A producer of multifilamentary NbTi superconducting composites must depend on specified and uniformly reproduced production quantities of NbTi. Characterization of commercial NbTi starting diameter rod has sug-tested that impurity content, nonmetallic inclusion content and mechanical properties of the NbTi may also affect superconducting performance of multifilamentary wire. This paper describes the superconducting performance, i.e. critical current density at field of commercial NbTi containing 45 to 46.5 wt. pct. Ti and analyzes the influence of NbTi mechanical properties on superconducting performance. Production billets with Cu/NbTi ratios of 1.25:1 to 5.5:1 which contain as many as 2046 NbTi filaments are included in this survey. All billets were extruded and drawn to final size with constant major heat treatment and 90-97% reduction after major heat treatment. Superconducting critical current density is shown to be a function of total reduction of NbTi under these conditions, with values of 2.4 × 105A/cm2(4T), 2.0 × 105A/cm2(5T), and 1.6 × 105A/cm2achieved reduction ratios from starting rod size to final filament size of 105or better. Superconducting performance of NbTi heats with tensile reductions in area of 65.5% to 80.5% do not support the suggestion that more ductile NbTi produces higher current density filaments. Mechanical properties appear to be too simple a parameter to predict superconducting performance of NbTi.

A quantitative theory for the magnetostriction due to exchange interactions in metals has been developed, based on the ideas of Zener. The theory is in good agreement with the measured values for pure nickel and iron, and it explains the anomalous thermal expansion in the iron-nickel Invar alloys, as well as more complex alloys containing manganese and/or cobalt.

The temperature and magnetic field dependence of the thermal expansion and forced magnetostriction of one cold worked and three annealed Fe–Ni–Co alloy samples has been measured at low temperatures. Sample A (Fe‐34% Ni‐8% Co annealed) and sample D (Fe‐39% Ni‐10% Co cold worked) had large forced magnetostrictions and large negative thermal expansion coefficients at low temperatures, similar to the behaviour of Fe‐36% Ni invar while samples B (Fe‐39% Ni‐10% Co annealed) and C (Fe‐30% Ni‐18% Co annealed) did not. The effect of cold work on the Fe‐39% Ni‐10% Co alloy is dramatic, increasing the forced magnetostriction by a factor of 5, and producing a strong invar thermal expansion anomaly. The samples with the strong invar anomalies exhibited the rapid decrease of the forced magnetostriction below 20K and the time dependent drifts that have been noted previously for the Fe‐36% Ni alloy.1 It appears that the failure of conventional thermodynamic relations to correlate the magnetovolume effects of the Fe‐36% Ni invar below 20K also applies to these alloys. The significance of these results with regard to present approaches to the invar problem and the treatment of magnetic contributions to thermal expansion will be discussed.

The influence of cobalt substitution for nickel on the Invar characteristics of iron‐nickel alloys has been determined. For equivalent electron‐to‐atom concentration, N, cobalt additions were found to increase both the Curie temperature and inflection temperature. The inflection temperature exceeded the Curie temperature for nickel‐rich binary alloys and for ternary alloys. Cobalt additions increased the atomic magnetic moment for equivalent N with the maximum in the Slater‐Pauling curve shifted slightly to lower N as well. The value of ho(0) was also increased by Co additions, with the relationship of ho(0) vs N approximating the form of the Slater‐Pauling curve.

Zener's Invar theory is based on repulsion of parallel electronic spins, such repulsion being directly related to overlap of 3d shells.1,2Using an Fe‐Ni‐Mn alloy with Ni/Fe ratio of 3/2, the substitution of up to 20% cobalt with intermediate 3d shell diameter has been made for either iron with its larger 3d shell or nickel with its smaller 3d shell; manganese concentrations of 2.7 at. pct. were present in all alloys to minimize any low temperature expansion anomalies. Spontaneous linear magnetostriction, ho, was increased by substitution of cobalt for nickel; some of this increase is attributable to a change of the Fe‐Fe spin coupling from antiferro‐magnetic to ferromagnetic as evidenced by an increase in the average atomic magnetic moment, μa. When cobalt was substituted for iron, it was found that ho was decreased as long as the Fe‐Fe spin coupling was antiferromagnetic; with ferromagnetic Fe‐Fe spin coupling for highest Co concentration, however, ho increases slightly. These results are discussed relative to their good agreement with Zener's Invar theory.

The temperature and magnetic field dependence of the thermal expansion coefficient and the magnetostriction of commercial Kovar, an Fe–Ni–Co glass-to-metal sealing alloy, has been measured below 50 K and well into the magnetic saturation region. The measured properties of Kovar show anomalous behavior similar to that found in other Invar-type alloys, and are markedly different from the properties of a pure alloy of about the same composition.

Metallurgical efforts to develop a soft magnetic material suitable for application in the rotor of a generator or motor in advanced aerospace electric systems are reviewed. Commercial materials which have been considered include AISI 4340 steel, H-11 steel, Nivco alloy, and 15- and 18-percent Ni maraging steels. Developments described have led to several new materials with combination of good mechanical and magnetic properties at elevated temperature. Such materials include an improved maraging steel a precipitation hardenable cobalt-base alloy, a carbide strengthened Co-W alloy, dispersion-strengthened soft magnetic alloys, and unidirectionally solidified Co-Nb-Fe eutectic alloys.

Eutectic composites of Co−Nb and Co−Nb−Fe were produced by unidirectional solidification. The structure consisted of aligned lamellae of NbCo3 in a stabilized fcc cobalt-rich matrix. NbCo3 lowered saturation induction, following a simple volume law of mixtures. Niobium in solid solution lowered both the saturation induction and the Curie temperature, whereas iron in solid solution raised the saturation induction while lowering the Curie temperature. A minimum in coercive force (H app parallel to growth direction) occurred near 8 at. pct Fe, attributable to a minimum in |K 1| for this composition.

