Octopus-Inspired Adaptable Soft Grippers Based on 4D Printing: Numerical Modeling, Inverse Design, and Experimental Validation

g ) of 65°C. Fused deposition modeling was employed to fabricate soft grippers using the 3D printer, Ultimaker 2 Extend þ with a nozzle temperature of 200°C. Bilayer structures were printed with a line width of 0.4mm and a print layer height of 80 μ m. The printed soft grippers were actuated by placing in a water bath, Lichen hh-8. The actuating temperature of the soft grip-per was set as 85°C while the initial temperature was 25°C. Finally, the deformation of the structures was measured with the OLYMPUS DSX110 digital microscope for further analysis.The fabrication, stimulation, and measurement processes are shown in Figure S7, Supporting Information.

DOI: 10.1002/aisy.202200384 Soft grippers based on stimuli-responsive materials show great promise to perform safe interaction and adaptive functions in unstructured environments. Hence, improving flexibility and designability of the stimuli-responsive soft grippers for better grasping ability are highly desired. Inspired by the biological structure of octopuses, a class of temperature-driven polylactic acid grippers is fabricated by 4D printing in this work, which consists of two types of bilayer structures, separately imitating the web and tentacles of an octopus. Compared with the traditional pure-bending soft grippers, the integrated structure results in a 1.5 times wider reachable area and enhanced flexibility. The complex grasping behaviors are predicted with the 4D printing parameters (i.e., printing paths and printing speeds) using the reduced bilayer plate theory. Moreover, a method for determining the printing parameters of the gripper is provided to realize the desired grasping behavior depending on different objects. The feasibility of the design method is experimentally verified by gasping three distinctive objects, including an egg, a weight, and a Metatron's cube, which is in good agreement with the simulation results. Herein, a promising strategy is provided to achieve the easy-fabricated, low-cost, and versatile soft graspers using smart materials.
SMP bilayer structures have provoked interest in soft grippers since Van Manen proposed a single-step production process of SMPs to simplify the fabrication of starting materials. [37] Various soft grippers based on SMP bilayer structures are proposed, including tendril-inspired grippers (imitating helical distortions for grasping), [9,38] shell-inspired designable grippers, [39] dual-stimuli-responsive grippers, [40] and fast thermal responsive grippers. [41] These reported grippers are all programmable and easily fabricated due to the functional properties of the SMP under external stimuli and the feasibility of 4D printing in gripper fabrication. Nevertheless, most of the reported soft grippers exhibit simple deformation patterns and lack design methods to control shape deformation. Design principles for traditional rigid robots are no longer applicable for modeling the dynamics of soft robotics as a consequence of their nonlinear material properties. Thus, developing new modeling methods is necessary to predict their behaviors.
Deformation simulation of 4D printed structures can benefit the behavior control of these self-morphing structures. Recent studies on simulating models in 4D printing mostly focus on a single deformation pattern of the bilayer structures, such as pure bending [42][43][44][45][46] or helix. [38,47] Most of these models have not considered of the influence of the printing process on deformation and are too simple to simulate complex deformations and therefore are not practically applicable for complex deformations. Thus, a simulating model of 4D printed SMP bilayer structures was reported by our group based on Bartels' work, with printing parameters as model inputs, and performed well in predicting coupling deformations of the printed bilayer structures. [48] The inverse design of the 4D printed bilayer structures, namely solving the inverse problem of determining the stimulated strain in bilayer structures based on the target stimulated shape, has also aroused great interest in recent years. Rees et al. established a mathematical equivalence between bilayers and curved monolayers, solving the pre-strain in an initial bilayer plate to form a target curve. [49] However, for common 4D printing techniques, such as material-extrusion based on 3D printing (ME-3DP) and digital light printing (DLP), it's a challenge to program the smart materials to obtain an ideal pre-strain distribution if the growth rate and paths of the materials are changeful. To overcome this, several 4D printing strategies has been developed, such as 2D hydrogels using DLP, [50,51] thin LCE sheets using the top-down microfabrication technique, [52] and the dielectric elastomer actuators using the active layers and stiff rings patterning methods. [53] As for ME-3DP, a widely used technique in 4D printing, [27,[37][38][39][40][41]54,55] few reports focused on the inverse design of this technique. Wang et al. proposed an inverse design method for the 4D printed composite plates with embedded continuous fibers, which is fabricated using ME-3DP, while this method is not suitable for the 4D printing of a single smart material based on ME-3DP. [56,57] Gladman et al. solved the inverse problem of designing the ME-3DP paths of hydrogel bilayer structure for the target shapes, but ignored the influence from printing parameters to the shrinkage of hydrogels, which is also necessary in solving the inverse problem. [58] In this study, a class of temperature-driven shape-adaptive grippers, inspired by octopuses, has been developed by a single-step 4D printing method using polylactic acid (PLA). The low stiffness of PLA at high temperature realizes adaptive grasping of objects while the high stiffness after being cooled ensures the stability of grasping. Two types of bilayer structures with different deformation patterns, separately imitating the behaviors of the web and tentacles of an octopus, are proposed and integrated into the grippers, which enable more flexible grasping behaviors. In addition, the simulating model in our previous work has been improved by encoding the printing parameters as inputs, which expands the adaptability of the model in practice. The explicit relationship between printing parameters and deformation of the gripper is also derived based on the principle of our simulating model. Thus, the inverse design of the proposed soft grippers is achieved with the desired grasping behavior, which considers the practical process constraints of ME-3DP. We provide the complete design process of the soft gripper, with which the grasping behavior of the gripper is preprogrammed in fabrication, to grab different objects with various shapes and even fragile items.

