By: Gary B. Howard, Chase P. Sturos and Theresa M. Ahlborn, Ph.D., P.E.
Infrared (IR) imaging has advantages over conventional techniques in the detection of bridge deck delaminations; for example, the process requires shorter inspection time, resulting in fewer or shorter lane closings, thus less disruption to traffic flow.
IR was established as a rapid, nondestructive detection method with the added capability to characterize individual bridge deck delaminations.
Technology developed in 1995 has been newly configured as a commercial bridge inspection tool by building on both recent developments in IR technology and internal supporting software tools. The resultant outcome, BridgeGuard, is able to capture deck conditions, readily analyze and store the data and allow future reference within its data management system. The rapid data collection eliminates lane closures and the labor-intensive effort common with current nondestructive methods. The simplicity, cost effectiveness and accuracy, along with the ability to analyze, manage and store this critical data, will be described in this two-part series.
IR gets the OK
Remote sensing is the science of measuring an object’s properties, which cannot be measured by traditional contact methods. It can further be defined as the collection of electromagnetic radiation, either emitted or reflected by the targeted scene, using a suitable receiver and data processing assets. A subset of remote sensing is IR imaging.
The science of IR technology can be applied to a variety of research and analytical activities. In fact, IR has been used successfully in a wide variety of industrial, commercial and environmental applications such as pipelines, roofs, electrical distribution systems, industrial facilities and, specific to this application, bridge decks. In fact, The American Society for Testing and Materials (ASTM) has developed a standard procedure for the bridge deck test process: ASTM Designation D4788-03—Standard Test Method for Detecting De-laminations in Bridge Decks Using Infrared Thermography.
Remote control
A thermal model of a concrete bridge deck can describe the normal flow of heat within a bridge deck. As this heat movement manifests itself at the surface of the bridge deck, it becomes a candidate for the application of IR as a remote-sensing modality. Using numerical techniques, a concrete slab can be modeled as a multilayered, semi-infinite solid in which time-dependent heat conduction occurs along a one-dimensional path. If a delamination were introduced to an otherwise homogeneous concrete slab, a disruption in thermal properties would occur at that local site. As a result, the normal flow of energy along the thermal path would be altered relative to its surroundings. This thermal disruption would eventually manifest itself at the slab surface and would be evidenced by a hot or cold spot in the IR imagery.
While the technology is weather dependent, time dependency is not limiting with the field use windows of opportunity encompassing most of a full diurnal cycle. The application windows can be segmented into two categories.
The first window begins a few hours after sundown. The daytime hours before the test should provide full irradiation by solar loading, and the night hours of the test should consist of a clear night sky. The homogeneous material will quickly begin to cool in response to the cooling effect of radiation exchange with the cold night sky, drawing heat from the lower layers of the concrete based on its uninterrupted thermal properties.
In contrast, the defect areas cannot draw this heat from the lower layers because of the thermal path disruption. The areas above the defect will respond much faster to this cooling effect. The normal areas will exhibit whiter (warmer) characteristics, and the defected areas will appear darker (cooler) in the thermal imagery. This condition will increase to a maximum point but then decrease throughout the night and into the morning hours until all surfaces drive toward thermal equilibrium.
The second window begins in the morning a few hours after the sun has risen. The night hours before the test should consist of a clear night sky, and the morning hours should be fully irradiated with solar energy. Full solar irradiation ensures a maximum transition from cold to warm in the upper concrete layers. Again exploiting the thermal transition into the concrete, we find that the homogeneous concrete warms much slower than defected areas where the discontinuity inhibits the ability to diffuse heat from the surface. Consequently, we look for whiter areas within the deck for anomalous indications.
Simulating results
As the 1995 system redesign process began, it was decided that a simulation of a bridge deck would be developed to complete the study. This simulator provides several uses, including:
- Validating and illustrating the theory;
- Developing an empirical prediction of what can be seen, when and how deep; and
- A training tool for future BridgeGuard users.
Typical concrete bridge decks in Michigan use a 9-in. structural deck and a 2-in. nonstructural concrete overlay. To quantify various thermal characteristics of bridge decks, a concrete slab with an appropriate framing system was designed and poured to replicate an actual bridge deck section. The design and pour were completed in conjunction with the Civil and Environmental Engineering Department at Michigan Technological University.
To simulate an elevated bridge deck, a slab was designed to mirror a typical highway bridge deck pour. The concrete consisted of an eight-sack mixture rated at 5,000 psi and was cast on I-beams. Additionally, five thermocouples were inserted resting at 2-in. intervals within the slab to quantify the depth and extent at which the thermal conditions at the surface are felt.
