Online%20Models%20for%20PEP-II

1
Online Models for PEP-II Status

• brief review of PEP modeling
• fully coupled normal form optics representation
• Orbit Response Matrix (ORM) model calibration
  fudge factors for quadrupole strengths
• continuing work

2
PEP Online Modeling Procedure

• prepare input files for MAD
• read reference input files
• set magnets to configuration file values
• apply ORM-derived fudge factors to quadrupole
  strengths
• run MAD
• use XCORs and YCORs to steer to measured absolute
  orbit
• compute effective transfer matrices (RMATs)
• generate model files for MCC
• extract coupled lattice functions from RMATs

3
PEP Online Modeling Process
MCC (VMS)
pepoptics (linux)
Input Files
AT
11
7
8
10
SCP
Matlab
1
5
2
6
9
SCP, SSH
MAD updated to v 8.51/15
4
DIMAD not used anymore
3
4
Lattice parameters in highly coupled systems

• MAD (v 8.51/15-SLAC) has been modified to output
  effective transfer matrices (first order
  expansion about the closed orbit includes feed
  down effects from sextupoles)
• Andy Wolskis normal form analysis1 is used to
  extract coupled lattice parameters from the
  transfer matrices
• 10 coupled lattice parameters (µ, ß, a, ?, ?? for
  modes 1 2) and 8 elements of the normalizing
  transformation (n13, n14, n23, n24, n31, n32,
  n41, n42) at each element are returned to be
  loaded into the MCC database

1See http//www-library.lbl.gov/docs/LBNL/547/74/P
DF/LBNL-54774.pdf
5
ORM analysis LER

• ORM analysis begins with the config lattice
  (actual magnet strengths, steered to the measured
  absolute orbit)
• only quadrupole strength errors are fitted (no
  sextupole strength errors)
• errors are assigned to quadrupole families (power
  supplies)
• sextupole feed down effects are not explicitly
  fitted (even though they are important for LER)
  the assumption is that the BPMs are correct (BBA)
  and that steering to the measured orbit is
  sufficient to model the feed down effects
• fudge factors are computed by comparing the
  fitted quadrupole strengths with their config
  values normal quadrupole fudge factors are
  multiplicative skew quadrupole fudge factors are
  additive (since most skew quads should nominally
  be at or near zero)
• the actual ORM analysis for LER is performed by
  Cristoph Steier (LBNL) using the Matlab-based
  version of the LOCO program

See PTs presentation on Recent ORM Results for
further details
6
ORM-derived fudge factors LER (1)
ORM data taken on December 11, 2003
7
ORM-derived fudge factors LER (2)
fudged
8
ORM-derived fudge factors LER (3)
unfudged
fudged
fudged
unfudged
9
ORM-derived fudge factors LER (4)
10
ORM-derived fudge factors HER (1)

• normal quadrupoles
• skew quadrupoles

ORM data taken on June 10, 2004
11
ORM-derived fudge factors HER (2)
fudged
unfudged
fudged
unfudged
12
ORM-derived fudge factors HER (3)
fudged
unfudged
fudged
unfudged
13
ORM-derived fudge factors HER (4)
14
Continuing work

• steering to absolute orbits when generating the
  model in order to properly account for sextupole
  feed down effects requires accurate knowledge of
  BPM offsets ? BBA1 many offsets for both HER and
  LER have been measured and are being routinely
  used to correct measured orbits LER BBA is
  ongoing (more on this in a minute )
• continue to fine tune the ORM analysis setup to
  avoid degeneracy in the variables
• fudge factors for individual magnets (?)
• develop more robust steering algorithms for model
  generation to take into account bad BPMs
  (especially for LER)
• create fudged design configs move toward
  design optics in both rings
• participate in the ILC design

1See Tonee Smiths presentation on New BBA
Hardware for further details
15
BBA at PEP-II Status

• large unexplained LER BPM offsets from BBA ?
  uncoupled analysis of orbits in a highly coupled
  machine
• new analysis algorithm
• BPM offsets revisited

16
Unexplained large (1 cm) LER BPM offsets from BBA
from Marc Ross summary at April MAC
17
BBA Analysis

• orbit fitting lies at the heart of our BBA
  analysis algorithm
• move the beam in a quadrupole (using a closed
  bump), change the strength of the quadrupole, and
  look at the orbit change
• if the orbit doesnt change when the quadrupole
  strength is changed, the beam is passing through
  the center of the quadrupole the reading on a
  nearby BPM under these conditions is the BPM
  offset
• if youre moving the beam in X, you look at the
  change in the X orbit, which should be
  proportional to the distance (in X) between the
  beam and the quadrupole center in an uncoupled
  ring
• if youre moving the beam in X, and the beam
  happens to be offset in Y to begin with, the
  previous statement remains true in an uncoupled
  ring
• if youre moving the beam in X, and the beam
  happens to be offset in Y, and your ring is
  highly coupled, you have to pay attention to
  whats happening in both planes simultaneously
  (well duh)

18
QDBM3 simulated X data -10 mm Y offset uncoupled
orbit fit
?x (mm)
blue o MAD red - - orbit fit
19
(No Transcript)
20
BBA Analysis
21
Coupled BBA Analysis Algorithm

• Change in closed orbit (?xco,?yco) due to a
  change in strength (K?K(1)) of a
• misaligned quadrupole (xbq,ybq)

• includes closed orbit effects of ?K (both kick
  and position shift)
• includes optics effects of ?K (change in closed
  orbit response matrix)
• fits both planes simultaneously, including any
  known coupling

A. Wolski and F. Zimmerman, Closed Orbit
Response to Quadrupole Strength Variation,
http//www-library.lbl.gov/docs/LBNL/543/60/PDF/LB
NL-54360.pdf
22
QDBM3 simulated X data -10 mm Y offset coupled
orbit fit
?x (mm)
?y (mm)
23
LER BPM X Offsets Then and Now
24
LER BPM Y Offsets Then and Now
25
Acknowledgements thanks!