High Transformer Ratio PWFA for Applications on XFELs
W. Lu,Weiming An,Chengkun Huang,C. Joshi,Warren Mori,Mark Hogan,T. Raubenheimer,K. S. Sim,P. Muggli,T. Katsouleas
2009-01-01
Abstract:The fourth generation of light sources (such as LCLS and the XFEL) require high energy electron drivers (1520GeV) of very high quality.We are exploring the possibility of using a high transformer ratio PWFA to meet their challenging requirements. This may have the potential to reduce the size of the electron drivers by a factor of 5 or more, therefore making these light sources much smaller and more affordable. In our proposed design, a high charge (5-10nC) low energy driver (1-3GeV) with an elongated current profile is used to drive a plasma wake in the relativistic blowout regime[1] with a high transformer ratio (5 or more). A second ultra-short beam that has high quality and lower charge ( 1nC) can be loaded into the wake at a proper phase and be accelerated to high energy (5-15GeV) in very short distances (10s of cms). The parameters can be optimized, such that high quality (0.1 percent energy spread and 1mm mrad normalized emittance) and high efficiency (60-80 percent) can be simultaneously achieved. In this paper, we will present some preliminary theoretical analysis and PIC simulations of this concept. HIGH TRANSFORMER RATIO PWFA CONCEPT Recently, PWFA experiments at SLAC have made significant progress. In the E164 experiment, the tail of the 28GeV electron beam had obtained more than 4GeV energy gain in a 10 centimeter long self-ionized plasma[2]. Later, in the E167 experiment, the tail of the 42GeV electron beam had obtained more than 42GeV energy gain in less than one meter[3]. These experiments show the great potential of PWFA as an effective method of energy transfer between charged particle bunches at least in sub 100 GeV level. In the future development of PWFA, using shaped particle bunches to optimize the acceleration and efficiency are very essential. In the long term, concepts such as PWFA based linear collider (PWFA-LC) can be explored, which may have impact on high energy physics. In the near term, we feel that it is also very interesting to explore how to shape the bunches to achieve high transformer ratio and high efficiency PWFA such that the size and cost of 10GeV level LINAC can be reduced significantly. This kind of PWFA may find applications in X-ray free electron lasers. For example, the fourth generation of light sources ( such as LCLS and the XFEL) require high energy electron drivers (15-20GeV) of very high quality ( charge 1nC, peak current 5kA, emittance 1mm mrad and energy spread ∗Work supported by DOE under contracts DE-FG03-92ER40727, DEFG52-06NA26195, DE-FC02-07ER41500, DE-FG02-03ER54721. 0.1percent), therefore a high transformer ratio PWFA that can meet these challenging requirements may have the potential to make the XFELs more compact and more affordable. The concept of large transformer ratio PWFA has been around for more than 20 years. The basic idea is to use a low energy high charge driver to accelerate a low charge trailing bunch to much higher energy. For this to happen, the transformer ratio must be high (the ratio between the accelerating field and the decelerating field). Furthermore, in order to achieve high transfer efficiency and low energy spread at the same time, the plasma wake must have uniform accelerating/decelerating fields within the bunches. In the linear regime, the optimal parameters have been explored by using the linear plasma wakefield theory. It can be shown using 1D linear theory, that the transformer ratio of the wakefield from a symmetric longitudinal beam profile is always less than two[4]. To increase the transformer ratio, an asymmetric profile should be used[5]. A particularly interesting example to achieve large transformer ratio as well as uniform decelerating field is to use a linearly ramped beam profile with a sharp termination and a δ function precursor[5]. Later, it was also shown that larger transformer ratios can be obtained in the 1D nonlinear regime[7]. However, it is now very well understood that the 1D linear and nonlinear regime are not easily accessible experimentally, instead a 3D highly nonlinear regime where plasma electrons blowout occurs is normally the case[1]. So it is natural to ask the same kind of questions for this highly nonlinear 3D regime: what kind of bunch shapes can lead to constant accelerating/decelerating fields and a high transformer ratio in the relativistic blowout regime? In the second section, we will answer this question through a simple theory and PIC simulations. The results show that a linear ramped profile is also the choice in the blowout regime, although with very different physical mechanism comparing with the linear case. In the third section, we will show a sample QuickPIC simulation based on the simple theory. The parameters of this simulation are relevant to the XFEL application. A SIMPLE THEORY OF TRANSFORMER RATIO IN THE BLOWOUT REGIME We can start from the hint of the linear theory. For a linearly ramped drive beam in the blowout regime, the transition from the 1D linear regime to the 3D nonlinear blowout regime can proceed as follows by assuming a fixed longitudinal current profile and fixed total charge. Initially, one can choose a large spot size such that nb/np 1 will occur, therefore reaching the blowout regime. So will the special properties based on predictions from linear theory still hold when blowout is reached? will they still have the same transformer ratio? We can get some insight from PIC simulations. In Fig.1, a linearly ramped electron beam driver is used to excite the plasma wake in the relativistic blowout regime. This beam has a normalized length kpL0 ≈ 22 and a flat top transverse profile (nb0 = 100np for kpr > 1, where rb(ξ) is the normalized blowout radius at each longitudinal slice. For a long electron beam with an adiabatically increasing current profile Λ(ξ) = ξ L0 Λ0, where Λ0 is calculated at the maximum beam density, the blowout radius rb(ξ) also increases adiabatically, and can be estimated from the theoretical formula kprb(ξ) ≈ 2 √ Λ(ξ) (2) With this expression of rb, we get ψ(0, ξ) ≈ ξ L0 Λ0 (3)