Probabilistic prediction of structural failure during 3D concrete printing processes

Structural failure during 3D concrete printing (3DCP) process due to the competition of two mechanisms, the elastic buckling and plastic collapse, has been largely observed in the experiments. Prediction of this phenomena has become as consequence an important task of this innovative construction technique. Due to its strong dependence on various parameters such as the properties of fresh concrete, the geometry of the printed structure and the printing parameters, the accurate prediction of the structural response during 3DCP is challenging. Specifically, the significant evolution in time of mechanical properties as well as the heterogeneous characteristics in nature of concrete at early age make it difficult to be determined and result in high uncertainty. This fact may be a reason for the important gap between the predicted failure of 3DCP and real experiments as usually stated in the literature. To improve the prediction, the probabilistic analysis is conducted in this work, to account for the uncertainty effect of fresh concrete properties on the structure's response. For this purpose, the Kriging metamodeling technique is chosen to estimate the probability of two failure modes of concrete structure during printing. The applicability and effectiveness of this probabilistic analysis for the 3DCP is demonstrated through a large numerical investigation that is conducted with different structure geometries and printing strategies. The present study elucidates the significant impact of uncertainties on the structural behavior during 3DCP, which has been ignored in the literature.