A Soft Gripper with Rigidity Tunable Elastomer Strips as Ligaments

Abstract

Like their natural counterparts, soft bioinspired robots capable of actively tuning their mechanical rigidity can rapidly transition between a broad range of motor tasks—from lifting heavy loads to dexterous manipulation of delicate objects. Reversible rigidity tuning also enables soft robot actuators to reroute their internal loading and alter their mode of deformation in response to intrinsic activation. In this study, we demonstrate this principle with a three-fingered pneumatic gripper that contains “programmable” ligaments that change stiffness when activated with electrical current. The ligaments are composed of a conductive, thermoplastic elastomer composite that reversibly softens under resistive heating. Depending on which ligaments are activated, the gripper will bend inward to pick up an object, bend laterally to twist it, and bend outward to release it. All of the gripper motions are generated with a single pneumatic source of pressure. An activation–deactivation cycle can be completed within 15 s. The ability to incorporate electrically programmable ligaments in a pneumatic or hydraulic actuator has the potential to enhance versatility and reduce dependency on tubing and valves.

Introduction

Natural organisms use striated muscles, catch connective tissues, or muscular hydrostats to alter the mechanical rigidity of their body and limbs.^1–3 Such stiffness tuning functionality is critical in motor tasks such as grasping that rely on reversible transitions between gentle, mechanically passive contact and heavy load bearing. Similarly, bioinspired and soft robotic grasping systems depend on active rigidity tuning to adapt their mode of operation and rapidly transition between different motor tasks. To match the mechanical properties and versatility of natural grippers, such mechanisms should be lightweight and capable of reversibly changing their effective elastic moduli between 0.1–1 MPa and 10–100 MPa, which correspond to the moduli of soft tissues and hard cartilages in humans.^1–4 Moreover, engineering these mechanisms to respond to moderate electrical voltages enables compatibility with conventional electronics and batteries and limits dependency on pneumatics, hydraulics, motors, internal combustion engines, or other typically bulky mechanical hardware.

In this article, we introduce an electrically controlled soft gripper (Fig. 1) consisting of three soft pneumatic fingers with rigidity tunable elastomer strips as ligaments attached on the sidewalls. Resistive heating through the strips can reversibly change the mechanical rigidity of the ligaments. By applying air pressure, pneumatic fingers bend in the opposite direction of the activated softer ligaments. Collectively, the three fingers can achieve not only grasping and releasing manipulations but also twisting. To validate our experimental results, we also perform finite element analysis (FEA) on a 3D model of the finger to predict the grasping and twisting behavior of the gripper, which is found to be in good agreement with the experimental results. The gripper can pick up a glass ball weighing 22 g, along with many other objects with complicated geometries, under an input pressure of 70 kPa. It can also twist a ping-pong ball by 18° while grasping it. All the grasping and twisting manipulations can be performed within seconds. In contrast to existing techniques for dexterous pneumatic grasping, this soft gripper (1) operates with only a single internal air pressure and (2) uses electrically stimulated ligaments to control the direction of finger motion. Such an approach greatly simplifies design by replacing pneumatic tubing, valves, and pressure sources with electrical wiring, switches, and a power supply such as a battery.

FIG. 1.

Principles of operation for the soft gripper. (a) Range of the tunable rigidity for cPBE. (b) The proposed soft gripper consists of three soft fingers, three finger holders, and two table top metal fixtures. (c) The soft finger consists of three PDMS phalanges, two Ecoflex joints, and three rigidity tunable ligaments. (d) Undeformed cross-sectional view of the soft finger. (e) Bottom view of the nonactivated (green) and activated gripper when the three outer ligaments (red) of the fingers are activated and generate a grasping configuration. (f) When the two inner ligaments of each finger are activated, a dropping manipulation is achieved. (g) When the outer and one of the inner ligaments of each finger are activated, a twisting manipulation is achieved. cPBE, conductive propylene-based elastomer; PDMS, polydimethylsiloxane. Color images available online at www.liebertpub.com/soro

