To: Distribution 7 Feb 96
From: Martin Nordby
Subject: Minutes of Near IR Engineering Meeting of 2 Feb 96
Q2 Chamber Design Update
Lou Bertolini presented the latest on the Q2 Chamber re-design to fit around the
P.M. Q2 magnet (v. 4) design. The new design incorporates the vertically-offset
Q2 magnet, the increased beam stay-clear due to moving the Q2 outboard, and
attempts to bring cooling channels as close as possible to the SR region.
With the new geometry, however, the cooling could not be brought much closer
than it was in the previous design. This allowed too much heat to flow into, and
around, the P.M. material, and caused an azimuthal gradient of 3-4 ¡C in the
P.M. material. Thermal analysis this week will show what the options are here.
With a Q2 inner collar of aluminum, the temperature variations should be less.
Lou also looked at the Z-space available for Q2. There appears to be just
enough room between the back of the Q2 envelope and the inside of the Q4
coils for 6-3/4" flanges on both beampipes. However, there is not room for bolt
clearances for in-situ installation of the chambers. The Q4 design does not
have a mirror plate, and the coils are the limiting feature. Jack Tanabe thought
that the coils may be able to be moved closer to the steel, but was not sure. Lou
will check with Johanna Swan on this.
Q2 Design Spec's
As a figure-of-merit, the lattice group has been using harmonics for Q4 of 0.4 G
for lower-order harmonics. The harmonics are assumed to be random in phase
and magnitude, up to this approximate value. Thus, for Q2, the target value of 1
G for lower harmonics is consistent with that for Q4.
Currently, for the 24-block Halbach configuration Q2, the n = 3 harmonic has a
value of about 5 G. However, unlike Q4, the phases of the Q2 harmonics are
NOT random, and some may cancel. Stan Ecklund will be providing the lattice
group with expected harmonic magnitudes and phases to see if the current Q2
fields are acceptable.
Q2 External Fields
David Humphries showed a status of Q2 shielding analysis. He sees
agreement among the three modelling methods: 2-D Halbach calc's, 1/8-model
3-D Amperes, and 2-D Pandera. For a 24-block magnet, with relative
permeability of 1.0 and no shielding, all three codes predict By = -50 G at 9 cm
from the LEB, and By = -5 G at 11 cm (HEB centerline).
However, for a permeability of 1.17 (David's assumed worst-case value), the
stray field increases by five-fold close-in to the LEB, and by over 20-fold near
the HEB centerline.
These field numbers are all "length-normalized," where the By-field is integrated
along a line parallel to the magnet axis at a given offset from the magnet center
(from the symmetry plane to z = infinity), then the integral is divided by the
length of the magnet. For longer models, the length-normalized values will
decrease, since the end effects become a smaller fraction of the total field.
However, the total integrated field will increase for longer magnets.
The next step is to look at adding shielding only around the HEB beampipe, and
using a mirror-plate-type disk at the ends of the magnets. Based on Tom
Mattison's investigations into the saturation effects of stray solenoid fields, any
shield design must be able to tolerate the 50-100 G of stray solenoid field
without saturating.
Stan Ecklund and Dave also showed an interference between the beam stay-
clear and the inner radius of the Q2 P.M. blocks. The beam stay-clear was
assumed to increase linearly to form a cone through Q2. However, it actually
bulges in the middle of the magnet by 1-2 mm. To rectify this overlap, the inner
radii of the P.M. blocks will have to increase somewhat. Dave Humphries will
look at where additional volume can be found to make up for this lost space.
Tom Mattison discussed preliminary results of some rough analysis of the Q2
stray field. He found that the stray field is due mostly to the irreducible intrinsic
harmonics of the magnet. The further away from the P.M. material (inside or
outside the material radius), the better the actual field agrees with that of a
simple N-dipole array.
In 3-D, at larger radii outside the magnet, the external fields at a given radius
tend to self-cancel, as the field orientation changes sign from within the length
of the magnet to off the ends. This is not true at smaller radii, but at the HEB
centerline (distance = 1.3 X Radius of Q2), the fields cancel.
These minutes, and agenda for future meetings, are available on the Web at:
http://www.slac.stanford.edu/accel/pepii/near-ir/home.html