Search

An asymptotic distributed transfer function method (TFM) was developed for conical shell analyses. Formulations of the TFM involved the use of partial differential equations, the analysis of eigenvalue problems and Fourier expansions as well as perturbation techniques. The static responses of different types of conical shells were determined using numerical analyses procedures. Results pointed to the convenience of asymptotic TFM for use in computer coding and numerical simulations.

A study was made to assess the effects of high-temperature gradients on sound transmission in an acoustic probe tube. The tube has its probe sensing end located in the hot environment and the transducer located at the cool end. The results show that a theoretically derived transfer-function can be used to correct acoustic data measured by a high-temperature probe tube. Moreover, the transfer-function model must have visco-thermal effects in the probe tube in order to be accurate. Agreement between the computed and measured attenuation is good over the frequency range from 0.4 to 6 kilohertz, with some discrepancies occurring at attenuation minima.

Structural damage method uses active member transfer function information, in coordination with artificial neural network, to detect damages in various structures. The neural network training uses the pole/zero transfer function data to specify damage measurements and predict the level of damage in the structure. The efficacy and validity of the method are demonstrated in a 10-bar truss structure.