Saeed Mollaee ^* Poul Fønss Nielsen Andrew James Taberner David Budgett Amir HajiRassouliha

• Auckland Bioengineering Institute, Faculty of Engineering, University of Auckland, Auckland, New Zealand

In this work, we present a simple, inexpensive, and scalable pneumatic silicone actuator array that can dynamically confirm to the skin surface, to relieve pressure concentrations in a prosthetic socket. We identify, parameterize, and validate the best constitute laws for creating a finite element model of our pneumatic actuators. Using the model, we evaluate the surface deformation field elicited by 270 different variations of soft actuator array design parameters, under realistic loading, and demonstrate an accuracy of better than 70 µm. Our results quantify how each design parameter of the soft actuator array affects the surface deformation, and thereby, the pressure distribution. Our findings are applied here to the design of a dynamically reconfigurable socket for transtibial amputees but are highly translatable to other robotics problems requiring soft, deformable, load-bearing surfaces. To address the complex non-linear deformation of the surface, a novel speckle imaging technique was employed, allowing for the precise measurement of surface displacement with an accuracy of 40 µm. These measurements facilitated the identification of the Ogden N3 model's capability to predict actuator deformation with an accuracy within 16%. In the case of the soft actuator array with an 8 mm void diameter and 8 mm thickness, the maximum shear displacement was approximately 2500 µm. However, this displacement increased to around 6000 µm when the void diameter expanded to 16 mm.