Accurate Prediction of Frequency Response Functions of Tobacco Composite Panel

Abstract

Accurate experimental and analytical identification of Frequency Response Functions (FRFs) of bio-based composites such as a tobacco composite panel (TCP) is challenging due to their anisotropy and heterogeneity, and variability in damping. Conventional experimental modal analysis (EMA) procedures, specially designed for isotropic materials, cannot be directly adopted to TCP. Similarly, the finite element (FE) modelling approach is limited by difficulties in representing damping, which often result in large discrepancies in both resonance and off-resonance FRF amplitudes. This study proposes a framework combining optimised EMA configurations with calibrated FE modelling to achieve accurate FRF prediction of TCP. Comparative EMA setups using different orientation suspensions are investigated and the outcomes evaluated. Vertical suspension at two corners is identified as the most suitable configuration due to its ability to reduce constraint interference. FE FRFs are compared with the EMA counterparts, and predictive accuracy is evaluated using the Modal Assurance Criterion (MAC) and Mean Squared Error (MSE). Damping calibration, through which modal damping values are tuned in the FE model to correlate EMA off-resonance amplitude trends, yields good agreement between the FE and EMA FRFs. The findings show that damping significantly affects the characteristics of off-resonance and resonance FRF peaks. The proposed framework establishes a validated and practical approach for accurate FRF prediction of TCP, enabling its broader use in engineering applications.