International Journal of Mechanical Sciences 275 (2024) 109309

Contents lists available at ScienceDirect

International Journal of Mechanical Sciences

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

4D-printed reusable metamaterial via shape memory effect for

energy dissipation
Di Zhang a, Meiyu Li a, Na Qiu b, Jie Yang b, Chi Wu c, Grant Steven c, Qing Li c,
Jianguang Fang a, *
a
School of Civil and Environmental Engineering, University of Technology Sydney, Sydney, NSW 2007, Australia
b
Mechanical and Electrical Engineering College, Hainan University, Haikou 570228, China
c
School of Aerospace, Mechanical and Mechatronic Engineering, The University of Sydney, Sydney, NSW 2006, Australia

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

Keywords: The present study aims at developing reusable metamaterials fabricated by 4D printing technology. Honeycomb
4D printing metamaterials were manufactured via fused deposition modeling (FDM) with shape memory polymers (SMPs).
Honeycomb metamaterials The reusability of these metamaterials was determined through cyclic cold programming experiments, where
Energy dissipation
each cycle involved a loading-unloading-heating (shape recovery)-cooling process. The novelty of this paper lies
Shape memory effect
not only in experimentally demonstrating the recoverability of metamaterials by reversing plastic deformation
Reusability
Mechanical degradation based on the shape memory effect of SMPs, but also in studying their reusability of SMP metamaterials under
cyclic programming and the effect of printing materials and unit-cell types on the mechanical degradation. The
results reveal that, under one single compression cycle, the polylactic acid (PLA) hexagonal honeycomb dissi­
pated 22% more energy than the polyethylene terephthalate glycol (PETG) counterpart because the higher elastic
modulus of PLA leads to a larger critical buckling load for segments in honeycomb structures. Furthermore, the
PETG re-entrant honeycomb dissipated 25% more energy than the hexagonal counterpart due to its negative
Poisson’s ratio and the overall uniform deformation pattern. More importantly, it is found that under multiple
compression cycles, the PETG hexagonal honeycomb maintained an energy dissipation capacity of 78.3% at
Cycle 6, nearly 3.5 times that of the PLA counterpart as a result of the better ductility of PETG. Moreover, the
PETG re-entrant honeycomb could be reused for 17 cycles, while the hexagonal counterpart could only be reused
for 12 cycles. This is because the re-entrant unit cells are failure-resistant and of less concentration in plastic
deformation. The results demonstrate that the constituent materials with better ductility and the unit-cells with
more failure resistance can reduce mechanical degradation, thereby exhibiting better reusability of
metamaterials.

1. Introduction manufacturing, or 3D printing, has made it possible to manufacture

metamaterials with complex topologies, greatly facilitating the studies
Mechanical metamaterials exhibit exceptional physical behaviors, in mechanical metamaterials. 3D printing is a computer-aided
usually unavailable in nature and counterintuitive in practice [1–5]. manufacturing process that enables the building of three-dimensional
Fueled by fast-expanding design and fabrication capacity, researchers structures by feeding, bonding, and solidifying constituent materials in
have been able to conceive some novel metamaterials over the last few a layer-by-layer fashion [19]. With the help of 3D printing, mechanical
decades, such as different auxeticities [6–9], negative mechanical metamaterials have been applied in automotive [9,14,20–22], aero­
properties [10–13], and extremal multifunctional characteristics (e.g., space [23–25], and biomedical engineering [3,26–28].
stiffness, strength, low mass density, etc.) [14–16]. One symbolic application for mechanical metamaterials is energy
Normally, traditional manufacturing technologies are not applicable dissipation because their light weight can potentially offer unrivaled
in fabricating mechanical metamaterials due to their sophisticated to­ specific energy absorption (SEA) [4]. The energy dissipation in tradi­
pological structures [4,17,18]. However, the emergence of additive tional mechanical metamaterials is typically achieved through severe

* Corresponding author.
E-mail address: [email protected] (J. Fang).

https://doi.org/10.1016/j.ijmecsci.2024.109309
Received 19 September 2023; Received in revised form 30 March 2024; Accepted 19 April 2024
Available online 20 April 2024
0020-7403/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

plastic deformation and material damage, both of which are permanent the design and fabrication of honeycomb metamaterials, as well as the
and irreversible to those materials [29,30]. Therefore, traditional me­ experimental protocols for the cyclic cold programming tests. Section 3
chanical metamaterials used for energy dissipation often exhibit char­ presents the mechanical analysis of the metamaterials under one-off
acteristics of one-off service life. In order to achieve the purpose of compression tests and elaborates the reusability of metamaterials
recurring energy dissipation in mechanical metamaterials, alternative under cyclic loading process. Section 4 summarizes the research findings
energy dissipation mechanisms have been explored [31], such as friction and provides an outlook for future research.
[30,32,33], instability-induced energy trapping and dissipation
[34–38], or combination of them [39,40]. However, while achieving 2. Materials and methods
reusability, instability-based metamaterials commonly exhibit poor
load-bearing capacity because they require structures with a small In this section, the design method for honeycomb metamaterials with
thickness ratio [30], and friction-based metamaterials may suffer from different Poisson’s ratios is introduced first. Then, the 4D printing pro­
surface wear [32]. More importantly, such metamaterials may offer cess of FDM fabricated honeycomb samples using SMPs as printing
limited energy dissipation, as in most cases, their constituent materials materials is explained. Last, the detailed steps involved in the cyclic cold
are not allowed to go beyond the elastic region [30,32,36]. programming process are presented.
To achieve the goal of reusability and significant energy dissipation,
one of the most straightforward approaches may be to find constituent
materials that are capable of reversing plastic deformation. The emer­ 2.1. Design of honeycomb structures
gence of 4D printing has made this approach possible [41–45]. Unlike
the traditional materials used in 3D printing, the materials adopted by The metamaterials to be investigated in this study are honeycomb
4D printing are normally stimulus-responsive, allowing fabricated structures for their characteristic features and practical popularity. They
samples to change their shape even after the printing process. Shape can be easily fabricated by FDM printers or other techniques. Honey­
memory polymers (SMPs) signify one of the most popular and widely combs with three different unit-cell structures were considered,
used constituent materials in 4D printing. SMPs are heat-responsive and including the hexagonal honeycomb with positive Poisson’s ratio, the
capable of recovering their plastic deformation through the shape hybrid honeycomb with zero Poisson’s ratio, and the re-entrant hon­
memory effect (SME) after a simple heating process [46–48]. eycomb with negative Poisson’s ratio. Fig. 1(a) illustrates the geometric
In recent years, some SMP metamaterials capable of dissipating en­ configurations of these three honeycombs. Note that a width (building
ergy by reversing plastic deformation through the shape memory effect height) of 12 mm was used here to avoid out-of-plane instability during
have been reported. For example, Namvar et al. [29] fabricated SMP in-plane compressions.
honeycomb metamaterials with different Poisson’s ratios, all of which
exhibited a 100% shape recovery after a simple heating process. Xu et al.
[49] also investigated these honeycomb metamaterials and found the 2.2. 4D fabrication materials and method
geometric configuration had an impact on their shape recoverability.
Wan et al. [50] fabricated cylindrical SMP metamaterials with Shape memory polymers (SMPs) were used as the printing materials
horseshoe-shaped unit cells and demonstrated their capability to recover in this study as they can recover plastic deformation through the shape
to their original shape after stretching or bending. Zeng et al. [51] also memory effect (SME), which enables the fabricated honeycombs to
studied horseshoe-shaped metamaterials and reported that using dissipate energy repetitively. The mechanism of the shape memory ef­
continuous fiber-reinforced composites (CFRCs) as the printing material fect on the molecular level is illustrated in Fig. 2, where SMPs are
not only provided remarkable shape recovery but also enhanced the regarded as networks composed of long molecular chains and netpoints.
energy dissipating capability for metamaterials. Dong et al. [52] re­ These netpoints are crosslinked either covalently or physically and
ported a shape recovery ratio above 90% over the first six cycles on the function as switches [58]. Below a critical temperature known as the
metamaterials printed by CFRCs. Wan et al. [53] fabricated SMP origami glassy transition temperature (Tg), the crosslinked netpoints are inactive
metamaterials capable of recovering from the temporary folded shapes and thus immobilize the movement of molecular chains, resulting in a
to the original shapes after reheating. Hamzehei et al. [54] revealed the glassy phase [46]. When SMPs are heated above Tg, the netpoints switch
good recoverability of the zero Poisson’s ratio (ZPR) metamaterials and open, and the polymeric chains become flexible, leading to a rubbery
recognized the mechanical degradation of metamaterials from the phase [47]. Therefore, the plastic deformation introduced in the glassy
force-displacement curves through two consecutive compression cycles. phase of SMPs is recoverable once they are heated above Tg because
However, the above-mentioned studies mainly focused on the shape polymeric chains are mobilized in the rubbery phase.
recoverability of metamaterials and have not extensively explored their PLA (JGMAKER, China) and PETG (Polymaker, China) filaments
reusability subject to multiple loading cycles. The reusability of meta­ were used to fabricate the honeycombs to investigate the effect of
materials indicates their capability to maintain the energy dissipation printing materials on the reusability of fabricated metamaterials. Ac­
level after being mechanically deformed and thermally recovered. In cording to the material datasheet provided by the manufacturers, PLA
other words, SMP-based metamaterials can be only regarded as reusable has a Tg of approximately 65 ◦ C, while PETG has a Tg of around 85 ◦ C.
when they exhibit minor mechanical degradation after loading- Ultimaker S3 FDM printer (Ultimaker, The Netherlands) was used to
unloading-heating cycles. To date, there have been few studies on the fabricate the designed honeycomb metamaterials. First, the CAD models
reusability of SMP metamaterials [55–57]. Moreover, the impact of the were created in SolidWorks and then exported as STL files. These STL
constituent material and the unit-cell geometry on the reusability of files were sliced using Cura (Ultimaker, Netherlands) to generate G-code
these metamaterials remains unclear. files for 3D printing. The key printing parameters were set as follows: a
This study aims to develop reusable energy-dissipating meta­ layer height of 0.4 mm, a line width of 0.4 mm, an infill line distance of
materials with different constituent materials and unit-cell configura­ 0.4 mm, a concentric infill pattern, a printing speed of 50 mm/s, a nozzle
tions by carrying out cyclic compression-recovery tests on the temperature of 200 ◦ C for PLA and 240 ◦ C for PETG, respectively, and a
metamaterials. Factors affecting the mechanical degradation of meta­ building plate temperature of 60 ◦ C for PLA and 80 ◦ C for PETG,
materials during cyclic usage are revealed by comprehensively respectively. Finally, the G-code files were uploaded to the FDM printer
recording and analyzing the performance of metamaterials in each for printing. Fig. 1(b) shows that a concentric infill pattern and an infill
cycle. It is anticipated to gain new insight and develop a framework for line distance of 0.4 mm result in a negative airgap value of − 0.05 mm,
the design of reusable lightweight metamaterials for energy dissipation. which allows better mechanical performance of the fabricated honey­
The remainder of this paper is organized as follows. Section 2 introduces combs [59].

