Control of Proton Beam Self-Modulation for AWAKE via Initial Beam Parameters

Plasma wakefield acceleration is a promising concept in the development of a new type of electron accelerators for high-energy physics experiments. The Advanced Wakefield Experiment AWAKE is based on this concept and has been, following the first successful proof of concept Run, under continuous and carefully planned development. In the future Run 2c, the experimental setup will consist of two plasma sources in which the highly relativistic proton driver bunch will undergo seeded self-modulation in the first plasma source, and electron acceleration will be performed in the second. Self-modulation is governed by transverse wakefields which have their origin in the self-modulation instability (SMI). At AWAKE, SMI can be controlled by either laser or electron seeding, which provides a wakefield peak that defines the phase of the fields and that of the emerging micro-bunch train. Consecutive bunches appear in the periodicity given by the plasma frequency such that they coherently drive the longitudinal wakefields. When laser seeding is applied, the part of the bunch in front of the seeding does not travel through plasma. For a successful application of laser seeding, this unmodulated front must not undergo self-modulation in the second plasma source. This front part will have a transverse bunch size of about 500 μm when arriving at the second plasma source. I studied in this thesis the self-modulation of proton beams of different transverse sizes. In the experiment the transverse proton beam sizes at the plasma entrance are measured by a code I developed which provides a description of the proton beam envelope. This characterisation of the proton beam is used to determine the parameter space of the numeric simulations. A detailed explanation of the code and propositions for noise reduction are given in this work. In this thesis for the first time the self-modulation growth rate of beams with transverse sizes larger than in the baseline design will be compared. It will be demonstrated through experiment that a proton beam with a transverse size of ∼560 μm and a length of ∼6.7 cm, focused at the plasma entrance, does not self-modulate in front of the bunch centre. This beam size corresponds, when laser seeding is applied, to the size of the unmodulated front at the second plasma source in Run 2c. These important results indicate that laser seeding could be possible in Run 2c. Before this work, comparisons of experimental results with numeric simulations have only been carried out for beam sizes according to the baseline design. I will show that the simulations performed with a quasi-static PIC code show good qualitative agreement with the experiment when comparing the micro-bunch train structure for a larger beam after seeded self-modulation. It is therefore shown that numerical simulations remain also for larger beams an important tool to discover and understand trends and phenomena. Additionally, I will show a previously unseen band-like radial structure that has been observed in the transverse projections of the beams. This radial structure is observed in numerical simulations after letting the bunch propagate through 10 m of plasma and is further enhanced as the beam size is increased. This could be used for future self-modulation diagnostics.