THROUGH ASPHALT DETECTION OF CORROSION DAMAGE IN CONCRETE USING GROUND-COUPLED GPR

Abstract

Many transportation departments (DOT) use Bridge Management Systems (BMS) to assist in planning, maintenance and repairs. BMS keep track of a number of aspects regarding a department’s bridge infrastructure that include location, data of construction, “original” drawings, and current state of the structure in addition to many other parameters. The most important of these is the current condition. This is generally assessed using visual inspection techniques. These inspection techniques are very subjective and lead to high contingencies for DOT’s and conservative bids from contractors. To assist in developing more accurate condition assessment of bridge deck deterioration, this thesis examines a number of methods that use a Ground Penetrating Radar (GPR) system with a single Ground Coupled Antenna (GCA) on bridge decks with an asphalt overlaid deck. Currently, when assessing bridge deck slabs with an asphalt overlay using GPR, air coupled antennas need to be used in order to determine the permittivity of the asphalt. If the permittivity of the asphalt is known then the permittivity of concrete can be determined using the reflection coefficient and then the losses due to geometric spreading can be calculated. A number of laboratory tests were conducted using asphalt and concrete samples to investigate the feasibility of determining the permittivity of the uppermost layer using a ground coupled antenna. Once the permittivity of the laboratory samples was accurately determined, field tests were conducted. The field testing consisted of assessing the conductivity of concrete for four unpaved bridge decks in Nova Scotia. The conductivity results were compared with a method being developed by another graduate student in the same research team. The field tests strongly indicated that the surface reflection for a GCA is unable to be used to accurately determine the permittivity of concrete. However one of the structures had a number of asphalt patches that produced reasonable permittivity and conductivity results. Field tests to confirm the accuracy of the method in determining the permittivity of asphalt using a GCA were unable to be done due to the lack of available test subjects; for this reason it was assumed from the laboratory results that a GCA is able to accurately determine the permittivity of asphalt. A number of methods were proposed for analysing bridges with an asphalt overlay; but due to the lack of ground truth data and limited sample size, success was assumed to be achieved if the estimated permittivity and conductivity value were within the generally accepted range. The two proposed methods tested were, a conductive model that assumed a constant signal velocity in concrete and a semi-conductive model that assumed the asphalt was non-conductive. The results from these tests indicated that asphalt may be treated as a non-conductive medium. The tests also raise the question of whether the asphalt/concrete interface reflection can be used to accurately determine the permittivity of concrete.