International Journal of Mechanical Sciences 272 (2024) 109189

Contents lists available at ScienceDirect

International Journal of Mechanical Sciences

journal homepage: [Link]/locate/ijmecsci

In pursuit of a high-performance mechanical metamaterial:

Simple-cubic-octahedral plate lattice
Manash Jyoti Baishya a, Nelson Muthu b, Prasenjit Khanikar a, *
a
Department of Mechanical Engineering, Indian Institute of Technology Guwahati, North Guwahati 781039, Assam, India
b
Department of Mechanical Engineering, Indian Institute of Technology Palakkad, Palakkad 678623, Kerala, India

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

Keywords: A comprehensive review by authors on well-known lattice cell topologies composed of metals and polymers
Mechanical metamaterial having sizes ranging from micro to nano-length scale revealed the simple-cubic (SC) plate lattice exhibits the
Additive manufacturing highest elastic modulus and yield strength. In this study, a novel simple-cubic-octahedral (SCO) plate lattice is
Plate lattice
proposed by reinforcing octahedral plates with SC-plate lattice. Finite element simulations demonstrated that the
Hashin-Shrtikman upper bound
Suquet bound
SCO-plate lattice exhibited the highest elastic modulus and yield strength among the strut, surface and plate
Isotropy lattices investigated previously. Furthermore, the SCO-plate lattice displayed a high degree of isotropy. Also, the
proposed lattice absorbed significantly higher energy compared to the SC-plate lattice for a relative density of
40%. A new semi-analytical framework is also presented to evaluate the elastic modulus of the SCO-plate lattice.
The homogenization scheme in Abaqus was used to evaluate the effective properties, resulting in a significant
reduction in computational time. The effects of geometrical parameters on the mechanical properties of the SCO-
plate lattice were also studied using a design of experiment approach. The quasi-static compression tests were
performed on 3D-printed lattices fabricated using a multi-jet fusion technique to validate the numerical simu­
lations. The proposed plate lattice can be an excellent choice for lightweight applications.

1. Introduction into bending and stretching dominated based on the nodal connectivity
of struts [8–10]. The bending-dominated behavior is characterized by its
Lattice structures are lightweight, multi-functional materials with low strength and stiffness, while stretching-dominated structures are
repeating patterns, offering significant advantages over traditional ma­ characterized by high strength and stiffness [10].
terials. These metamaterials have gained substantial interest due to ad­ The primary focus of the researchers working on mechanical meta­
vancements in additive manufacturing technology, enabling the precise materials has been strut lattices in the last decade. Various methods to
and efficient production of intricate structures through 3D printing. This enhance the mechanical properties of strut lattices have been explored
class of advanced materials possesses exceptional properties, including recently. These methods include tailoring pre-existing lattice topologies
high strength, energy absorption, thermal insulation, noise reduction and to create novel designs, designing functionally graded lattices, using
low weight, thereby gaining popularity for various purposes such as tapered struts, incorporating a soft phase into the pores of the lattice and
aerospace, automotive, biomedical, machinery and construction in­ topology optimization [1,11–18]. Another recent design approach,
dustries [1,2]. The mechanical behavior of lattice structures depends on introduced by Pham et al. [19], has combined multiple lattice topologies
their parent material, unit cell topology and relative density [3–5]. to improve mechanical performance [13,20–23]. Researchers have also
Lattice structures are divided into strut, surface and plate lattices enhanced the usefulness of this approach by using the classical com­
based on their geometric configuration. Among them, the most exten­ posite rule of mixture to tailor the mechanical properties of the
sively studied are the strut-based lattices, having strut thicknesses multi-phase lattice structures [20,24–26].
ranging from millimeters to nanometers [6,7]. Strut-based lattices Surface-based lattice structures are another category of lattices that
comprise struts connected at different nodes or joints, and their nodal have been widely studied. These lattice structures can be divided into
connectivity determines their load-bearing capacity, structural stability two types− triply periodic minimal surface (TPMS) lattice and shellular
and energy absorption abilities. Maxwell classified strut-based lattices lattice. The TPMS lattices are designed using level-set approximation

* Corresponding author.
E-mail address: pkhanikar@[Link] (P. Khanikar).

[Link]
Received 1 December 2023; Received in revised form 25 February 2024; Accepted 9 March 2024
Available online 11 March 2024
0020-7403/© 2024 Elsevier Ltd. All rights reserved.
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

equations in the form of trigonometric functions creating surfaces with best of the authors’ knowledge, this study compares all plate lattice
zero mean radius of curvature [4,5,27–33]. In contrast, shellular lattices topologies reported in the literature, including hybrid plate lattice to­
have surfaces with a non-zero mean radius of curvature [34]. pologies. It is noteworthy that multi-phase lattice structure, which are
Plate lattices, which are composed of interconnected plates in 3D also called as multilattice, are not considered for comparison due to their
space, have recently gained popularity due to their specific advantages inability to surpass the elastic modulus of its higher-stiffness constituent
over strut and surface lattices. Although strut lattices have been the lattice phase as they follow the rule of mixture used for conventional
prime focus of the researchers working on mechanical metamaterials but composite materials [20,24,25].
structural stability of these lattices is substantially low [35]. The pri­ Although lattice structures offer numerous advantages, most of these
mary factor contributing to the limited stability of the strut lattices is structures exhibit anisotropic behavior, which is determined by their
that the nodes or the joints of the struts act as hinges under external arrangement and the orientation of the struts or plates in the unit cell. In
loads, thereby exhibiting nodal bending and twisting. In the case of the applications like impact energy absorption and bone implants, achieving
plate lattices, the three-dimensional intersections of the plates effec­ isotropy is crucial not only for structural integrity but also for human
tively prevent the formation of such kinematic mechanisms ensuring safety [44]. Researchers have developed a few isotropic lattices [41,44],
high structural efficiency [36,37]. In addition, plate lattices can store but the mechanical properties of these lattices lie well below the
high strain energy and require significantly higher loads to undergo both SC-plate lattice.
elastic and plastic deformation [38,39]. The plate lattices can achieve The present study aims to design the best-performing lattice by
stiffness and strength values close to or even beyond the maximum reviewing the mechanical properties of the commonly studied strut,
theoretical bounds— Hashin-Shtikman upper bound (HSU) and Suquet surface and plate lattices having different unit cell topologies. By
upper bound, respectively [36,38]. In general, plate lattices typically investigating these lattices and understanding their underlying defor­
exhibit a higher elastic modulus and yield strength compared to most mation mechanisms, a novel plate lattice was designed and developed to
strut lattices and the majority of surface lattices [34,36,40]. Although outperform the mechanical properties of the existing lattices. A new
plate or surface lattices outperform strut lattices, a drawback in their semi-analytical framework to obtain the elastic modulus of the novel
fabrication is powder entrapment within the enclosed spaces during lattice was also presented. In addition, the quasi-static mechanical
powder-bed 3D printing. However, the entrapped powder can be properties and the deformation mechanics of the proposed lattice were
removed by creating holes in the plates or surfaces though the presence evaluated by performing finite element (FE) simulations in Abaqus, and
of holes marginally reduces the mechanical properties [41–43]. To the mechanical tests were performed to corroborate the numerical

( ) ( ) ( ) ( )
E
Fig. 1. (a) Normalized elastic modulus , (b) normalized yield strength σy,sy ρ , normalized bulk modulus KKs ρ , (c) normalized shear modulus GGs ρ , and (d) total
σ
Es ρ
stiffness values of various lattice unit cell topologies made of metals and polymers obtained from previously published literature [16,38,41,45–55].

