Wall damage due to oblique high velocity dust impacts

Runaway electron termination on plasma facing components can trigger material explosions that are accompanied by the expulsion of fast solid debris. Due to the large kinetic energies of the ejected dust particles, their subsequent mechanical impacts on the vessel lead to extensive cratering. Earlier experimental studies of high velocity micrometric tungsten dust collisions with tungsten plates focused exclusively on normal impacts. Here, oblique high velocity tungsten-on-tungsten mechanical impacts are reproduced in a controlled manner by a two-stage light gas gun shooting system. The strong dependence of the crater characteristics and crater morphology on the incident angle is documented. A reliable empirical damage law is extracted for the dependence of the crater depth on the incident angle.

💡 Research Summary

The paper addresses a critical yet under‑explored aspect of plasma‑facing component (PFC) degradation in tokamaks: the non‑local damage caused by high‑velocity tungsten (W) dust ejected during runaway‑electron (RE) termination events. While RE‑driven explosions can launch micrometric W debris at speeds of 1–3 km s⁻¹, most prior experimental work on dust impacts has been limited to normal incidence. This study systematically investigates the effect of impact angle on crater formation by using a two‑stage light‑gas gun to accelerate nearly monodisperse, 63 µm spherical W particles to velocities of 2.0–3.0 km s⁻¹ and striking planar W targets tilted from 0° (normal) to 80° relative to the surface normal.

Key experimental details: the dust is produced by RF‑induction plasma melting and solidification, yielding high sphericity, low porosity, and a narrow size distribution (±3 µm). The target plates are 1 mm thick, polished, and effectively semi‑infinite for the shallow craters observed. Impact speeds are measured by time‑of‑flight between two laser sheets with <1 % uncertainty; the angle of incidence is set by the target holder with <0.5° error. After impact, crater depth (H), major‑axis length (L) and minor‑axis width (W) are quantified using scanning electron microscopy (SEM), optical microscopy, and a precision profilometer. For angles ≤60° SEM images are used; for steeper angles optical images are employed, leading to a 10–20 % measurement uncertainty for L and W, while H is measured with ±3 µm accuracy.

Results: At a representative speed of ~2 km s⁻¹, the average crater depth drops dramatically from 43 ± 7 µm at 0° to 6 ± 1 µm at 80°, roughly a factor of seven reduction. The major‑axis length grows from ~119 µm (0°) to a maximum of ~152 µm around 60°, then declines to 85 µm at 80°. The minor‑axis width follows a monotonic decrease from 119 µm (0°) to 51 µm (80°). The morphology evolves from circular (low angles) to elliptical (intermediate angles) and finally to a “fish‑like” shape with a shallow head at grazing incidences, which the authors attribute to secondary fragments generated during the primary impact.

A simple empirical law for depth versus angle is extracted: H(θ) ≈ H₀ cosⁿ(θ), with the data closely following a cosine dependence (n≈1). Length and width display approximately linear trends with cos θ and sin θ, respectively, consistent with the geometric projection of the impact velocity onto the surface. The study also notes that overlapping craters (≈50 % of impacts) were excluded from statistical analysis, emphasizing the need for isolated events to obtain reliable metrics.

The authors compare their findings with molecular‑dynamics (MD) simulations of W‑on‑W impacts, which predict distinct regimes: plastic deformation (200–500 m s⁻¹), impact bonding (500–1000 m s⁻¹), partial disintegration with sticking (1000–2500 m s⁻¹), and partial disintegration without sticking (2500–4000 m s⁻¹). The experimental velocities fall within the latter two regimes, where significant crater formation and partial fragmentation occur, matching the observed depth‑angle trends.

Implications for tokamak operation and design are substantial. First, the angle‑dependent damage law enables more accurate estimation of material loss from RE‑induced dust clouds, especially for components not directly facing the RE termination zone. Second, witness plates with known orientation can serve as diagnostics: measured crater dimensions can be inverted to infer dust speed and size distributions, providing a benchmark for RE‑driven fragmentation models. Third, the data furnish validation points for advanced simulation tools (e.g., smoothed particle hydrodynamics, peridynamics) that aim to predict macroscopic wall erosion from microscopic impact physics.

In conclusion, this work fills a critical gap by delivering the first systematic experimental dataset on oblique high‑velocity W‑dust impacts on W targets, establishing a clear quantitative relationship between impact angle and crater depth, and documenting the accompanying changes in crater geometry. These results constitute a practical foundation for incorporating non‑normal dust impact effects into the safety analysis, lifetime assessment, and mitigation strategies for future fusion reactors.