National Physical Laboratory

AC Conductivity Standards

Introduction

AC standards set

Conductivity standards are required for the calibration of eddy-current conductivity meters widely used in the aerospace industry to determine the state of heat treatment and effects of work hardening of aluminium alloys. This is particularly important nowadays when parts of aircraft are regularly produced by different companies in different countries.

Conductivity is gaining increasing use in coin vending and coin handling and a sustainable infrastructure for the Euro currency therefore requires conductivity references values that not only have the required value but also the correct dimensions.

Although it is possible to calibrate conductivity meters using material calibrated by a DC method (and this is the procedure in the USA and generally on a global scale), the value so measured is an average for the total thickness of the material. It is difficult to prove that the conductivity of the surface layers, as measured by a conductivity meter operating at a typical frequency of 60 kHz, is the same as this average. For this reason it is necessary to carry out the calibration of reference standards at the working frequency.

NPL has recently piloted a European Framework V on conductivity measurements that included a rigorous investigation into the differences between AC and DC measured conductivities. The project also investigated methods for harmonising conductivity measurements and materials with improved properties for conductivity reference standards. Please contact NPL using the details below if you require further information.

If you have a conductivity measurement challenge, please contact NPL to discuss the details and the possibility of establishing a consultancy or small research project.

Transfer standards

In the NPL system, the AC conductivity of large annular rings enclosed by a toroidal winding is accurately determined using a mutual-inductance bridge. The conductivity of a ring is related to the observed resistance of the inductor and the bridge can be calibrated in terms of conventional electrical standards.

These annuli, however, are expensive, and simpler working standards are needed. They can conveniently be blocks 80 mm square and about 10 mm thick, whose conductivities can be compared directly at any desired frequency with those of annuli of similar conductivities. The comparison can be made either with a proprietary meter or with a solenoidal mutual inductor designed to stand on the surface to be measured and at a well-defined distance from it. This distance is critical, especially in the measurement of inductance. The mutual inductance and resistance can be measured with the same Heydweiller bridge used for the annular ring standards.

Boxed sets of AC conductivity standards are available. They comprise up to 9 square blocks of material, each approximately 80 mm x 80 mm x 10 mm, having nominal conductivities - 59, 35, 29, 22, 19, 14, 11, 9.6 and 2.2 megasiemens/metre (where 58.0 MS/m = 100.0% on the International Annealed Copper Scale, IACS).

Each set is supplied complete with a Certificate of Calibration accredited by UKAS quoting the conductivities measured at a frequency of 60 kHz. Measurements at other frequencies in the range 10 kHz to 100 kHz can be made on request.

Design of annular specimen

The metal annulus is made in two or more parts so that it can be assembled as a complete ring within permanent windings on a coil former, also made in two sections. It has flat surfaces on which the probe of a transfer instrument can be used. It should therefore have a rectangular cross-section. Experiments with commercial conductivity meters showed that consistent readings were obtained if no part of the probe was within one probe diameter of an edge of the specimen. The flat surface of the annulus has therefore been made 80 mm wide. The diameter must be such that the mean magnetic path length is not seriously in doubt as a result of averaging for various radii. Also there must be enough room for the winding (preferably as a single layer) on the inner diameter of the coil former. Hence a mean diameter of 300 mm was chosen, with outer diameter 380 mm and inner diameter 220 mm.

The thickness of the metal is such that at the lowest working frequency the effects of the flux penetration into the specimen remain calculable. This condition is always met by a very wide lamination, but not by one in which a significant part of the eddy-current path is influenced by the corners of the cross-section. This influence might extend laterally to a few times the skin depth, which at 10 kHz is 0.7 mm in copper, 0.9 mm in aluminium. The thickness chosen was 10 mm, which is about as large as is practical.

Design of the inductor and bridge network

The coil former, made in two halves, carries windings which do not have to be disturbed. The windings are in two sections on each half, each of 45 turns primary and 45 turns secondary, which can be connected in series or parallel to provide pairs of windings of 45, 90 or 180 turns; the 45-turn windings are suitable for frequencies from 10 kHz to about 100 kHz. Any eddy currents induced in the windings will be measured as if they were eddy currents in the specimen; for this reason, the windings of the two inductors must not merely be alike, but must be of very fine wire, and to keep the resistance small the wire must be multi-stranded.

Fig 1: The Heidweiller bridge with T-network for measuring small mutual resistances
Fig 1: The Heidweiller bridge with T-network
for measuring small mutual resistances

The mutual resistance of the windings, which is a measure of the eddy-current conductivity, is measured by a Heidweiller bridge (see Figure 1), which has the unusual property that both the oscillator and the detector have one side connected to earth. The connections between the inductors and the bridge network are made through coaxial cables.

The preferred method of calibration is to use the bridge to measure some known low resistance. The resistors constructed for this purpose are coaxial, each consisting of a short length of 0.525 mm Zeranin wire in a copper tube mounted on a coaxial plug. They are of approximately 0.1 W and 0.2 W , and there is also a short-circuited plug. The measurements are of the difference between either resistor and the short circuit. The resistors have been calibrated at 400 Hz on an NPL-designed bridge intended for resistance thermometry, itself calibrated to within ┬▒ 5 ppm. The resistances at frequencies up to 150 kHz increase with frequency because of skin effect in the copper of the outer conductor, but the skin effect in the wire itself is negligible. The temperature coefficient of the resistivity of Zeranin is about 3 ppm/K and therefore negligible in the present context.

Further Reading

  1. Lynch A C, Drake A E, and Dix C H, Measurement of eddy-current conductivity, Proc IEE A 1983, Vol 130 pp 254-260.
  2. Drake A E and Lynch A C, AC conductivity standards for the calibration of eddy-current conductivity meters, JPhysE:Sci Instruments 1987, Vol 20 pp 137-139.
  3. Gert Rietveld, Ch V Koijmans, L C A Henderson, M J Hall, S Harmon, P Warnecke and B Schumacher, DC Conductivity Measurement in the Van der Pauw Geometry, IEEE transactions on Instrumentation and Measurement, 52, pp 449-453, 2003.
  4. S A C Harmon, M J Hall, L C A Henderson, P P Munday, Calibration of commercial conductivity meters for measuring small items, IEE Proc-Sci. Meas. Technol., Vol 151, No 5, 2004.

Contact

Customer Service tel: +44 20 8943 7166
E-mail: electromagnetics_enquiries@npl.co.uk

Last Updated: 18 Aug 2015
Created: 11 Oct 2007

Registration

Please note that the information will not be divulged to third parties, or used without your permission

Login