DPP 2025 Summary

As a California-based corporation dedicated to advancing plasma modeling, fusion engineering, and high-performance simulation technologies, General Physics closely followed the broad scientific landscape presented at this year’s DPP discussions. Across magnetic confinement physics, energetic particles, MHD stability, and turbulence, the community emphasized fundamental questions aligned with our mission: how to optimize confinement, mitigate instabilities, and couple predictive modeling with emerging experimental capabilities. Beginning with progress in negative triangularity (NT) plasmas, researchers from MIT and General Atomics highlighted NT’s promise for ELM-free regimes, improved exhaust handling, and well-defined pedestal transport barriers. Although historically abandoned for MHD-stability concerns, renewed experiments on TCV, ASDEX Upgrade, and DIII-D now reveal improved microstability characteristics across ITG and TEM regimes, along with refined λq scaling. For us as a modeling-focused corporation, these results underscore the value of shape-dependent optimization frameworks that incorporate both equilibrium topology and nonlinear turbulence predictions.

A central theme of the meeting was MHD modeling, where the broad range of talks—from Z-pinch dynamics to advanced eigenvalue solvers—emphasized the multi-scale nature of plasma stability. Presentations describing shock propagation in Z-pinches, inner-layer matching, Riccati-transformed ODE formulations, and hypergeometric analytical treatments reiterated the necessity of accurate dissipative and resistive MHD descriptions. Meanwhile, work from DIII-D and Princeton demonstrated how toroidal angle field measurements, rotation of modes in nonuniform plasmas, and 3D magnetic topology can be combined to design optimized field coils and minimize error fields. Researchers such as Jong-Kyu Park and Matthew Parr further connected quasi-symmetry, coil torque matrices, and Monte-Carlo-augmented reconstruction tools, reinforcing our view that next-generation machines must integrate diagnostics, simulation, and coil optimization into a single workflow—precisely the integrated modeling strategy General Physics develops.

Stellarator and hybrid-topology research also occupied a substantial part of the discussion. New work on the C2A pulsed quasi-axisymmetric stellarator, STRUPHY simulations, and Mercier and ballooning-mode assessments highlights the growing role of flexible stellarator design frameworks. These tools make heavy use of Python-based linear MHD solvers, open-source workflows, and rapid configuration scanning, allowing laboratories to generate thousands of plasma shots for model validation. From our corporate perspective, such reproducibility and rapid design iteration point toward a future where commercial engineering groups can meaningfully collaborate with national laboratories and universities using shared simulation toolchains.

Energetic particles and current-drive physics formed another major thread, especially concerning gyrokinetic modeling of EP geodesics, alpha heating, and mode excitation. The GENE, Tango, CHEASE, and GRILLIX/GENE-X toolsets illustrate how multi-code coupling now allows device-scale predictions in high-field machines like SPARC. The community’s emerging ability to reconcile turbulent fluxes with Fokker–Planck-based classical simulations is crucial for scaling energetic-particle interactions with TAEs, zonal flows, and pedestal evolution—core issues for any future burning-plasma design. These developments align with our ongoing efforts to integrate gyrokinetic solvers and reduced-model transport codes into full-device system simulations.

The interplay between energetic particles and ELM dynamics was explored in depth. Work using MEGA MHD simulations demonstrated how fast ions reshape spatial mode spectra, alter ballooning-mode coupling, and shift pedestal stability boundaries. Combined with FILD measurements and nonlinear multi-n modeling, these studies suggest that EP populations may either suppress or enhance edge instabilities depending on resonance conditions. For General Physics, these findings emphasize the importance of incorporating EP-dependent response functions and resonance-driven transport into our edge-modeling software, especially for customers designing ELM-mitigation strategies.

Equally important was the broad suite of presentations on turbulence and transport, both neoclassical and gyrokinetic. Studies of NT turbulence using XGC, JOREK, CGYRO, and GENE-X confirm that NT plasmas exhibit reduced filamentary behavior, altered radial correlation lengths, and modified skewness statistics. Meanwhile, Type One Energy’s T1E stellarator analysis, W7-X validation efforts, and ARC/ignition-regime scaling studies explore the complex parameter space linking ITG/ETG turbulence, pellet fueling, isotope effects, and core KBM thresholds. These results reinforce a central message: traditional scaling laws lose predictive power in burning plasmas where fast-ion pressure and non-linear saturation dominate. This need for new scaling frameworks parallels the surrogate-modeling and reduced-order transport solvers we continue to develop.

Disruption and runaway-electron physics rounded out the meeting with significant advances in adjoint-formulated Fokker–Planck solvers, PINN-based PDE learning, halo-current reconstruction tools, and 3D tearing-mode simulations. Machine-learning PINNs capable of solving RE dynamics without mesh discretization present a compelling approach for real-time scenario optimization. Additionally, REMC-based mitigation, passive-coil–mediated diffusion enhancement, and stochastic runaway-electron transport modeling provide the engineering path toward safer high-current operation. As a corporation invested in high-reliability modeling software, we see these developments as essential for commercial-scale reactor risk assessment.

Across all topics—triangularity optimization, MHD stability, stellarator design, energetic-particle dynamics, turbulence modeling, and disruption mitigation—the DPP discussions underscored themes central to General Physics’ mission: **the necessity of integrated multi-physics modeling, geometry-dependent optimization, EP-aware stability analysis, AI-accelerated simulation, and validated experiment–theory pipelines**. Negative triangularity’s revival, advanced coil-design optimization, multi-code gyrokinetic frameworks, nonlinear EP–ELM coupling, and emergent ML-driven disruption tools collectively point toward a future where predictive modeling becomes the backbone of fusion development. As a California corporation committed to advancing these capabilities, we view this year’s DPP themes as a clear roadmap for the computational and engineering challenges we must continue to tackle.