[Paper Review] Beam Loss in Linacs

M. Plum|arXiv (Cornell University)|Feb 17, 2016

Particle accelerators and beam dynamics3 citations

TL;DR

This paper investigates beam loss mechanisms in high-intensity H⁻ and H⁺ linacs, emphasizing the greater complexity in H⁻ systems due to mechanisms like residual gas stripping, H⁺ capture, field stripping, black-body radiation, and intra-beam stripping. Using empirical tuning of quadrupole gradients and RF phases at the SNS, beam loss was reduced by a factor of two, demonstrating that model-based adjustments combined with real-time optimization are essential for minimizing loss in next-generation accelerators.

ABSTRACT

Beam loss is a critical issue in high-intensity accelerators, and much effort is expended during both the design and operation phases to minimize the loss and to keep it to manageable levels. As new accelerators become ever more powerful, beam loss becomes even more critical. Linacs for H- ion beams, such as the one at the Oak Ridge Spallation Neutron Source, have many more loss mechanisms compared to H+ (proton) linacs, such as the one being designed for the European Spallation Neutron Source. Interesting H- beam loss mechanisms include residual gas stripping, H+ capture and acceleration, field stripping, black-body radiation and the recently discovered intra-beam stripping mechanism. Beam halo formation, and ion source or RF turn on/off transients, are examples of beam loss mechanisms that are common for both H+ and H- accelerators. Machine protection systems play an important role in limiting the beam loss.

Motivation & Objective

• To identify and analyze the distinct beam loss mechanisms unique to H⁻ linacs compared to H⁺ linacs.
• To quantify the impact of residual gas stripping, intra-beam stripping, and field stripping on beam loss in high-intensity accelerators.
• To evaluate the effectiveness of empirical tuning of quadrupole gradients and RF phases in minimizing beam loss at the SNS.
• To assess the limitations of simulation codes in predicting beam loss and the need for flexible machine lattices and mitigation systems.
• To establish best practices for beam loss minimization in next-generation spallation neutron sources and proton accelerators.

Proposed method

• Empirical measurement of beam loss as a function of gas pressure in the SNS CCL section to isolate residual gas stripping effects.
• Use of beam profile measurements and wire scanners to characterize beam tails and halo formation in the DTL and HEBT sections.
• Systematic adjustment of quadrupole gradients and RF cavity phases to identify low-loss operating tunes.
• Comparison of beam loss under design-based settings versus empirically optimized settings, particularly in the SCL and HEBT.
• Application of model-based tuning with updated simulation codes to reduce discrepancies between simulation and actual accelerator performance.
• Implementation of vacuum improvements and additional ion pumps to mitigate gas stripping losses, informed by in-situ measurements.

Experimental results

Research questions

• RQ1 What are the primary beam loss mechanisms specific to H⁻ linacs that are absent in H⁺ linacs?
• RQ2 How does residual gas stripping contribute to beam loss, and what is its dependence on beam energy and gas composition?
• RQ3 To what extent can beam loss be reduced through empirical tuning of RF phases and quadrupole gradients in existing linacs?
• RQ4 Why do modern simulation codes fail to accurately predict beam loss, and what design strategies compensate for this limitation?
• RQ5 How do beam halo formation and mismatched Twiss parameters affect beam loss in the DTL and HEBT sections?

Key findings

• Residual gas stripping is a dominant loss mechanism in H⁻ linacs, with beam loss increasing significantly under elevated gas pressures, as demonstrated in the SNS CCL and HEBT sections.
• Empirical tuning of RF phases and quadrupole gradients reduced beam loss by a factor of two at the SNS, with even 1° phase shift causing a doubling of loss at sensitive locations.
• The low-loss operating tune at SNS exhibits non-Gaussian beam tails (up to 30% of peak) from the first wire scanner in the DTL, indicating halo formation despite design-based settings.
• Beam loss in the HEBT injection dump line is highly sensitive to HEBT quadrupole gradient settings, especially at lattice transition regions.
• Intra-beam stripping, field stripping, and black-body radiation are significant H⁻-specific loss mechanisms not present in H⁺ linacs.
• Despite accurate simulation models, discrepancies between predicted and actual beam loss necessitate real-time empirical optimization and flexible lattice designs.