Predicting Buildability Using the Surface Texture of 3D Printed Concrete Elements

Abstract

The buildability of three-dimensional (3D) printable concrete is commonly measured by the number of layers that can be printed before a collapse or excessive deformations. Buildability depends on rheological properties and their dynamic changes over time due to hydration and evaporation. These variations influence strength development and lead to dimensional changes in individual layers, potentially resulting in failures such as plastic or buckling collapse. Despite the importance of dimensional changes, no studies have monitored these variations in real time to predict buildability failure. In this study, dimensional changes are indirectly assessed by tracking surface texture changes using two-dimensional (2D) cameras and computer vision techniques. Entropy standard deviation (ESD) is introduced as a metric to quantify temporal textural changes and assess buildability collapse. Results indicate that significant variations in surface texture values of individual layers are observed in collapsed elements, allowing for failure prediction before the collapse. The limiting ESD value for a concrete mix can be identified by carrying out a set of experimental prints. This value could be used for the early prediction of buildability collapse. Experimental data show that buildability collapse can be predicted with 100% accuracy by monitoring the maximum ESD values of all the printed layers. Based on this concept, a methodology has been developed for real-time, nonintrusive buildability assessment of 3D printed elements, offering the potential for feedback control systems to enhance quality, reduce material wastage, and improve the sustainability of concrete 3D printing technology.