2
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

simulation results. The homogenization technique, in which the het­ 2. Design motivation, material and experiment
erogeneous material was substituted with a continuous material, was
used to evaluate the effective elastic modulus of the novel lattice to 2.1. Design motivation
minimize computational effort significantly. The mechanical properties
of the novel lattice exceeded the properties of the best-performing pre­ The motivation of this proposed lattice design is to attain a higher
viously studied lattice structure, i.e. the SC-plate lattice, as shown in the elastic modulus than the best-performing SC-plate lattice. To achieve
normalized elastic modulus and normalized yield strength plots of Fig. 1 this purpose, the von Mises stress distribution and strain energy density
(a) and (b), respectively. Although this study primarily focuses on the were investigated through FE simulations by applying 3% compressive
elastic modulus, yield strength and energy absorption ability, the com­ strain on an SC-plate unit cell. The unit cell has a dimension (LSC)of 7.5
parison of bulk modulus, shear modulus and total stiffness (Ω) of the mm × 7.5 mm × 7.5mm and a relative density (RD) of 40%. The vertical
proposed lattice with other previously studied lattices is also presented plates of the SC-plate oriented in the loading direction had maximum
in Figure (c), (d) and (e), respectively. The evaluation of the total stiff­ stress distribution and strain energy density than those of horizontal
ness and HSU bounds for bulk modulus and shear modulus can be found plates (cf. Fig. 2(a)). Based on this observation, adding structural
in literature [38]. members with high strain energy density can further augment the load-
Furthermore, the energy absorption ability of this proposed lattice bearing capacity. Since plate members can store strain energy density
exceeded the SC plate lattice. Additionally, the three-dimensional elastic more than strut members of the same mass, additional plate members
surface of the SCO-plate lattice displayed a high degree of isotropy as were chosen as reinforcement members instead of strut members to
opposed to the strong direction-dependent elastic modulus of the SC- enhance the mechanical performance of the SC-plate lattice. As already
plate lattice. known, the strain energy density due to bending is higher than that of

Fig. 2. Isometric view, top/bottom view, von Mises stress and strain energy density distributions of (a) SC-plate unit cell, (b) SCO-plate unit cell without holes and (c)
SCO-plate unit cell with holes of 1.5mm diameter, and (d) SC-plate lattice and SCO-plate lattice with holes having 4 × 4 × 4 unit cells in x × y × z directions. (All unit
cells and lattices have a relative density of 40%).

3
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

axial compression, therefore eight inclined plates, each having an Table 2

identical thickness (toct), were added in the SC-plate unit cell to increase Mechanical properties of parent material PA12 obtained from ASTM-D68 tensile
the strain energy density (cf. Figure 2(b)). The resulting unit cell is test:.
named as SC-octahedral (SCO) plate, and it had the same dimensions Elastic modulus Yield strength Ultimate tensile Elongation at
and relative density as those of SC-plate unit cell. The design parameters (MPa) (MPa) strength (MPa) break (%)
of the SC-plate and the SCO-plate unit cells were chosen based on the 1214 29 48 15.5
printing feasibility as listed in Table 1. The strain energy density and the
von Mises stress values of the SCO-plate unit cell were higher than the
SC-plate unit cell when 3% compressive strain was applied as illustrated remain unchanged regardless of the printing direction [55]. The PA12
in Figure 2(b). The higher values of von Mises stress in the SCO-plate powder having an average particle size of 60μm used for the 3D printing
unit cell than in the SC-plate unit cell indicated that, for the same had a bulk density of 0.425gm/cm3 and a melting temperature of 187
deformation, the SCO-plate unit cell would require higher force to
◦
C. The layer thickness of printing was 80μm. Three samples of each
deform it, thereby exhibiting higher axial stiffness. design were created to check the repeatability.
The SC-plate lattice and the SCO-plate lattice were fabricated using
the multi-jet powder bed fusion technology. The overall size of all lattice 2.4. Experimental setup for quasi-static compression test
samples under investigation was chosen as 30 mm × 30 mm × 30mm
with 4 × 4 × 4 unit cells in x × y × z directions. The relative density of The quasi-static compressive response and the deformation mecha­
the lattice structures was 40%, which was the same as that of the unit nism of lattices investigated through the FE simulation were validated
cells considered for the FE simulations. To address the problem of using mechanical tests. Three samples of each design, measuring 30 mm
powder entrapment in lattice structures, holes having a diameter of × 30 mm × 30 mm with 40% RD, were compressed using a Shimadzu
1.5mm were created in the octahedral plates and the horizontal plates AGS-X 100 kN Universal Testing Machine (UTM). In the experimental
(cf. Figure 2(c) and (d)). For the SC-plate lattice, the holes created in the setup shown in Fig. 4, the samples were positioned on the stationary
horizontal plates were sufficient for powder removal (cf. Figure 2(d)). bottom platen while the top platen was allowed to move downward at a
The subsequent sections present a comparison of the mechanical constant velocity of 1.8 mm/min, resulting in a strain rate of 10− 3 S− 1
properties of the SCO-plate lattice with the SC-plate lattice having the during the experiment. Compression continued until a deformation of 21
same relative density. The investigation will utilize a theoretical mm was achieved, slightly exceeding the point of visible lattice
approach combined with semi-analytical analysis, homogenization densification.
method, FE simulation and experiments.
3. Theoretical analysis and finite element model

2.2. Parent material 3.1. Theoretical analysis to evaluate the elastic modulus of SC-plate and
SCO-plate
To determine the mechanical properties of the parent material PA12
fabricated by multi-jet powder bed fusion, quasi-static tensile tests were
In this section, a theoretical framework to calculate the elastic
conducted on three specimens using an Instron-8801 100KN universal modulus of the SC-plate and the SCO-plate was established. As seen in
testing machine following the ASTM-D638 standard. The results ob­
Fig. 5(a) and (c), the elastic modulus of the SC plate and the SCO-plate
tained from the tensile tests, including the elastic modulus, yield unit cell in the x, y, and z-directions are identical due to cubic symmetry.
strength, ultimate tensile strength and elongation at break, are listed in
Therefore, the theoretical analysis of elastic modulus in the z-direction
Table 2. The engineering stress-strain curves of the three tensile speci­
(EZZ) is presented for each case.
mens are shown in Fig. 3.
3.1.1. Theoretical elastic modulus of the SC-plate unit cell
2.3. Sample fabrication To determine elastic modulus in the z-direction (EZZ), unidirectional
compressive traction in the z-direction (σZ)was applied to induce an
In this study, polymeric lattice samples made of PA12 were fabri­ axial defection (δz) in the SC-plate unit cell (cf. Figure 5(a)). From the
cated in a Hewlett-Packard (HP) made 3D-printer using a multi-jet free body diagram (FBD) illustrated in Figure 5(b), the generated
fusion technique. To determine the mechanical properties used in the compressive load (Pz) due to σz was resisted by the two vertical plates in
FE simulations, three tensile specimens of ASTM-D638 standard were the z-direction. Therefore, the two vertical plates were considered as a
fabricated along with the lattice structures. Briefly, the multi-jet fusion single bar having length LSCwhere the load Pz was applied on the cross-
technique involves spreading powder evenly on the build platform, 2
sectional area of A = 2LSC tSC − tSC (cf. Fig. 5(b)).
preheating it and selectively melting it with the help of a fusing agent. A The axial strain energy stored due to the compressive load (U), is
detailing agent is also used to print sharp edges for high resolution. The expressed as,
part is built by repeatedly laying down successive layers of powder, with
the bed moving down after adding each new layer [56,57]. This method
ensures high precision and the mechanical properties of the product

Table 1
Design parameters of the different lattice structures.
Lattice unit cell Relative density Unit cell dimensions Plate dimensions
(RD) (%) (LSC) (mm)
Horizontal and vertical plate Octahedral plate thickness Slenderness ratio
thickness (tSC) (mm) (toct) (mm) (t )
SC
LSC

Simple cubic (SC) plate 40 7.5 × 7.5 × 7.5 1.2 – 0.16

Simple cubic octahedral (SCO) plate 40 7.5 × 7.5 × 7.5 0.9 0.7 0.12
(without hole)
SCO-plate (with holes of 1.5 mm 40 7.5 × 7.5 × 7.5 1 0.7 0.13
diameter)

4
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

Fig. 3. Quasi-static tensile test of parent material PA12 using ASTM-D68 standard: (a) experimental setup of Instron 8801 machine, and (b) stress-strain response
obtained from three repeated experiments.

Fig. 4. The experimental setup for quasi-static compression of the lattice structures.

