Soft Robotics

Aspects of Materials Processing and Engineering for the Development of Soft Robotic Devices

Shib Shankar Banerjee¹, ², Injamamul Arief¹, Andreas Fery¹, Gert Heinrich³, Amit Das¹, *

¹ Leibniz-Institut für Polymerforschung Dresden e. V., D-01069, Dresden, Germany

² Department of Materials Science and Engineering, Indian Institute of Technology Delhi, New Delhi-110016, India

³ Technische Universität Dresden, Institut für Textilmaschinen und Textile Hochleistung-swerkstofftechnik, Hohe Straße 6, 01069 Dresden, Germany

Abstract

The development of soft materials for various smart applications, such as in soft robotic devices derived from elastomers, naturally occurring rubbers and polymers, can be traced back to 1940s. Contrary to conventional robotics, soft robotic research was primarily dedicated to reproducing and mimicking the elasticity and compliance of biological substrates. In the last few decades, however, subsequent progress in soft robotic devices and manipulators has been witnessed within the subcategories of material & processing, sensor and actuators. In this chapter, a comprehensive introduction to soft robotics engineering is highlighted exclusively from the perspective of materials development. Soft material prototypes are eventually exploited towards designing biomimetic prosthetic devices with sensing capabilities and soft actuating bodies. We present an elaborate discussion on soft materials engineering, including elastomers, polymer composites containing fillers and associated processing methodologies. Despite high flexibility and inherent biocompatibility, conventional soft elastomer-based (silicone and polyurethane) matrices often suffer from either poor mechanical strength or reduced mechanical output at higher temperatures. This review briefly summarizes and critically discusses the different types of compliant materials, i.e. elastomers, fluids, gels and others, which are currently being employed in soft robotics. This chapter also highlights new opportunities and paves some potential avenues to address several challenges, particularly the ones concerning and soft robotic materials. Moreover, utilization of soft rubber for sensors and other applications could ameliorate the versatility of soft material applications. Therefore, challenges and future direction in soft robotics research have to be explicitly integrated within the subdomain of materials development and processing and, therefore, addressed in-depth in this chapter, with an emphasis on autonomous soft robotic arms capable of stimuli-responsive operations.

Keywords: Crosslinking of elastomers, Elastomer composites, Elastomer processing, Mechanical properties, Network stability, Soft elastomeric materials.

* Corresponding author Amit Das: Leibniz-Institut für Polymerforschung Dresden e. V., D-01069, Dresden, Germany; Fax: +49 351 4658-362; Tel: +49 351 4658-579; E-mail: [email protected]

1. INTRODUCTION

A robot is an automatically controlled, programmable machine having desired abilities to execute movement, perception, and cognition [1-4]. The field of rob-otics has emerged in an effort to create fully adaptable autonomous machines capable of independent movement and interaction with surrounding environments in response to changing conditions. Its ultimate goal is to enhance robotic capabilities in terms of precision movement, accurate motion control, and finally, the level of intelligence. Mainstream robots are mostly designed to be rigid, hard, and inflexible. In such robotic systems, commonly used robotics segments, such as, actuators and sensors are made of rigid electromagnetic components, steel, alloys or semirigid plastics [5-8]. These can bear and support high loading and sustain reliable electrical and mechanical properties under extreme conditions (see chapter 1, Zimmermann et al.). However, in order to create universal and customizable machines capable of performing a wide range of tasks inspired by nature, robots built from soft materials need to be engineered. Furthermore, there is no single machine versatile enough to adequately perform all possible tasks.

Soft robotics is an advanced and sophisticated area of robotics dealing with building robots from easily deformable and highly compliant materials such as fluids, gels, and soft elastomers [1, 9-11]. Soft robots are capable of performing various actions such as squeezing, stretching, growing, and climbing that is not possible for conventional robots [12, 13]. Recently, exploratory research on compliant materials and their functioning mechanism has gained tremendous attention for the soft-matter technologist [13-18]. Development of functional compliant materials for robotic applications is one of the potential areas. Research on compliant materials has been conducted for several applications, e.g. to replace human skin, soft manipulation, development of soft sensors, actuators, etc [13, 19-21]. In the recent past, significant work has been carried out on anthropomorphic and multi-fingered robot hands to develop functional capabilities similar to human hands [22-29]. Most of them used metals and plastics for such application. The important parameters toward selecting a suitable material for the robotic hand are impact force attenuation, conformability and repetitive strain dissipation [30]. Unlike a rigid finger made of metal and semirigid plastics, the use of a compliant material for such applications will be advantageous for several reasons. During object manipulation, a finger made of soft matter will not leave marks on the objects, and its grasping stability will be enhanced. Finally, repetitive strain can be eliminated. However, development of electronic/mechanical skin-based materials mimicking human skin is very challenging.

