Many damage models have been developed over the past decades, but most of them still struggle to simulate damage without additional ad-hoc criteria or the need for difficult implementation. In particular, brittle fracture models that represent a mathematical singularity are not easy to be simulate correctly. Recently discovered phase field models prove to be advantageous in many ways. These are variationally consistent models that show a high degree of robustness. Problems such as tensile cracking, cracking of a specialized structure in 3D, crack branching or crack initiation are all described by the same set of equations. Since the model uses a regularized damage field, its implementation is also efficient. Laminated glass is a material that needs to be investigated even after cracks appear. Undamaged laminated glass is a stunning material itself, but its undisputed advantage is its cohesiveness and load-bearing capacity even in post-breakage domain. And phase-field models are suitable for simulating this residual load capacity, which allow investigating both the initiation of cracks and their branching and the impact on the structural response. This work is supposed to bring the combination of such phenomena - the simulation of laminated glass before and after the appearance of cracks. Models of different degrees of spatial reduction and structures under different loading are investigated. The work first investigates the quasi-static response of the beams, then stochastically investigates the response when the strength changes and is concluded by investigating the low-velocity impact. The paper primarly tries to create fully applicable approaches for engineering practice and reveals the benefits and shortcomings of such an approach for glass modeling.