Abstract

The application of additive manufacturing (AM) technology in architecture and structural engineering has been extended due to recent development of metal 3d printing. In space frame structures, a set of bars is configured in three dimensions, with bars connected by nodes. This article presents two methods to design metal 3d-printed multi-branch nodes to accommodate any number of incident bars at arbitrary angles. Resulting node designs are intended to be smooth and lightweight. A multi-branch node is sketched using the dimensional information of the blank space between the converging bars in a pre-designed space frame and then parameterized by two different approaches to perform structural optimization. The first design method, namely the curve parameter method, which is semi-automated approach, the distances between the control points of the spline curves between node branches and the node branch intersection point are the optimization parameters. For the other method, called fatness parameter method, which is a fast and automated approach, the fatness parameters of the center part of the node and the root radiuses of each branch are chosen as the main parameters of optimization. The optimization procedure is accomplished using a genetic algorithm to minimize the maximum von Mises stress as the objective function subjected to the mass of the node as a constraint function. Finally, functional tests are conducted on 3D printed metal nodes in order to compare the strength and stiffness of the nodes designed by the two form-finding approaches.