Considerable progress has been made in recent years in the field of soft robotics, with efforts focused on novel methods of actuation, sensing, design and fabrication [17, 31-41]. This new trend is moving towards soft robotic material systems driven by several objectives like refine potentialities, proprioception capacity, ability to assist very sophisticated tasks like surgery, and finally increased safety for human interaction. The transition from a hard-conventional robot to its soft counterpart mainly depends on its underlying materials. This transition goes along with several challenges on the material side. Materials for soft robotics have, in addition to their functionality, to fulfill several minimum requirements related to high stretchability, flexibility, durability, low power consumption, adaptability for accomplishing tasks and lightweight. Despite the recent progress, in order to advance further in the area of soft robotics and its related applications, new multifunctional compliant materials with a broader range of advanced functionalities must be explored, which are still challenging today. On the other hand, a much deeper understanding of the properties and dynamics of soft materials and deeper insight into their interaction behavior with control systems and the environment have to be explored. These are crucial to achieving the desired robotic behavior in real contexts. There have been some excellent review reports on soft materials for different smart applications [9, 11, 20]. However, these pioneering research works do not specifically discuss in depth the ‘soft and compliant materials’ for robotic system applications. Therefore, the aim of this review is to serve (as an essential background or state-of-the-art) as a guide to materials for the researchers who are working on soft robotics and the corresponding sub-research areas. This review briefly summarizes and critically discusses the different types of compliant materials, such as, elastomers, fluids, gels and others, which are currently being used in soft robotics. This review will also highlight new opportunities and pave some potential avenues to address the robotic material’s challenges. As the dynamic response of the soft materials are often very nonlinear in nature with many degrees of freedom, specific material modeling and development of the constitutive relations for the soft materials as well as their characterization are very crucial to the design of soft robotic components. Therefore, this review also includes a comprehensive survey on the various existing analytical modeling approaches (e.g., pseudo-rigid body model, reduced order model, etc.) used for small-scale deformations and relatively simple geometries as well as numerical modeling based on the finite element analysis (FEA) used for large-scale deformations and the design of relatively complex robotic components. Finally, current challenges and prospects of soft materials and their dynamics for robotic application are critically discussed.

1.1. Classification of Compliant Materials for Soft Robotics

The softness of a matter or device is generally measured according to its compliance value [42]. Therefore, compliance matching, i.e., similar mechanical rigidity, is one of the most important parameters in soft robotics technology [43-46]. The mechanical rigidity of a material can be evaluated by its modulus of elasticity or Young’s modulus (E) which is a measure of resistance to elastic deformation of material upon tensile loading. E is generally estimated in the small deformation (< 0.2% elongation for metals) region, where stress varies linearly with strain. In conventional engineering applications, applied strain is typically small and E is a useful parameter in such cases [47]. However, E has limited relevance in soft robotics and other soft-matter technologies where materials undergo large elastic and/or inelastic and irregular (non-prismatic) shape deformations, with a typically non-linear stress response. Nevertheless, E is a useful parameter for comparing the mechanical rigidity of materials that are applicable in soft robotics. The elastic behavior, yield point and plastic deformation behavior of various classes of materials is shown in Fig. (1A), where the gap between rigid and soft material is clearly highlighted. Fig. (1B) shows Young’s moduli of various materials [48]. Conventional robots are made of rigid and semi-rigid materials such as metals and hard plastics, which have a modulus value greater than 10⁹ Pa [49]. In contrast, soft materials in natural organisms, such as human skin and cartilage, have a modulus in the order of 10²–10⁶ Pa [50, 51]. Therefore, the materials in conventional robots have higher rigidity (3–10 orders of magnitude higher) than the materials in natural organisms. This mismatch in mechanical compliance is one of the major factors for the incompatibility of conventional robots for intimate human interaction.

Fig. (1))

(a) Elastic properties, yield behavior, and plastic deformation of various materials, (b) Young’s modulus of various classes of materials (the transition from soft to rigid materials) [48].

1.2. Elastomers

An elastomer is a type of amorphous crosslinked polymer of low glass-transition temperature, which has a unique property of ‘elasticity’ along with low elastic modulus and high failure strain [52, 53]. A material is ideally elastic if it returns to its initial shape and size after it is stretched or compressed under an applied load. Elastomers are generally considered as an elastic material [54] since its stress-strain response is typically nonlinear and their elastic strain can completely recover over a period of time after removal of a transient load. Elastomers are also considered to be hyperelastic materials since they can respond elastically when subjected to very large strains.

