Experimental Investigation of Topology-Optimized Deep Reinforced Concrete Beams with Reduced Concrete Volume

Abstract

© 2020, RILEM. This paper presents an experimental investigation of digitally manufactured, reinforced concrete beams designed with topology optimization. The backbone of the current work is a hybrid mesh topology optimization algorithm that automatically generates strut-and-tie layouts. The resulting designs have tensile truss elements describing the reinforcing phase and compressive continuum force flow elements that illustrates how the concrete is carrying load. The aim of this work is to investigate the effect of removing a percentage of the non-load carrying concrete phase. A beam is designed with a standard, by-hand approach and the same steel amount is used in to generate a topology-optimized design. This work considers three beam designs; (i) the standard, (ii) a topology-optimized beam with a prismatic section (i.e. 100% concrete), and (iii) the topology-optimized steel layout in a beam with a reduced concrete volume (herein 75%). An alternative reinforcement method is used in which steel plates are cut by waterjet. To improve the bond quality between concrete and reinforcement, corrugations and anchors are added to the steel layouts. However, as opposed to previous experimental tests conducted by the authors, a poor bond quality is achieved, leading to premature failures of all test specimens. Due to the lack of proper bonding, comparison can only be made in the early elastic range. Here, a significant trend is that the by-hand and the topology-optimized specimens with 75% concrete exhibit near identical behaviors.