Search NASA⌕ Search

SEARCH · Search NASA

Results for “MAGNET”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 631 records · Page 35

Short-Term Aging of NeFeB Magnets for Stirling Linear Alternator Applications

NeFeB type magnets have been proposed for use in free piston Stirling engine driven, linear alternators to generate electric power during long duration space missions. These type of materials provide the highest energy product commercial magnets, thus minimizing alternator size or mass, but do not provide the high temperature stability of magnetic properties found in the SmCo type magnets. Therefore, to apply the NeFeB type magnets at elevated temperatures to multiyear space missions, their long-term aging characteristics must be determined. This report presents 200 hr aging data for six types of NeFeB magnets selected from three manufacturers. Aging was performed under vacuum at 150 C, with a steady demagnetizing field of 5 kOe applied. From the data produced by this short-term aging run, candidate magnet types were selected for a planned 12,000 hr long-term run. Depending on the manufacturer's magnet type, remanence losses observed ranged from 0 to 7%, when measured at 120 C on an established recoil line. Also, intrinsic coercivity losses up to about 4% were observed for the M-H curve at 120 C. In some cases, these coercivity losses were not recoverable by recharge of the magnet, indicating a structural change of the material.

Niedra, Janis M.↗

Flux Compression Magnetic Nozzle

In pulsed fusion propulsion schemes in which the fusion energy creates a radially expanding plasma, a magnetic nozzle is required to redirect the radially diverging flow of the expanding fusion plasma into a rearward axial flow, thereby producing a forward axial impulse to the vehicle. In a highly electrically conducting plasma, the presence of a magnetic field B in the plasma creates a pressure B(exp 2)/2(mu) in the plasma, the magnetic pressure. A gradient in the magnetic pressure can be used to decelerate the plasma traveling in the direction of increasing magnetic field, or to accelerate a plasma from rest in the direction of decreasing magnetic pressure. In principle, ignoring dissipative processes, it is possible to design magnetic configurations to produce an 'elastic' deflection of a plasma beam. In particular, it is conceivable that, by an appropriate arrangement of a set of coils, a good approximation to a parabolic 'magnetic mirror' may be formed, such that a beam of charged particles emanating from the focal point of the parabolic mirror would be reflected by the mirror to travel axially away from the mirror. The degree to which this may be accomplished depends on the degree of control one has over the flux surface of the magnetic field, which changes as a result of its interaction with a moving plasma.

Thio, Y. C. Francis↗

Current-Produced Magnetic Field Effects on Current Collection

Current collection by an infinitely long, conducting cylinder in a magnetized plasma, taking into account the magnetic field of the collected current, is discussed. A region of closed magnetic surfaces disconnects the cylinder from infinity. Due to this, the collected current depends on the ratio between this region and the plasma sheath region and, under some conditions, current reduction arises. The current collection along a realistic "bare wire" space tether is considered. A number of factors are taken into account, including the resistance of the wire and shielding resulting from the current-induced magnetic field produced by current flow in the tether. The plasma density, tether length and radius, the geomagnetic field strength and angle to the orbital velocity vector were all used as parameters in the study. It is shown that magnetic shielding for certain tether system configurations, when combined with particular values of the governing parameters, significantly reduces the collected current. Specifically, it is shown that an electrodynamic tether in the "thruster" mode suffers greater reduction from magnetic shielding than a tether with the same characteristics deployed in the "generator" mode. We find that, for both modes, current-induced magnetic shielding becomes more significant as plasma density and wire radius increase. The same is true for the dependence on the angle of the geomagnetic field to the orbital velocity vector and the motion-induced electric field for the generator mode For the thruster mode, the effect is larger for smaller angles. In both operating modes, the shielding is more important for smaller angles between the tether and magnetic field. In addition to the above dependencies, the effect for the thruster mode essentially depends on the tether length. In general, any parametric change that increases tether current, relative to the strength of the electric field between the tether and the ambient plasma, will increase the shielding effect (thereby reducing overall tether current compared to the case without considering these effects). The effect of the induced magnetic shielding can be significant in some cases.