Elastomers are the most potentially promising materials for soft robotics due to their low elastic moduli, very high stretchability and high elastic resilience. The elastic modulus of elastomers is typically in the range of 0.1-10 MPa, although soft robots are normally composed of highly deformable and compliant materials which typically have moduli in the range 0.1-1.0 MPa [55-57]. In soft robotics, a wide range of elastomers, such as silicones, liquid metal-embedded elastomers, dielectric elastomers, etc. are being widely used for both academic prototypes and initial commercial products due to their compliance, stretchability, elastic resilience, electromechanical efficiency, etc. Poly(di-methylsiloxane) (PDMS) is a class of silicone elastomer that is one of the most popular elastomers in soft robotics due to their low modulus, high strain limit, and relatively low hysteresis between loading and unloading cycles. There are several reports where PDMS is being used as a soft robust material for various soft robotics applications [59, 60]. However, PDMS poses some serious drawbacks, such as contamination issues from the curing process, low tear strength, high air permeability, poor abrasion resistance and limited service-life. Therefore, long-term reliability and performance of robot segments made from PDMS will be poor. In addition, their potential use for practical purposes is also under debate, because of their inherent time and temperature dependent viscoelastic properties, which can lead to unpredicted changes in the performance of the material. Natural and synthetic rubbers can be potential alternatives to address and overcome such problems. However, commercially available rubbers are not suitable for soft robotics due to a lack of adequate softness and functionality. There is a significant challenge to develop soft functional elastomeric materials from commercially available rubbers. By in situ modification of mineral-based filler structures, functional and mechanically adaptive natural elastomeric materials have been recently demonstrated, as shown in Fig. (2) [55, 58]. When robot skin becomes wet, the adaptive rubber skin will adapt its modulus automatically. Super-tough and ultra-soft liquid metal embedded natural rubber based composite materials using industrially viable solid-state mixing and vulcanization methods have recently been developed [61]. It was found that, in addition to substantial boosts in mechanical and elastic performance, the fracture energy of the material was enhanced up to six times after incorporation of liquid metal compared to control vulcanizate without liquid metal, as illustrated in Fig. (3). The balance between high flexibility and enhanced fracture toughness in commercially available elastomers will eventually provoke practical implications for future innovative materials for soft robotics.

Fig. (2))

(a) Mechano-Adaptive crosslinked rubber composites comprised of epichlorohydrine (ECO) rubber and calcium sulphate. (b) The composite can be made either soft or hard by heat or water treatment. (c) Storage modulus as a function of CaSO4 content. (reproduced with permission from ACS Publications) [58].

Fig. (3))

(a) Fracture toughness properties of super tough and ultra-soft liquid metal embedded functional natural rubber. (b) Comparison of crack growth rate vs tearing energy between unfilled natural rubber (NR) vulcanizate and 50 phr Eutectic Gallium-Indium (EGaIn)-loaded NR vulcanizate (NR−LM50) (reproduced with permission from ACS Publications) [61].

1.3. Dielectric Elastomers

Dielectric elastomers (DEs) belong to the group of electroactive polymers (EAP) - a relatively new class of compliant active materials, which can change its size or shape in response to an electric field [62-64]. EAPs are generally classified into two major classes, dielectric elastomers and ionic EAPs. The later has limited application in soft robotics due to its slow electric response and poor coupling efficiency. On the other hand, dielectric elastomers can be used as soft actuators to be integrated into robotic systems due to their rapid electric response (few milliseconds) and high electromechanical efficiency. DEAs are made of incompressible soft dielectric elastomer membranes, typically silicones or acrylic tape with the thickness of several microns, covered with compliant electrode layers on their top and bottom side to form dynamic capacitors. When an electric field is applied across the electrodes, the elastomer layer becomes thinner in response to the Maxwell stress derived from the electrostatic attraction between the charges on the opposing electrodes. A large electric field-induced deformation is generally obtained from the elastomeric actuator compared to conventional capacitor materials due to the elastomer’s lower mechanical stiffness. As the elastomer is incompressible, in-plane expansion of the DEA can be observed, which leads to voltage-tunable mechanical actuation, i.e. electrical energy We is converted into mechanical energy in the form of electrostatic pressure across the electrodes.

Dielectric elastomer actuators (DEAs) are one of the promising candidates for soft robotics due to their large voltage-induced deformation, fast response, high energy density, lightweight, compact structure and its physical nature like a human muscle [65, 66].

DEAs are being used in various sophisticated soft robotic applications, such as soft robotic gripper structures, aerial robot and humanoid robot applications. Grippers typically comprise an actuator segment with painted electrodes (see chapter 4, Sindersberger & Monkman). Gripper functions are generally controlled by applying a voltage to the actuator. The gripper can be brought closer to the object for grasping when voltage is applied to the actuator. Similarly, the gripper will close and grasp the object for manipulation when the voltage is removed. DEAs