Uniaxial anisotropy by "radiomagnetic" treatment; controlling factors in a new process

A new process-an electron-"radiomagnetic" treatment-for obtaining high-remanence, low-coercive-force loops in magnetic alloys was recently announced. As an example, 2-MeV electron irradiation of 6-mil-thick ring laminations of polycrystalline 5-80 Mo Permalloy with 10¹⁷e/cm²in an applied circumferential magnetic field of 0.2 Oe atsim100degC produced record highs in remanence (∼6700 G) for this material. Additional studies of this process have been made to determine some of the controlling factors and the range of application. In particular, the effects of the dose (number of e/cm²) and of the preirradiation magnetic properties were examined. The results show that: 1) for a given dose of1.1 times 10^{17}2-MeV e/cm², the relative change in remanence (DeltaB_{r}/B_{r}) is always positive, ranging from 10 to 50 percent, but varies inversely with the preirradiation value of remanence (Br); 2) for the same dose, the relative change in coercive force (DeltaH_{c}/H_{c}) also depends upon the preirradiation value of remanence, but in a different way. ForB_{r} < 5000G,DeltaH_{c}/H_{c}is either negative or zero. ForB_{r} > 5000G,DeltaH_{c}/H_{c}is positive, ranging from 20 to 150 percent, and increases linearly withB_{r}; 3) if the dose is reduced tosim0.8 times 10^{17}e/cm², thenDeltaH_{c}/H_{c}is reduced to a tolerable level (∼10 percent) with no significant sacrifice in the positive gain in remanence and rectangularity. Hence, there are optimum dose ranges in the "radio-magnetic" treatments of alloys, where significant gains in remanence may be obtained without appreciable increases in coercive force.     

Models of elastic and inelastic flux switching

The totaldot{phi}(t)waveform of a square-loop magnetic core switched by MMFF(t)from negative remanencephi = -phi_{r}is composed of elasticdot{phi}_{epsilon}(t), decaying inelasticdot{phi}_{i}(t), and bell-shaped main inelasticdot{phi}_{ma}(t). The worst delta noise in a coincident-current memory core pair is essentiallydot{phi}_{i}(t). The sources of these components and the staticphi(F)curve are explained qualitatively by means of the random variations of the energy gradient vs. domain-wall position: elastic wall displacements and elastic rotation of magnetization inducedot{phi}_{epsilon}(t); minor inelastic wall displacements of essentially constant wall areas inducedot{phi}_{i}(t); and major inelastic wall displacements (involving domain collisions) of varying wall areas inducedot{phi}_{ma}(t). Semiempirical models for thedot{phi}components are based on the characteristics of these displacements. A satisfactory agreement is obtained between these models anddot{phi}(t)oscillograms of a thin ferrite core.     

Static and anhysteretic magnetic properties of tapes

The anhysteretic remanencebar{M}_{ar}(H_{o},T)of solidified suspensions of magnetic particles with predominant shape anisotropy is calculated from first principles for small dc fields Hoand arbitrary temperatureT < T_{B}(blocking temperature), describing the particle interactions by a mean field and assuming constant decrement of the ac field,2H_{d}per cycle. ForH_{d}< 2H_{o}, the anhysteretic distribution of particle magnetizations is found to be subject to the condition that the net internal dc fieldbar{H}_{i}is a minimum, and, for small Ho, to the condition,bar{H}_{i} = 0. The theory yieldsbar{M}_{ar}(H_{o},T)as a unique function of independently measurable static magnetic material properties, i.e., it contains no adjustable parameters and is hence quantitatively related to experimental data. Further, according to theory, ifbar{M}_{ar}(H_{o},T,T_{m})denotesbar{M}_{ar}as acquired in Hoat T and measured atT_{m}, bar{M}_{ar}(H_{o},T,T_{m} = T)is independent ofTforH_{d} ll 2H_{o}, andbar{M}_{ar}(H_{o},T,T_{m} neq T) = [M_{s}(T_{m})/M_{s}(T)] cdot bar{M}_{ar}(H_{o},T,T_{m} = T). The thermoremanent magnetization acquired in Hoand measured at a temperatureT_{m} ll T_{B},bar{M}_{thr}(H_{o},T_{m}), is related tobar{M}_{ar}(H_{o},T = T_{m}, T_{m})bybar{M}_{thr}(H_{o},T_{m}) = [M_{s}(T_{m})/M_{s}(T_{B})]bar{M}_{ar}(H_{o},T=T_{m},T_{m}), where TBis the blocking temperature below whichbar{M}_{thr}becomes thermally stable. Up to a constant factor of about 2, the theoretical results agree quantitatively with the experimental data on all materials that correspond to the premises of the theory, i.e., solidified suspensions, tapes in particular, of particles having predominant shape anisotropy.     

