Evolution of SPI-induced disruptions in ASDEX Upgrade

This paper characterizes the evolution of Shattered Pellet Injection (SPI)-induced disruptions in ASDEX Upgrade, detailing how varying injection parameters and neon assimilation influence the sequence of disruption phases and mitigation efficiency, as evidenced by changes in disruption time scales and plasma current profiles.

The Gist

Imagine a fusion reactor like the ASDEX Upgrade as a giant, super-hot pot of soup (the plasma) that we are trying to keep boiling without spilling over or burning the kitchen. The goal is to create clean energy, but sometimes the soup gets too unstable and threatens to explode, damaging the pot. This is called a disruption.

To save the pot, scientists need to cool the soup down instantly and safely. They use a technique called Shattered Pellet Injection (SPI). Think of this like firing a frozen snowball into the hot soup. But instead of a single solid ball, the snowball is smashed into millions of tiny ice chips right before it hits the soup. This creates a massive cloud of cold particles that spreads out quickly, cooling the soup down before it can cause damage.

This paper is a detailed report on what happens when they fire these "ice clouds" into the ASDEX Upgrade machine. Here is the story of what they found, broken down simply:

1. The Setup: The "Ice Cannon"

The scientists built a very flexible cannon at the machine. They can shoot pellets made of different things:

• Pure Deuterium: Like a snowball made of regular ice.
• Neon-doped: Like a snowball with little bits of "glow-in-the-dark" neon mixed in.
• Pure Neon: A snowball made entirely of neon.

They also changed the "shatter head" (the hammer that smashes the pellet). Some heads made tiny, slow chips, while others made larger, faster chips. They wanted to see which combination cooled the soup best.

2. The Chain Reaction: A Movie in Slow Motion

When they fired the pellets, they didn't just see a simple "cool down." They saw a specific movie play out, scene by scene. The paper describes these scenes:

• Scene 1: The First Light (FL). Before the main snowball hits, a few tiny chips and gas sneak in. It's like seeing the first few snowflakes hit the window.
• Scene 2: The Main Crash (MFA). The big cloud of chips hits the soup. This causes a huge flash of light (radiation) as the cold material starts to melt and interact with the hot plasma.
• Scene 3: The Wobble (PME). As the cold hits, the soup gets confused. The pressure drops, and the whole blob of soup wobbles and shifts position, moving toward the center of the pot.
• Scene 4: The Glow-Up (MARFE). A bright, glowing ring forms near the edge of the soup and starts creeping upward, like a fire spreading along a wall.
• Scene 5: The Big Cool-Down (TQ). The soup finally loses its heat. The temperature crashes.
• Scene 6: The Current Spike (IP-spike). As the heat vanishes, the electrical current in the soup tries to compensate, causing a sudden, sharp jump in power (like a car engine revving up before stalling).
• Scene 7: The Fall (CQ & VDE). The current dies out, and the soup blob falls toward the bottom of the pot.

3. The Big Discovery: It's All About the "Neon"

The most important finding is that how much neon is in the pellet changes the entire movie.

• The "Boring" Movie (Pure Deuterium): If they shoot pure ice (Deuterium), the movie is long and dramatic. You see all the scenes clearly: the wobble, the creeping glow, the spike, and the fall. The soup takes a long time to cool down (up to 15 milliseconds). The current curve looks like a convex hill (slow at first, then fast).
• The "Fast-Forward" Movie (High Neon): If they shoot a pellet with a lot of neon, the movie speeds up. The scenes merge together. The soup cools down almost instantly (in less than 1 millisecond). The current curve looks like a concave bowl (fast at first, then slowing down).

The Analogy:

• Convex (Pure Deuterium): Imagine a car braking on a dry road. It takes a while to stop, and the last bit of stopping is the hardest. This is a "conduction-dominated" disruption (heat moves slowly).
• Concave (High Neon): Imagine a car hitting a giant wall of water. It stops almost instantly. This is a "radiation-dominated" disruption (heat is radiated away instantly).

