Additive Manufacturing for Bioinspired Structures: Experimental Study to Improve the Multimaterial Adhesion Between Soft and Stiff Materials

Abstract

The fabrication of bioinspired structures has recently gained an increasing popularity: mimicking the way in which nature develops structures is a vital prerequisite in soft robotics to achieve multiple benefits. Stiff structures connected by soft joints (recalling, for instance, human bones connected by cartilage) are highly appealing: several prototypes have been manufactured and tested, demonstrating their full potential. In the present research, the material extrusion (MEX) additive manufacturing technology has been used to manufacture stiff–soft bioinspired structures activated by shape memory alloy (SMA) actuators. First, three commercially available stiff composite plastic materials were investigated and linked to different 3D printing infills. Surprisingly, we found that the “gyroid” infill was correlated to the mechanical properties, demonstrating that it produces better results in terms of Young's modulus and ultimate tensile strength (UTS) than the widely studied “lines” infill. The primary focus of the research is an experimental study aimed at improving the adhesion at the interface between stiff and soft materials using an inexpensive method (i.e., MEX). Three different variables that have significant effects on the interface bonding were studied: (1) the interface geometry between stiff and soft parts, (2) the mesh overlapping process parameter, and (3) the annealing post-treatment. By optimizing the three variables, a Young's modulus of 48.8 MPa and a UTS of 3.8 MPa were achieved, when nylon+glass fiber (a stiff material) and thermoplastic polyurethane (a soft material) were 3D printed together. In particular, the 3.8 MPa UTS is 48% higher than the highest adhesion between the soft and stiff material (thermoplastic polyurethane [TPU] and acrylonitrile butadiene styrene) reported in literature. Finally, taking advantage of the improved stiff–soft adhesion, a bioinspired robotic finger has been fabricated and tested using an SMA actuator, showing an enormous potential for the proposed additive manufacturing approach in realizing bioinspired systems.

Introduction

Bioinspiration is one of the cornerstones around which soft robotics is established^1–5: the possibility to mimic animals, plants, and human beings can bring numerous advantages, including the ability to perform complex motions and modulate stiffness. Bioinspired caterpillar,⁶ pangolin,^7,8 octopus,^9,10 jellyfish,^11,12 and frog¹³ soft robots are great examples showing all the bioinspired benefits in terms of crawling, stiffness modulation, swimming, and grasping.

Several ingenious examples of stiff structures (such as bones) connected with soft matters (such as cartilage) are found in nature; the use of soft joints and stiff structures leads to many advantages, including the following: (1) the ability to perform complex motions, (2) the reduction of stress concentration, and (3) the capacity to withstand high loads.¹⁴

The exploitation of the stiff–soft mechanism motivated researchers to come up with different solutions. In 2021, Yamamura and Iwase¹⁵ manufactured a 4D-printed walking soft robot based on a shape memory polymer (SMP)-based soft joint activated utilizing heat and showed 500-folding cycles without failure. A similar approach was also used by Zolfagharian et al; they used a photothermal stimulus to bend the main structure.¹⁶ Al-Rubaiai et al¹⁷ fabricated a PneuNet actuator equipped with three SMP joints inside the inextensive layer: when activated, the range of motion completely changed leading to more sophisticated finger movements recalling human fingers. Other different examples of stiff–soft structures able to perform several bending movements when activated using twisted coiled polymer actuators are provided in Ref.^18,19

Additive manufacturing (AM) addresses very well soft robotic requirements as shown in Refs.,^20–23 and it has the potential to become the dominant manufacturing technique for the fabrication of functional bioinspired soft robots.

Fused filament fabrication (FFF), which is well known for being inexpensive, widespread, and easy to customize,^24–26 has recently been used extensively in the soft robotic field: due to the ability to extrude multimaterials²⁷ in the same printing cycle using commercial 3D printers,²⁸ stiff–soft structures can be manufactured.^29,30

Improving the adhesion among two different materials extruded utilizing FFF is a significant challenge, and overcoming this issue would push the role of the FFF in the manufacturing of biomimetic parts composed of stiff and soft materials. In the past few years, several studies have sought to correlate process parameters with the final multimaterial bonding in an attempt to improve the understanding of multimaterial adhesion.