Background

Soft grippers abound in nature, such as lizard tongues, octopus arms, and elephant trunks. These grippers can work with high degrees of freedom and generate a broad range of motions, including bending, extending, and twisting.⁵ The flexibility and adaptability of these natural soft grippers exhibit huge advantages over their engineered rigid counterparts, for which accurate displacement and force control are typically the priority. The potential benefits of biomimetic soft robotics have been highlighted in application domains ranging from healthcare to military.⁴ Such systems are lightweight, elastically deformable, maneuverable in confined spaces, and safe for human–robot interaction. Moreover, their high mechanical compliance allows for intrinsic and robust force control that can accommodate large displacements or positioning inaccuracies. Driven by the increasing demand on soft robotics, there has been recent growing interest in soft multifunctional materials that can actively tune their mechanical properties and be used in applications such as soft gripping.^6,7

Inspired by nature, many soft grippers have been designed using pneumatic actuation mechanisms.^8–11 In these soft grippers, mechanical asymmetry created by air pressure control provides bending motion of individual fingers, which is used collectively to generate the grasping motion of the grippers.^12,13 For example, the pneumatic soft gripper developed by Suzumori et al. consisted of four elastomeric fingers and each contained three isolated pressurized channels that allowed for independent pneumatic control.¹³ Kwok et al. used fingers with a single pneumatic channel for their soft gripper design, in which the sidewall of the pneumatic channel was composed of two elastomeric halves with different rigidity such that mechanical asymmetry and thus bending are achieved when pressurized. They used magnetic joints that can assemble multiple modules of soft robotics.¹⁴ In another study, a pneumatic universal robotic gripper was introduced based on jamming of granular materials, which is capable of picking up objects with a wide range of sizes and shapes.¹⁵ Because they are soft, these grippers can grasp a broad range of objects with a single motion. However, more sophisticated manipulation typically requires a dedicated pump or valve for each additional grasping mode or degree of freedom.

In another group of studies, soft actuators are designed with electrically activated dielectric thin films embedded in elastomeric matrix.^16–19 For example, Araromi et al. introduced an electrically activated gripper based on prestretched dielectric elastomers.¹⁷ The gripper consists of four “fingers,” which are prestretched dielectric elastomers bonded on a flexible frame whose bending angle can be tuned by high voltage input. In another study, a soft gripper was introduced in which stiff fibers are embedded in a dielectric elastomer film for wrapping motion generation.¹⁸ Most recently, a two-finger soft gripper was proposed based on electrostatic actuation and intrinsic electroadhesion force.²⁰ The gripper consists of a prestretched elastomer membrane, patterned compliant electrodes, and two passive silicone layers. It can pick up a wide range of objects, including fragile and deformable ones. However, the high voltages required for these grippers (∼1–10 kV) can sometimes pose safety issues and impose restrictions on the operating environment.

Other than the pneumatically and electrically activated grippers described above, soft grippers based on the combination of two or more mechanisms have also been introduced.^21–23 For example, to pick up objects with flat surfaces, Yoshimi et al. designed a soft gripper consisting of two parallel soft fingers with a soft nail mounted on one of the fingers.²¹ Cheng et al. designed and fabricated an elephant trunk-shaped gripper that works with an evacuation pump and tension cables.²³ The gripper has four tension cables connected to off-board spooler motors that can control and position the gripper. Local areas with tunable stiffness achieved by reversible jamming of granular media are used to fixate the motion. Giannaccini et al. also introduced an elephant trunk-shaped gripper that picks up and holds roughly cylindrical objects.²² Finally, pick-and-place manipulation has been demonstrated using adhesion-controlled techniques based on rate-dependent adhesion hysteresis²⁴ and bioinspired adhesion.^25,26

Inspired by nature, materials with tunable stiffness have recently been used for the design of wearable devices and soft robotics. There have been different mechanisms to achieve tunable rigidity, including electrical,^27–29 thermal,³⁰ magnetical,³¹ pneumatic,³² and chemical³³ approaches. In a recent work, Shan et al. engineered an electrically conductive propylene-based elastomer (cPBE) that can change its rigidity rapidly and reversibly through resistive heating.³⁴ This elastomer can switch its elastic rigidity between two states—from 175 MPa at room temperature to roughly 1 MPa at activated state, as shown in Figure1a—when it is heated above or cooled off its glass transition temperature at 75°C.³⁴ Such a range is similar to the change in stiffness exhibited by human skeletal muscles and catch connective tissue in sea cucumbers.

