ABSTRACT

The Electric Potential Sensor (EPS) technology was invented at Sussex and is best understood by treating the sensor as if it were a ‘‘perfect’’ voltmeter. This would imply that it requires no input current for stable operation and that it exhibits an infinite input impedance. Clearly this is not attainable in a real sensor, however we have approached this ideal situation far more closely than has been possible previously. The EPS consists of an electrometer grade operational amplifier with an external bias circuit and associated feedback systems which have been described previously in the literature1. The net effect of this combination is to produce a broadband sensor (in the range 100 µHz to 1 GHz) with extremely high input impedance (∼1018) and low effective input capacitance (∼10−15 F). The EPS is capable of monitoring, via the displacement current, changes in potential due to currents flowing within the material even through an insulating surface layer. This makes it ideally suited to composite materials, such as the examples studied here, where the carbon mat is embedded in an insulating matrix. The end result is a sensor which is capable of measuring spatial electric potential via a weak capacitive coupling analogous to a magnetometer measuring spatial magnetic field. Due to the extremely weak interaction between adjacent sensors array formats may be built, which enables real-time imaging of potential distributions to be achieved. Usually a dielectric material is used to provide a stable mounting and a constant distance between the sensor and the sample. The system has proved capable of yielding local conductivity information for a range

of materials which are not well suited to eddy current techniques2 due to their poor electrical conductivity. As such we believe that this method will provide a complementary technique to those commonly in use for conducting materials. Some of these non-destructive testing methods, such as thermography3 and acoustic methods4, are being developed for use with composites. Damage resulting from impact or excessive loading will typically lead to delamination in these materials with a resulting reduction in the conductivity in the damaged regions of the sample. Monitoring conductivity via surface electrodes5,6, in the laboratory indicates that both reversible and irreversible changes occur in the local resistance of the samples, in response to applied loads. The irreversible changes remain measurable in the unloaded state and are associated with fibre fracture rather than elastic strain within the material. These results indicate that local conductivity will yield information about delamination in the unloaded state even where the small voids created when loaded have closed. This creates the possibility of detecting delamination when other methods such as ultrasonic and radiographic are unable to do so.