radiation hardening carries two distinct and easily confused meanings: in engineering, the practice of designing a component to keep functioning under neutron and gamma exposure; in materials science, the increase in a metal's yield strength — and the matching loss of ductility — that irradiation itself inflicts.
Both senses apply to the same 📝fusion machine, which is why the collision matters. In the engineering sense, hardening is something a designer does: select materials and dopants that degrade slowly, interpose 📝shielding, derate the component, or build in a recovery step. In the materials sense, hardening is something radiation does without anyone's consent — 📝neutron damage knocks atoms off their lattice sites, and the resulting defect clusters pin dislocations, so an alloy's yield strength rises while its uniform elongation collapses. Embrittlement is the practical consequence, exposure is counted in 📝displacements per atom (dpa), and structural alloys such as 📝RAFM steel are qualified against it alongside the activation criteria that define 📝low-activation materials. Nothing in that process makes a component more radiation-tolerant. The same word points in opposite directions depending on who is speaking, and neither usage is going to yield.
In a compact high-field tokamak the components with the least radiation margin are the ones nearest the 📝plasma that cannot be replaced. 📝REBCO 📝HTS tape loses 📝critical current as fast-neutron fluence accumulates, and unlike a projection about plasma performance this is a measured quantity: reactor irradiation campaigns on tape samples establish the fluence at which degradation sets in. That measurement becomes a lifetime dose limit for the magnet, and the shield is then sized to keep the coil beneath it.
Optical fiber is the other pressure point, because fiber-optic sensing is the natural way to catch a 📝quench inside a winding and the fiber necessarily runs where the radiation is. Ionizing radiation creates color centers in silica that absorb light — radiation-induced attenuation — so the sensor slowly goes blind. 📝Commonwealth Fusion Systems and the Paihau-Robinson Research Institute at Victoria University of Wellington irradiated germanium-doped and radiation-hard silica fibers with gamma rays at cryogenic temperature and showed that photobleaching at 1550 and 970 nm mitigates the induced attenuation (Journal of Applied Physics 134, 2023). Follow-on work applied optical annealing to quench-detection sensors (IEEE Transactions on Applied Superconductivity 34, 2024) and resolved the attenuation along the fiber's length (IEEE TAS 35, 2025). "Radiation-hard fiber" in that literature is a procurement specification — typically a pure-silica core with few dopants — and is the engineering sense of the term in its purest form.