Electrical conductivity measurement of metal plates using broadband eddy-current and four-point methods

Abstract

Electrical conductivity of metal plates is measured by two distinct methods and the uncertainty associated with each method is evaluated. First, the impedance of an air-cored eddy-current coil is measured in the frequency range 100 Hz to 20 kHz. Corrections are made to account for the fact that the coil is not a pure inductor but exhibits finite resistance and capacitance in and between the windings. Then, the conductivity of brass and stainless steel plates is determined with 3 and 2% uncertainty (68% confidence level) by seeking the best fit (least-mean-square error) between experimental measurements of coil impedance and values calculated theoretically. The residual error in the fitting process is found to be the main indicator of uncertainty in the conductivity measurement. Second, four-point alternating current potential drop measurements are made on the same samples in the frequency range 1–100 Hz. Conductivity is determined from these measurements by means of a simple analytic formula, valid in a quasi-static regime, with an uncertainty approximately 0.5%. The main source of uncertainty in the four-point conductivity measurement is scatter in the voltage measurements. Both of these techniques give rise to smaller uncertainties in the measurement of conductivity than a MIZ-21A eddy-current instrument (2% and 40% for brass and stainless steel, respectively) and without the need for calibration specimens. In addition, the four-point approach is independent of magnetic permeability below a certain characteristic frequency and can be used to measure conductivity of ferrous metals. As an example, the conductivity of a spring steel plate is also determined.