Principles of Operation

The schematic of the soft gripper is presented in Figure 1b, where three soft fingers are attached to a rigid fixture with sliding positioners. The soft fingers are made of silicone elastomers—Sylgard 184 polydimethylsiloxane (PDMS) and Ecoflex—as well as three cPBE-PDMS rigidity tunable multifunctional composites as ligaments. Each finger is composed of three phalanges made of PDMS and two joints made of Ecoflex similar to the human hand fingers (Fig. 1c). Both the phalanges and the joints are hollow such that each finger has a hollow chamber inside (Fig. 1d). There are three extensor ligaments made of cPBE-PDMS rigidity tunable elastomers positioned axisymmetrically around the finger sidewall; this feature provides the capability of bending in three as well as the reverse directions for the soft finger, enabling it to bend with even more flexibility than human fingers.

When its hollow chamber is inflated, the soft finger tends to expand in both the radial and the axial directions. Since the ligaments pose constrains for axial expansion of the finger, expansion mostly occurs in the radial direction. Furthermore, since the Ecoflex joints are about 20× softer than the phalanges, the expansion mostly occurs in the joints. Due to the symmetry in geometry and mechanical properties of the finger, the axial force will be equally balanced among the three ligaments. The radial forces applying on the inner surface of a symmetric chamber are in balance as well.

When one of the ligaments is activated, its rigidity decreases by about 35× . This disrupts the original symmetry in mechanical properties of the soft finger and causes the activated ligament to extend more than the nonactivated pair. Once the activated strip starts extending, the distal phalange will tilt toward the original axis of symmetry of the finger chamber, in the bending symmetry plane of the activated strip. The finger will continue to bend until it reaches a configuration where force balance in the bending symmetric plane of the activated ligament is reached again. Thus, by selectively activating different combinations of the ligaments, different manipulations such as grasping, releasing, and twisting can be achieved by the gripper, as illustrated in Figure 1e–g.

To restore the finger to its original configuration before activation, the air pressure inside the hollow finger should be decreased to atmospheric pressure while the ligaments are still in activated state such that they can return to their original length due to the hyperelasticity of the PDMS component of the cPBE-PDMS composite.

Materials and Methods

Materials

The soft pneumatic finger is made of PDMS phalanges, Ecoflex joints, and cPBE-PDMS rigidity tunable composite ligaments. The PDMS and the Ecoflex joints are composed of Sylgard 184 (Dow Corning, Inc.) and Ecoflex 0050 (Smooth-on, Inc.), respectively. Both segments are produced with elastomer casting using 3D printed molds (Objet 24; Stratasys, Ltd.).

The cPBE was produced by blending a propylene/ethylene copolymer with a percolating network of structured carbon black, for which details about the manufacturing process can be found in Shan et al.³⁴ Flattened sheets of cPBE were patterned with a CO₂ laser (Epilog Helix 24; Epilog Laser, Inc.) to form shapes with electrical terminals to supply current (Fig. 2c). With their ends wrapped by thin copper shim wires, the U-shaped cPBE strips were attached onto the bottom of a Petri dish using small pieces of double-sided tape (VHB, 3M, Inc.) such that the cPBE strips were 0.4 mm above the bottom surface of the Petri dish. Then, uncured PDMS solution was poured into the Petri dish to submerge the cPBE strips such that PDMS can embed the cPBE right at halfway of its depth. Next, the composite was cured in a vacuum oven (Across International, Inc.) for 1.5 h at 80°C. Finally, cPBE-PDMS composite strips were carved out of the cured sheet (Fig. 2f).

FIG. 2.

Fabrication steps for rigidity tunable strips made of cPBE-PDMS composite. (a) Start with a flat smooth metal substrate. (b) Cover with a layer of cPBE. (c) Pattern with a CO₂ laser. (d) Remove the excess cPBE, attach thin strips of copper shim to the terminals, and attach the strips with double-sided tape (VHB, 3M, Inc.) to the substrate. (e) Embed with uncured PDMS and then cure PDMS. (f) Cut edges and release. Color images available online at www.liebertpub.com/soro

Fabrication of the soft finger

The PDMS phalanges and Ecoflex joints were designed to have female and male surface features that can fit with each other. This mating is intended to increase the bonding between different materials during assembly. When attaching a joint to a phalange, uncured PDMS was applied to the interface of the phalange and then the joint was placed on top of that and cured in the oven for 40 min at 90°C. To avoid flowing of uncured PDMS onto the joint surface during curing in the oven, which could affect the finger's functionality, the PDMS segment was always kept on the bottom, underneath the Ecoflex joint.

