An advanced computational framework is presented for the geometrically accurate simulation of 3D Concrete Printing (3DCP) across multiple spatial and temporal scales, encompassing extrusion, layer deposition, and buildability assessment. The framework couples a robust finite element environment with a novel constitutive formulation based on the Saramito elasto-viscoplastic model [1], here extended to cementitious materials for the first time. Using a limited set of experimentally identifiable parameters, the model reproduces elastic behaviour in the solid regime while enabling a smooth, thermodynamically consistent transition between fluid- and solid-like states. Pressure-dependent yielding is captured through a Drucker-Prager criterion, while thixotropy and early-age structuration are incorporated via time-dependent material properties. Large deformations and free-surface evolution, particularly relevant during extrusion and deposition, are resolved using the Particle Finite Element Method (PFEM), combining an updated Lagrangian formulation with an efficient Delaunay-based remeshing strategy. Dedicated numerical techniques are introduced to address the complex boundary conditions and contact interactions inherent to the printing process [2,3]. The framework is validated against experimental benchmarks conducted at TU/Eindhoven, including multilayer vertical and inclined walls. Laser-scanned geometries show strong agreement with numerical predictions, highlighting the influence of elastic and plastic effects on shape retention. Out-of-plane buckling and collapse mechanisms are also successfully reproduced. Finally, the framework is applied to the full-scale simulation of a printed structural component featuring interlayer merging and junction formation, demonstrating its predictive capability for complex 3DCP scenarios.