2
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 1. (a) Structural profiles and dimensions of honeycomb metamaterials: the hexagonal honeycomb with positive Poisson’s ratio (left), the hybrid honeycomb
with zero Poisson’s ratio (middle), and the re-entrant honeycomb with negative Poisson’s ratio (right). (b) Infill pattern and air gap value for the FDM print­
ing process.

Fig. 2. Molecular mechanism of the thermally induced shape memory effect (SME) of shape memory polymers, Tg = glass transition temperature related to
phase changes.

2.3. Experiment procedure and equipment setup • Step 2: Unloading process – The applied force was removed, allowing
the metamaterial samples to reverse elastic deformation.
The reusability of metamaterials was characterized through a series • Step 3: Shape recovery process – The metamaterial samples were
of cold programming cycles. Fig. 3(a) illustrates the process of a single heated above Tg to trigger the shape memory effect of the SMPs,
cold programming cycle, comprising four major steps [46]: leading them to recover their pre-programmed shape.
• Step 4: Cooling process - The heated samples were cooled down to
• Step 1: Loading process – The metamaterial samples were com­ their initial state (room temperature).
pressed at room temperature, below Tg of the SMPs.

3
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 3. (a) Schematic of a single cold-programming cycle. (b) Experimental setup for the compression tests. The cold programming of metamaterials is a loading-
unloading-recovery-cooling process.

Fig. 3(b) illustrates the experimental setup for the compression tests. printing materials and unit-cell structures.
The loading in Step 1 and unloading in Step 2 were performed on an
Autograph AGX universal testing machine (Shimadzu, Japan) at a speed 3.1.1. Metrics for mechanical performance analysis
of 3 mm/min to ensure quasi-static conditions and eliminate viscosity The force-displacement responses obtained from the one-off cold
dependency by applying a low strain rate [60]. The deformation of programming tests were used to analyze the mechanical characteristics
metamaterial samples during the loading-unloading process was recor­ of honeycombs. Fig. 4(a) illustrates a typical force-displacement curve
ded using a MONET-3D camera (Sobriety, Czech Republic) at a rate of after one loading-unloading cycle. The loading curve starts from an
0.5 frames per second, ensuring that shape changes resulting from every elastic region, but reduces after the force reaches a peak, and then ex­
0.1 mm displacement were captured. In Step 3, the heating of the met­ periences a long plastic plateau. The load starts to be removed once the
amaterial samples was achieved by placing the samples into a heating maximum displacement is reached, which also marks the beginning of
medium (water or air) with a temperature of 95 ◦ C (above Tg). The the unloading curve. The unloading curve ends with a displacement
heating process concluded when no further shape recovery was significantly greater than zero, indicating that the metamaterial has
observed from the samples. This study selected water-bath heating as the undergone significant plastic deformation and, therefore, cannot
heating method for the cyclic cold programming tests as the cyclic re­ recover most of the compressed height.
sults revealed that water was not only more effective but also allowed Four metrics are used to quantify the mechanical performance of
for better shape recovery of honeycombs than air when working as the honeycombs, including structural elastic stiffness (Kstru), initial peak
heating medium, as detailed in Section 3.2.2. force (Fpeak), plateau force (Fpl), and total energy dissipation (Ediss). Kstru
is determined from the elastic region of the force-displacement curve.
3. Results and discussion Fpeak is the first peak reached by the force. Fpl is the average force in the
plateau region, which is defined as the region between the displacement
In this section, the effect of printing materials and unit-cell structures at Fpeak and the maximum loading displacement. Ediss is defined as the
on the mechanical performance and reusability of honeycomb meta­ area between the loading and unloading curves of the force-
materials are investigated. Two polymers (i.e., PLA and PETG) and three displacement response, as shadowed in Fig. 4(a).
unit-cell structures (i.e., hexagonal, hybrid, and re-entrant honeycombs)
were considered. Section 3.1 presents the results of all honeycombs 3.1.2. Effect of printing materials
under the first cycle of compression to analyze the effect of these two The tensile properties of PLA and PETG were determined from uni­
factors (constituent materials and unit-cell structures) on the mechani­ axial tensile tests of material specimens in line with ASTM standard
cal performance of metamaterials for one-off use. Section 3.2 analyzes D638. All the uniaxial tension specimens were designed and fabricated
the data derived from cyclic cold programming tests to investigate the by using type IV geometry with a thickness of 4 mm. The same printing
effect of them on the reusability of metamaterials under multiple cycles. parameters and FDM printer used for honeycomb samples (as detailed in
Section 2.2) were adopted to fabricate uniaxial tension specimens,
except that a line infill pattern with a raster angle of 0◦ (along the length
3.1. One-off mechanical performance of metamaterials direction) was used. Fig. 4(b) compares the tensile behavior between
PLA and PETG, where PLA exhibits a greater elastic modulus (1.47 GPa)
This section discusses the effects of printing materials and unit-cell than PETG (0.70 GPa) and a higher tensile strength (43.5 MPa) than
structures on the one-off mechanical characteristics of honeycomb PETG (34.5 MPa). However, PETG shows better ductility with an elon­
metamaterials. All the experimental data was derived from one gation at a break of 7.6%, whereas PLA breaks at 4.8% elongation.
compression cycle with a loading distance of 12 mm (approximately The effect of printing materials on the mechanical behavior of met­
34% loading strain). The metrics used for analyzing the mechanical amaterials was characterized by the testing results of honeycomb
behavior are introduced first, followed by scrutinizing the effect of

4
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 4. (a) Typical force-displacement curve and the metrics for one-off mechanical performance analysis. (b) Strain-stress responses of the PLA and PETG tensile
samples. PLA has higher elastic modulus and strength, while PETG offers better ductility. (c) The force-displacement responses of the PLA hexagonal honeycomb and
the PETG hexagonal honeycomb from Cycle 1. The PLA hexagonal honeycomb has better mechanical performance than the PETG hexagonal honeycomb. (d) The
force-displacement responses of the PETG hexagonal honeycomb, the PETG hybrid honeycomb, and the PETG re-entrant honeycomb from Cycle 1. The PETG re-
entrant honeycomb has the best mechanical performance, followed by the PETG hybrid honeycomb, and then the PETG hexagonal honeycomb.

samples with the same unit-cell structure but printed using different from these two different honeycombs during the first cold programming
materials, i.e., the PLA hexagonal honeycomb and the PETG hexagonal cycle. Fig. 5 compares the mechanical metrics between these two hon­
honeycomb. Fig. 4(c) plots the force-displacement responses obtained eycombs derived from the force-displacement curves in the first cycle.

Fig. 5. Comparison of the mechanical metrics of hexagonal honeycombs printed with different constituent materials in the first cold programming cycle: (a)
structural stiffness, (b) initial peak force, (c) plateau force, (d) energy dissipation. The PLA hexagonal honeycomb outperforms the PETG counterpart, indicating
constituent materials have an effect on the mechanical performance of metamaterials.