Khazanov, George V.↗

Operation of A Sunpower M87 Cryocooler In A Magnetic Field

The Alpha Magnetic Spectrometer-02 (AMS-02) is an experiment that will be flown as an attached payload on the International Space Station to detect dark matter and antimatter. It uses large superconducting magnets cooled with superfluid helium to bend the path of cosmic particles through a series of detectors, which then measure the mass, speed, charge, and direction of the particles. Four Sunpower M87N Stirling-cycle cryocoolers are used to extend the mission life by cooling the outer vapor-cooled shield of the dewar. The main magnet coils are separated by a distance of approximately 1 m and the coolers are located approximately 1.5 m from the center line of the magnet, where the field is as high as 925 gauss perpendicular to the cryocooler axis and 400 gauss along the cryocooler axis. Interactions between the applied magnetic field and the linear motor may result in additional forces and torques on the compressor piston. Motion of the compressor arid displacer pistons through the magnetic field spatial gradients will generate eddy currents. Additional eddy currents are created during magnet charge, discharge, and quench by the time-varying magnetic field. The results of tests to determine the magnitude of the forces, torques, and heating effects, as well as the need for additional magnetic shielding, are presented.

Breon, S. R.↗

Blood Pump Having a Magnetically Suspended Rotor

A blood pump preferably has a magnetically suspended rotor that rotates within a housing. The rotor may rotate about a stator disposed within the housing. Radial magnetic bearings may be defined within the stator and the rotor in order to suspend the rotor. The radial magnetic bearings may be passive magnetic bearings that include permanent magnets disposed within the stator and the rotor or active magnetic bearings. The pump may further include an axial magnetic bearing that may be either a passive or an active magnetic bearing. A motor that drives the rotor may be disposed within the housing in order to more easily dissipate heat generated by the motor. A primary flow path is defined between the rotor and the stator, and a secondary flow path is defined between the stator and the rotor. Preferably, a substantial majority of blood passes through the primary flow path. The secondary flow path is large enough so that it provides adequate flushing of the secondary flow path while being small enough to permit efficient operation of the radial magnet bearings across the secondary flow path.

James F Antaki↗

Blood Pump Having a Magnetically Suspended Rotor

A blood pump preferably has a magnetically suspended rotor that rotates within a housing. The rotor may rotate about a stator disposed within the housing. Radial magnetic bearings may be defined within the stator and the rotor in order to suspend the rotor. The radial magnetic bearings may be passive magnetic bearings that include permanent magnets disposed within the stator and the rotor or active magnetic bearings. The pump may further include an axial magnetic bearing that may be either a passive or an active magnetic bearing. A motor that drives the rotor may be disposed within the housing in order to more easily dissipate heat generated by the motor. A primary flow path is defined between the rotor and the stator, and a secondary flow path is defined between the stator and the rotor. Preferably, a substantial majority of blood passes through the primary flow path. The secondary flow path is large enough so that it provides adequate flushing of the secondary flow path while being small enough to permit efficient operation of the radial magnet bearings across the secondary flow path.

James F Antaki↗

Can MARSIS Measure the Low-Altitude Components of the Mars Magnetic Field?

Measuring the magnetic field anomaly of Mars at low altitudes (e.g. 100-200 km) can be an interesting application of Mars Advance Radar for Subsurface and Ionospheric Sounder (MARSIS). Due to a low HF operation frequency, the radio wave propagating in the ionosphere of Mars, over the magnetic anomaly regions, will be affected and distorted by the localized magnetic field. This distortion in the sounder signal is due to the Faraday rotation and provides information about the strength of the magnetic field. MARSIS is especially sensitive to the radial magnetic field at altitudes where the electron density in the ionosphere peaks (i.e. 100-200 km). Consequently, MARSIS is potentially capable of providing measurements for the radial component of the magnetic field at altitudes between 100 to 200 km that are normally out of reach for orbital magnetometers (with the exception of the aero-braking phase). Such low-altitude measurements would be complementary to already existing measurements at 400 km by MAG-ER on Mars Global Surveyor. This paper will explain the sensitivity of MARSIS as a magnetometer and the method envisioned to measure the radial magnetic field component. MARSIS (Picardi et al.), the first major planetary radar sounder, is the result of an international collaboration between NASA, the Italian Space Agency (ASI), and European Space Agency (ESA), and will arrive at Mars in early 2004 for a two-year mission. MARSIS has a frequency range between 0.1-5.5 MHz and is designed to penetrate the subsurface to a depth of a few kilometers. MARSIS primary objective is to map and characterize the subsurface geological structure of Mars, and search for subsurface liquid water reservoirs. The secondary objective of MARSIS is to study the ionosphere of Mars providing the most extensive amount of data on Martian ionosphere to date. In addition to MARSIS, a second radar sounder named SHARAD (SHallow RADar) with operation frequency of 15-25 MHz is under development. SHARAD is an Italian instrument (Seu et. al) that will fly on NASA s Mars Reconnaissance orbiter in 2005. SHARAD can also provide magnetic measurements, however, it is not expected to be as sensitive as MARSIS to magnetic field variations.