The effects of low temperature fatigue on the RRR and strength of pure aluminum

Low temperature fatigue effects on residual resistivity ratio (RRR = rho_{273 K}/rho_{4.2K}) and strength of 300 and 1000 RRR aluminum are reported. The objective of this investigation is to select the best initial purity for the stabilizer aluminum used in energy storage magnets. Monolythic centimeter diameter specimens were fatigued at 4.2 K to strains (ε) reaching 0.3 percent. The resistivity ratio rapidly decreases during the first 100 cycles and approaches saturation (RRRf) after about 1000 cycles for all strains tested. The RRRfvalues are different for different initial resistivity ratio (RRRi) values, but all tend to come together at 0.3% strain independent of RRRi. The maximum specimen stress (sigma_{max}) is reached after about 1000 cycles also, and approaches a common value (sigma_{max} = εE/2, where ε is the strain range and E the elastic modulus) independent of RRRi. Thus high purity aluminum becomes "fully hard" at equilibrium and behaves elastically. The impact of fatigue damage on conductor design and choice of stabilizer purity is considered.     

Measurements of superconducting Nb3Sn cavities in the GHz range

Superconducting Nb3Sn Cavities have potential advantages over rf cavities with Nb surfaces To test possible applications and to improve the understanding of Nb3Sn coatings on Nb, rf cavities have been measured between 1.5 and 8K and between 0.1 and 7GHz. The temperature dependence of the surface resistance R(T) indicates weak superconducting spots with transition temperaturesTmin{c}max{ast} < 1K andTmin{c}max{ast} simeq 2.5K. The normal conducting spotsTmin{c}max{ast} lsim 1K cause the large rf residual lossesR'_{res} propto f^{2}observed up to date. The spots withTmin_{c}max_{ast} simeq 2.5K cause temperature dependences ofR'(T)between 2 and 6K, where RBCS(Nb3Sn) is still negligible. In line withR_{res} propto f^{2}, the lowest rf lossesR_{res} < 2.10^{-9}Omegaand the highest field strengthB_{crit} = 83 m^{T}(wedgeE_{peak} = 29have been observed at the lowest frequency 0.1GHz measured. Surface resistance and penetration depth measurements have shown that grain boundaries or hydrogen clusters do not cause the weak spots observed withTmin{c}max{ast} < 2.5K. The origin and the chemistry of the weak spots withTmin{c}max{ast} lsim 1K, which cause the largeR_{res} propto f^{2}and the lowB_{crit} (T) simeq const, are still not clear. They seem related to the Nb3Sn surface. The weak spots withTmin{c}max{ast} simeq 2.5K consist most likely of Nb6Sn5, which in cooling below 950°C precipitates due to the excess Sn present in Nb3Sn coatings grown in Sn vapor.     

Effects of spacer layer elastic properties on metallic thin film edge-induced stress gradients in bubble memory devices

It is well-known that metallization edge-induced stresses can change the uniaxial magnetic anisotropy of a liquid phase epitaxial (LPE) garnet film near the metallization edge. We have investigated this magnetostrictive interaction of patterned metallic films with ion-implanted LPE films by using several different spacer layers such as polyimide, SiO2, Si3N4, and combinations of polyimide and SiO2beneath a Cr-Cu-Cr conductor pattern. It is concluded that the stress eliminating capability of a spacer depends on the hardness parameterK = frac{E_{s}(1-numin{f}max{2})}{E_{f}(1-numin{s}max{2})}whereE_{s} , E_{f}are Young's moduli andnu_{s}, nu_{f}are Poisson's ratios for the spacer and metallic film, respectively. The polyimide spacer withE_{s} < 10^{11}dyn/cm²and withK leq 0.1transmits an order of magnitude of smaller stress than a SiO2spacer withK geq 1with the stress being more uniformly distributed across the spacer.     