∫L Now, substituting Eqs. (4) and 6 in Eq. (5),

P2Z dz
U= (1) { ( ) ( )}
2
2AES tSC tSC
0 EZZ = 2 − ES (7)
LSC LSC
where, ES represents the elastic modulus of the parent material.
2 Substituting, tSC = 1.2mm, LSC = 7.5mm and Es = 1214MPa in Eq.
Substituting the area of cross-section, A = 2LSC tSC − tSC and inte­
(7), EZZ = 357.40MPa
grating Eq. (1),

P2Z L 3.1.2. Theoretical elastic modulus of the SCO-plate unit cell

U= (2)
2
2(2LSC tSC − tSC )ES Deriving the effective elastic modulus of the SCO-plate unit cell
purely through theoretical methods is challenging due to the out-of-
The δZ can be evaluated using Castigligano’s second theorem, plane deformation of inclined octahedral plates. Hence, a semi-
∂U PZ LSC analytical approach inspired by the finite element method was used to
δz = = 2
(3) calculate the effective elastic modulus of the SCO-plate lattice having
∂PZ (2LSC tSC − tSC )ES
40% RD. To simplify calculations, 1/8th of the geometrically symmetric
The strain in the z-direction εZ is expressed as structure was considered (cf. Figure 5(c)). The semi-analytical proced­
δZ PZ ure is as follows:
εZ = = 2
(4) Step 1: Calculation of the stiffness ‘K1’ (N/mm) of the SC-plate by the
L (2LSC tSC − tSC )ES
theoretical method
By applying Hooke’s law, the elastic modulus Ezz can be expressed as The mathematical expression K1for the 1/8th part of the SCO-plate
σZ lattice (cf. Figure 5(c)) as follows:
EZZ = (5)
εZ ( 2
)
PZ 2LSC tSC − tSC ES
K1 = = (8)
where, the compressive stress σZ is expressed as 2δZ 2LSC

Pz Substituting, values of LSC = 7.5mm, tSC = 0.9mm (cf. Table 1) and ES

σz = (6) = 1214MPa in Eq.8,
L2SC

5
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

Fig. 5. Theoretical procedure to evaluate elastic modulus of the SC-plate and SCO-plate unit cells: (a) compressive stress applied to the effective area L2SC of SC-plate
unit cell, (b) free body diagram showing generated compressive load Pz resisted by the two vertical constituent plates considered as a single bar having a length LSC
and cross-sectional area A, (c) 1/8th part of the SCO-plate, and (d) local coordinates of the triangular plate for the calculation of stiffness.

K1 = 1027.90 MPa (9) detJ = x13 y23 − x32 y13

Step 2: Calculation of the stiffness ‘K2’ (N/mm) of the octahedral plate by
the FE method where, x and y are local co-ordinates of the triangular plate
The stiffness matrix [K]of the triangular plate having a thickness ′toct′ xij = xi − xj (14)
and area ′Aoct′subjected to plane stress condition can be expressed as
[58]: yij = yi − yj (15)
T
[K] = [B] [D][B]toct Aoct (10) Substituting Eqs. (11)–(13) in Eq. (10), the stiffness matrix K can be
obtained as
Figure 5(d) illustrates the inclined triangular plate with local coor­
⎡ ⎤
dinate axes x and y , having base and heights measuring 5.3 mm and 4.9 552.03 0 − 276.02 157.94 − 276.02 − 157.94
mm, respectively. Due to the small thickness, a plane stress condition ⎢
⎢ 0 184.95 160.33 − 92.43 − 160.33 − 92.43 ⎥
⎥
was assumed for the analysis. ⎢ − 276.02 160.33 277.04 − 159.13 − 0.997 − 1.19 ⎥
[K] = ⎢ ⎥(N/mm)
⎢ 157.94 − 92.43 − 159.13 461.13 3.53 − 371.78 ⎥
The plane stress constitutive matrix for linear elastic isotropic ma­ ⎢ ⎥
⎣ − 276.02 − 160.33 − 0.997 − 1.19 277.03 159.13 ⎦
terial is
− 157.94 − 92.43 − 1.19 − 368.70 159.13 461.13
⎡ ⎤
1 ϑ 0
ES ⎢⎢ϑ 1
⎥ Since the base of the triangular plate is fixed, the stiffness K2 along
0 ⎥ (11)
[D] = ⎢
1 − ϑ2 ⎣
⎥ the x the direction was considered the stiffness of the triangular plate.
1 − ϑ⎦
0 0
2 Therefore, K2 = 552.03MPa (16)
The strain displacement (B) matrix is Step 3: The effective stiffness combining K1 and K2 along the loading
⎡ ⎤ direction is, therefore,
y 0 y31 0 y12 0
1 ⎣ 23
[B] = 0 x32 0 x13 0 x21 ⎦ (12) K = K1 + K2 cos2 θ (17)
detJ
x32 y23 x13 y31 y21 y12
Where θ is the angle of inclination of the triangular plate with the
where the Jacobian (J) is given by vertical axis z (cf. Figure 5(c)).
[ ] Step 4: Equating the strain energy of this SCO-plate unit cell to that of
x
[J] = 13
y13
(13) homogenized material of the same volume to calculate the effective elastic
x23 y23
modulus of the SCO-plate unit cell

6
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

The strain energy of the 1/8th part of the SCO-plate in terms of PZ, K1 However, to determine the effective modulus, the strain energy of a
and K2 can be expressed as: homogenized RVE (U) is employed instead of that of a heterogeneous
RVE, which is expressed as:
PZ 2
USCO = (18)
32 × (K1 + K2 cos2 θ) 1
U = σij εij VRVE (25)
2
Now, let us consider a cubic homogenized pseudo-continuum ma­
terial having a length of L2SC , which is equivalent to the size of the 1/8th where, σ ij and εij are average stress and strain respectively.
part of the SCO-plate unit cell. Under the identical compressive load of P4Z To determine the mechanical properties of an RVE, the periodic
, the strain energy of the homogenized continuum can be expressed as: boundary conditions (PBCs) are imposed. The displacement field ui(x1,
x2,x3), of an RVE is expressed as [64]:
PZ 2
UHomogenized = (19) ui (x1 , x2 , x3 ) = εij xj + u∗i (x1 , x2 , x3 ) (26)
16 × LSC × Eeff

Equating strain energies from Eqs.18 and 19, where εij = average strain tensor of the RVE. The periodic function
2 u∗i (x1 , x2 , x3 ) represents the modification of the displacement field due to
2(K1 + K2 cos θ)
Eeff = (20) the heterogeneous structure.
LSC
To ensure the accurate representation of a continuous body through
Substituting values of K1, K2 from Eqs. (9) and (16), and LSC = the periodic array of RVEs, it is necessary to impose the following con­
7.5mm, θ = 35.260 in Eq. (20), ditions at the boundaries between neighboring RVEs:
Eeff = 371.15 MPa (21)
1. Continuity of displacements must be maintained, and there should be
Eq. (21) represents the lower bound estimation of the SCO-plate no separation or overlap of neighboring RVEs at the boundaries after
lattice [59]. deformation.
Thus, the force boundary conditions applied to the SCO-plate resul­ 2. The distribution of traction must be consistent at opposite parallel
ted in the lower bound of the elastic modulus, which is higher than the boundaries of an RVE.
theoretical elastic modulus of the SC-plate with 40% RD. Table 3 rep­
resents the comparison of the elastic modulus of SCO-plate and SC-plate These conditions are necessary to accurately capture the behavior of
obtained from theoretical analysis and homogenization scheme in the microstructure in relation to the macroscopic response of the
Abaqus. material.
The first condition for the displacement field, as outlined in the two
3.2. Homogenization theory conditions, is fulfilled by Eq. (8). However, the direct implementation of
this equation is not possible as the periodic function u∗i (x1 , x2 , x3 ) not
In design analysis, engineering materials exhibit heterogeneity due known. To maintain parallel boundary surfaces for any RVE, the
to their microscopic composition. To simplify analysis, it is more prac­ displacement of a pair of opposite boundary surfaces can be represented
tical to consider the overall or effective properties of the material, which as follows:
are averaged across a certain volume of the heterogeneous material. The
Homogenization theory suggests that a material with a complex uk+ k+ ∗
i = εij xj + ui (27)
microstructure can be approximated by a homogeneous pseudo-
continuum material. This method uses a representative volume uk−i = εij xk−j + u∗i (28)
element (RVE) to represent periodic microstructures and establish a
relationship between volume-averaged stresses and strains at the where the indices ′k+′ and ′k− ′ denotes the kth pair of opposite boundary
macroscopic level. An RVE contains enough micro-architectural char­ surfaces of an RVE. Since u∗i (x1 , x2 , x3 ) remains the same at the two
acteristics to represent the entire physical response of the lattice when parallel boundaries, the difference between the Eqs. (27) and (28) gives:
subjected to external loads [60–63]. ( )
To determine macroscopic stress and strain, it is necessary to perform uk+ k− k+
i − ui = εij xj − xj
k−
= εij Δxkj (29)
volume averaging of the stress and strain over the RVE, which is
The term ′Δxkj is constant for each pair of parallel boundaries. By
′
expressed below:
∫ defining εij , the right-hand side of Eq. (29) becomes constant, which can
1
σ ij = σij (x, y, z)dV (22) now be used as a nodal displacement constraint equation in FE element
VRVE
analysis [64,65]. The effective elastic constants of the proposed
1
∫ SCO-plate lattice are evaluated using a homogenization algorithm
εij = εij (x, y, z)dV (23) developed by [66], and the results are detailed in the supplementary
VRVE
section (cf. Table S1).
The strain energy (U) of a heterogeneous RVE can expressed as:
∫
1
U= σij εij dV (24)
2

