Characterization of Clear Urethane Plasticizer for Use in Optical Waveguide Strain Sensors in Soft Robotics

One of the key challenges in soft robotics is proprioceptive sensing due to the robots’ deformability and infinite degrees of freedom. Most often, soft robots rely on motion capture systems, which require external hardware like cameras. One of the proposed benefits of soft robots is their ability to navigate unstructured environments, however, the nature of these environments often leads to occlusions that interfere with motion capture sensing. Recently, researchers have proposed methods to integrate sensors into the robot body which have the potential to perform better in these environments; examples include galinstan and other liquid-metal resistive strain sensors. However, these sensors are prone to oxidizing, leading to hysteresis in sensor measurements over time.<br/>We propose using optical strain sensors to sense motions within the robot body. These sensors are made by cladding a fiber made of clear, stretchable, high refractive index material with a low refractive index material. Light travels differently along the length of the fiber when it is deformed by stretching, bending, or locally applying lateral pressure. Prior work in this area has established methods to differentiate deformation modes based on the intensity and time of flight through these fibers.<br/>Optical fiber strain sensors have little hysteresis or degradation over time, they are indifferent to electromagnetic interference and have large bandwidth. However, prior efforts to implement these sensors have been hampered by a mechanical mismatch between the core and cladding materials. One of the more stretchable versions of this type of sensor is made from aurethane core material with silicone cladding. Most optically clear urethanes, however, are somewhat stiff. Clear Flex 30™ made by <i>Smooth-On</i> is a clear urethane with lower stiffness than most optically clear urethanes. It has been used in soft robotics in our work and the work of others. It is commercially available in shore hardnesses of 30, 50, or 90 A. Usually, the silicone is picked to match this material’s mechanical properties as closely as possible. This avoids issues like fiber slip and non-uniform robot inflation due to the fibers acting as a local reinforcement to the robot wall. Matching silicone cladding stiffness to the urethane core fiber limits the material design options. Higher stiffness robots allow for greater lifting capacity, while lower stiffness robots are more deformable. It is also sometimes desirable to match the stiffness of something like human skin or an object to be manipulated.<br/>To address these issues, new formulations of optically clear urethanes were created to match the mechanical properties of a range of cladding or substrate materials. Optical and mechanical analyses of urethanes modified with a commercial plasticizer (<i>Smooth-On</i> So-Flex II™) were performed to measure the basic properties of the new formulations and to determine suitability for sensing applications. Digital image correlation tensile testing shows that this additive can modify the 100% modulus from 380 kPa for the unmodified material to 200 kPa at a concentration of 20% by weight in Clear Flex 30™. At that concentration, spectrophotometer testing shows negligible optical loss compared to the unmodified material meaning the additive does not negatively affect usable fiber length. A refractive index test at 632.8 nm found that the refractive index of the modified material differed from the unmodified material by 0.01, dropping from 1.48 to 1.47. Further testing will establish a larger spectrum of refractive indices. A tensile specimen made of Dragon Skin 10™ silicone cladding and a Clear Flex 30™ core with So-Flex II™ showed adequate total internal reflection for a 12 cm long sensor; having approximately 0.14 dB additional loss under 50% strain.