5
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Apparently, the PLA hexagonal honeycomb exhibits certain advantages first cold programming cycle for the hexagonal, the hybrid, and the re-
in terms of all the metrics over the PETG counterpart. Fig. 5(a) shows entrant honeycombs.
that the PLA hexagonal honeycomb had a structural stiffness of 502 N/ Fig. 6 compares the derived mechanical metrics of these three PETG
mm in the first cycle, which was 39% higher than that of the PETG honeycombs in the first cycle. Fig. 6(a) shows that the re-entrant hon­
hexagonal honeycomb (302 N/mm). Fig. 5(b) indicates the initial peak eycomb exhibited the highest structural stiffness of 692 N/mm. In
force of the PLA hexagonal honeycomb in the first cycle was 734 N, 25% comparison, the hybrid and the hexagonal honeycombs were of struc­
higher than that of the PETG counterpart (548 N). Fig. 5(c) shows the tural stiffnesses of 493 N/mm and 305 N/mm, respectively, which were
PLA hexagonal honeycomb exhibited a plateau force of 621 N in the first 29% and 56% lower than the re-entrant counterpart. Fig. 6(b) shows
cycle, 24% higher than that of the PETG counterpart (472 N). Fig. 5(d) that the re-entrant honeycomb had an initial peak force of 766 N in the
reveals that the energy dissipation capacity of the PLA hexagonal hon­ first cycle, which was 10% and 28% higher than those of the hybrid (686
eycomb during the first cycle was 6.09 J, which is 22% higher than that N) and the hexagonal (548 N) counterparts, respectively. Fig. 6(c) in­
of the PETG hexagonal honeycomb (4.72 J). dicates that the re-entrant honeycomb provided a plateau force of 595 N,
The PLA metamaterial exhibited a higher structural stiffness might which was 1% and 21% higher than those of the hybrid (591 N) and
be due to its higher elastic modulus (c.f. Fig. 5(b)), which also achieved hexagonal (472 N) counterparts, respectively. Fig. 6(d) reveals that the
a higher initial peak force and higher plateau force. 3D-printed meta­ re-entrant honeycomb dissipated 6.30 J energy during the first cycle. In
materials are believed to inherit the mechanical properties of their contrast, the hybrid and hexagonal honeycombs dissipated 6.19 J (2%
constituent materials. The slender wall segments of honeycombs expe­ lower) and 4.72 J (25% lower) of energy, respectively.
rienced a series of local bending/buckling during compression, which The quantitative comparison reveals that the re-entrant honeycomb
can be observed from both experimental results and numerical results exhibited the best mechanical performance, followed by the hybrid and
(to be shown in Figs. 7 and 8). Since the critical buckling load is pro­ then the hexagonal counterparts. The main reason is that these three
portional to the elastic modulus of constituent material, the PLA hon­ honeycombs have distinct deformation patterns due to different Pois­
eycomb can withstand a higher load than the PETG honeycomb prior to son’s ratios.
local buckling, resulting in a greater initial peak force and greater For the hexagonal honeycomb, the material expanded horizontally
plateau force. Note that the higher structural stiffness, greater initial under vertical compression due to a positive Poisson’s ratio. This led to
peak force and plateau force of the PLA honeycomb contribute to a the formation of a shear band along the diagonal direction of the hon­
larger area enclosed by the loading-unloading curve. As a result, the PLA eycomb structure during compression, as shown in Fig. 7(a). The pres­
honeycomb dissipated more energy than the PETG counterpart during ence of the shear band significantly reduced the stiffness and strength of
the first cold programming cycle. the hexagonal honeycomb, as the majority of plastic deformation was
concentrated in the shear band. As a result, the hexagonal honeycomb
3.1.3. Effect of unit-cell structures dissipated the least energy in the first cycle of compression among these
In addition to the printing materials, the unit-cell structures also play three honeycomb configurations.
a significant role in impacting the mechanical characteristics of hon­ For the re-entrant honeycomb, however, its overall negative Pois­
eycomb metamaterials. The effect of unit-cell structures was analyzed son’s ratio caused the structure to contract horizontally when being
based on the PETG honeycomb samples with different unit-cell struc­ compressed vertically. Fig. 7(c) shows that the material continuously
tures. Fig. 4(b) plots the force-displacement responses obtained from the flew towards the central region of the re-entrant honeycomb during the

Fig. 6. Mechanical metrics for PETG metamaterials printed with different unit-cell configurations in the first cold programming compression cycle: (a) structural
stiffness, (b) initial peak force, (c) plateau force, (d) energy dissipation. The re-entrant honeycomb outperforms the hybrid honeycomb, which in turn outperforms the
hexagonal honeycomb, indicating unit-cell geometry has an effect on the mechanical performance of metamaterials.

6
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 7. Deformation patterns of the honeycombs with different unit-cell configurations during compression: (a) the hexagonal honeycomb with deformation
concentrated on a shear band, (b) the hybrid honeycomb with deformation occurred in the top and bottom layers of unit-cells, (c) the re-entrant honeycomb with a
more uniformly distributed deformation.

compression process. This auxetic characteristic allowed for the for­ structure, which influence how the structure deforms under stress. As
mation of a uniform deformation pattern in the re-entrant honeycomb. shown in Fig. 8(a), the hexagonal honeycomb has convex cells with
Since plastic deformation was uniformly distributed throughout the outward angles. Therefore, the hexagonal unit cell has its inclined seg­
honeycomb instead of being concentrated on the shear band, more ments bending outward after buckling until they come into contact with
material in the re-entrant honeycomb experienced plastic deformation inclined segments from adjacent cells to achieve localized compactness.
than in the hexagonal counterpart overall. This led the re-entrant hon­ During this process, the inclined segments horizontally push neigh­
eycomb to exhibit higher structural stiffness and strength, thereby boring cells away, resulting in a positive Poisson’s ratio. In comparison,
dissipating more energy than other honeycomb counterparts. the re-entrant honeycomb has concave cells with inward angles.
On the other hand, the hybrid honeycomb underwent no lateral Accordingly, the inclined segments of the re-entrant unit cell bend in­
contraction or expansion under compression due to a zero overall ward after buckling and eventually contact the horizontal segments
Poisson’s ratio. This is because the hybrid honeycomb has an alternating from the same cell to achieve localized compactness. Consequently, the
arrangement of the hexagonal and re-entrant uni-cells in each layer. neighboring cells are pulled closer horizontally, leading to a negative
When being compressed, the hexagonal cells expand laterally whilst the Poisson’s ratio of the re-entrant structure.
re-entrant cells contract. Once the unit cell initiates deformation, it Numerical simulations of the deformation process of honeycombs
triggers its adjacent cells in the same layer (with different unit-cell under quasi-static compression were conducted using the finite element
structures) to deform significantly as well, as shown in Fig. 7(b). This analysis (FEA) method to validate the experimental analysis. The details
led to most plastic deformation being concentrated in the top and bot­ of the FEA setup can be found in the Appendix. Fig. 8 indicates the
tom layers of the hybrid honeycomb during the compression. Since the agreement between experimental and numerical results. The deforma­
overall plastic deformation in the hybrid honeycomb lies in between the tion patterns of the hexagonal and re-entrant honeycombs at the
hexagonal honeycomb and the re-entrant honeycomb, it exhibited some maximum compression displacement during the experiment (in Fig. 8
intermediate mechanical performance overall. (a)) closely match that from the simulation (in Fig. 8(b)).
The different Poisson’s ratios exhibited by different honeycombs
stem from the different orientations of struts within the honeycomb

7
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 8. Comparison of deformation patterns between the hexagonal honeycomb and the re-entrant honeycomb. (a) The experimental result. (b) The numerical result.
The plastic strain concentrates in diagonal cells within the hexagonal honeycomb but spreads more uniformly within the re-entrant honeycomb. Numerical results
show that the maximum plastic strain in the hexagonal honeycomb (1.24) is approximately 40% higher than that in the re-entrant honeycomb (0.75).

3.2. Reusability of metamaterials with shape memory polymers, these samples have the potential for
recurring energy dissipation as the residual plastic strain is recoverable
The above analysis on the effects of printing materials and unit-cell after unloading in each cycle simply following a heating process. Thus,
structures on the mechanical performance of metamaterials was based the honeycomb samples were performed with consecutive cold pro­
on the experimental data obtained from the first cold programming gramming cycles for the reusability study.
cycle of compression. As all the honeycomb samples were fabricated In this section, the experimental data of each honeycomb sample was

Fig. 9. A typical force-displacement response of honeycomb metamaterials from the cyclic cold programming process. The energy dissipating capability (the area
between the loading and unloading curves) of metamaterials is expected to decrease, indicating the mechanical degradation of metamaterials during the process.

8
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

derived from multiple cold programming cycles with a constant recovery height, Hr, which becomes a new initial height, H0, for the next
compression distance of 12 mm (approximately 34% loading strain). cycle.
The metrics used for the reusability analysis are introduced first, fol­ Therefore, Rr at each cycle is calculated by Rr = HelaDcomp
+HSME
× 100%,
lowed by the study on the influence of different heating media on the where Hela = Hunloading − Hloading is the height recovered from the elastic
shape recovery of honeycombs to determine an optimal heating method. deformation during unloading, HSME = Hr − Hunloading is the height
After that, the force-displacement responses of honeycomb samples recovered through SME during heating. Thus, the numerator Hela + HSME
derived from the cyclic cold programming tests and the reusability = Hr − Hloading represents the total height recovery that the honeycomb
metrics obtained will be presented, followed by the analysis of the effect achieves at the end of each programming cycle. Dcomp is the compression
of printing materials and unit-cell structures on the reusability of the distance. Rr reflects the height recovery of honeycombs at the end of the
honeycomb metamaterials based on the reusability metrics. cold programming process.