A Safaeinili↗

Near-Field Magnetic Dipole Moment Analysis

This paper describes the data analysis technique used for magnetic testing at the NASA Goddard Space Flight Center (GSFC). Excellent results have been obtained using this technique to convert a spacecraft s measured magnetic field data into its respective magnetic dipole moment model. The model is most accurate with the earth s geomagnetic field cancelled in a spherical region bounded by the measurement magnetometers with a minimum radius large enough to enclose the magnetic source. Considerably enhanced spacecraft magnetic testing is offered by using this technique in conjunction with a computer-controlled magnetic field measurement system. Such a system, with real-time magnetic field display capabilities, has been incorporated into other existing magnetic measurement facilities and is also used at remote locations where transport to a magnetics test facility is impractical.

Harris, Patrick K.↗

Solar Magnetic Explosions, Spicules, and the Heliosphere

We present an example of each of the following observed characteristics of the magnetic origins of quiet-region coronal heating, spicules, macrospicules, and coronal mass ejections (CMEs). (1) In quiet regions, the luminosity of the corona is roughly proportional to the edge length of the underlying photospheric magnetic network. (2) Spicules and EUV explosive events are concentrated at the edges of the magnetic network. (3) Many macrospicules have the magnetic structure of a surge rooted around an inclusion of opposite-polarity magnetic flux. (4) CMEs and eruptive flares are driven by explosions of sheared magnetic fields rooted along polarity dividing lines (neutral lines) in the photospheric magnetic flux. These characteristics together suggest that the mainstay of the heliosphere, the corona/solar wind rooted in quiet regions and coronal holes, may be driven by myriads of tiny magnetic explosions at the network edges, explosions like those that drive CMEs but of vastly smaller scale. If so, the steady solar wind and the CMEs that disrupt it both have the same root cause: explosions of initially-closed, strongly-sheared, bipolar magnetic fields. The photospheric vector magnetograms, chromospheric filtergrams, EUV spectra, and coronal images from Solar-B are expected to have sufficient sensitivity, spatial resolution, and cadence to test this scenario for coronal heating in quiet regions and coronal holes. This work was supported by NASA/OSS through its Solar and Heliospheric Physics SR and T Program and Sun-Earth Connection GI Program.

Moore, Ronald L.↗

An Overview of Magnetic Bearing Technology for Gas Turbine Engines

The idea of the magnetic bearing and its use in exotic applications has been conceptualized for many years, over a century, in fact. Patented, passive systems using permanent magnets date back over 150 years. More recently, scientists of the 1930s began investigating active systems using electromagnets for high-speed ultracentrifuges. However, passive magnetic bearings are physically unstable and active systems only provide proper stiffness and damping through sophisticated controllers and algorithms. This is precisely why, until the last decade, magnetic bearings did not become a practical alternative to rolling element bearings. Today, magnetic bearing technology has become viable because of advances in micro-processing controllers that allow for confident and robust active control. Further advances in the following areas: rotor and stator materials and designs which maximize flux, minimize energy losses, and minimize stress limitations; wire materials and coatings for high temperature operation; high-speed micro processing for advanced controller designs and extremely robust capabilities; back-up bearing technology for providing a viable touchdown surface; and precision sensor technology; have put magnetic bearings on the forefront of advanced, lubrication free support systems. This paper will discuss a specific joint program for the advancement of gas turbine engines and how it implies the vitality of magnetic bearings, a brief comparison between magnetic bearings and other bearing technologies in both their advantages and limitations, and an examination of foreseeable solutions to historically perceived limitations to magnetic bearing.

