Session: SYMP 2-4: Structure and Performance of Shape Memory Polymer Actuators
Paper Number: 140421
140421 - Data Driven 1d - Dispersion Estimation in Hydraulic Fluid Media
Acoustics are commonly analyzed using a source-path-receiver paradigm, in which the source and receiver are often pre-determined by design requirements. Consequently, the characteristics of the path typically have the most significant effect on the acoustic profile developed by a dynamic system. In particular, the wave propagation characteristics of the medium are critical for analyzing the propagation of noise through a system. As a result, understanding the behavior of wave propagation within a medium is essential to characterize the acoustic properties exhibited by a system. The most relevant of these characteristics is the speed at which a wave propagates, which can be derived from the relationship between spatial wave number and cyclic frequency. Current testing approaches rely on time of arrival, or "Pitch and Catch", analysis which is limited to higher frequencies where reflection from boundaries do not interfere with time estimation. To accurately estimate arrival time, a sufficient number of wavelengths are required between measurement locations to allow for measurable time to pass. This results in cumbersome and impractical testing conditions that are often unobtainable for low frequency measurements. To experimentally determine the relationship between wave number and frequency without relying on time of arrival methods, estimated acoustic mode shapes are utilized to compute their spatial wave number. This wave number is plotted against the corresponding resonance frequency to form a dispersion curve. Multiple harmonic modes are measured to more accurately estimate dispersion properties across a range of frequencies. To capture adequate measurements for dispersion estimation, acoustic modes are measured using a variable position pressure transducer within a medium-filled wavetube. Such a method removes the necessity for multiple wavelengths between measurement locations as well as allowing for measurements to be taken at the lowest resonance frequency of the wavetube, enabling a compact test bench which can be utilized across a wider range of frequencies and mediums. By taking advantage of standing waves within the wavetube rather than avoiding them, a simpler testing method can be employed to extract clearer dispersion information across more frequencies when compared to traditional estimation techniques. By applying this data-driven dispersion estimation approach to hydraulic oil, an experimentally determined dispersion curve is generated which captures the non-linear characteristics of such a medium. This demonstrates not only the effectiveness of the technique to estimate dispersion, but also to characterize non-linear behavior. Such a method has the potential to fundamentally shift how non-solid media are simulated and utilized across a multitude of engineering disciplines.
Presenting Author: Matthew David Beals Michigan Technological University
Presenting Author Biography: Graduate Research Assistant at Michigan Technological University, working in the ViTALS Research Group within the Mechanical Engineering-Engineering Mechanics department.
Authors:
Matthew David BealsHrishikesh Gosavi
Vijaya v N Sriram Malladi
Data Driven 1d - Dispersion Estimation in Hydraulic Fluid Media
Paper Type
Technical Paper Publication