After assembling all the phalanges and joints, cPBE-PDMS strips were attached onto the sidewall of the phalanges and cured one by one. To avoid flowing of the PDMS in the assembly process, a mixture of half-cured and uncured PDMS as bonding agent was used. Finally, a soft latex rubber tube (McMaster-Carr, Inc.) was inserted into a small hole in one end of the finger and sealed with uncured PDMS. This tube was connected to the air pressure regulator.

Assembly of the gripper

As shown in Figure 1, three fingers were inserted into three 3D printed finger holders, which were bolted to the rails on an L-shaped metallic fixture to form a gripper. The opening between the fingers and the vertical position of the gripper can be manually adjusted. Next, the copper leads of the fingers' ligaments were connected to an electrical board, where electrical circuits and switches can be designed to selectively activate any combination of the nine ligaments (strips). The electrical board was connected to an electrical power supply (GPR-30H10D; GW Instek, Inc.), which can be used to activate selected ligaments. Finally, the three rubber tubes coming out of the fingers' chambers were connected to a single air pressure regulator (PneumaticPlus, Inc.), which was connected to a source of high-pressure air in the laboratory.

Mechanical behavior of the soft finger

To characterize the effective force and deflection of the fingertip under different air pressure inputs, a soft finger was mounted horizontally using the aforementioned fixtures on the table such that one of the ligaments was positioned horizontally on the top, parallel to the table surface (Fig. 3a, b). A supporting horizontal fixture was used to make sure that at the beginning of the tests the finger was at a horizontal position.

FIG. 3.

(a) Mechanical testing setup for individual soft finger bending force measurements. (b) Mechanical testing setup for individual soft finger fingertip deflection measurements. (c) Assembled gripper with all supporting hardware, including a pressure gauge, rubber tubes, a power supply, and an electrical board. Color images available online at www.liebertpub.com/soro

To measure the bending force at the fingertip after activation of the finger, a digital scale (CPA224S; Sartorius, Inc.) was positioned under the fingertip such that the scale pan barely touched the fingertip before removing the supporting horizontal fixture. To simulate the frictionless support at the interface of the scale pan and the fingertip, olive oil was applied on the surface of the scale pan. Finally, the top ligament was activated, the supporting fixture was removed, and the gauge pressure was increased from 0 to 70 kPa in 10 kPa increments to measure the effective forces at different air pressure inputs (Fig. 3a).

To measure the fingertip deflections under different air pressure inputs, the top ligament was activated, the supporting fixture was removed, and then the gauge pressure was increased from 0 to 70 kPa in 10 kPa increments. The deflections were recorded at each increment with a vertical ruler (Fig. 3b). Due to the self-weight of the finger, in both force and deflection measurements, there will be an initial value when the gauge pressure is zero.

Mechanical characterization of elastomers

A Young's modulus value of 1.04 MPa was reported in Shan et al.³⁴ for cured PDMS. However, in this study, PDMS phalanges were cured in the oven several times (about 5 h in total at 90°C) during the process of assembling different components. Therefore, a series of tensile tests were independently performed on homogeneous PDMS specimens cured multiple times in the oven. The Young's modulus is found to be E_PDMS = 4.12 ± 0.20 MPa based on testing of four samples, which is about 4 × higher than the reported values in other literatures.³⁴

Due to the large deformation regime encountered in the finger joints during their functioning, the nonlinear hyperelastic neo-Hookean model is used to capture the stress/strain behavior of the Ecoflex parts.³⁵ Tensile tests were performed on homogeneous oven-cured Ecoflex samples. Using the relationships in Equations 1 –3,³⁵ the initial shear modulus μ_Eco is found to be 0.0638 ± 0.00637 MPa based on testing of three samples. The incompressibility parameter d is thus found to be 0.00626 (MPa)⁻¹ using a Poisson ratio ν_Eco = 0.499. Here in the equations, σ is the stress, λ is the stretch, ɛ is the strain, and κ is the bulk modulus: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \sigma = \mu ( { \lambda ^2 } - \frac { 1 } { \lambda } ) \tag { 1 } \end{align*} \end{document} \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \lambda = 1 + \varepsilon \tag{2} \end{align*} \end{document} \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \kappa = \frac { 2 } { d } = \frac { \mu } { { 1 - 2 \nu } } . \tag { 3 } \end{align*} \end{document}

