Improvement of acoustic and vibration models by temporal error metrics

Abstract

Analyses and examples are presented that explore the limits and accuracies of a technique for improving acoustic and vibration models by temporal comparisons to experimental data. In a previous presentation, the authors proposed the use of impulsive excitations followed by time windowing of responses. This approach allows comparisons between experimental data and model predictions over an isolated spatial region whose volume is a fraction of the entire system volume. The advantage of this spatial isolation is that it significantly reduces the number of model parameters that must be varied to bring the model predictions into agreement with the experimental data. In the present work, the method is analyzed in detail to quantify the limits and accuracies of the method relative to window size, number of measurement locations, and excitation. Two types of examples will be presented to illustrate these findings. The first involves the improvement of material properties for a homogeneous region. The second involves the improvement of the coupling conditions between two homogeneous regions. Results of these examples will be reviewed in the context of existing and emerging measurement technologies, such as digital image correlation.