Prior work demonstrates that an attached subordinate oscillator array (SOA) can attenuate vibration of a host structure. The distribution of masses and stiffnesses of the attached oscillators can craft a flat frequency response over a desired band. This response modification can be a significant improvement over classical dynamic vibration absorbers (DVA) that attenuate response at one target frequency while increasing the frequency response amplitude at nearby side frequencies. Performance of the SOA can be highly sensitive to the uncertainty or disorder in the mechanical properties of the system. This paper shows that use of piezolectric SOAs (PSOAs) has the potential to address and ameliorate such sensitivity to off-design situations. It is important to note that the design strategy is simple and effective: it can be carried out without optimization techniques by choosing simple or well-known distributions of electromechanical properties.

Dispersion relations describe the frequency-dependent nature of elastic waves propagating in structures. Experimental determination of dispersion relations of structural components, such as the floor of a building, can be a tedious task, due to material inhomogeneity, complex boundary conditions, and the physical dimensions of the structure under test. In this work, data-driven modeling techniques are utilized to reconstruct dispersion relations over a predetermined frequency range. The feasibility of this approach is demonstrated on a one-dimensional beam where an exact solution of the dispersion relations is attainable. Frequency response functions of the beam are obtained numerically over the frequency range of 0–50kHz. Data-driven dynamical model, constructed by the vector fitting approach, is then deployed to develop a state-space model based on the simulated frequency response functions at 16 locations along the beam. This model is then utilized to construct dispersion relations of the structure through a series of numerical simulations. The techniques discussed in this paper are especially beneficial to such scenarios where it is neither possible to find analytical solutions to wave equations, nor it is feasible to measure dispersion curves experimentally. In the present work, actual experimental data is left for future work, but the complete framework is presented here.

Steady-state traveling waves in structures have been previously investigated for a variety of purposes including propulsion of objects and agitation of a surrounding medium. In the field of additive manufacturing, powder bed fusion (PBF) is a commonly used process that uses heat to fuse regions of metallic or polymer powders within a loose bed. PBF processes require post-process removal of loose powder, which can be difficult when blind holes or complex internal geometry are present in the fabricated part. Here, a preliminary investigation of a simple part is conducted examining the use of traveling waves for post-process de-powdering of additively manufactured specimens. The generation of steady-state traveling waves in a structure is accomplished through excitation at a frequency between two adjacent resonant frequencies of the structure, resulting in two-mode excitation. This excitation can be generated by bonded piezoceramic elements actuated by a sinusoidal voltage signal. The response of the structure is affected by the parameters of the excitation, such as the particular frequency of the voltage signal, the placement of the piezoceramic actuators, and the phase difference in the signals applied to different actuators. Careful selection of these parameters allows adjustment of the quality, wavelength, and wave speed of the resulting traveling waves. In this work, open-top rectangular box specimens composed of sintered nylon powder and coated with fine sand are used to represent freshly fabricated parts yet-to-be cleaned of un-sintered powder. Steady-state traveling waves are excited in the specimens while variations in the frequency content and phase differences between actuation points of the excitation are used to affect the characteristics of the dynamic response. The effectiveness of several response types for the purpose of moving un-sintered nylon powder within the specimens is investigated.

Current acoustoelastic-based stress measurement techniques operate at the high-frequency, weakly-dispersive portions of the dispersion curves. The weak dispersive effects at such high frequencies allow the utilization of time-of-flight measurements to quantify the effects of stress on wave speed. However, this comes at the cost of lower sensitivity to the state-of-stress of the structure, and hence calibration at a known stress state is required to compensate for material and geometric uncertainties in the structure under test. In this work, the strongly-dispersive, highly stress-sensitive, low-frequency flexural waves are utilized for stress measurement in structural components. A new model-based technique is developed for this purpose, where the acoustoelastic theory is integrated into a numerical optimization algorithm to analyze dispersive waves propagating along the structure under test. The developed technique is found to be robust against material and geometric uncertainties. In the absence of calibration experiments, the robustness of this technique is inversely proportional to the excitation frequency. The capabilities of the developed technique are experimentally demonstrated on a long rectangular beam, where reference-free, un-calibrated stress measurements are successfully conducted.