Table 3
Comparison of the elastic modulus values obtained from the theoretical and the FE simulations for SCO-plate and SC-plate having a relative density of 40%.
Lattice type Elastic modulus (MPa)

Theoretical analysis Homogenization (PBC) Percentage increase in homogenization

SCO-plate 371.15 433.62 19.39%

SC-plate 357.40 363.20

7
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

3.3. Finite element modeling for quasi-static compression Table 4

Input parameters are chosen at three operating range.
The commercial software Abaqus/Explicit was used for the finite Input parameters Operating range
element (FE) simulations to study the mechanical properties and DoE levels
− 1 0 1
deformation behavior of the proposed lattice design under quasi-static
(t ) 0.77 1.135 1.5
compressive loading. The complexity of lattice deformation necessi­ TR
oct
tSC
tates using explicit solvers due to material, geometrical and contact RD (%) 20 30 40
nonlinearities. Material nonlinearity arises from plastic deformation
when local stress exceeds the elastic limit, while geometrical nonline­
arity results from large deformations during compression. Contact Table 5
nonlinearity emerges from the self-contact between the constituent Analysis of Variance (ANOVA) results.
plates of the lattice during compression [52]. The FE model comprised a Models F-value P-value R2 Predicted Adjusted
lattice structure of dimensions 30 mm × 30mm × 30mm and two R2 R2
compression rigid plates (cf. Fig. 6). The bottom rigid plate was fixed, Elastic modulus 15,127.08 <0.0001 0.9997 0.9992 0.9997
and the top plate underwent quasi-static compression at a constant ve­ Yield strength 9371.02 <0.0001 0.9995 0.9994 0.9990
locity. The lattice samples were discretized using quadratic 10-node Energy absorption 114.25 <0.0001 0.9879 0.9792 0.8889
up to yield
tetrahedral solid elements (C3D10M), while the compression plates
strength
were discretized using rigid shell elements (R3D4). The mechanical Energy absorption 875.23 <0.0001 0.9966 0.9824 0.9954
properties of the parent material PA12 were applied using an isotropic up to
hardening material law obtained from the tensile test results (cf. densification
Table 2). A general contact algorithm was used to define the contact
between the lattice and compression plates and the self-contact between
4. Results and discussions
the constituent plates of the lattice to prevent contact surface penetra­
tion during compression. Mesh sensitivity analysis determined that a
4.1. Statistical analysis
0.9mm mesh size was appropriate for the FE simulation.

The statistical significance of the models (cf. Eq. (i), (ii), (iii) and
3.4. Parametric study using the design of experiment (DoE) technique
(iv)), which were developed using the RSM (Response Surface Meth­
odology) technique based on the face-centered central composite design,
Design of Experiments (DoE) is a cost-effective statistical approach to
was assessed using Analysis of Variance (ANOVA). To assess the sig­
analyze the impact of various factors on responses. The Response Sur­
nificance of the models, it is necessary to evaluate three criteria: (i) the
face Methodology (RSM) is used in this study which generates a math­
probability (p-value) should be lower than 0.05, (ii) the value of the
ematical model between the responses and input parameters involving
coefficient of determination (R2) should approach 1, and (iii) the dif­
higher-order polynomial equations. Analysis of Variance (ANOVA) is
ference between the adjusted R2 and the predicted R2 should not exceed
employed to check the accuracy of the model [67,68].
0.2 [67]. Table 5 shows that all of these criteria are satisfied, demon­
The main factors influencing the mechanical properties of the SCO-
strating that the models are highly significant. In addition, the DoE
plate lattice are the thickness of the vertical-horizontal plates (tSC) and
simulation scheme is detailed in Table 6. The detailed ANOVA tables for
octahedral plates (toct), as well as the relative density (RD). To simplify
( ) responses are presented in Supplementary Information (cf. Table S2, S3,
the study and reduce dependent factors, the thickness ratio TR = ttoct SC S4 and S5). The F-value, which is also known as the Fisher ratio, is
was considered as an input factor. The input parameters were set at three defined as the ratio of the mean sum of squares among responses to the
levels (− 1, 0, 1) as shown in Table 4. Three responses (elastic modulus, mean sum of squares within responses. When the F-ratio significantly
yield strength and energy absorption ability) were obtained from the FE exceeds one, it indicates that the distinctions between responses hold
simulations. A three-level-two-factor face-centered central composite statistical significance and are unlikely to be solely attributable to
design (FCCD) based on the RSM was used to analyze the effects of input random chance. Furthermore, the method for computing p-value is
factors on responses. outlined in [69] . A low p-value (<0.0001) signifies that there is a low
probability (0.01%) that the observed F-value is due to random chance,
demonstrating statistical significance [70].

Fig. 6. Finite element model for quasi-static compression.

8
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

Table 6
Responses obtained from the finite element (FE) simulations for the input parameters suggested by the design of experiments (DoE) .
Serial number of FE Input parameters Responses
simulation
RD TR Elastic modulus Yield strength Energy absorption ability up to yield Energy absorption ability up to
(%) (t ) (MPa) (MPa) strength (J/cm3) densification (J/cm3)
oct
tSC

1 30 1.5 292.7 8.5 0.204 6.66

2 20 0.77 212.8 6.5 0.163 3.45
3 40 1.5 401.7 11.5 0.224 10.90
4 30 0.77 320.3 9.5 0.214 7.58
5 30 1.135 308.4 9 0.215 7.30
6 30 1.135 308.4 9 0.215 7.30
7 30 1.135 308.4 9 0.215 7.30
8 40 1.135 413.8 12 0.24 11.08
9 30 1.135 308.4 9 0.215 7.30
10 20 1.5 186.5 5.5 0.138 3.29
11 20 1.135 199.4 5.8 0.145 3.32
12 30 1.135 308.4 9 0.215 7.30
13 40 0.77 433.6 12.5 0.256 12.38

The mathematical relationship between the responses and the input slight strain hardening was visible in the FE results of the SCO lattice,
parameters, namely relative density (RD in %) and thickness ratio (TR) contrary to the long-plateau behavior observed in the experiments as
is as follows: shown in Fig. 8. A noticeable stress-drop in the experimental response
Elasticmodulus = 27.17 + 10.84 × RD − 39.18 × TR (i) was also observed at a strain of 27.8%. The SC-plate lattice exhibited
Yieldstrength = 1.44 + 0.30 × RD − 1.37 × TR (ii) plateau behavior in the FE simulation, however, stress-drops at a strain
Energyabsorptionuptoyieldstrength = − 0.086 + 0.016 × RD + of 19.5% and 34.5% were observed in the plateau region of the exper­
0.014 × TR − 0.013 × TR2 (iii) imental stress-strain curve. This reduction in the stress level in the
Energyabsorptionuptodensification = − 6.55 + 0.506 × RD + 1.507 experimental response of the lattices was attributed to the inherent
× TR − 0.089 × RD × TR (iv) porosity and surface defects often present in 3D-printed samples, which
The effective elastic moduli of the SCO-plate lattice with different TR reduces the overall strength of the lattice structures [73,74]. The hon­
and RD were determined using the homogenization technique, and the eycomb lattice displayed prolonged constant plateau behavior with no
results are presented in the supplementary Table S1. Notably, the SCO strain hardening, in both FE simulation and experiment. There was a
plate lattice having TR and RD values of 0.77 and 40%, respectively, good agreement in the stress-strain response between the FE simulation
demonstrated the highest mechanical properties (cf. Table 6, and Fig. 7 and experimental results of the honeycomb lattice. The third or final
(b), (c), (d) and (e)). stage of the stress-strain behavior of the lattice structures, i.e., the
Fig. 7(a) demonstrates a satisfactory agreement between the model densification stage, was categorized as the steep rise of stress values
predictions of the responses achieved using the RSM technique and the because of the fully-collapsed lattice with very low porosity as shown in
actual values derived from the FE simulations. From Figure 7(b), (c), (d) Fig. 8. These three lattices exhibited densification starting from almost
and (e), it is evident that the relative density (RD) has a significant effect the same strain as shown in Fig. 11(c).
on the elastic modulus, yield strength and energy absorption ability
compared to the thickness ratio (TR). The thickness of the vertical plates
4.3. Comparison with theoretical maximum stiffness and strength limits
had a stronger impact on the elastic modulus, yield strength and energy
for cellular solids
absorption than the octahedral plates. The highest elastic modulus, yield
strength and energy absorption ability were achieved when the vertical
The maximum elastic limit, Hashin-Shtrikman upper bound, (HSU)
plate thickness was relatively higher (TR = 0.77). Conversely, increasing
and the maximum yield strength limit, Suquet upper bound, are the most
the thickness ratio (TR = 1.5) resulted in lower elastic modulus, yield
popularly used theoretical bounds for the mechanical behavior of
strength and energy absorption at all relative densities. The relation
cellular solids. These bounds were used in this study to compare the
between the input parameters and output response was linear for elastic
mechanical performance of the proposed SCO-plate lattice with the SC-
modulus, yield strength and energy absorption up to densification, while
plate lattice, which is the only lattice found in the literature that exceeds
for energy absorption ability up to yield strength, the relationship was
these limits. The mathematical expressions for HSU for elastic modulus
quadratic (cf. Figure 7(b), (c), (d) and (e)).
(Eq. (30) ) and Suquet upper bound (Eq. (31)) are as follows [36,75]:
EHSU 2ρ(5ϑ − 7)
= (30)
4.2. Stress-strain response ES 13ρ + 12ϑ − 2ρϑ − 15ρϑ2 + 15ϑ2 − 27