Khudiavoka et al³¹ demonstrated that the double cantilever beam method is the best (among the analyzed methods) for studying the adhesion between materials characterized by different stiffness degrees, namely polylactic acid (PLA) and PLA+carbon fiber (CF). Using the tensile test method, Lopes et al³² demonstrated that the multimaterial printing of PLA-TPU (thermoplastic polyurethane) and PLA-PET (polyethylene terephthalate) is negatively affected by the presence of the zebra pattern and that the chemical affinity between two materials plays a key role. In Ribeiro et al,³³ the interface geometry between two 3D-printed materials was analyzed and varied, proving that it produces a remarkable effect to take into consideration to improve the material bonding.

Tamburrino et al³⁴ show that the printing order influences the adhesion strength (the stiff material must be printed first) and that the thermal aspects are crucial to improving the adhesion. The latter consideration has been also studied by Yin et al who linked the thermally driven diffusion at the interface with the final part bonding strength.³⁵ The importance of the thermal factors for the intralayer bond strength is also pointed out in Moetazedian et al.³⁶

In this article, a commercial FFF machine was used to extrude stiff and soft materials at the same time to fabricate bioinspired structures. First, the influence of the infill pattern on three different stiff composite materials has been studied, afterward, a way to improve the adhesion strength between a stiff material (nylon+glass fiber [NGF]) and a soft material (TPU) was analyzed.

Three different factors were studied: interface geometries between the two materials, a process parameter called mesh overlap, and thermal postprocessing (annealing). All the studied factors produced a significant effect on the multimaterial adhesion. In particular, by setting the best set of the three studied variables, an ultimate tensile strength (UTS), 48% higher than the highest found in literature (when stiff and soft materials are extruded), was achieved. Finally, a 3D-printed bioinspired finger made up of stiff parts connected by means of soft joints was fabricated and actuated by using a shape memory alloy (SMA) actuator.

Materials and Methods

The objective of the present research is the monolithic manufacturing of bioinspired structures using FFF additive manufacturing technology extruding two materials with completely different material properties in the same printing cycle: a very stiff material and a soft material. The study on the adhesion mechanism between the two materials is a crucial aspect to take into consideration to create functionalized 3D-printed structures activated by an external actuation system such as a spring SMA. The research objective is illustrated in Figure 1.

FIG. 1.

Summary of the research and bioinspired joint: (a) human bone (example of the combination of stiff–soft materials found in nature), (b) study of the interface between stiff and soft materials, (c) main elements of the proposed stiff-soft structure, and (d) proposed finger actuated using coiled SMA actuator at 3.4 A. SMA, shape memory alloys.

A commercial-grade dual-extruder 3D printer was used throughout the whole research (Tenlog 3D), and commercial materials were used: specifically, three composite materials and a soft material were tested. The authors investigated the composite materials as stiff materials to identify the material with the best mechanical properties and they are as follows:

• Nylon+glass fiber (GF), henceforth NGF,

• Polyethylene terephthalate glycol (PETG)+CF, henceforth PETGCF, and

• Polycarbonate (PC)+CF, henceforth PCCF.

As a soft material, a commercial TPU (NinjaFlex, NinjaTek, CA) was used, which is widely recognized as one of the softest materials on the market (shore hardness equal to 85 A, elongation at break equal to 660%, and UTS equal to 26 MPa). After studying which composite commercially available material has the best mechanical properties in conjunction with different printing patterns, a method for improving its adhesion with the TPU soft material has been studied. An investigation was performed on three interface geometries and two parameters (a process parameter and a postprocessing parameter).

Effect of printing pattern on composite materials

In scientific literature, several studies have been conducted to correlate the mechanical properties of FFF dog bones and process parameters such as layer height, printing orientation, and printing speed. In this research, three different printing patterns were studied: lines (well known in the scientific literature for providing an increase in the mechanical properties when a 45° orientation is set), gyroid (a new infill pattern available on slicing software that has been proved to provide good results in terms of compressibility³³), and cross 3D (a printing pattern based on the 3D fabrication of cruxes and not studied in scientific literature). All the infill patterns have been generated by the slicing software Ultimaker Cura 4.11.

The three composite materials (NCF, PETGCG, and PCCF) were used, and for each material, a total of nine dog bones (three samples for each printing pattern) were fabricated and tested under tensile loading using a universal testing machine (Model 5969; INSTON, Inc., Norwood, MA). Standard ASTM D638 was utilized for the fabrication and testing of dog bones. Each material's process parameters are described in Supplementary Table S1 in detail. Figure 2 depicts three different printing patterns and 3D-printed samples.

FIG. 2.

