Application of Eurocode 2, ACI, and RILEM model in predicting shrinkage of 3D-printed concretes: Challenges and model optimization

Abstract

3D-printed concrete (3DPC) cures under nonstandard conditions due to the absence of formwork and external curing, which leads to rapid moisture evaporation. Combined with a unique mix design – characterized by high binder content and reduced water-to-cement ratio – this results in significantly different shrinkage behavior compared to conventional concrete, particularly at early ages. Measuring shrinkage in 3D-printed elements is technically challenging and often technically challenging, which emphasizes the need for reliable predictive models. Since well-established models exist for traditional concrete, this study focuses on estimating updated coefficient values for three widely used models – Eurocode 2 (EC2), RILEM B4, and ACI 209-92 – to enable their application to 3DPC. Experimental shrinkage measurements were performed on multi-layer printed specimens consisting of six layers, each with a 40×10 mm cross-section and a total length of 500 mm. Both noncontact laser sensors and LVDT devices were used. Model parameters were estimated using nonlinear regression techniques. The results show that all three models can accurately describe the shrinkage behavior of 3DPC after recalibration, with coefficients of determination exceeding 0.94. This confirms the potential of adapting existing shrinkage models – originally developed for conventional concrete – to the specific characteristics of 3D-printed materials.