In this section, the stress-strain behavior of the SCO-plate lattice σ SU 2ρ

= √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ (31)
having 40% RD obtained from the FE simulation and the experiment is σ y,s 4 + 11 (1 − ρ)
compared. Furthermore, the stress-strain responses of the conventional
3

honeycomb lattice known for its excellent energy absorption ability [71,
where EHSU, σSU and ρ represent theoretical Hashin-Shrtikman upper
72] and the SC-plate lattice, both having 40% RD, were also compared
bound, Suquet upper bound and relative density, respectively, of a
with the SCO-plate lattice. The stress-strain behavior of the SCO-plate
cellular solid. ES,σy,sandϑrepresent elastic modulus, yield strength and
lattice was categorized into three distinct stages. The first stage was
Poisson’s ratio of the solid parent material, respectively.
the linear region of elastic deformation, similar to the behavior observed ( )
in the SC-plate lattice and the honeycomb lattice. It was followed by a The normalized elastic modulus EEs ρ values of the SCO-plate lattice
non-linear plastic deformation zone in the second stage. In this stage, the having elastic modulus E with relative densities of 20%, 30%, and 40%

9
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

Fig. 7. (a) Comparisons of elastic modulus, yield strength and energy absorption ability of SCO-plate lattice predicted by the RSM (Response Surface Methodology)
models with the actual values obtained from the FE simulations, and the surface plots of the responses (b) elastic modulus, (c) yield strength, (d) energy absorption
ability up to yield strength, and (e) energy absorption ability up to densification showing the effects of thickness ratio (TR) and the relative density (RD) on
the responses.

exhibited a 19.5–20.0% higher elastic modulus than the highest- 4.4. Mechanical performance by Gibson-Ashby model
performing SC-plate lattice at the aforementioned relative densities
( )
(cf. Fig. 9(a)). Similarly, the normalized yield strength σy,sy ρ values of
σ The classical power law theory proposed by Gibson and Ashby in­
dicates that the mechanical properties of lattice structures are pre­
the SCO-plate lattice having yield strength σy with 20%, 30%, and 40%
dominantly influenced by their relative density (ρ) [61].
RD exhibited 22.51–25.85% higher yield strength than the SC-plate
The following are the Gibson-Ashby power law for elastic modulus
lattice (cf. Figure 9(b)).
and yield strength, respectively [76],

10
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

Fig. 8. (a) Comparison of stress-strain behavior of the SCO-plate and SC-plate lattices, both having holes in their constituent plates, and honeycomb lattices at a
relative density of 40%, and (b) conventional hexagonal honeycomb lattice showing the loading direction.

Fig. 9. Comparison of mechanical performance of SCO-plate lattice with the highest-performing SC-plate lattice based on the maximum theoretical limits: (a)
( ) ( )
normalized elastic modulus EEs ρ vs. relative density and (b) normalized yield strength σy,sy ρ vs. relative density.
σ

Table 7
The Gibson-Ashby power-law expressions with scaling exponent values for elastic modulus, yield strength, and strain energy of the SCO-plate lattice were obtained
from curve fitting of FE results.
Thickness ratio (TR) of the lattice Gibson-Ashby model
Deformation mechanism
Elastic modulus Yield strength
( )0.985 ( )1.04
0.77 E ρ σY ρ Stretching dominated
∝ ∝
ES ρS σY,s ρS
( )1.053 ( )1.04
1.135 E ρ σY ρ
∝ ∝
ES ρS σY,s ρS
( )0.97 ( )0.98
1.5 E ρ σY ρ
∝ ∝
ES ρS σY,s ρS

( )n
E
∝
ρ
(32) where E, σY and ρ are the elastic modulus, yield strength, and density of
ES ρ S the lattice structure, respectively, whereas, ES, σY,s and ρS are the elastic
( )n modulus, yield strength, and density of the parent material, respectively.
σY ρ The exponent ′n′ depends on the mechanical behavior of the lattice
∝ (33)
σ Y,s ρS structure.

11
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

Fig. 10. Gibson-Ashby relationship: (a) elastic modulus vs relative density, (b) yield strength vs relative density.

Fig. 11. Energy absorption ability: (a) the evaluation of densification strain from the highest peak of efficiency line, (b) comparison of energy absorption ability, and
(c) densification strains of SCO-plate lattice with holes, SC-plate lattice with holes and honeycomb lattice, all having a relative density of 40%.

In this study, the power law for the SCO-plate lattice was established 4.5. Energy absorption ability
through FE simulations having relative densities of 20%, 30%, and 40%.
The obtained power-law expressions from curve fitting indicate that the The energy absorption per unit volume (ξ) was calculated by the area
deformation behavior of the SCO-plate lattice is stretching-dominated, under the stress-strain curve up to the densification strain and can be
with elastic modulus and yield strength scaling exponents (n) close to expressed as below,
1, as per the Gibson-Ashby model [8] (cf. Table 7). ∫ εD
ξ= σ(ε)dε (34)
0

12
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

where ε and σ are strain and stress, respectively, and εD stands for the where σ0 is the nominal stress, respectively at a particular instant.
densification strain. To determine the εD, the energy efficiency curve The densification strain can be evaluated as follows [56]:
⃒
(Eq. (35)) was plotted alongside the stress-strain curve, and the εD was dη(ε)⃒⃒
obtained by identifying the highest peak on the energy efficiency curve ε = εD = 0 (36)
dε ⃒
(cf. Fig. 10(a)). The mathematical expression for energy efficiency (η) is
[18,56]. Fig. 11(a) depicts the graphical procedure used to evaluate the
∫ ε0 densification strain through the energy efficiency method. The energy
σ(ε)dε absorption abilities obtained from the FE simulations and the experi­
η= 0 (35)
σ0 mental investigations of the SCO-plate lattice with holes were compared
to the SC-plate lattice with holes and the conventional honeycomb

Fig. 12. Comparison of deformation mechanisms of (a) the SC-plate lattice, (b) the SCO-plate lattice, all having a relative density of 40%, obtained from FE sim­
ulations and experiments, and (c) SCO-plate having a relative density of 20% and 30%, where buckling tendency is higher at lower relative densities, obtained from
FE simulations.

