fabricated with the
adjacent single-layer core together. This process was shown in Fig. 2(c). The bolts were used to fasten the molds and the
single-layer core of the reentrant honeycomb sandwich plates. Subsequently, the assembled structure was cured at 125 ◦ C
for 1.5 h, as shown in Fig. 2(d). After curing, the molds and the single-layer core of composite auxetic sandwich plate with
a face sheet were gradually cooled to the room temperature. And then, the molds were removed from the single-layer core,
as shown in Fig. 2(e). The two single-layer cores with a face sheet obtained by the same method were glued together using
epoxy resin adhesive, as shown in Fig. 2(f). Then composite auxetic sandwich plates with reentrant honeycomb core were
fabricated, as shown in Fig. 2(g). Fig. 3 presents the fabricated composite sandwich plates with three relative densities.

2.2. Experimental methods

The modal experiments of sandwich plates with three different thicknesses were carried out using hammer impulse
method to research the natural frequencies and corresponding mode of the sandwich plates under simply supported bound-
ary conditions, as shown in Fig. 4(a, b).
The tested plate was put into the designed fixture to simulate the simply supported boundary condition. The input
excitation which was applied on the tested plate by an impact hammer with a force transducer, and the output response
which was gotten by an acceleration transducer were collected and transferred by a dynamic signal analyser [34]. The multi-
input single-output method was applied to gain the modal characteristics of composite auxetic honeycomb sandwich plates
with reentrant honeycomb cores, as shown in Fig. 4(c). Each input point should be measured more than three times. At
present paper, only six modes were considered in these modal testing.

3. Theoretical investigation

3.1. The equivalent moduli of the unit cell of reentrant honeycomb core

The equivalent moduli of the unit cell of reentrant honeycomb core were deduced referring several textbooks [35–37] and
references [33,38].

549
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 2. The fabrication processes of composite auxetic honeycomb sandwich panel: (a) put prepregs into molds, (b) assemble the molds, (c) put the face
sheet, (d) fasten the molds and cure the assembled structure, (e) remove molds from the single layer plate, (f) glue the two single layer plates, (g) fabricated
sandwich plate.

Fig. 3. The fabricated composite auxetic honeycomb sandwich panels: (a) ρ̄ = 7.29%, (b) ρ̄ = 14.57%, (c) ρ̄ = 21.86%, (d) unit cell with ρ̄ = 14.57%.

3.1.1. Elastic moduli

The deformation of the unit cell with stress σ 1 along x-direction for E1 can be seen in Fig. 5. The deformation vertical
to BC is
 
m2 1
δ1 = NBC m + (2)
12EI 2kGtL2
and the deformation along BC is
FBC m
δ2 = (3)
Es L2 t
where m = h/(2 sin θ ), EI=D11 L2 , k = 0.86, G = 3500 MPa, Es =48,400 MPa [33]. D11 , G and Es were the bending stiffness,
shear module and compressive module of the laminate with [0°/90°]ns , respectively. For laminate, the bending stiffness can

550
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 4. Experimental setup of modal test for composite auxetic sandwich panel under simply supported boundary condition: (a) sketch map of experimental
setup, (b) physical map of experimental setup, (c) arrangement of tested points.

be written as
1    3 
n
Di j = Q̄i j zk − zk3−1 (4)
3 k
k=1

where Qij is the stiffness coefficient of lamina, zk is the coordinate of the kth layer along z-direction, which is shown in
Fig. 5(b). In Fig. 5(d), the deformation along CD is
FCD l
δ3 = (5)
Es L2 t
where l = b/2 + t/sin θ . As shown in Fig. 5(e), the deformation along AB is
FAB b
δ4 = (6)
Es L2 t
The total deformation of D along x-direction relative to A is
δDx = δ1 sin θ + δ2 cos θ + δ3 + δ4 (7)
The total deformation of D along z-direction relative to A is
δDz = δ1 cos θ − δ2 sin θ (8)
The relation of the forces can be written as

FCD = 2FBC cos θ + 2NBC sin θ
FBC sin θ − NBC cos θ = 0 (9)
FAB = FBC cos θ + NBC sin θ

551
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 5. The deformation map of unit cell for E1 : (a) stress diagram of unit cell, (b) sketch map of laminate, (c) deformation of CB, (d) deformation of CD,
(e) deformation of AB.

The stress of the unit cell is

FCD
σ1 = (10)
hL2
The strain of the unit cell is
δDx
εx = (11)
e
δDz
εz = (12)
h/2
where e = b − mcos θ . The equivalent modulus E1 of the unit cell is
σ1 e
E1 = = 
εx h msin2 θ  m2 + 1  +
(13)
mcos2 θ +2b+2l
2 12D11 2kGt 2Es t

The Poisson’s ratio is

 m3 
e sin θ cos θ − 12D11 − 2kGt + Es t
m m
εz
υ13 = − = 2   (14)
εx h sin2 θ m3 + m + mcos2 θ +2l+b
12D 2kGt Et
11 s

552
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 6. The deformation map of unit cell for G13 : (a) force diagram of unit cell, (b) deformation of AB, (c) deformation of BC, (d) deformation of CD.

The equivalent modulus E2 of the unit cell is [38]

E2 = ρ̄ Es (15)

3.1.2. Shear moduli

Fig. 6 shows the deformation of the unit cell with shear force N for G13 . The deformation of AB along z-direction is
3
NAB (b/2 )
δ1 = (16)
3EI
As shown in Fig. 6(c), the deformations of BC along x’-direction and z’-direction are
FBC m
δ2 = (17)
Es tL2

NBC m3
δ3 = (18)
3EI
The rotation of BC is
δ3 NBC m2
ϕ = tan ϕ = = (19)
m 3EI
As shown in Fig. 6(d), the deformation of CD along x-direction caused by shear force is
NCD l 3
δ4 = (20)
3EI
And the deformation caused by rotation is
NBC m2
δϕ = l tan ϕ = l (21)
3EI
The total deformation of D along z-direction relative to A is
δDz = δ1 + δ2 sin θ + δ3 cos θ + δ4 + δϕ (22)

553
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 7. The map of unit cell for shear modules: (a) G12 , (b) G23 .

The relation of the forces can be written as

⎧
⎪N =N
⎨ CD 1
NAB = 2 NCD
(23)
⎪FBC = 12 NCD sin θ
⎩
NBC = 12 NCD cos θ
The shear stress of the unit cell is
N
τ= (24)
L2 h
The shear strain of the unit cell is
δDz
γ= (25)
e
The equivalent shear modulus G13 of the unit cell is
2e
G13 =   (26)
sin θ cos θ +(b/2 ) +2l 3 +m2 l sin θ
3
m3 m sin θ cos θ
h 3D11
+ Es t

Fig. 7(a) shows the map for shear modulus G12 . The shear strain of the unit cell is γ 12 . The strain energy of the equivalent
body is
U = G12 γ12
2
heL2 (27)
The strain energies of AB, BC and CD are

UAB = 14 Gγ12
2
btL2
UBC = 2 Gγ12 cos2 θ mtL2
1 2 (28)
UCD = 12 Gγ12
2
ltL2
The strain energy of the unit cell is equal to that of the equivalent body, so there is
U = 2UAB + 4UBC + 2UCD (29)
The shear modulus G12 can be written as
 
Gt b/2 + l + 2cos2 θ m
G12 = (30)
he
Fig. 7(b) shows the map for shear modulus G23 . Assuming the unit cell has a deformation δ along y-direction under a
shear force N23 . The deformation of the equivalent body is
N23 h
δ= (31)
2G23 eL
The deformation of BC can be written as
N23 m
δ= (32)
GtL
The Eq. (28) is equal to the Eq. (29), so the shear modulus G23 is
Gt sin θ
G23 = (33)
e

554
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 8. The deformation map of unit of the sandwich panel using Zig-Zag theory.

