Safe Shutdown Sequences
The controlled procedures that bring a plasma to rest with minimal stress on the machine, whether planned or triggered by fault.
Two ways to stop
A discharge ends either by planned ramp-down or by a triggered safe shutdown when something goes wrong. Both aim for the same outcome: the plasma current and stored energy brought to zero without a disruption and without exceeding any force, heat, or field limit on the way down. A good stop is uneventful.
The controlled soft landing
The preferred shutdown is a soft landing - a controlled ramp-down that keeps the plasma inside all limits until the current is low enough that a disruption would be benign. This requires maintaining shape and vertical control with diminishing authority and reducing density in step with current, so it is an active, designed maneuver rather than simply cutting power.
When soft landing is not possible
If the plasma is already too far from safe - vertical control lost, a mode locked, a limit exceeded - a soft landing may not be reachable in time. Then the stack chooses the least-damaging alternative: mitigation to spread the energy dump, or a fast but controlled termination. The decision logic for which path to take is pre-defined by severity.
Layered triggers
- Operator command: a planned or discretionary stop
- Exception handler: an off-normal condition beyond adaptation
- Disruption predictor: imminent disruption, trigger avoidance then mitigation
- Reflex layer: a hard hardware limit crossed, immediate protective action
Leaving the machine safe
A shutdown is not complete when the current reaches zero. The sequence also brings coils, power supplies, and fueling to their standby states and confirms the machine is in a defined, safe configuration ready for the next discharge or for maintenance. The end state is explicit and verified, not assumed.