Collisional-radiative data for tokamak disruption mitigation modeling

This paper presents high-fidelity collisional-radiative data for hydrogen, helium, neon, and argon plasma species generated using ATOMIC and FCR codes, which are represented as efficient B-spline surfaces to support accurate tokamak disruption mitigation modeling and runaway electron minimization.

The Gist

Imagine a giant, glowing donut of super-hot gas called a tokamak. This is our best hope for creating clean, limitless energy by mimicking the sun. But sometimes, this donut gets unstable and "disrupts"—it's like a car engine suddenly seizing up. When this happens, the energy inside has to go somewhere, and if it's not managed carefully, it can melt the walls of the machine, causing massive damage.

To stop this disaster, scientists have a plan: they inject "impurities" (like neon or argon gas) into the plasma. Think of these impurities as fire extinguishers. When they mix with the hot gas, they glow brightly and radiate the dangerous heat away safely, spreading it out over the whole wall instead of letting it burn a single spot.

However, to use these fire extinguishers effectively, we need to know exactly how they behave. This is where the paper comes in.

The Problem: A Complex Puzzle

The plasma inside the tokamak is a chaotic dance of particles. Electrons are zipping around, bumping into atoms, knocking electrons off (ionization), or sticking back on (recombination). To predict how much heat the impurities will radiate away, scientists need to solve a massive, complex math puzzle called a Collisional-Radiative (CR) model.

Think of the CR model as a giant, multi-layered traffic map.

• The Cars: Electrons and ions.
• The Intersections: Collisions where energy is exchanged.
• The Traffic Lights: Rules that determine if an electron stays attached or flies off.

For a long time, scientists used "simplified maps" (like the Coronal Equilibrium model). These maps assume the traffic is light and predictable. But in a tokamak disruption, the traffic is gridlocked and chaotic. The simplified maps give wrong directions, leading to bad predictions about how much heat will be radiated.

The Solution: A High-Fidelity GPS

The authors of this paper built a super-detailed, high-fidelity GPS for this plasma traffic. They used two powerful computer codes (ATOMIC and FCR) to calculate the behavior of four key "fire extinguisher" gases: Hydrogen, Helium, Neon, and Argon.

They didn't just guess; they calculated the exact probabilities of every possible collision and energy jump for these atoms across a huge range of temperatures and densities.

• For Hydrogen and Helium: They used a "fine-structure" approach, looking at the traffic down to the individual lane changes.
• For Neon and Argon: Because these atoms are more complex (like a city with more intersections), they used a "configuration-average" approach, looking at the flow of traffic by neighborhood.

The Challenge: Too Much Data

The problem with this high-fidelity GPS is that it generates terabytes of data. If you tried to plug this raw data directly into a simulation of a tokamak disruption, the computer would take years to finish the calculation. It's like trying to navigate a city by reading every single street sign in real-time while driving at 100 mph. You'd crash.

The Innovation: The "Smooth Map" (B-Splines)

To solve this, the authors created a clever shortcut. They took their massive, detailed data and compressed it into a smooth, mathematical "rubber sheet" (called a Tensor-Product B-Spline surface).

Imagine you have a crumpled piece of paper covered in millions of data points. Instead of keeping the crumpled paper, you smooth it out into a perfect, continuous curve that passes through all the important points.

• Why this is great: You can now ask the computer, "What happens at this specific temperature and density?" and it can instantly "read" the smooth curve to give you the answer. It's fast, accurate, and easy to use.
• The Result: They created a set of compact "coefficient tables" (like a small instruction manual) that any simulation code can read instantly to get the right physics data.

Why This Matters

This paper is a toolkit for safety.

1. Safety First: By providing accurate data, engineers can design better "fire extinguisher" strategies to protect the massive, expensive tokamak reactors (like ITER) from melting during a disruption.
2. Stopping Runaway Electrons: The data helps ensure that when the plasma cools down, it doesn't create a dangerous beam of "runaway electrons" that could punch a hole through the reactor wall.
3. Community Resource: The authors have made these "smooth maps" available to everyone. It's like giving every physicist a pre-calculated, high-quality map so they don't have to spend years drawing their own.

In a Nutshell

The authors took a messy, chaotic physics problem (how atoms behave in a crashing plasma), solved it with extreme precision using supercomputers, and then compressed that massive solution into a simple, fast-to-use format. This allows scientists to design safer fusion reactors, ensuring that when things go wrong, the "fire extinguishers" work perfectly to save the day.

Technical Summary

1. Problem Statement

Tokamak disruptions pose a severe threat to fusion reactors (like ITER) due to rapid thermal and current quenches. Effective mitigation requires injecting high-Z impurities (e.g., Neon, Argon) to radiate away plasma thermal energy.

