The decks of highway bridges under intensive traffic use are frequently exposed to attack by chlorides through the use of de-icing or marine salts. Deck surfacing layers are typically thin and joints are often the weak spot that allows access for the aggressive elements that can shorten the life of both the surfacing materials and the bridge structure.
Reinforcement closest to either the top or bottom surfaces of the concrete is usually the first to corrode, and once damage is under way the expansion this causes usually exacerbates the problem. The result is a pattern of delamination with large areas of loose concrete, further failure of the deck waterproofing and damage to key structural components such as post-tensioning cables.
Highway authorities in the USA have become increasingly aware of the scale of the problem during the 1990s, and current estimates suggest a possible US$3 trillion bill for repair to damaged concrete bridge decks. Similarly in the UK many of the major motorway structures have exceeded their design lives and now require major refurbishment. High profile repairs have been required on elevated sections of the M4 motorway in west London and the elevated motorway network in Birmingham.
Repairs are not only expensive but have wider implications such as delays and increased health and safety risks - so it is vital to effectively target resources. In the UK and USA plans for using advanced ground radar systems to map concrete defects before they reach the surface are already well developed.
The traditional solution has been to hammer tap the surface and detect areas that sound hollow, but this approach is expensive on the soffit due to the difficulty of access, and impossible on the deck without removal of the blacktop. Deck testing has usually been squeezed into repair programmes, meaning that results are provided too late to enable advance planning of repairs and only the shallowest, most obvious defects have been picked up.
Faced with these challenges and ever greater pressure to reduce down-time, improve sustainability and most significantly to deliver best value, bridge managers have sought alternative ways to prioritise resources. Ground radar systems have been used for some years to map delamination and voids on a bridge-by-bridge basis, but faster data processing and improved antennas now offer the prospect of regular inspection of whole highway networks.
Ground radar systems work by firing low power radio signals into a structure and analysing the velocity and strength of reflections from material boundaries and embedded targets such as reinforcement bars. Ground-penetrating radar is effective at detecting delamination or debonding between materials by mapping changes in reflection amplitude at a layer boundary (debonding), or by identifying planar reflections within an otherwise homogeneous material (delamination).
When applied to bridge decks the system can map laminar defects in the asphalt surfacing and within the reinforced concrete. As well as condition surveys, radar is used to map reinforcement, locate tensioning tendons and check layer thickness.
Analysis of the radar data requires careful consideration of factors such as potentially misleading signals from thin layers of material close to the defects, variations in moisture, and construction changes such as variations in reinforcement placement. These factors complicate automated processing. This technology in the UK has developed significantly over the last five years. Surveys of the Tyne Bridge conducted by Aperio in 1999 for Newcastle City Council Engineering Services were successful in mapping defects and demonstrated the value of the approach.
The last major maintenance works were carried out in the early 1980s, when the extent of defects was only discovered once the repairs were under way. The condition was found to be worse than expected, and as a result, the project ran over. This time around, GPR surveys were conducted six months before the repairs and the condition reports were sent to contractors bidding for the project, enabling much better estimation of materials, programme and costs.
A more recent project to benefit from advanced processing software was a trial survey of the deck of the Titford Viaduct section of the M5 Midlands Links in Birmingham, UK. The survey was conducted for WS Atkins on behalf of the Highways Agency and the objective was to evaluate the effectiveness of GPR in mapping delamination caused by expansive corrosion of the deck reinforcement.
The GPR data went through six stages of processing to remove enough of the unwanted signals from reinforcement and internal boundaries to resolve coherent areas of apparent separation between the rebars. The concentration of these features around the deck joints added weight to the interpretation that they represented delamination requiring treatment.
However, both these projects required temporary road closures because the speed at which data could be collected was little more than walking pace.
If bridge managers are to enjoy the full benefits of the approach there is a need for faster profiling that can operate without disrupting traffic. Now a new generation of radar systems that can collect data at up to 800 scans per second has made this possible. In a recent survey for one of the UK Highways Agency managing agents on a motorway in northern England Aperio completed what is thought to be a UK first by surveying six underbridges to map delamination at survey speeds of up to 110km/hour. The survey was completed in a few hours with no need for traffic management and the results were used alongside more traditional testing.
Meanwhile in the USA the problem of corrosion in concrete decks was highlighted by the Strategic Highway Research Program as long ago as 1993 and a group of 19 states is funding the development of a new ground penetrating radar system to address this problem.
The aim of the development team working on this new system, which has the acronym Hermes II, is to map bridge deck defects from a survey vehicle driven at traffic speed. Similar technology was integrated into the first generation Hermes I system, which was designed to use an array of 64 antennas set out in a staggered pattern to collect data over the entire width of one lane in a single pass.
Processing software in Hermes I and II can reconstruct data using a wavefield back-propagation algorithm to produce sub-surface tomographic images of a bridge deck in a short duration of time. These images remove much of the subjective interpretation that has limited the use of GPR in engineering applications. Defect mapping using the system is being studied with much anticipation. Working with defect maps produced using this type of imaging technology, it is anticipated that bridges across entire highway networks can be cost-effectively surveyed. Periodic coverage that determines the rate of deterioration and targets maintenance and repair spending may also be possible. The Hermes II system was designed as an exercise to improve the resolution of the original Hermes technology, for consistent defect detection. Results from the new system will be forthcoming.