Fabrication of Octopus-Inspired Grippers by 4D Printing
In this study, we focus on two critical parts of the octopus: web and tentacles, since they play an important role in the daily activities of an octopus, such as swimming or preying. The web of an octopus contracts and unfolds (red-dotted lines in Figure 1a), generating thrust underwater or wrapping up its prey. The tentacles can bend or twist (blue-dotted lines in Figure 1a), driven by the web, for walking or twining. Many complex behaviors of an octopus are completed by the cooperation between the web and tentacles. The variable stiffness of the muscular hydrostats in an octopus also ensures flexible and powerful deformation in these motions. Inspired by the octopuses, this work proposes a soft gripper structure consisting of a half-circle and three rectangles, which correspond to the web and tentacles, respectively, as shown in Figure 1b,c.
The soft grippers were printed by a 3D printer with a bilayer structure in different printing speed using PLA materials. The thermosensitive characteristics of PLA make it a good candidate to trap the stretching stress caused by printing and release it at high temperature, which benefits the deformation control of bio-inspired grippers. The softness in high temperature and stiffness at room temperature of the PLA resemble the variable stiffness of the muscular hydrostats in an octopus. The printing paths of the half-circle were concentric, while the printing paths of the rectangles were along the lengths of the rectangles, as shown in Figure 1b. The stress mismatch was generated due to different elongations along same stretching direction between the top and the bottom layers after cooling down. As a consequence, the deformation of the soft grippers can be well controlled: the half-circle imitates the web from expansion to contraction, and the rectangles exhibit pure bending with the same curvatures, following the behavior of the octopus (Figure 1c). The grasping process of the bio-inspired 4D-printing gripper is just like the predation of an octopus, which can be described as follows: when the gripper was actuated by heating, the half-circle bent and drove the three rectangles to curve while the rectangles also bent themselves, the synergy of which helps to grasp the objects in a flexible way. Once the stimulus was removed and the gripper was cooled, the PLA bilayer became stiff and sturdy. Finally, the gripper hoisted the object tightly and the grasping is finished.

Actuation Principle of the PLA Bilayer Rectangles
The residual stress remains in printed structures due to the repeated stretch and storage from the heating and cooling cycles during the 3D printing, providing an opportunity to control the programming of soft grippers. [35] Therefore, the 4D printed PLA bilayer structures will shrink and recover under thermal stimulus without an extra program process. The thermal shrinkage of the bilayer structure was modeled with the coefficient of thermal expansion (CTE) difference between the top and bottom layers. In this work, we mainly focus on the CTE difference along the printing paths in the bilayer structure, denoted with Δα. CTE differences in other directions are ignored as there is no difference in the printing process of the other directions. There are lots of printing parameters which will influence the prestrain regime of PLA including printing speed, nozzle temperature, printing line width, etc. [59] And it has been proved that the printing speed has a major influence on the prestrain regime of PLA along the www.advancedsciencenews.com www.advintellsyst.com printing paths. [60] Because exploring the relationships between the prestrain and the related printing parameters is not the purpose of this work, it's obvious that using a single variable to control the actuated deformation of PLA is most convenient and stable. Thus, the printing speed of the top layer was chosen as the only variable, denoted with v t , while other parameters were set as constants.
According to Timoshenko's thermal bending theory, [61] when triggered, the curvature radius R T of the bending rectangles in a gripper can be calculated with the CTE difference Δα between the two layers, by where t refers to the thickness of the pure-bending bilayer rectangles and Δt is the temperature change when triggered. With the measured curvature radii of the actuated rectangles in the grippers and Equation (1), the relationship between v t and Δα can be explored, which is expressed with a linear regression model in this work. Thus, the behavior of the octopus-inspired gripper can be designed by changing the printing speed of the top layer.

