An overview and comparison of structural form finding methods for general networks

Abstract

This paper discusses and compares existing form finding methods for discrete networks. Well-known methods such as the force density method, dynamic relaxation, updated reference strategy and others are discussed by mathematically structuring and presenting them in the same way, using the same notation and combining terminology. Based on this, a single computational framework using a sparse branch-node data structure is presented. It is shown how each method approaches the initial equilibrium problem, defines and linearizes the equilibrium equations applied to linear elements, and uses particular solving strategies. This framework marginalizes any differences related to operating platforms, programming language and style, offering a better baseline for independent comparison of performance and results. As a consequence, it is possible to more clearly relate, distinguish and compare existing methods, allow for hybrid methods and identify new avenues for research.