To: Distribution 4 Feb 97
From: Martin Nordby
Subject: IR Engineering and Physics Meeting Minutes of 31 January 97
Hard-Copy Distribution:
Bob Bell | 41 | Nadine Kurita | 18 |
Gordon Bowden | 26 | Jim Krebs | 41 |
Pat Burchat | 95 | Harvey Lynch | 41 |
David Coward | 95 | Tom Mattison | 17 |
Scott Debarger | 17 | James Osborn | LBL B71J |
Hobey DeStaebler | 17 | Andy Ringwall | 17 |
Jonathan Dorfan | 17 | John Seeman | 17 |
Stan Ecklund | 17 | Mike Sullivan | 17 |
Alex Grillo | 95 | Uli Wienands | 17 |
John Hodgson | 12 | Mike Zisman | LBL B71J |
Hank Hsieh | LBL B71J | ||
David Humphries | LBL 46-161 | Lew Keller | 41 |
Roy Kerth | LBL 50-340 | ||
David Kirkby | 95 |
Electronic Distribution:
Curt Belser | Kay Fox | Jeff Richman | Joe Stieber |
Lou Bertolini | Fred Goozen | Natalie Roe | Jack Tanabe |
Catherine Carr | J. Langton | Ross Schlueter | Rick Wilkins |
Al Constable | Georges London | Knut Skarpaas VIII | Fran Younger |
David Coupal | Joseph Rasonn | Ben Smith |
Shielding Plug Area Layout
Scott Debarger reported on the final layout of the Q2/4/5 Rafts and Forward and Backward Shielding Plugs. The Back Plug clears the coil leads for Q2 by 0.27", at the point of closest approach. At z = 5250 mm, the water connections for the Q4 magnet stick into the path of the Back Plug as it is extracted. This happens at 2:00 and 8:00 o'clock, with an interference at both places of 0.25". Since the fittings are stacked two or three deep, elbows cannot be used, and the packaging cannot be made tighter.
Furthermore, the layout used the actual Plug dimensions, not a stay-clear value, to account for alignment and fabrication errors in the Plug. When this 5 mm is added on, the bore of the backward Plug must be increased by 1 cm in radius to ensure there is no interference during withdrawal. Also, at the 8:00 o'clock position, an additional rabbet is needed to allow room for the coil leads. The iron removed from the move-able Plug can be added to the fixed portion, so there is no net loss of iron. Although this is only a 1 cm wide piece, there was strong opposition by BaBar to moving this material. We will meet off-line with Johanna Swan, the Q4 Magnet engineer to look into moving the coil leads, if possible (and schedule permits).
The Q5 LEB Chamber has been notched,
and the ion pump on the Q4 LEB Chamber reduced from a 220 L/s
pump to a 60 L/s Pump to allow room for the Plug extraction. The
LEB corrector between Q4 and Q5 no longer interferes with extraction,
since it fits into the notched chamber. All magnets will be shielded,
so the power supplies will NOT NEED to be locked and tagged (however,
they can be for access, if desired by BaBar personnel).
Plug Magnetic Analysis
Lew Keller reported on analysis of
the "final" Plug configurations, which include changes
proposed by Kawasaki Heavy Industries, and by PEP-II.
Back End
On the Back End, increasing the Plug
radius by 1 cm decreases the peak radial field in the Q1 Magnet
from 2.4 kG to 2.35 kG (using 2-D axisymmetric MERMAID model).
The axial force on the solenoid cryostat is reduced by 700 lbs.
However, the back end bucking coil will most likely have to be
increased in strength, and Georges London needs to investigate
the effects on stray fields in the DIRC PMT's. Lew will also be
looking at field uniformity in the Drift Chamber active volume,
but felt that this should not be significantly impacted.
Conclusion and Decision
--Open up the Back end Plug bore
by 1 cm in radius to 37 cm.
Forward End
3-D modelling done by Stefan Mikhailov showed that, for differing depths of chamfer on Q2, the fraction of axial flux converted to octupole in Q2 changes. For a 1 cm chamfer, 30% of the axially field resulted in skew octupole. This is far less than the 100% previously assumed.
Lew used this scaling to correlate results from his 2-D axisymmetric model with the expected octupole due to the real, 3-D fields. He ran three models.
1. Baseline design.
2. KHI Mod 1: included proposed changes by KHI, which included closing the gap between fingers 2 and 3, making the bore out of stepped cylinders, rather than frustrums, and making the second and third finger tips be planar washers, rather than gently sloping dishes as called-for in the baseline design.
3. KHI Mod 2: included changes proposed, above, plus an increase to the radius of the Plug bore and third finger by 3 mm. This accommodated the expected increase in the wall thickness of the Q2/4/5 Raft (see below).
Results for the three runs are shown,
below, for a 1 cm chamfer on Q2, with a mirror plate. Bz is the
field measured at the mirror plate. The central field for these
runs is 17 kG, unless otherwis noted.
Model | |||
PEP-II CDR | |||
Baseline | |||
KHI Mod 1 | |||
KHI Mod 2 | |||
KHI Mod 2 (B0=15 kG) |
For KHI Mod 2, field values for all
magnets are listed below. This includes a 17 kG main solenoid
field, , one main bucking coil around the DIRC SOB, but no coils
around Q2 at either end, and no "Q4 Bucking Coil" on
the forward end door.