13
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

lattice, all having 40% RD. The experiments conducted on the SCO-plate observed at strains 19.8% and 34.61% within the plateau region, which
lattice showed a reduction of 6.86% in energy absorption ability was attributed to the fracture of vertical plates occurred consecutively in
compared to the FE simulations, while the experiments on the SC-plate two different horizontal unit cell layers (cf. Figs. 8 and 12(a)).
lattice exhibited a reduction of 3.74%. However, the energy absorption In the SCO-plate lattice, the failure modes included the bending failure
abilities of the honeycomb lattice obtained from FE simulations and of the octahedral plates and the buckling failure of the vertical plates.
experiments showed good agreement. Importantly, the octahedral plates of the SCO-plate lattice experienced
The SCO-plate lattice showed significantly high energy absorption both axial compressive forces and bending moments. Since the strain
ability, surpassing the SC-plate and honeycomb lattices by 26.84% and energy stored due to bending moments is significantly higher than that
100.32% in FE results, respectively, and 23.37% and 93.63% in exper­ from the axial forces, the octahedral plates predominantly experienced
iments, respectively (cf. Figure 11(b)). Despite the differences in energy bending stresses rather than axial stresses. Figure 12(b) represents the
absorption abilities, the densification strain remained nearly the same in deformation behavior of the SCO-plate lattice with the von Mises stress
these three lattice structures (cf. Figure 11(c)). plot. As compression continued, localized buckling was observed at ∼14%
strain in the vertical plates in the FE simulation, as opposed to ~10.8%
4.6. Deformation behavior strain for the SC-plate lattice. Interestingly, even though the slenderness
ratio, i.e., the ratio between the thickness to the length of the horizontal
This section compares the deformation behavior of the SCO-plate and vertical plates, of the SCO-plate lattice (0.12) was smaller than that of
and SC-plate lattices having 40% RD obtained from the FE simulations the SC-plate lattice (0.16) ([Link] 1), the SCO-plate lattice exhibited
and the experiments. Considering the printing feasibility of these delayed buckling. This was attributed to the compressive forces shared by
structures, lattices were fabricated only at 40% RD and the mechanical the octahedral plates in the SCO-plate lattice.
test was performed only at this relative density. In the SC-plate, as severe buckling occurred, the stress level
In the SC-plate lattice, the buckling failure of the vertical plates was decreased in the stress-stain curve as the buckling members were unable
observed as a prominent failure mode, as depicted in Fig. 12(a). The to sustain the compressive stress. Therefore, flat plateau behavior was
localized buckling became noticeable at ∼10.8% strain in the FE simu­ observed at a strain of 1–3%. In the SCO-plate lattice, the compressive
lation. Severe buckling failure of the vertical plates can be seen at a load was shared by the both octahedral plates and the vertical plates
strain of 25% for both FE simulation and experiment, which is consistent which accounted for the slight strain hardening observed in the stress-
with the high-stress concentrations observed in the von Mises stress plot strain curve obtained from finite element results. In contrast, the
(cf. Figure 12(a)). In the stress-strain curves shown in Fig. 8, the SC-plate experimental stress-strain response of the SCO-plate lattice showed a
lattice displayed a plateau region of constant stress level without any plateau behavior due to defects associated with 3D printing. Notably,
fluctuations in the FE simulation. However, in the experimental stress- the SCO-plate lattice tends to have higher buckling when the relative
strain curve of the SC-plate lattice, slightly depressed troughs were density is lower, as seen in Fig. 12(C).

Fig. 13. Representation of degree of elastic isotropy: (a) Zener anisotropy plot, and (b), (c), (d) the spatial visualization elastic modulus of SC-plate lattice, SCO-plate
lattice and isotropic-strut lattice, respectively, all having a relative density of 40%.

14
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

4.7. Degree of isotropy of the SCO-plate lattice and its comparison with line perform similarly and those above the line are considered more
other lattices suitable than those below it [77].
The properties of SCO-plate lattice structures having 20%, 30% and
This section compares the degree of isotropy of the SCO-plate lattice 40% RD were plotted in the Ashby chart to compare its performance
structures with the SC-plate lattice and the isotropic-strut lattice through with the parent material PA12 and other known materials. For the
the Zener anisotropy ratio. In addition, the elastic modulus of these uniaxial loading condition, the performance index lines representing the
lattices in the 3D space was visualized with the help of homogenization parameter Eρ, i.e., the maximum elastic modulus for a minimum weight
and MATLAB code. design, and the parameter ρ, i.e., the maximum yield strength for
σy
The Zener anisotropy ratio (Z) for cubic symmetry materials can be
minimum weight design, in Fig. 14(a) and (b), respectively, were drawn
determined as follows [42]:
through the points corresponding to the parent material PA12 to
G compare the performance of the lattice structures with the parent ma­
Z= − 1 (37)
Giso terial. Based on the performance index line provided in Figure 14(a), the
elastic moduli of SCO-plate lattices aligned perfectly with the parent
where G is the shear modulus obtained from FE simulation (cf. material, exhibiting the same performance as the parent material and
Table S1). The Giso is the theoretical shear modulus evaluated from the outperforming most of the foams. Similarly, for yield strength, the SCO-
elastic modulus E and Poisson’s ratio ϑ relationship for isotropic mate­ plate lattice surpassed both the parent material and foams (cf. Fig. 14
rials as follows: (b)). Therefore, the SCO-plate lattice has the potential for lightweight
E applications requiring high stiffness and high strength.
Giso = (38)
2(1 + ϑ)
5. Conclusions
The material exhibited elastic isotropy when its value is equal to or
close to the isotropic line (Z = 0). In Fig. 13(a), the Z values of the SCO-
The study successfully designed a novel SCO-plate lattice that out­
plate lattices at the three aforementioned RDs, which overlap with each
performed the existing well-known lattices. The following are concluded
other, lie very close to the isotropic line.
in the study:
The three-dimensional spatial visualization of the elastic modulus of
The normalized elastic modulus and the normalized yield strength
the SCO-plate lattice, the SC-plate lattice and the isotropic-strut lattice at
values of the SCO-plate lattice exceeded the maximum theoretical
40% RD is also illustrated in Fig. 13(b). The SCO-plate lattice displays a
bounds, the Hashin-Shtrikmn upper bound and the Suquet upper bound,
nearly uniform surface of elastic modulus contrary to the highly
respectively. Moreover, this proposed lattice outperformed the highest-
direction-dependent elastic modulus of the SC-plate lattice. Although
performing SC-plate lattice by 19.5–20.0% in elastic modulus and
nearly isotropic behavior indicated by almost spherical elastic modulus
22.5–25.9% in yield strength in the finite element simulations for rela­
surface was achieved in the hybrid-plate lattices [41], modified
tive densities 20%, 30% and 40%. This is consistent with the experi­
octet-lattice [44] and the isotropic strut lattice, the elastic modulus and
ments, where the SCO-plate lattice exhibited 18.1% higher elastic
yield strength of these lattices are much lower than the proposed
modulus and 28.0% higher yield strength compared to the SC-plate
SCO-plate lattice. Thus, the proposed lattice not only outperformed the
lattice for a relative density of 40%.
SC-plate lattice in mechanical properties but also displayed a higher
Based on performance index lines in the Ashby charts for elastic
degree of isotropy compared to the SC-plate lattice.
modulus and yield strength, the SCO-plate lattice exhibited equal per­
formance in terms of elastic modulus and slightly higher yield strength
4.8. Ashby plot for elastic modulus and yield strength compared to the parent material PA12.
The energy absorption ability of the SCO-plate lattice was 26.84%
Ashby chart is a popular tool for visualizing the relationship between higher in the finite element simulation and 23.37% higher in the
two material properties in a two-dimensional space. These charts divide experiment compared to the SC-plate lattice with 40% relative density.
materials into distinct domains based on their properties and use Additionally, it outperforms the widely used honeycomb lattice by
guidelines or performance index lines for different loading conditions to 100.32% in finite element simulation and 93.63% in experiments at the
make material selection more efficient. Materials that lie on the same same density.

Fig. 14. Ashby charts showing the mechanical property of the SCO-plate lattice, having relative densities of 20%, 30% and 40%− (a) elastic modulus vs density, and
(b) yield strength vs density [78].

