Non destructive testing methods for integrity determination of concrete structures

Abstract: The ageing of concrete structures can be a threat to human safety as well as being a huge economical factor in the near future. Lifetime prediction and damage assessment of concrete structures will be a growing part of infrastructure planning. Thus the field of Non Destructive Testing (NDT) of concrete structures is an important part of the future in civil engineering. A substantial amount of Ground Penetrating Radar (GPR) data has been measured on concrete structures to define some of the limiting factors that prevent GPR from being used more extensively in connection with NDT of concrete. A Ramac CU2- GPR system with three sets of high frequency antennas has been used to collect data. The main difficulties found in concrete GPR surveys are: limiting depth penetration; large amount of spurious signals causing difficulties when interpreting the data; significant low frequency "noise" blurring the appearance. A concrete test block containing various reinforcement-meshes and cable ducts situated at different depths was constructed. The result of the initial GPR study on the test block showed that the resolution of an 800MHz antenna was poor and that there was a need for a better understanding of the ray paths involved. To explore the ray paths involved one of the GPR antennas was split into a separate transmitting, Tx and receiving, Rx antenna and the electronics were modified to transmit energy at a higher frequency. Initial measurements also suggested that the varying properties of different concrete structures have a significant influence on the response and thus complicate comparison between structures. To overcome this negative influence of the varying properties a simulation tank was built. The matrix used to simulate concrete was a mixture of rapeseed oil, water and table salt. During measurements in the simulation tank it was found that a surprisingly large portion of the energy in the radar wave is transmitted through the reinforcement mesh. The result suggests that ground penetrating radar should be more successful than measurements on real concrete structures has indicated. Part of this difference is likely to be caused by complex ray paths in the concrete that distorts the received data and makes it difficult to interpret. One of these ray paths was found when measuring on a single reinforcement bar, which produced a "ghost" hyperbola situated at the bottom interface of the simulation-tank. This ghost response could have the shape of a "double" hyperbola or a horizontally stretched hyperbola depending on the geometry and permittivity of the host medium. The simulation tank also revealed that the antenna used has two perpendicular polarisation directions. The additional polarisation will have a huge influence on the measured response in concrete measurements since concrete structures involves linier targets such as reinforcement bars and cable ducts. From the measurements done it is clear that there is a need to use different antenna configuration, particularly the in-line mode, as well as different antenna offset when measuring on reinforced concrete. Measurements on real world structures suggest that GPR can be successful but the technique should be applied with caution due to its sensitivity of host environment.