APPROXIMATE STOCHASTIC TECHNIQUES FOR DIVERSE ENGINEERING DYNAMICS APPLICATIONS



Gazis, N
(2018) APPROXIMATE STOCHASTIC TECHNIQUES FOR DIVERSE ENGINEERING DYNAMICS APPLICATIONS. PhD thesis, University of Liverpool.

[img] Text
201008254_Nov2018.pdf

Download (77MB)

Abstract

Generally, deterministic approaches are used in practice to analyze dynamic systems. Variations in loading conditions and material properties are taken into account by either selecting high, low or average values. Consequently, the uncertainty inherent in almost every dynamic analysis is considered just intuitively. To realistically capture the behavior of a dynamic system the intrinsic randomness must be appropriately modeled requiring concepts and methods of mathematical statistics and probability theory, as well as, random vibration theory. Undeniably, stochastic dynamics based approaches provide a more realistic modeling of the dynamic response of engineered systems allowing for enhanced design solutions. The prevailing approach used in the industry is the Monte Carlo simulation method. However, a well-known shortcoming of the method is the extensive computational cost required. Further, the class of problems of nonlinear random vibrations that lend themselves to exact solutions (e.g., via the associated Fokker-Planck-Kolmogorov equation) is extremely limited. Therefore, approximate approaches are desired for solving nonlinear stochastic dynamics problems. The current thesis seeks to exploit approximate stochastic dynamics tools to solve engineering dynamics problems encountered in practice. In particular, the primary focus is directed towards the recently developed Wiener path integral technique, which has been shown to poses certain advantages over alternative well-established solution methodologies, namely, computational efficiency and accuracy. Two applications are investigated: the stochastic response of nonlinear vibratory energy harvesters, and, the depth determination of ice gouging events. The accuracy/reliability of the approximate approaches is demonstrated via comparisons with pertinent Monte Carlo simulation data.

Item Type: Thesis (PhD)
Divisions: Fac of Science & Engineering > School of Engineering
Depositing User: Symplectic Admin
Date Deposited: 30 Nov 2018 15:04
Last Modified: 29 May 2019 07:20
URI: http://livrepository.liverpool.ac.uk/id/eprint/3028920
Repository Staff Access