Effects of different rare earths on the temperature dependence of rare earth cobalt magnets

In many applications, it is necessary to have permanent magnets whose characteristics vary little around a temperature Tr. We have developed a calculation to describe the thermal variation of the magnetization of the rare earth (RE) in such magnets including the effects of crystalline field. The exchange field of cobalt has been deduced at all temperatures from magnetization measurements on a single crystal of YCo5. From the calculation we can deduce the concentration x0, which gives dM/dT = 0 around room temperature, the resulting magnetization and the correspondingd^{2}M/dT^{2}.     

A new principle for fast DC-sensitive potential-independent current comparators

A new principle of operation makes feasible small, cheap potential-independent current comparators whose performance reaches or exceeds the best properties of bulky magnetic amplifiers of the second-harmonic type. Connection of a ferrite-core coil in parallel with a suitable negative resistance generates relaxation oscillationsg(t). Inherently, at large oscillations the coil (i,phi)- characteristic is perfectly symmetric about the origin:phi(-i) = -phi(i). Provided that the negative-resistance (i,u)- characteristic has the same symmetry:i(-u) = -i(u), the oscillationsg(t)are perfectly symmetrical in the sense thatg(t+T/2) = -g(t). Heregrepresents either the coil flux φ, the currentior the voltageu. The time istandTthe oscillation period. External magnetomotive force shifts the coil (i,phi)-characteristics, invalidating (1) and therefore (3). Consequently, the deviation from 50% duty-ratio of the oscillatingq(t)is a sensitive and extremely stable measure of this MMF. Neither core temperature, pressure nor magnetic creep cause any zero drift. Orders of magnitude reached experimentally are: Short-time zero instability and noise:leq 10^{-5}ampere-turn, time resolution:leq 10^{-4}s, zero drift from -70 to +100°C:< 10^{-4}ampere-turn.     

The influence of applied tensile stress on power loss in Co-rich amorphous Co-Fe-Si-B ribbons with induced magnetic anisotropy

The influence on power loss PTof applied tensile stress σ in amorphous (Co0.89Fe0.11)72Mo3Si15B10(lambda_{s} > 0) and Co73Mo2Si15B10(lambda_{s} < 0) ribbons with different induced magnetic anisotropy Kuis reported. The losses are measured under sinusoidal flux conditions atf = 50Hz,J_{max} = 0.57T and atf = 400Hz,J_{max} = 0.10T. Measurements are carried out on samples in a stress-relieved state and with magnetic anisotropies induced by stress or field annealing. Atf = 50Hz, a minimum m Ptversus σ is observed. The σ-value (sigma_{min}) corresponding to the minimum PTincreases with increasing |Ku|. Atf = 400Hz, a minimum in PTversus σ is observed in the samples with induced magnetic anisotropy, whereassigma_{min} = 0in the stress-relieved samiales. However, no correlation between andsigma_{min}and Kuis possible from the present data.     

Magnetic properties of evaporated CoCr films

Perpendicular magnetic CoCr films were prepared on glass substrates using electron beam evaporation. The magnetic properties depend strongly on the substrate temperature, Ts. The anisotropy field,H_{k}^{eff}, and the perpendicular coercivity, Hc(perp) show maximum values at Tsaround 250°C. In this case, the alignment of c-axis is optimal and the grain size is smallest. The lattice spacing of the c-planes increases wlth Tsuntil 300°C. When the films prepared below 200°C were annealed around 300°C in high vacuum, the saturation magnetization, Ms, increased. The value of Ms, however, decrease by annealing above 400°C. In spite of the decrease of Msdue to the annealing above 400°C,H_{k}^{eff}and Hc(perp) increase by annealing above 400°C. If the films were bombarded by argon ions during film growth, Msdecreased and the internal stress changed from tensile to compressive. For low substrate temperature (below 150°C),H_{k}^{eff}decreased due to ion bombardment.     

