The unfamiliarity of metallurgists with the materials that in 1942 suddenly became in demand for atomic energy purposes was not due so much to their rarity as to their general metallurgic cussedness. Uranium, thorium, and beryllium are about as abundant in the earth's crust as are copper and lead, but they are less well concentrated by geological processes and are extremely difficult to reduce to a pure metallic form. If they were not so reactive with oxygen and nitrogen, they would have been made at least a generation ago and thoroughly studied. Had their oxides been reducible with carbon, they would have been known centuries ago. The metallurgist still seems to limit his interest to those materials that are easily prepared and avoids even the abundant ones which have desirable propeties if they need unusual means of preparation. Metallic uranium is highly reactive, though when exposed to air, in clean massive form, it quickly develops a nearly black oxide coating that prevents continuous attack except at somewhat elevated temperatures. As a powder, uranium is highly pyrophoric, particularly in moist air, and in water it corrodes rapidly. For any severe service (as in air-cooled or water-cooled piles) it has to be clad or canned with a completely protective material. Uranium metal, even in solid form, reacts avidly with hydrogen to give an extremely fine gray powder of the hydride, UH₃. This reaction is rapid and reversible, the hydride having a well-marked dissociation pressure at any temperature (760 mm pressure at 430° C and 1 mm at 212° C). The hydride, when slightly heated, provides an excellent source of high-purity hydrogen, which is evolved to restore the equilibrium pressure whenever gas is removed from the system, and of uranium powder which remains after the hydrogen has left.

Because of its strongly electropositive character, uranium is difficult to reduce to the metallic form, and once prepared is difficult to keep pure. The reduction process cannot be used as in common metallurgical practice to separate impurities in the ore, and some compound has first to be prepared in a sufficiently pure form by chemical methods. Though uranium can be made electrolytically, most metal to date has been prepared by various modifications of the historic method of Peligot who, in 1840, first prepared the metal by reducing the tetrachloride with sodium at a red heat. Reductants that have been named include calcium and magnesium and other halides have been substituted for the chloride. The process is similar in principle to the well-known Goldschmidt process. Refractory problems are severe, but with proper control the method can give almost quantitative recovery of metal.

The reduced metal is generally of slightly lower purity than the reactants used, and needs to be remelted, preferably in vacuum, to remove the last traces of the reductant and (as with any other metal) to cast it either into billets for subsequent mechanical working or into appropriate molds to give directly objects of the final desired shape. Though uranium can be melted under a heavy flux cover, it is sim­pler to melt and cast it under a fairly good vacuum. This operation is quite straightforward if properly designed apparatus is used, with pumps of large enough capacity. As with all metals, the experienced skill of the founder is essential for good results, but granted this and proper mold materials that do not liberate gas, sound and strong castings of almost any desired shape can be made. The metal machines well with tungsten carbide-tipped tools and copious coolant. Machining is a completely safe operation despite the occasional spectacular but nearly smokeless fires that occur if the turnings get hot enough to ignite.

Though uranium is not very plastic at room temperatures, it does have reasonable ductility and can be shaped to a limited extent. For extensive working, it is necessary to heat it, either high within the alpha range or into the gamma range where it has extreme plasticity, comparable with beta brass but even softer. Deformed grains (particu­larly those of certain critical orientations in relation to the deforma­tion) will recover by "polygonization" before recrystalization starts; hence it is difficult to produce very fine grain size. Such sub-grains are formed also on transformation, and the large, grains that at first glance appear in the microstructures of castings will often be found on closer examination to consist of many sub-grains differing slightly in orientation.

Uranium metal can be fabricated by most of the usual metal-shaping methods. It has been forged, made into rod by both rolling and extrusion, into wire by drawing, and into sheet by rolling. Sheet can be cupped once its peculiarities have been mastered. Its welding is difficult.

Uranium exists in three different allotropic modifications, with transi­tions at about 668° and 774° C. It melts at 1,133°C. The alpha phase, stable below 668° C, has an orthorhombic unit cell. It has a layer structure composed of corrugated sheets of atoms, the distance between atoms in the sheets being much less than the distance between sheets, in a similar manner to the structure of antimony, arsenic, and bismuth. Uranium has high electrical resistivity (about 60 microhm-cm., some six times that of iron) which would be expected in a material with so much covalent bonding. It is highly anisotropic in its properties. In the c direction, its thermal expansion coefficient is actually negative (-1.4 x 10^-6 per deg C), while transverse to this direction it is larger than the coefficient of any common metal except the alkalies (28 and 22 x 10^-6 per deg C for the a and b axes). Obviously, therefore, the measured value for the thermal expansion of a sample of polycrystalline metal will be greatly dependent on its degree of preferred orientation. Almost all the chemical and physical properties of uranium are similarly strongly anisotropic. This directionality is generally troublesome, but it could be useful as it gives metallurgists an additional variable to utilize in designing materials with a specific combination of properties. There is need for imaginative work on anisotropic materials generally, and both fundamental and applied research is certain to be rewarded.

The beta phase, existing between 668° and 774° C, is of complex structure that has only recently been fully determined. It has the same structure as that annoying brittle low-temperature phase in the iron-chromium system known as sigma, and, like its counterpart, is brittle. The gamma phase, existing between 774° C and the melting point, is the only phase to have a really normal metallic structure — it is body-centered cubic — and high ductility.

Though the two high-temperature phases cannot be retained at room temperature on quenching, it is possible by appropriate alloying and quenching to retain them both and to study their structure. There are indications that a uranium alloy with enough molybdenum may be stable gamma at room temperature, and will provide an alloy which, being body-centered cubic in structure is both stronger and more ductile than pure uranium as well as being more nearly isotropic and free from the complications due to transformation. Unfortunately, many nuclear applications require as little dilution as possible, and there remains a large field for the development of new alloys to meet the exacting tolerances.

Pure uranium, when wrought and annealed, has a tensile strength of about 80,000 psi and an elongation in the vicinity of 15 pct. Castings are somewhat less attractive because of their large grain size. Strains due to thermal anisotropy result in a low proportional limit. Because of the extremely high density (19.0 grams per cubic centimeter) large pieces must be carefully supported if they are not to distort locally under their own weight. The strength is greatly improved by cold-working and alloys having strengths of over 200,000 psi in the worked and heat-treated condition have been developed.