Fluid-structure interaction (FSI) simulations can be quite challenging due to a variety of reasons, and their success is often dependent on their specific application. One such challenging application is the simulation of the effects of blast-loading on thin steel plates that undergo large deformations all the way up to complete failure. To achieve accurate simulations, robust and specific FSI algorithms are required. Fortunately, recent advancements in experimental and numerical frameworks have allowed for detailed studies on the FSI during the dynamic response of blast-loaded steel plates. Therefore, combining experiments and numerical simulations can provide greater insight into the underlying physics of the blast-structure interaction. This work presents ongoing research on the influence of FSI effects on the ductile crack growth in perforated, thin steel plates subjected to blast loading. Despite extensive studies, we do not fully understand the significance of FSI effects in these loading scenarios. Thin steel plates can undergo large, inelastic strains and ductile fracture, making them suitable for studying this phenomenon. However, accurately representing the loading and damage localization requires a fine mesh size, which can significantly increase computational (CPU) costs. To address this challenge, we evaluate the effectiveness of adaptive mesh refinement (AMR) in reducing CPU costs while maintaining solution accuracy. Specifically, we investigate the performance of FSI- and damage-based AMR in predicting the dynamic response and failure of coarsely meshed shell structures subjected to blast loading. Numerical simulations are conducted in the EUROPLEXUS software. To predict the observed fracture patterns accurately, it was found that the most crucial aspect is having a precise description of the blast loading during the FSI phase of the dynamic response. Moreover, the use of immersed FSI algorithms in conjunction with finite volume discretization of the compressible flow was found to yield highly promising results.