15
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

The SCO-plate lattice displayed a higher degree of isotropy than the [7] Meza LR, Das S, Greer JR. Strong, lightweight, and recoverable three-dimensional
ceramic nanolattices. Science 2014;345(6202):1322–6.
direction-dependent elastic behavior of most other lattices, including
[8] Deshpande VS, Ashby MF, Fleck NA. Foam topology: bending versus stretching
the SC-plate lattice, which makes the proposed lattice appropriate for dominated architectures. Acta Mater 2001;49:1035–40.
applications requiring isotropic characteristics. Thus, the proposed [9] Deshpande VS, Fleck NA, Ashby MF. Effective properties of the octet-truss lattice
simple-cubic-octahedral (SCO) plate lattice is suitable for applications material. J Mech Phys Solids 2001;49:1747–69.
[10] Leary M, Mazur M, Elambasseril J, McMillan M, Chirent T, Sun Y, Qian M,
that demand high elastic modulus, yield strength, energy absorption Easton M, Brandt M. Selective laser melting (SLM) of AlSi12Mg lattice structures.
ability and isotropic characteristics. Mater Des 2016;98:344–57.
The semi-analytical method used the traction boundary condition to [11] White BC, Garland A, Alberdi R, Boyce BL. Interpenetrating lattices with enhanced
mechanical functionality. Addit Manuf 2021;38:101741.
predict the elastic modulus of the SC-plate and SCO-plate lattices. [12] Vangelatos Z, Komvopoulos K, Grigoropoulos CP. Regulating the mechanical
However, this method has limitations in accurately determining all behavior of metamaterial microlattices by tactical structure modification. J Mech
elastic constants. To overcome this, periodic boundary conditions were Phys Solids 2020;144:104112. 2.
[13] Alberdi R, Dingreville R, Robbins J, Walsh T, White BC, Jared B, Boyce BL. Multi-
applied within the finite element framework. In addition, although the morphology lattices lead to improved plastic energy absorption. Mater Des 2020;
finite element simulations were performed for three relative densities, 194:108883.
the experiments were conducted only for one relative density due to [14] Osman MM, Shazly M, El-Danaf EA, Jamshidi P, Attallah MM. Compressive
behavior of stretched and composite microlattice metamaterial for energy
manufacturing constraints. Performing experiments at lower relative absorption applications. Compos Part B Eng 2020;184:107715.
density with design modifications can provide additional insights into [15] Zhang X, Jiang L, Yan X, Wang Z, Li X, Fang G. Regulated multi-scale mechanical
the deformation behavior of the lattice. performance of functionally graded lattice materials based on multiple bioinspired
patterns. Mater Des 2023;226:111564.
[16] Cao X, Duan S, Liang J, Wen W, Fang D. Mechanical properties of an improved 3D-
CRediT authorship contribution statement printed rhombic dodecahedron stainless steel lattice structure of variable cross
section. Int J Mech Sci 2018;145:53–63.
[17] Novak N, Krstulović-Opara L, Ren Z, Vesenjak M. Mechanical properties of hybrid
Manash Jyoti Baishya: Writing – original draft, Validation, Meth­ metamaterial with auxetic chiral cellular structure and silicon filler. Compos Struct
odology, Investigation, Formal analysis, Conceptualization. Nelson 2020;234:111718.
Muthu: Writing – review & editing, Visualization, Supervision, Soft­ [18] Alawwa F, Barsoum I, Abu Al-Rub RK. Modeling, testing, and optimization of novel
lattice structures for enhanced mechanical performance. Mech Adv Mater Struct
ware, Methodology, Conceptualization. Prasenjit Khanikar: Writing –
2023;0:1–24.
review & editing, Visualization, Supervision, Methodology, [19] Pham MS, Liu C, Todd I, Lertthanasarn J. Damage-tolerant architected materials
Conceptualization. inspired by crystal microstructure. Nature 2019;565:305–11.
[20] Xiao L, Xu X, Song W, Hu M. A multi-cell hybrid approach to elevate the energy
absorption of micro-lattice materials. Materials (Basel) 2020;13:4083.
Declaration of competing interest [21] Guo W, Huang Y, Ritchie RO, Yin S. Dissipative dual-phase mechanical
metamaterial composites via architectural design. Extreme Mech Lett 2021;48:
101442.
The authors declare that they have no known competing financial [22] Yin S, Guo W, Wang H, Huang Y, Yang R, Hu Z, Chen D, Xu J, Ritchie RO. Strong
interests or personal relationships that could have appeared to influence and tough bioinspired additive-manufactured dual-phase mechanical metamaterial
composites. J Mech Phys Solids 2021;149:104341. 1.
the work reported in this paper. [23] Baishya MJ, Sahariah BJ, Muthu N, Khanikar P. High-Stiffness Metamaterial
Composite Structure with Plate-Reinforced Strut-Microlattice. TMS 2022 151st
Data availability annu meet exhib suppl proc. Springer; 2022. p. 1551–68.
[24] Sahariah BJ, Namdeo A, Khanikar P. Composite-inspired multilattice metamaterial
structure: an auxetic lattice design with improved strength and energy absorption.
Data will be made available on request. Mater Today Commun 2022;30:103159.
[25] Baishya MJ, Sahariah BJ, Muthu N, Khanikar P. Composite strut-plate lattice: a
high-stiffness design of cellular metamaterial having excellent strength and energy
absorption ability. Mater Today Commun 2022;33:104939.
Acknowledgement [26] Sahariah BJ, Baishya MJ, Namdeo A, Khanikar P. A novel strategy to design lattice
structures with zero Poisson’s ratio. Eng Struct 2023;288:116214.
[27] Feng J, Fu J, Yao X, He Y. Triply periodic minimal surface (TPMS) porous
This work was partially supported by the Science and Engineering structures: from multi-scale design, precise additive manufacturing to
Research Board [Project No. CRG/2022/007441]. The authors also multidisciplinary applications. Int J Extreme Manuf 2022;4:022001.
acknowledge the use of Shimadzu AGS-X 100 kN UTM, at the Depart­ [28] Han L, Che S. An overview of materials with triply periodic minimal surfaces and
related geometry: from biological structures to self-assembled systems. Adv Mater
ment of Mechanical Engineering, IIT Guwahati, procured through DST-
2018;30:1705708.
FIST (Project No. SR/FST/ET-II/2017/111 (C)). [29] Liu F, Gan D, Zhang K, Zhang P, Zhang C, Li S, Xie H, Tang K. A selective strategy
for determining suitable structure and fatigue behavior study of minimal surface
scaffolds fabricated by EBM. Int J Fatigue 2023;167:107380.
Supplementary materials [30] Ma X, Zhang DZ, Zheng X. Revealing the excellent properties of minimal surface
lattice structures based on additive manufacturing through the principle of least
Supplementary material associated with this article can be found, in action. Int J Adv Manuf Technol 2022;121:1575–88.
[31] Yang L, Yan C, Han C, Chen P, Yang S, Shi Y. Mechanical response of a triply
the online version, at doi:10.1016/[Link].2024.109189.
periodic minimal surface cellular structures manufactured by selective laser
melting. Int J Mech Sci 2018;148:149–57.
References [32] Yang X, Yang Q, Shi Y, Yang L, Wu S, Yan C, Shi Y. Effect of volume fraction and
unit cell size on manufacturability and compressive behaviors of Ni-Ti triply
periodic minimal surface lattices. Addit Manuf 2022;54:102737.
[1] Nian Y, Wan S, Wang X, Zhou P, Avcar M, Li M. Study on crashworthiness of
[33] Almesmari A, Baghous N, Ejeh CJ, Barsoum I, Abu Al-Rub RK. Review of Additively
nature-inspired functionally graded lattice metamaterials for bridge pier protection
Manufactured Polymeric Metamaterials: design, Fabrication, Testing and
against ship collision. Eng Struct 2023;277:115404.
Modeling. Polymers (Basel) 2023;15:3858.
[2] Ngo TD, Kashani A, Imbalzano G, Nguyen KTQ, Hui D. Additive manufacturing (3D
[34] Almesmari A, Sheikh-Ahmad J, Jarrar F, Bojanampati S. Optimizing the specific
printing): a review of materials, methods, applications and challenges. Compos
mechanical properties of lattice structures fabricated by material extrusion
Part B Eng 2018;143:172–96.
additive manufacturing. J Mater Res Technol 2023;22:1821–38.
[3] Ashby MF. The properties of foams and lattices. Philos Trans Royal Soc A Philos T R
[35] Xue R, Cui X, Zhang P, Liu K, Li Y, Wu W, Liao H. Mechanical design and energy
Soc A 2006;364:15–30. Jan 151838.
absorption performances of novel dual scale hybrid plate-lattice mechanical
[4] Abou-Ali AM, Al-Ketan O, Lee DW, Rowshan R, Al-Rub RKA. Mechanical behavior
metamaterials. Extreme Mech Lett 2020;40:100918.
of polymeric selective laser sintered ligament and sheet based lattices of triply
[36] Crook C, Bauer J, Guell Izard A, Santos de Oliveira C, Martins de Souza e Silva J,
periodic minimal surface architectures. Mater Des 2020;196:109100.
Berger JB, Valdevit L. Plate-nanolattices at the theoretical limit of stiffness and
[5] Shen M, Qin W, Xing B, Zhao W, Gao S, Sun Y, Jiao T, Zhao Z. Mechanical
strength. Nat Commun 2020;11:1579.
properties of 3D printed ceramic cellular materials with triply periodic minimal
[37] Hu J, Tan ATL, Chen H, Hu X. Superior compressive properties of 3D printed plate
surface architectures. J. Eur. Ceram. Soc. 2021;41:1481–9.
lattice mechanical metamaterials. Int J Mech Sci 2022;231:107586.
[6] Zhang X, Vyatskikh A, Gao H, Greer JR, Li X. Lightweight, flaw-tolerant, and
ultrastrong nanoarchitected carbon. Proc Natl Acad Sci 2019;116:6665–72.

