Applicability of Fatigue Crack Detection with Infrared Thermography Camera for Bridges in Denmark ^†

Abstract

This paper reports on the applicability of fatigue crack detection with an infrared thermography camera, the T-gap method, to a steel bridge in Denmark. The T-gap method is a non-destructive test developed in Japan and does not require approaching close to the bridge members, unlike visual inspections. The principle of the T-gap method is to measure the thermal profile of the welding point. One of the crucial factors generating the temperature gap is the solar altitude, and there is a smaller solar altitude in high-latitude areas, so it is unclear whether the T-gap method is more applicable in higher-latitude areas than Japan. Thus, the trial of the T-gap method in Denmark, which is located at a higher latitude than Japan, was planned to grasp the method’s applicability. As the results of trials, the T-gap method successfully detected both locations and the length of cracks even in Denmark.

1. Introduction

U-rib steel deck plates are widely adopted in structures such as long-span bridges in straits and urban expressways with stringent construction constraints due to their lightweight and short construction period. In recent years, several types of fatigue cracks caused by increased traffic of heavy vehicles have been reported in U-rib steel deck plates. Among these, cracks that propagate and penetrate from the root of the U-rib welded joint to the bead surface under wheel loads (hereinafter referred to as “bead penetration cracks”) have been frequently observed [1].

Furthermore, fatigue cracks can be mitigated by early detection, but early detection at stages where no paint film cracking or rust formation occurs is difficult with visual inspections, which are generally performed. Non-destructive testing methods, such as ultrasonic testing and eddy current testing, can be used to detect cracks before coating cracking occurs. However, these methods often require the installation and removal of scaffolding for close inspection, and considering the vast number of weld lines to be inspected, they are not efficient for inspecting the entire weld line [2].

Honshu-Shikoku Bridge Expressway Company (hereinafter referred to as “HSBE”) has researched and developed a method for detecting bead penetration cracks using infrared thermography-based temperature distribution measurements (hereinafter referred to as the “T-gap method”) (see Figure 1). The key feature of this method is that it can detect cracks non-destructively and from a distance at stages where no coating cracking has occurred. This method enables the measurement of the entire weld line by moving the camera while taking measurements [3,4].

The T-gap method is adopted practically by Honshu-Shikoku Bridges and other Japanese bridges; however, the method has never been applied to bridges outside Japan. A previous study showed that the detectable period is from March to November in Japan, with the period from March to September being preferable when the temperature difference between the deck plate and the U-rib is 0.7 °C or higher. However, the measurement period is based on results from bridge locations, Seto Inland Sea region, Japan [5]. When applying this to regions with different climates, the measurement period must be set based on data such as solar altitude. Then, a test of the T-gap method in Denmark, which is located at a higher latitude than Japan, was planned to grasp its applicability.

2. Test Object

The test of the T-gap method was carried out on the New Little Belt Bridge, which is a suspension bridge located in Denmark. Since 2017, cracks have been observed on the orthotropic deck by visual inspection. The bridge owner is the Danish Road Directorate.

The Little Belt Suspension Bridge carries the E20 motorway over the crossing between the island of Funen and the mainland of Jutland, connecting east and west Denmark. The bridge was opened in 1970, and it carries three lanes of traffic in each direction. The average daily traffic number is around 100,000 vehicles, which is close to maximum capacity. The suspension bridge has a main span of 600 m and two side spans of 240 m for a total length of 1080 m. The approach spans have a total length of 620 m, giving a total length of 1700 m for the entire connection.

The orthotropic steel deck structure with closed troughs acts as the upper flange in the closed steel box girder. The troughs have a height of 250 mm, a thickness of 6 mm, a width at the deck of 300 mm, and a spacing of 600 mm (Figure 2). Troughs are welded with a partial penetration butt weld with up to 2 mm root gap. The plated bulkheads have a spacing of typically 3.0 m, but the spacing can vary down to 1.0 m at site connections. There are R = 35 mm cope holes at the trough-to-deck weld.

During a general visual routine inspection in the summer of 2017, indications of fatigue cracks in the orthotropic steel deck were first observed, and non-destructive testing (hereinafter referred to as “NDT”) inspections were carried out later in 2017. All crack indications were in the trough-to-deck plate welds of No. 14 and 17, which correspond to the typical wheel positions of heavy vehicles in the slow lane (Figure 3). A total of 10 cracks were found. The cracks initiate at the weld root near the bulkhead cope hole and propagate through the weld bead to the face. Since 2017, a number of fatigue cracks, at approximately 160 locations, have been observed in orthotropic steel deck [6].

3. Method and Conditions

Considering the locations of detected cracks in 2017, No. 14 and 17 of the bead lines were tested by the T-gap method in this test. The test uses the infrared thermography camera, a tripod, and a PC, which has the measuring system. The tripod is moved at intervals of 500 mm in the longitudinal direction while taking still images via the thermography camera. Specifications of the employed infrared camera are shown in

Table 1. Specifications of infrared thermography.

Manufacturer’s Name Product Name	Type of Detector	Pixels	Spatial Resolution	Temperature Resolution
Teledyne FLIR Boson	Microbolometer	640 × 512 px	0.9 mrad	0.050 °C

. The test was carried out in July 2024. After the test, the detected cracks were confirmed and compared using magnetic particle testing (MT).

In order to detect cracks of approximately 40 mm, a temperature difference of more than 0.5 °C between the deck plate and the U-rib and 1 mm/Pixel shooting resolution are necessary [5]. Furthermore, a temperature difference of more than 0.7 °C is desirable. This means that March to September is desirable in Japan to obtain the temperature difference of 0.7 °C.

One of the crucial factors generating the temperature difference is the solar altitude. Among the road temperature, air temperature, and solar altitude, the solar altitude showed the highest correlation with the temperature difference of R2 = 0.983 (see Figure 4). Figure 5 shows the comparison of solar altitude in Japan and Denmark. Denmark has a lower solar altitude because Denmark is located in a higher-latitude area. Then, based on the desirable duration of March to September in Japan, May to August is desirable in Denmark.

4. Results

The test, three cracks were detected using the T-gap method. Results are shown in Figure 6 and

Table 2. Result comparison with the T-gap method and MT.

Method	T-Gap Method	MT
Case 1	48 mm	50 mm
Case 2	121 mm	120 mm
Case 3	73 mm	80 mm

. All detected cracks are located at the weld root near the bulkhead cope hole.

A more-than-0.5 °C temperature difference is necessary for the T-gap method, but the temperature difference was from about 0.15 to 0.33 °C in the test. Nevertheless, the T-gap method can accurately detect cracks, as confirmed by a comparison with MT results. This implies the possibility of changing the recommended value of the temperature gap. However, further testing and research are necessary.