The concrete slab was designed with intentional damage at known locations through the placement of objects replicating delaminations within the slab.
Three 6- x 12-in. concrete cylinders were cast according to ASTM C31– Standard Method for Making and Curing Concrete Test Specimens in the Field. The cylinders were transferred to the Cement and Concrete Research Facility at Michigan Tech, an AASHTO, CCRL-accredited laboratory, and allowed to cure for 28 days. Compression strength tests were conducted according to ASTM C39–Standard Test Method for Compressive Strength of Cylinder Concrete Specimens. The average maximum compressive strength was 4,140 ksi. These test results, along with documentation from the ready-mix concrete supplier, indicates that this concrete is similar to what would be cast in a typical bridge deck.
By exposing the simulator to the atmospheric and meteorological conditions that are expected on a bridge deck, the goals for the simulator can be quickly achieved both for instantaneous validation of the thermal responses expected in a concrete deck and for the long-term goal of modeling the empirical results for future study. Therefore, the bridge deck simulator was subjected to actual measured meteorological conditions for two complete diurnal cycles. The first test occurring on Aug. 4, 2009, considered a cool summer day, and the second test occurring on Aug. 13, 2009, considered a warm summer day.
As the slab was exposed to the diurnal conditions, an IR camera was mounted to record the surface every 15 minutes throughout the 24-hour cycle. Rudimentary on-site meteorological characterization was made at each interval. The IR data was used as a qualitative measure of the slab’s thermal response over the cycle and as a quantitative measurement of the equivalent blackbody temperatures at the surface. As more diurnal data is gathered, it is the eventual goal that an empirical model be developed to act both as a training aid and as an aid for users to determine optimum times to test bridge decks using simple input variables.
Defect items 1, 2, 8 and 9 are apparent in both the daytime and nighttime imagery. Items 3 and 7 are beginning to appear in the imagery from both tests but never become an obvious defect.
Because defect item 10 is very thin, it is expected to act as a worst-case delamination indicator. It is considered worst-case simply because it was integral to the concrete pour, therefore the thermal discontinuity (or measurable break) expected at a true delamination site is not as delineated. Note that when reviewing the total data set, item 10 is detectable, even without further image processing.
To quantify the time of day that yielded the best contrast results, as determined by the measured contrasts at the defect locations, two methods were used. First, a visual evaluation of the contrast over the entire surface was conducted. The larger the visual difference, the greater the contrast score. Second, a simple statistical evaluation was carried out. Because defect item 1 provides the most obvious contrast, the mean temperature value in that region of interest was determined. The mean temperature of the background was generated in the same way. The percent mean contrast between item 1 and its background for each image was determined, and the results were plotted versus time of day.
Because the detection of delaminations within a concrete slab is based on heat transfer in and out of the slab, it is important to know how the environmental exposure at the surface affects the thermal conditions deep within the slab. Radiometric temperatures were recorded over the diurnal cycle for both tests. Note that the temperature plot begins at the onset of the diurnal test, which is not the same time for each test. The orange plot, indicated as the slab surface temperature, is the temperature reading from the thermal imagery. As expected, this temperature vector leads the thermocouple temperatures in response to environmental exposure.
For this technology to detect thermal discontinuities within a concrete slab there must be a thermal driver to generate a temperature gradient between the slab layers. This driver is provided by the sun and by the cold, cloud-free sky. The slab was raised off the ground to simulate the thermal boundary condition that exists on the underside of an actual bridge.
The question remains: Is there enough thermal gradient within the slab to see delaminations near the bottom of the slab?
If this is the case, a portable IR camera can be used under a bridge to detect delaminations at the deepest depth of the decking. As evidenced by the thermocouple readings, the thermal effect at the surface is well tracked at the deepest thermocouple position. An IR image taken with a handheld camera during the initial development of the system in 1995—with the thermocouple data as a further indicator—clearly indicates that these delaminations can be detected under the bridge.
No generic substitute
For IR to play a significant role in the detection, mapping and subsequent reporting of bridge deck delaminations, a system had to be developed to ensure that the IR sensor implementation is the best it can be and not just another adaptation of a generic system.
The background data presented in this first article provides the empirical justification for the claim that IR can play this significant role and, consequently, a total product development phase is fully merited. The next piece in this series will describe how this development proceeded and the results.
About The Author: Howard is director of engineering at Talon Research Inc., Hancock, Mich. Sturos is an engineer at Talon Research. Ahlborn is the director of the Center for Structural Durability at MTU.
