Author: Potratz, C.
Paper Title Page
MOPC080 First Considerations Concerning an Optimized Cavity Design for the Main Linac of BERLinPro 259
 
  • B. Riemann, T. Weis
    DELTA, Dortmund, Germany
  • W. Anders, J. Knobloch, A. Neumann
    HZB, Berlin, Germany
  • H.-W. Glock, C. Potratz, U. van Rienen
    Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock, Germany
  • F. Marhauser
    JLAB, Newport News, Virginia, USA
 
  Funding: work supported by BMBF under contracts 05K10PEA and 05K10HRC
The Berlin Energy Recovery Linac Project (BERLinPro) is designed to develop and demonstrate CW linac technology and expertise required to drive next-generation Energy Recovery Linacs. Strongly HOM-damped multicell 1.3 GHz cavities are required for the main linac. The optimization of the cavities presented here is primarily based on the CEBAF 1.5 GHz 5-cell high-current cavity design, including HOM waveguide couplers. The cavity was scaled to 1.3 GHz and extended to 7 cells. Modifications to the end group design have also been studied. An effort was also made to reduce the ratio Epk/Eacc while still permitting HOMs to propagate.
 
 
WEPC098 Automatic Pole and Q-Value Extraction for RF Structures 2241
 
  • C. Potratz, H.-W. Glock, U. van Rienen
    Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock, Germany
  • F. Marhauser
    JLAB, Newport News, Virginia, USA
 
  The experimental characterization of RF structures like accelerating cavities often demands for measuring resonant frequencies of Eigenmodes and corresponding (loaded) Q-values over a wide spectral range. A common procedure to determine the Q-values is the -3dB method, which works well for isolated poles, but may not be applicable directly in case of multiple poles residing in close proximity (e.g. for adjacent transverse modes differing by polarization). Although alternative methods may be used in such cases, this often comes at the expense of inherent systematic errors. We have developed an automation algorithm, which not only speeds up the measurement time significantly, but is also able to extract Eigenfrequencies and Q-values both for well isolated and overlapping poles. At the same time the measurement accuracy may be improved as a major benefit. To utilize this procedure merely complex scattering parameters have to be recorded for the spectral range of interest. In this paper we present the proposed algorithm applied to experimental data recorded for superconducting higher-order-mode damped multi-cell cavities as an application of high importance.  
 
WEPC099 Coupler Design and Optimization by GPU-Accelerated DG-FEM 2244
 
  • C. Potratz, H.-W. Glock, U. van Rienen
    Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock, Germany
 
  The numerical optimization of rf-components like couplers is a common task during the design phase of particle accelerators. Typically, these optimizations involve the simulation of a multitude of very similar structures with minor geometric variations. Nevertheless, this process is in its entire extend rather demanding on both the invested time and hardware budget. With recent advancements in the field of numerical electromagnetic field simulation and consumer graphic processors, an interesting alternative for the time-consuming simulation part of the optimization is available. In this contribution we show, how the Discontinuous Galerkin FEM method in conjunction with consumer graphic cards can be used to build moderately prized cluster solutions for the parallel simulation of rf-components. The contribution will mainly focus on, but is not limited to, Higher Order Mode couplers as a typical application example, where the DG-FEM method accelerated by a graphic processor might be used to significantly reduce the overall time necessary for the optimization.