A bi-level hierarchical method for shape and member sizing optimization of steel truss structures

• We introduce the BiOPT method for shape and sizing optimization of steel structures.
• A unique combination of algorithms is used to operate on shape and sizing variables.
• We benchmark BiOPT against existing methods using numerical examples.
• BiOPT compares favorably in terms of solution quality and computational efficiency.
• The method is applied to an industry project to demonstrate the practical benefits.

This paper describes a new bi-level hierarchical method for optimizing the shape and member sizes of both determinate and indeterminate truss structures. The method utilizes a unique combination of algorithms that are organized hierarchically: the Fully Constrained Design (FCD) method for discrete sizing optimization is nested within SEQOPT, a gradient-based optimization method that operates on continuous shape variables. We benchmarked the method against several existing techniques using numerical examples and found that it compared favorably in terms of solution quality and computational efficiency. We also present a successful industry application of the method to demonstrate its practical benefits.