WO2020222705A1

WO2020222705A1 - A variable stiffness device for soft robotics actuation

Info

Publication number: WO2020222705A1
Application number: PCT/SG2020/050258
Authority: WO
Inventors: Chen Hua YEOW; Tiana MILLER-JACKSON; Rainier NATIVIDAD; Yi Sun
Original assignee: National University of Singapore
Current assignee: National University of Singapore
Priority date: 2019-05-02
Filing date: 2020-04-29
Publication date: 2020-11-05
Anticipated expiration: 2021-11-02

Abstract

A variable stiffness device for soft robotics actuation comprises a plurality of pressurizable members and a holder for retaining the plurality of pressurizable members, the plurality of pressurizable members being arranged so that at least one pressurizable member is in contact with at least one other pressurizable member, wherein pressurizing the plurality of pressurizable members increases the stiffness of the device.

Description

A VARIABLE STIFFNESS DEVICE FOR SOFT ROBOTICS ACTUATION

Field

The present disclosure relates, in general terms, to a variable stiffness device for soft robotics actuation.

Background

The use of soft robotic actuators is on the rise because these systems typically offer the advantage of being highly flexible and compliant. This may provide safer robot- environment interactions, and the gentleness necessary to handle delicate objects. In addition, these systems typically closely mimic biological tissues, which can provide safer human-robot interactions. However, this advantage becomes a shortcoming in high-force applications, where flexible components typically fold and fail under large loads. Typically, soft robotic actuators offer other advantages over traditional rigid robotic actuators, such as impressive force-to-weight ratios and considerably lower fabrication costs.

One prime class of such soft robotic actuators is the soft pneumatic actuator (SPA), driven by compressed fluid to actuate upon pressurization. SPAs are typically fabricated through silicon rubber molding, fabric-based manufacturing, or additive manufacturing of flexible materials. Furthermore, SPAs are generally cheaper and simpler to fabricate, safer to operate, and exhibit more flexibility when compared to other types of soft actuator such as shape memory alloys (SMAs) and electroactive polymers (EAPs).

However, one advantage of the SPA, its extreme degree of compliance, can instead prove to be a limitation when handling relatively heavy loads. The exterior wall of the SPA is compliant to compressive loads, which leads to local wrinkling and eventually transverse buckling. Therefore, structural collapse occurs at significantly smaller loads for SPAs, as compared to traditional rigid components, precluding their use in high force actuation tasks.

A range of stiffening methods have been introduced to achieve variable stiffness in soft actuator components. One method is the use of antagonistic actuators. In this method two actuators are set in opposing arrangement such that their respective activations would induce opposing motions. Antagonistic actuators may take the form of bending actuators, contracting and elongating actuators, or other actuators which have opposing motions. An example of such a system is the granular jamming method, which involves filling a soft containing structure with relatively small particles (granules). Vacuum is applied to the containing structure, causing it to decrease in volume, which forces the particles into tight contact (jamming).

A drawback of such systems is the need for both positive and negative pressure delivery, for example by providing a system that includes both a compressor pump and a vacuum pump. This results in a bulky system with large energy consumption. Further, the granules typically add weight, and are hard elements in an otherwise soft system.

Alternative stiffening methods include embedding fibers or mesh in the walls of silicon rubber actuators. However, fabrication of these actuators is a sensitive process, as fibers must be aligned correctly to avoid uneven stress distributions. Other systems employ stiffer fabrication materials resulting in an increase of force output but at the cost of actuator flexibility.

Summary

In accordance with the present invention, there is provided a variable stiffness device for soft robotics actuation including a plurality of pressurizable members and a holder for retaining the plurality of pressurizable members, the plurality of pressurizable members being disposed so that at least one pressurizable member is frictionally engaged with at least one other pressurizable member, wherein pressurizing the plurality of pressurizable members increases the stiffness of the device.

Typically, each of the plurality of pressurizable members are individually pressurizable.

In some embodiments, the holder includes a first portion and a second portion arranged at an angle to the first portion, the first and second portions being coupled at a junction, wherein at least one pressurizable member is an actuating member that is located at and/or that spans the junction, and wherein pressurizing the actuating member causes the first portion to pivot relative to the second portion.

The first portion may be a first arm and the second portion may be a second arm that is coupled to the first arm. In some embodiments, a straight member is coupled to both the first arm and the second arm, and pressurizing the actuating member causes the device to stiffen and the angle to decrease.

In some embodiments, at least one pressurizable member is tubular.

The present invention also provides a soft robotic system comprising :

a plurality of variable stiffness devices as disclosed herein;

a controller; and

at least one pressure source in fluid communication with the plurality of variable stiffness devices;

wherein the controller is configured to cause the at least one pressure source to provide positive pressure to one or more pressurizable members of the plurality of variable stiffness devices to stiffen and/or actuate one or more of the plurality of variable stiffness devices.

The system may further comprise at least one pressurizable wedge-shaped member coupled to one or more variable stiffness devices; wherein pressurizing the at least one pressurizable wedge-shaped member causes pivoting motion of the one or more variable stiffness devices to which it is coupled.

In some embodiments, the at least one pressurizable wedge-shaped member comprises a plurality of wedge-shaped members arranged in a concertina configuration.

In some embodiments, the plurality of wedge-shaped members are coupled to each other by a resilient retaining component.