3.2.1. Metrics for reusability analysis 3.2.2. Heating method for SMP recovery
The reusability analysis of metamaterials is based on the force- Shape memory polymers have a certain degree of hydrophilicity,
displacement curves obtained from the cyclic cold programming pro­ meaning they can absorb moisture when placed in a humid environ­
cesses. Therefore, all the mechanical metrics introduced in Section 3.1.1 ment, such as water. Most literature studies adopted water-bath heating
(e.g., energy dissipation (Ediss)) at each cycle are used for the reusability as the triggering mechanism for the SME of SMPs [29,51,54,61,62].
analysis. It is worth noting that in some later cycles, the initial peak force However, there has been limited discussion on the influence of heating
may not be observed (which is believed to be caused by accumulated media on the shape recovery of SMPs. In this study, the honeycomb
material damage and inadequate shape recovery). Thus, in these later metamaterial samples were heated using water and air to investigate the
cycles, the force at the transition from the elastic phase to the plastic influence of heating media.
phase was taken as Fpeak. Fig. 10(a) shows the water-heated cyclic cold programming results,
Fig. 9 depicts the typical loading-unloading curves of the first two in which the sample was submerged in a water bath (Joanlab, China)
cold programming cycles. The degradation in mechanical metrics of with a water temperature of 95 ◦ C to heat for 30 s. Fig. 10(b) plots the
honeycomb metamaterials occurs with increasing cycle numbers. For air-heated cyclic cold programming results, in which the sample was
example, the energy dissipation capacity of the honeycomb in Cycle 2 placed into a preheated oven to conduct fan-forced heating at 95 ◦ C for 3
(represented by the area embraced by the loading and unloading curves) min. The shape recovery ratios for each sample at the first four cycles
is significantly lower than that in Cycle 1. were investigated. Fig. 10(c) exhibits that the water-bath heating
To characterize the reusability of honeycomb metamaterials, the method provided better shape recovery than the air-heating counter­
remaining percentage of mechanical metrics (e.g., remaining energy part, with around 6% more recovery in each cycle (e.g., the shape re­
dissipation percentage (Ediss rem %)) was used to quantify the rate of covery ratio of the air-heating method in Cycle 4 was 48.5%, whereas
mechanical degradation during cyclic testing. The remaining percentage that of the water-bath heating method was 54.7%). Besides, the water-
for each mechanical metric is determined by dividing the current me­ bath heating method was way more efficient than the air-heating
chanical metrics obtained for honeycombs in a given cycle by the cor­ method in triggering the SME of SMPs. The deformed honeycomb
responding initial metrics measured in the first cycle. For example, reversed most of its plastic deformation after about 180 s when heated
Ediss
Ediss rem % is calculated by Ediss rem % = Ediss 1st
× 100%, where Ediss rep­ with air (see Video 1 in the Supplementary material). In comparison, the
resents the dissipated energy at the current cycle, and Ediss 1st denotes same honeycomb achieved similar shape recovery within 8 s when
the initial energy dissipation at the first compression cycle. Ediss rem % can heated with water (see Video 2 in the Supplementary material), which
be used to reflect the ability of honeycombs to maintain the level of was 20 times more efficient than the air-heating method.
energy dissipation as the number of cycles increases. Likewise, the According to the principle of heat transfer in thermodynamics, the
remaining percentage of other mechanical metrics can be calculated in a heat flux is proportional to the convective heat transfer coefficient (h) of
similar way. media. The coefficient h of boiling water in free convection conditions is
In addition, the shape recovery ratio (Rr) is used to quantitatively 2500− 25,000 W/(m2 ⋅ K), whereas the h of air in forced convection
describe the shape recovery of honeycomb metamaterials in cyclic conditions is just 10− 500 W/(m2 ⋅ K). Therefore, water has a much
testing. The height variation of the honeycomb during cyclic testing can greater heat transfer coefficient, allowing much faster heating on the
be determined from the force-displacement curves in Fig. 9, where the samples. Considering shape recovery and experimental efficiency, the
honeycomb starts at an initial height, H0, reaches Hloading after water-bath heating method was adopted as the means to trigger the
compressive loading, rebounds to Hunloading after unloading, and finally shape memory effect of SMP honeycombs in this study.
recovers to Hr after heating. In Cycle 1, H0 is the initially designed height
of the honeycomb sample. In the subsequent cycles, the H0 becomes Hr 3.2.3. Effect of printing materials
from the previous cycle. Due to incomplete shape recovery, H0 normally In this section, the experimental data of the PLA hexagonal honey­
decreases in the subsequent cycles. This can typically be reflected in the comb and the PETG hexagonal honeycomb from the first six cold pro­
rightward shift of the initial displacement value in the loading curve, gramming cycles of compression is analyzed to investigate the effect of
indicating that the loading platen needs to move a longer distance to the printed constituent materials on the reusability of honeycomb
contact the upper surface of the sample from a fixed initial clearance. metamaterials.
Note that the Hloading remains the same for all the cycles because the Figs. 11(a) and (b) display the force-displacement responses ob­
cyclic compression employs a constant compression strain (34%), where tained from the first six cold programming cycles of the PLA and the
the compression distance (Dcomp) equals 12 mm. Since Hunloading is the PETG hexagonal honeycombs, respectively. Significant degradation is
height recovered by the honeycomb through elastic deformation after observed in the PLA hexagonal honeycomb because the increase in cycle
unloading, its value generally decreases due to cumulative material numbers causes the force-displacement curve to shrink towards the
damage during the cycle. This is reflected in the rightward shift of the bottom right, as indicated in Fig. 11(a). In contrast, the PETG hexagonal
final displacement value in the unloading curve, indicating that the honeycomb exhibits a marginal decrease in mechanical performance as
loading platen moves a shorter distance from the maximum compression the difference between the cyclic force-displacement curves is much less
position, then no longer senses the reaction force from the honeycomb. significant, as shown in Fig. 11(b). An interesting finding is that an
In the final programming stage, the honeycomb undergoes stress-free increasing force is observed at a displacement of around 9 mm in the
shape recovery stimulated by heating, reaching the shape memory force-displacement response of the PLA hexagonal honeycomb in later

9
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 10. Force-displacement responses of the PLA hexagonal honeycomb metamaterials from the different heating methods under the cyclic cold programming
process. (a) The water-bath heating method. (b) The air-heating method. Colored solid lines in (a) and (b) refer to the force-displacement responses in the different
cycles, while colored dashed lines depict the shape recovery due to the SME during the heating process at the last cycle. (c) Shape recovery ratios. The water-bath
heating method provides better shape recoverability for the metamaterials than the air-based heating method.

Fig. 11. Force-displacement responses of the hexagonal honeycomb metamaterials from the cyclic cold programming process. (a) The PLA hexagonal honeycomb.
(b) The PETG hexagonal honeycomb. The PLA hexagonal honeycomb experiences more severe mechanical degradation than the PETG hexagonal honeycomb as the
energy dissipation capacity (area between the loading-unloading curve) decreases more obviously. (c) Buckling occurred in the unit-cells on the secondary diagonal
of the PLA hexagonal honeycomb in Cycle 2, causing the increase in force in the honeycomb.

cycles. This is believed to be related to the localized compactness that increase in force until the buckling took place at around 12 mm of
occurred on shear-band unit cells. As depicted in Fig. 11(c), when the displacement.
PLA honeycomb was compressed by 10 mm at Cycle 2, the inclined wall The reusability metrics for each honeycomb in each cycle were
segments of cells along the secondary diagonal started undergoing more calculated and summarized in Fig. 12(a) to (e). Fig. 12(a) shows that the
deformation to withstand the external load. This led to a continuous structural stiffness of the PLA honeycomb experienced a significant

10
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 12. Calculated reusability metrics for the hexagonal honeycomb metamaterials in each cold programming cycle based on corresponding cyclic force-
displacement responses: (a) structural stiffness and its remaining percentage, (b) initial peak force and its remaining percentage, (c) plateau force and its remain­
ing percentage, (d) energy dissipation and its remaining percentage, (e) shape recovery ratio and height recovery via elasticity and shape memory effect. The PLA
hexagonal honeycomb exhibits more severe mechanical degradation than the PETG hexagonal honeycomb under cyclic cold programming.

