Durham School of Architectural Engineering and Construction


First Advisor

Professor Ece Erdogmus

Date of this Version


Document Type



A THESIS Presented to the Faculty of The Graduate College at the University of Nebraska In Partial Fulfillment of Requirements For the Degree of Master of Science, Major: Architectural Engineering, Under the Supervision of Professor Ece Erdogmus. Lincoln, Nebraska: December 2020

Copyright © 2020 Theresa Marie McCabe


This thesis presents utilization of Ground Penetrating Radar (GPR) for the assessment of highway pavements to gain a better understanding of this NDE method’s capabilities in terms of early detection of consolidation-caused air voids in brittle materials. The objectives are to first, determine the earliest time during the dormant, setting, and early hardening periods of concrete’s set that GPR can be used to detect shallow air voids. Secondly, the change in dielectric constant will be tracked over the first 24 hours of concrete setting in order to quantify the impact of the dielectric constant assumptions in early detection. Two different sets of experiments were conducted – the first to determine how early after casting GPR could be used to detect shallow air voids, and the second to track the change in dielectric constant over the first 24 hours of a specimen’s lifetime.

To accomplish the first objective, two reinforced concrete slabs were cast in the lab environment and two types of artificial air voids were implanted at different depths and locations. Type I voids were made of insulation foam sprayed into oblong and roughly circular shapes with dimensions of 2” x 1.75”, 2” x 2.5”, and 5” x 1”. Type II voids were created by injecting pressurized air at different depths and locations, creating air voids of unknown size and shape. Using GPR, the presence of these two types of air voids was detected as early as three hours after pouring; their locations and depths could be mapped out within an average error margin of 9%, 5%, and 19%, in the x, y, and z directions, respectively. Since this establishes a detection time earlier than any that has been previously reported in the literature, various post-processing and display techniques regarding color schemes and gains (other than those that are default and typically used, such as grayscale display) were tested and chosen to best identify the presence of air voids. The slabs cast and tested were cored to validate the GPR-based void locations. The work here presents the potential of GPR for early quality control of concrete pavements to minimize consolidation-related air voids and subsequent problems.

To accomplish the second objective, a third slab was cast without artificial voids implanted inside. A steel plate was laid at the base for calibration purposes, and the hourly change in dielectric constant was monitored and recorded for the first 24 hours after casting using GPR. It was found that the dielectric constant decreased at an average rate of 2.0% for the first 4.5 hours, 4.4% from hours 5.5 – 8.5, and leveled off between hours 10.5 and 13.5, before experiencing a drop and leveling off again at hour 15.5. Even though this concrete mix was different than the one used for the first two specimens, when the variation of dielectric constant is considered, the error margin previously established remains at less than 3%. The particular curve generated by the change in dielectric constant over time is unique to the specific type of concrete mix used in this experiment, however, similar curves can be constructed and for other mix designs. Should researchers or DOTs test specimens with mix designs they frequently utilize in highway pavement, they will have a growing archive of data and dielectric constant curves for reference should calibration of GPR equipment in the field be necessary. This would allow detection of subsurface voids in fresh and setting concrete pavement to be accomplished quickly and accurately, mitigating the need to calculate two-way-travel while in the field to determine an accurate dielectric constant for void detection and provide a quality control resource for reference both in further research and on construction sites.

Advisor: Ece Erdogmus