Skip to content

H⁻ stripping

For H⁻ ion linacs (PIP-II, ESS, J-PARC), the loosely-bound second electron can detach via two mechanisms en route, producing unwanted neutrals or H⁰ that get lost downstream. HELIX has two analytic stripping-rate analyzers.

Magnetic (Lorentz) stripping

Implements Folsom et al. 2021 (Phys. Rev. Accel. Beams 24:074201):

\[ \frac{1}{N}\frac{dN}{ds} = \frac{|B_\perp|}{A_1} \exp\left(-\frac{A_2}{\gamma\,\beta\,c\,|B_\perp|}\right) \quad [\text{1/m}] \]

with empirical constants A₁ = 3.073 × 10⁻⁶ s·V/m, A₂ = 4.414 × 10⁹ V/m (Stinson 1969 / Scherk 1979 fits, also used by TraceWin and PyORBIT).

Mode A: post-tracking deterministic analyzer (no per-particle MC):

from linac_gen.analysis.magnetic_stripping import magnetic_stripping_loss
from linac_gen.core.config import BeamConfig
from linac_gen.core.lattice import Lattice
from linac_gen.core.simulation import Simulation
from linac_gen.distributions.factory import create_beam
from linac_gen.elements.drift import Drift
from linac_gen.elements.solenoid import Solenoid

# H⁻ beam through a small solenoid line (stripping needs B fields):
lattice = Lattice()
lattice.add(Drift(name="D1", length=200.0))
lattice.add(Solenoid(name="SOL1", length=300.0, field=0.30))
lattice.add(Drift(name="D2", length=200.0))
beam_config = BeamConfig(species="H-", n_particles=1000)
beam_cfg = beam_config           # both names used in this chapter
beam = create_beam(beam_config, seed=11)
results = Simulation(lattice, beam).run()

result = magnetic_stripping_loss(results, lattice, beam_config)
# MagStripResult fields (all s-indexed arrays; s in METRES):
#   result.s                      — m
#   result.loss_rate_per_m        — (1/N) dN/ds  [1/m]
#   result.integral_loss          — ∫ rate ds (dimensionless)
#   result.power_loss_per_m_W     — W/m
#   result.integral_power_loss_W  — W

Keyword options: duty_factor (default beam_config.duty_cycle/100), current_mA (default beam_config.current), quad_b_scale ("2sigma" default / "1sigma" / "pole"), fieldmap_sample ("max" default / "on_axis"). Raises ValueError unless beam_config.species == "H-".

The walk:

  1. For each step, look up the element type at element_names[i].
  2. Extract the perpendicular field |B_⊥| — only the component ⟂ to v strips (Folsom Eq. 4). Per element: |B_⊥| = √( B_transverse² + (B_axial·θ_rms)² ), where B_transverse is the dipole bend field, the quad G·r (at 2σ by default), or a field-map B_r/B_x/B_y, and B_axial is an on-axis longitudinal field (a solenoid's B_z, a field-map F_z) that is ∥ to v on axis and only couples through the rms beam divergence θ_rms (from the σ-matrix; a solenoid does not strip a paraxial beam, matching Folsom/TraceWin, which model only dipoles and quads).
  3. Evaluate Folsom Eq. (5).
  4. Integrate: cumulative_loss = 1 − exp(−∫ rate ds).

Intrabeam stripping (IBS)

The bunch's own electromagnetic field can strip electrons via beam-beam collisions. Implements the form-factor model:

from linac_gen.analysis.intrabeam_stripping import ibs_loss

result = ibs_loss(results, beam_config)
# IbsResult fields (s-indexed arrays; s in METRES):
#   result.s               — m
#   result.sigma_strip_cm2 — cross-section used at each step [cm²]
#   result.loss_rate_per_m — (1/N) dN/ds  [1/m]
#   result.integral_loss   — cumulative ∫ rate ds (dimensionless)

Keyword options: duty_factor, current_mA, non_constant_xs (per-step β-dependent σ_H instead of a constant σ_max), and theta_z_convention. The longitudinal velocity coordinate is convention-flagged: the default is "dp_over_p" (the paper's convention); pass theta_z_convention="dv_over_v" to reproduce Ostiguy's override (dv/v = dp/p / γ²). Raises ValueError unless the species is H⁻.

Visualising stripping losses

import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
from linac_gen.analysis.magnetic_stripping import magnetic_stripping_loss
from linac_gen.analysis.intrabeam_stripping import ibs_loss

mag = magnetic_stripping_loss(results, lattice, beam_cfg)
ibs = ibs_loss(results, beam_cfg)

fig, ax = plt.subplots()
ax.plot(mag.s, mag.integral_loss * 100, label="magnetic (Lorentz)")
ax.plot(ibs.s, ibs.integral_loss * 100, label="intrabeam")
ax.set_xlabel("s [m]"); ax.set_ylabel("cumulative loss [%]")
ax.legend()

Both result objects carry s in metres (converted from the recorder's mm internally), so no unit juggling is needed for the x axis.

Engineering decisions

Stripping fraction Implication
< 0.01 % typical PIP-II target — acceptable
0.1 % needs review; likely localised hot spot
> 1 % redesign required

For PIP-II's HEBT, magnetic stripping in the dump dipoles and the beam-bunching cavities are the dominant contributors — both are analyzed here.

Cross-references

  • Aperture profile — the other loss mechanism.
  • linac_gen/analysis/magnetic_stripping.py:1
  • linac_gen/analysis/intrabeam_stripping.py:1

References:

  • Folsom, B. et al. "Stripping mechanisms and remediation for H⁻ beams", Phys. Rev. Accel. Beams 24, 074201 (2021), arXiv:2103.16195.
  • Stinson, G.M. et al. (1969).
  • Ostiguy, J.-F. Internal note on dv/v vs dp/p convention.

← Aperture · Continue to Phase advance →