reduction with increasing cycle numbers, starting from 502 N/mm in plateau force during cold programming cycles, the PETG hexagonal
Cycle 1 and down to 187 N/mm (63% drop) in Cycle 6. In comparison, honeycomb exhibited better capability in maintaining energy dissipa­
the structural stiffness of the PETG counterpart hexagonal honeycomb tion than the PLA counterpart. Fig. 12(d) shows that the PLA meta­
only decreased by 11% from 305 N/mm in Cycle 1 to 272 N/mm in Cycle material provided an energy dissipation of 1.37 J in Cycle 6, which was
6. Although the PLA honeycomb had an initial structural stiffness 39% only 22.5% of that in Cycle 1 (6.09 J). In contrast, while the PETG
higher than the PETG counterpart, from Cycle 4 onwards, the structural metamaterial dissipated 1.37 J less energy than the PLA counterpart in
stiffness of the PETG metamaterials started outperforming the PLA Cycle 1, it offered more energy dissipation from Cycle 3 onwards. The
metamaterials. PETG honeycomb achieved an energy dissipation of 3.70 J, which was
A similar result can be observed in the initial peak force. Fig. 12(b) still 78.3% of that in Cycle 1.
indicates that the initial peak force of the PLA hexagonal honeycomb All the above evidence indicates that the degradation of mechanical
reduced by 76.5% from 734 N to 172 N after six programming cycles, performance under cyclic working conditions is much more severe in the
whereas the PETG counterpart only decreased by 25% from 548 N to PLA hexagonal honeycomb than in the PETG counterpart. The main
411 N. The slow degradation of PETG metamaterial hexagonal honey­ reason is that PLA has lower ductility than PETG, as demonstrated in
comb enabled its initial peak force to surpass the PLA counterpart since Section 3.1.2. This makes the PLA honeycomb more prone to fracture
Cycle 3. than the PETG counterpart when undergoing the same level of plastic
Regarding the plateau force, the PETG hexagonal honeycomb also deformation.
outperformed the PLA counterpart. Fig. 12(c) shows that the plateau Fig. 13(a) shows that cracks appeared in the shear-band unit-cells of
force of the PLA metamaterial decreased from 621 N in Cycle 1 to 303 N the PLA hexagonal honeycomb only after two cycles. The cell at the
in Cycle 6 (dropped by 51%). In comparison, the PETG metamaterial bottom-left corner even had a complete nodal fracture. In addition, more
only decreased 16% in the plateau force from 472 N in Cycle 1 to 395 N and more cracks and complete nodal fractures occurred on the shear
in Cycle 6. band as the programming cycle increased. In contrast, cracks accumu­
Due to the slower degradation of the stiffness, initial peak force, and lated much slower in the PETG counterpart. Only two minor cracks and

11
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 13. Crack initiation and propagation of hexagonal honeycomb metamaterials at different cycles after unloading. (a) The PLA hexagonal honeycomb. (b) The
PETG hexagonal honeycomb. The PLA hexagonal honeycomb is more prone to material failure than the PETG hexagonal honeycomb.

one major crack were observed in the PETG honeycomb in Cycle 6, and deformation and 8.27 mm from the SME). Furthermore, the PLA hon­
no cracks developed into complete nodal fractures (see Fig. 13(b)). The eycomb experienced a more significant reduction in the shape recovery
presence of nodal cracks and complete fractures significantly lowered ratio than the PETG counterpart. The shape recovery ratio of the PLA
the structural stiffness and strength of honeycombs. Thus, the PETG metamaterial decreased from 74.1% in Cycle 1 to 49.3% in Cycle 6, a
metamaterial exhibited much slower mechanical degradation than the reduction of 24.8%, whereas the PETG hexagonal honeycomb had a
PLA counterpart because the better ductility of PETG slowed down the shape recovery ratio of 92.3% in Cycle 6, only a 1.7% reduction
initiation and propagation of cracks. compared to the 94.0% in Cycle 1.
From the perspective of shape recovery ratio, the PETG hexagonal Material damage in the node regions also played a dominant role in
honeycomb showed better shape recovery capability than the PLA the overall shape recovery of honeycombs. As discussed in Section 2.2,
counterpart. Fig. 12(e) compares the shape recovery ratios and height the long molecular chains drive the shape memory effect of SMPs.
recoveries between these two honeycombs during the cyclic tests. It can Cracks and complete fractures at the nodes lead to the rupture of mo­
be observed that the PETG honeycomb achieved not only slightly higher lecular chains, making it impossible to trigger the SME. As shown in
recovered heights through elasticity but also significantly higher Fig. 14(a), the shear-banding cells with severe deformation in the PLA
recovered heights through the SME compared to the PLA counterpart, hexagonal honeycomb failed to recover to a hexagonal shape from Cycle
exhibiting a superior overall shape recovery performance. For example, 1 because of the failure of the constituent material. Moreover, the plastic
the shape recovery ratio of the PLA honeycomb in Cycle 1 was 74.1%, deformation recovered by the SME in these cells became increasingly
with a total height recovery of 8.89 mm (i.e., 2.66 mm from elastic limited as the number of cycles increased. In contrast, Fig. 14(b) shows
deformation and 6.23 mm from the SME). In contrast, the PETG hon­ that the cells in the shear band of the PETG hexagonal honeycomb
eycomb achieved a much larger shape recovery ratio of 94.0%, with a almost recovered to a perfect hexagonal shape even at the end of Cycle 6.
total height recovery of 11.28 mm (i.e., 3.01 mm from elastic Thus, the PETG hexagonal honeycomb always exhibited better shape

Fig. 14. Shape recovery of the hexagonal honeycomb metamaterials at different cycles after heating. (a) The PLA hexagonal honeycomb. (b) The PETG hexagonal
honeycomb. The PLA hexagonal honeycomb fails to recover its initial shape from Cycle 1 due to severe material damage, while the PETG hexagonal honeycomb can
still reverse to its initial shape in Cycle 6 due to less material damage.

12
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

recovery than the PLA counterpart in the same cycle due to the slower Fig. 16(a) shows that the structural stiffness of the hexagonal hon­
accumulation of material damage. The inadequate shape recovery re­ eycomb gradually decreased from 305 N/mm in Cycle 1 to 74 N/mm in
duces the mechanical performance of the PLA hexagonal honeycomb. Cycle 12 (only 24% of Cycle 1). In comparison, the hybrid counterpart
This is because the wall segments stay in a buckling state even after exhibited a decreasing structural stiffness from 493 N/mm in Cycle 1 to
heating, which reduces the overall structural stiffness and strength of the 293 N/mm in Cycle 12 (59% of Cycle 1), then further down to 149 N/
honeycomb. mm in Cycle 15 (30% of Cycle 1). The structural stiffness of the PETG re-
Therefore, the PETG material shows excellent potential in cyclic entrant honeycomb dropped from 692 N/mm in Cycle 1 to 532 N/mm in
working conditions due to its high fracture resistance. However, the Cycle 12 (77% of Cycle 1), then to 278 N/mm in Cycle 15 (40% of Cycle
significant mechanical degradation of the PLA material greatly hinders 1), and finally to 159 N/mm in Cycle 17 (23% of Cycle 1). The re-entrant
its reusable application in energy dissipation. honeycomb exhibited the highest structural stiffness in each cycle, while
the hexagonal honeycomb showed the lowest stiffness, with the hybrid
3.2.4. Effect of unit-cell structures honeycomb in between.
This section presents the results of PETG honeycombs with the In terms of maintaining structural stiffness for recurring use, the re-
hexagonal, the hybrid and the re-entrant configurations from cyclic cold entrant honeycomb outperformed the hybrid honeycomb and the hex­
programming tests. All the honeycomb samples were continuously agonal honeycomb. More specifically, based on a reference of approxi­
programmed until their remaining energy dissipation capacity dropped mately 30% of the remaining structural stiffness, the re-entrant
to around 40% of the first cycle. The reusability metrics will be pre­ metamaterial was reusable for 16 cycles (reduced to 35%), the hybrid
sented, followed by an analysis of the effect of unit-cell structures on the metamaterial for 15 cycles (reduced to 30%), and the hexagonal meta­
reusability of metamaterials. material for only 12 cycles (reducing to 24%), which was four cycles less
Figs. 15(a) to (c) exhibit the results of the PETG hexagonal, the PETG than that of the re-entrant counterpart.
hybrid and the PETG re-entrant honeycombs under multiple cycles, Similar results can be observed in other reusability metrics. As shown
respectively. The hexagonal honeycomb was reused with 12 cycles in Fig. 16(b), the re-entrant honeycomb consistently exhibited the
before the remaining capacity of energy dissipation dropped to around highest initial peak force at any given cycle, while the hexagonal hon­
40%, which was three cycles and five cycles less than the hybrid hon­ eycomb displayed the least force, with the hybrid honeycomb in the
eycomb and the re-entrant honeycomb, respectively. Significant me­ middle. Moreover, taking approximately 30% of the remaining initial
chanical degradation was observed in the last few cycles for all peak force percentage as a reference, the re-entrant honeycomb was
honeycombs, manifested by a force decrease in the force-displacement reused for 15 cycles (reducing to 32%), which was one cycle more than
responses. Fig. 15(d) demonstrates the abrupt drop in the force- the hybrid honeycomb (reducing to 33%) and four cycles more than the
displacement curve of the re-entrant honeycomb in Cycle 16 caused hexagonal honeycomb (reducing to 37%). Fig. 16(c) demonstrates that
by a complete fracture at the node. The reusability metrics for each the re-entrant honeycomb consistently presented a higher plateau force
honeycomb at different cycles are summarized in Figs. 16(a) to (e). at every cycle than the hybrid honeycomb, which, in turn, performed

Fig. 15. Force-displacement responses of the PETG honeycomb metamaterials from the cyclic cold programming process. (a) The PETG hexagonal honeycomb. (b)
The PETG hybrid honeycomb. (c) The PETG re-entrant honeycomb. Mechanical degradation is the most severe in the hexagonal honeycomb, followed by the hybrid
honeycomb, and then the re-entrant honeycomb because it takes the hexagonal, the hybrid, and the re-entrant honeycombs 12, 15, and 17 cycles of cold pro­
gramming, respectively, to have their remaining energy dissipation percentage decrease to approximately 40%. (d) A complete nodal fracture at the marked area in
the re-entrant honeycomb causing a sudden drop in the force-displacement curve in (c).

