NDT.net • Aug 2005 • Vol. 10 No.8

NDT Inspection of Railway Tracks Embedded in Concrete

Johannes Hugenschmidt*
EMPA, Überlandstrasse 129, CH-8600 Dübendorf
Carl N. Kasa, VAP, Suracherstrasse 6
CH-8142 Uitikon/Zürich

*Corresponding Author Contact:
Email: johannes.hugenschmidt@empa.ch, Internet: www.empa.ch
Published in Proceedings of RAILWAY ENGINEERING-2005, The Eighth International Conference
"Maintenance & Renewal of Permanent Way; Power & Signalling; Structures & Earthworks", Internet: www.railwayengineering.com


Information on the present condition of infrastructure is of particular interest within the framework of rehabilitation or in the event of damage. In order to obtain the required information an increasing number of non-destructive methods is available. Many of these methods are still in a phase of development, whereas the applicability of others has been shown for many different problems. Ground-Penetrating-Radar (GPR) has been applied successfully to a wide range of objects and applications, some of which are closely related to tracks embedded in concrete, such as the inspection of concrete bridge decks [1] or the inspection of railway ballast [2]. GPR is a method of interest because of its versatility and because of the fact that results are quasi-continuous. Moreover it has been proven to be economical for many applications.

The case-study presented here describes the application of GPR for the inspection of tracks embedded in concrete on an industrial site. The radar survey was carried out because traditional methods were unable to provide the information required for explaining causes of damage. In addition, the process of discussion between the building contractor and the client had come to a halt. In this pilot study with a very basic set-up for data acquisition, the combination of the radar survey with a very limited number of boreholes provided all the information required for the assessment of the problems on the whole site.

KEYWORDS: NDT, GPR, tracks embedded in concrete


Industrial railway tracks have to meet many requirements [3]. Crossing the tracks with lorries is becoming increasingly important.

The simple method of covering the railway ballast with asphalt pavement (figure 1) is in many cases not providing sufficient weight-bearing-capacity for vehicles crossing the rails. An alternative are tracks embedded in concrete (figure 2). In the last century, numerous systems have been developed. For industrial tracks, where trains are operated at low speed, simple solutions can be sufficient provided that.

  • Tracks are placed at the correct geometry and are fixed properly during concreting
  • Suitable materials are used
  • The concrete is laid on a properly prepared subbase

At first glance it seems that those requirements can be met easily. However, it has to be stated that corrections are very expensive after the concrete has been laid.

Figure 1: Railway ballast covered with asphalt pavement [3] Asphalt pavement Ballast

Figure 2: Tracks embedded in concrete [3] Pavement (asphalt or concrete) Concrete

As far as the track geometry is concerned, the following parameters have to meet required measures:

  • Track gauge
  • Level of track/ track distortion
  • Width of groove for tracks with guard rail or grooved rails
  • Longitudinal profile.

In order to repair defects efficiently and with long-lasting-effect their causes have to be known. On the track system described in this case study the rolling performance of vehicles was unacceptable. This led to various examinations and finally to the use of a track recording car. The gauges registered were far below the acceptable limits. In order to obtain further information on possible causes of the problems observed, a radar survey was carried out.


GPR is an electromagnetic investigation method. It is also known as Surface Penetrating Radar or Electromagnetic Reflection Method. Mostly it is used in reflection mode where a signal is emitted via an antenna into the structure under investigation. Reflected energy caused by changes in material properties is recorded (figure 3) and analysed.

Figure 3: Radar principles

A sketch of the emitted and recorded signal in wiggle and grey scale mode is presented in figure 4. The signal recorded is usually referred to as a scan or a trace. Please note that the vertical axis is a time axis. In order to obtain depths the signal velocities in the different materials under investigation have to be known. There are several different ways to obtain these velocities. In the examples presented below signal velocities were defined by a comparison between radar data and borehole information.

Figure 4: Comparison between emitted and recorded signals

Daniels [4] gives a detailed description of Ground Penetrating Radar, the application of the method for non-destructive testing in civil-engineering is described in an instruction leaflet published by the German Society for Non-Destructive Testing [5].


The radar survey was carried out by two persons in January 2004. It was completed within 4 hours. During this time, radar data were recorded along 2 lines on each of the 4 tracks as shown in figure 5. Each track was about 300 metres long leading to a survey length of about 2400 metres.

Figure 5: Position of data acquisition lines A and B

As this was a pilot study, the set-up for data acquisition was rather basic. A 400 MHz antenna (GSSI, model 5103) was placed in a sledge and pulled manually (figure 6). The radar unit (GSSI, SIR-20) and the rest of the equipment were placed in a vehicle moving at walking pace.

Figure 6: Data acquisition


The data analysis was carried out in two steps:

1. Preliminary analysis and decision on locations for boreholes
2. Final analysis using boring results

The boreholes were important as they allowed for a check of the preliminary results and a calibration of thicknesses. The locations of the two boreholes were defined with the help of the radar data. The position of an additional probing slit of a length of 5 metres was defined during a visual inspection on the site. The slit was used for a check of radar results and to obtain information that could not be provided by the radar survey such as the inspection of suspected cavities directly beneath the base of rails. This problem could not be addressed with GPR as the radar signal is unable to penetrate the steel of the rails.