Deformation Simulation of 4D Printed PLA Bilayer Structures
Our group has applied the reduced bilayer plate theory to simulate thermally actuated bilayer plates. [48] Compared with the commonly used finite-element analysis software, this model has the advantages of stable convergence, high efficiency, and easily measured parameters. In this work, the simulating model is further upgraded for more complex situations such as a bilayer structure with varying strain principal axes. The inputs of the simulating model are also related to the printing process for adaptability. The improved model can simulate the deformation of arbitrary 4D printed PLA bilayer structures and is used for the behavior prediction of the octopus-inspired gripper. ω t ¼ ω Â t ⊂ R 3 refers to the domain of the bilayer structure, and the mappingm∶ω ! R 3 models the deformation of the midsurface in the bilayer structure before and after actuated, as shown in Figure 2. The reduced elastic energy function of the bilayer structure isẼ where μ is the second Lamé constant of the material in the bilayer structure, II is the second fundamental quantities of the deformed curve, ε r1 and ε r2 are the effective strain per unit thickness of the top and bottom layers, respectively, and C is a constant (see Section S1, Supporting Information). The thermosensitive characteristics of PLA are modeled with CTEs in our work, allowing the strain in a layer to be expressed as a product of the CTE tensor and the temperature difference. The effective strain per unit thickness is expressed as where α is the effective CTE tensor in the plane with the components along the thickness omitted, as shown in Figure 2a.
As analyzed before, the printing paths and speeds mainly determine the deformation of a 4D printed bilayer structure. Based on the transformation between two Cartesian coordinate systems (Figure 2a), the effective CTEs per unit thickness can be expressed as where α T is the CTE along the printing path and β is the angle between the printing direction and the horizontal direction. The difference between ε ⋅ r1 and ε ⋅ r2 can be further simplified as where α T1 and α T2 are the linear CTEs along the printing paths of the top and bottom layers, respectively. Thus, it is easy to find that the elastic energy function can be controlled with the CTE Figure 2. Mapping of a bilayer structure before and after deforming: a) the coefficient of thermal expansion (CTE) tensor transformation between two coordinate systems; b) modeling of the curvilinear printing path in the bilayer structure; and c) the deformation of the mid-surface in the bilayer structure.
www.advancedsciencenews.com www.advintellsyst.com difference, Δα, and the printing direction angle, β, which are related to the actual 4D printing process of PLA. The deformation simulation of the bilayer structure is actually minimizing the elastic energy function ! EðmÞ according to the minimum energy principle, giving for simplification. The weak form of the Euler-Lagrange equation for the elastic energy function is derived as with the isometric mapping constraint ½dm T dm ¼ I 2 , wherew is the test function satisfying the condition The displacement mappingm which minimizes the potential function can be obtained by solving Equation (7). The time and space discrete methods proposed by Bartels are used to solve Equation (7) (see Section S3, Supporting Information). [62] The simulating model on complicated deformations of 4D printed PLA structures is thus built by visualizing the solved mid-surface displacementm.
If the printing paths and speeds are known, by obtaining the angle β and the CTE difference Δα, the actuated deformation of arbitrary 4D printed PLA bilayer structure can be simulated. Behavior prediction of the octopus-inspired grippers was also realized using the improved model in our work.

