Skip to content
Technology How it works Breeder — Hyperion Burner — Aegis Burner — MetroVolt AI-Native Architecture Magnets Fuel cycle Safety Roadmap
Solutions AI & Data Centers Defense & Government Grid & Baseload Neutron Detection Quantum
Learn Technical Library
Proof Publications Whitepapers Technical Library Open Science & Reproducibility The Honest Gates
Company About / Mission Leadership Environment Health & Safety Investors Careers Press Contact
3D Model
Hyperion › Materials & Components
Materials & Components

Plasma-Facing Component Design

First wall, divertor, and their attachments form one system that must join a refractory surface to a cooled, replaceable structure.

Surface plus structure

A plasma-facing component is more than its surface material. It is a refractory armor bonded or attached to a cooled heat sink, mounted so it can be removed and replaced remotely. The hard problem is the join: connecting a brittle, high-temperature carbide to a cooled metallic structure that expands differently when heated.

Refractory armorBond /interlayerCooled heatsinkSupport structureRemote attachmentPlasma-facing component stack, surface to mount

The bond is the risk

Differential thermal expansion between a carbide armor and a metal heat sink concentrates stress at the interface. Under cycling this is a fatigue-and-delamination risk. Interlayers, graded joints, and mechanical attachment are all studied to manage it. A component that survives steady state can still fail at the bond under thermal cycling.

Designed to be replaced

Because armor erodes and neutron damage accumulates, plasma-facing components are consumables on a maintenance schedule. Their design is inseparable from the remote-handling system that will replace them, which is why component design and maintenance are treated as one program.

Standardized interfaces

So they can be replaced remotely, plasma-facing components share standardized mechanical, coolant, and mounting interfaces that a manipulator can disconnect and remake. Standardization trades some per-component optimization for maintainability, and it is decided early because a bespoke component that cannot be handled remotely is unserviceable once the machine is activated.

This page documents a design and simulation study, not a built machine. Construction begins Q2 2027; first-of-a-kind first tritium is targeted near 2030. Figures are computed, reproducible targets, not measurements.

Content reviewed August 2026 · design-and-simulation stage