In Gibson’s equations [38], the equivalent moduli of the unit cell of honeycomb core are
t 3
b/m + sin β
E1 = Es (34a)
m cos3 β

E2 = Es ρ̄ (34b)

(b/m + sin β ) sin β

ν13 = (34c)
cos2 β
ν23 = ν21 = νs (34d)

b/m + sin β 1 t

b/m + 2sin β
2
Gs ≤ G12 ≤ Gs (34e)
(1 + 2b/m) cos β 2 m (b/m + sin β ) cos β

cos β
t
G23 = Gs (34f)
b/m + sin β m
t 3
b/m + sin β
G13 = Es (34g)
m (b/m ) (1 + 2b/m) cos β
2

where the angle β is negative for reentrant core and the relation of θ and β is
π
β=θ− (35)
2
In Eq. (34), the orientation and symbols have been translated to the representation of this work.

3.2. Free vibration control equation

In this section, the Zig-Zag theory was employed to deduce the governing equations of free vibration of sandwich plates.
Then, natural frequencies of composite auxetic honeycomb sandwich plates under simply supported could be obtained. In
the Zig-Zag theory, the deformation of the core obeyed the first-order shear deformation theory, but the deformation of face
sheets obeyed the classic laminated plate theory as shown in Fig. 8.
The constitutive equations for the kth of an orthotropic laminate are [37]
⎧ ⎫k ⎡ ⎤k ⎧ε ⎫k
⎪σ1 ⎪ Q11 Q12 ⎪
⎪ ⎪
1
⎪
⎨σ2 ⎪ ⎬ ⎢Q21 ⎪
Q22 ⎥ ⎨ε2 ⎬
τ 23 = ⎢ Q44 ⎥ γ 23 (36)
⎪
⎪ ⎪ ⎣ ⎦⎪ ⎪
⎩τ 31 ⎪
⎭ Q55 ⎪
⎩γ 31 ⎪
⎭
τ 12 Q66 γ 12
555
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

k = E1k k = E2k k = ν21

k Ek ν12
k Ek
k = Gk ,Q k = Gk , Q k = Gk . When the orientation of the
where Q11 , Q22 , Q12 1
= 2
, Q44
1−ν12
k νk
21
1−ν12
k νk
21
1−ν12
k νk
21
1−ν12
k νk
21
23 55 13 66 12
kth layer is α , the transformed stress-strain relations of an orthotropic lamina in a panel state of stress are as follows:
⎧ ⎫k ⎡Q̄ ⎤k ⎧ ⎫k
⎪σx ⎪ 11 Q̄12
⎪εx ⎪
⎪
⎨σy ⎪ ⎬ ⎢Q̄21 Q̄22 ⎥⎪ ⎨εy ⎪ ⎬
⎢ ⎥
τ yz = ⎢ Q̄44 ⎥ γ yz (37)
⎪
⎪ ⎪ ⎣ ⎦⎪ ⎪
⎩τ zx ⎪
⎭ Q̄55 ⎪
⎩γ zx ⎪
⎭
τ xy Q̄66 γ xy
where
Q̄11 = Q11 cos4 α + 2(Q12 + Q66 )sin α cos2 α + Q22 sin α
2 4

Q̄12 = (Q11 + Q22 − 4Q66 )sin α cos2 α + Q12 sin α + cos4 α
2 4

Q̄22 = Q11 sin α + 2(Q12 + Q66 )sin α cos2 α + Q22 cos4 α

4 2
(38)
Q̄44 = Q44 cos α + Q55 sin α
2 2

Q̄55 = Q44 sin α + Q55 cos2 α

2
 4 
Q̄66 = (Q11 + Q22 − 2Q12 − 2Q66 )sin α cos2 α + Q66 sin α + cos4 α
2

The displacement field can be written in forms of the middle plane displacement w0 , rotation angles θ x and θ y . The
rotations angles θ x and θ y are real rotation angles normal to the middle plane with respect to x and y axes, respectively.
The displacement field can be written as follows [12]
⎧ ∂w  
⎨−z ∂ x0 + h2c θx − ∂∂wx0 − H2 ≤ z ≤ − h2c
u = −zθx − h2c ≤ z ≤ h2c (39a)
⎩ ∂ w0 hc  
−z ∂ x − 2 θx − ∂∂wx0 h2c ≤ z ≤ H2
⎧ ∂w  
⎨−z ∂ y0 + h2c θy − ∂∂wy0 − H2 ≤ z ≤ − h2c
v = −zθy − h2c ≤ z ≤ h2c  (39b)
⎩ ∂ w0 hc
−z ∂ y − 2 θy − ∂∂wy0 h2c ≤ z ≤ H2

w = w0 (39c)
where u, v, w, w0 ,θ x ,θ y are the function of coordinates(x, y, z) and time T, H is the height of the sandwich plates including
the height of the core hc and the thicknesses of face sheets 2t
The plane-strain can be written as:
⎧ 
⎪
⎪−z ∂ 2 w0
+ hc ∂θx ∂ 2 w0 H
≤ z ≤ − h2c
⎨ ∂x 2 2 ∂x − ∂ x2 − 2
∂u
εx = = −z ∂θ hc
∂x − 2 ≤
x hc
z ≤ 2 2 (40a)
∂x ⎪⎪
⎩−z 20 − 2c ∂ xx − ∂ w20
∂ 2
w h ∂θ hc
≤z≤ H
∂x ∂x 2 2
⎧ 
⎪ ∂θ
−z ∂∂ yw20 + h2c ∂ yy − ∂ 2 w0
2
⎪ − H
≤ z ≤ − h2c
⎨ ∂ y2 2
∂v
εy = = −z ∂θ hc
∂y − 2 ≤
hc
z ≤ 2 2
y
(40b)
∂y ⎪⎪
⎩−z ∂ w20 − hc
2 ∂θ y
− ∂ w20 hc
≤z≤ H
∂y 2 ∂y ∂y 2 2

∂ w0
εz = =0 (40c)
∂x
⎧ 
⎪
⎪−2 ∂ 2 w0
+ hc ∂θx
+
∂θy ∂ 2 w0 H
≤ z ≤ − h2c
⎪ z ∂ x − 2 ∂ x∂ y −
⎨  ∂ x∂ y 2 ∂ y 2
∂ u ∂v ∂θy
γxy = + = −z ∂θ hc
∂y + ∂x − 2 ≤ z ≤ 2
x hc
(40d)
∂y ∂x ⎪ ⎪ 
⎪
⎩−2z ∂ 2 w0 − hc ∂θx + ∂θy − 2 ∂ 2 w0 hc H
∂ x∂ y 2 ∂y ∂x ∂ x∂ y 2
≤z≤ 2
⎧
⎨0 − H
≤ z ≤ − hc
∂w ∂u 2 2
γxz = + = ∂∂wx0 − θx − h2c ≤ z ≤ hc
(40e)
∂ x ∂ z ⎩ hc H
2
02 ≤z≤ 2
⎧
⎨0 − H
≤ z ≤ − hc
∂ w ∂v 2 2
γyz = + = ∂∂wy0 − θy − h2c ≤ z ≤ hc
(40f)
∂ y ∂ z ⎩ hc H
2
02 ≤z≤ 2

