[PyTokMHD] 2D Morrison bracket limits realistic 3D scenarios

[callme-YZ]

Problem

v2.0 uses 2D Morrison bracket, limiting 3D realism

Current design:

• 2D bracket: (r, θ) spatial grid
• z-direction: Fourier modes (periodic)
• Toroidal but not truly 3D

Code evidence:

• toroidal_bracket.py - 2D operators
• z handled via FFT (spectral)

Limitations

Cannot model:

1. ❌ Full 3D magnetic reconnection
2. ❌ Realistic divertor geometry (X-point in poloidal plane only)
3. ❌ 3D RMP coil fields (simplified to 2D + Fourier)
4. ❌ Edge localized modes (ELMs) with full 3D structure

Can model:

• ✅ Ballooning modes (2D + periodic is sufficient)
• ✅ Toroidal coupling (ε∂/∂φ terms included)
• ✅ RMP control (simplified 3D forcing)

Impact on v3.0

If v3.0 goal = better ballooning benchmark:

• ✅ Current 2D framework sufficient
• ⏸️ This issue can be deferred

If v3.0 goal = realistic EAST/DIII-D scenarios:

• ⚠️ May need full 3D
• Depends on which physics phenomena to model

Proposed Solution

Option A: Full 3D grid (Major upgrade)

• Replace 2D + Fourier with 3D spatial grid
• Effort: 2-3 months
• Benefit: General 3D capability
• Cost: Memory × N_z, compute × N_z

Option B: Keep 2D + enhance Fourier

• Better z-resolution (more modes)
• Improved spectral operators
• Effort: 2-3 weeks
• Benefit: Better accuracy within 2D limit
• Limitation: Still fundamentally 2D

Option C: Defer to v4.0

• v3.0 focuses on 2D ballooning (proven strength)
• Full 3D as major future version
• Recommended ✅

Decision Gate

YZ decision needed: v3.0 scenarios require full 3D?

• If NO → Close as deferred ✅
• If YES → Choose Option A or B

Priority

P3-low 🟢 - Current 2D sufficient for ballooning

Defer unless v3.0 explicitly requires 3D ⏸️

---

Reported by: 小P ⚛️
Date: 2026-03-23