4. Why Does This Matter?

The scientists are testing this to prepare for ITER, a massive fusion reactor being built in France. ITER will be much bigger and more powerful. If ITER has a disruption, it could cause serious damage.

• The Goal: They want to find the perfect "ice pellet" recipe that stops the disruption as fast and safely as possible.
• The Problem: If the pellet has too little neon, the disruption takes too long, and the machine might get damaged. If it has too much neon, it stops too fast, which might be hard to control.
• The Sweet Spot: They found that even a tiny bit of neon (0.085%) makes the disruption happen much faster. This is good for safety, but it means the timing has to be perfect.

5. The "Shape" Tells the Story

The scientists realized they don't need to look at every single camera to know how well the mitigation worked. They just need to look at the shape of the current curve as it dies out:

• Convex shape? The mitigation wasn't great; the soup took too long to cool.
• Concave shape? The mitigation was excellent; the soup cooled down instantly and safely.

Summary

This paper is like a driver's manual for fusion reactors. It tells us that when we try to stop a runaway fusion reaction, the "ingredients" of our emergency brake (the pellet) matter immensely. By adding the right amount of "neon spice" and smashing the pellet into the right size of chips, we can turn a slow, dangerous crash into a quick, safe stop. This knowledge is crucial for building the future of clean fusion energy.

Technical Summary

1. Problem Statement

Disruptions in tokamak fusion reactors pose a severe threat to machine integrity due to the sudden release of thermal energy and electromagnetic forces. If unmitigated, these events can cause localized heat loads, structural damage, and the generation of runaway electron (RE) beams.

• The Challenge: For the ITER reactor, the Shattered Pellet Injection (SPI) technology has been selected for the Disruption Mitigation System (DMS). Unlike Massive Gas Injection (MGI), SPI injects a cryogenic pellet that shatters into fragments near the plasma edge, increasing the surface-to-volume ratio for rapid ablation and assimilation.
• The Knowledge Gap: While SPI is the chosen technology, the complex dynamics of how fragment size, velocity, and material composition (specifically neon content) influence the disruption chain of events, mitigation efficiency, and plasma current decay profiles were not fully characterized prior to this study.

2. Methodology

The study utilizes data from the ASDEX Upgrade (AUG) tokamak, which was equipped with a highly flexible, custom-built SPI system to support ITER DMS design and modeling.

• Experimental Setup:
  • SPI System: Three independent injection barrels with different shatter heads installed for the 2022 campaign:
    • GT1 (25°, rectangular, long): Produces collimated, smaller, and slower fragments.
    • GT2 (25°, circular, short): Produces a plume with increased spatial spread and swirling motion.
    • GT3 (12.5°, rectangular, long): Produces collimated, larger, and faster fragments.
  • Target Plasmas: H-mode discharges at 1.8 T and 800 kA.
  • Injection Variables: Pellets composed of varying mixtures of Deuterium ($D_2$) and Neon ($Ne$), ranging from 100% $D_2$ to 100% $Ne$.
  • Diagnostics: A comprehensive suite including foil bolometers (5 toroidal sectors), AXUV diodes, ultra-high-speed (UHS) cameras (top-down and side views), interferometry, and Thomson scattering.
• Analysis Approach: The authors analyzed the chronological evolution of disruption phases, correlating injection parameters (fragment size/velocity via shatter head geometry and neon content) with observable plasma responses (radiation peaks, current quench rates, and vertical displacement).

3. Key Contributions

The paper provides a comprehensive characterization of the SPI-induced disruption chain of events and establishes a link between injection parameters and mitigation outcomes.

