On the statistical mechanics of structural vibration

• Published in: Journal of sound and vibration Vol. 466; p. 115034
• Main Author: Langley, R.S.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Ltd 03.02.2020; Elsevier Science Ltd
• Subjects: Dynamic structural analysis; Energy flow; Entropy; Equations of motion; Granulation; Initial conditions; Mechanics; Nonlinear systems; Nonlinearity; Random systems; Statistical analysis; Statistical energy analysis; Statistical mechanics; Structural vibration; Vibration; Vibration analysis; Energy flow; Random systems; Entropy; Vibration; Statistical mechanics
• ISSN: 0022-460X; 1095-8568
• DOI: 10.1016/j.jsv.2019.115034

The analysis of the structural dynamics of a complex engineering structure has much in common with the subject of statistical mechanics. Both are concerned with the analysis of large systems in the presence of various sources of randomness, and both are concerned with the possibility of emergent laws that might be used to provide a simplified approach to the analysis of the system. The aim of the present work is to apply a number of the concepts of statistical mechanics to structural dynamic systems in order to provide new insights into the system behaviour under various conditions. The work is foundational, in that it is based on employing the fundamental equations of motion of the system in conjunction with various definitions of entropy, and no recourse is made to emergent laws that are accepted in thermodynamics. The analysis covers closed (undamped and unforced) and open (forced and damped) systems, linear and nonlinear systems, and both single systems and coupled systems. The fact that the system itself can be random leads to a number of results that differ from those found in classical statistical mechanics, where the initial conditions might be considered to be random but the Hamiltonian is taken to be well defined. For example, the occurrence of a stationary state in a closed system normally requires nonlinearity and coarse-graining of the statistical distribution, but neither condition is required for a random system. For coupled systems it is shown that under certain conditions both Statistical Energy Analysis (SEA) and Transient Statistical Energy Analysis (TSEA) are emergent laws, and insights are gained as to the validity of these laws. The analysis is supported by a number of numerical examples to illustrate key points.