3D-printed concrete (3DPC) using toothed nozzles exhibits improved mechanical properties, attracting significant research attention. This study investigates the automated enhancement effect of novel toothed nozzles on interlayer tensile strength. The interlayer tensile strength is influenced by many factors, and key parameters, such as the area proportion coefficient (the ratio of cohesive failure area to total tension failure area at the interlayer interface), are difficult to determine, making strength prediction challenging. To address this, a hybrid CTGAN-XGBoost model is developed. CTGAN augments limited experimental data, and XGBoost accurately predicts tensile strength and area proportion coefficient. Explanatory analysis identifies edge length, depth, and bottom width as the most significant factors. Based on interpretability analysis, optimized nozzle parameters were experimentally validated, demonstrating enhanced performance and model effectiveness. This study supports automated strength prediction and nozzle design for 3D concrete printing and extends to other digital manufacturing processes in concrete. • Toothed nozzles significantly improve interlayer tensile strength in 3DPC. • The CTGAN effectively augments small datasets for 3D concrete printing. • The XGBoost accurately predicts tensile strength and area proportion coefficient. • Edge length, depth, and bottom width dominate toothed nozzle performance.
Jiang et al. (Sun,) studied this question.