224. NONLINEAR SEMI-ANALYTICAL FINITE ELEMENT ALGORITHM FOR INTERNAL RESONANCE ANALYSIS IN COMPLEX WAVEGUIDES - APPLICATION TO THERMAL STRESS AND BUCKLING DETECTION IN RAILS
Name: Claudio Nucera
Grad Year: 2012
Nonlinear guided wave propagation has received considerable attention in the last few decades because of the large sensitivity of nonlinear waves to structural condition (defects, quasi-static loads, instability conditions, etc...). However, the mathematical framework governing the nonlinear guided wave phenomena becomes extremely challenging in the case of waveguides that are complex in either materials (damping, anisotropy, heterogeneous, etc...) or geometry (multilayers, geometric periodicity, etc...). The present work develops predictions of nonlinear second-harmonic generation in complex waveguides by extending the classical Semi-Analytical Finite Element formulation to the nonlinear regime, and implementing it into a highly flexible, yet very powerful, commercial Finite Element code. Once formulated correctly, the proposed analysis can easily take into account damping effects, anisotropic multi-layered properties, periodic geometries and other complex waveguide properties in a computational efficient and accurate manner. As a final result optimum combinations of primary wave modes and resonant double-harmonic nonlinear wave modes are identified. Knowledge of such combinations is critical to the implementation of structural monitoring systems for these structures based on higher-harmonic wave generation. A Continuous Welded Rail track (CWR) was selected as case-study to benchmark the potential of the aforementioned analysis. In these structures the almost total absence of expansion joints and consequent rail thermal stresses can result in derailments in hot weather (sun-kinks) and breakage in cold weather (pull-apart). These reasons emphasize the importance of collecting periodically reliable information on the level of stress acting in the rail to ensure safety of operation in CWR. An innovative technique is under study at UCSD. It correlates the level of thermal stress acting in the rail and the eventual incipient buckling condition with some peculiar nonlinear features of guided wave propagation. It is apparent how in this framework the proposed nonlinear algorithm is of paramount importance to discover optimal combinations of resonant primary and secondary modes able to maximize the nonlinear response and, consequently, the efficiency of the proposed technique. The successful identification of these modes is discussed in detail. Later this crucial information guided the execution of a series of unique experimental tests on a 70'-long full rail track aimed at corroborating the theoretical/numerical predictions and designing a wayside prototype for Neutral Temperature/Incipient Buckling detection in CWR rails.