Date: Thu, 09 Sep 2004 10:18:29 -0500
From: Jim Birchall 
To: Juliette Mammei , Roger Carlini ,
     Klaus Grimm , Mark Pitt ,
     Neven Simicevic , Greg Smith ,
     Shelley Page , Tony Forest ,
     Norman Morgan , Mike Finn ,
     David Armstrong , Dave Mack ,
     Yongguang Liang ,
     Allena Opper ,
     Richard Jones 
Subject: Tilting the coils
Parts/Attachments:
   1 Shown    33 lines  Text
   2   OK     83 KB     Application
----------------------------------------
                                                                                
The following pdf contains a couple of diagrams that summarize results
of tilting the upstream end of the QTOR coils in and out radially,
moving the magnet relative to the collimator and shortening and
lengthening QTOR. I showed this at Santa Fe.
                                                                                
The first page shows the standard deviation of a Gaussian fit to the
distribution across the width of a bar vertically and the position of
the beam downstream of the centre of QTOR horizontally. The full width
at the base is about 6.1*sigma. The coils (all of them) were rotated up
to 4 degrees outward at the upstream end, keeping the downstream end
fixed; coils were moved out up to 4 cm from the axis (front and back
equally); and the magnet was moved up to 1 m upstream of its nominal
position. Coils were also shortened and lengthened up to 50 cm. All of
the points correspond to the position of the waist of the beam, i.e.
where the Gaussian sigma of the fit to the width is the narrowest. The
phi acceptance of the collimator was 28 degrees. The points lie on an
almost universal curve - QTOR is acting somewhat as an optical lens
with the size of the image proportional to the image distance.

The second page shows the relation between length and width of the
image on the bar for the points on the first plot. As this was for a
phi acceptance of 28 degrees, the lengths are quite large.
                                                                                
My conclusion was that there was no clear advantage to tilting or
moving the coils. You can certainly decrease the length of the image
for the same collimator, but by increasing the current and moving the
image and the bars in closer. You can also decrease the azimuthal
defocussing by dropping the field, but then the image moves out further
and grows larger anyway.
                                                                                
Jim
                                                                                
Jim's attachment:
May2003Talk.pdf