Geometry of Almost-Conserved Quantities in Symplectic Maps. Part III: Approximate Invariants in Nonlinear Accelerator Systems

1. Problem Statement

The fundamental theory of motion in ring-shaped particle accelerators has historically relied on linear dynamics, particularly the Courant-Snyder (CS) theory, developed in the 1950s. Although the CS theory provides a powerful framework for systems only weakly perturbed from harmonic oscillations, it is increasingly insufficient for modern accelerator challenges, including:

• Higher beam intensities and smaller emittances.
• The ubiquitous presence of strong nonlinear magnetic elements (e.g., in the Integrable Optics Test Accelerator (IOTA), Muon Collider, and FCC).
• The need for tight stability margins.

Existing methods for handling nonlinearities, such as Lie-algebraic normal forms or resonant normal forms, suffer from specific limitations:

• Ambiguity: They often yield multiple invariants of motion conserved to the same order, requiring manual selection of a specific normal form (often by isolating a single resonance).
• Computational Cost: They can be computationally intensive and less transparent.
• Limited Scope: Traditional approaches often struggle to describe global stability boundaries or handle multiple resonances simultaneously without losing accuracy.

There is a critical need for a general, analytical extension of CS theory to strongly nonlinear regimes that is conceptually transparent, computationally efficient, and capable of describing global phase space structures (islands, separatrices) without requiring extensive particle simulations.

2. Methodology

The authors propose a systematic perturbative framework for constructing approximate invariants of motion directly from discrete symplectic maps (one-turn maps).

Basic Principles:

• Approximate Invariance: Instead of seeking an exact invariant $K(T(p,q)) - K(p,q) = 0$ (which exists only for integrable systems), the method seeks a function $K^{(n)}$ such that the deviation is of order $O(\epsilon^{n+1})$, where $\epsilon$ is a small parameter related to the oscillation amplitude.
• Polynomial Approach: The approximate invariant is constructed as a series of homogeneous polynomials:
  $$K^{(n)}[p, q] = K_0 + \epsilon K_1 + \dots + \epsilon^n K_n$$
  where $K_0$ is the standard linear Courant-Snyder invariant.
• Averaging Strategy: To resolve the underdetermination of coefficients (where the invariant is known only up to arbitrary powers of $K_0$), the authors introduce an averaging procedure.
  • The map is transformed into the eigenbasis (action-angle variables).
  • The residual error $R_n$ is integrated over the angle variable $\psi$.
  • The coefficients are determined by minimizing the fluctuation of the invariant: $\frac{d}{dC_k} \int_0^{2\pi} R_n^2 d\psi = 0$.
• Reversibility Symmetry: The method explicitly preserves discrete reversibility symmetries up to the perturbative order and acts as a discrete analogue to Noether's theorem. This ensures geometric accuracy and consistency with the system's symmetries.

Key Technical Steps:

1. Linearization: Derivation of the zeroth-order invariant ($K_0$) from the linear part of the one-turn map (Twiss parameters).
2. Stepwise Construction: Solving for higher-order coefficients ($K_1, K_2, \dots$) by enforcing the conservation condition up to the desired order.
3. Averaging: Application of residual fluctuation minimization to fix arbitrary coefficients and effectively handle rational tune values and singular resonances.
4. Extraction of Nonlinear Tune: Use of Danilov's Theorem to extract the nonlinear betatron tune ($\nu_n$) and action-angle variables directly from the constructed invariant.

3. Main Contributions

• Nonlinear Generalization of CS Theory: The method provides a compact, closed-form function that generalizes the Courant-Snyder invariant to nonlinear regimes. It summarizes the cumulative effect of all nonlinear lattice elements into a single approximate invariant.
• Handling Multiple Resonances: Unlike traditional resonant normal forms that isolate specific resonances, this averaging technique retains multiple resonances up to a given perturbative order. This enables accurate description of complex phase space structures, including island chains and separatrices.
• Computational Efficiency: The method requires no particle simulation or extensive numerical simulations. It relies solely on the coefficients of the discrete time transformation (the one-turn map) and enables rapid diagnostics.
• Prediction of Global Stability: The approach successfully reconstructs global phase space structures (stability boundaries, chaotic regions) often overlooked by local linearization or isolated resonance approximations.
• Scalability: Although demonstrated for 1D systems, the framework is theoretically extendable to higher dimensions (coupled motion) and offers a path for analyzing complex systems with multiple degrees of freedom.

4. Results and Applications

The authors validated the method using realistic dynamic models from three different facilities at the Fermi National Accelerator Laboratory (FermiLab):

1. Mu2e Delivery Ring (Slow Extraction):
   • Scenario: Dominance of a singular third-order resonance utilized for slow extraction.
   • Result: A low-order approximation ($n=4$) showed excellent agreement with full numerical tracking and accurately captured the structure of the resonance island and stability boundaries.
2. Integrable Optics Test Accelerator (IOTA):
   • Scenario: Features of a non-singular third-order resonance driven by octupole-like nonlinearities, characterized by a chain of islands.
   • Result: The method ($n=7$) successfully reproduced the characteristic chain of islands and the global phase space topology, matching numerical trajectories perfectly.
3. Main Injector (MI):
   • Scenario: Exhibits high-order resonances and chaotic behavior.
   • Result: Even a sixth-order approximation ($n=6$) accurately captured the stability region and general phase space structure, outperforming traditional methods in global stability analysis.

Comparison with Existing Methods:

• Compared to the Square-Matrix Method (equivalent to full normalization), the proposed method showed comparable local accuracy but superior global stability and resonance performance.
• It avoids the problem of the "vanishing domain" of higher-order twist expansions in non-integrable systems by providing a coherent expression valid over a broad parameter range.

5. Significance

This work introduces a new paradigm for describing nonlinearity in accelerator physics:

• Diagnostic Power: It enables rapid, interpretable diagnostics of nonlinear behavior, allowing physicists to visualize stability boundaries and identify resonant structures immediately without performing time-consuming simulations.
• Design and Optimization: By reducing complex nonlinear lattices to a series of approximate invariants (analogous to Twiss parameters in linear theory), the method offers a scalable tool for nonlinear design, optimization, and control of next-generation machines (e.g., Muon Collider, FCC).
• Theoretical Bridge: It serves as a conceptual and computational bridge between practical machine design and modern dynamical systems theory, offering a geometrically transparent understanding of almost-conserved quantities in symplectic maps.

In summary, the paper presents a robust, efficient, and conceptually clear method to extend the century-old Courant-Snyder formalism into the nonlinear regime, providing a powerful tool for the analysis and optimization of modern high-performance particle accelerators.