16
M.J. Baishya et al. International Journal of Mechanical Sciences 272 (2024) 109189

[38] Berger JB, Wadley HNG, McMeeking RM. Mechanical metamaterials at the [57] O’Connor HJ, Dickson AN, Dowling DP. Evaluation of the mechanical performance
theoretical limit of isotropic elastic stiffness. Nature 2017;543:533–7. of polymer parts fabricated using a production scale multi jet fusion printing
[39] Li X, Yu X, Chua JW, Lee HP, Ding J, Zhai W. Microlattice metamaterials with process. Addit Manuf 2018;22:381–7.
simultaneous superior acoustic and mechanical energy absorption. Small 2021;17: [58] Bathe KJ. Finite element procedures. Klaus-Jurgen Bathe 2006.
2100336. [59] Terada K, Hori M, Kyoya T, Kikuchi N. Simulation of the multi-scale convergence
[40] Bonatti C, Mohr D. Smooth-shell metamaterials of cubic symmetry: anisotropic in computational homogenization approaches. Int J Solids Struct 2000;37:
elasticity, yield strength and specific energy absorption. Acta Mater 2019;164: 2285–311.
301–21. [60] Qu J, Cherkaoui M. Fundamentals of micromechanics of solids. Wiley Online
[41] Tancogne-Dejean T, Diamantopoulou M, Gorji MB, Bonatti C, Mohr D. 3D plate- Library 2006.
lattices: an emerging class of low-density metamaterial exhibiting optimal isotropic [61] Sun CT, Vaidya RS. Prediction of composite properties from a representative
stiffness. Adv Mater 2018;30:1803334. volume element. Compos Sci Technol 1996;56:171–9.
[42] Liu Y. Mechanical properties of a new type of plate–lattice structures. Int J Mech [62] Wan Hu J, Park T. Continuum models for the plastic deformation of octet-truss
Sci 2021;192:106141. lattice materials under multiaxial loading. J Eng Mater Technol 2013;135.
[43] Li T, Jarrar F, Al-Rub RA, Cantwell W. Additive manufactured semi-plate lattice [63] Kumar A, Muthu N, Narayanan RG. Equivalent orthotropic properties of periodic
materials with high stiffness, strength and toughness. Int J Solids Struct 2021;230: honeycomb structure: strain-energy approach and homogenization. Int J Mech
111153. Mater Des 2023;19:137–63.
[44] Feng J, Liu B, Lin Z, Fu J. Isotropic octet-truss lattice structure design and [64] F. Ye, H. Wang. A simple Python code for computing effective properties of 2D and
anisotropy control strategies for implant application. Mater Des 2021;203:109595. 3D representative volume element under periodic boundary conditions. ArXiv
[45] Almesmari A, Barsoum I, Abu Al-Rub RK. Modelling, optimization, and testing of Prepr. ArXiv170303930. (2017).
novel cuboidal spherical plate lattice structures. Virtual Phys Prototyp 2024;19: [65] Kurukuri S. A comprehensive study: boundary conditions for representative
e2308514. volume elements (RVE) of composites. Inst Struct Mech 2013.
[46] Abueidda DW, Abu Al-Rub RK, Dalaq AS, Lee DW, Khan KA, Jasiuk I. Effective [66] Omairey SL, Dunning PD, Sriramula S. Development of an ABAQUS plugin tool for
conductivities and elastic moduli of novel foams with triply periodic minimal periodic RVE homogenisation. Eng Comput 2019;35:567–77.
surfaces. Mech Mater 2016;95:102–15. [67] El-Sayed MA, Essa K, Ghazy M, Hassanin H. Design optimization of additively
[47] Yan C, Hao L, Hussein A, Young P. Ti–6Al–4V triply periodic minimal surface manufactured titanium lattice structures for biomedical implants. Int J Adv Manuf
structures for bone implants fabricated via selective laser melting. J Mech Behav Technol 2020;110:2257–68.
Biomed Mater 2015;51:61–73. [68] Sahariah BJ, Dalakoti M, Khanikar P. A high-performance design of tubular lattice
[48] Gümrük R, Mines RAW, Karadeniz S. Static mechanical behaviours of stainless steel structure having zero Poisson’s ratio. Mech Adv Mater Struct 2024;0:1–21.
micro-lattice structures under different loading conditions. Mater Sci Eng A 2013; [69] Toutenburg H. Shalabh. statistical analysis of designed experiments. 3rd ed. New
586:392–406. York, NY: Springer; 2009.
[49] Namdeo A, Bhandare V, Sahariah BJ, Khanikar P. Tetrahedral and strut-reinforced [70] Chen L, Liu Z, Sun P, Huo W. Formulation of a fuel spray SMD model at
tetrahedral microlattices: selectively laser melted high-strength and high-stiffness atmospheric pressure using Design of Experiments (DoE). Fuel 2015;153:355–60.
cellular metamaterials. Mater Sci Eng A 2022;855:143878. [71] Yang C, Xu P, Yao S, Xie S, Li Q, Peng Y. Optimization of honeycomb strength
[50] Khiavi SG, Sadeghi BM, Divandari M. Effect of topology on strength and energy assignment for a composite energy-absorbing structure. Thin-Walled Struct 2018;
absorption of PA12 non-auxetic strut-based lattice structures. J Mater Res Technol 127:741–55.
2022;21:1595–613. [72] Mohammadi H, Ahmad Z, Petrů M, Mazlan SA, Faizal Johari MA, Hatami H,
[51] Li P, Ma YE, Sun W, Qian X, Zhang W, Wang Z. Mechanical behaviors and failure Rahimian Koloor SS. An insight from nature: honeycomb pattern in advanced
modes of additive manufactured Ti6Al4V lattice structures under compressive structural design for impact energy absorption. J Mater Res Technol 2023;22:
load. Thin-Walled Struct 2022;180:109778. 2862–87.
[52] Ataee A, Li Y, Fraser D, Song G, Wen C. Anisotropic Ti-6Al-4V gyroid scaffolds [73] Chen LY, Liang SX, Liu Y, Zhang LC. Additive manufacturing of metallic lattice
manufactured by electron beam melting (EBM) for bone implant applications. structures: unconstrained design, accurate fabrication, fascinated performances,
Mater Des 2018;137:345–54. and challenges. Mater Sci Eng R Rep 2021;146:100648.
[53] Yan C, Hao L, Hussein A, Bubb SL, Young P, Raymont D. Evaluation of light-weight [74] Echeta I, Feng X, Dutton B, Leach R, Piano S. Review of defects in lattice structures
AlSi10Mg periodic cellular lattice structures fabricated via direct metal laser manufactured by powder bed fusion. Int J Adv Manuf Technol 2020;106:2649–68.
sintering. J Mater Process Technol 2014;214:856–64. [75] Hashin Z, Shtrikman S. A variational approach to the theory of the elastic
[54] Arabnejad S, Burnett Johnston R, Pura JA, Singh B, Tanzer M, Pasini D. High- behaviour of multiphase materials. J Mech Phys Solids 1963;11:127–40.
strength porous biomaterials for bone replacement: a strategy to assess the [76] Gibson LJ, Ashby MF. Cellular solids: structure and properties. 2nd ed. Cambridge:
interplay between cell morphology, mechanical properties, bone ingrowth and Cambridge University Press; 1999.
manufacturing constraints. Acta Biomater 2016;30:345–56. [77] MF Ashby. Materials selection in mechanical design. 5th ed. Elsevier; 2017.
[55] Yan C, Hao L, Hussein A, Raymont D. Evaluations of cellular lattice structures [78] Fleck NA, Deshpande VS, Ashby MF. Micro-architectured materials: past, present
manufactured using selective laser melting. Int J Mach Tools Manuf 2012;62:32–8. and future. Proc R Soc Math Phys Eng Sci 2010;466:2495–516.
[56] Habib FN, Iovenitti P, Masood SH, Nikzad M. Fabrication of polymeric lattice
structures for optimum energy absorption using Multi Jet Fusion technology. Mater
Des 2018;155:86–98.

17