13
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 16. Calculated reusability metrics for the PETG honeycomb metamaterials in each cold programming cycle based on corresponding cyclic force-displacement
responses: (a) structural stiffness and its remaining percentage, (b) initial peak force and its remaining percentage, (c) plateau force and its remaining percentage, (d)
energy dissipation and its remaining percentage, (e) shape recovery ratio and height recovery via elasticity and shape memory effect. Under cyclic cold programming,
the mechanical degradation of the hexagonal honeycomb is the most severe, followed by the hybrid honeycomb, and then the re-entrant honeycomb.

better than the hexagonal honeycomb. Based on approximately 45% of Fig. 16(d) indicates that, at each cycle, the re-entrant honeycomb
the remaining plateau force percentage, the re-entrant honeycomb was consistently provided the highest energy dissipation, followed by the
reused for 17 cycles (reducing to 45%), which was two cycles more than hybrid honeycomb and then the hexagonal honeycomb. Moreover, even
the hybrid honeycomb (reducing to 46%) and five cycles more than the after 17 reuse cycles, the re-entrant honeycomb still achieved a 39%
hexagonal honeycomb (reducing to 44%). remaining energy dissipation percentage. In comparison, the remaining

14
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

percentage of the hybrid honeycomb was reduced to a similar level Less strain concentration made the re-entrant honeycomb more
(40%) after 15 cycles, two cycles less than the re-entrant honeycomb. failure-resistant as the initiation and propagation of cracks were much
The hexagonal honeycomb was reused for 12 cycles, five cycles less than slower. Fig. 17 illustrates the crack formation and propagation to com­
the re-entrant counterpart, before its remaining energy dissipation plete nodal fractures in different honeycombs under cyclic cold pro­
percentage decreased to 38%. gramming. Fig. 17(a) shows that the hexagonal honeycomb experienced
Fig. 16(e) shows that, in terms of shape recovery ratio, the re-entrant significant deformation in shear-banding cells. As a result, eight major
honeycomb performed the best, followed by the hybrid honeycomb, cracks and three complete nodal fractures occurred by Cycle 12. In
whereas the hexagonal honeycomb performed the worst. For example, comparison, severe deformation occurred in the cells at both the top and
the shape recovery ratio of the re-entrant honeycomb remained bottom layers of the hybrid honeycomb. The cracks accumulated and
impressively at 94.6% in Cycle 12, with a total recovery height of 11.35 propagated to fractures in the hybrid honeycomb much slower than in
mm (2.85 mm from elastic deformation and 8.51 mm from the SME). In the hexagonal counterpart, with only five major cracks and one com­
comparison, the shape recovery ratio of the hybrid honeycomb in that plete fracture occurring by Cycle 12, and twelve major cracks and three
cycle was slightly lower at 94.0%, with a total recovery height of 11.28 complete fractures by Cycle 15, as shown in Fig. 17(b). Due to an overall
mm (i.e., 2.35 mm from elastic deformation and 8.94 mm from the uniform deformation, the crack accumulation and propagation was the
SME). The hexagonal honeycomb exhibited a shape recovery ratio of slowest in the re-entrant honeycomb, with only two major cracks and
only 88.0%, with a total recovery height of 10.56 mm (i.e., 2.57 mm one complete fracture occurring by Cycle 12, nine major cracks and two
from elastic deformation and 8.00 mm from the SME), 0.79 mm less than complete fractures by Cycle 15, and thirteen major cracks and five
the re-entrant counterpart. In addition, the shape recovery ratio of the complete fractures by Cycle 17, see Fig. 17(c). Thus, in terms of me­
re-entrant honeycomb decreased from 96.4% in Cycle 1 to 94.6% in chanical performance and resistance to mechanical degradation, the re-
Cycle 12, only a decrease of 1.8%. The hybrid honeycomb showed a entrant honeycomb outperformed the hybrid honeycomb, whereas the
similar reduction from 95.7% to 94.0%, a decrease of 1.7%. In contrast, hexagonal honeycomb was the worst.
the hexagonal honeycomb decreased from 94.0% to 88.0%, a decrease
of 6.0%. 4. Conclusions
Another interesting finding is that with increasing cycles, the height
recovered through elastic deformation in the re-entrant honeycomb In this study, shape memory polymer (SMP) honeycomb meta­
remained similar, while that in the other two honeycombs gradually materials were fabricated with PLA and PETG materials by using FDM
decreased. This suggests the advantage of the re-entrant structure in technology. Cyclic cold programming tests were performed on prepared
resisting material failure (i.e., with better remaining elastic recover­ honeycomb samples to investigate the effect of printed materials and
ability) over the other two structures during cyclic cold programming. unit-cell structures on the mechanical characteristics and reusability of
The differences in reusability between honeycombs with different honeycomb metamaterials. Based on the experimental data obtained
unit-cell structures are due to the strut angle between wall segments. As during cyclic tests, the following conclusions are drawn within its
illustrated in Fig. 8(a), the buckling segments of the re-entrant honey­ limitations:
comb require a rotation angle of 60◦ to achieve localized compactness. The water-bath heating method not only provided 6% more shape
In contrast, the buckling segments of the hexagonal honeycomb require recovery for honeycomb metamaterials than the air-based heating
a rotation angle of 120◦ . This is because the re-entrant honeycomb method but also showed 20 times more efficiency in triggering the shape
achieves localized compactness after its inclined segments contact the memory effect of SMPs.
horizontal segments within the same unit cell. In comparison, the in­ Under the one-off compression, the PLA hexagonal honeycomb
clined segments in the hexagonal honeycomb need to rotate until exhibited better mechanical performance than the PETG counterpart
making contact with inclined segments from other cells to achieve because PLA has a more than double elastic modulus than PETG.
localized compactness. Due to a larger rotation angle, the wall segments Moreover, the PETG re-entrant honeycomb showed the best mechanical
of the hexagonal honeycomb experience more severe deformation than performance because it presented an overall uniform deformation
that of the re-entrant honeycomb, potentially leading to more significant pattern due to its negative Poisson’s ratio. The PETG hybrid honeycomb
material damage. exhibited an intermediate performance because it showed a deformation
The localized compactness prevents the deformed unit cell from pattern with most plastic deformation residing in the top and bottom
further buckling, allowing the cell to transfer the applied force to other layers due to its zero Poisson’s ratio. The PETG hexagonal honeycomb
cells that have not undergone significant deformation, thereby facili­ exhibited the worst performance because it presented a shear-band
tating subsequent buckling behavior. Due to the auxetic design, the re- deformation due to its positive Poisson’s ratio.
entrant honeycomb can achieve local densification and make the Under the cyclic cold programming, the PETG hexagonal honeycomb
transfer of force application more easily, which allows the cells of the re- exhibited better reusability than its PLA counterpart, as the PETG hon­
entrant honeycomb to avoid excessive deformation and form a uniform eycomb showed a remaining energy dissipation percentage of 78.3% at
distribution of plastic deformation at the maximum displacement (see Cycle 6, nearly 3.5 times of the 22.5% of the PLA counterpart. This was
Fig. 7). In comparison, the hexagonal metamaterial has plastic defor­ because the better ductility of PETG not only slowed down mechanical
mation concentrated on shear-banding cells, while the hybrid meta­ degradation but also achieved better shape recovery to maintain struc­
material presents the deformation in the cells at the top and bottom tural stiffness and strength.
layers. Under the cyclic cold programming, the PETG re-entrant honeycomb
To quantitatively analyze the deformation, the plastic strains of the showed the best reusability in the three PETG honeycombs. The re-
hexagonal and the re-entrant honeycombs at the maximum compression entrant honeycomb was reusable for 17 cycles to maintain about 40%
displacement were calculated using the FEA method. From Fig. 8(b), the of energy dissipation capacity, two cycles more than the hybrid coun­
maximum plastic strain of the hexagonal honeycomb is computed as terpart and five cycles more than the hexagonal counterpart. This was
1.24, while that of the re-entrant honeycomb is only 0.75 (about 40% because the re-entrant honeycomb experienced an overall uniform
lower). Moreover, the plastic strain of the hexagonal honeycomb is deformation pattern that caused the least strain concentration, thereby
concentrated in the nodal regions of the diagonal cells. This means that delaying crack formation and propagation. In contrast, the hybrid
material failure will initiate along the diagonal direction. In comparison, honeycomb had strain residing in the top and bottom layers of cells, thus
the plastic strain of the re-entrant honeycomb is distributed more uni­ showing an intermediate strain concentration. Due to the presence of the
formly, leading to less severe material failure over the honeycomb shear band, the hexagonal honeycomb experienced the most severe
structure. strain concentration that exacerbated the formation and propagation of

15
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

Fig. 17. Crack initiation, propagation, and development to complete nodal fractures at different cycles in different PETG honeycomb metamaterials. (a) The PETG
hexagonal honeycomb. (b) The PETG hybrid honeycomb. (c) The PETG re-entrant honeycomb. The hexagonal honeycomb is the most prone to material failure,
followed by the hybrid honeycomb and then the re-entrant honeycomb.

cracks. Data availability

Therefore, when applying metamaterials for reusable purposes,
using constituent materials with better ductility and choosing failure- Data will be made available on request.
resistant structural designs (i.e., designs that introduce less strain con­
centration) are both considered effectivein achieving good reusability.
In future work, other design strategies to enhance the reusability of Acknowledgments
metamaterials can be investigated, e.g., embedding honeycombs into
foams to create composite materials [63], fabricating honeycombs with The support from the Australian Research Council (ARC) Discovery
multimaterials [60,64–66], introducing disorder into honeycomb cells scheme (DP190103752) is acknowledged. The first author is a recipient
[67], adopting TPMS (Triply Periodic Minimal Surfaces) lattice struc­ of the FEIT Scholarship and IRS Scholarship at the University of Tech­
tures [68]. Furthermore, it is necessary to employ advanced constitutive nology Sydney. The last author is a recipient of the ARC Discovery Early
models [69,70] to simulate the shape memory effect of SMPs, and to Career Research Award (DE210101676).
explore the applications of reusable metamaterials in various fields such
as aerospace [61] and soft robotics [71]. Supplementary materials