Brief Description of the Drawings

Embodiments of the present disclosure are hereafter described, by way of non-limiting example only, with reference to the accompanying drawings in which :

Figure 1 shows an example variable stiffness device in an unpressurized state and a pressurized state;

Figure 2 shows another example of a variable stiffness device in unpressurized and pressurized states; Figure 3 shows an example of a pressurizable device usable in conjunction with variable stiffness devices as part of a soft robotics system;

Figure 4 shows the device of Figure 3 coupled to the device of Figure 1;

Figure 5 is a block diagram of a soft robotic system according to certain embodiments; Figure 6 shows the bending orientation of an example variable stiffness device;

Figure 7 is a graph showing the effects of bending orientation on stiffness for a device consistent with that in Figure 1;

Figures 8a-8c are graphs showing the range of stiffness for variable stiffness devices according to certain embodiments;

Figure 9 is a graph showing the effect of packing ratio (to be defined below) on stiffness; Figure 10 shows the pressure-stiffness relationship for an example variable stiffness device;

Figure 11 is a graph showing the effects of tubule size on stiffness;

Figure 12 shows sensor positions for a tubule force distribution test for a variable stiffness device;

Figure 13 is a graph of bending load across tubules as measured by the sensor configuration of Figure 12;

Figure 14 is a graph showing torque output of a variable stiffness device consistent with that of Figure 2, with varying number N of tubule hinges;

Figure 15 shows bending actuation performance for a variable stiffness device consistent with that of Figure 2; and

Figure 16 is a graph showing angular displacement for a variable stiffness device consistent with that of Figure 3.

Detailed Description

Referring initially to Figure 1, a first example of a variable stiffness device 100 according to certain embodiments includes a plurality of pressurizable members 112 which are retained within a holder 114. The pressurizable members 112 may each, for example, be tube-like structures (also referred to herein as tubules) which are flat when unpressurized, as shown in Figure 1(a), but substantially cylindrical when pressurized, as shown in Figure 1(b).

The plurality of pressurizable members 112 are arranged within the holder 114 so that at least one pressurizable member 112 is in contact with at least one other pressurizable member 112. Preferably, each pressurizable member 112 is in contact with at least one other pressurizable member 112, and/or an internal surface of the holder 114. When at least one of the plurality of pressurizable members 112 is pressurized, this increases the stiffness of the device 100, due at least in part to frictional engagement between the pressurizable member 112 and/or the holder 114. For example, as shown in Figure 1(b), all pressurizable members 112 may be pressurized, thereby maximizing the stiffness of the device 100.

Each of the plurality of pressurizable members 112 may be individually pressurizable, thus enabling greater control of variation in the stiffness of device 100, by selective pressurization or de-pressurization of the pressurizable members 112. To this end, each pressurizable member 112 may be connected to a pressure source (not shown) via tubing 116 that is in fluid communication with the interior of the pressurizable member 112. In some embodiments, a single pressure source may be connected to all of the pressurizable members 112, for example via a hub to which tubes 116 are connected. In some embodiments, tubes 116 may be connected to the pressure source via respective programmable valves that may be selectively opened and closed such that individual pressurizable members 112 may be selectively pressurized or de-pressurized. In another example, multiple pressure sources may be connected to pressurizable members 112. For example, each pressurizable member 112 may be connected to a separate pressure source.

The variable stiffness device 100 operates according to a principle that may be referred to as "tubular jamming", since in this embodiment, stiffening is provided (at least partly) by pressurization of tube-like structures 112. This is in contrast to previously known "granular jamming" soft robotic structures which, as discussed above, have various drawbacks including the need to add hard structures and to use both positive and negative pressure.

Tubular jamming creates a jamming effect similar to granular jamming but rather than vacuum pressure, positive pressure is used; instead of inducing vacuum to tighten the structure surrounding the inbound particles, the tubular jamming method expands elements 112 inside of a substantially fixed-volume structure 114.

This is realized through pneumatically sealed soft tubules 112, which are equivalent to the particles used in granular jamming, set inside a non-expandable sleeve housing 114. By supplying the tubules 112 with compressed air (or another fluid), the tubules 112 inflate to fill the sleeve 114 area. Due to the inextensibility of the sleeve housing 114, tubules 112 are forced into larger contact with one another and higher normal force as they inflate. Similar to granular jamming, a large friction force between tubules 112 is induced, which prevents tubules 112 from changing position relative to one another. This results in a stiffening effect for the overall structure 100.

In general, the dimension of the tubules and the number of tubules to pack into the outer sleeve may depend on the desired stiffness profile for the device 100. However, the minimum number of tubules required in order to achieve jamming is governed by the "minimum packing circle," that is, the minimum bounding circle when pressurized tubules 112 are tangential to one another. This quantity can be determined from the optimal circle packing densities. The cross-sectional area of the outer sleeve 114 in this embodiment must be smaller than the area of the minimum packing circle of the tubules in order for the tubules 112 to come into contact and be pressed against one another as they inflate to provide the jamming effect.

The variable stiffness device 100 is a soft structural member that may be used in conjunction with other soft components. As a beam, its primary role is to transmit loads or withstand external loads without deflecting. Accordingly, embodiments consistent with the device 100 shown in Figure 1 may be referred to as a Tubular Jammed Beam (TJB).

Embodiments of the variable stiffness device 100 may provide the same load-bearing or load-transmitting qualities as rigid components, yet provide flexibility that enables it to adapt to the environment. Thus, embodiments provide optimal load transmission while maintaining a level of flexibility, in the form of a soft actuator component with variable stiffness capabilities. The variable stiffness device 100 is designed such that, when unpressurized, it is compliant and flexible, but when pressurized, it increases in stiffness and is able to resist deflection due to external loads.