Behavior Modeling of the Octopus-Inspired Gripper
The behavior modeling was put forward as the basis for the controllable deformation of the octopus-inspired gripper. Figure 3a illustrates an actuated octopus-inspired gripper with internal radius r and outside radius R of the half-circle (Figure 1b).
The hollow truncated cone, which is deformed from the half-circle, is quantified with the top diameter d, the bottom diameter D, and the half vertex angle γ, as shown in Figure 3a. The angle γ c , as the complementary angle of γ, is set as the quantitative indicator of the deformed half-circle, which increases as the deformation increases. The opening angle θ in the top view is the denoted measurement to calculate γ c for its easily measurable characteristic. With the isometry assumption used in our model, relations between the measurement θ and the deformation angle γ c can be deduced as To simplify the behavior design process of a gripper, the grasping space is defined with radius R g indicated by a sphere with a red border in Figure 3a, tangent to the generatrices of the truncated cone. Then, the design process of the bilayer half-circle determines the deformation angle γ c based on the grasping space radius R g , which is mainly dictated by the shape and size of the grasped object and the thickness of the gripper, and the length a, which is the distance from the grasping space to the cone point of the hollow truncated cone as shown in Figure 3a. An equation can be listed based on the geometric relationship thus, the grasping space radius R g can be indicated with the half vertex angle γ as where a is a constant selected at the beginning of design. For intuitive expression of the distance a, a factor k is defined as resulting in Equation (11) becoming Figure 3. Grasping behavior modeling of the octopus-inspired gripper: a) grasping space and parameters of an actuated octopus-inspired gripper; b) reachable workspace of the octopus-inspired gripper and the pure-bending gripper. The yellow parts show the deformation process of the pure-bending gripper to the limit state and the faint yellow area is the swept area indicating the reachable workspace of the pure-bending gripper. The blue and purple parts are two limit states of the octopus-inspired gripper (deformed from two different octopus-inspired grippers) and the faint blue area indicates the reachable workspace of the octopus-inspired gripper.
www.advancedsciencenews.com www.advintellsyst.com Then the deformation angle γ c can be solved with the determined grasping space radius R g , the internal radius r, and the factor k using Equation (9) and (13). The deforming behavior of the half-circle in a gripper can therefore be designed. As for the design of the rectangles in a gripper, the critical parameters of the rectangles are the length L (Figure 1b) and the deformation curvature radii R g (Figure 3a), which are designed to adapt to the shape of the grasped objects.
Most proposed stimuli-responsive soft grippers only used actuators with a single deformation pattern, usually pure bending, to grasp objects. [20][21][22]24,[26][27][28][29]31,35,36] Figure 3b shows the comparison of the reachable workspace between an octopus-inspired gripper and a pure-bending soft gripper. The octopus-inspired gripper and the pure-bending soft gripper have the same length for comparability, where the length of an octopus-inspired gripper refers to R þ L. Defining the limit state as the state where the interference of the gripper occurs at the center line, the reachable workspace of the pure-bending gripper is calculated as the swept area from the initial state to the limit state, as denoted by the faint yellow area in Figure 3b. There are two critical limit states of the octopus-inspired gripper, limit state 1 where the rectangle completely deforms, and limit state 2 where the half-circle completely deforms. Similarly, the reachable workspace of the octopus-inspired gripper is obtained as the swept area from the initial state to the limit states, as denoted by the faint blue area in Figure 3b. By comparing the areas of the two workspaces (1697.6 vs 1120.7 mm 2 while R = 20 mm and L = 30 mm, measured by Autocad 2023), the synergy of two actuators in the octopus-inspired grippers gains a roughly 1.5 times wider reachable area than the proposed pure-bending grippers, due to the increased flexibility.