(a) Three different printing patterns, (b) manufactured dog bones using three different composite materials (PCCF = polycarbonate+carbon fiber, NGF = nylon+glass fiber, PETGCF = polyethylene terephthalate glycol+carbon fiber), and (c) Young's modulus versus infill pattern for every material obtained experimentally. The sample size is 165 mm along x-axis and 19 mm along y-axis.

In particular, considering the material composition (plastic+CF or GF), a brass nozzle with a diameter of 0.6 mm was chosen; in this way, problems related to nozzle clogging have been avoided. It is noted that the use of smaller nozzles (i.e., 0.4 mm) is allowed if particular attention is paid to the careful selection of process parameters (i.e., high extrusion temperature), nozzle material (i.e., hardened steel and ruby tip), and nozzle geometries (i.e., 50-mm-long nozzle volcano version). The same nozzle diameter (0.6 mm, brass nozzle) has also been used for the extrusion of TPU, in the Stiff–Soft Material Adhesion section.

The results obtained from the tensile test are summarized in Table 1, as well as the following conclusions can be drawn:

Table 1.

Infill Pattern Impact on the Three Different Stiff Materials

Material	Pattern	E (MPa)	UTS (Mpa)	Elongation at break (%)
PC+CF	Lines	720	27	0.082
	Cross 3D	421	8	0.028
	Gyroid	845	30	0.047
Nylon+GF	Lines	1116	49	0.11
	Cross 3D	622	18	0.05
	Gyroid	1455	58	0.066
PETG+CF	Lines	952	39	0.07
	Cross 3D	620	16	0.035
	Gyroid	1335	49	0.06

CF, carbon fiber; GF, glass fiber; PETG, polyethylene terephthalate glycol; UTS, ultimate tensile strength.

• As a function of the printing pattern, the behavior of every material is the same: the lowest mechanical properties (E and UTS) are obtained for the cross-3D patter, while the highest mechanical properties are obtained for the gyroid pattern. Because of the lack in scientific literature about the effect of gyroid pattern studies, the present result might pave the way for more investigations to create mathematical models.

• Considering the gyroid pattern (the pattern with the best mechanical properties), it is evident that the NGF material has the best mechanical properties, with a Young's modulus of 1.45 GPa and a UTS of 57.48 MPa.

• The manufacturing process is robust: for each material and pattern, three samples were fabricated, resulting in a standard deviation calculated on the Young's modulus <5%.

In conclusion, the best composite material in terms of mechanical properties is NGF, printed using a gyroid pattern: it will be used throughout the present research as a stiff material in conjunction with the soft TPU.

Stiff–soft material adhesion

Several studies have been performed to improve the adhesion between two materials: in particular, Yin et al³⁵ mathematically described the material strengths between two materials in multimaterial FFF printing:

where ${\sigma }_{A∕B}$ is the interfacial bonding strength (UTS) between the two materials (A and B), t_p is the total printing time, $R\left(t\right)$ is the average strength ratio of the whole interface between the two materials, and ${\sigma }_A$ is the UTS of the material A.

In particular, $R\left(t\right)$ can be explained as recurring to the intermolecular diffusion theory occurring at the interface between two polymers. Treating the FFF interfacial bonding as a two-dimensional and grown process; considering $\Omega$ a certain domain at the interface and that once contact of two polymers occurs, areas with intermolecular diffusion (wetted area) are nucleated at random locations, and thus, the strength of domain $\Omega$ will be the sum of all the intermolecular diffusion initiated within $\Omega$.

In this way, $R\left(t\right)$ can be written as follows:

where R_s is the strength ratio, $dA\left(\tau ,\overrightarrow{X}\right)$ indicates the wetted area at the position $\overrightarrow{X}$, at time $\tau$, and A is the total wetted area.

In the present research, three different adhesion geometries among the stiff and soft materials were studied: a wave geometry, a T-shape geometry, and a sandwich geometry (Fig. 3). The three geometries have been designed in accordance with the results obtained in Ribeiro et al.³³ For each geometry, two different parameters have been studied to understand if there is a correlation between them and the mechanical properties of the dual material structures:

FIG. 3.

(a) Three different adhesion mechanisms (white = stiff, pink = soft). (b) Manufactured dog bones with three different adhesion mechanisms (gray = stiff, black = soft). (c) Side view of the sandwich mechanism. (d) T-shape dog bone during the tensile test. (e) Mesh overlap equal to 0 mm. (f) Mesh overlap equal to 0.4 mm. (g) Manufactured dog bones, repetition 1.