• The Challenge: Accurate modeling of the Current Quench (CQ) phase is critical. During this phase, the plasma temperature stabilizes at low values, and the timescale (~100 ms) is long enough to assume a quasi-steady-state condition.
• The Gap: Existing models often rely on simplified approximations, such as the Coronal Equilibrium (CE) model or Superconfiguration models (e.g., FLYCHK). These approximations often fail to capture:
  • Density-dependent effects (crucial at $n_e \ge 10^{15} \text{ cm}^{-3}$).
  • The role of metastable states and transitions between excited states.
  • Three-body recombination processes.
  • Fine-structure details for light elements.
• Computational Bottleneck: High-fidelity Collisional-Radiative (CR) models are computationally expensive, making them difficult to couple directly with time-dependent plasma transport codes used in disruption simulations.

2. Methodology

The authors developed and utilized high-fidelity atomic data to model the plasma parameters for Hydrogen (H), Helium (He), Neon (Ne), and Argon (Ar) over a wide range of tokamak-relevant conditions ($n_e \in [10^{12}, 10^{17}] \text{ cm}^{-3}$, $T_e \in [1, 4\times10^4] \text{ eV}$).

• Codes Employed:
  • ATOMIC: A suite of codes using the LANL atomic physics database. It employs fine-structure resolution for H and He, and configuration-average models for Ne and Ar.
  • FCR (Fusion Collisional-Radiative): A newly developed code incorporating Convergent Close-Coupling (CCC) data and LANL structure data, used specifically for Hydrogen.
  • FLYCHK: Used as a benchmark for comparison, utilizing the superconfiguration approximation.
• Physics Models:
  • Fine-Structure (H, He): Resolved up to principal quantum number $n=10$, including 101 levels for H and 1,046 levels for He.
  • Configuration-Average (Ne, Ar): Used to manage computational complexity for complex systems. The Ne model includes 1,508 configurations; the Ar model includes 11,274 configurations.
  • Processes Included: Electron-impact excitation/ionization, radiative recombination, dielectronic recombination, autoionization, and three-body recombination.
• Data Representation (Surrogate Modeling):
  • To enable efficient coupling with plasma simulations, the authors fitted the precomputed CR data using Tensor-Product B-spline surfaces.
  • The data was transformed into logarithmic space ($x = \log_{10} n_e$, $y = \log_{10} T_e$).
  • An adaptive knot placement strategy was used to capture rapid variations in the data, resulting in compact coefficient tables.

3. Key Contributions

1. High-Fidelity Atomic Data: Generated comprehensive CR data for H, He, Ne, and Ar, covering the specific parameter space relevant to the CQ phase of tokamak disruptions.
2. Model Comparison: Systematically compared fine-structure/configuration-average models against superconfiguration (FLYCHK) and Coronal Equilibrium (CE) models, quantifying the errors introduced by simplifications.
3. Efficient Coupling Framework: Developed a B-spline surrogate representation of the CR data. This allows plasma simulation codes to access high-fidelity atomic data via fast, smooth function evaluations rather than running expensive CR solvers in real-time.
4. Open Data: Made the C++ code and spline coefficient files available for community use and benchmarking.

4. Key Results

• Radiative Power Loss ($P_{rad}$):
  • Density Dependence: Unlike CE models, the full CR models show strong dependence on electron density ($n_e$). At high densities, non-radiative de-excitation channels (like three-body recombination) significantly alter the ionization balance and cooling rates.
  • Model Discrepancies: Superconfiguration models (FLYCHK) and CE models significantly underestimate or misrepresent $P_{rad}$, particularly for Ne and Ar, because they neglect $\Delta n = 0$ transitions and complex autoionization pathways.
  • Structure: The radiative loss curves exhibit "hump-like" structures corresponding to the ionization of specific electron shells (K, L, etc.).
• Average ($Z_{avg}$) and Effective ($Z_{eff}$) Charge States:
  • $Z_{avg}$ and $Z_{eff}$ increase with temperature but show plateaus at closed-shell configurations.
  • At high densities ($n_e \ge 10^{15} \text{ cm}^{-3}$), the population of metastable states increases, leading to higher effective ionization compared to CE predictions.
• Spline Fitting Accuracy:
  • The B-spline fits achieved extremely high accuracy, with $R^2$ scores > 0.999 for all species and parameters (Table I).
  • The Mean Squared Error (MSE) was negligible, confirming the surrogate model preserves the fidelity of the original CR calculations while being computationally lightweight.

5. Significance

• Safety and Design: This work provides the necessary atomic data to accurately predict radiative cooling and runaway electron generation during tokamak disruptions. Accurate $Z_{eff}$ and $P_{rad}$ are essential for designing mitigation strategies that protect the reactor wall and prevent runaway electron beams.
• Operational Regime: The study validates the use of steady-state CR models for the Current Quench phase, provided that relativistic corrections are not required (runaway avoidance regime).
• Community Resource: By providing pre-fitted spline coefficients, the authors remove the computational barrier for integrating high-fidelity atomic physics into large-scale plasma transport codes. This facilitates more realistic simulations of disruption scenarios, aiding in the design of future fusion reactors like ITER and DEMO.
• Model Validation: The paper highlights the limitations of widely used simplified models (CE and Superconfiguration), urging the community to adopt more rigorous CR approaches for high-density, low-temperature plasma regimes.