A combined Q and heterodyne sensor incorporating real-time DSP for reinforcement imaging, corrosion detection and material characterisation

Abstract

An inductive sensor is described that is used to image steel reinforcement bars in concrete and to visualise surface corrosion. When a conductive target is placed within the vicinity of the sensing coil, an EMF is induced that opposes the change producing it. This causes a reduction in the Q-factor of the coil and a fall in amplitude of the excitation signal. When a non-conducting ferrous target such as corrosion product is present, a local increase in magnetic flux density occurs that in turn causes the inductance of the sensing coil to increase, reducing the resonant frequency of the tuned-system. The changes in Q-factor and resonant frequency are measured to image both the parent steel and surface corrosion. The circuit is designed to respond to very small changes in the received signal amplitude via a sensitive gain system, and to small changes in frequency, through the application of heterodyning. The excitation frequency has been selected so that the skin effect restricts the flow of eddy currents to the surface of the target. In this manner, surface features such as cracks and corrosion may be identified. The measured parameters are ultimately expressed as voltages before being applied to a purpose-designed DSP system. This averages signals in real-time, allowing high speed scanning. Feedback stabilisation is also applied between the DSP system and sensor to minimise voltage drift. At present it is possible to detect a steel bar to a depth of 200 mm and to image steel and surface corrosion to a depth 60 mm. Additionally, it is possible to image a range of materials for which non-destructive testing is important, such as graphite rods used in the nuclear industry. Since eddy currents are disrupted by vertical cracks, these features are readily detected by the system.