High-entropy alloys are unusual in more ways than one. Because several elements share the crystal lattice almost equally, the atoms interfere with one another’s movement, producing severe lattice distortion, sluggish diffusion, and what materials scientists call the cocktail effect, in which the mixture behaves in ways its individual components do not predict. When such an alloy is designed around elements with high heats of combustion, notably titanium and zirconium, it becomes an energetic structural material: a substance that can hold together as a structural component and, under violent impact, fragment and react with oxygen to release substantial chemical energy. That dual personality is precisely what makes these alloys interesting for impact applications, since they deliver both kinetic penetration and a secondary thermal and overpressure effect that conventional penetrators cannot provide.

The team fabricated three candidate alloys, TiZrWV, TiZrNbWV, and TiZrNbV, using vacuum arc melting. High-purity raw materials were weighed in equiatomic ratios, melted under strong electromagnetic stirring, and remelted six times, with each melt held liquid for twenty minutes to ensure chemical homogeneity. The ingots were then rapidly solidified under argon-assisted forced convection and machined into cylindrical specimens by wire electrical discharge machining. Tungsten was added to boost solid-solution strengthening, while niobium and vanadium were introduced to stabilize the solid-solution phases and refine the grains. The choice of composition was deliberate: each alloy represents a different balance between strength, ductility, and reactivity.

Quasi-static compression tests on an MTS Landmark universal testing machine, run at strain rates of 0.001 and 0.01 per second, revealed a clear hierarchy of mechanical behavior. TiZrNbWV emerged as the strongest of the three, with a yield strength rising from 1418 to 1469 megapascals as the strain rate increased, and a fracture strength reaching 2000 megapascals at the higher rate while still sustaining a fracture strain of roughly thirteen percent. Its yield strength exceeded that of TiZrWV by more than twenty-five percent under identical conditions. TiZrNbV, by contrast, showed the best plastic deformation capability and the highest quasi-static fracture strength, climbing from 1960 to 2021 megapascals, while TiZrWV sat in the middle with intermediate values in both categories.

Scanning electron microscopy of the fractured specimens explained these differences at the microstructural level. TiZrWV failed through quasi-cleavage planes characteristic of brittle fracture. TiZrNbWV displayed a layered, stepped surface with folds and tear ridges, indicating a mixed brittle-ductile process in which crack propagation was accompanied by significant plastic flow. TiZrNbV showed river-like patterns and abundant ductile dimples of moderate depth, the signature of fracture driven by the nucleation, growth, and coalescence of microvoids. In all three alloys, cracks tended to propagate along shear planes oriented at roughly forty-five degrees to the compression axis, and bright flashes of light were recorded at the fracture surfaces, evidence that even slow, quasi-static loading can trigger measurable energy release.

Dynamic compression on a split Hopkinson pressure bar pushed the strain rates into the 1000 to 3100 per second range and exposed pronounced strain-rate strengthening in every alloy. TiZrWV saw its yield strength jump by more than fifty percent relative to its quasi-static values, reaching 1723 megapascals at a strain rate of 2105 per second, while TiZrNbWV reached 1967 megapascals at nearly 3000 per second. The relative strengthening followed the order TiZrWV, then TiZrNbV, then TiZrNbWV, a pattern the authors attribute to the different ways niobium and tungsten contribute to baseline strength and strain-rate sensitivity. Under the microscope, the dynamically fractured specimens told an even more dramatic story: all three alloys developed localized molten structures on their fracture surfaces, from large melted blocks in TiZrWV and TiZrNbWV to fine molten droplets and flow traces in TiZrNbV, confirming that adiabatic heating during high-strain-rate shear had raised local temperatures high enough to melt metal.