When tubules 112 are unpressurized, they are highly flexible and easily move past one another, making the overall beam pliable. When tubules 112 are pressurized, they increase in stiffness, thereby increasing the stiffness of the overall beam 100. Furthermore, adjacent tubules 112 press against one another, providing an opposing shear force which prevents tubules 112 from sliding past one another. This frictional shear force acts to oppose bending, further increasing stiffness of the overall beam 100. Thus, variable stiffness is achieved. In the device 100, tubular jamming has been found to increase bending stiffness by nearly three-fold at the maximum pressure and packing ratio tested (as will be discussed below), compared to a conventional SPA beam of the same dimensions. It has also been found that the device 100 requires less supply pressure to achieve the same performance as a conventional SPA, and that it performs better in maintaining the vertical position of a borne object. It has also been found that a triangular support configuration made from devices 100 can support a load of nearly 33 times its own weight.

Referring now to Figure 2, an alternative embodiment of a variable stiffness device 200 will be described.

As shown in Figure 2, the device 200 comprises a holder 214 having a first portion 214a and a second portion 214b that meet at a junction 215. The second portion 214b is disposed at an angle Q to the first portion 214a. The first and second portions 214a, 214b may be arm-like structures, for example. Arranged within holder 214, as shown in Figure 2(c), are a plurality of pressurizable members 212. At least one of the pressurizable members 212 is located at, and/or spans across, the junction 215. In the example depicted in Figure 2(c), each pressurizable member 212 spans across junction 215.

Similarly to the embodiment of Figure 1, each of the plurality of pressurizable members 212 may be individually pressurizable, thus enabling greater control of variation in the stiffness of device 200, by selective pressurization or de-pressurization of the pressurizable members 212. To this end, each pressurizable member 212 may be connected to a pressure source via fluid supply tubing (not shown) that is in fluid communication with the interior of the pressurizable member 212. In some embodiments, a single pressure source may be connected to all of the pressurizable members 212, for example via a hub to which the fluid supply tubing is connected. In some embodiments, the fluid supply tubing may be connected to the pressure source via respective programmable valves that may be selectively opened and closed such that individual pressurizable members 212 may be selectively pressurized or de-pressurized. In another example, multiple pressure sources may be connected to pressurizable members 212. For example, each pressurizable member 212 may be connected to a separate pressure source. Pressurizing the plurality of pressurizable members 212 causes the first portion 214a to pivot relative to the second portion 214b. In particular, pressurizing those members 212 that are at, or that span, the junction 215 causes relative pivoting motion of first and second portions 214a, 214b. For example, when the resting state of the first portion 214a is 90° as shown in Figure 2(a), the device 200 may be deformed in such a way that Q increases to more than 90°, thereby causing the first portion 214a to pivot relative to the second portion 214b.

The device 200 operates in accordance with a jamming principle that is similar to that by which device 100 operates, but with rotational motion being caused by the jamming and consequent stiffening of the device 200. Accordingly, the device 200 may be referred to herein as a tubular jammed hinge (TJH).

In the TJH structure 200, tubular jamming has been found to have a compound effect on torque output, as three jammed tubule hinges produce approximately four times the torque of a single tubule hinge. The TJH 200 may form part of a wearable elbow flexion device as will be described below.

The TJH structure 200 is a soft rotary actuator intended to generate torque and provide rotational motion. Similar to the TJB 100, the TJH 200 is designed to be highly pliable when unpressurized and to exhibit increased stiffness when pressurized. However, in the case of the TJH 200, stiffness and actuation are coupled.

Each of the pressurizable members 212 may themselves be of hinged configuration, as shown in Figure 2(b). Accordingly, the TJH 200 may comprise several tubule hinges 212 and an encasing hinge 214 sleeve. When unpressurized, the arms 212a, 212b of a tubule hinge 212 can be moved apart to form angles greater than 90°. For example, at 180°, the tubule hinge arms 212a, 212b form a straight line. As the arms are moved apart, folds form in the tubule hinge 212 at the dashed lines shown in Figure 2(a). Upon pressurization, the internal air pressure exerts a stress on these folded areas to move the tubule hinge 212 back to its resting state with arms 90° apart, for example. A plurality of hinge tubules 212 are packed together into the encasing hinge sleeve 214 to form a TJH 200 as shown in Figure 2(c). When the TJH 200 is deformed then pressurized, it is able to exert a torque as it moves back to resting position of arms 90° apart. The jamming of the internal tubule hinges 212 helps to transmit this torque through the arms providing increased torque output. In general terms, tubular jamming may be implemented by providing an encasing structure 114 or 214 which is packed with smaller inflatable tubules 112 or 212. Within that general principle, tubular jamming structural design is customizable and can be modified according to intended use to fit each application.

Preferably, the encasing sleeve component (holder) 114 or 214 is made from a flexible but inextensible material. Pressurizable members 112 or 212 may be made from the same material or from an extensible material. However, an inextensible material for tubules may be advantageous because its greater resistance to deformation will further help to increase stiffness. For example, all components (sleeves and tubules) may be made from nylon fabric coated with thermoplastic polyurethane (TPU), (Jiaxing Inch ECO Materials Co., LTD, Zhejiang, China). This material may be advantageous because the nylon layer is inextensible, exhibiting robustness and the necessary restriction of tubule volume to support jamming. Additionally, the TPU layer is able to form a hermetic seal when heated, allowing for ease of construction of sleeve and tubules. However, this is only one example of many flexible and inextensible fabrication materials which are suitable for tubular jamming.

Referring now to Figure 3, an example of a pressurizable device that can be used in conjunction with one or more variable stiffness devices as part of a soft robotic system will be described.