556
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

The governing equations were derived by hamilton principle [12].

 t2
δ (U + V − T )dt = 0 (41)
t1

where δ is the sign of variation, U is the strain energy of structure, V is the potential energy caused by external force, T is
the kinetic energy of structure. In free vibration, the potential energy caused by external force is zero. The strain energy and
kinetic energy of sandwich plate are

1
U= (σx εx + σy εy + τxy γxy + τzy γzy + τxz γxz )dV (42)
2


  
1
T = ρ (u˙ )2 + (v˙ )2 + (w˙ )2 dV (43)
2


Take Eqs. (37), (39), (40) into Eqs. (42), (43), the variation of strain energy and kinetic energy are
⎧⎧  2 ∂ 4 w 0 ⎫ ⎫
⎪
⎪⎪
⎪ 2Dt11 + 2Bt11 hc + 2At11 h2c 4 +
⎪
⎪ ⎪
⎪
⎪
⎪⎪
⎪ ∂ x ⎪
⎪ ⎪
⎪
⎪
⎪⎪
⎪   ⎪
⎪ ⎪
⎪
⎪⎪
⎪
⎪⎪ 2 D t
+ 2 B t
h + 2 h c
2
A t ∂ 4
w 0
+ ⎪
⎪ ⎪
⎪
⎪
⎪⎪ 
22 22 c 2 22 ∂y 4
 ⎪
⎪ ⎪
⎪
⎪
⎪⎪
⎪ ∂ 4 ⎪
⎪ ⎪
⎪
⎪
⎪⎪
⎪ 4
D t
+ 4 B t
h c + h 2
A
c 12 
t
+ 8 D t
+ 8 B t
h c + 2 h 2
A t
c 66 ∂ x ∂ y ⎪
w 0
⎪ ⎪
⎪
⎨ 12 12 66 66
⎬
2 2
⎪
⎪ − h Bt + hc At ∂ θx − Bt h + hc At ∂ θy ⎪
⎪
⎪
⎪
2 3 2 3

δ ⎪
⎪
⎪
⎪⎪ 
c 11 2 11 ∂ x 3 22 c 2 22 ∂ y 3
⎪
w 0 ⎪
⎪
⎪
⎪⎪
⎪ ⎪
⎪ ⎪
⎪
⎪
⎪⎪
⎪ − h B t
+
h 2
c
A t
+ 2 B t
h + h 2 t
A ∂ θx
3
⎪
⎪ ⎪
⎪
⎪
⎪⎪
⎪ 
c 12 2 12 66 c c 66 ∂ x ∂ y 2
⎪
⎪ ⎪
⎪
⎪
⎪⎪
⎪ ⎪
⎪ ⎪
⎪
 ⎪ ⎪
2 ∂ 3
θ
⎨⎪ ⎪
h
⎪
⎪
− h c B t
+ c
A t
+ 2 B t
h c + h 2
A t
c 66 ∂ x ∂ y
y
⎪
⎪ ⎬
 12 2 12 66

2

δU =
1 ⎪
⎪ ⎪
⎪ dxdy (44a)
2 ⎪⎩+Ac ∂ w20 − y + Ac ∂ w20 − ∂θx
2 ∂θ 2
⎭ ⎪
⎪ ⎧
⎪ 44
 ∂y ∂y 55 ∂x ∂x
⎫ ⎪ ⎪
⎪
⎪
⎪   ∂ 3 w0   3 ⎪
⎪
⎪ ⎪
⎨ hc −2B11 − hc A11 ∂ x3 + −2
t t
Bt12 − hc At12 − 4Bt66 − 2At66 hc ∂∂x∂wy02 ⎪ ⎬ ⎪ ⎪
⎪
⎪   ⎪
⎪
⎪
⎪+ 2 t ∂ θx 2
t ∂ θx 2
+hc A11 ∂ x2 + hc A66 ∂ y2 + hc A12 + hc A66 ∂ x∂ y t t ∂ 2
θ y
δθ x⎪
⎪
⎪
⎪ ⎪   ⎪ ⎪
⎪
⎪ ⎩
⎪ +Dc 2x + Dc 2x + Dc + Dc
∂ θ ∂ θ ∂ 2
θ ∂ ⎭ ⎪
⎪
⎪ θ ⎪
2 2 w
⎪
y
− A c 0
+ A c
⎪
⎪
⎪ ⎧  
11 ∂ x 66 ∂ y 12 66 ∂ x ∂
 y 55
∂ x 55 x
  ⎫ ⎪
⎪
⎪
⎪ ⎪
⎪ −2 B t
− h A t
− 4 B t
− 2 A t
h ∂ 3
w
+ −2 Bt
− h At ∂ 3
w ⎪
⎪ ⎪
⎪
⎪ ⎪
0 0

⎪
⎪ ⎨ h c
 12 c 12
 66 66 c ∂ x 2∂y 22 c 22 ∂ y3 ⎬ ⎪
⎪
⎪
⎪+
2
+ h A t
+ h A t ∂ θx
2
+ h A t ∂ θy
2
+ h A t ∂ θy
2
δθ ⎪
⎪
⎪
⎪
c
 c ∂ ∂ c ∂y c ∂x ⎪
⎪ ⎪
12 66 x y 22 2 66 2 y
⎪
⎩ ⎩⎪
⎪   ⎪ ⎪
⎭
∂ 2 θx c ∂ θy
+ D12 + D66 ∂ x∂ y + D22 ∂ y2 + D66 ∂ x2 − A44 ∂ y + A44 θy
c c
2
c ∂ θy
2
c ∂ w0 c ⎭

⎡⎡  ⎤ ⎤
h2c ∂ 2 ẅ0 h2c ∂ θ̈x
2I2t + 2I1t hc + I0t ∂ x2 − I1t hc + I0t ∂x +
⎢⎣ 2
 2
 t c  ⎦δ w 0 ⎥
⎢ ∂ θ̈y ⎥
 ⎢ 2I2t + 2I1t hc + I0t 2c ∂∂ yẅ20 − I1t hc + I0t 2c
h2 2 h2
1 ∂y + 2I0 + I0 ẅ0 ⎥
δT = ⎢     ⎥dxdy (44b)
⎢ ⎥
2 ⎢+ − I1t hc + I0t 2c ∂∂ẅx0 + I0t 2c + I2c θ̈x δθx
h2 h2
⎥
⎣    2  ⎦
+ − I1t hc + I0t 2c ∂∂ẅy0 + I0t 2c + I2c θ̈y δθy
h2 h

Take Eq. (44) into Eq. (41), the governing equation of motion for sandwich plates which derived by using the Zig-Zag
theory can be written as
 
h2c t ∂ 4 w0 h2c t ∂ 4 w0
2Dt11 + 2Bt11 hc + A
2 11 ∂ x4 + 2Dt22 + 2Bt22 hc + A
2 22 ∂ y4 +
 
