Automatic Truss Design Based on Topology Optimization and Image Processing Techniques

Abstract

Topology Optimization (TO) methods have been well developed for structural and multidisciplinary designs in terms of finding the optimal objective functions subject to constraints of loading conditions and other design requirements. In practical uses, clear black-and-white optimal designs are desired but may not be delivered at the end of the TO procedures. Therefore, many additional treatments such as regularization and filtering have been developed to produce black-and-white designs. Some of the said procedures were also found in the fields of computer vision and image processing. This paper presents a method to automatically generate truss designs by applying image-processing techniques to TO designs. The presented post-processing procedure includes noise removal, pattern thinning, and automatic determination of truss nodes linages. Several numerical examples were shown to demonstrated the presented methodology for automatic truss designs.