Form Finding Nodal Connections in Grid Structures

Nodes for grid structures are often manufactured in a rather material intensive and inefficient way, increasing the weight of the structure and thus the load. Recent development of additive manufacturing techniques, have resulted in a rising interest in large-scale metal 3D printing. Topology optimization has become the obvious companion in the design of structural parts for 3D printing, and rightfully so. The technique is demonstrably able to provide material efficient solutions and is well suited for a manufacturing technique with few formal restrictions. However, from a designer’s perspective one could argue that topology optimization have some limitations. Like other “automated processes”, it tends to take over and does not leave much room for other form drivers. This paper presents an alternative method for designing material efficient nodes in grid structures that builds on the conventional form-finding techniques, usually applied to create minimal surface tensile structures or gravity shell like structures. The technique works by modelling the node as a hollow shell with a mesh, applying a set of tensile forces derived from the structural action from elements adjacent to the node (where compression is converted to tension) and running a form finding simulation. After the simulation, the shell is then thickened and analysed for the real load case (which consider both tension and compression) using FE-analysis. The benefit of such technique is that the designer has control over the topology of the design which enables more creative control and free exploration of a range of design variations. The form finding is done using dynamic relaxation and introduces spline elements with bending capability to control deviation from the pure spring network solution