ITER's Divertor: A Practical Engineering Decision for Exhausting Plasma Heat

What you are trying to achieve

You need a divertor configuration that removes a tokamak's exhaust power without destroying the surfaces that receive it. The divertor is the component that intercepts plasma from the scrape-off layer (SOL — the region outside the last closed flux surface, where field lines run into the wall instead of looping back into the core) and puts it somewhere other than the vacuum vessel wall.

ITER's divertor is a useful reference because it is a documented, supported design decision rather than a concept sketch. This guide walks through the decision as a task: fix the heat-flux target, spread the load with geometry, choose materials and cooling, then verify before hardware is committed.

Prerequisites and assumptions

• A defined exhaust target: steady-state and transient heat flux the target plates must survive, plus pulse length.
• Working knowledge of the SOL, magnetic flux expansion, and strike-point behaviour.
• Material property data for the plasma-facing material and the heat sink at operating temperature.
• Access to a thermal-hydraulic solver, and ideally a linear plasma device for coupon-level exposure.
• Version assumption: the figures below follow the ITER baseline as described in publicly available design documentation. ITER's divertor material baseline has been revised during the project's history, so confirm which baseline currently applies before you commit to a material.

The design decision, step by step

1. Fix the exhaust target. ITER's divertor is specified for steady-state heat removal on the order of 10 MW/m² on the vertical targets. Treat that as the number your cooling and material choices must satisfy — not as a value to copy into a different machine, whose SOL width, power and pulse length will differ.
2. Choose geometry that spreads the load. The ITER geometry uses inner and outer vertical targets plus a dome. The targets are tilted relative to the field lines so magnetic flux expansion spreads arriving flux over a larger area. The dome shields the private flux region and intercepts neutrals.
3. Choose the plasma-facing material and heat sink. The planning brief behind this article describes tungsten-coated carbon fibre composite plates with a path toward monoblock tungsten. Tungsten has a high melting point and low sputtering yield but is brittle at low temperature; carbon tolerates transients better but erodes and traps tritium. Record explicitly which baseline you are designing to.
4. Specify active cooling. Pressurised water flows through copper-alloy heat sinks embedded behind the plasma-facing components. The heat sink's job is to hold the bond line and the surface below their limits at the specified flux.
5. Make it replaceable. The modular cassette design lets a whole divertor unit be removed and replaced remotely. That is what makes inspection schedules and material upgrades practical rather than a shutdown-scale project.

Worked example: a design review checklist

The table below is a review aid, not a specification. Fill the right-hand columns with your own analysis results.

Decision	ITER reference choice	What to check	Failure mode if wrong
Heat-flux target	Order of 10 MW/m² steady state on vertical targets	Reconcile with your own SOL model and pulse length	Surface melting or bond failure
Geometry	Inner and outer vertical targets, dome, tilted to the field	Flux expansion factor and strike-point sweep range	Localised hot spot
Plasma-facing material	Tungsten-based PFCs (baseline revised over the project)	Confirm current baseline; check behaviour near the ductile-to-brittle transition	Cracking during disruptions
Heat sink	Copper alloy with pressurised water	Flow rate, pressure, subcooling margin	Critical heat flux / burnout
Maintenance	Modular cassettes, remote handling	Remote handling envelope and alignment tolerance	Extended downtime

Expected checks

• Thermal-hydraulic simulation. Run your specified flow rates and confirm component temperatures stay below material limits. The planning brief names ANSYS CFX and OpenFOAM as validated codes; use whichever your organisation has qualified, and document the boundary conditions.
• Prototype exposure. Linear plasma devices such as Magnum-PSI expose prototype cassettes or coupons to representative fluxes to measure erosion and check bonding integrity. This is where you learn whether the joint survives, not whether the concept does.
• Documentation reconciliation. Compare your parameters against the ITER baseline design documents (for example the ITER Physics Basis and the Divertor Design Report) rather than secondary summaries, which may describe superseded baselines.

If the checks fail

This is an analysis and design task, so there is no deployed state to roll back. Recovery means revising the design before hardware is committed.

• Peak heat flux above the target limit: increase flux expansion, adjust strike-point position or sweep, or move toward a partially detached regime using impurity seeding, which radiates power before it reaches the target.
• Thin heat-sink margin: raise flow rate or lower inlet temperature, staying inside the cooling loop's design envelope.
• Plasma-facing material cracking: control preheating to keep tungsten out of its brittle regime, or reconsider the material for transient tolerance.
• Bonding failure in prototype testing: revisit the joining process before freezing the cassette design.

Limitations and how to verify the result

The parameters here come from a planning brief, not from a live source check, and ITER's divertor design has been revised more than once — the material baseline in particular. Two cautions from that brief are worth carrying into any review: tungsten's brittleness at low temperature demands careful preheating and temperature control to avoid cracking during disruptions, and radiation damage plus transmutation can change tungsten's thermal conductivity over time, which is why periodic inspection and replacement schedules exist rather than a one-off qualification.

To verify: confirm material specifications and cooling parameters against current ITER baseline documents, reproduce the thermal-hydraulic case in a validated solver, and, where possible, test a prototype in a linear plasma device. Treat the 10 MW/m² figure as ITER's design point, not a universal requirement.