4Dt12 + 4Bt12 hc + h2c At12 + 8Dt66 + 8Bt66 hc + 2h2c At66 ∂∂x2w
4
0
  ∂ y2
h 2
∂ 3
θ h 2 ∂ 3
θ
− hc Bt11 + 2c At11 ∂ x3x − Bt22 hc + 2c At22 ∂ y3 y

 
∂θ
+ 2Bt66 hc + h2c At66 ∂∂x∂θyx2 + Ac44 ∂∂ yw20 − ∂ yy
h2c t 3 2
− hc Bt12 + A (45a)
 2 12

h2c t ∂3θ ∂ 2 w0 ∂θx
− hc Bt12 + A + 2Bt66 hc + h2c At66 ∂ x2 ∂yy + Ac55 ∂ x2 − ∂ x =
 2 12

h2c ∂ ẅ0
2 h2c ∂ θ̈x
2I2t + 2I1t hc + I0t ∂ x2 − It hc + I0t ∂x +
 2
1 2
 t c
h2c ∂ 2 ẅ0 h2c ∂ θ̈y
2I2t + 2I1t hc + I0t 2 ∂ y2 − I1t hc + I0t 2 ∂ y + 2I0 + I0 ẅ0

557
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

  3   3 
−2Bt11 − hc At11 ∂∂ xw30 + −2Bt12 − hc At12 − 4Bt66 − 2At66 hc ∂∂x∂wy02
hc
  ∂2θ
+hc At11 ∂∂ xθ2x + hc At66 ∂∂ yθ2x + hc At12 + hc At66 ∂ x∂ yy
2 2 2

  ∂2θ (45b)
+Dc11 ∂∂ xθ2x + Dc66 ∂∂ yθ2x + Dc12 + Dc66 ∂ x∂ yy − KAc55 ∂∂wx0 + KAc55 θx
2 2

 
h2c ∂ ẅ0 h2c
= − I1t hc + I0t 2 ∂x + I0t 2
+ I2c θ̈x
   ∂ 3 w0
 3 
−2Bt12 − hc At12 − 4Bt66 − 2At66 hc ∂ x2 ∂ y + −2Bt22 − hc At22 ∂∂ yw30
hc
 t  ∂2θ ∂2θ
+ hc A12 + hc At66 ∂∂x∂θxy + hc At22 ∂ y2y + hc At66 ∂ x2y
2 2

   ∂ 2 θx ∂2θ ∂2θ (45c)

+ Dc12 + Dc66 ∂ x∂ y + Dc22 ∂ y2y + Dc66 ∂ x2y − KAc44 ∂∂wy0 + KAc44 θy
 
h2c ∂ ẅ0 h2c
= − I1t hc + I0t 2 ∂y + I0t 2
+ I2c θ̈y
where K is the shear correction coefficient. The stiffnesses and the mass moments of intera can be give as
⎧   − hc  
⎪
⎪ Ati j , Bti j , Dti j = h 2 Q̄itj 1, z, z2 dz, (i, j = 1, 2, 4, 5, 6)
⎪
⎪  0
−
 
⎪
⎪
2

⎨ Aci j , Bci j , Dci j = − h2c Q̄icj 1, z, z2 dz, (i, j = 1, 2, 4, 5, 6)

N 
 (46)
⎪
⎪Irt =
zk+1
ρ k zr dz, (r = 0, 1, 2)
⎪
⎪
zk
⎪ k=1
⎩ c  2c
⎪ h
Ir =
− h2c
ρ c zr dz, (r = 0, 2)
where the subscripts t and c represent the top face sheet and core, respectively.
The simply boundary conditions can be expressed as [37]
∂ w0 ∂ w0
x = 0, L1 : w0 = 0, Mx = B11 ∂θ
2 2

∂ x − D11 ∂ x2 − D12 ∂ y2 = 0
x

∂θy ∂ w0
θy = 0, Nx = A11 ∂θ
2

∂ x + A12 ∂ y − B11 ∂ x2 = 0
x

∂θ (47)
y = 0, L2 : w0 = 0, My = −B11 ∂ yy − D12 ∂∂ xw20 − D22 ∂∂ yw20 = 0
2 2

∂θy ∂ 2 w0
θx = 0, Ny = A12 ∂θ
∂ x + A11 ∂ y + B11 ∂ y2 = 0
x

The solution of the theoretical formulations satisfying the simply supported boundary conditions can be selected as
⎧ ∞  ∞
⎪
⎪θx = xmn (T ) cos mLπ1 x sin nLπ2y
⎪
⎪
⎨ n=1 m=1
∞  ∞
θy = ymn (T ) sin mLπ1 x cos nLπ2y (48)
⎪
⎪ n =1 m =1
⎪
⎪ 
∞  ∞
⎩w0 = wmn (T ) sin mLπ1 x sin nLπ2y
n=1 m=1
where m and n are the half wave number of the vibration mode along the x and y axes, respectively. The mode of composite
sandwich plate is represented as (m, n). In Eq. (48), the functions of time T can be written as
⎧
⎨xmn (T ) = θ̄x eiωT
y (T ) = θ̄ eiωT (49)
⎩wmn T = w̄y eiωT
mn ( ) 0

where θ̄x , θ̄y and w̄0 were coefficients, i is an imaginary number and ω is the natural frequency of sandwich plate at mode
(m, n). Substituting Eq. (48) into Eq. (45), there will be a matrix
 
[K ] − ω2 [M] {δ} = {0} (50)
where [K] is the structural stiffness matrix, [M] is the mass matrix, {δ } is the displacement. In free vibration, the force vector
is zero. In order to ensure the Eq. (50) solvable, the determinant of linear stiffness matrix is zero.
[K ] − ω 2 [M ] = 0 (51)
The natural frequency ω can be solved through Eq. (51), the smallest one is selected to consider [39].

4. Finite element analysis

4.1. The equivalent moduli of the unit cell

Finite element analysis for the material properties of unit cell were performed using ABAQUS/Standard version 6.16. Fig. 9
presents the FEM model of the structure for calculating the moduli of the unit cell. The model was established with 9 × 9
unit cells, and the extruded length was 300 mm. Table 1 lists the boundary conditions of the structure for calculating the
equivalent moduli.

558
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 9. FEM model for calculating the modules of the unit cell.

Table 1
The boundary conditions of structure for calculating the modules of the unit cell.

A B C D E F

U1 = U2 = 0 U1 = 0.1, U2 = 0
E1 – – coupling U3 coupling U3
U R1 = U R2 = U R3 = 0 U R1 = U R2 = U R3 = 0
U1 = U2 = U3 = 0 U2 = 0.1, U1 = U3 = 0
E2 – – – –
U R1 = U R2 = U R3 = 0 U R1 = U R2 = U R3 = 0
U1 = U2 = U3 = 0 U1 = 0.1, U2 = U3 = 0
G12 – – – –
U R1 = U R2 = U R3 = 0 U R1 = U R2 = U R3 = 0
U1 = U2 = U3 = 0 U1 = 0.1, U2 = U3 = 0
G13 U3 = 0 U3 = 0 – –
U R1 = U R2 = U R3 = 0 U R1 = U R2 = U R3 = 0
U1 = U2 = U3 = 0 U3 = 0.1, U1 = U2 = 0
G23 – – – –
U R1 = U R2 = U R3 = 0 U R1 = U R2 = U R3 = 0

Fig. 10. FEM models of composite auxetic honeycomb sandwich panels with different relative densities.

