As more advanced concrete based infrastructure materials find their way into practice, also increasing is the demand of computational models capable of simulating and predicting their behavior at early age and beyond. In this work, a computational aging framework coupling a hygro-thermo-chemical (HTC) theory and a comprehensive mesoscale Lattice Discrete Particle Model (LDPM) is proposed. The proposed A-LDPM model is then utilized to simulate and predict various mechanical behaviors of Ultra High Performance Concrete at early age with calibrated mesoscale material parameters based on a comprehensive experimental campaign. Size effect studies are also carried out with experiments and A-LDPM modeling, which shows great capabilities of accurately capturing the age dependent size effect and fracture characteristics of UHPC. With the final goal of modeling structural scale responses, LDPM is utilized to simulate mechanical behavior of five meters long shear beams with full reinforcement and various prestress levels. The simulated results, failure types, major crack locations and magnitudes have high agreement with those of experiments. However a gap exists between the experimental and simulated stiffness of response due to not including creep and shrinkage effects. Thus, an extended A-LDPM framework is proposed to also take into account the coupled effects of creep, shrinkage and steel relaxation.