Clark, Daniel J.↗

The Effects of Differential Rotation on the Magnetic Structure of the Solar Corona: MHD Simulations

Coronal holes are magnetically open regions from which the solar wind streams. Magnetic reconnection has been invoked to reconcile the apparently rigid rotation of coronal holes with the differential rotation of magnetic flux in the photosphere. This mechanism might also be relevant to the formation of the slow solar wind, the properties of which seem to indicate an origin from the opening of closed magnetic field lines. We have developed a global MHD model to study the effect of differential rotation on the coronal magnetic field. Starting from a magnetic flux distribution similar to that of Wang et al., which consists of a bipolar magnetic region added to a background dipole field, we applied differential rotation over a period of 5 solar rotations. The evolution of the magnetic field and of the boundaries of coronal holes are in substantial agreement with the findings of Wang et al.. We identified examples of interchange reconnection and other changes of topology of the magnetic field. Possible consequences for the origin of the slow solar wind are also discussed.

Lionello, Roberto↗

Conical Magnetic Bearings Developed for Active Stall Control in Gas Turbine Engines

Active stall control is a current research area at the NASA Glenn Research Center that offers a great benefit in specific fuel consumption by allowing the gas turbine to operate beyond the onset of stall. Magnetic bearings are being investigated as a new method to perform active stall control. This enabling global aviation safety technology would result in improved fuel efficiency and decreased carbon dioxide emissions, as well as improve safety and reliability by eliminating oil-related delays and failures of engine components, which account for 40 percent of the commercial aircraft departure delays. Active stall control works by perturbing the flow in front of the compressor stage such that it cancels the pressure wave, which causes the compressor to go into stall. Radial magnetic bearings are able to whirl the shaft so that variations in blade tip leakage would flow upstream causing a perturbation wave that could cancel the rotating stall cell. Axial or thrust magnetic bearings cannot be used to cancel the surge mode in the compressor because they have a very low bandwidth and thus cannot modulate at a high enough frequency. Frequency response is limited because the thrust runner cannot be laminated. To improve the bandwidth of magnetic thrust bearings, researchers must use laminations to suppress the eddy currents. A conical magnetic bearing can be laminated, resulting in increased bandwidth in the axial direction. In addition, this design can produce both radial and thrust force in a single bearing, simplifying the installation. The proposed solution combines the radial and thrust bearing into one design that can be laminated--a conical magnetic bearing. The new conical magnetic bearing test rig, funded by a Glenn fiscal year 2002 Director's Discretionary Fund, was needed because none of the existing rigs has an axial degree of freedom. The rotor bearing configuration will simulate that of the main shaft on a gas turbine engine. One conical magnetic bearing replaces the ball bearing in front of the compressor, and the second replaces the roller bearing behind the burner. The rig was made operational to 10,000 rpm under Smart Efficient Components funding, and both position and current adaptive vibration control have been demonstrated. Upon program completion, recommendations will be made as to the efficacy of the conical magnetic bearing for active stall control.

Trudell, Jeffrey J.↗

Numerical Investigation of Plasma Flows in Magnetic Nozzles

Magnetic nozzles are used in many laboratory experiments in which plasma flows are to be confined, cooled. accelerated, or directed. At present, however, there is no generally accepted theoretical description that explains the phenomena of plasma detachment from an externally-imposed magnetic field. This is an important problem in the field of plasma propulsion, where the ionized gas must detach from the applied, solenoidal magnetic field to realize thrust production. In this paper we simulate a plasma flowing in the presence of an applied magnetic field using a multidimensional numerical simulation tool that includes theoretical models of the various dispersive and dissipative processes present in the plasma. This is an extension of the simulation tool employed in previous work by Sankaran et al. The new tool employs the same formulation of the governing equation set. but retains the axial and radial components of magnetic field and the azimuthal component of velocity that were neglected in other works. We aim to compare the computational results with the various proposed magnetic nozzle detachment theories to develop an understanding of the physical mechanisms that cause detachment. An applied magnetic field topology is obtained using a magnetostatic field solver and this field is superimposed on the time-dependent magnetic field induced in the plasma to provide a self-consistent field description. The applied magnetic field and model geometry match those found in experiments by Kurtki and Okada. We model this geometry because there ts a substantial amount of experimental data that can be compared to our computations, allowing for validation of the model. In addition, comparison of the simulation results with the experimentally obtained plasma parameters will provide insight into the mechanisms that lead to plasma detachment, revealing how the 3 scale with different input parameters.