4.2. Free vibration of composite sandwich plates

The two face sheets and the reentrant honeycomb cores were modelled as solid bodies and meshed as C3D8R elements. It
is assumed that face sheets and the reentrant honeycomb cores were perfectly combined in the calculation. Fig. 10 presents
the models of composite sandwich plates with varied relative densities. A grid convergence analysis is carried out to guar-
antee that the selected mesh schemes was reasonable, as shown in Fig. 11.
The natural frequencies and mode shapes of sandwich plates can be finally calculated by a linear perturbation analysis.
Also, the influences of dimensions of unit cells on the modal characteristics had been studied systematically. In order to
ensure that the mass of sandwich structures are similar, the relative densities of unit cells with different dimensions are

559
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 11. Mesh convergence analysis of composite auxetic honeycomb sandwich panel (ρ̄ = 7.29%).

Table 2
The modulus of composite auxetic unit cell (MPa).

ρ̄ = 7.29% ρ̄ = 14.57% ρ̄ = 21.86%

FEM Theory Error (%) Gibson Error (%) FEM Theory Error (%) Gibson Error (%) FEM Theory Error (%) Gibson Error (%)

E1 12.09 13.49 11.57 8.56 −29.25 80.64 94.10 16.69 68.45 −15.12 204.36 224.36 9.79 231.01 13.04
E2 3488.40 3528.36 1.15 3528.36 1.15 6861.58 7051.88 2.77 7051.88 2.77 10,120.01 10,580.24 4.55 10,580.24 4.55
G12 277.96 196.22 −29.41 193.12 −30.52 546.58 398.62 −27.07 386.24 −29.33 805.84 607.21 −24.65 579.37 −28.10
G13 0.35 0.33 −4.71 0.18 −48.40 2.27 2.38 4.85 1.45 −36.03 5.89 5.73 −2.66 4.90 −16.87
G23 59.53 61.87 3.93 61.87 3.93 125.89 123.74 −1.70 123.74 −1.70 195.52 185.62 −5.07 185.62 −5.07

same. Thus, when the angle of the unit cell is changed, the thickness, platform length or height is changed. The FEM models
of different dimensions of the unit cell with same relative density are presented in Fig. 12.

5. Results and discussion

5.1. The equivalent moduli of the unit cell

The equivalent moduli of the unit cell with three different relative densities calculated by theory and Gibson’s equations
[38] are listed in Table 2 and compared with FEM results. The moduli E2 and G23 calculated by theory in this paper were
same with Gibson’s results. The absolute errors of E2 and G23 between theoretical results and FEM solutions reached the
maximum value of 4.55% and 5.07%, respectively, when the relative density of the unit cell was 21.86%.
The moduli E1 and G13 calculated by theory had great differences with Gibson’s results. In Gibson’s equation, the bend-
ing stiffness of metal was expressed by the product of young’s modulus and moment of inertia of section. However, the
bending stiffness per unit length of laminate was expressed as Eq. (4). The values of bending stiffness per unit length of
laminate with [0°/90°]s calculated by D11 L2 and Es I were 421.81 and 258.13 (N·mm2 ), respectively. Thus, the moduli E1 and
G13 calculated in this paper differed from Gibson’s results. The absolute errors of E1 and G13 between theoretical results and
FEM results had maximum values of 16.69% and 4.85% respectively when the relative density of the unit cell was 14.57%.
The difference of G12 between the values calculated by theory and Gibson’s results was mainly caused by the length of
CD. The length CD was b/2 in Gibson’s equation, but it was b/2 + t/sin θ in this paper. It was because that there were three
parts: CB, CD and CE at point C, and the thickness t of BC was taken into consideration as a part of the length of CD. The
maximum absolute error of G12 between theoretical results and FEM solutions appeared with a value of 29.41% when the
relative density was 7.29%.
The moduli E1 , E2 , G12 , G13 and G23 increased as the thickness increased, and the values obtained by theory were similar
to FEM results.

560
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 12. FEM models of sandwich panels with different dimensions and same relative densityρ̄ = 7.29%: (a) angle θ and thickness t, (b) angle θ and
platform length b, (c) angle θ and height h.

Table 3
Poisson’s ratio of the unit cell with different dimensions.

t b h ρ̄ ν 13
(mm) (mm) (mm) θ (%)
theory Gibson FEM

Relative 0.40 24.00 16.00 45° 7.29 −1.96 −2 −1.94

density 0.80 24.00 16.00 45° 14.57 −1.85 −2 −1.84
(change t) 1.20 24.00 16.00 45° 21.86 −1.76 −2 −1.72
Change t and θ 0.53 24.00 16.00 60° 7.29 −1.24 −1.40 –
0.63 24.00 16.00 75° 7.29 −0.62 −0.73 –
Change b and 0.40 16.68 16.00 60° 7.29 −0.85 −0.87 –
θ 0.40 11.92 16.00 75° 7.29 −0.32 −0.33 –
Change h and 0.40 24.00 8.70 60° 7.29 −2.51 −2.85 –
θ 0.40 24.00 7.62 75° 7.29 −1.33 −1.62 –

5.2. Poisson’s ratio

As known, the reentrant core is attractive because it has a negative Poisson’s ratio. So, the Poisson’s ratios of reentrant
core of present composite sandwich plates were discussed in this section. The Poisson’s ratios of the reentrant core with
different dimensions are listed in Table 3. Since the Poisson’s ratios are negative, absolute values of Poisson’s ratios were
discussed. When relative density of the reentrant core increased, Poisson’s ratio ν 13 obtained by theory and FEM decreased.
And the results calculated by theory was close to the FEM results. In Eq. (34c), the Poisson’s ratio is relative with platform

561
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 13. The six typical mode shapes of composite auxetic honeycomb sandwich panels under simply supported boundary condition (ρ̄ = 14.57%).

length b, angle θ and height h. The influence of thickness t on the Poisson’s ratio of reentrant core has been ignored.
However, relative density of reentrant core varies with thickness in present work. So, the Poisson’s ratios calculated by
Gibson’s equation were consistent when relative densities varied. As thickness increased and angle increased, the Poisson’s
ratio of reentrant core with same relative density decreased. Similar, the Poisson’s ratio of reentrant core with same relative
density decreased as platform length decreased and angle increased. There had no obvious regularity on the Poisson’s ratio
of reentrant core with same relative density when height and angle differed.

5.3. Modal characteristic of sandwich plates with different relative densities

Six typical mode shapes including (1,1), (2,1), (3,1), (4,1), (2,2) and (2,3) were selected to discuss. Natural frequencies of
sandwich plates with different relative densities were compared at same mode shape. And the natural frequencies between
the experimental results, numerical results and theoretical results of sandwich plates with same relative density at same
mode shape were analysed.

5.3.1. Mode shapes of sandwich plates

The six typical mode shapes of composite auxetic sandwich plates with relative density ρ̄ = 14.57% under simply sup-
ported boundary conditions obtained from FEM were compared to the experimental ones in Fig. 13. Mode shapes of com-
posite sandwich plates with three relative densities under simply supported boundary conditions were similar. The mode
shapes of the sandwich plates under simply supported boundary condition were bending mode. The half wave number of
the Y axe was usually no more than that of the X axe. It was because that reentrant honeycomb cores of sandwich plates
were arrayed in the x-direction. So the bending stiffness of sandwich plates in the x-direction was smaller than that in the
y-direction.

