Stochastic identification of temperature effects on the dynamics of a smart composite beam: assessment of multi-model and global model approaches

The temperature effects on the dynamics of a smart composite beam are experimentally studied via conventional multi-model and novel global model identification approaches. The multi-model approaches are based on non-parametric and parametric VARX representations, whereas the global model approaches are based on novel constant coefficient pooled (CCP) and functionally pooled (FP) VARX parametric representations. The analysis indicates that the obtained multi-model and global model representations are in rough overall agreement. Nevertheless, the latter simultaneously use all available data records offering more compact descriptions of the dynamics, improved numerical robustness and estimation accuracy, which is reflected in significantly reduced modal parameter uncertainties. Although the CCP-VARX representations provide only 'averaged' descriptions of the structural dynamics over temperature, their FP-VARX counterparts allow for the explicit, analytical modeling of temperature dependence exhibiting a 'smooth' deterministic dependence of the dynamics on temperature which is compatible with the physics of the problem. In accordance with previous studies, the obtained natural frequencies decrease with temperature in a weakly nonlinear or approximately linear fashion. The damping factors are less affected, although their dependence on temperature may be of a potentially more complex nature.