Polzin, K. A.↗

The Maximum Free Magnetic Energy Allowed in a Solar Active Region

Two whole-active-region magnetic quantities that can be measured from a line-of-sight magnetogram are (sup L) WL(sub SG), a gauge of the total free energy in an active region's magnetic field, and sup L(sub theta), a measure of the active region's total magnetic flux. From these two quantities measured from 1865 SOHO/MDI magnetograms that tracked 44 sunspot active regions across the 0.5 R(sub Sun) central disk, together with each active region's observed production of CMEs, X flares, and M flares, Falconer et al (2009, ApJ, submitted) found that (1) active regions have a maximum attainable free magnetic energy that increases with the magnetic size (sup L) (sub theta) of the active region, (2) in (Log (sup L)WL(sub SG), Log(sup L) theta) space, CME/flare-productive active regions are concentrated in a straight-line main sequence along which the free magnetic energy is near its upper limit, and (3) X and M flares are restricted to large active regions. Here, from (a) these results, (b) the observation that even the greatest X flares produce at most only subtle changes in active region magnetograms, and (c) measurements from MSFC vector magnetograms and from MDI line-of-sight magnetograms showing that practically all sunspot active regions have nearly the same area-averaged magnetic field strength: =- theta/A approximately equal to 300 G, where theta is the active region's total photospheric flux of field stronger than 100 G and A is the area of that flux, we infer that (1) the maximum allowed ratio of an active region's free magnetic energy to its potential-field energy is ~1, and (2) any one CME/flare eruption releases no more than a small fraction (less than ~10%) of the active region's free magnetic energy. This work was funded by NASA's Heliophysics Division and NSF's Division of Atmospheric Sciences.

Moore, Ronald L.↗

Numerical Investigation of Plasma Flows in Magnetic Nozzles

Magnetic nozzles are used in many laboratory experiments in which plasma flows are to be confined, cooled, accelerated, or directed. At present, however, there is no generally accepted theoretical description that explains the phenomena of plasma expansion in and detachment from an externally-imposed magnetic field. The latter is an especially important problem in the field of plasma propulsion, where the ionized gas must detach from the applied, solenoidal magnetic field to realize thrust production. In this paper we simulate a plasma flowing in the presence of an applied magnetic field using a multidimensional numerical simulation tool that includes theoretical models of the various dispersive and dissipative processes present in the plasma. This is an extension of the simulation tool employed in previous work by Sankaran et al. The new tool employs the same formulation of the governing equation set, but retains the axial and radial components of magnetic field and the azimuthal component of velocity that were neglected. We aim to compare the computational results with the various proposed magnetic nozzle detachment theories to develop an understanding of the physical mechanisms that cause detachment. An applied magnetic field topology is obtained using a magnetostatic field solver, and this field is superimposed on the time-dependent magnetic field induced in the plasma to provide a self-consistent field description. The applied magnetic field and model geometry match those found in experiments by Kuriki and Okada. 4 A schematic showing the setup used in those experiments is shown. We model this geometry because there is a substantial amount of experimental data that can be compared to our computations, allowing for validation of the model. In addition, comparison of the simulation results with the experimentally obtained plasma parameters will provide insight into the mechanisms that lead to plasma detachment, revealing how they scale with different input parameters.

