Automated discovery and optimization of large irregular tensegrity structures

  • Authors:

  • John Rieffel;Francisco Valero-Cuevas;Hod Lipson

  • Affiliations:

  • Department of Mechanical and Aerospace Engineering, Cornell University, 138 Upson Hall, Ithaca, NY 14853, United States;Department of Mechanical and Aerospace Engineering, Cornell University, 138 Upson Hall, Ithaca, NY 14853, United States;Department of Mechanical and Aerospace Engineering, Cornell University, 138 Upson Hall, Ithaca, NY 14853, United States

  • Venue:
  • Computers and Structures
  • Year:
  • 2009

Quantified Score

Hi-index 0.01

Visualization

Abstract

Tensegrities consist of disjoint struts connected by tensile strings which maintain shape due to pre-stress stability. Because of their rigidity, foldability and deployability, tensegrities are becoming increasingly popular in engineering. Unfortunately few effective analytical methods for discovering tensegrity geometries exist. We introduce an evolutionary algorithm which produces large tensegrity structures, and demonstrate its efficacy and scalability relative to previous methods. A generative representation allows the discovery of underlying structural patterns. These techniques have produced the largest and most complex irregular tensegrities known in the field, paving the way toward novel solutions ranging from space antennas to soft robotics.