• A process parameter called “Mesh Overlapping,” refers to the overlapping at the interface among the two parameters. Two levels of the following parameter were studied: low level (0 mm of overlapping) and high level (0.4 mm of overlapping). In Figure 3e and f, the Mesh Overlapping parameter is depicted.

• A postprocessing parameter, namely the Annealing Temperature: three levels were studied, no annealing (“As printed”), 50°C, and 70°C. The samples were annealed for 1 h in a furnace at the desired temperature and after they were cooled down in the air for 1 h.

A total of 54 dog bone-shaped samples were fabricated (three repetitions for every combination) and tested, with the Young's modulus chosen as a measure of the material's adhesion.

As shown in Figure 4, the main results can be summarized as follows:

FIG. 4.

Annealing effect on mesh overlapping (a) 0 mm overlapping and (b) 0.4 mm overlapping.

• The annealing postprocessing has a remarkable effect when the mesh overlapping parameter is set to 0.4 mm: for each geometry, switching from no annealing (as printed) to 50°C and then to 70°C increases material adhesion (increase in Young's modulus). Specifically, the E increase for the wave, T-shape, and sandwich from “As printed” to 50°C annealing is 15.91%, 47%, and 2%, while the E increase for the same geometries from “As printed” to 70°C is 21.9%, 51.6%, and 19%. The opposite behavior (reduction of E when the annealing temperature increases) occurs when the mesh overlapping parameter is set to 0 mm: in this case, a reduction of E of 7%, 5%, and 4.5% occurs for wave, T-shape, and sandwich switching from “As printed” to 50°C, whereas switching from “As printed” to 70°C, the reduction is, respectively, of 12.8%, 31%, and 4.7%.

At the state of the art, some preliminary studies have been conducted to correlate the annealing process with the mechanical properties of FFF parts, and it has been determined that the annealing process improves the mechanical properties of several materials, such as PLA and acrylonitrile butadiene styrene (ABS). Due to a lack of scientific literature regarding the correlation between annealing and dual material bonding, the following results are difficult to explain: according to the authors, when the overlap parameters are set (0.4 mm), annealing increases the mechanical properties because heat facilitates intramolecular diffusion and the number of wetted areas is higher³⁵ [Eqs. (1) and (2)]. It is worth mentioning that the complexity of the problem is also related to the composition of the studied materials: the stiff material is a composite material made up of nylon and GFs, which in turn contribute to the interfacial bonding.

While the obtained results are clear, their explanation from a chemical/material point of view requires more tests and examination considering multiple variables.

• For every geometry, when no annealing is performed (as printed), the best results in terms of E are obtained when the mesh overlapping was set as 0 mm. On the opposite side, after annealing at 70°C, all samples fabricated with a mesh overlapping of 0.4 mm had a higher E than the “As printed” structures fabricated with a mesh overlapping of 0 mm.

Considering Table 2, E of A2-70 is higher than E of A1-AP; E of B2-70 is higher than B1-AP, and E of C2-70 is higher than E of C1-AP.

Table 2.

Results of the Stiff–Soft Material Adhesion Analysis

Sample name	Geometry	Mesh overlapping (mm)	Annealing temperature (°C)	E (MPa)	Rupture
A1-AP	Wave	0	As printed	31.2	Interface
A1-50	Wave	0	50	29.8	Interface
A1-70	Wave	0	70	27.2	Interface
A2-AP	Wave	0.4	As printed	28.1	TPU
A2-50	Wave	0.4	50	32.55	TPU
A2-70	Wave	0.4	70	34.2	TPU
B1-AP	T-shape	0	As printed	47.5	TPU
B1-50	T-shape	0	50	47.1	TPU
B1-70	T-shape	0	70	34.3	TPU
B2-AP	T-shape	0.4	As printed	32.2	TPU
B2-50	T-shape	0.4	50	47.5	TPU
B2-70	T-shape	0.4	70	48.8	TPU
C1-AP	Sandwich	0	As printed	41.8	Interface
C1-50	Sandwich	0	50	39.8	Interface
C1-70	Sandwich	0	70	39.8	Interface
C2-AP	Sandwich	0.4	As printed	35.1	Interface
C2-50	Sandwich	0.4	50	35.7	Interface
C2-70	Sandwich	0.4	70	42.1	Interface

TPU, thermoplastic polyurethane.

In conclusion, for every geometry, the highest value of E was achieved by setting a mesh overlapping of 0.4 mm and annealing the sample at 70°C for 1 h.

