Abstract

The error in narrow-band analytical or physical model predictions is estimated for the example of a harmonic force-driven rectangular plate. It is shown that a 5% uncertainty in input parameters defining structural dimensions, material properties, and source location may account for errors in a predicted transfer function exceeding 10 dB over a reasonably broad bandwidth. It is also shown that this error increases 6 dB per doubling of the (percent) uncertainty. These results imply that in the absence of incoherent averaging, either temporal, spatial, or ensemble, expectations for favorable comparisons of structural acoustic predictions among models, or between model and prototype, may often be unrealistic.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call