Atomic-Like Packing for Scalable and Efficient Spheroid Pneumatic Artificial Muscles

Soft artificial muscles hold promise for wearable devices due to their flexibility, adaptability, and biocompatibility. However, scaling these actuators remains challenging, particularly in optimising spatial efficiency and performance. Bubble Artificial Muscles (BAMs) are soft spheroid pneumatic actuators that emulate biological muscle contraction, achieving high strains and stresses. This study evaluates BAM scalability by examining force and strain generation across three sizes (small, medium, and large).

Inspired by atomic packing, we propose compact multi-BAM configurations. Unlike conventional parallel and series arrangements, our approach enhances packing efficiency while maintaining low complexity. Additionally, we present an improved model of BAM actuation, accounting for the effect of actuator length on performance, alongside models of the atomic packing configurations.

Results show smaller BAMs deliver higher stress, and three small BAMs generate greater tensile force than a single large BAM (263.0 N vs 225.9 N at 70 kPa), while occupying 89% of the actuator cross-section and half the volume. Atomic-like packing optimises volume use, allowing flexibility for either high strain or high stress. The body-centered rectangular prism achieved the highest contraction (38.1%), while the face-centered cubic arrangement generated the highest stress (0.083 N/mm2). These configurations improve BAM adaptability, supporting applications in robotics, wearables, and soft exosuits.