5.3.2. Natural frequencies of sandwich plates

The corresponding natural frequencies of composite auxetic honeycomb sandwich plates with three relative densities
obtained from test, theory and FEM were shown in Figs. 14–16. The first order frequencies of sandwich plates with relative
densities of 7.29%, 14.57% and 21.86% were 943.5 Hz, 1024.8 Hz and 1095 Hz respectively in the test. The natural frequencies
of composite sandwich plates at same mode shape increased as the relative density increased. It was obvious that the

562
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 14. Frequencies of composite auxetic honeycomb sandwich panel (ρ̄ = 7.29%).

Fig. 15. Frequencies of composite auxetic honeycomb sandwich panel (ρ̄ = 14.57%).

stiffness and strength of sandwich plates with higher relative density were higher. When the relative density of sandwich
plate was 7.29%, the maximum absolute error of natural frequencies between the results calculated by test and theory was
appeared at mode shape (2,2) with a value of 8.46%, while the maximum absolute error between experimental results and
numerical results was appeared at mode shape (4,1) with a value of 3.65%.
For sandwich plates with relative density of 14.57%, the maximum absolute errors between experimental results and
theoretical results, experimental results and numerical results were appeared at mode shape (4,1) with value of 9.44% and
mode shape (3,2) with value of 7.49%, respectively.
The values of the maximum absolute errors between results obtained by test and theory, test and FEM were 17.68% and
10.41%, respectively. They appeared at mode shape (1,1) and (4,1), respectively, when the relative density of sandwich plate
was 21.86%.
The errors between results obtained by test, theory and FEM mainly caused by the following reasons. First, the dimen-
sions of fabricated composite sandwich plates had errors with those of FEM and theory. And there were some unavoidable

563
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 16. Frequencies of composite auxetic honeycomb sandwich panel (ρ̄ = 21.86%).

defects of sandwich plates during the fabrication processes. Second, the moduli of unit cell had errors between the results of
theory and FEM, as discussed in Section 5.1. Moreover, the supposed displacement field in Zig-Zag theory could not reflect
the real displacement of sandwich plates accurately. Although there had errors between the results of test, theory and FEM,
the results calculated by this theory and FEM were both credible. So the theory and FEM were reasonable to further study
the effects of dimensions on natural frequencies of sandwich plates.

5.4. Dimensions analyses

To explore the effects of dimensions of the unit cell on modal characteristic of composite auxetic honeycomb sandwich
plates, natural frequencies of three groups of sandwich plates with different dimensions but same relative density had been
compared theoretically and numerically. The geometrical characteristic of the unit cell was determined by the thickness,
height, platform length and angle. In order to keep the relative density of the unit cell unchanged, other one must be
changed when one of the dimensions was changed. So, when the angle of the unit cell was changed, the thickness, height
or platform length was also changed to ensure that the relative density was consistent. The dimensions of these three groups
of sandwich plates are listed in Table 3. The relative density of these sandwich plates were 7.29%.
Fig. 17 gives the natural frequencies of sandwich plates with varied angle and thickness but same relative density. The
results showed that the frequencies at same mode shape increased as the angle and thickness raised. And the results calcu-
lated by theory were relatively consistent with those obtained by FEM. The maximum absolute errors between theoretical
results and numerical results were 9.39%, 20.39% and 16.51% at mode shape (1,1), (4,1) and (1,1) when the angles of unit cell
were 45°, 60° and 75° respectively. The first order frequencies obtained by FEM for sandwich plates with angles of 60° and
75° were 988.9 Hz and 1034.7 Hz respectively.
Fig. 18 presents the natural frequencies of sandwich plates with different angle and platform length but same relative
density. It was obviously that the frequencies at the same mode shape increased as the angle increased and the platform
length decreased. The values calculated by theory were similar to those obtained by FEM, and the absolute errors between
theoretical results and numerical results were no more than 22.87%. The first order frequencies of sandwich plates with
angles of 60° and 75° were 1101.2 Hz and 1210.2 Hz respectively by FEM.
Fig. 19 displays the natural frequencies of sandwich plates with varied angle and height but same relative density. As
the angle increased and height decreased, the natural frequencies of sandwich plates at the same mode shape were similar.
The absolute errors between the FEM results and theoretical results were less than 16.92%. The first order frequencies of
sandwich plates with angles of 60° and 75° were 912.4 Hz and 899.2 Hz respectively by FEM. When the angle and relative
density of unit cell were same, the natural frequencies of sandwich plates at same mode shape with varied platform length
were maximum compared to those of sandwich plates with varied thickness and height. The natural frequencies of sandwich
plates with same angle and relative density but varied height were minimum at same mode shape.

564
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 17. Frequencies of composite auxetic honeycomb sandwich panels with different thickness and angle (ρ̄ = 7.29%).

Fig. 18. Frequencies of composite auxetic honeycomb sandwich panel with different platform length and angle (ρ̄ = 7.29%).

5.5. Boundary conditions

The effects of boundary conditions on free vibration behaviours of composite sandwich plates with different dimensions
were researched numerically. Four boundary conditions included CCCC, CCSS, CSCS and SCSC were taken into consideration,
shown in Fig. 20. Symbol C represents clamp boundary condition and symbol S represents simply boundary condition.
Natural frequencies of composite sandwich plates with different dimensions under different boundary conditions were
listed in Tables 4–7. As relative density of reentrant cores increased, the natural frequency of composite sandwich plates
under same boundary conditions at same mode increased. Similar with SSSS boundary conditions, natural frequency of
composite sandwich plates with same relative density increased as angle increased and thickness decreased under CCCC,
CCSS, CSCS and SCSC boundary conditions. Also, when relative density was same, natural frequency of composite sandwich
plates with larger angle and shorter platform length was higher under same boundary conditions. And natural frequencies

565
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Fig. 19. Frequencies of composite auxetic honeycomb sandwich panel with different height and angle (ρ̄ = 7.29%).

Fig. 20. Four different boundary conditions of composite sandwich plates.

Table 4
Natural frequencies of composite sandwich plates with different dimensions under CCCC boundary conditions (Hz).

Mode Relative density ρ̄ Angle and thickness Angle and platform length Angle and height
shape
7.29% 14.57% 21.86% 60°/0.53 mm 75°/0.63 mm 60°/16.68 mm 75°/11.92 mm 60°/8.70 mm 75°/7.62 mm

(1,1) 1039.9 1096.3 1156.4 1119.0 1169.2 1272.4 1437.0 1030.5 1006.6
(2,1) 1079.1 1165.3 1268.9 1186.0 1242.0 1313.8 1487.7 1082.7 1063.6
(3,1) 1145.1 1282.5 1458.3 1299.3 1362.6 1383.1 1567.0 1171.9 1164.2
(4,1) 1236.8 1445.6 1716.2 1447.4 1528.1 1478.5 1675.0 1291.1 1296.8
(2,2) 2057.7 2233.3 2380.7 2284.7 2394.9 2588.6 2980.7 2040.4 2014.0
(3,2) 2144.3 2357.8 2549.7 2425.8 2532.2 2676.4 3068.8 2146.8 2138.1

Table 5
Natural frequencies of composite sandwich plates with different dimensions under CCSS boundary conditions (Hz).

