Magnet material oe -1 . BondcdNdFeb 34.0 ~ SinctcrdNdFce 1.44 The magnetic p r o p d e s of emerging rare-earth permanent magnets are beginning to exceed the limits for which their hysteresis loop can be measured to an acceptable accuracy in systems employing iron-yoked electromagnets, eg. the dc permcameter. However the most widely used alternative, which is capable of producing magnetic fields far in excess of those obtainable with iron-yoked systems, is the superconducting Vibrating sample magnetometer (VSM), which suffers from the drawbacks of having high capital and running costs and a limited pcak field capability. Therefon. there is a growing interest in developing a mom economic and flexible method based on the production of pulsed fields by the discharge of capacitm into a conventional resistive solenoid Although pulsed field magnetometry for material charactcrivltion is well established in basic materials research, when magnetic fields in excess of 4OT are now commonly employed, at present their use for the routine characterisation of rare-earth permanent magnets, which a~ produced in a wide range of geometries, is limited. One major consideration in applying pulsed field magnetometry is the effect of eddy currents which are induced, to a greater or lesser degree, in elecmcally conductive rare-earth magnet materials, particularly the more highly conducting sintered grades. In the majority of existing high field pulsed magnetomem the eddy currents are made negligible by employing small samples of material as well as a large stored energy in the capacitors so as to produce a long pulse time. However a general purpose measuring system must catcr for a wide range of magnet gtomemes, whilst the stored energy should be minimised so as to reduce the capital cost In order to establish the minimum pulse times which arc suitable for a given range of magnet gcomehies and grades of material, as well as to gain an insight into the modifying effect of eddy currents on the measured parameters, a transient simulation method has been developed. The simulation is based on a time-stepped, non-linear, axisymmemc finite element solution of the aansient diffusion quation which governs the magnetic field dismbution. Fig 1 shows an idealised model of a pulsed magnetometry system in which a prescribed uniform but time-varying flux density is imposed at the axial boundary surfaces. The resulting field dismbution is calculated f i s t with the electrical conductivity of the magnet set to zero and then to its experimentally determined value. Subsequent comparison of the two field solutions enables a detailed assessment of the effect of eddy currents to be made., for example on the flux linking the pick-up coils which encircle the magnet, from whose output the magnet parameters are derived. Fig 2 shows some representative results for a sintered NdFeB magnet when subjected to a transient sinusoidally time-varying magnetic field having an amplitude of 1OT. It shows that significant errors can arise if the signal from a pick-up coil is not corrected