The inspection of the concrete thickness aimed at answering two questions:

1. Is the required concrete thickness of 0.4m really existing?
2. Are there major variations in thickness?

The radar survey showed that the concrete thickness is around 0.4m everywhere on the site and that no major variations are existing. Figure 7 presents a section 21m long from a dataset recorded in the centre between the rails (position A). A borehole (vertical line) revealed a thickness of 0.4m. The concrete-subbase reflection is marked with an arrow and it can be seen over the whole length of the section presented and beyond. Thus it was possible to determine the concrete thickness over the whole length of the lines inspected. This led to the conclusion that the concrete thickness is around 0.4m everywhere and that there are only minor variations.

Figure 7:
Section from dataset (top), length=21m,
position A and result for concrete thickness (bottom)


The reflection at the concrete-subbase is caused by the contrast between the physical properties of the materials on both sides of the interface. Cavities can be considered an anomaly and can influence the reflection, in particular the reflection amplitude. However, it has to be stated that reflection amplitudes can be influenced by other parameters such as changes in humidity. The borehole position in figure 7 was chosen in a place with an increased reflection amplitude. In figure 8 an enlarged section of the dataset (length=10m) is shown. The arrow points at the borehole position. No cavity was found. The same is true for the second borehole which has been chosen based on the same principle. This led to the conclusion, that the radar data do not provide indications on cavities at the concrete-subbase interface.

Figure 8:
Detail from section presented in figure 7, length=10m (bottom)


The radar survey provided final and conclusive information on the presence and type of reinforcement. In figure 9 (left) a 4m long section from position A is presented. The reflections from single re-bars are clearly visible, some of them are marked with dots on the right hand side of figure 9. Thus the radar survey showed the existence of reinforcement on all the lines surveyed in a depth of around 0.38m and with a spacing between single bars varying between 0.15m and 0.20m. As expected, no information was obtained concerning the diameter of single bars as this would exceed the spatial resolution of the 400MHz antenna used for this survey.

Figure 9:
Detail from position A, length=4m (left) and same section with marked re-bars (right). (bottom)


Before the radar survey the type of sleepers and the distances in between them were unknown to the owner of the site.


Because of its limited spatial resolution, the 400 MHz antenna used for this survey is not the ideal instrument for the distinction between sleepers or spacing bars. The width W in figure 10 can be expected to be greater than 0.2m for sleepers, whereas the diameter D for a steel spacing bar can be assumed to be less than 0.05m. In figure 11 an enlarged section of the dataset acquired in position A is shown. Although it is not possible to decide conclusively based on the radar data alone, it seems most likely that the reflection is caused by a steel bar rather than a sleeper. On all other lines recorded in position A the situation was similar. As these results were presented to the building contractor, the company made clear that no continuous sleepers had been used on the whole site.

Figure 10: Continuous sleeper, double block sleeper and steel spacing bar without support


In figure 12, a 10m long section is presented. Six reflections resulting from the steel spacing bars (arrows) can be distinguished easily. The distance between the spacing bars is varying around 1.8m. This is true for all lines inspected.

Figure 11:
Detail, position A, length=1.5m

Figure 12:
Section from dataset, position A, length=10m

Figure 13:
Section from dataset, position B, length=10m

Figure 14:
Probing slit


The section presented in figure 13 is running parallel (position B) to the section shown in figure 12 (position A). The reflections caused by the steel spacing bars are visible in both datasets. No additional reflections caused by single blocks in between the known bars are visible. From that it was concluded that additional blocks do not exist. This is true for all lines inspected.


After the radar survey and the boring had been completed a probing slit was opened (figure 14). The distance between the three spacing bars were 1.8m and 1.95m. No single blocks were found in between them. Thus the probing slit confirmed the results obtained by the radar survey.


  • GPR is an effective method for the non-destructive, quasi-continuous and economic testing of tracks embedded in concrete.
  • The concrete thickness, type and depth of reinforcement and the distance between steel spacing bars were established with certainty. The existence of additional blocks not connected by steel spacing bars was ruled out conclusively.
  • The close cooperation and the open exchange of information between the responsible engineer and the radar specialist were important for the success of the radar survey.
  • The radar survey supported the process of discussion between the owner of the site and the building contractor.


    1. Hugenschmidt J.: "Accuracy and Reliability of radar results on bridge decks", Proc. 10 th International Conference on Ground Penetrating Radar, Delft/ NL, June 21-24, 2004
    2. Hugenschmidt J., "Railway track inspection using GPR", Journal of Applied Geophysics, 43, 2000, pp.147-155
    3. Carl N. Kasa, Frank Furrer: Industriegleise ein komplettes Vademekum, Verband Schweizerischer Anschlussgeleise- und Privatgüterwagenbesitzer, Uitikon/CH; 1995
    4. Daniels D. J.: Ground Penetrating Radar, The institution of Electrical Engineers, London/UK, 2004
    5. Deutsche Gesellschaft für Zerstörungsfreie Prüfung e.V. „Merkblatt über das Radarverfahren zur Zerstörungsfreien Prüfung im Bauwesen", Merkblatt B10, November 2001
  • © NDT.net |Top|