The prevention of unfavourable machine-induced vibrations represents a crucial issue for the design of industrial facilities. A special attention is required for the structural assessment of the load-bearing members, that should be optimally designed with the support of specific input parameters. The characterization of the expected vibration sources, together with a reliable structural model, is in fact a key step for the early design stage. In this paper, a case-study eyewear factory is investigated. Its layout takes the form of a two-span, two-story precast concrete frame. The lack of customer/designer communication resulted in various non-isolated Computer Numerical Control (CNC) vertical machinery centers mounted on the inter-story floor. Accordingly, the floor started to suffer for severe resonance issues. This research study focuses on the dynamic investigation of the structure. An efficient, coupled experimental-numerical approach is presented and validated for early predictive studies. Based on field experiments on the floor, but also on the machinery components, the most unfavourable conditions are first detected and characterized with the support of accelerometers and video-tracking displacement acquisitions. The experimental outcomes are then further assessed with Finite Element (FE) numerical models, giving evidence of the accurate predictability of resonance issues.
