The technique, known as shear assisted processing and extrusion, or ShAPE, belongs to a family of solid-phase processing methods in which metal is deformed plastically at elevated temperature without ever melting. A rotating, non-consumable die is plunged into solid feedstock at ambient starting temperature. Frictional and adiabatic heating soften the alloy locally, and scrolls machined into the die face draw the plasticized material into the die orifice and around a mandrel, where it emerges as a tube. The process operates at roughly half to nine-tenths of the melting point, and because the material never liquefies, the operator retains unusually direct control over the final microstructure through temperature regulation alone. Aluminum and magnesium alloy tubes have been made this way before, but zirconium alloys posed an open question.

That question matters because Zircaloy-4 is metallurgically awkward. Its hexagonal close-packed crystal structure offers few slip systems, so conventional thermomechanical routes impose a strong crystallographic texture on the finished tube. Certain grain orientations allow hydrides, which form when the metal picks up hydrogen in service, to align circumferentially or radially, and those hydride alignments are notoriously damaging to mechanical properties. Producers therefore invest in elaborate processing sequences to steer grain orientations toward safer configurations. A single-step process that instead produces a nearly random texture would sidestep much of that complexity, cutting cost, time, and energy while potentially improving performance.

In the new study, published in the Journal of Materials Science: Metallurgy, the researchers began with the simpler of two experiments: shear assisted processing without extrusion. A die without an orifice was plunged into Zircaloy-4 billets roughly 10 millimeters tall and 31.75 millimeters in diameter, deforming the material in place. By varying the die rotational speed between 100 and 300 revolutions per minute while holding the plunge rate constant, the team reached four distinct processing temperatures, measured with a thermocouple attached to the tool. Those temperatures, 710, 802, 904, and 1002 degrees Celsius, deliberately sampled three different phase fields of the alloy: the low-temperature alpha phase, the two-phase alpha-plus-beta region, and the high-temperature beta phase, whose boundaries sit near 810 and 977 degrees Celsius respectively.

The microstructural payoff was striking. Billets processed at the two highest temperatures, which crossed into the beta phase field, developed lath-type structures: needle-shaped alpha grains roughly 1.4 to 1.6 micrometers thick, evidence that the material had been quenched from the beta phase as the tool withdrew. Notably, such fine laths appeared even though the billets were not deliberately quenched and cooled far more slowly than the water quench rates used in reference studies, which the authors attribute to the extreme shear deformation generating dense dislocations that multiply nucleation sites for new alpha grains. Billets processed at 710 and 802 degrees Celsius, below the beta transus, instead showed equiaxed grains. Grain sizes ranged from about 1.2 micrometers at the lowest temperature to roughly 5.1 micrometers at the highest, giving the team a direct dial connecting die temperature to grain morphology.

Secondary phase particles told a complementary story. Scanning electron microscopy with energy-dispersive spectroscopy revealed iron- and chromium-rich precipitates, consistent with Zr(Fe,Cr)2 particles, distributed along phase boundaries, prior grain boundaries, and grain interiors in all processed conditions. Because the billets cooled without controlled quenching, the sizes of these particles were not regulated, but the authors point out that controlled cooling can readily be incorporated into future experiments to tailor secondary phase dimensions and, with them, specific properties. The temperature mapping exercise also carried a caveat: the thermocouple sat about 3 millimeters from the deformation zone, so the true processing temperature likely exceeded the recorded die temperature, a discrepancy that sacrificial thermocouples or smoothed-particle hydrodynamics modeling could resolve in future work.

Armed with that temperature-microstructure map, the team attempted the headline achievement: extruding actual Zircaloy-4 tubes from solid billets in a single ShAPE step. The billets, about 10 millimeters tall with a 31.75 millimeter outer diameter and a 10 millimeter inner hole, were processed with a tungsten-lanthana die and an IN718 nickel-alloy mandrel, targeting a tube with 1 millimeter wall thickness, 10 millimeter inner diameter, and 12 millimeter outer diameter. After tuning rotation speed and plunge rate across several trials, the best tube emerged from a run at 200 revolutions per minute and a 7.62 millimeters-per-minute plunge rate. Die temperatures during steady-state extrusion climbed from about 850 degrees Celsius near the start of the tube to about 920 degrees Celsius near the end, and the finished walls measured 1.1 and 1.13 millimeters thick at those two locations, close to the design target.

Microscopy of the tube revealed a gradient that mirrors the temperature map established in the first experiments. Near the start of the tube, where the die was cooler, the microstructure was fully recrystallized with fine equiaxed alpha grains averaging roughly 3.6 to 4.5 micrometers across longitudinal, transverse, and planar sections. Near the end, where temperatures approached the beta transus, the grains coarsened dramatically, reaching 16.7 to 24.1 micrometers, and adopted a lath-shaped morphology with an average alpha lath size of about 5.4 micrometers. Precipitate distributions shifted along the tube as well, from predominantly intragranular particles near the start to inter-lath particles near the end, and limited X-ray diffraction detected small peaks of zirconium hydrides. Oxidation along the tube’s inner surface near the exit, a consequence of processing without inert atmosphere protection, accounted for some unindexed regions in the electron backscatter diffraction data.

Perhaps the most consequential result came from texture analysis. Using Kearns parameters, which quantify the fraction of basal crystal planes aligned with a given direction, the team found values between 0.28 and 0.44 near the tube start and 0.29 to 0.37 near the end, close to the 0.33 value that indicates a fully random texture. By contrast, conventionally recrystallized Zircaloy-4 typically shows Kearns values spanning 0.10 to 0.58, signaling preferred orientation. A randomly textured tube, produced without any of the intermediate annealing steps that conventional pilgering demands, would be expected to resist the anisotropic hydride alignment that degrades cladding performance over a reactor’s life.

As a final proof of concept, the researchers co-extruded Zircaloy-4 with pure nickel, sandwiching the zirconium alloy between inner and outer nickel layers, half the billet cross-section being nickel and half Zircaloy-4. The resulting cladded tube emerged in a single step, with a pure nickel inner diameter, a zirconium middle layer, and an outer region containing mixed nickel-zirconium plus a pure nickel skin. The mixed zone reflects an initial, untuned parameter set rather than a fundamental limitation, and the demonstration establishes that ShAPE can fabricate multi-material zirconium tubing that would otherwise require separate cladding operations. Funded through the National Nuclear Security Administration’s tritium modernization program, the work suggests a future in which reactor cladding, tritium-producing rod components, and cladded tubes are extruded directly from solid feedstock, with grain size, morphology, and texture controlled not by exhaustive post-processing but by the temperature of a spinning die.

Subject of Research: Solid-state shear assisted processing and extrusion of Zircaloy-4 for nuclear-grade tube fabrication

Article Title: Shear assisted processing of Zircaloy-4

Article References: Komarasamy, M., Canfield, N., Darsell, J., Garcia, D., Zhang, D., Guzman, A., Casella, A., & Senor, D. (2026). Shear assisted processing of Zircaloy-4. Journal of Materials Science: Metallurgy, 1(1), Article 3. https://doi.org/10.1007/s44492-025-00003-9

Image Credits: AI Generated

DOI: 10.1007/s44492-025-00003-9

Keywords: Zircaloy-4, shear assisted processing and extrusion, solid phase processing, zirconium alloys, nuclear fuel cladding, microstructure, texture, co-extrusion, advanced manufacturing, grain refinement, hydrides, materials science