BNL, Upton, Long Island, New York
In the quest for brigthness, the horizontal emittance remains one of the main performance parameters for modern synchrotron light sources. A control theory approach that takes the nonlinear dynamics aspects into account, by a few simple (linear) optics guidelines, at an early stage generates robust designs. Modern analytic- and computational techniques enables the optics designer to avoid the fallacy of the traditional approach guided by the Theoretical Minimum Emittance (TME) cell: the "chromaticity wall". In particular, by an interleaved computational approach with the nonlinear dynamics analyst/model. We also outline how to implement the correction algorithms for a realistic model so that they can be re-used as part of an on-line model/control server for commissioning- and operations of the real system.
UCLA, Los Angeles, California
New emerging multi-core technologies can achieve high performance, but algorithms often need to be redesigned to make effective use of these processors. We will describe a new approach to Particle-in-Cell (PIC) codes and discuss its application to Graphical Processing Units (GPUs). We will conclude with lessons learned that can be applied to other problems. Some of these lessons will be familiar to those who have programmed vector processors in the past, others will be new.
CERN, Geneva
After a intense and hectic code development during the LHC design phase the MAD-X program (Methodical Accelerator Design – Version X) is going through a period of code consolidation. To this end the development on the core has been frozen and most effort are concerned with a solid debugging in view of a trustworthy production version for the LHC commissioning. On the other hand, the demand on further code development from the LHC pre-accelerators and CLIC are dealt with PTC related parts of the code where the implementation is in full swing. Having reached a mature state of the code the question arises what kind of future can be envisaged for MAD-X.
ORNL, Oak Ridge, Tennessee
SLAC, Menlo Park, California
XAL is a Java programming framework for building high-level control applications related to accelerator physics. The core of XAL consists of a GUI framework to provide common “look and feel” and functionality for all XAL applications, a hardware representation of the machine for connectivity and control, and a beam simulation model termed the "online model" for model reference and comparison to the hardware operation. The structure, details of implementation, and interaction between these components, auxiliary XAL packages, and applications are discussed. A general overview of applications created for the SNS project and based on XAL is presented.
BNL, Upton, Long Island, New York
The on-line models for RHIC and the RHIC pre-injectors (the AGS and the AGS Booster) can be thought of as containing our best collective knowledge of these accelerators. As we improve these on-line models we are building the framework to have a sophisticated model-based controls system. Currently the RHIC on-line model is an integral part of the controls system, providing the interface for tune control, chromaticity control, and non-linear chromaticity control. What we will discuss in this paper is our vision of the future of the on-line model environment for RHIC and the RHIC pre-injectors. Although primarily these on-line models are used as Courant-Snyder parameter calculators using live machine settings, we envision expanding these environments to encompass many other problem domains. We will also discuss the importance of the modeling infrastructure and organization as well as interfacing to controls, power supply, and magnetic measurement infrastructure and organizations. The model engines themselves will be discussed and our own evolution toward incorporating more sophisticated simulation filters, such as PTC and UAL, into the on-line model infrastructure.
BNL, Upton, Long Island, New York
LBNL, Berkeley, California
Multiobjective Genetic Algorithm(MOGA) is a technique for optimization, and based on the populations and natural selections. We have integrated this algorithm with linear lattice calculation code to make the lattice design more robust.
RadiaBeam, Marina del Rey
UCLA, Los Angeles, California
PSI, Villigen
The development of the code RadTrack is based on the need to model accelerator system diagnostics. The code is built using a modular approach with a strong emphasis on intuitive user interface. The operations of trajectory calculation and radiation field solving are segregated; currently the tracking is handled by Q-Tracker and the field solving is executed by a modified version of QUINDI. Additionally, the RadTrack user interface allows for seamless start-to-end stitching of I/O exchange between certain codes, and the visualization canvas reinforces user directives in a near-real-time environment.