Permanent magnet utilization criteria in a DC motor

In a dc motor with permanent magnet stator, the mean no-load working point of the magnet material can, after stabilization by stall current, be represented by a pointB_{m},H_{m}on a recoil line inside the demagnetization curve. The point results from the application during stall of an effective mean field Hadue to armature reaction in addition to the self-demagnetizing field due to circuit reluctance. It is shown that the motor specification and sizes lead to a specific value forH_{m}/H_{a}. The limiting values of this ratio (Hmzero or Hazero) imply that the corresponding optimum magnet designs for minimum magnet volume should be based either on maximum recoil energy or on(BH)_{max}. In practice, the best extreme working point during stall should lie between the points for these two criteria, dependent on the actualH_{m}/H_{a}. In some existing motorsH_{m}/H_{a}has been found to be between 0.5 and 1.5. For such values ofH_{m}/H_{a}, the variation of Bmand ofB_{m},H_{m}with working point is illustrated for high coercivity ferrite and for two grades of cast alnico alloy. A note on design methods is appended     

Polymer-Coated Fiber Bragg Grating Sensors for Simultaneous Monitoring of Soluble Analytes and Temperature

A new fiber-optic sensor for simultaneous measurement of water-soluble analytes and temperature with polymer-coated fiber Bragg gratings (FBGs) is proposed. As an application of the approach, simultaneous monitoring of the concentration of sugar or potassium chloride (KCl) and temperature has been achieved. Changes in these environmental parameters result in different extents of either red- or blue-shifts of the Bragg resonance wavelengths of the gratings. It has been found that polyimide-coated FBG responds to variations of both temperature and concentrations of soluble analytes, while acrylate-coated FBG is sensitive to environmental temperature only. The experimental results showed that the temperature sensitivity of the acrylate-coated FBG, temperature, sugar, and KCl concentration sensitivities of the polyimide-coated FBG are 0.0102 $~hbox{nm}/^{circ}hbox{C}$ , 0.0094 $hbox{nm}/^{circ}hbox{C}$, 0.0012 $hbox{nm}/^{circ}hbox{Bx}$ , and 0.0126 nm/M, respectively. The sensing mechanism of the polyimide-coated FBG lies in the hygroscopic properties of the polyimide coating, which result in the change of the strain of the fiber and, thus, the optical properties of the grating. Since the sensor detects the analytes that swell the polyimide coating and different analytes induce different swelling effects, the sensor can detect different analytes without prior knowledge once a calibration curve is developed.      

Helical anisotropic magnetic film

The magnetic coupling between the magnetization in two nonmagnetostrictive Ni-Fe layers separated by a SiO layer has been investigated by means of a transverse susceptibility measurement. The main results are that 1) the coupling energy Ec per Unit area of the multilayered film has a form ofE_{c}= -A cos (phi_{1}-phi_{2}), wherephi_{1}-phi_{2}is the angle between the magnetization vectors in the two Ni-Fe layers, and 2) the dependence of the coupling constant on the thickness b of the intermediate SiO layer can be interpreted quantitatively by the combination of the coupling energy due to Néel's topography model and that due to the magnetostatic interaction between the magnetic free poles appearing at the edges of the two Ni-Fe layers. The former coupling energy is given byE_{c1} = -frac{p}{2sqrt{2}}omega^{2}M^{2} exp(-sqrt{2}pb) cos (phi_{1}-phi_{2})wherep=2pi/LandLandware the wavelength and the amplitude of the undulation of the interface between Ni-Fe and SiO layers, respectively. The latter is given byE_{c2} = frac{2M^{2}D^{2}}{R} ln (frac{R}{D+b}) cos (phi_{1}-phi_{2})whereDis the thickness of each Ni-Fe layer, andRis the radius of the film.     

