TECHNOTE 97-01.

Application of nickel as a magnetic flux calibration standard.

High-purity nickel is well-suited for use as a magnetic flux reference material. Since its coercivity is extremely low, it has a highly reproducible magnetic flux density at fairly low applied fields. The intrinsic saturation magnetization of approximately 6115 Gauss at 10000 Oersteds and +20ºC falls in a practical range near the remanence of hard ferrites. It has only about one-half the value of the remanence of Sm₂Co₁₇ or Nd-Fe-B, but this is much more useful than the principal alternative material iron, which saturates at about 21,580 Gauss. Most test systems can be calibrated to higher resolution without loss of accuracy using nickel rather than iron. Since nickel is chemically stable, it will not rust or corrode thus avoiding the attendant changes in magnetic characteristics.

Nickel certainly has its drawbacks. It is a very soft metal and thus requires reasonably careful handling. As long as it is not physically damaged, however, its properties do not change. If it is crushed or otherwise damaged, it can be ground or machined to a new shape and, if properly annealed for stress-relief, can again be certified to its original magnetic specifications. KJS Associates, Inc. offers this service.

There is some disagreement worldwide about the correct magnetization value for nickel. Nickel and cobalt occur together in nature and are difficult to separate. Small amounts of cobalt in nickel can cause the saturation magnetization to vary quite a bit. That is why high purity is necessary. The nickel provided by KJSA comes from a single bar of 99.995% pure nickel obtained from the INCO Corporation. Its chemistry was certified by INCO. Samples were then taken from several places in the bar and magnetically tested by the University of Dayton using a magnetometer calibrated against a nickel sphere obtained from NBS. The saturation magnetization is certified as uniform and NBS-traceable through this series of tests.

In a hysteresigraph system, flux density is measured using a search coil which surrounds the magnet sample near its neutral zone. The sensitivity of this search coil must be calibrated in some way.

One method is to calibrate the search coil area-turns at time of manufacture. The search coil is wound, potted using epoxy or other potting compound, and then its area-turns value is measured using a known magnetic field and a calibrated fluxmeter. That area-turns value is then attached to the search coil and used to calculate the sensitivity of the B channel in the hysteresigraph. This is very convenient because you do not need an additional calibration standard to use the coil. Although the method is theoretically correct, it is our experience that the area turns value of the coil changes with time due to handling, mechanical deformation, temperature and humidity changes, etc. The assigned area-turns value is then no longer correct.

A second method is to use a material reference standard to calibrate the search coil. This requires having a magnetic material with known and reproducible magnetic properties. It is possible to use a permanent magnet, but the reversible temperature dependence and long-term stability of PM's tend to cause problems. Nickel and iron are very suitable, since their magnetic saturation is stable. Because of their low coercivity, they are easy to saturate and this leads to reproducible behavior at low to moderate applied fields. Iron tends to rust when exposed to moisture and this changes the physical and magnetic characteristics. Nickel does not exhibit this behavior to any significant degree. To use nickel for calibrating a surrounding search coil, the nickel standard should have approximately the same size and shape as the magnet to be tested. The empty search coil is placed in a region of near-zero field and the integrator (B-channel fluxmeter) is turned on. Fluxmeter drift is adjusted at this time. After the fluxmeter is stable, it is reset. Then, the nickel standard is placed in the search coil and the search coil with nickel is placed in the gap of the system electromagnet. The gap is closed to produce a near-zero air gap at the interface between the electromagnet pole faces and the nickel standard. (About 0.001" or less of air gap is typically required for an accurate calibration.) Blocks or shims should be used to hold the electromagnet poles apart to prevent them from crushing the nickel. The required magnetizing field is applied to the nickel/search coil set, as stated in the nickel reference sheet. The gain of the B-channel electronics is adjusted so that the B-channel of the hysteresigraph indicates the stated magnetic induction of the nickel at that applied field and temperature. After the adjustment, the field is turned off, the nickel is removed from the search coil and the accumulated drift in the fluxmeter should be checked. If it is too high, the drift must be adjusted and the procedure repeated. When this calibration process can be repeated two or three times in a row without requiring any gain adjustments, the B-channel is considered calibrated. Any changes in the search coil are "calibrated out" by this method each time the search coil is used, since its sensitivity is corrected by the gain in the test system.

For embedded pole coils there are some additional concerns. A complete treatment of embedded pole coils is beyond the scope of this summary, but a few of the basic conditions that must be considered will be mentioned. Since the embedded search coils contain an iron or iron cobalt core, the exact magnetization state of the core is generally not well known at any given time. The residual flux in the core can be significant and this will affect the starting point ("zero") of the measurement. That is one reason why most embedded pole coils are used in a compensated arrangement, where the B and H coils are connected in series opposition to produce a (B-H) signal. Assuming that the B and H coils have the same area turns and that the residual flux in each of their cores is the same, this cancels out the residual flux and makes the system independent of the exact starting magnetization state of the poletips. For additional stability, the electromagnet poles can be dynamically demagnetized prior to test. (This is time consuming). Applied field H is usually measured with a Hall probe near the sample so that the H measurement is independent of the residual flux in the coil core. The remanent induction of nickel cannot be used as a calibration reference point. The calibration must again be carried out at specified field values, typically between 2000 and 4000 Oersteds. Often the use of a well-characterized permanent magnet is the best way to perform this calibration. The rest of the procedure is substantially as described in the previous paragraph.