Simulation of soft regolith dynamic anchors for celestial exploration

Recent exploration missions to celestial bodies have shown an increasing demand for surface based landers and rovers designed to perform experiments on the ground, rather than relying purely on traditional orbiting observatories. Many of the scientifically interesting locations have proven hazardous and difficult to reach and traverse, driving the need for different methods of locomotion. Some of these locations lie in deep, permanently shadowed craters or in rocky, highly uneven landscapes. Various wheeled, flying, jumping, and legged rovers have been proposed. Those chosen for development have experienced both success and problems alike. Even stationary landers, such as the Philae lander which attempted to perform a controlled landing onto a comet surface, encountered unforgiving terrain causing it to bounce multiple times due to the ineffectiveness of its two on-board anchoring mechanisms. A new generation of legged rovers and landers is envisioned to utilize dynamic anchors on the feet of its legs to claw into the surface, engaging and disengaging with each step or landing. A method for simulating and evaluating the performance of these dynamic anchors is proposed to aid in-progress surface missions with relatively quick response to new target data. Discrete Element Method software is used to simulate a lunar-like regolith medium and the interaction of a dynamic anchor with this medium. The engagement, holding, and disengagement forces are recorded during this simulation. Physical testing was performed by using a robotic arm to engage a series of anchors with a lunar regolith simulant while measuring the same three forces as the simulation. The actual test data efficient anchor geometry as determined during testing is compared to predicted data to evaluate the simulation accuracy. Calibration testing to determine suitable simulation parameters is also presented. Results show the applicable forces can be predicted well within an order of magnitude, but improvements are possible to predict soil behavior more accurately.