Inverse Design on Printing Speed of the Octopus-Inspired Grippers
Exploring the relationship between printing parameters and grasping behaviors is necessary and critical for the inverse design of the octopus-inspired gripper. In our work, the printing speed of the top layer was chosen as the design variable to control the behavior of the gripper and the relationship between v t and Δα was explored. Figure 4a shows the print sequences of the octopus-inspired gripper, first printing the half-circle followed by the three rectangles. Considering the actual printing process, the printing speed of the printer was controlled with G-code and therefore an uneven speed phase during each speed change can be assumed; hence, fewer changes in speed results in better stable speed control. The printing speeds of the top layer of the half-circle and the rectangles are thus set as two constants, denoted with v H and v R separately. Accordingly, the CTE differences along the printing paths of the half-circle and the rectangles are Δα H and Δα R . The printing speed of rectangles v R can be easily determined with the desired bending curvature radii R T using Equation (1) and v t -Δα relationship. The remaining constant, v H , denotes the printing speed of the half-circle with a complicated deformation. Figure 4b shows a half-circle surface with the parameters (u 1 , v 1 ), and Figure 4c shows the deformed truncated conical surface with the parameters (u 2 ,v 2 ). The second fundamental quantities of the truncated conical surface in the parameters (u 1 ,v 1 ) can be derived as (see Section S2, Supporting Information) Equation (2) indicates that the stable state of the deforming curve is them to minimize the ∫ ω jII þ Zj 2 . The relationship between v 1 and the printing direction angle β in the half-circle can be easily found as β = À(π/2À v 1 ). The transformation matrix from the global coordinate system in Figure 2a to the polar coordinate system (u 1 ,v 1 ) is Thus, the Z matrix of the half-circle, also regarded as a tensor, can be expressed in the polar coordinate system (u 1 ,v 1 ) as Then, the integral ∫ ω jII H þ Z H j 2 can be explicitly expressed as By solving the equation of the derivative of F the relationship between the half vertex angle γ and the CTE difference Δα H of the half-circle can be calculated as Then, the printing speed of the half-circle v H can be obtained with the desired half vertex angle γ and the v t -Δα relationship.
Thus, for the complicated target curve with integrable second fundamental quantities, the actual constraint of the printing process can be coded with matrix Z and solved. This provides a solution for the inverse design problem of the kind of 4D printed bilayer structures, namely calculating the printing parameters based on the target deformation under practical ME-3DP process constraints.

Grasping Behavior Control with Printing Parameters
Eight classes of the soft grippers were fabricated at different printing speeds of the top layer, from 500 to 1200 mm min À1 with an interval of 100 mm min À1 , while the bottom layer printing speed was set as 300 mm min À1 . Other parameters of the grippers are the same, including the outside radius R = 20 mm, the internal radius r = 10 mm, the rectangle length L = 30 mm, and the total thickness t = 1.6 mm with a thickness ratio of 1:1. Each kind of gripper was printed in three copies, namely nine pure-bending rectangles, and three half-circles were actuated in each class of experiments. The curvature radii of all rectangles were measured and converted into Δα using Equation (1), with the results listed in Table S1, Supporting Information. A linear regression model was established to figure out the relationship between v t and Δα as Δα ¼ 5.989 Â 10 À4 À 1.688 Â 10 À6 v t (20) as shown in Figure 5a. The opening angles θ of the bilayer halfcircles in all actuated grippers were measured. Figure 5b and Table S2, Supporting Information, made the comparison of the experimental γ c , theoretically calculated γ c with Equation (9) and (19), and the simulated γ ' c based on the calculated Δα' with Equation (20), demonstrating that the theoretical values and the simulation model were consistent with the experimental data. Some images of experiments and simulations of the octopus-inspired grippers are given in Figure 5c,d, demonstrating that the simulations are in good agreement with the corresponding experiments. The consistency between theoretical values, simulations, and experiments also showed that the linear approximation of the relationship between v t and Δα is reasonable, and changing printing speeds is a reliable way to control the grasping behaviors of the gripper.

Object-Based Design Process of the Octopus-Inspired Gripper
To design the soft gripper, some parameters of the grippers need to be predetermined first based on the grasping object, including the outside radius R, the internal radius r, and the total thickness t. The grasping space radius R g was then decided according to the shape and size of the object. The half top angle γ was determined based on R g by Equation (11) with the chosen scale factor k. With these parameters, the desired deforming half-circle and the grasped object can be drawn in the CAD software. Thus, the critical sizes of the bilayer rectangles including the length L and the appropriate deforming curvature radius R T can be also determined by drawing.
The printing speed of the top layer of the soft grippers was varied, while the printing speed of the bottom layer was set at a constant 300 mm min À1 . For the half-circles, the top-layer printing speed v H can be obtained with γ by combining Equation (19) and (20). As for the rectangles, the top-layer printing speed v R can be calculated with the desired curvature radius R T using Equation (1) and (20) and therefore the soft grippers can be fabricated with the designed parameters. The grasping behavior of the soft gripper can also be predicted with the extended simulating model for verification.

