Topology Optimized Reinforced Concrete Walls Constructed with 3D Printed Formwork

Abstract

The construction industry continually evolves to adapt to gains in knowledge, market pressures and new technologies. However, two promising new technologies, 3D printing and computational topology optimization, have not yet penetrated the civil engineering industry despite being important drivers of change in other fields. The aim of this study was the potential to overcome the major barriers to adoption of both technologies by using them in combination. Both theoretical and practical problems must still be addressed, but the potential impacts are significant: lightweight, architecturally pleasing, reduced volume structures.

Two small-scale specimens were constructed and tested to demonstrate the feasibility of using additively manufactured (3D printed) formwork to construct complex reinforced concrete (RC) structures. The concept was shown to be viable. Areas were identified where further development is necessary before 3D printing can be used for large-scale cost-competitive formwork. An approach, based on the rule of mixtures, was proposed for applying computational topology optimization to RC structures. This was necessary because the computational topology optimization algorithm employed in this study assumes a structure is homogenous but RC structures are not. The approach was shown to work for optimizing an RC wall for force demands within the linear-elastic range of response.

The sensitivity of optimization outputs to modeling parameters was investigated. The effects and interdependencies of mesh size, element type, number of optimization cycles, and target volume ratio on optimization outcome were demonstrated. The importance of ISO and “percent reduction” parameters on the process of importing the optimized geometry to ABAQUS was also demonstrated.

Finally, a parametric study was conducted to examine the relationships between volume ratio and member strength and stiffness (volume ratio refers to the volume of the optimized structure divided by the volume of the original structure). The study used finite element models of topology optimized slender structural walls subjected to pseudo-static lateral force. It was shown that reductions in volume are not proportional to reductions in stiffness, as expected for slender walls that are flexure-dominated. Reductions in volume of 10 to 20% cause only approximately 3 to 7% reductions in uncracked member stiffness. These reductions in stiffness can be compensated for with use of modestly higher-strength concrete.