Finite element analysis

To better understand the mechanical behavior of the proposed finger, simulations of FEA were conducted with the commercial FEA package ANSYS™. A 3D model of the finger with the real finger dimensions was used to simulate the stress/strain distribution during functioning as well as the deflection and reaction force at the fingertip. Due to the twofold symmetry of the finger geometry, only a half of the finger was modeled in ANSYS. A fixed support boundary condition was applied at the base of the finger and the fingertip was free to move. Air pressure was applied inside the finger chamber normal to the chamber surfaces. Frictionless support boundary condition was implemented on the symmetry plane of the finger. Frictionless contacts were defined between the contacting surfaces of the ligaments and the finger phalanges such that the ligaments can freely slide on the finger without penetration. The finger model was reproduced using PDMS, Ecoflex, and cPBE-PDMS composite properties for finger phalanges, joints, and ligaments, respectively. Isotropic, homogeneous, and linear elastic properties were assigned to the finger phalanges and ligaments due to small strain regime (<15%) in them. Poisson's ratio of the PDMS was set to be 0.499 and its Young's modulus was set to be E_PDMS = 4.12 MPa as measured. The cPBE-PDMS composite strips were treated as a homogenous material with a Young's modulus of 30.12 MPa before activation, and 0.86 MPa after activation, calculated using methods and data from Shan et al.³⁴

Taking the self-weight of the finger into account, the model was run for two scenarios corresponding to the mechanical characterization testing conducted for individual fingers: (1) there was a horizontal frictionless support at the fingertip with no vertical deflection allowed, to find the effective bending force of the fingertip at various pressure inputs and (2) there was no constraint on displacement of the fingertip, to compute the deflection of the fingertip at various pressure inputs.

Results and Discussion

The soft fingers were fabricated with two different distal phalange lengths using the methods described earlier and the gripper was assembled as shown in Figure 3c. Fingers with a shorter distal phalange were used for both characterization testing and FEA. For the deflection test, where there was no constraint on the displacement of the fingertip of a horizontally positioned finger, an FEA simulation was conducted taking the self-weight of the finger into consideration. The stress and strain distributions within the half finger model at p = 70 kPa are shown in Figure 4, which confirm the small deformation assumption for PDMS region and finite deformation assumption for Ecoflex region. The highest strain and stress at p = 70 kPa are 92% on the inner surface of the second joint and 0.59 MPa at the attaching point of the strips to the body, respectively.

FIG. 4.

(a) Stress distribution in the half finger model for the deflection test when the input air pressure is 70 kPa. (b) Strain distribution in the half finger model for the deflection test when the input air pressure is 70 kPa. Color images available online at www.liebertpub.com/soro

The deflections at the fingertip under increasing air pressure inputs predicted by the FEA and measured in the experiments are compared in Figure 5a. As shown, the computational predictions agree well with the experimental results. A 1.93 mm initial deflection due to self-weight at the fingertip was predicted by the FEA and verified by the experiment before applying the pressure. The fingertip deflection at p = 70 kPa predicted by FEA is 8.29 mm, whereas the measured value is 9.60 mm, which is a 15.8% difference based on the FEA results. The experiment was stopped at 70 kPa because for higher pressure, leaking from the interface between the joints and the phalanges might happen. What is more, the slope of the curve increases with increasing air pressure, which suggests that improving the interfacial strength between the joints and the phalanges will enable larger deflections.

FIG. 5.

(a) Comparison between the experimental and modeling results of fingertip deflections versus different air pressure inputs; the error bars are all based on three tests and the deflection direction is downward. (b) Deformed half finger when the input air pressure is 70 kPa showing the displacement distribution of the finger body. Color images available online at www.liebertpub.com/soro

Figure 6 shows the experimental and modeling results of the fingertip bending force when the fingertip was not allowed to move in the vertical direction. As observed, the body of the finger moved upward due to constraint applied at the fingertip. The maximum strain and stress predicted by the FEA at p = 70 kPa are 88% on the inner surface of the second joint and 0.66 MPa at the attaching point of the strips to the body, respectively. The highest deflection is also seen on the inner surface of the second joint to be 4.29 mm, which is expected due to the fixed support at the finger base and the frictionless support on the fingertip in the horizontal plane. Comparing the experimental and modeling results, the maximum difference is found to be 11.5% when the input air pressure is p = 30 kPa. Similar to the deflection plot in Figure 5, the bending force at the fingertip increases nonlinearly as the pressure inside the finger chamber increased. So once again, it implies that improving the interfacial strength between the joints and the phalanges becomes critical in achieving higher gripping forces.