Applications to Grasp Universal Objects
To exhibit the grasping ability of the soft gripper, three distinctive objects were chosen to be grasped, including a delicate egg, a heavy weight, and an irregular Metatron's cube. All the grippers in applications had the same total thickness of 1.6 mm with the thickness ratio of 1:1 and the same bottom layer printing speed of 300 mm min À1 . Table 1 shows the critical parameters (meanings of these parameters have been declared in Section 2.4 and 2.5) of the grippers designed for grasping the three objects. The design, simulation results of the grippers, and the grasping   www.advancedsciencenews.com www.advintellsyst.com process of these three objects are shown in Figure 6. The grasping experiments are implemented in a water bath, Lichen hh-8. The object is first put into the water bath. When the water has been warmed to 85°C, which usually takes about 30 min, the soft gripper is laid down and actuated by the hot water, grasping the object in 1 min. After being cooled down to 45°C to ensure the stiffness of the soft gripper, which usually takes another 30 min, the object is hoisted with the soft gripper within 30 s and the grasping is completed. The successful grasping of the egg, the weight of 200.0 g (which is 150 times the gripper's weight), and the Metatron's cube, respectively, exhibit the soft gripper's abilities of grabbing fragile objects, lift heavy objects, and grasp irregularities. Good agreement between experimental tests and numerical simulations are achieved, which demonstrates the feasibility of the inverse design (see Section S5 and S6, Supporting Information, for more details of the design, simulation, and experiment results of the grippers in applications).

Conclusion
In this study, a class of octopus-inspired soft grippers was proposed based on 4D printing technology. The gripper consists of a bilayer half-circle and three bilayer rectangles, imitating the behaviors of the web and tentacles of an octopus, respectively. The soft grippers were made by PLA, the characteristics of which ensure variable stiffness of the bio-inspired gripper for flexible and stable grasping, like the muscular hydrostat in an octopus. The deformation mechanism of the gripper is attributed to the change of the difference in CTEs between two layers caused by the printing process. The complex deformation stimulation of the 4D printed bilayer structures was performed by the reduced bilayer plate theory, as well as predicting the behavior of the soft grippers. The relationship between the printing speeds of the top layer and the difference of CTEs between the two layers were explored and derived for a better deformation control of the gripper. The motions of the soft grippers corroborated well with the simulated results and the theoretical analysis results, verifying the validity of both the inverse design and the simulating method. Based on the design process, a bio-inspired gripper was fabricated that was able to conform to an object's shape and actuate to achieve grasping in a simple and efficient way. Grasping tests with objects of different properties including an egg, a weight, and a Metatron's cube were carried out to demonstrate the ability to lift fragile, heavy, and irregular items. While the proposed PLA-based soft grippers exhibit promising capabilities, they do have some limitations: they can't be actuated repeatedly, thus are disposable, and have a relatively long cycle time for grasping operation, which can be attributed to the properties of PLA. However, it is worth mentioning that the bio-inspired structure and the design process proposed in this study are universal: the structure, behavior simulating, and the inverse design process can be implemented with other functional materials which can be programmed with ME-3DP, not only limited to PLA. And the performance of soft gripper can be greatly enhanced by using functional materials with superior properties. The unrepeated limitation can be overcome by using responsive reversible materials such as hydrogel, while the cycle time can be shortened by using conductive PLA. Thus, the easily fabricated 4D printing technique and effective design method show great potential for universal grasping for industrial applications as the development of functional materials.
There are several significant points we plan to research further in the future: 1) the relationships between the programmed CTE of PLA and the related printing parameters, such as nozzle temperature, printing line width, and layer thickness, still need to be explored by the characterization or experimentation on PLA; 2) based on the explored relationships, the optimization model can be established to decide the optimal printing parameters for more accurate control of the actuated deformation of PLA; and 3) the sensing performances of some electrical smart materials, such as graphene or conductive PLA, can be researched, which shows potential to be printed and integrated in the 4D printed structures, to sense the complex actuated deformation.

Experimental Section
The PLA used in our work to fabricate the bilayer grippers was purchased from Ultimaker (Tough PLA) with a glass transition temperature (T g ) of 65°C. Fused deposition modeling was employed to fabricate soft grippers using the 3D printer, Ultimaker 2 Extendþ with a nozzle temperature of 200°C. Bilayer structures were printed with a line width of 0.4 mm and a print layer height of 80 μm. The printed soft grippers were actuated by placing in a water bath, Lichen hh-8. The actuating temperature of the soft gripper was set as 85°C while the initial temperature was 25°C. Finally, the deformation of the structures was measured with the OLYMPUS DSX110 digital microscope for further analysis. The fabrication, stimulation, and measurement processes are shown in Figure S7, Supporting Information.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.