Speaker
Description
A 300–600 MeV scaling fixed-field alternating gradient (FFA) ring is proposed as a multipurpose proton driver for the China Spallation Neutron Source (CSNS) Phase-II upgrade under strict engineering constraints ($R_0 = 12\text{ m}$, radial beam excursion $\Delta r \le 0.8\text{ m}$, and peak field $B \le 1.66\text{ T}$). To overcome these tight boundaries, a global statistical design framework covering over 360,000 parameter configurations with realistic multisource error budgets was performed. While low super-periodicities suffer from orbit-scallop-induced parameter bottlenecks and severe resonance overlap, an $N=16$ spiral FD doublet lattice suppresses nonlinear multipoles and opens a broad, resonance-sparse stability island. An interpretable machine learning attribution model (XGBoost–SHAP) further confirms that the spiral edge angle ($\zeta$) and the full-ring vertical tune ($Q_z$) are the primary drivers of dynamic aperture (DA) and survival boundaries. Taking into account injection foil integration and magnet cross-talk mitigation, an $N=16$ spiral FD baseline was established and verified, achieving a radial excursion of 0.75 m, a peak field of 1.64 T, and a perturbed DA exceeding $1000\ \pi\cdot\text{mm}\cdot\text{mrad}$