FIG. 6.

(a) Comparison between the experimental and modeling results of fingertip bending forces versus different air pressure inputs; the error bars are all based on three tests. (b) Deformed half finger when the input air pressure is 70 kPa showing the strain distribution of the finger body. Color images available online at www.liebertpub.com/soro

After the individual finger behavior was calibrated by FEA and experiments, we used the assembled gripper to conduct grasping, releasing, and twisting experiments with various objects. A simple electrical circuit was used to control the activation sequences of the strips, with two switches designated for activating grasping and twisting, respectively (Supplementary Movie S1; Supplementary Data are available online at www.liebertpub.com/soro). To achieve grasping, the three outer ligaments of the soft gripper (Fig. 1e) were activated at the same time. Figure 7a and b shows that the gripper picked up a ping-pong ball and a roll of tape with ease. The weights of the ping-pong ball and the tape roll are 2.1 and 10.2 g, respectively. The gripper was not able to pick up a glass ball of 22 g weight due to the lack of friction between the PDMS fingertip surface and the glass surface. To enhance the friction coefficient at the fingertip, an elastomer band (Ecoflex 0050; Smooth-On) was added to the fingertip and the grasping was successful (Fig. 7c). We also used the gripper to pick up objects with complicated geometries such as a plastic nylon spring clamp (Fig. 7d), a paper clip (Fig. 7e), a plastic Petri dish (Fig. 7f), and a paper box (Fig. 7g). In all cases, the pickup processes were finished within 10 s (Supplementary Movie S1).

FIG. 7.

Grasping, releasing, and twisting manipulations by two soft grippers, one with “shorter” distal phalanges and the other with “longer” distal phalanges. (a) Grasping of a 2.1 g ping-pong ball by the “shorter” soft gripper. (b) Grasping of a 10.2 g tape roll by the “shorter” soft gripper. (c) Grasping of a 22.0 g glass ball by the “longer” soft gripper. (d) Grasping of a 7.3 g plastic nylon spring clamp by the “longer” soft gripper. (e) Grasping of a 26.3 g paper clip by the “longer” soft gripper. (f) Grasping of an 8.5 g plastic Petri dish by the “longer” soft gripper using the inverse bending mechanism. (g) Grasping of a 13.2 g paper box by the “longer” soft gripper. (h) Twisting of a 2.1 g ping-pong ball immediately after picking it up by the “longer” soft gripper. More details can be found in the video indexed Supplementary Movie S1 in supporting materials. Color images available online at www.liebertpub.com/soro

Object twisting manipulation has seldom been shown in recent soft gripper designs, with previous demonstrations relying on multiple pneumatic chambers within each finger.¹³ In our study, twisting manipulation begins by first grasping the object with a finger closing motion. Then, while the object is being held, another set of ligaments of the fingers as shown in Figure 1g is activated to generate the twisting motion. Here the gripper composed of soft fingers with longer distal phalanges is used to twist a ping-pong ball immediately after picking it up (Fig. 7h, Supplementary Movie S1). Figure 8 shows a series of images of the twisted ping-pong ball, together with the twisting angle versus time plot. After activating the second ligament of each finger to rotate the ball, the effective force of the fingertip holding the ball is observed decreased, as predicted by FEA simulation on the deflection of individual fingers. However, as the effective force of the fingertip in the gripping configuration is still higher than the required force to hold the ball, the twisting manipulation can be successfully completed. As we see from Figure 8, the twisting process was almost done in 15 s. And a twisting angle ∼18° was achieved using the current version of the fingers. More twisting could be achieved by optimizing the geometry of the fingers as well as by improving the interfacial bonding between the joints and phalanges as discussed earlier.

FIG. 8.

(a) Series of snapshots of twisting of a 2.1 g ping-pong ball. (b) Twisting angles versus activation time plot indicate that the twisting manipulation can be finished in roughly 15 s. Color images available online at www.liebertpub.com/soro

It is worth mentioning that a transient heat transfer simulation taking temperature-dependent resistivity of cPBE into account using ANSYS Fluent can be performed to estimate the time needed for ligament activation, taking the thermal properties of cPBE estimated by an inverse scheme as input.³⁶ In the experiments, we have used a constant direct current voltage of 110 V to activate the ligaments within seconds, given that both 100 and 150 V have been used earlier in activating cPBE strips with similar dimensions.³⁴ If a lower voltage had been used, the activation time would be longer. If a higher constant voltage had been used, the activation time would be less; however, without good control of heating time, the cPBE ligaments might be burned if the temperature goes too high. In addition, the gripper has been shown to exhibit consistent mechanical behavior after more than 30 activation cycles at a constant voltage of 110 V over a span of more than half a year, which shows the robustness of the gripper.