As shown in Figure 3(a), a pressurizable device 300 may include at least one pressurizable member 312a that is wedge-shaped, wherein pressurizing the wedge- shaped member 312a causes pivoting movement, for example of another device such as variable stiffness device 100. For example, the device 300 may include a plurality of wedge-shaped members 312a, 312b, 312c arranged in a concertina configuration, and held together by a holder 314, which may be a resilient restraint, for example. Accordingly, when wedges 312a, 312b, 312c expand due to pressurization, holder 314 also expands (Figure 3(b)), whereas when the wedges are depressurized and contract, the holder 314 also contracts. Device 300 may also be referred to herein as an expansion actuation wedge (EAW).

Each wedge 312a-c can be inflated individually to achieve discrete rotary movements. Once a wedge 312a-c is fully inflated, additional pressurization stiffens the wedge body. Each wedge-shaped pressurizable member 312a-c is in fluid communication with a pressure source that supplies positive pressure to members 312a-c via respective fluid supply tubes 316a, 316b, 316c.

Although three wedge-shaped pressurizable members 312a, 312b, 312c are shown in Figure 3, it will be appreciated that fewer or greater numbers of wedge-shaped members may be included in the device 300.

The expansion actuation wedge (EAW) 300 is a soft rotary actuator which may work in conjunction with soft actuator beams to form a bending actuator. This is illustrated in Figure 4, in which a TJB 100 is attached to an EAW 300, thereby forming a bending actuator capable of transmitting torque from the base 302 of EAW 300 to a distal end of TJB 100. In Figure 4, a surface 301 of EAW 300 is attached to an immovable support, and in Figure 4(a), the wedges 312a-c of EAW 300 are unpressurized. In Figure 4(b), wedges 312a and 312b have been inflated, thus causing rotational motion of TJB 100.

Referring now to Figure 5, a block diagram of an example soft robotic system 500 that employs multiple variable stiffness devices is shown. The soft robotic system comprises a controller 502 in communication with a pressure source 504, such as a compressor, via signal lines 506. Controller 502 sends commands to the pressure source 504 via the signal lines 506, such as commands to open or close valves that regulate flow of fluid (typically air) to a plurality of variable stiffness devices that are in fluid communication with the pressure source 504. For example, the system 500 may comprise two TJBs 100a, 100b that are connected by an EAW 300 to form a robotic arm.

TJBs 100a, 100b are supplied by fluid supply tubing 116, and EAW 300 is supplied by fluid supply tubing 316. The controller 502 may be programmed to selectively pressurize one or more pressurizable members of the TJBs 100a, 100b and EAW 300. For example, controller 502 may inflate tubules 112 of the TJBs 100a, 100b to stiffen them so that they can function as load-bearing/load-transfer beams, and then inflate one or more wedges 312 of EAW 300 to cause TJB 100b to pivot relative to TJB 100a.

It will be appreciated that many other system configurations are possible, with greater or fewer variable stiffness devices, and/or including devices that are not variable stiffness devices but which are coupled to same. Experimental results

Pressurization and geometrical configuration for the TJB and TJH were varied and the resulting tubular jamming properties were assessed. Being a structural member, the TJB does not exhibit movement, but is rather intended to be used in conjunction with other components to form a robotic assembly. Tubular jamming enhances the basic beam component with variable stiffness capabilities. The TJB was assessed under variation of orientation, packing ratio, pressurization, and tubule size, each of which are explained in more detail below. For further understanding of the inner workings of the TJB, its internal force distribution was also assessed. Finally, the TJB was used to demonstrate the efficacy of a soft beam with varying stiffness capabilities through several weight-bearing demonstrations. Tubular jammed beam (TJB1

To assess the efficacy of the TJB design in comparison to the conventional soft pneumatic actuator (SPA) beam design, a standard SPA comprising only a soft exterior wall was fabricated.

Since there may be any size of tubule and sleeve, it is desirable to develop a standardization in order to compare TJBs. To normalize TJB stiffness, we introduce the packing ratio (F) per Equation 1, where N is the number of tubules, Asieeve and Atubuie are the cross sectional areas of the sleeve and of a single tubule, respectively, and r_sieeve and rtubuie are the radii of the sleeve and of a single tubule, respectively.

F = N -Atubule / Asieeve ⁼ N TT(rtubule)^ / TT(rsleeve)^ ⁼ N (rtubule)^ / (rsieeve)^ ( 1 )

The TJB components were formed through folding and heat-sealing. Flat sections of fabric (TPU) were cut according to the dimensions shown in the table below:

To make a TJB device 100, masking tape was placed over one half of the TPU surface to prevent heat-sealing over that area. The fabric was rolled into four quarter folds, with the taped area being the two inner folds and the un-taped area being the two outer folds. The fabric was secured with clamps and heat was applied over the entire outer surface in order to form a seal. This produced a cylinder. By rolling the fabric into four quarter folds, the cylinder walls are comprised of two layers of nylon over their entire circumference, which enables more uniform stiffness characteristics. The sleeve 114 is completed at this step. The tubules 112, however, still need to be made hermetic. In order to accomplish this, a 1 mm heat seal was applied to both ends and an air inlet hole was punched in the fabric cylinder of each tubule 112.

Of course, the shape of the tubule 112 and sleeve 114 may be something other than cylindrical, such as ellipsoidal, rectangular, triangular, and so on.

A silicon tube 116 of larger diameter than the punched hole was forced into the punched hole to create a sealed inlet path for pressurized air, and the distal end of the tube was connected to a compressed air source for inflation. The tubules 112 are complete at this step. Once the tubules 112 were fabricated, they were aligned and packed into the sleeve 114 to form a completed TJB 100. The aforementioned standard SPA beam was fabricated through an identical process to that used for the tubules. The standard SPA was made to be of the same length and diameter as the TJB for comparison purposes.