• The rupture position is a powerful tool for comprehending the adhesion dynamics between the stiff and the soft material. Regarding the T-shaped sample, the geometric interlocking mechanism was robust enough to produce a rupture every time within TPU and not at the interface. For the sandwich geometry, rupture always occurred at the interface between the two materials, whereas for the wave sample, a rupture occurred within the TPU when the overlapping parameter was set to 0.4 mm and never at the interface (unlike for the case in which the mesh overlap was set to 0 mm).

• In conclusion, the best geometry and combination of process parameters for improving the adhesion between the stiff and soft material is, respectively, T-shape (in accordance with Ribeiro et al³³), 0.4 mm of mesh overlapping, and 70°C of annealing. For the following configuration, a UTS of 3.85 MPa was achieved, this value is 48.05% higher compared with the highest one found in the scientific literature when a combination of stiff and soft materials was extruded (Yin et al³⁵ obtained 2 MPa using ABS and TPU).

Moreover, as shown in literature,³² material affinity is an important factor to consider to improve the multimaterial adhesion: if two materials (stiff and soft) having a similar chemical composition are jointly extruded, the following results in terms of adhesion can be improved. The development of materials having different degrees of stiffness, but a similar chemical composition, will provide many benefits to the 3D printing field.

The outcomes of the present article can also be used as a benchmark to improve the multimaterial adhesion in several AM technologies able to process more materials at the same time such as material jetting³⁷ and hybrid AM technologies.³⁸

A B2 sample, the best one in terms of adhesion, was examined using a Nikon X-ray microcomputed tomography (μCT) C1 system with a scanning energy level of 65 kV at 115 μA current, and a cubic voxel size of 28.733 μm. After reconstruction from the μCT scan, the volumetric images were visualized in ORS Dragonfly software to observe and analyze the internal structure of the 3D-printed structure made of stiff–soft materials. Inspection of the volumetric images determined that the printed sample has no internal defects within the scanning resolution used.

In particular, the following two observations are made (Fig. 5): (1) at the interface between the two materials (0.4 mm overlap), a straight region in which the two materials are completely bonded together is clearly seen, demonstrating the effectiveness of the proposed approach; and (2) after sectioning the sample, a weak adhesion in the TPU material among the external perimeters and the gyroid infill is revealed, as the gap circled in red in Figure 5d shows. Therefore, future research will focus on the bonding improvement in TPU to further increase the overall mechanical properties.

FIG. 5.

X-ray μCT volumetric images of the proposed sample: (a) 3D-printed sample scanned using the X-ray μCT apparatus, (b) 3D rendering of the entire B2 sample from the μCT scan, (c) close-up view showing the overlapping between the two materials, and (d) 3D rendering revealing the weak wall-infill adhesion. μCT, microcomputed tomography.

AM of Bioinspired Structures

Soft robotic, and the biomedical field in general, can benefit from the proposed manufactured approach in several ways: the fabrication of bioinspired structures that comprised very stiff and soft materials is very appealing due to the possibility to achieve complex motions using 3D-printed structures recalling biological structures.

Setting the mesh overlap parameter equal to 0.4 mm, annealing the stiff–soft structure at 70°C for 1 h, and using the T-shape interface geometry, a remarkable improvement at the interface was achieved (UTS of 3.8 MPa and E of 48.8 MPa), making structures manufactured in accordance with the proposed approach more reliable, improving the life span as well.

To demonstrate the potentialities of the stiff–soft (NGF and TPU) additive manufacturing approach, two bioinspired structures were designed and fabricated: a robotic finger and a multidirectional bender. For both the objects, the working mechanism is the same: the joints fabricated using TPU are connected to the passive structure (made up of NGF) recurring to the T-shape interface. After the fabrication, the proposed structures were annealed for 1 h at 70°C, to improve the interfacial adhesion in accordance with the results obtained in the Stiff–Soft Material Adhesion section

It is worth mentioning that no assembly tasks were involved for the fabrication of the proposed bioinspired structures, leading to significant advantages when compared with the traditional stiff–soft structure, which requires multiple assembly steps and different manufacturing technologies or an expensive additive manufacturing setup.¹⁹

The proposed finger was designed in accordance with Mutlu et al,³⁹ as shown in Figure 6. It has the following properties: (1) it is equipped with a 4 mm hole for an SMA actuator insertion (Supplementary Table S2) after the annealing process, (2) made up of three TPU joints, namely metacarpophalangeal joint (MCP), distal interphalangeal joint (DIP), and proximal interphalangeal joint (PIP), and (3) it is equipped with a terminal block to fix it to the experimental setup.