SLAC, Menlo Park, California
An x-ray Free-Electron Laser (FEL) calls for a high brightness electron beam. Generically, such a beam needs to be accelerated to high energy on the GeV level and compressed down to tens of microns, if not a few microns. The very bright electron beam required for the FEL has to be stable and the high quality of the electron beam has to be preserved during the acceleration and bunch compression. With a newly developed model independent global optimizer [*], here we report study for the control and error diagnostics of such a generic machine: magnetic elements, and RF cavities, and the electron beam parameters: the peak current, centroid energy, and trajectory. Collective effects, such as coherent synchrotron radiation, space charge, and various wakefields are incorporated in a parametric approach. Applicability and verification are detailed for the LINAC Coherent Light Source, an x-ray FEL project being commissioned at SLAC.
KIT, Karlsruhe
FZK, Karlsruhe
ANKA is a synchrotron radiation source located in Karlsruhe, Germany. While the control system has always provided access to technical parameters, like power supply currents or RF frequency, direct access to physical parameters like tune or chromaticity has been missing. Thus the operator has to change and monitor the technical parameters manually and to calculate the physical parameters using separate tools. Therefore effort has been made to integrate the monitoring of physical parameters and simulation tools into the control system. At ANKA the MATLAB-based Accelerator Toolbox is used for simulation purposes, however the control system framework ("ACS") does not support MATLAB natively. For this reason, a software bridge has been created, which provides direct access to control system components from MATLAB. Thus operators can write their own MATLAB code simultaneously using simulation code and components from the control system. This system has already been used to automate measurements, thus allowing unattended long-term measurements, which have not been possible before. Future plans include creating a graphical user interface and various monitoring and stabilization loops.
LBNL, Berkeley, California
Tracy is an accelerator modeling and simulation code originally developed at LBNL in Pascal two decades ago*. Tracy evolved to Tracy2** which served as the basis for several derivative codes at other synchrotron light sources, including PSI, SSRL and Soleil. In most of these cases, the accelerator physics library was extracted and translated in C. At the ALS the library was re-written in C++ (Goemon***) in an object-oriented manner. Later this version was converted to C# with some effort spent on optimizing its performance****. Tracy# is the latest C# version upgraded to take advantage of the new features of the .NET Framework 3.5 and 4.0. It efficiently uses the modern language features of the C# and the standardized libraries of the .NET Framework for database, XML and networking. It also works with other .NET languages, such as IronPython and F# for interactive scripting. Although it is developed on Windows, MONO makes it portable to other operating systems including Linux.
ANL, Argonne
Use of SDDS, the Self-Describing Data Sets file protocol and toolkit, has been a great benefit to development of several accelerator simulation codes. However, the serial nature of SDDS was found to be a bottleneck for SDDS-compliant simulation programs such as parallel elegant. A parallel version of SDDS would be expected to yield significant dividends for runs involving large numbers of simulation particles. In this paper, we present a parallel interface for reading and writing SDDS files. This interface is derived from serial SDDS with minimal changes, but defines semantics for parallel access and is tailored for high performance. The underlying parallel IO is built on MPI-IO. The performance of parallel SDDS and parallel HDF5 are studied and compared. Our tests indicate better scalability of parallel SDDS compared to HDF5. We see significant I/O performance improvement with this parallel SDDS interface.
Fermilab, Batavia
The Recycler is a fixed 8 GeV kinetic energy storage ring using permanent gradient magnets. A phase trombone straight section is used to control the tunes. For ProjectX , the H-particle extracted from the Linac will be striped and painted in the Recycler Ring and then the protons will be extracted into the Main injector. A long drifting space is needed to accommodate the injection chicane with stripping foils. In this paper, the existing FODO lattice in rr10 straight section being converted into doublet will be described. Due to this change, the phase trombone straight section has to be modified to bring the tunes to the nominal working point. On the other hand, a toy lattice of recycler ring is designed to simulate the end-shim effects of each permanent gradient magnet to add the flexibility to handle the tune shift to the lattice during the operation of 1.6·1014 with KV distribution of the proton beam to give ~0.05 of space charge tune shift . The comparison or the combinations of the two modification ways for the Recycler ring lattice will be presented also in this paper.