Stiffness was quantified as the measure of beam deflection per applied load. The majority of loads incurred by a robotic member during actuation are transverse in nature, for example the weight of an object being lifted by a robotic gripping arm. Therefore, bending stiffness was the stiffness mode used to observe the effects of varying TJB parameters. Three-point flexural bend testing was performed using a universal test machine (Model 3345, Instron, MA, USA). The test configuration used a span length of 140 mm, and supports and loading arm were fitted with adapters of 5 mm diameter.

To allow for tubules 112 to be packed into the sleeve 114, tubules were flattened into the shape of highly eccentric ellipses and stacked such that their long axes are aligned when inserted into the sleeve. This causes some eccentricity in the overall TJB cylinder making it slightly elliptical. To quantify anisotropic stiffness properties of the TJB, bending stiffness was measured for two orthogonal orientations, with load applied through the long axis (BL) and load applied through the short axis (Bs) of the elliptical TJB cross section, as shown in Figure 6.

Bending stiffness through the short axis (Bs) was notably higher than through the long axis (BL) of the TJB, as shown in Figure 7. This trend was consistent even when the number of tubules installed was varied (N=6 to N = 14). On average, the relationship between Bs and BL can be expressed by Equation 2.

Bs = 1.4B_L (2)

Bending through the short axis provides higher stiffness since more material lies along the plane of deformation resisting movement, meaning more material must be deformed in order to achieve the same deflection.

Addition of tubules increases beam stiffness, even in the unpressurized state, due to the increased amount of material inside. To account for this effect, stiffness of unpressurized TJBs was measured using the same flexural bending test procedure as discussed above. The magnitude of the range of stiffness between pressurized and unpressurized states of a TJB configuration was used as the primary measure for quantification, and is denoted as | Kp-u | , as shown in Equation 3, where Kp and Ku are the pressurized and unpressurized stiffnesses, respectively.

I Kp-u I = I Kp - Ku I (3) The pressurized and unpressurized stiffnesses of the TJBs across varying number of tubules installed are shown in Figure 8a, 8b and 8c for both the short and long axis TJB orientations. In the short axis orientation, | Kp-u | plateaus around N = 10, after which it remains within 8.5±0.4 N/mm, whereas the long axis | Kp-u | plateaus earlier at N=8, remaining at 6.0±0.4 N/mm thereafter. In general, the short axis orientation exhibits a larger | Kp-u | than the long axis, showing the short axis configuration to yield greater versatility. As stated, the short axis orientation also provides higher stiffness. Due to its superior performance, the stiffness of the short axis orientation is reported as the overall TJB stiffness in the following discussion.

The apparent plateau of | Kp-u | in both configurations indicates that the subsequent inclusion of additional tubules will not influence | Kp-u | . To maximize versatility and simplicity in a design involving TJB members, the packing ratio wherein | Kp-u | begins to plateau is preferably selected; this will yield the largest range of stiffness while utilizing the least amount of tubules. Alternatively, if range of stiffness is less important to the design than overall stiffness, more tubules may be installed to achieve maximum possible stiffness.

To explore the stiffening capabilities of the TJB design 100 compared to the standard SPA design, the bending stiffness of a standard SPA beam was compared to that of the TJB 100 as the amount of tubules packed inside the TJB was incrementally increased. Initially, the TJB was packed with N=6 tubules, the minimum amount necessary for tubules to jam for the chosen dimensions of the TJB and tubules. Subsequently the amount of tubules was incremented by one, up to N = 14 tubules. The influence of packing ratio on beam bending stiffness was measured under a constant pressurization of 100±5kPa for all tubules and the standard SPA beam.

To demonstrate the superior stiffness exhibited by the TJB 100 as compared to the standard SPA, the stiffness values of the TJB are normalized by those of the standard SPA, as shown in Figure 9. The TJB 100 displays higher stiffness than the standard SPA even from the minimal packing ratio of F=0.67 (N=6). Furthermore, the increased number of installed tubules increases TJB stiffness considerably beyond that of the standard SPA. At the maximum packing ratio tested (0=1.56, N = 14), TJB stiffness reached a nearly threefold improvement to the standard SPA stiffness (13.3N/mm compared to 4.7N/mm). The slope of the curve formed by the TJB stiffness appears to be asymptotic. This asymptotic behavior suggests that each addition of a tubule has less impact than the previous tubule added and the stiffness value will approach some upper limit. This upper limit is theorized to be the stiffness of a beam composed entirely of TJB fabric (where tubules are packed so tightly they have essentially no room to inflate). As packing ratio is increased, tubules have less volume available in which to inflate, therefore the effects of pressurization become less significant, and the intrinsic stiffness of the tubule fabric becomes the more significant factor.

Pressurization is an integral factor in increasing the stiffness of any soft inflatable structure. In order to determine the relationship between pressure and stiffness, a TJB of F= 1.11 was incrementally pressurized from 0 to lOOkPa while subjected to bending load. For comparison, the standard SPA beam was similarly pressurized and loaded. The resultant stiffness was measured.

As expected, the stiffness of the TJB increases with increased pressure (Figure 10). TJB stiffness increases approximately 0.09N/mm per unit of pressure increase (kPa). Stiffness of the standard SPA also increases with pressure, but the TJB stiffness increases at a much higher rate. As shown in Figure 10, on average, the TJB is 4.3 times stiffer than the traditional SPA beam across the range of pressures tested.

The TJB stiffness is able to be tuned by adjusting pressurization and can provide a much larger gamut of stiffness than the standard SPA beam. The TJB stiffness trend has a larger variance compared to the standard SPA, likely caused by the shifting of tubule position during inflation and deflation cycles and handling.