• Definition of Disruption Phases: The authors identified and detailed a specific sequence of events in SPI disruptions:
  1. First Light (FL): Initial radiation from tiny fragments/gas.
  2. Main Fragment Arrival (MFA): Large radiation peak as the bulk of fragments enters.
  3. Plasma Movement Event (PME): A distinct phase (observed in long pre-TQ durations) involving plasma shape change and a secondary radiation peak, likely caused by MHD activity breaking central flux surfaces.
  4. MARFE: Multifaceted Asymmetric Radiation from the Edge, moving upstream.
  5. Thermal Quench (TQ) & IP-Spike: Rapid loss of thermal energy and a transient plasma current spike.
  6. Current Quench (CQ) & Vertical Displacement Event (VDE): Decay of plasma current and vertical movement.
• Evolution of Mitigation Efficiency: The study demonstrates that disruption behavior is a continuous process evolving from "conduction-dominated" (unmitigated) to "radiation-dominated" (well-mitigated) as assimilated neon increases.
• CQ Shape as a Proxy: The paper identifies the shape of the plasma current decay during the CQ as a critical figure of merit:
  • Convex: Poorly mitigated (conduction dominated).
  • Linear: Intermediate mitigation.
  • Concave: Well-mitigated (radiation dominated).

4. Key Results

• Impact of Neon Content ($f_{neon}$):

  • Radiation Peaks: In 100% $D_2$ injections, up to four distinct radiation peaks are visible (MFA, PME, TQ, VDE). As neon content increases, these peaks merge, eventually resulting in a single, large radiation peak.
  • Time Scales: Increasing neon assimilation drastically reduces disruption time scales.
    • Pre-TQ duration: Reduced from ~15 ms (100% $D_2$) to < 0.5 ms (high neon).
    • Early CQ duration ($\Delta t_{100\to80}$): Reduced from ~13.3 ms to ~8.2 ms.
  • IP-Spike: The amplitude of the plasma current spike decreases as neon content increases, due to the transition from conduction to radiation dominance.
  • Halo Currents: Shunt current measurements indicate a reduction in halo currents flowing through the divertor tiles as neon content increases, suggesting better mitigation of VDE forces.

• Impact of Fragment Size and Velocity:

  • Fragment size and velocity (determined by shatter head geometry) significantly affect material assimilation, particularly at low neon doping levels.
  • Larger, slower fragments (produced by the 12.5° head) generally lead to deeper penetration and better assimilation, resulting in more linear CQ shapes.
  • Trend Reversal: A "roll-over" effect was observed around 1.25% neon. Below this threshold, larger fragments shorten pre-TQ times; above it, larger fragments may delay the TQ by preventing edge radiative collapse, allowing more core assimilation.

• Mechanism of CQ Shape Transition:

  • Conduction Dominated (Convex): Core remains hot; resistivity increases as the plasma shrinks during VDE, accelerating current decay.
  • Radiation Dominated (Concave): Plasma volume remains constant due to broad halo currents stabilizing the centroid. The resistivity profile becomes flat, leading to an exponential (concave) current decay typical of an RL circuit.

5. Significance

• ITER Design Validation: The results provide crucial input for the ITER DMS design, confirming that SPI can effectively mitigate disruptions but highlighting the sensitivity to injection parameters.
• Operational Constraints: The study reveals that even small amounts of neon (0.085%) drastically shorten pre-TQ durations (< 4 ms). This poses a challenge for ITER's planned staggered injection scheme, as the timing window for deuterium assimilation (to suppress runaway electrons) becomes extremely narrow.
• Diagnostic Proxy: The paper establishes the CQ shape (convex vs. concave) and the number of radiation peaks as immediate, shot-to-shot indicators of mitigation efficiency. This allows operators to assess the success of a mitigation attempt in real-time or for rapid post-shot analysis, even without valid density measurements.
• Physics Understanding: The identification of the Plasma Movement Event (PME) and its likely connection to MHD-induced flux surface break-up adds a new layer of understanding to the pre-TQ dynamics of SPI disruptions.

In conclusion, this paper maps the complex landscape of SPI-induced disruptions, demonstrating that while increasing neon content improves mitigation (reducing forces and thermal loads), it simultaneously compresses the available time for control actions, necessitating precise optimization of fragment size and injection timing for future reactors like ITER.