Conclusions

We presented a soft gripper using a combination of experiments and simulations in the design process. The soft gripper is composed of three soft fingers, which use rigidity tunable elastomer strips as ligaments that can be activated by resistive heating. A single central source of pressure is used to control the air pressure inside the finger chambers, which greatly reduces the complexity of supporting hardware when generating complicated manipulations. We have demonstrated that this soft gripper can grasp and release various objects in seconds. This kind of soft gripper can potentially handle objects in narrow channels and holes the same way as human fingers. More importantly, in this design, not only can all the fingers be individually controlled but they can also be actively tuned for functionality after being fabricated thanks to the rigidity tunable ligaments. This enables the soft gripper to perform closing, twisting, and opening motions previously demonstrated by multichamber pneumatic grippers using only a single internal air pressure.

Footnotes

Acknowledgments

This work was financially supported by the University of Nevada, Reno. The authors thank Eduardo Torres from Chemical Engineering Department, and Patrick Stampfli from Mechanical Engineering, both from the University of Nevada, Reno, for their assistance in the experiments in this study.

Author Disclosure Statement

No competing financial interests exist.

References

Hunter

, Lafontaine

. A comparison of muscle with artificial actuators. In: Technical Digest IEEE Solid-State Sensor and Actuator Workshop, Hilton Head Island, SC, 1992, pp. 178–185.

Jung

DWG

, Blange

, de Graaf

, Treijtel

. Elastic properties of relaxed, activated, and rigor muscle fibers measured with microsecond resolution. Biophys J, 1988; 54:897–908.

Trotter

, Tipper

, Lyons-Levy

, Chino

, Heuer

, Liu

, et al. Towards a fibrous composite with dynamically controlled stiffness: Lessons from echinoderms. Biochem Soc Trans, 2000; 28:357–362.

Majidi

. Soft robotics: A perspective-current trends and prospects for the future. Soft Robot, 2014; 1:5–11.

Trivedi

, Rahn

, Kier

, Walker

. Soft robotics: biological inspiration, state of the art, and future research. Appl Bionics Biomech, 2008; 5:99–117.

Hammock

, Chortos

, Tee

BCK

, Tok

JBH

, Bao

. 25th anniversary article: the evolution of electronic skin (e-skin): A brief history, design considerations, and recent progress. Adv Mater, 2013; 25:5997–6038.

Fassler

, Majidi

. Liquid-phase metal inclusions for a conductive polymer composite. Adv Mater, 2015; 27:1928–1932.

Ilievski

, Mazzeo

, Shepherd

, Chen

, Whitesides

. Soft robotics for chemists. Angewandte Chemie Intern Ed, 2011; 50:1890–1895.

Kim

, Laschi

, Trimmer

. Soft robotics: A bioinspired evolution in robotics. Trends Biotechnol, 2013; 31:287–294.

10.

Seok

, Onal

, Wood

, Rus

, Kim

. Peristaltic locomotion with antagonistic actuators in soft robotics. In: IEEE International Conference on Robotics and Automation (ICRA). Anchorage, AK, 2010, pp. 1228–1233.

11.

Shepherd

, Ilievski

, Choi

, Morin

, Stokes

, Mazzeo

, et al. Multigait soft robot. Proc Nat Acad Sci, 2011; 108:20400–20403.

12.

Robinson

, Davies

JBC

. Continuum robots-a state of the art. Proc IEEE Int Conf Robot Autom, 1991; 4:2849–2854.

13.

Suzumori

, Iikura

, Tanaka

. Development of flexible microactuator and its application to robotic mechanism. Proceedings of 1991 IEEE International Conference on Robotics and Automation, Sacramento, CA.

14.

Kwok

, Morin

, Mosadegh

, So

, Shepherd

, Martinez

, et al. Magnetic assembly of soft robots with hard components. Adv Funct Mater, 2014; 24:2180–2187.

15.

