Development of a Vibroacoustic Stochastic Finite Element Prediction Tool for a CLT Floor

Abstract

Low frequency impact sound insulation is a challenging task in wooden buildings. Low frequency prediction tools are needed to access the dynamic behavior of a wooden floor in an early design phase to ultimately reduce the low frequency impact noise. However, due to the complexity of wood and different structural details, accurate vibration predictions of wood structures are difficult to attain. Meanwhile, a deterministic model cannot properly represent the real case due to the uncertainties coming from the material properties and geometrical changes. The stochastic approach introduced in this paper aims at quantifying the uncertainties induced by material properties and proposing an alternative calibration method to obtain a relative accurate result instead of the conventional manual calibration. In addition, 100 simulations were calculated in different excitation positions to assess the uncertainties induced by material properties of cross-laminated-timber A comparison between the simulated and measured results was made in order to extract the best combination of Young’s moduli and shear moduli in different directions of the CLT panel.