Professor Adrian Oeftiger

Abstract:

At CERN we have ramped up a program to investigate space charge effects in the LHC pre-injectors with high brightness beams and long storage times. This is in view of the LIU upgrade project [1] for these accelerators. These studies require massive simulation over large number of turns. To this end we have been looking at all available codes and started collaborations on code development with several laboratories: MAD-X frozen & adaptive mode [2] and integration into the main branch of the MAD-X in-house development [3] code, PyORBIT [4] from SNS, SYNERGIA [5] from Fermilab, MICROMAP [6] from GSI . We have agreed with our collaborators to bench-mark all these codes in the framework of the GSI bench-marking suite [7], in particular the main types of frozen space charge and PIC codes are being tested. We also include a study on the subclass of purely frozen and the adaptive frozen modes both part of MAD-X in comparison with the purely frozen MICROMAP code. Last, we will report on CERN’s code development effort to understand and eventually overcome the noise issue in PIC codes.

Abstract:

The presentation will cover approaches and strategies of modeling and implementing collective effects in modern simulation codes. We will review some of the general approaches to numerically model collective beam dynamics in circular accelerators. We will then look into modern ways of implementing collective effects with a focus on plainness, modularity and flexibility, using the example of the PyHEADTAIL framework, and highlight some of the advantages and drawbacks emerging from this method. To ameliorate one of the main drawbacks, namely a potential loss of performance compared to the classical fully compiled codes, several options for speed improvements will be mentioned and discussed. Finally some examples and applications will be shown together with future plans and perspectives.

Abstract:

Hollow longitudinal phase space distributions have a flat profile and hence reduce the impact of transverse space charge. Dipolar parametric excitation with the phase loop feedback systems provides such hollow distributions under reproducible conditions. We present a procedure to create hollow bunches during the acceleration ramp of CERN’s PS Booster machine with minimal changes to the operational cycle. The improvements during the injection plateau of the downstream Proton Synchrotron are assessed in comparison to standard parabolic bunches.

Abstract:

PyHEADTAIL is a 6D tracking tool developed at CERN to simulate collective effects. We present recent developments of the direct space charge suite, which is available for both the CPU and GPU. A new 3D particle-in-cell solver with open boundary conditions has been implemented. For the transverse plane, there is a semi-analytical Bassetti-Erskine model as well as 2D self-consistent particle-in-cell solvers with both open and closed boundary conditions. For the longitudinal plane, PyHEADTAIL offers line density derivative models. Simulations with these models are bench-marked with experiments at the injection plateau of CERN’s Super Proton Synchrotron.