Brown

, Rodenberg

, Amend

, Mozeika

, Steltz

, Zakin

, et al. Universal robotic gripper based on the jamming of granular material. Proc Nat Acad Sci, 2010; 107:18809–18814.

16.

Kofod

, Wirges

, Paajanen

, Bauer

. Energy minimization for self-organized structure formation and actuation. Appl Phys Lett, 2007; 90:081916.

17.

Araromi

, Gavrilovich

, Shintake

, Rosset

, Richard

, Gass

, et al. Rollable multisegment dielectric elastomer minimum energy structures for a deployable microsatellite gripper. IEEE/ASME Trans Mechatron, 2015; 20:438–446.

18.

Shan

, Bertoldi

, Clarke

. Dielectric elastomer based grippers for soft robotics. Adv Mat, 2015; 27:6814–6819.

19.

Suo

. Theory of dielectric elastomers. Acta Mechanica Solida Sinica, 2010; 23:549–578.

20.

Shintake

, Rosset

, Schubert

, Floreano

, Shea

. Versatile soft grippers with intrinsic electroadhesion based on multifunctional polymer actuators. Adv Mater, 2016; 28:231–238.

21.

Yoshimi

, Iwata

, Mizukawa

, Ando

. Picking up operation of thin objects by robot arm with two fingered parallel soft gripper. In: IEEE Workshop on Advanced Robotics and its Social Impacts (ARSO), Munich, 2012, pp. 7–12.

22.

Giannaccini

, Georgilas

, Horsfield

, Peiris

, Lenz

, Pipe

, et al. A variable compliance, soft gripper. Autonomous Robots, 2014; 36:93–107.

23.

Cheng

, Lobovsky

, Keating

, Setapen

, Gero

, Hosoi

, et al. Design and analysis of a robust, low-cost, highly articulated manipulator enabled by jamming of granular media. In: IEEE International Conference on Robotics and Automation (ICRA), Saint Paul, MN, 2012, pp. 4328–4333.

24.

Kim

, Wu

, Carlson

, Jin

, Kovalsky

, Glass

, et al. Microstructured elastomeric surfaces with reversible adhesion and examples of their use in deterministic assembly by transfer printing. Proc Nat Acad Sci, 2010; 107:17095–17100.

25.

Song

, Sitti

. Soft grippers using micro-fibrillar adhesives for transfer printing. Adv Mater, 2014; 26:4901–4906.

26.

Song

, Majidi

, Sitti

. Geckogripper: A soft, inflatable robotic gripper using gecko-inspired elastomer micro-fiber adhesives. In: Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2014, pp. 4624–4629.

27.

Anderson

, Gisby

, McKay

, Obrien

, Calius

. Multi-functional dielectric elastomer artificial muscles for soft and smart machines. J Appl Phys, 2012; 112:041101.

28.

Shan

, Lu

, Wang

, Majidi

. Thermal analysis and design of a multi-layered rigidity tunable composite. Int J Heat Mass Trans, 2013; 66:271–278.

29.

Shan

, Lu

, Majidi

. Soft-matter composites with electrically tunable elastic rigidity. Smart Mater Struct, 2013; 22:085005.

30.

Follador

, Cianchetti

, Arienti

, Laschi

. A general method for the design and fabrication of shape memory alloy active spring actuators. Smart Mater Struct, 2012; 21:115029.

31.

Varga

, Filipcsei

, Zrinyi

. Magnetic field sensitive functional elastomers with tuneable elastic modulus. Polymer, 2006; 47:227–233.

32.

Shiva

, Stilli

, Noh

, Faragasso

, De Falco

, Gerboni

, et al. Ligament-based stiffening for a pneumatically actuated soft manipulator. 2016; 1:632–637.

33.

Maeda

, Hara

, Sakai

, Yoshida

, Hashimoto

. Self-walking gel. Adv Mater, 2007; 19:3480–3484.

34.

Shan

, Diller

, Tutcuoglu

, Majidi

. Rigidity-tuning conductive elastomer. Smart Mater Struct, 2015; 24:065001.

35.

Finney

. Finite element analysis. In: Engineering with Rubber, 3rd Edition. Gent

(Ed). Munich, Germany: Hanser, 2012, pp. 295–343.

36.

Tutcuoglu

, Majidi

, Shan

. Nonlinear thermal parameter estimation for embedded internal Joule heaters. Int J Heat Mass Trans, 2016; 97:412–421.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

16.75 MB

0.00 MB