We developed a shape-changing constructive kit, named Mimosa1. A key component of the toolkit is the modular hinges, each of which is equipped with two antagonistic shape memory alloy (SMA) wires. One wire deforms the hinge to approach its predetermined angle at high temperature, and another wire drives the hinge back when it cools down. Hinge leaves are available in different materials including acrylic, cardboard and textile, which increases the versatility of the toolkit. Every hinge weighs 2.1-5.4 g, and generates up to 5.7 N actuation force. A Bluetooth control module was developed, enabling remote control of the shape-changing objects. Mimosa aims to inspire designers to explore and create interactive shape-morphing objects with SMAs. A few examples are given such as a gripper, a rolling robot, a butterfly, an airplane and a self-closing pocket. A workshop study with 6 participants showed that Mimosa indeed motivated and inspired the participants to create new ideas.

Conventional hinge actuators often face limitations including excessive weight, large size and unpleasant noise. Shape memory alloys (SMAs) offer a solution to address these issues due to their favorable characteristics, such as lightweight, high actuation force and small form factor. However, most existing SMA-based hinge actuators rely on the tension loading mode. Achieving an ideal actuation angle thereby necessitates the inclusion of long SMA wires, which inadvertently constrains the actuator size. Notably, the full potential of SMAs’ deformation capacities, encompassing torsion and bending, remains largely untapped and underutilized. In this research, a reversible torsion SMA hinge actuator is studied, which can reversibly open 60° during heating and cooling. The actuator weighs 2 g, and can produce actuation forces of up to 5 N. The mechanical performances of nitinol at different temperatures are measured. Based on the measurements, a model which can predict the opening and closing angle is proposed, with deviations of 13.5 ± 8.2 %. Gripper and butterfly demonstrators constructed by the hinge actuators are given as application examples. The actuators hold potential in many fields like soft robotics, aerospace and medical instruments.

Devices delivering sophisticated and natural haptic feedback often encompass numerous mechanical elements, leading to increased sizes and wearability challenges. Shape memory alloys (SMAs) are lightweight, compact, and have high power-to-weight ratios, and thus can easily be embedded without affecting the overall device shapes. Here, a review of SMA-based haptic wearables is provided. The article starts with an introduction of SMAs, while incorporating analyses of relevant devices documented in the literature. Haptic and SMA materials fields are correlated, with haptic perception insights aiding SMA actuator design, and distinct SMA mechanisms offering diverse haptic feedback types. A design process for SMA haptic wearables is proposed based on material-centered approach. We show SMAs hold potential for haptic devices aiding visually impaired people and promise in immersive technology and remote interpersonal haptic communication.

Shape memory alloy (SMA) wires are excellent candidates for wearable actuators since they are thin, low weight and have a high actuation force. The main drawbacks are that the wire should be kept straight and needs to be relatively long to enable a large enough actuation stroke. Embedding the SMA wire in a flexible tube largely enhances its applicability since then the counter forces are transferred by the tube material and the tube can be rolled up or attached to flexible surfaces or clothing layers. The performance of such tube-guided SMA actuators is, however, more complicated since it not only depends on the SMA behaviour but also on the tube materials and the actuator construction. In this research, a simple end-state model for a tube-guided SMA actuator system is proposed. We measure and model both the SMA and tube material properties, including tube creep effects, and derive an approximate prediction for the actuator stroke. Validation experiments showed that the predicted stroke during the second heating and cooling experiments agreed well with the measurements and that the average deviation is 9.6%, even though the deviation is much larger (27.3%) for the maximum applied force.