Computer simulation of the operation of a magnetic bubble domain propagation circuit modeled after a current-access device

The operation of a bubble-domain straight-line propagation circuit has been simulated successfully. This simulation has been achieved by our approximating the motion of an s = 0 frozen-azimuth bubble placed under a drive fieldH_{Z}(X, Y, T)= -H_{p} cdot cos [2pi(X/R_{X} - n(T)/4)] cdot exp [-(Y/R_{Y})^{2}]. The simulation has been generated from a previously developed numerical scheme to simulate the motion of a bubble, whose domain shape and magnetization structure along its domain wall were variable. The drive field has been modeled after a dual conductor-sheet, current-access propagation structure, which has a bit period RXand a transverse width on the order of2R_{Y}. The entire field contour has been advanced stepwise in the positiveXdirection by an increase of the integern(T), which represents the drive-phase number. The bubble motion has been observed during the first six drive phases to produce operating margin diagrams for drive frequencies of 250 KHz, 796 KHz and 1 MHz. The method of calculation and the results of the simulation are given.     

Factors affecting the sensitivity of gamma-level ring-core magnetometers

A theoretical and experimental analysis has been made of the geometrical and magnetic factors affecting the sensitivity of the ring-core magnetometer first introduced by Geyger in 1961. Using a drive-to-pickup transfer function derived from an assumed dynamic hysteresis loop and a given dc input signal, the second-harmonic output voltage was derived for both current-source and voltage-source driving conditions. Results of the analysis, which were confirmed by experiment, show the following. 1) When a single, diametrically wound (solenoid-type) pickup winding is used with a ring-core sensor, simultaneous minimization of both fundamental and second-harmonic feedthrough can be achieved by a simple rotational adjustment. This optimization is possible only with a circular core. 2) Sensitivity can be increased by increasing frequency, number of pickup turns, dynamic differential permeability, cross-sectional area, and effective sensor length, and by decreasing the dc initial permeability (μ0). 3) The dominant factor affecting the sensitivity is the demagnetizing termmu_{0}K = mu_{0}l^{-1.72} (4A/pi)^{0.86}, which determines the entrance attenuation of the dc signal to be measured. Highest sensitivity is obtained whenmu_{0}Kis small, which requires a low initial permeability, since a large dimensional ratiol/A^{1/2}is usually not attainable. When the demagnetizing term is not small, the sensitivity advantages of the other factors are negated by the input signal attenuation. Power considerations show that increasing sensitivity by raising area, length, or frequency results in increased power consumption. However, increasing sensitivity by using a pickup coil with a larger number of turns or by a sensor core with lower initial permeability can be achieved without additional power.     

Establishment of All Digital Closed-Loop Interferometric Fiber-Optic Gyroscope and Scale Factor Comparison for Open-Loop and All Digital Closed-Loop Configurations

This paper covers the design details of an all digital closed-loop interferometric fiber-optic gyroscope (ADCL-IFOG) prototype, constructed in TUBITAK UME, and scale factor comparison between open-loop and ADCL-IFOG prototypes with sine wave biasing modulation. The output of demodulation circuit, proportional to the applied rotation rate, was sampled by AD7714YN analog-to-digital converter (ADC), operated in 16 bit resolution. Error voltage, generated by microcomputer – controlled LTC 1667CG, 14 bit digital to analog converter (DAC), was sent to the phase modulator through a linear summing circuit to make Sagnac Phase Shift zero, depending on the rotation direction. For this implementation, the ultimate rotation rate of 1.84 ($^{circ}/{hbox{h}}$ ) was nullified. The averaged sensitivity of the proposed closed-loop IFOG in unit of error voltage applied to the phase modulator was determined as 132.65 $mu hbox{V}/(^{circ}/{hbox{h}}$ ). The scale factors of both the open-loop and ADCL-IFOG prototypes were compared in a range of 1–15270 ( $^{circ}/hbox{h}$) rotation rate, corresponding to Sagnac Phase Shifts varying from 0.00115 ( $^{circ}$) to 17.57448 ( $^{circ}$). The maximum peak to peak noise and the bias stability of ADCL-IFOG prototype were determined as 4.97 ($^{circ}/hbox{h}$ ) and 1.48 ($^{circ}/hbox{h}$ ) at 23.0$~^{circ}hbox{C}$ , respectively.      

