RH Lyon (United States)

It is well known that the real part of a drive-point function, averaged in frequency, can be estimated from the modal density and mass of a structure. Furthermore, the results are often the same as when the modal overlap is large due to damping or when the system boundaries are infinitely extended. The present note examines the possible use of modal density in estimating the imaginary part of the same system drive-point function.

The sound radiation from a guitar contains about 50 fundamental semitones and about 200 harmonics. The resonant modes of the guitar body are insufficient in number and of insufficient bandwidth to support all of these frequencies, so some frequencies are coincident with resonances and some are not. It is the purpose of this paper to discuss the role of resonances as they occur in the guitar top, sides, and back in affecting the radiation. Conjectures about the role of the nonresonant radiation in the attack of the sounds and the hearing mechanism on the perception of the sound are presented in a tentative manner.

When forces are applied to a flexible structure, a worker in structural acoustics who is concerned with the radiated sound will normally be interested in how much of the sound is arising from the resonant response of the structure compared to the nonresonant response. Lightweight panels below their critical frequency are likely to have a significant amount of sound arising from nonresonant response due to mass inertial response localized around the location of force application. Since a guitar top is such a structure, it is natural to inquire about the comparison between resonant and nonresonant radiation from a guitar over its normal range of frequencies. Measurements have been made on a student Yamaha and a concert quality Modranas/Connor guitar. The results of these measurements and their interpretation will be discussed.

An unresolved issue in the author’s Ph.D. thesis concerned the relation between turbulent flow speed and the maximum response of a string excited by the turbulence. The maximum response occurred at a flow speed significantly less than that predicted. This discrepancy was not resolved at the time, but can be resolved if one notes that the flow excitation is dispersive, with longer scale disturbances that flow faster than smaller scale disturbances. It is suggested that string response to turbulence may be a useful way to determine small-scale properties of a turbulent flow.

Curt Holmer was always interested in the practical applications of statistical energy analysis, and the author had many discussions with him regarding its uses, and areas of transition to deterministic analysis. In some of the latest discussions, the topic of the ‘‘mid‐frequency range’’ dominated, referred to by the author as ‘‘the awkward zone.’’ In this transition range between deterministic and fully probabilistic models, the greatest uncertainty may be in the joints—their actual location, and their rigidity. This suggests a hybrid approach, in which the substructures are treated as deterministic, and the joint locations and rigidities are uncertain, or in popular terms, fuzzy. This approach is exemplified by the joining of two plates by a link having a location and stiffness that are variable over a small range. The results of the analysis support such an approach, in that they show the kind of transition that has been noted in the awkward zone.