CRediT authorship contribution statement Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.ijmecsci.2024.109309.
Di Zhang: Data curation, Methodology, Software, Writing – original
draft. Meiyu Li: Data curation, Investigation. Na Qiu: Data curation, Appendix
Investigation, Methodology, Resources. Jie Yang: Data curation,
Investigation. Chi Wu: Data curation, Investigation. Grant Steven: The finite element analysis of honeycombs was performed using the
Funding acquisition, Supervision, Writing – review & editing. Qing Li: explicit solver of LS-DYNA. The honeycombs were modeled using the
Funding acquisition, Supervision, Writing – review & editing. Jian­ solid element with an element size of 0.2 mm. The perfect plasticity
guang Fang: Methodology, Project administration, Supervision, material model was considered for simplification. The material prop­
Writing – review & editing, Conceptualization, Funding acquisition. erties of PETG (Polymaker, China), including a density of 1.25 g/cm3, an
elastic modulus of 0.70 GPa, a Poisson’s ratio of 0.35, and a yield stress
Declaration of competing interest of 34.5 MPa, were adopted.
To numerically investigate the deformation process of honeycombs
The authors declare that they have no known competing financial during the quasi-static compression in the vertical direction, the top
interests or personal relationships that could have appeared to influence rigid plate moved downward at a speed of 0.2 m/s, while the bottom
the work reported in this paper. rigid plate was fully constrained. The "automatic surface to surface"
contact algorithm was used to model the contact between honeycombs

16
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

and rigid plates, and the "automatic single surface" contact algorithm [29] Namvar N, Zolfagharian A, Vakili-Tahami F, Bodaghi M. Reversible energy
absorption of elasto-plastic auxetic, hexagonal, and AuxHex structures fabricated
was used to model the self-contact of honeycombs. The static and dy­
by FDM 4D printing. Smart Mater Struct 2022;31(5). https://doi.org/10.1088/
namic coefficients of friction were both set to 0.25. 1361-665X/ac6291.
[30] Lv W, Yu P, Li D. An energy dissipation metamaterial based on Coulomb friction
References and vibration. Int J Mech Sci 2024;263. https://doi.org/10.1016/j.
ijmecsci.2023.108764.
[31] Deng F, Nguyen QK, Zhang P. Liquid metal lattice materials with simultaneously
[1] Zadpoor AA. Mechanical meta-materials. Mater Horiz 2016;3(5):371–81. https:// high strength and reusable energy absorption. Appl Mater Today 2022;29. https://
doi.org/10.1039/c6mh00065g. doi.org/10.1016/j.apmt.2022.101671.
[2] Surjadi JU, et al. Mechanical Metamaterials and Their Engineering Applications. [32] Garland AP, et al. Coulombic friction in metamaterials to dissipate mechanical
Adv Eng Mater 2019;21(3). https://doi.org/10.1002/adem.201800864. energy. Extreme Mech Lett 2020;40. https://doi.org/10.1016/j.eml.2020.100847.
[3] Dogan E, Bhusal A, Cecen B, Miri AK. 3D Printing metamaterials towards tissue [33] McCrary A, Hashemi MS, Sheidaei A. Programmable Bidirectional Mechanical
engineering. Appl Mater Today 2020;20. https://doi.org/10.1016/j. Metamaterial with Tunable Stiffness and Frictional Energy Dissipation. Adv Theor
apmt.2020.100752. Sep. Simul 2022;5(7). https://doi.org/10.1002/adts.202200135.
[4] Fan J, et al. A review of additive manufacturing of metamaterials and developing [34] Shi J, Mofatteh H, Mirabolghasemi A, Desharnais G, Akbarzadeh A. Programmable
trends. Mater Today 2021;50:303–28. https://doi.org/10.1016/j. Multistable Perforated Shellular. Adv Mater 2021;33(42):e2102423. https://doi.
mattod.2021.04.019. org/10.1002/adma.202102423. Oct.
[5] Jiao P, Alavi AH. Artificial intelligence-enabled smart mechanical metamaterials: [35] Tan X, Chen S, Wang B, Zhu S, Wu L, Sun Y. Design, fabrication, and
advent and future trends. Int Mater Rev 2020;66(6):365–93. https://doi.org/ characterization of multistable mechanical metamaterials for trapping energy.
10.1080/09506608.2020.1815394. Extreme Mech Lett 2019;28:8–21. https://doi.org/10.1016/j.eml.2019.02.002.
[6] de Jonge CP, Kolken HMA, Zadpoor AA. Non-auxetic mechanical metamaterials. [36] Yang H, Ma L. Multi-stable mechanical metamaterials by elastic buckling
Materials (Basel) 2019;12(4). https://doi.org/10.3390/ma12040635. Feb 20. instability. J Mater Sci 2018;54(4):3509–26. https://doi.org/10.1007/s10853-
[7] Ren X, Das R, Tran P, Ngo TD, Xie YM. Auxetic metamaterials and structures: a 018-3065-y.
review. Smart Mater Struct 2018;27(2). https://doi.org/10.1088/1361-665X/ [37] Giri TR, Mailen R. Controlled snapping sequence and energy absorption in
aaa61c. multistable mechanical metamaterial cylinders. Int J Mech Sci 2021;204. https://
[8] Kolken HMA, Zadpoor AA. Auxetic mechanical metamaterials. RSC Adv 2017;7(9): doi.org/10.1016/j.ijmecsci.2021.106541.
5111–29. https://doi.org/10.1039/c6ra27333e. [38] Hua J, Zhou Y, Chen CQ. Design and analysis of a tunable multistable mechanical
[9] Balan P M, Mertens A J, Bahubalendruni MVAR. Auxetic mechanical metamaterials metamaterial. Int J Mech Sci 2024;272. https://doi.org/10.1016/j.
and their futuristic developments: a state-of-art review. Mater Today Commun ijmecsci.2024.109170.
2023;34. https://doi.org/10.1016/j.mtcomm.2022.105285. [39] Wu L, Xi X, Li B, Zhou J. Multi-Stable Mechanical Structural Materials. Adv Eng
[10] Yu X, Zhou J, Liang H, Jiang Z, Wu L. Mechanical metamaterials associated with Mater 2018;20(2). https://doi.org/10.1002/adem.201700599.
stiffness, rigidity and compressibility: a brief review. Prog Mater Sci 2018;94: [40] Fu K, Zhao Z, Jin L. Programmable Granular Metamaterials for Reusable Energy
114–73. https://doi.org/10.1016/j.pmatsci.2017.12.003. Absorption. Adv Funct Mater 2019;29(32). https://doi.org/10.1002/
[11] Nicolaou ZG, Motter AE. Mechanical metamaterials with negative compressibility adfm.201901258.
transitions. Nat Mater 2012;11(7):608–13. https://doi.org/10.1038/nmat3331. [41] Tibbits S. 4D Printing: multi-Material Shape Change. Archit Des 2014;84(1):
May 20,. 116–21. https://doi.org/10.1002/ad.1710.
[12] Li H, Li Y, Li J. Negative stiffness devices for vibration isolation applications: a [42] Yang H, D’Ambrosio N, Liu P, Pasini D, Ma L. Shape memory mechanical
review. Adv Struct Eng 2020;23(8):1739–55. https://doi.org/10.1177/ metamaterials. Mater Today 2023;66:36–49. https://doi.org/10.1016/j.
1369433219900311. mattod.2023.04.003.
[13] Lakes RS, Lee T, Bersie A, Wang YC. Extreme damping in composite materials with [43] Qi J, et al. Recent Progress in Active Mechanical Metamaterials and Construction
negative-stiffness inclusions. Nature 2001;410(6828):565–7. Principles. Adv Sci (Weinh) 2022;9(1):e2102662. https://doi.org/10.1002/
[14] Zheng X, et al. Ultralight, ultrastiff mechanical metamaterials. Science 2014;344 advs.202102662. Jan.
(6190):1373–7. https://doi.org/10.1126/science.1252291. Jun 20. [44] Montgomery SM, Kuang X, Armstrong CD, Qi HJ. Recent advances in additive
[15] Meza LR, Das S, Greer JR. Strong, lightweight, and recoverable three-dimensional manufacturing of active mechanical metamaterials. Curr Opin Solid State Mater Sci
ceramic nanolattices. Science 2014;345(6202):1322–6. https://doi.org/10.1126/ 2020;24(5). https://doi.org/10.1016/j.cossms.2020.100869.
science.1255908. Sep 12. [45] Demoly F, Dunn ML, Wood KL, Qi HJ, André JC. The status, barriers, challenges,
[16] Jang D, Meza LR, Greer F, Greer JR. Fabrication and deformation of three- and future in design for 4D printing. Mater Des 2021;212. https://doi.org/
dimensional hollow ceramic nanostructures. Nat Mater 2013;12(10):893–8. 10.1016/j.matdes.2021.110193.
https://doi.org/10.1038/nmat3738. Oct. [46] Behl M, Lendlein A. Shape-memory polymers. Materials Today 2007;10(4):20–8.
[17] Askari M, et al. Additive manufacturing of metamaterials: a review. Add Manuf https://doi.org/10.1016/s1369-7021(07)70047-0.
2020;36. https://doi.org/10.1016/j.addma.2020.101562. [47] Hager MD, Bode S, Weber C, Schubert US. Shape memory polymers: past, present
[18] Ryan KR, Down MP, Banks CE. Future of additive manufacturing: overview of 4D and future developments. Prog Polym Sci 2015;49-50:3–33. https://doi.org/
and 3D printed smart and advanced materials and their applications. Chem Eng J 10.1016/j.progpolymsci.2015.04.002.
2021;403. https://doi.org/10.1016/j.cej.2020.126162. [48] Xia X, Spadaccini CM, Greer JR. Responsive materials architected in space and
[19] Ngo TD, Kashani A, Imbalzano G, Nguyen KTQ, Hui D. Additive manufacturing (3D time. Nat Rev Mater 2022;7(9):683–701. https://doi.org/10.1038/s41578-022-
printing): a review of materials, methods, applications and challenges. Compos 00450-z.
Part B Eng 2018;143:172–96. https://doi.org/10.1016/j. [49] Xu P, et al. Compression behavior of 4D printed metamaterials with various
compositesb.2018.02.012. Poisson’s ratios. Int J Mech Sci 2024;264. https://doi.org/10.1016/j.
[20] Zhang W, Xu J. Advanced lightweight materials for Automobiles: a review. Mater ijmecsci.2023.108819.
Des 2022;221. https://doi.org/10.1016/j.matdes.2022.110994. [50] Wan M, Yu K, Sun H. 4D printed programmable auxetic metamaterials with shape
[21] Yin H, Zheng X, Wen G, Zhang C, Wu Z. Design optimization of a novel bio-inspired memory effects. Compos Struct 2022;279. https://doi.org/10.1016/j.
3D porous structure for crashworthiness. Compos Struct 2021;255. https://doi. compstruct.2021.114791.
org/10.1016/j.compstruct.2020.112897. [51] Zeng C, Liu L, Bian W, Leng J, Liu Y. Temperature-dependent mechanical response
[22] Lee JY, An J, Chua CK. Fundamentals and applications of 3D printing for novel of 4D printed composite lattice structures reinforced by continuous fiber. Compos
materials. Appl Mater Today 2017;7:120–33. https://doi.org/10.1016/j. Struct 2022;280. https://doi.org/10.1016/j.compstruct.2021.114952.
apmt.2017.02.004. [52] Dong K, Wang Y, Wang Z, Qiu W, Zheng P, Xiong Y. Reusability and energy
[23] Zhang Z, Zhang L, Song B, Yao Y, Shi Y. Bamboo-inspired, simulation-guided absorption behavior of 4D printed continuous fiber-reinforced auxetic composite
design and 3D printing of light-weight and high-strength mechanical structures. Compos Part A Appl Sci Manuf 2023;169. https://doi.org/10.1016/j.
metamaterials. Appl Mater Today 2022;26. https://doi.org/10.1016/j. compositesa.2023.107529.
apmt.2021.101268. [53] Wan M, Yu K, Gu J, Zeng H, Sun H, Khatibi AA. 4D printed TMP origami
[24] Zhilyaev I, Krushinsky D, Ranjbar M, Krushynska AO. Hybrid machine-learning metamaterials with programmable mechanical properties. Int J Mech Sci 2023;
and finite-element design for flexible metamaterial wings. Mater Des 2022;218. 250. https://doi.org/10.1016/j.ijmecsci.2023.108275.
https://doi.org/10.1016/j.matdes.2022.110709. [54] Hamzehei R, Serjouei A, Wu N, Zolfagharian A, Bodaghi M. 4D Metamaterials with
[25] Wang Z, Li K, He Q, Cai S. A light-powered ultralight tensegrity robot with high Zero Poisson’s Ratio, Shape Recovery, and Energy Absorption Features. Adv Eng
deformability and load capacity. Adv Mater 2019;31(7):e1806849. https://doi. Mater 2022;24(9). https://doi.org/10.1002/adem.202200656.
org/10.1002/adma.201806849. Feb. [55] Wang K, Sun G, Wang J, Yao S, Baghani M, Peng Y. Reversible energy absorbing
[26] Wang C, Vangelatos Z, Grigoropoulos CP, Ma Z. Micro-engineered architected behaviors of shape-memory thin-walled structures. Eng Struct 2023;279. https://
metamaterials for cell and tissue engineering. Mater Today Adv 2022;13. https:// doi.org/10.1016/j.engstruct.2023.115626.
doi.org/10.1016/j.mtadv.2022.100206. [56] Sun G, Wang J, Wang K, Baghani M, Peng Y, Rao Y. Repeatable compressive
[27] Lin C, Liu L, Liu Y, Leng J. 4D Printing of Bioinspired Absorbable Left Atrial functionality of 3D printed shape-memory thin-walled corrugated structures. Int J
Appendage Occluders: a Proof-of-Concept Study. ACS Appl Mater Interfaces 2021; Mech Sci 2023;257. https://doi.org/10.1016/j.ijmecsci.2023.108552.
13(11):12668–78. https://doi.org/10.1021/acsami.0c17192. Mar 24. [57] Zhang X, et al. Bio-inspired 4D printed intelligent lattice metamaterials with
[28] Shi J, Zhu L, Li L, Li Z, Yang J, Wang X. A TPMS-based method for modeling porous tunable mechanical property. Int J Mech Sci 2024;272. https://doi.org/10.1016/j.
scaffolds for bionic bone tissue engineering. Sci Rep 2018;8(1):7395. https://doi. ijmecsci.2024.109198.
org/10.1038/s41598-018-25750-9. May 9,.