Sankaran, Kamesh↗

Magnetic-Field-Response Measurement-Acquisition System

A measurement-acquisition system uses magnetic fields to power sensors and to acquire measurements from sensors. The system alleviates many shortcomings of traditional measurement-acquisition systems, which include a finite number of measurement channels, weight penalty associated with wires, use limited to a single type of measurement, wire degradation due to wear or chemical decay, and the logistics needed to add new sensors. Eliminating wiring for acquiring measurements can alleviate potential hazards associated with wires, such as damaged wires becoming ignition sources due to arcing. The sensors are designed as electrically passive inductive-capacitive or passive inductive-capacitive-resistive circuits that produce magnetic-field-responses. One or more electrical parameters (inductance, capacitance, and resistance) of each sensor can be variable and corresponds to a measured physical state of interest. The magnetic-field- response attributes (frequency, amplitude, and bandwidth) of the inductor correspond to the states of physical properties for which each sensor measures. For each sensor, the measurement-acquisition system produces a series of increasing magnetic-field harmonics within a frequency range dedicated to that sensor. For each harmonic, an antenna electrically coupled to an oscillating current (the frequency of which is that of the harmonic) produces an oscillating magnetic field. Faraday induction via the harmonic magnetic fields produces an electromotive force and therefore a current in the sensor. Once electrically active, the sensor produces its own harmonic magnetic field as the inductor stores and releases magnetic energy. The antenna of the measurement- acquisition system is switched from a transmitting to a receiving mode to acquire the magnetic-field response of the sensor. The rectified amplitude of the received response is compared to previous responses to prior transmitted harmonics, to ascertain if the measurement system has detected a response inflection. The "transmit-receive-compare" of sequential harmonics is repeated until the inflection is identified. The harmonic producing the amplitude inflection is the sensor resonant frequency. Resonant frequency and response amplitude are stored and then correlated to calibration data.

Woodward, Stanley E.↗

Flux Transport and the Sun's Global Magnetic Field

The Sun s global magnetic field is produced and evolved through the emergence of magnetic flux in active regions and its transport across the solar surface by the axisymmetric differential rotation and meridional flow and the non-axisymmetric convective flows of granulation, supergranulation, and giant cell convection. Maps of the global magnetic field serve as the inner boundary condition for space weather. The photospheric magnetic field and its evolution determine the coronal and solar wind structures through which CMEs must propagate and in which solar energetic particles are accelerated and propagate. Producing magnetic maps which best represent the actual field configuration at any instant requires knowing the magnetic field over the observed hemisphere as well as knowing the flows that transport flux. From our Earth-based vantage point we only observe the front-side hemisphere and each pole is observable for only six months of the year at best. Models for the surface magnetic flux transport can be used to provide updates to the magnetic field configuration in those unseen regions. In this presentation I will describe successes and failures of surface flux transport and present new observations on the structure, the solar cycle variability, and the evolution of the flows involved in magnetic flux transport. I find that supergranules play the dominant role due to their strong flow velocities and long lifetimes. Flux is transported by differential rotation and meridional flow only to the extent that the supergranules participate in those two flows.

Hathaway, David H.↗

Toroidal-Core Microinductors Biased by Permanent Magnets

The designs of microscopic toroidal-core inductors in integrated circuits of DC-to-DC voltage converters would be modified, according to a proposal, by filling the gaps in the cores with permanent magnets that would apply bias fluxes (see figure). The magnitudes and polarities of the bias fluxes would be tailored to counteract the DC fluxes generated by the DC components of the currents in the inductor windings, such that it would be possible to either reduce the sizes of the cores or increase the AC components of the currents in the cores without incurring adverse effects. Reducing the sizes of the cores could save significant amounts of space on integrated circuits because relative to other integrated-circuit components, microinductors occupy large areas - of the order of a square millimeter each. An important consideration in the design of such an inductor is preventing magnetic saturation of the core at current levels up to the maximum anticipated operating current. The requirement to prevent saturation, as well as other requirements and constraints upon the design of the core are expressed by several equations based on the traditional magnetic-circuit approximation. The equations involve the core and gap dimensions and the magnetic-property parameters of the core and magnet materials. The equations show that, other things remaining equal, as the maximum current is increased, one must increase the size of the core to prevent the flux density from rising to the saturation level. By using a permanent bias flux to oppose the flux generated by the DC component of the current, one would reduce the net DC component of flux in the core, making it possible to reduce the core size needed to prevent the total flux density (sum of DC and AC components) from rising to the saturation level. Alternatively, one could take advantage of the reduction of the net DC component of flux by increasing the allowable AC component of flux and the corresponding AC component of current. In either case, permanent-magnet material and the slant (if any) and thickness of the gap must be chosen according to the equations to obtain the required bias flux. In modifying the design of the inductor, one must ensure that the inductance is not altered. The simplest way to preserve the original value of inductance would be to leave the gap dimensions unchanged and fill the gap with a permanent- magnet material that, fortuitously, would produce just the required bias flux. A more generally applicable alternative would be to partly fill either the original gap or a slightly enlarged gap with a suitable permanent-magnet material (thereby leaving a small residual gap) so that the reluctance of the resulting magnetic circuit would yield the desired inductance.

Lieneweg, Udo↗