This website is no longer maintained. Its content may be obsolete. Please visit http://home.cern for current CERN information.
The basis for all LHC machine settings are the data generated by the offline program MAD (Methodical Accelerator Design) - a large and complex suite of software. This program takes as input a sequence file which describes the layout of the accelerator and a set of files containing the strengths of all magnetic elements. Typical basic output includes the Twiss parameters (describing the beam characteristics) and the strengths in all magnetic elements. This output is imported into the controls database and forms the basis of the machine settings generation process. Each instance resulting from a MAD run is known in this context as an "optics".
At present the generation of the MAD output files is done manually. The import into the controls database is done via the manual invocation of a PERL script. Error handling is very rudimentary. The number of optics files could eventually number in the low hundreds. Careful management of the optics generation and transfer is required.
Work to be performed in the LSA programming environment: Java, Spring, Oracle, Eclipse, CVS. Linux and/or Windows.
1. Provide an interface to configure and run MAD on-line, providing the user with a set of appropriate choices (sequence files, strength files, etc.). Provide error reporting from the MAD execution. Stream output files to a designated repository. Provide environment for MAD on-line execution. (In this context MAD is run in its simplest possible mode - there are clear opportunities for extensions.)
Estimate: 3 man months (with some lead time for familiarization)
2. Provide an interface to select MAD output files and load to the controls database. All exceptions and errors to be handled appropriately.
3. In the above task ensure full database consistency including cross-check of machine element configuration.
Estimate: 1-2 man months
4. Improve existing settings generation interface to allow easy selection of the available optics. Extend existing sofware to provide generation for different equipment classes. The list of different equipment classes is finite, but this is relatively complex domain with options for extensibility.
Estimate: 2 ++ months.
Task 1-3 form a well-defined work package which should cover a 6 month period. Task 4 indicates that there are possibilities for extensions within this domain.