Mode Relative density ρ̄ Angle and thickness Angle and platform length Angle and height
shape
7.29% 14.57% 21.86% 60°/0.53 mm 75°/0.63 mm 60°/16.68 mm 75°/11.92 mm 60°/8.70 mm 75°/7.62 mm

(1,1) 1002.9 1043.5 1089.0 1080.5 1128.2 1223.5 1373.2 990.4 974.6
(2,1) 1036.1 1110.4 1198.2 1146.1 1201.5 1264.6 1426.5 1037.5 1030.7
(3,1) 1100.7 1225.6 1386.4 1257.4 1322.6 1333.6 1508.0 1122.4 1130.9
(4,1) 1185.6 1384.1 1638.6 1401.5 1486.7 1427.5 1618.2 1236.1 1261.8
(2,2) 1977.1 2146.0 2279.6 2225.2 2343.9 2516.7 2907.4 1947.4 1964.3
(3,2) 2048.2 2263.1 2441.1 2362.0 2478.2 2599.2 2993.1 2047.2 2088.4

566
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

Table 6
Natural frequencies of composite sandwich plates with different dimensions under CSCS boundary conditions (Hz).

Mode Relative density ρ̄ Angle and thickness Angle and platform length Angle and height
shape
7.29% 14.57% 21.86% 60°/0.53 mm 75°/0.63 mm 60°/16.68 mm 75°/11.92 mm 60°/8.70 mm 75°/7.62 mm

(1,1) 1039.1 1094.1 1151.8 1116.9 1167.4 1271.3 1435.8 1029.3 1005.2
(2,1) 1075.0 1153.6 1244.9 1177.0 1233.5 1308.7 1482.3 1076.7 1057.2
(3,1) 1137.7 1263.0 1420.0 1282.3 1347.9 1373.9 1556.9 1160.7 1152.8
(4,1) 1223.0 1410.0 1649.4 1417.4 1500.2 1461.7 1657.0 1270.1 1278.3
(2,2) 2054.7 2226.5 2368.3 2278.1 2388.7 2585.4 2977.5 2037.4 2011.4
(3,2) 2138.7 2345.0 2526.1 2413.9 2521.2 2670.4 3062.7 2140.9 2133.3

Table 7
Natural frequencies of composite sandwich plates with different dimensions under SCSC boundary conditions (Hz).

Mode Relative density ρ̄ Angle and thickness Angle and platform length Angle and height
shape
7.29% 14.57% 21.86% 60°/0.53 mm 75°/0.63 mm 60°/16.68 mm 75°/11.92 mm 60°/8.70 mm 75°/7.62 mm

(1,1) 924.3 954.2 987.26 991.51 1037.1 1103.1 1212.0 914.3 901.1
(2,1) 967.2 1034.7 1120.8 1070.5 1125.3 1155.9 1281.0 968.4 966.9
(3,1) 1036.5 1165.3 1333.2 1198.7 1264.1 1240.2 1383.7 1063.1 1079.2
(4,1) 1129.3 1340.9 1611.4 1358.6 1446.1 1350.7 1516.7 1188.0 1221.8
(2,2) 1913.9 2075.3 2203.6 2176.7 2303.2 2454.9 2844.0 1866.4 1920.9
(3,2) 1984.0 2106.1 2371.1 2315.2 2440.4 2537.0 2931.2 1964.7 2047.6

of composite sandwich plates with same relative density and varied height and angle were similar. In summer, the influence
of dimensions of composite sandwich plates on natural frequencies was the same under different boundary conditions.

6. Conclusions

The equivalent moduli of composite reentrant honeycomb core were deduced. The results calculated by theory were
similar to those obtained by FEM. The free vibration control equation was derived by Zig-Zag theory and the natural fre-
quencies of composite sandwich plates were gained through this theory. The vibration behaviours of sandwich plates with
three relative densities were studied experimentally, theoretically and numerically. It was obvious that the frequencies of
the sandwich plate were raised as relative density increased. The results calculated by theory and FEM were similar to the
values tested by experiment, which indicated the theory and FEM could forecast the natural frequencies of sandwich plates
effectively. Furthermore, the influence of dimensions of the unit cell on natural frequencies of sandwich plates were re-
searched theoretically and numerically. The natural frequencies of sandwich plate with same relative density raised at same
mode shape as angle increased and thickness increased. When angle increased and platform length decreased, the natural
frequencies of sandwich structure with same relative density increased at same mode shape. As angle increased and height
decreased, the frequencies of the sandwich plate was similar at same mode shape. When the angle and relative density of
unit cell were same, changing the platform length could obtain the maximum values of natural frequencies, and changing
the height could obtain the minimum results of natural frequencies. In summary, the dimensions of the unit cell had a great
influence on the frequencies of sandwich plates.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have
appeared to influence the work reported in this paper.

Data Availability

No data was used for the research described in the article.

Acknowledgments

The present work is supported by National Science Foundation of China under Grant No. 12172140.

References

[1] I. Cielecka, J. Jedrysiak,

˛ A non-asymptotic model of dynamics of honeycomb lattice-type plates, J. Sound Vib. 296 (1) (2006) 130–149.
[2] S.S. Shi, Z. Sun, X.Z. Hu, H.R. Chen, Carbon-fiber and aluminum-honeycomb sandwich composites with and without Kevlar-fiber interfacial toughening,
Compos. Part A Appl. Sci. Manuf. 67 (2014) 102–110.
[3] B. Wang, J.Q. Hu, Y.Q. Li, Y.T. Yao, S.X. Wang, L. Ma, Mechanical properties and failure behavior of the sandwich structures with carbon fiber-reinforced
X-type lattice truss core, Compos. Struct. 185 (2018) 619–633.

