Denis Branthonne
Ira Dyer
Dept. of Ocean Eng., MIT, Cambridge, MA 02139
The objective of this research is to explore experimentally the effects of adding dynamic absorbers to a three-dimensional truss and to design analytical models which predict the phenomena of attenuation. The truss is built with cells made of aluminum rods and connected in series. 110 dynamic absorbers are mounted on the rods in the center cell to achieve a local mass ratio of 3. The truss is excited at one end with white noise to measure the spatial attenuation averaged on octave bands as a function of axial distance. The data are compared to the spatial attenuation for the undamped structure. The differences determine the effect of the dynamic absorbers. The overall shape of the differential curves of attenuation versus axial distance is a step function, with the step located in the dynamic absorbers attachment area. Assuming the equipartition of energy between the different wave types and using the classical theory of attenuation of waves propagating on a semi-infinite rod loaded with a continuous layer or dynamic absorbers, one predicts stronger attenuation than experimental data at low frequencies. The assumption of equipartition of energy proves to be incorrect and the applicability of the semi-infinite model is questioned. Subsequently, a new theory is derived and validated by experiment to describe the attenuation of waves on finite rods loaded with a layer of dynamic absorbers. At low frequencies, this model achieves a better estimation of the axial attenuation along the truss. [Work supported by ONR.]