To observe the effects of tubule size variation, two sizes of tubules 112 were compared : small tubules of circumference 40 mm, and large tubules of circumference 50mm. To equitably compare the results obtained with two different tubule sizes, a similar packing ratio was selected. A set of 7 large size tubules (F = 1.22) and a set of 10 small size (F = 1.11) were used. The pressure was varied from 0 to 100 kPa while the TJB, containing large or small tubules, was subjected to a flexural bending test.

As displayed in Figure 11, the slope of the stiffness curve showed a similar trend for both sets of tubules, but the large size tubules produced less stiffness. On average, the large size tubules yielded 43% lower stiffness in bending than small size tubules. On the other hand, the large size tubules produced a stiffness trend with smaller standard deviation than the small tubules.

The larger standard deviation of TJB stiffness shown in the small size tubule configuration is due to the variation of possible tubule positions within the sleeve. Since the TJB configured with small size tubules contains a greater number of tubules and therefore more possible tubule positions, it follows that there will be more variation in TJB stiffness than when employing the large tubule configuration. The smaller size tubules provide more surface area and therefore larger frictional force to resist buckling, which leads to smaller size tubules producing a higher stiffness than large size tubules under similar packing ratio.

Since the TJB is made from several individual parts and not a single solid material as a traditional rigid beam, it is useful to examine internal force distribution dynamics among those parts to determine if they differ from the behavior of a solid material beam.

The forces experienced by individual tubules in different positions within the TJB during bending were determined using a strain-force relation. An expression of tubule force as a function of tubule strain was defined as shown in Equation 4, where F and e are force and strain experienced by the tubule, respectively, and a and b are experimentally determined coefficients. A strain sensor was installed on a single tubule and the tubule was subjected to three-point bending. The resulting force versus strain data was used to calculate coefficients a and b. The tubule with strain sensor was then installed into the TJB among the other tubules, first at the top layer, then the middle layer, and finally at the bottom layer, as shown in Figure 12. For each case, the TJB was subjected to three-point bending, and the resultant strain was converted to force to determine the force experienced by the tubule in each of the three layer positions.

F = ae + b (4)

The bending load is not evenly distributed across the TJB cross-section as would be expected in a solid beam. Instead, the tubules on the bottom layer bear a greater percentage of the load as shown in Figure 13. When minimal load is applied, all tubules incur approximately equal load. As load and deflection on the beam increase, tubules near the top of the TJB experience compressive load while lower tubules sustain tensile load. Around 7.5mm deflection, topmost tubules experience around 3. ON of compressive force, while bottommost tubules experience around 6.9N of tensile force, and tubules near the center experience almost no force since they are at the neutral axis. At around 15mm deflection, the discrepancy between upper and lower tubule force is even more pronounced. Upper tubules experience up to around 4.6N of compressive force, while lower tubules experience up to around 15.3N of tensile force, the bottom tubules incurring more than three times the force on the top tubules. This force gradient shifts the neutral axis such that central tubules experience compressive force at around 2N.

Knowledge of the manner in which an external bending force is distributed amongst internal tubules can be used to optimize TJB design. Since tubules at the bottom support a larger load than tubules at the top, using a stiffer material only in select bottom tubules could significantly impact the stiffness of the overall TJB without the need to increase the stiffness of all tubules, and thus, better preserving the flexibility of the unpressurized TJB.

One performance indicator of a soft actuator beam is its ability to successfully support a load without buckling. To observe the performance of the TJB in this ability, both the TJB 100 and a standard SPA beam were loaded with 4kg while beam pressurization was incrementally decreased from lOOkPa to beam failure.

At the mid-point of 50kPa, the TJB 100, although identical in size, was able to support the load while the standard SPA was incapable of supporting the load at the same pressure. The standard SPA failed at a pressure of 85kPa, while the TJB was able to sustain the load until a pressure of 8kPa. This demonstrates a strength of the TJB in that it can perform the same functions as a traditional SPA but at a significantly lower pressure. Replacing standard SPA beams with TJBs 100 is therefore able to greatly increase system efficiency.

Another valuable performance indicator of a robotic system is positional consistency. To evaluate the vertical positional consistency of the TJB, both the TJB and an SPA were incorporated into a soft robotic gripper arm. This configuration represents a typical example of the role of a soft beam in a soft robotic assembly. The distal end of the soft beam (either TJB or standard SPA beam) was outfitted with a soft gripper end effector which is capable of bending and straightening to grasp and release an object, respectively. The proximal end of the soft beam was secured by a rotating arm and the soft robotic gripper arm was made to transfer an object from one platform to another platform of identical height. The objective of the task was to pick up a 429g item from the first platform, and to maintain the vertical position of the item while being rotated by the rigid rotating arm, then successfully transport the object onto the second platform. In each case, the beam was pressurized to 40kPa.

The TJB 100 was able to provide the necessary stiffness to minimize beam deflection when the item was airborne and was thus able to complete the desired task and deliver the item onto the second platform. Conversely, the SPA was not able to exhibit sufficient stiffness, leading to excessive beam deflection, which resulted in the item colliding with the second platform rather than being set on top. Although taking up the same outward dimensions and being made from the same material, the TJB design is able to exhibit a much higher stiffness than the SPA at equivalent supply pressure. The stiffening capabilities of the TJB make it a critical member of soft robotic systems where stiffness and flexibility are both important.

