Difference between revisions of "Model April 12 2016"
(4 intermediate revisions by the same user not shown) | |||
Line 43: | Line 43: | ||
realistic gradient distribution and rematch all the spreaders. | realistic gradient distribution and rematch all the spreaders. | ||
− | Currently studying the implications of this. See attached | + | Currently studying the implications of this. See attached [[Media:Lemrffocusing04122016.pdf | powerpoint slides]] |
Latest revision as of 17:23, 12 April 2016
Previous action items (see march 8th minutes )
- ATLIS created to study FOPT versus 30 Hz. (D.T) done.
Found that the RF cavities attached to the C100 zone equipped with the vernier
become unstable when using the AC vernier mode.
Some of them have increasing GMES, some have decreasing GMES.
It might be wise to switch the vernier back to a C50 zone.
- Future release of eDT will set the separator angle. (D.T) Done
- eDT will load setpoint2 for the MCP1P01 (amd MCP3P01) magnet to the actual BDL needed for the chicane (D.T) Done
- 2R optics error. a FOPT custom kick file will be prepared to perform a LOCO of the area during beam studies (Y.R)
Contents
girders in the model
In the real machine, quadrupoles are arranged on girders (With a few exceptions) featuring a specific layout of diagnostics and correctors.
There is a finite number of girder types. We discussed whether one should capture this in the model.
The first step would be to write scripts to check that the element placement is indeed accurately captured.
diagnostic and corrector lengths
Currently we captured the diagnostics and correctors as zero length elements.
For correctors, this location in the center of it, for the bpms, the location in that of the antennas (3 inches from end of can).
There are benefits to actually model the true length of all of these elements. This is a significant undertaking and likely will have to be done
via scripting to post process the existing decks. It was felt that while it is useful, it is a medium to long term project.
M56 cancelation in ARCS
Recent observations show that there is some remaining M56 content in the arcs, probably in the early arcs. the arc models have exactly zero M56.
Hence, it is likely that the source of M56 observed in the machine is coming from an imperfect knowledge of the quadrupole gradients present in the dipoles.
One can study this effect in ELEGANT and attempt to reconcile it with the measurements. Another thing to do is to check that by using the M56 quadrupoles one can correct
for this without destroying the match.
dipole and quads lengths
We have a few locations in the machine for which the dipole elements have an erroneous length.
i.e one should capture the trajectory length which itself is derived from steel length plus extra fringe field length obtained from HGAP.
One needs to review the model to adjust it wherever we have a deviation.
. Quadrupoles also need to be reviewed. The focal lengths and placement is accurate, however one might want to adjust slightly their total length (and
therefore their setpoints ) to capture the fringe fields.
Effect of linac gradient distribution on the transverse match
The design model assumes a simple gradient distribution where all the cavities are set at 3.6875 MeV and the C100 are set at 12.5 MeV yielding 1090 MeV of total acceleration.
An utility was written by D. Turner to retrieve the actual distribution from the machine and load it into ELEGANT (lem2sdds). Using this, one can set the design model to use a
realistic gradient distribution and rematch all the spreaders.
Currently studying the implications of this. See attached powerpoint slides