Grainger College of Engineering, All Events

View Full Calendar

William W. Hay Railroad Engineering Seminar: John Popovics, University of Illinois Urbana-Champaign

Event Type
Seminar/Symposium
Sponsor
Rail Transportation and Engineering Center (RailTEC)
Virtual
wifi event
Date
Feb 5, 2021   12:30 pm  
Registration
Registration
Contact
RailTEC
E-Mail
RailTEC-Central@illinois.edu
Phone
217-300-1340
Views
27
Originating Calendar
RailTEC

Assessment of Rail Track Condition Using Novel Nondestructive Evaluation Technologies - Recently developed nondestructive evaluation (NDE) technologies provide opportunities for more sensitive and effective evaluation of track structure condition. Dr. John Popovics will present recent work from his research group to develop and apply new NDE technology for two different track condition issues. First, he will describe contactless non-destructive evaluation technology for fast and continuous condition assessment of concrete railroad ties. Ultrasonic surface waves are generated in railroad ties using an air-coupled ultrasonic transmitter and receiver set, where concrete railroad ties with different types of damage are evaluated. A set of signal parameters are used to generate two- and four-dimensional decision spaces computed using the support vector machine (SVM) algorithm. The results demonstrate that high quality signal data are obtained, and specific decision spaced can identify healthy and damaged railroad ties with an accuracy of around 80%. Next, he will describe in-progress work to evaluate the use of impulse vibration test data collected from continuous welded rail (CWR) to monitor stress state in the rail. In this study, vibration data are collected from an instrumented section on a revenue-service rail line, where true axial stress state and temperature are continuously monitored over the long term. Resonant vibration data are collected across different temperatures and stress states demonstrate that certain high-frequency modes of vibration are affected in different ways by temperature-induced thermal stress in the rail. Signal processing algorithms that identify, extract, and monitor pertinent vibrational modes within the rich experimental vibration spectra are described, and identifiable temperature and stress-induced behaviors are discussed. Approaches to invert the collected data to predict rail stress state are discussed.

link for robots only