Summary notes of the LCU meeting on 21/10/2008

Present: IA, MA, RB, HB, RC, MG, NH, JJ, YL, AM, TR, FR, FS, RT, SW, SY, FZ, W. Fischer/BNL

Report from meetings

• SLM and general information -> MG
  Chamonix 2009 is back on the schedule, now for the first week of February. The organisation will be presented at the next LHCPerC meeting.
  The repair of sector 3-4 will be organized as project led by R. Schmidt
  for the other sectors the activities will be co-ordinated by team with S. Baird (shutdown co-ordination), R. Bailey (cold check out), and R. Schmid (hardware commissioning).
  A special introductory level CERN accelerator school will be organised in March 2009 in Divonne. It will be targeted mostly for newly hired technical engineers.
  There has been further follow up on the clarification of hardware versus MAD-X sign conventions involving S. Russenschuck. A feedback concerning the correctness of the polarity flag is awaited.
  There has also been more work on consistency of apertures and layouts together with TR.

Analysis of LHC Injection into sector 78 data -> YS (pdf)

YS analyses data taken on the 24 August 2008 during the injection tests into LHC sector 78. Trajectories were recorded for large initial kicks. This allows checks of signs, not only of corrector magnets but in some cases also for quadrupoles and sextupoles circutis. While most corrector signs were found as expected, some further analysis and checks with the actual settings as recorded in the settings database will be needed to clarify the signs in few cases. In particular, the comparison between the results obtained using Beam2 and Beam 4 requires still clarifications. For a number of circuits a good agreement between measurement and model is obtained in one plane, but not in the other.

Status of the Thin-lens Tracking module in MAD-X -> YS (pdf)
YS reported from work with MAD-X thin lens tracking. Thin lens tracking with MAD-X is well adapted for injection studies and commissioning with few turns. Thin lens tracking in MAD-X was improved by A. Koschik. Among the remaining things to do is to find the 6d closed orbit using the computed by Twiss as an initial guess, only. The crucial point is that the physics implemented in the Twiss and Tracking module is not the same. Hence, exporting results between the two module is not correct. The implementation of the closed orbit computation is still in progress.
YS looked closer to the MAD-X tracking routine trrun.F and the use of apertures in MAD-X. Indeed, the racetrack shape was not implemented in the tracking module. This is fixed, now.
It is raised the question whether all the aperture types available in the aperture module are indeed useful and should be implemented in the rest of the code. the general opinion is that the basic shapes (circle, ellipse, rectangle) should be kept together with rectellipse and racetrack. Aperture types like lhcscreen (not even used by us) and marguerite might be frozen. This will be followed up by FS with JBJ.

FS reported that while working on the Beam2 and Beam 4 conversion, TR found a bug in the micado module. This turned out to be generated by an utomatic translation from FORTRAN66 (original code) to FORTRAN77 (current code in MADX). FS also reports the proposal by WH to change the default for closed orbit correction (it is currently micado). This proposal did not received particular support and it was decided to leave the situation as it is now.
MG reports that in a future release of MADX the thresholds for SVD correction will be adjustable parameters.

SAD-MADX converter -> AM (pdf)
AM looked into the possibility of automatic translation between optics descriptions for MAD-X and SAD. He explained that this is not straight forward due to some fundamental differences between MAD-X and SAD. Different conventions for the field strength are used, but more than this, SAD allows using fancier field configurations in a lattice. Furthermore, SAD allows for offsets as element attributes and for overlapping elements, which MAD-X does not. This explains why a one-to-one translation is in general not possible. In fact, the additional functionality in SAD makes it hard to translate SAD input in to MAD-X input. On the other hand, MAD-X to SAD conversion is easier. Essentially, the convertor developed allows generating a SAD lattice for LHC.
MG asks about the status of the beta* tuning knob, which were developed in SAD. AM replies that the core of the knobs is the matching algorithm and it cannot be converted into MADX. FZ suggests to document the algorithm used and mentioned that YS could take care of preparing the beta* tuning knobs. This proposal is retained.

AOB

None

Last update: 22-October-08

MG & HB

Back to LCU Minutes