Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock
Precise and fast 3D space-charge calculations for bunches of charged particles are still of growing importance in recent accelerator designs. A widespread approach is the particle-mesh method computing the potential of a bunch in the rest frame by means of Poisson's equation. Whereas an adaptive discretization of a bunch is often required for efficient space charge calculations in practice, such a technique is not implemented in many computer codes. For instance, the FFT Poisson solver that is often applied allows only an equidistant mesh. An adaptive discretization following the particle density is implemented in the GPT tracking code (General Particle Tracer, Pulsar Physics). The disadvantage of this approach is that jumps in the distribution of particles are not taken into account. In this paper we present a new approach to an adaptive discretization which is based on the multigrid technique. The goal is that the error estimator needed for the adaptive distribution of mesh lines can be calculated directly from the multigrid procedure. The algorithm will be investigated for several particle distributions and compared to that adaptive discretization method implemented in GPT.
TU Darmstadt, Darmstadt
TEMF, TU Darmstadt, Darmstadt
The problem of self-consistent simulations of short relativistic particle bunches in long accelerator structures exhibits a pronounced multi-scale character. The adequate resolution of the THz space charge fields excited by short ultra-relativistic bunches requires mesh spacings in the micrometer range. On the other hand, the discretization of complete accelerator sections using such fine meshes results in a vast number of degrees of freedom. Due to the spatial concentration of the particles and the excited space charge fields, the application of time-adaptive mesh refinement is an emerging idea. We reported on the implementation of time-adaptive mesh refinement for the Finite Integration Technique (FIT)*. Based on this work, we implemented an hp-adaptive discontinuous Galerkin (DG) code. The twofold refinement mechanisms of the hp-adaptive DG method offer maximum modeling freedom. We present details of the h- and p-adaptations for the DG method on Cartesian grids. Special emphasis is put on the stability and efficiency of the adaptation techniques.
DESY Zeuthen, Zeuthen
INRNE, Sofia
The production of electron beams suitable for the successful operation of the European XFEL is studied at the Photo-Injector Test Facility at DESY, Zeuthen site (PITZ). The PITZ beamline is equipped with three dedicated stations for transverse emittance measurements and in the forthcoming shutdown period a section for transverse phase-space tomography diagnostics will be installed. The module contains four observation screens and therefore only four projections can be used in order to reconstruct an underlying phase-space density distribution. This work presents the performance of a number of reconstruction algorithms on limited projection sets using numerical data applied to the PITZ operating conditions. Different concepts for comparison between an original phantom and the reconstructed distribution are presented.