The centerpiece of the study was a series of ballistic experiments in which cylindrical flat-nosed projectiles of each alloy were fired from a 14.5-millimeter-bore ballistic gun at a novel three-layer spaced target: a five-millimeter-thick 2A12 aluminum front plate, a 1.6-millimeter circuit board populated with conventional electronic components, and a three-millimeter aluminum rear plate, with 100-millimeter gaps between layers. A high-speed camera running at 30,000 frames per second captured the impacts at velocities ranging from roughly 548 to 1012 meters per second. The results showed a striking division of labor. Penetration of the front aluminum plate was dominated by kinetic effects, with classic plug-shear failure and circular perforation, and many reactive fragments passed through unreacted, producing only weak flashes. The main energy-release reactions occurred later, when the fragmented projectile struck the rear circuit board, shattered further, and generated a high-temperature fragment cloud that reacted violently with atmospheric oxygen.

To quantify this behavior, the researchers introduced two optical indicators drawn from the high-speed imagery: flash duration, which reflects how long the energy release persists, and flash intensity, a normalized grayscale measure of instantaneous emission. TiZrNbWV proved the most persistent reactor, with flash durations reaching 73.92 milliseconds at high impact velocity, nearly eighteen times its low-velocity value. TiZrWV followed with a maximum of 55.08 milliseconds, while TiZrNbV lasted only 15.02 milliseconds even at 984 meters per second, showing little sensitivity to velocity. The intensity picture inverted at different speeds: at low velocities TiZrNbV produced the highest peak flash intensity and the steepest rise, indicating strong instantaneous release, whereas at high velocities TiZrWV delivered the most intense and rapid response. TiZrNbV’s intensity curves at 856 and 984 meters per second even showed a bimodal profile, attributed to secondary fragmentation generating a second wave of reactions.

The damage left behind on the targets provided macroscopic confirmation of these differences. The third-layer aluminum plates perforated by TiZrWV and TiZrNbWV projectiles bore irregular, jagged, tear-like holes with rough boundaries, consistent with dispersed fragment populations and sustained, spatially nonuniform loading. The TiZrNbV projectile, with its better deformability and shorter energy-release duration, left a comparatively regular circular perforation with smooth edges, indicating a more concentrated residual fragment population. At the highest velocities, all three alloys produced a quasi-detonation phenomenon, in which extensive fragmentation enlarged the reactive surface area and triggered intense thermochemical release within a very short time, and in several cases energy release continued behind the fully perforated target as a pronounced post-penetration effect.

The authors are careful to note the limits of the work. The ballistic tests covered a restricted velocity range, and flash duration and intensity are image-based comparative parameters rather than direct measurements of absolute energy, so the quantitative link between microstructure, fragmentation, and reaction kinetics remains to be established. Even so, the study delivers a practical design guideline: TiZrNbWV suits applications demanding high structural strength, strong penetration, and sustained post-impact energy release; TiZrWV fits scenarios that prize rapid, intense release under high-velocity impact; and TiZrNbV offers a balance of deformability, concentrated penetration, and fast response. By turning the fleeting glow of an impact into numbers, the work gives designers of energetic structural materials a quantitative language for tuning composition to the mission at hand.

Subject of Research: Mechanical properties and impact-induced energy release of TiZr-based high-entropy alloy projectiles against multi-layer circuitry targets

Article Title: Mechanical properties of reactive high-entropy alloys and energy-release characteristics of projectiles impacting a multi-layer circuitry target

Article References: Shi, Z., Wu, C., Qiao, W., Mao, L., Pang, Z., & Liu, J. (2026). Mechanical properties of reactive high-entropy alloys and energy-release characteristics of projectiles impacting a multi-layer circuitry target. Mechanical Sciences, 17(2), 901-916. https://doi.org/10.5194/ms-17-901-2026

Image Credits: AI Generated

DOI: 10.5194/ms-17-901-2026

Keywords: high-entropy alloys, energetic structural materials, TiZr-based alloys, ballistic impact, energy release, vacuum arc melting, split Hopkinson pressure bar, flash duration, flash intensity, penetration mechanics, strain-rate strengthening, multi-layer target