The stiffness levels attained here are sufficiently high to enable use of soft beams in a structural application. To illustrate this, three TJBs were configured into a triangular support apparatus and loaded with 1.6 kg weights. The total mass of the apparatus was 0.49 kg, yet it was able to support a quantity of ten 1.6 kg weights (total of 16 kg), a load-bearing-capacity-to-weight ratio of 32.9. Assuming this figure can be extrapolated to estimate the load-bearing-capacities of larger soft beam assemblies, then tubular jamming will open new possibilities for soft beams which have not been explored before as lightweight structural supports.

Tubular Jammed Hinge (TJI-0

A TJH 200 was formed through a combination of heat-sealing and binding using hook and loop fasteners. To form tubule hinges 212, flat sheets of fabric were cut into two 90° hinge shaped pieces (as in Figure 2(a)) having mirrored dimensions. Masking tape was used to cover the TPU side of the hinge shaped pieces except for a perimeter area of 1mm width, which was left exposed. The masking tape was applied in this way to ensure that when heat was applied to the fabric, only the exposed perimeter area would become heat-sealed while the covered area would be kept from sealing in order that it could be inflated upon pressurization. An air inlet hole was punched into one of the sheets. The two hinge-shaped sheets were aligned with the TPU coated faces in contact, and heat was applied to seal their exposed perimeters together. A silicon tube was forced into the previously punched hole, and the distal end of the tube was connected to a compressed air source. The tubule hinges 212 are complete at this step. The encasing sleeve 214 is comprised of four sections, two center segments and two arm sleeves. To form an arm sleeve, a rectangular sheet of fabric was cut to size such that it could wrap twice around an arm of the tubule hinges 212. To form a center sleeve, a section of fabric was cut such that it could wrap around the center section of the hinge tubules 212 once. Hook and loop fasteners were adhered to the fabric sections for both arm sleeves and center sleeves. To form the TJH 200, tubule hinges 212 were bundled, and one arm sleeve was wrapped and secured around each arm of the tubule hinge bundle. Two center sleeves were then secured around the center section of the tubule hinge bundle. This produced a completed TJH 200.

The TJH 200 is designed to generate torque upon pressurization, and the output torque is thought to be affected by the number of tubule hinges packed into the TJH 200 and the separation angle of the TJH arms 214a, 214b. Therefore, these parameters were varied and resulting torque output was measured.

The TJH 200 was pressurized to 100±5 kPa and the resulting torque generated was measured with a torque meter. The fixed arm of the torque meter is rigidly attached to the base of a torque sensor (Forsentek, FT01), while the adjustable arm of the torque meter is rigidly attached to the sensing head of the torque sensor, such that torque generated by the TJH 200 will be transmitted along the rigid arm and recorded by the torque sensor. Number of tubule hinges was varied from N = 1 to N=4. For each configuration, measurements were recorded for three torque meter arm orientations: 180°, 150°, and 120°.

TJHs of different tubule hinge configuration exhibit somewhat linear torque output with respect to arm angle, as shown in Figure 14. For each case, 180° arm angle produces the highest TJH torque, and torque decreases as arm angle decreases toward 90°. In general, the TJH produces more torque as more tubule hinges are inserted. However, a TJH containing N=3 tubule hinges was able to produce slightly more torque than a TJH of N=4 tubule hinges at an arm angle of 180° (1.45 and 1.34 N-m, respectively).

Due to its geometry, the tubule hinge resting state is a 90° angle between hinge arms. When hinge arms are forced into a larger angle, a resistive stress from the fabric is generated. Therefore it follows that a 180° arm angle produces the highest TJH torque since this position induces the most strain in the TJH fabric. Conversely, it also follows that, as arm angle increases toward 90°, TJH torque output decreases. The increase in torque output with increased number of tubule hinges may be due to the increase in beam stiffness of each hinge arm which enables a larger transmission of torque. The surprising result that a TJH containing three tubule hinges is able to outperform a TJH with four tubule hinges suggests there may be a packing limit above which addition of tubule hinges actually decreases torque output. This behavior is perhaps due to insufficient volume for tubule hinge inflation within the packed sleeve area when the sleeve is over packed. Evidence for this theory is also seen at the 120° arm position, where the TJH of N=4 tubule hinges begins to exhibit greater force than N=3; due to sleeve geometry, at this position, tubule hinges have more volume into which they can expand within the sleeve, so the four tubule hinges no longer prohibit one another's inflation and are able to exert more torque. Notably, the torque produced by a TJH with multiple tubules is larger than the sum of the torque of individual tubule hinges. For example, the N=3 configuration produces an average of 4.04 times the torque of the N = 1 configuration. This fourfold increase in torque for a threefold increase in tubule hinges gives evidence that the tubular jamming effect goes beyond mere coupling of individual parts and has a compounding effect on torque output.

According to the results presented, the TJH 200 is at least effective for exerting torque at angles of 180° to 120°. Advantageously, the TJH 200 is capable of exerting torque for any angle apart from 90°. To form an actuator assembly capable of autonomous motion in two directions, two differently-shaped members can be coupled. For example, as shown in Figure 15A, a straight member 1500 may be coupled to both the first arm and the second arm, wherein pressurizing the straight member 1500 causes the angle Q to increase and pressurizing the plurality of pressurizable members 212 causes the device 200 to stiffen and the angle Q to decrease. When the straight member 1500 is pressurized, the arms of the TJH 200 are moved apart and the assembly moves toward a position of 180°. When the TJH 200 is pressurized, the assembly moves in the opposite angular direction and approaches a position of 90°. Figure 15A shows the resultant motion when the TJH 200 and the straight member 1500 are alternatingly pressurized and depressurized.