Occurrence of the Invar anomaly in thermal expansion characteristics of antiferromagnetic Fe–Mn–(Ni) alloys having constant Mn/Fe atomic ratios of ⅔ has been investigated. Linear spontaneous magnetostriction, Néel temperature, and inflection temperature were found to be well described by electron concentration; Ni substitution decreased all of these parameters. Such results are shown to be consistent with Zener's Invar theory based on competing exchange interactions.

Transfer impedance of a variety of soft magnetic materials (Ni, Fe, Ni–Fe and Fe–Si alloys) has been measured for the frequency range 0.01 to greater than 100 kHz. At frequencies above about 10 kHz, it is shown that ferromagnetic materials can have lower transfer impedance than copper. In magnetic materials, the occurrence of low impedance appears to be related to magnetic softness, whether attained through choice of composition or through annealing treatments.

The martensite ⇌ austenite transformations were investigated in Fe-Ni-Co alloys containing about 65 wt pct Fe and up to 15 wt pct Co. A change in morphology of martensite from plate-like to lath-type occurred with increasing cobalt content; this change in morphology correlates with the disappearance of the Invar anomaly in the austenite. The martensite-to-austenite reverse transformation differed depending on martensite morphology. Reversion of plate-like martensite was found to occur by simple disintegration of the martensite platelets. Reverse austenite formed from lath-type martensite was not retained when quenched from much aboveA s, with microcracks forming during theM→γ→M transformation.

An Fe–Ni–Co alloy containing acicular martensite has been partially reversed in the temperature region 525° to 725°C and the influence of the reversed microstructure on the coercive force-temperature (Hc-T) relationship has been determined up to 500°C. In all cases, an increase in Hc was observed at θγ, the Curie temperature of the austenite. For reverse annealing temperatures up to 590°C, Hc was increased at all temperatures; for higher reverse annealing temperatures, Hc decreased at all temperatures, but the increase in Hc at θγ was found to be more abrupt. These observations are correlated with the microstructure resulting from the shear-like martensite-to-austenite reverse transformation.

The reverse martensitic transformation was investigated in three Fe-Ni-Co alloys containing acicular martensite by means of dilatometry and coercive force (Hc) measurements. Two maxima were observed in the Hc -T relationships: the first Hc peak occurred at the Curie temperature of the austenite and is attributable to magnetic effects in a material containing two ferromagnetic phases; the second Hc peak, with values near 400 oe, is caused by reverse transformation wherein the remaining ferromagnetic martensite has increased coercivity due to decreased particle size and increased shape anisotropy. Dilatometric measurements display a gradual decrease in expansivity in the early stages of reversal. This observation is explained by a two-step process for shear-type transformation in which the first austenite formed has high magnetostriction and low expansivity. As the magnetostriction is lost by further heating, a large expansion occurs which counteracts contraction due to the crystallographic phase change.

Highest initial permeabilities are observed in 49% Ni–Fe alloy sheet which contains no strong deoxidizers. The adverse effect of strong deoxidizers such as silicon and aluminum is attributed to increased susceptibility to internal oxidation in a dry-hydrogen annealing atmosphere. For alloys containing no deoxidizers and a secondary recrystallized structure, μ40 values of 15 000 were observed by melting to an oxygen level of <5 ppm and annealing to effect removal of sulfur to a value of 1 ppm.

The origin of the large volume magnetostriction responsible for the Invar anomaly is discussed with respect to several recent theories. Magnetic and thermal expansivity measurements of Fe–Ni–Mn alloys with constant Ni/Fe ratio of 2/3 indicate qualitative agreement with Zener's theory based on competing exchange interactions. Coercive force (Hc) measurements of an Fe–Ni–Co alloy containing mixed martensite and austenite ferromagnetic phases display a maximum Hc at the Curie temperature (θγ) of the austenite which is known to display the Invar anomaly. The sharp change in the value of Hc at θγ is contrary to theories based on concentration fluctuation.

The saturation magnetization of fcc Fe–Ni–Mn alloys was measured by a force technique at temperatures from 77°K to the Curie temperature. Additions of Mn were found to increase the atomic moment only for alloys with Ni-to-Fe ratio of 4/1. In this series, a maximum in atomic moment was observed for Mn concentration between 2.5 and 5.0 at.%. For Ni-to-Fe ratios of 3/2 and 2/3, Mn additions decreased the atomic moment, such decrease being more severe as iron content is increased. The effect of Mn on the Curie temperature is similar, except that no increase was observed in alloys with Ni-to-Fe ratio of 4/1. Results are satisfactorily explained using a Heitler-London model for the 3d electrons in which the atomic moment associated with each atom is determined by the principle of maximum spin and the sign of the interaction between neighboring pairs.

Attainment of high initial permeability in commercial 49% Ni–Fe alloys is shown to be a function of the sulfur content and grain size. Low sulfur material was obtained by desulfurization during melting or by strip annealing. μ40 of the low sulfur material was significantly higher than that of normal commercial material at annealing temperatures below 1200°C. When annealed at 1200°C, μ40 for the initially low sulfur material was slightly lower because of a smaller grain size or a higher final sulfur level. Analysis of sulfur content of annealed material by the methylene-blue technique allowed a direct correlation between μ40 and sulfur content in secondary recrystallized structures. In general, μ40 values in excess of 10 000 cannot be obtained unless final sulfur content is less than 12 ppm. Highest initial permeabilities result from a coarse-grained secondary recrystallized structure and low impurity contents.