Structural vibration analysis and design of glued timber truss intermediate floors

This study investigates the vibration performance of intermediate floors constructed from glued timber trusses, the so-called TK-beams. A full-scale 10.5 m × 5.5 m TK‑beam floor system was tested through ten sequential construction stages, ranging from bare trusses to a completed layered assembly including OSB sheathing and a 50 mm concrete topping. A high‑fidelity finite element model was developed and calibrated using the full‑scale measurements. The numerical simulations were carried out following the sequence of construction stages to capture construction‑stage effects, composite action, and realistic boundary conditions as well as highlight the structural parameters that influence the most to the vibration response. Additionally, simplified analytical predictions, that are commonly used when designing such floor systems were compared against the measured as well as the simulated response. The finite element results showed deviations of only 0–15% above the measured values at beam mid span, indicating close agreement with experimental data. In contrast, the analytical predictions overestimated the measured values by 16–61%, demonstrating substantially larger discrepancies. Additionally, one of the key findings of the study is that in practical floor assemblies, the interpretation of the numerical results may not be trivial due to multiple nearly coincident vibration modes. • Full-scale vibration tests were conducted on glued timber truss intermediate floors across 15 construction stages. • Adhesive-only TK-beams demonstrated reliable stiffness and acceptable vibration performance in layered floor assemblies. • A calibrated FEM model accurately reproduced modal behavior, capturing composite action and mode switching effects. • Interlayer shear stiffness dominated frequency response, while boundary rotational stiffness had minor influence. • The finite element (FE) results showed deviations of only 0–15% above the measured values at beam mid span, indicating close agreement with experimental data. In contrast, the analytical predictions overestimated the measured values by 16–61%, demonstrating substantially larger discrepancies. This comparison highlights the superior accuracy and reliability of the FE model in capturing the system behavior relative to the analytical approach.