Files in this item

FilesDescriptionFormat

application/pdf

application/pdfEVANI-THESIS-2017.pdf (15MB)
(no description provided)PDF

Description

Title:Air-coupled non-destructive evaluation technology for condition assessment of railroad ties
Author(s):Evani, Sai Kalyan
Advisor(s):Popovics, John S
Department / Program:Civil & Environmental Eng
Discipline:Civil Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:M.S.
Genre:Thesis
Subject(s):Non-destructive evaluation (NDE)
Railroad ties
Micro-Electro Mechanical sensors (MEMs)
Condition assessment
Abstract:This study describes the development of fully air-coupled non-destructive evaluation (NDE) technology for condition assessment of railroad ties. Ultrasonic surface waves are generated in railroad ties using an air-coupled ultrasonic sender. Surface waves are sensed at the receiving end using an array of Micro-Electro Mechanical sensors (MEMs). The resulting wave field is visualized in time-space (t-x domain) and frequency-wavenumber domains (f-k domain) to derive parameters indicative of the extent of damage in crossties. A robust “wave front fitting algorithm” is developed to estimate surface wave speed from a B-scan image generated using multiple time domain signals. Five signal parameters, namely surface wave speed from t-x domain, energy of the wave field in t-x domain, coherent surface wave speed from f-k domain, maximum value in f-k domain, and area under the k-plot at excitation frequency, are investigated for sensitivity to damage in crossties. Crossties with different types of damage; transverse cracking, longitudinal cracking, loss of cross section, and rail seat damage (RSD) are tested in this study. Signal parameters computed from the damaged ties are compared with signal parameters obtained from ties in a good structural condition (crossties without any visible damage) to generate decision spaces. The ability of these decision spaces to distinguish “good” crossties and “damaged” crossties is investigated using two parameters: hit-rate and reliability factor. The decision spaces are ranked based on these values to determine the most reliable decision spaces for predicting the structural state of a crosstie beyond a reasonable confidence level. It is observed that two-dimensional decision spaces performed better than one-dimensional decision spaces in predicting the structural state of a crosstie. The results demonstrate that the 2-D decision space of coherent wave speed from f-k domain vs. maximum value in f-k domain performed the best among all other decision spaces discussed in this study, followed by 2-D decision space of surface wave speed from f-k domain vs. coefficient of variation for surface wave speed from f-k domain, 2-D decision space of surface wave speed from t-x domain vs. energy in t-x domain, and 2-D decision space of surface wave speed from t-x domain vs. maximum value in f-k domain. A prediction scheme is developed for predicting the structural condition of a crosstie based on the location of signal parameters in various 2-D decision spaces.
Issue Date:2017-07-18
Type:Text
URI:http://hdl.handle.net/2142/99115
Rights Information:Copyright 2017 Sai Kalyan Evani
Date Available in IDEALS:2018-03-02
2020-03-03
Date Deposited:2017-08


This item appears in the following Collection(s)

Item Statistics