The study, led by Lingyan Shen and Yonggui Liu with Xiahui Pan, builds on two decades of experimental work showing that the onset of frictional sliding is mediated by rupture fronts. In landmark experiments, researchers visualized the real contact area between polymethyl methacrylate (PMMA) blocks using laser illumination at angles beyond the critical angle for total internal reflection, revealing three distinct classes of detachment fronts: slow fronts, sub-Rayleigh fronts, and intersonic fronts that travel faster than shear waves but slower than longitudinal waves. Macroscopic sliding begins only after these fronts have swept across the whole interface, breaking micro-contacts one after another like a line of falling dominoes.

The new observations do not fit that sequential picture. In the experiments, two PMMA blocks forming a rough interface were loaded by a split-Hopkinson pressure bar, which delivers a well-defined compressive pulse to the trailing edge of the upper slider. Arrays of miniature piezoelectric film sensors, just 3 millimeters in active diameter with a 50-megahertz center frequency, were bonded along two orthogonal directions: one set parallel to the interface at a fixed distance of 2.5 millimeters, and another set perpendicular to it at depths ranging from 2.5 to 28 millimeters. A trigger strain gauge on the impact face defined the absolute time origin with sub-microsecond precision.

What the sensors recorded was startling. All five transducers in the parallel array detected three distinct wave packets at approximately 0.34, 0.70, and 1.00 microseconds after impact, with nearly identical amplitudes and waveforms. At the earliest of those instants, the incident compressive wave, traveling at roughly 2,520 meters per second, could have covered less than a millimeter, nowhere near the 20 millimeters separating the first sensor from the loaded edge. Simultaneous detection across the full array therefore points to a distributed, non-local source: the interface as a whole lit up at once, rather than being activated edge-first by a propagating disturbance. Control experiments, including specimen-removed electrical cross-talk tests and independent optical trigger calibration showing channel-to-channel jitter of only about 5 nanoseconds, confirmed that the synchronization was a physical reality rather than an instrumentation artifact.

The wave packets carried further signatures that set them apart from anything in the established catalog of interfacial dynamics. Their propagation speed, extracted from the trajectories of the wavefronts across the sensor arrays, matched the longitudinal wave speed of PMMA, about 2,540 meters per second experimentally. That is far above both the Rayleigh wave speed of roughly 1,237 meters per second and the shear wave speed of about 1,400 meters per second, the ceilings that classical fracture mechanics imposes on rupture fronts propagating along an interface. Moreover, the waves radiated perpendicularly into the bulk of the material, whereas rupture fronts are confined to the interfacial plane itself. The team conducted 150 experiments varying impact velocity from 5 to 15 meters per second, interface roughness, and loading geometry, and found that the characteristic arrival times, inter-packet intervals, and peak amplitudes varied by less than 4.2 percent across all conditions, a robustness that implicates collective interfacial dynamics rather than stochastic asperity-scale contact mechanics.

To probe the mechanism, the researchers built explicit dynamic finite element simulations in ABAQUS, modeling two elastic blocks pressed together along an interface with idealized asperity geometries: a smooth baseline, periodic isosceles-triangular asperities, and sharper right-triangular asperities. Using four-node quadrilateral elements with a refined mesh of 1-micrometer characteristic length and a time step of about 0.35 nanoseconds, the simulations reproduced the essential phenomenology. Von Mises stress contours revealed spherical wavelets emanating from micro-contact regions whose collective envelope formed planar wavefronts parallel to the interface, a structure that persisted across all three asperity configurations. The simulated propagation speed of 2,536 meters per second agreed with the experimental value to within 0.2 percent, and the simulated inter-packet interval of about 0.30 microseconds matched the experimental 0.30 to 0.34 microseconds, with the small offset attributable to the finite rise time of the real loading pulse and trigger uncertainty.

