Application of Self-Diagnosis and Self-Repair on a Truss Prototype That Adapts to Loading Through Shape Morphing

Abstract

This paper presents experimental testing of self-diagnosis and self-repair strategies on an adaptive truss prototype that counteracts the effect of loading through shape morphing. The prototype is a simply supported spatial truss with a span of 6 m and is equipped with 12 linear actuators. The structure is designed to adapt to external loads through shape morphing—that is, by undergoing large shape changes to achieve configurations that are optimal for load-bearing. A damage event is replicated via the removal of a truss element, which simulates a loss of stiffness caused by buckling or fracture. A damage detection and localization algorithm is implemented based on the similarity evaluation of numerical and empirical redundancy matrices. Testing results demonstrate the efficacy of this method, with up to 81% and 79% accuracy for detection and localization, respectively, obtained considering all scenarios including false alarms (false positives) in the nondamaged state. For damaged states, the detection accuracy is 100% (no false negative). A self-repair strategy based on shape morphing is proposed. The structure is controlled into a shape that is optimal to carry the external load, achieving a significant stress redistribution to mitigate the effect of damage. Experimental results demonstrate that when an element of the structure is removed to simulate damage, the stress increases by up to 22% compared to the undamaged condition. This increase is fully recovered through shape adaptation. Actuator faults were also analyzed. With all actuators in operation, shape adaptation reduces stress by up to 22% under peak load (in the absence of damage). When two actuators are simulated as faulty, a stress reduction of up to 11% is still achieved, demonstrating the effectiveness of the proposed shape morphing–based control strategy.