Dielectric behaviour of a ferrofluid subjected to a uniform magnetic field

The electric susceptibility of samples of ferrofluids subjected to a uniform magnetic fieldHwas measured. The electric susceptibilitychiis dependent on the magnitude of the magnetic field and on the relative direction between the electric fieldEand the magnetic fieldH. 1) WhenEis perpendicular toH, frac{partialchi_{perp}}{partialH} < 02) WhenEis parallel toH, frac{partialchi_{parellel}}{partialH} > 0These results have been interpreted as a magneto-electric directive effect. A model is proposed, based on the assumption that the magnetic particles are roughly ellipsoidal and conducting grains.     

Limits of validity of the one-dimensional wall theory for bubble materials

It is shown that the one-dimensional studies of the wall surrounding a bubble domain do not violate some necessary self-consistency requirements. Moreover, it is shown that the ratio of the magnetostatic self energy (which is neglected in these studies) to the total one-dimensional wall energy isT/Q, whereTis of the order of 1 for typical film thickness of a typical bubble material. This justifies the use of the one-dimensional wall for these materials, as long as the quality factorQ = K/(2piMmin{s}max{2})is large.     

High field properties of ternary metal-molybdenum-sulfides

Measurements of the upper critical fields are presented for ternary metal-molybdenum-sulfides as a function of temperature, T, with dc magnetic fields up to ∼215kG and pulsed fields to ∼500kG. A nearly linear variation of HC2versus T is observed up to the highest dc fields for all the compounds. The materials (nominal composition), Tc(measured);(dH_{C2}/dT)_{T=T_{c}}(measured), and HC2(4.2 K) (calculated) for a dirty type II superconductor assuming no paramagnetic limiting are respectively: (a) Pb0.9Mo5.1S6: 11.7 K, 60 kG/K, 390 kG; (b) SnAl0.5Mo4.5S6: 14.2 K, 32 kG/K; 275 kG; (c) SnMo5S6:13.4K, 37kG/K; 290 kG; (d) PbMo5.1S6: 14.4, 60kG/K, 515kG. Pulsed field measurements yield HC2(4.2 K) = 390 kG for (a), 275 kG for (b), 290 kG for (c), and ≃510kG for (d). The PbMo5.1S6has the highest value of HC2measured to date.     

Magnetic properties of amorphous (fe, ni)80B20, (fe, ni)80B19si1, and (fe, ni)80P14B6alloys

Magnetic measurements were performed on amorphous (Fe, Ni)80B20, (Fe, Ni)80B19Si1, and (Fe, Ni)80P14B6alloy ribbons to yield the values of the magnetic moment per transition metal atom at 0 K and the Curie temperature. A close fit to the moment data, obtained by allowing not only the moments on Ni and Fe atoms to vary with Fe concentrationx, as revealed by neutron diffraction in crystalline alloys, but also the number of electrons per atom npand nbthat P and B atoms, respectively, donate to the transition metaldbands as described by the relationsn_{p} = 1.74 + 0.75 (1 - x)andn_{b} = 0.58 + 1.20 (1 - x), suggests a moment of 0 andsim0.13 mu_{B}per Ni atom in amorphous Ni80P14B6and Ni80B20alloys, respectively. The functional dependence of the Curie temperature onx, when followed closely for low Fe concentrations, gives the critical concentration xcabove which ferromagnetism appears in amorphous FexNi80-xB20and FexNi80-xP14B6alloy series asx_{c} cong 5and 8 at % Fe, respectively. The results of a theory based on coherent-potential approximation have been fitted to the measured values of Curie temperature in order to arrive at the Fe concentration dependence of the exchange interaction (JNiNi) between Ni-Ni pairs. The physical significance of the variation of JNiNiwithxobtained thereby has been discussed.