The simulations also exposed a remarkable discrete structure in the interfacial response. Monitoring six adjacent material points along the interface showed statistically indistinguishable time histories, confirming synchronous activation, while the normal stress and displacement traces revealed wave packets arriving at equally spaced intervals of roughly 0.3 microseconds. When the plateau amplitudes were normalized, they followed a striking quantization law: the amplitude at the nth plateau grows as 3n(n − 1), a closed-form rule indicating that the interface accumulates wave energy in discrete, integer-multipled steps. Spectral analysis of the oscillatory signal showed a dominant carrier frequency of about 34.4 megahertz, with higher harmonics, whose amplitude but never its frequency changed from packet to packet. The normal displacement evolved as a staircase of flat plateaus, each step coinciding with the arrival of an additional packet, echoing discrete force plateaus previously observed in stick-slip systems.

To explain these features from first principles, the authors propose a non-local micro-oscillation model that extends Huygens’ principle to elastodynamics. In this picture, the frictional interface behaves as a distributed secondary source: once the incident plane wave arrives, every material point on the interface acts as a point emitter of spherical wavelets, and the envelope of those wavelets is precisely the observed planar longitudinal wavefront. The model’s Green’s function formalism naturally yields the characteristic 1/r geometric decay of the radiated amplitude, which the finite element results match closely; the experimental amplitudes fall systematically lower, a discrepancy the authors attribute to viscoelastic damping in the real PMMA that the conservative theory deliberately excludes. A Taylor expansion of the discrete-time source term reproduces the step-wise amplification, with the theory predicting a proportionality coefficient of π where the simulations give 3, a modest deviation likely arising from the discrete sampling of the continuous source distribution.

Perhaps the deepest implication concerns the foundations of solid mechanics itself. Classical continuum theory rests on the principle of local action, the assumption that a material point responds only to conditions in its immediate neighborhood, which underpins the standard hyperbolic wave equation and guarantees that disturbances propagate at finite speeds. The synchronous activation of the entire interface within about 0.3 microseconds, when acoustic transit across the 40-millimeter contact would take roughly 16 microseconds, suggests instead that elastic coupling among micro-asperities establishes a correlated source density across the whole interface almost instantaneously. The authors formalize this by making the source term of the bulk wave equation satisfy an elliptic constraint, a modified Helmholtz equation with an exponential-decaying Green’s function, so that the wave operator remains hyperbolic while its source behaves non-locally. The temporal discreteness, meanwhile, challenges the assumed smoothness of classical elastodynamic solutions and points toward delay-differential and pseudodifferential operator frameworks for interface-mediated wave propagation.

The practical stakes are considerable. Because the discrete wave packets are intrinsic to the interface rather than artifacts of loading or morphology, they offer a quantitative fingerprint that could underpin non-destructive evaluation of bolted joints and friction assemblies, where hidden interfacial degradation is a persistent safety concern. The authors also point to seismic precursor monitoring as a potential application, since the same fracture-wave coupling operates at laboratory faults and may scale to natural ones. Much remains to be done: a self-consistent derivation of the 0.3-microsecond characteristic timescale from material and geometric parameters is still outstanding, and interface curvature effects await study. But the core message is already clear and likely to reverberate well beyond tribology: under impulsive loading, a frictional interface is not merely a passive plane where cracks nucleate. It is a coherent, quantized radiator, a collective oscillator whose synchronized voice challenges how physicists have modeled contact for generations.

Subject of Research: Non-local, quantized stress wave emission from frictional interfaces under dynamic shock loading

Article Title: Observation of non-local waves at frictional interface

Article References: Shen, L., Liu, Y., & Pan, X. (2026). Observation of non-local waves at frictional interface. Results in Engineering, 32, Article 113150. https://doi.org/10.1016/j.rineng.2026.113150

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.113150

Keywords: friction, rupture fronts, stress waves, frictional interfaces, PMMA, split-Hopkinson pressure bar, finite element simulation, non-locality, Huygens principle, acoustic emission, earthquake nucleation, non-destructive evaluation