Katelyn Knopf

Date of Award

Fall 2019

Document Type

Open Access Dissertation


Civil and Environmental Engineering

First Advisor

Dimitris Rizos


Continuous Welded Rail (CWR) has become the standard in modern railway track construction around the world because it alleviates well-documented disadvantages of rail joints in a track. CWR practice results in long segments of continuous rail in the track that will develop significant thermal elongation. To avoid the use of impractical large thermal expansion joints and limit the expected thermal elongation, the rail is anchored to the ties. Consequently, the rail is exposed to higher thermal stress demands as the rail temperature varies. At the time a CWR is laid, the rail is free of thermal stresses; the temperature at that time is known as the Rail Neutral Temperature (RNT). As temperatures deviate from the RNT, significant tensile or compressive thermal stresses are introduced longitudinally, leading to potential failures, including pull-apart and buckling that compromise the integrity of the track and the safety of train operation. Although the CWR installation procedures control RNT, it generally decreases over time. Since a decrease in RNT increases the risk of buckling due to moderate temperature increases, there is a need to determine the state of stress in the rail at different temperatures, as well as changes in the RNT.

Existing methods meet this need with various success levels but are not free of shortcomings. Disadvantages are related to the ease of implementation, system complexity, practicality, reliability, simplicity, cost, and instrumentation demands. All methods rely on data collected through means of contact with the rail. We propose a novel concept for measuring stress in rail and determining the RNT. The proposed technology is a non-contacting method that uses stereo-vision and three-dimensional (3D) Digital Image Correlation (StereoDIC) technology for full-field deformation and strain measurements. The StereoDIC system acquires rail deformations within a thermal cycle and processes the data in two steps. Initially, a reference free estimate of the RNT is obtained. Subsequently, the longitudinal stress in the rail is estimated as a function of the temperature. The StereoDIC hardware and software proposed is a technology developed at the University of South Carolina (USC). The method to process the StereoDIC measurements is the true innovation presented in this work.