Methodology for surface defect assessment in 3D concrete printing using computer-vision and ultrasonic testing considering structural build-up

Abstract

This study quantifies how interlayer time gaps and mixture rheology jointly influence interfacial integrity and mechanical performance in 3D concrete printing (3DCP). Two mixtures incorporating identical binders, but differing limestone fineness (4 and 50 mu m) were printed with interlayer delays of 0, 2, and 5 min, corresponding to stacking rates of approximately 9, 0.5, and 0.2 m/h. The finer limestone mixture exhibited higher plastic viscosity (13.2 vs. 11.5 Pa s) and faster structural buildup (58 vs. 38 Pa/min), resulting in accelerated early stiffening. Deposition yield stress derived from buildup measurements remained low for 0-2 min (1400-1500 Pa for the higher-thixotropy mixture; 1100-1500 Pa for the lower one) but increased markedly at 5 min to 4100 and 2900 Pa, respectively. Six single-wall prints were evaluated via surface defect imaging, layer-resolved ultrasonic S-wave velocity mapping, and 28-d compressive testing of specimens extracted from lower (layers 1-5) and upper (6-10) regions. Defect density rose with both delay and height, reaching 14.5% in the upper layers of the higher-thixotropy mixture at 5 min, while the 0-min condition remained <= 2%. S-wave velocity declined from 2100 to 2150 (0 min) to 1,800-1,870 m/s (5 min), accompanied by a strength reduction from 50 to 25-30 MPa. The findings define a practical deposition window, yield stress of 1100-1500 Pa, that minimizes interlayer defects under realistic delays, and demonstrate that layer-wise S-wave monitoring sensitively captures vertical stiffness gradients, providing a viable process-control metric for large-scale additive construction.