FIG. 6.

(a) Proposed finger, (b) finger position in x-y space, (c) bending angle, (d) proposed multidirectional bender, and (e) motions performed by the bender as a function of the activated SMA.

The SMA actuator was provided with 3.4 A current for 2 s (in accordance with the SMA datasheet), and an off period of 20 s was set to cool the SMA down; in addition, during the cooling period, a deadweight of 100 g was attached to the SMA crimp to fully restore its original position.

Actuation was done for five cycles, and it was found to be highly repeatable; among the five cycles the standard deviation for the MCP, DIP, and PIP joints was in the (x, y) space, respectively (0.18mm, 0.19 mm), (0.2 mm, 0.25 mm), and (0.33 mm, 0.29 mm). The characterization setup is shown in Supplementary Figure S1. An average bending angle (calculated as shown in Supplementary Fig. S2) of 120° was found, and a very low standard deviation of 0.5° was calculated over five consecutive cycles. The proposed finger recalls human fingers not only because it is made up of stiff and soft materials but also in terms of flexibility and performed motions (Fig. 6b), paving the way for the fabrication of 3D-printed stiff robotic hands equipped with soft joints for the biomedical field (see Supplementary Movie S1).

With regard to the multidirectional bender, it was designed and fabricated to demonstrate that the proposed manufacturing approach can be applied to multiple fields, including surveillance and walking/crawling soft robots.

The multidirectional bender is shown in Figure 6d and e and it is made up of four different elements: (1) a stiff passive structure made up of NGF, (2) a soft TPU joint connected to the stiff structure by means of the T-shape mechanism, (3) two terminal plates where the stiff structure is anchored and the SMA actuators are positioned, and (4) eight SMA spring actuators.

The SMA actuators were placed in a square configuration and labeled from 1 to 8, as shown in Figure 6e. Each SMA actuator was individually activated (3.4 A for 2 s), and eight different bending motions were achieved due to the soft joint geometry, as shown in Figure 6e.

The proposed multidirectional bender is the first prototype that demonstrates how many motions a single 3D-printed object can achieve: by changing the design and connecting more benders together, an extremely wide range of motions can be obtained, and several kinds of robots can be easily manufactured.

Conclusions

In conclusion, the present article focuses on the multimaterial extrusion of stiff–soft structures and reports a method for improving their interfacial adhesion. Using the “gyroid” infill, the NGF composite material was found to be the stiffest commercially feasible material (among the three analyzed) having Young's modulus (E) of 1.4 GPa and tensile strength of 58.48 MPa. The stiff material was extruded in conjunction with a commercial soft TPU to fabricate bioinspired structures. Three different interface geometries were studied along with a process parameter (mesh overlapping) and a postprocessing (annealing). The optimal combination (maximizing the Young's modulus) of the three variables was determined to be as follows: T-shape, 0.4 mm, and 70°C for 1 h.

Tensile tests were performed for all the manufactured dog bones as a measure of the interface adhesion, and in the best case, E of 48.8 MPa and UTS of 3.8 MPa were achieved. In particular, the latter is 48% higher than the highest adhesion value between soft and stiff materials available in scientific literature. Finally, a bioinspired robotic finger actuated using SMA actuators was fabricated and tested, mimicking very well the human finger motion. The subsequent results in terms of stiff–soft adhesion pave the way for enormous exploitation of AM material-extrusion technology for the fabrication of soft robots: humanlike biomimetic joints, such as prosthetic hand and arm-wrist systems for humanoid robots, and other biomedical applications.

Moreover, future works will focus on the application of the findings on several baseline materials (such as PLA and ABS, widely used for hobbyist aims) in conjunction with soft TPU, to provide a 3D printing guideline useful at different levels, from makers to researchers.

Authors' Contributions

G.S.: Conceptualization, methodologies, and writing—original draft. S.M.A.I.O.: Conceptualization, methodologies, and writing—original draft. G.P.: Supervision, conceptualization, and writing—review and editing. R.Z.: Resources and editing. H.L.: Resources and editing. Y.T.: Supervision, conceptualization, and writing—review and editing.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No external funding have been used for the present research paper.

Supplementary Movie

3D printed robotic finger:

https://www.youtube.com/watch?v=TOMXni7mTeE

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Table S1

Supplementary Table S2

Supplementary Movie S1