17
D. Zhang et al. International Journal of Mechanical Sciences 275 (2024) 109309

[58] Liu C, Qin H, Mather PT. Review of progress in shape-memory polymers. J Mater [65] Cheng Y, Li J, Qian X, Rudykh S. 3D printed recoverable honeycomb composites
Chem 2007;17(16). https://doi.org/10.1039/b615954k. reinforced by continuous carbon fibers. Compos Struct 2021;268. https://doi.org/
[59] Gao G, Xu F, Xu J, Tang G, Liu Z. A Survey of the Influence of Process Parameters 10.1016/j.compstruct.2021.113974.
on Mechanical Properties of Fused Deposition Modeling Parts. Micromachines [66] Ren L, et al. 4D printing of customizable and reconfigurable mechanical
(Basel), 13; 2022. https://doi.org/10.3390/mi13040553. Mar 30. metamaterials. Int J Mech Sci 2024;270. https://doi.org/10.1016/j.
[60] Bodaghi M, Serjouei A, Zolfagharian A, Fotouhi M, Rahman H, Durand D. ijmecsci.2024.109112.
Reversible energy absorbing meta-sandwiches by FDM 4D printing. Int J Mech Sci [67] Bhuwal AS, Pang Y, Ashcroft I, Sun W, Liu T. Discovery of quasi-disordered truss
2020;173. https://doi.org/10.1016/j.ijmecsci.2020.105451. metamaterials inspired by natural cellular materials. J Mech Phys Solids 2023;175.
[61] Liu K, Wu J, Paulino GH, Qi HJ. Programmable Deployment of Tensegrity https://doi.org/10.1016/j.jmps.2023.105294.
Structures by Stimulus-Responsive Polymers. Sci Rep 2017;7(1):3511. https://doi. [68] Novak N, et al. Quasi-static and dynamic compressive behaviour of sheet TPMS
org/10.1038/s41598-017-03412-6. Jun 14. cellular structures. Compos Struct 2021;266. https://doi.org/10.1016/j.
[62] Bodaghi M, Damanpack AR, Liao WH. Self-expanding/shrinking structures by 4D compstruct.2021.113801.
printing. Smart Mater Struct 2016;25(10). https://doi.org/10.1088/0964-1726/ [69] Zolfagharian A, Jarrah HR, Xavier MS, Rolfe B, Bodaghi M. Multimaterial 4D
25/10/105034. printing with a tunable bending model. Smart Mater Struct 2023;32(6). https://
[63] Zolfagharian A, Picken P, Bodaghi M, Fard M, Rolfe B. Additive manufacturing of doi.org/10.1088/1361-665X/accba8.
composite foam metamaterial springs for vibration isolation. Adv Eng Mater 2023; [70] Roudbarian N, Jebellat E, Famouri S, Baniasadi M, Hedayati R, Baghani M. Shape-
25(20). https://doi.org/10.1002/adem.202300356. memory polymer metamaterials based on triply periodic minimal surfaces. Euro J
[64] Dong J, et al. Advanced design and fabrication of dual-material honeycombs for Mech ASolids 2022;96. https://doi.org/10.1016/j.euromechsol.2022.104676.
improved stiffness and resilience. Micromachines (Basel) 2023;14(11). https://doi. [71] Mohammadi M, et al. Sustainable robotic joints 4D printing with variable stiffness
org/10.3390/mi14112120. Nov 18. using reinforcement learning. Robot Comput Integr Manuf 2024;85. https://doi.
org/10.1016/j.rcim.2023.102636.

18