To observe the efficacy of the TJH 200 in bending actuation of a hinge joint, the TJH was attached to the inner surface of the elbow joint of a human arm model and used to flex the joint, as shown in Figure 15B. A straight member 1500 was attached to the outer surface of the elbow and was used to extend the joint. The TJH 200 and straight member 1500 were alternately pressurized to 75kPa and HOkPa, respectively, and the resultant angular motion of the arm was measured using Tracker Video Analysis and Modeling Tool. An arm angle of 180° corresponds to a straight elbow (zero flexion). The pressure of the TJH and the subsequent angular position of the arm (Q) are shown in Figure 15C. When the TJH air supply is activated, there is a resultant decrease in the arm position angle. The duration of TJH pressurization is varied so that its effect can be observed. An increased duration of TJH pressurization may lead to an increased degree of angular motion. TJH 200 may provide sufficient output torque to effectively flex a hinge joint. For example, the TJH design may increase torque output with the intention of achieving flexion of the elbow joint of a real human subject as well as flexion of other biological joints such as the knee.

Referring now to Figure 16, there is shown a graph of angular displacement as a function of number of activated (pressurized) tubule wedges for an example EAW 300. As can be seen from the graph, the angular displacement is linear with the number of pressurized wedges 312a-c.

The tubular jamming method is an effective means of enabling a soft actuator component with variable stiffness, as demonstrated in this disclosure. Firstly, the TJB design has shown that, with the simple addition of tubules, any standard SPA beam can be upgraded to a sophisticated soft beam, capable of varying stiffness, and capable of achieving stiffness several times that of the original beam. Tubular jamming provides a promising new method for employing soft members in high force applications such as structural support. Soft beams could be used to support massive yet delicate objects such as optics or large, precious gems.

Inflatable beams have long been used in aerospace applications. Since tubular jamming can be implemented on inflatable beams, it could potentially be implemented on such beams in aerospace applications to enhance them with variable stiffening capabilities and potentially enable their use in higher force applications than have been performed thus far. The TJH design has further expanded the tubular jamming concept into the realm of bending actuation. It has been found that, by packing together multiple bending actuators, it is possible to achieve a soft robotic mechanism greater than the sum of its parts, with torque output being amplified by the jamming effect. Tubular- jamming-based actuators such as the TJH are flexible enough to be worn by a human, yet strong enough to provide bending actuation.

The tubular jamming method has been demonstrated to be versatile, widely applicable, and easily implemented. Tubular jamming is a positive-pressure jamming method- employing positive air (or other fluid) pressure, rather than vacuum— which makes it highly compatible with existing soft pneumatic systems utilizing positive air pressure. Since tubular jamming involves only soft materials, it enables the use of an entirely soft system. The necessary materials for jamming add only minimal weight to a system, which maintains the lightweight feature of a soft robotic solution. Furthermore, the flexibility of the system is maintained, since unpressurized tubules are compliant.

A few limitations may be seen in the tubular jamming design, in the TJB for example, the magnitude of the range of stiffness between pressurized and unpressurized states approaches an upper limit. This can be addressed by tailoring tubule design based on the range of stiffness required by the intended application. Another restraint is the diminishing efficacy of installing additional tubules, forming an upper limit to the jamming effect (be it stiffness, torque output, or otherwise). To address this challenge, the size of tubules relative to encasing sleeve may be adjusted to increase the upper limit of jamming efficacy.

Tubular jamming opens new possibilities for soft components to achieve the stiffness needed to perform high force tasks such as weight bearing and large-scale actuation, while retaining the suppleness to enable a safe robot-to-environment interface.

Throughout this specification, unless the context requires otherwise, the word "comprise", and variations such as "comprises" and "comprising", will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

The reference to any prior art in this specification is not, and should not be taken as, an acknowledgment or any form of suggestion that the prior art forms part of the common general knowledge.

Claims

1. A variable stiffness device for soft robotics actuation, comprising a plurality of pressurizable members and a holder for retaining the plurality of pressurizable members, the plurality of pressurizable members being arranged so that at least one pressurizable member is in contact with at least one other pressurizable member, wherein pressurizing the plurality of pressurizable members increases the stiffness of the device.

2. The device of claim 1, wherein each of the plurality of pressurizable members are individually pressurizable.

3. The device of claim 1 or claim 2, wherein the holder includes a first portion and a second portion arranged at an angle to the first portion, the first and second portions being coupled at a junction, wherein at least one pressurizable member is an actuating member that is located at and/or that spans the junction, and wherein pressurizing the actuating member causes the first portion to pivot relative to the second portion.

4. The device of claim 3, wherein the first portion is a first arm and the second portion is a second arm that is coupled to the first arm.

5. The device of claim 4, wherein a straight member is coupled to both the first arm and the second arm, and wherein pressurizing the actuating member causes the device to stiffen and the angle to decrease.

6. The device of any one of the preceding claims, wherein at least one pressurizable member is tubular.

7. A soft robotic system comprising :

a plurality of variable stiffness devices in accordance with any one of claims 1 to

6;

a controller; and

wherein the controller is configured to cause at least one pressure source to provide positive pressure to one or more pressurizable members of the plurality of variable stiffness devices to stiffen and/or actuate one or more of the plurality of variable stiffness devices.

8. A system according to claim 7, further comprising at least one pressurizable wedge-shaped member coupled to one or more variable stiffness devices; wherein pressurizing the at least one pressurizable wedge-shaped member causes pivoting motion of the one or more variable stiffness devices to which it is coupled.

9. A system according to claim 8, wherein the at least one pressurizable wedge- shaped member comprises a plurality of wedge-shaped members arranged in a concertina configuration.

10. A system according to claim 9, wherein the plurality of wedge-shaped members are coupled to each other by a resilient retaining component.