567
W. Jiang, J. Zhou, J. Liu et al. Applied Mathematical Modelling 116 (2023) 547–568

[4] M. Li, H.L. Fan, Free vibration behaviors and vibration correlation technique of hierarchical Isogrid stiffened composite cylinders, Thin Wall Struct. 159
(2021) 107321.
[5] J. Mei, J.Y. Liu, W. Huang, Three-point bending behaviors of the foam-filled CFRP X-core sandwich panel: experimental investigation and analytical
modelling, Compos. Struct. 284 (2022) 115206.
[6] H. Li, H.Y. Lv, T.N. Zhang, Q.K. Han, J.G. Liu, J. Xiong, Z.W. Guan, Modeling and evaluation of dynamic degradation behaviours of carbon fibre-reinforced
epoxy composite shells, Appl. Math. Model. 104 (2022) 21–33.
[7] J. Lou, L.Z. Wu, L. Ma, J. Xiong, B. Wang, Effects of local damage on vibration characteristics of composite pyramidal truss core sandwich structure,
Compos. B. Eng. 62 (2014) 73–87.
[8] J.S. Yang, L. Ma, K.U. Schröder, Y.L. Chen, S. Li, L.Z. Wu, R. Schmidt, Experimental and numerical study on the modal characteristics of hybrid carbon
fiber composite foam filled corrugated sandwich cylindrical panels, Polym. Test. 68 (2018) 8–18.
[9] T.M. Tu, L.N. Thach, T.H. Quoc, Finite element modeling for bending and vibration analysis of laminated and sandwich composite plates based on
higher-order theory, Comput. Mater. Sci. 49 (4) (2010) S390–S394.
[10] R. Sahoo, B.N. Singh, A new trigonometric zigzag theory for buckling and free vibration analysis of laminated composite and sandwich plates, Compos.
Struct. 117 (2014) 316–332.
[11] A. Mahi, E I A A Bedia, A. Tounsi, A new hyperbolic shear deformation theory for bending and free vibration analysis of isotropic, functionally graded,
sandwich and laminated composite plates, Appl. Math. Model. 39 (9) (2015) 2489–2508.
[12] J.E. Chen, W. Zhang, M. Sun, M.H. Yao, J. Liu, Free vibration analysis of composite sandwich plates with different truss cores, Mech. Adv. Mater. Struct.
25 (9/16) (2018) 701–713.
[13] C. Luo, C.Z. Han, X.Y. Zhang, X.G. Zhang, X. Ren, Y.M. Xie, Design, manufacturing and applications of auxetic tubular structures: a review, Thin Wall
Struct. 163 (2021) 107682.
[14] Y.G. Guo, J. Zhang, L.M. Chen, B. Du, H.C. Liu, L.L. Chen, W.G. Li, Y.Z. Liu, Deformation behaviors and energy absorption of auxetic lattice cylindrical
structures under axial crushing load, Aerosp. Sci. Technol. 98 (2020) 105662.
[15] H.Y. Zhou, K.C. Jia, X.J. Wang, M.X. Xiong, Y.H. Wang, Experimental and numerical investigation of low velocity impact response of foam concrete filled
auxetic honeycombs, Thin Wall Struct. 154 (2020) 106898.
[16] J.H. Zhang, X.F. Zhu, X.D. Yang, W. Zhang, Transient nonlinear responses of an auxetic honeycomb sandwich plate under impact loads, Int. J. Impact
Eng. 134 (2019) 103383.
[17] Y.L. Wang, W.Z. Zhao, G. Zhou, C.Y. Wang, Analysis and parametric optimization of a novel sandwich plate with double-v auxetic structure core under
air blast loading, Int. J. Mech. Sci. 142-143 (2018) 245–254.
[18] C. Li, H.S. Shen, H. Wang, Nonlinear bending of sandwich beams with functionally graded negative Poisson’s ratio honeycomb core, Compos. Struct.
212 (2019) 317–325.
[19] C. Li, H.S. Shen, H. Wang, Z.F. Yu, Large amplitude vibration of sandwich plates with functionally graded auxetic 3D lattice core, Int. J. Mech. Sci. 174
(2020) 105472.
[20] N.D. Duc, S.E. Kim, P.H. Cong, N.T. Anh, N.D. Khoa, Dynamic response and vibration of composite double curved shallow shells with negative Poisson’s
ratio in auxetic honeycombs core layer on elastic foundations subjected to blast and damping loads, Int. J. Mech. Sci. 133 (2017) 504–512.
[21] N.D. Duc, S.E. Kim, N.D. Tuan, P. Tran, N.D. Khoa, New approach to study nonlinear dynamic response and vibration of sandwich composite cylindrical
panels with auxetic honeycomb core layer, Aerosp. Sci. Technol. 70 (2017) 396–404.
[22] W. Zeng, W.M. Jiang, J.Y. Liu, W. Huang, Fabrication method and dynamic responses of composite sandwich structure with reentrant honeycomb cores,
Compos. Struct. 299 (2022) 116084.
[23] W.C. Yang, R.X. Huang, J.Y. Liu, J.X. Liu, W. Huang, Ballistic impact responses and failure mechanism of composite double-arrow auxetic structure, Thin
Wall Struct. 174 (2022) 109087.
[24] D.U. Yang, S. Lee, F.Y. Huang, Geometric effects on micropolar elastic honeycomb structure with negative Poisson’s ratio using the finite element
method, Finite Elem. Anal. Des. 39 (3) (2003) 187–205.
[25] Y.Y. Zhou, Y.F. Li, D. Jiang, Y. Chen, M.X. Yi, L.J. Jia, In-plane impact behavior of 3D-printed auxetic stainless honeycombs, Eng. Struct. 266 (2022).
[26] T. Mukhopadhyay, S. Adhikari, Effective in-plane elastic properties of auxetic honeycombs with spatial irregularity, Mech. Mater. 95 (2016) 204–222.
[27] C. Qi, A. Remennikov, L.Z. Pei, S. Yang, Z.H. Yu, T.D. Ngo, Impact and close-in blast response of auxetic honeycomb-cored sandwich panels: experimental
tests and numerical simulations, Compos. Struct. 180 (2017) 161–178.
[28] Z. Ru, R. Xin, Y.Z. Xiang, L. Chen, Z. Yi, M.X. Yi, Mechanical properties of concrete composites with auxetic single and layered honeycomb structures,
Constr. Build. Mater. 322 (2022) 126453.
[29] K.Y. Singh, H.N. Sri, B.N. Dhar, S.S. Kumar, Free vibration analysis of PLA based auxetic metamaterial structural composite using finite element analysis,
Mater. Today Proc. 56 (2022) 1063–1067.
[30] V.Q. Nguyen, V.T. Nguyen, Q.T. Quoc, D.N. Dinh, Nonlinear forced vibration of sandwich cylindrical panel with negative Poisson’s ratio auxetic honey-
combs core and CNTRC face sheets, Thin Walled Struct 162 (2021) 107571.
[31] Y.L. Chen, D.W. Wang, L. Ma, Vibration and damping performance of carbon fiber-reinforced polymer 3D double-arrow-head auxetic metamaterials, J.
Mater. Sci. 56 (2020) 1–18.
[32] D.D. Ngo, Q.Q. Tran, D.D. Nguyen, Vibration analysis of auxetic laminated plate with magneto-electro-elastic face sheets subjected to blast loading,
Compos. Struct. 280 (2022) 114925.
[33] J. Mei, P.J. Tan, T. Zhang, J.Y. Liu, B. Wang, W. Huang, Fabrication and mechanical characterization of CFRP X-core sandwich panels, Thin Wall Struct.
158 (2021) 107144.
[34] Z. Li, M.J. Crocker, Effects of thickness and delamination on the damping in honeycomb–foam sandwich beams, J. Sound Vib. 294 (3) (2006) 473–485.
[35] Q. Ni, G.Q. LI, Q. Qian, Mechanics of Materials, Huazhong University of Science and Technology Publishing, 2006 in Chinese.
[36] G.T. Yang, Introduction to Elasticity and Plasticity, 2nd ed., Tsinghua University Press, 2013 in Chinese.
[37] G.L. Shen, G.K. Hu, B. Liu, Mechanics of Composite Materials, 2nd ed., Tsinghua University Press, 2013 in Chinese.
[38] L.J. Gibson, M.F. Ashby, Cellular solids: Structure and Properties, 2nd ed., Cambridge University Press, Cambridge, UK, 1997.
[39] D.D. Nguyen, S.E. Kim, H.C. Pham, T.A. Nguyen, D.K. Nguyen, Dynamic response and vibration of composite double curved shallow shells with negative
Poisson’s ratio in auxetic honeycombs core layer on elastic foundations subjected to blast and damping loads, Int. J. Mech. Sci. 133 (2017) 504–512.

568