Location | |||
Forward End | |||
Q2 Front | |||
Q2 Rear | |||
Q4 Front | |||
Backward End | |||
Q2 Front | |||
Q2 Rear | |||
Q4 Front |
Conclusions and Decisions:
--Open up radius of Forward Plug bore and third finger inside profile by 3 mm
--Open up radius of Back Plug bore by 1 cm to 37 cm
--Eliminate Forward End Q4 Bucking Coil
--Eliminate Back End Q2 bucking coils
Q2/4/5 Raft Analysis
Scott Debarger and John Hodgson reported on analysis of the Forward Q2/4/5 Raft, and design modifications to increase its stiffness. The analysis showed that the box beam and Y transition region, as well as the forward transition and cylinder/cone parts, needed stiffening. Some stiffening was accomplished by thickening members, improving the cross-bracing, and filling in the Y. It was found necessary to increase the thickness of the cylinders and cone from 3/8" to 1/2" to reduce the peak stress at the cone/cylinder welded joint to a safe level during installation. To maintain safe stress levels in this region during a seismic event, the half-cylinders were closed-out with structurally integrated tops.
Below is a summary of the results.
All models include "real" magnets, with actual struts
modelled, to pick up modes which include individual magnet motions.
The loaded Raft weighs 25,900 lbs (22,450 lbs of equipment plus
3450 lbs for the Raft itself), and is supported at 2 points under
Q5 and 1 point under Q4 during operation. Unless otherwise indicated,
deflections and stresses listed below are for this support condition
and 1 g gravity loading. Natural frequencies listed below also
assume operational support conditions. In an earthquake, the support
under Q4 breaks away, and a "Seismic Bracket" mounted
to the BaBar door picks up the very in-board end of the Raft.
Earthquake (EQ) stress values shown below are for this support
case.
"Baseline" design:
Includes a small gusset under the Q2 Magnet and Septum Chamber. Box beams under Q5 are 6 x 6 x 1/4" wall with a 4 x 6 x 1/4" wall channel welded under it, and partial cylinders under Q2 and septum are 3/8" thick.
Deflection at in-board end: 0.107"
f1: 6.0 Hz
Mode shape: Q5 pitches axially, distorts Raft locally, Raft kinks vertically between Q4 and Q5.
f2: 8.8 Hz
Mode shape: Q4 and Q5 pitch transversely
on Raft, Q4 more so than Q5; Raft racks under Q5, twists between
Q5 and Q4.
"Stiff Raft" design, 3/8"-thick nose:
Replaced box beams and channels under Q5 with 1"-top/bottom-plate, 10" x 1/2"-side-plate box beams. Added 3/4" plates to close out top, bottom of the Y-transition region between Q4 and Q5, and connect top to bottom with a shear plate. Removed gusset under septum, increased thickness of canoe-to-cylinder transition piece from 3/8" to 1/2".
Deflection at in-board end: 0.113"
f1: 9.0 Hz
Mode shape: full-body deflection, with in-board end kinking at cone/cylinder joints
f2: 9.9 Hz
Mode shape: similar to mode 1 but in-board end kinking more pronounced
Max Von Mises stress in nose with
distributed support tube load: 23.7 ksi (at welded cone/cylinder
joint)
"Stiff Raft" design, 1/2"-thick nose
Increased thickness of both partial cylinders under Q2 and septum, and cone connecting cylinders, to 1/2". NOTE: Through bore of Plug, cylinder is now 1/2" thick.
Deflection at in-board end: 0.086"
f1: 9.2 Hz
Mode shape: full-body deflection, with in-board end motion subdued
f2: 11.1 Hz
Mode shape: in-board end bobbing vertically
Max Von Mises stress in nose with
distributed support tube load: 13.6 ksi (at welded cone/cylinder
joint)
"Covered Cylinder" design
Added a cover to complete the cylinders and cone around Q2 and septum.
Deflection at in-board end: 0.066"
f1: 9.3 Hz
Mode shape: full-body deflection, with in-board end nearly stationary
f2: 12.3 Hz
Mode shape: vertical bobbing of in-board end
Max Von Mises stress in nose, with distributed support tube load, under vertical EQ load (2 g vertical acceleration + 1 g gravity): 20.9 ksi
Max Von Mises stress in nose, with
distributed support tube load, under lateral EQ load (2 g lateral
acceleration + 1 g gravity): 20.6 ksi
Analysis Summary
Deflection | ||||
f1 | ||||
f1 shape | ||||
f2 | ||||
f2 shape | ||||
Max oper. stress | ||||
Max EQ stress |
Conclusions
--Use 1/2" plate on all cones and cylinders on in-board end, to reduce
stresses at joints
--Do NOT use any gusset under Q2.
--Plan on using a cover over Q2 to increase stiffness and reduce stresses
and deflections under earthquake
loading.
These minutes, and agenda for future meetings, are available on the Web at:
http://www.slac.stanford.edu/accel/pepii/near-ir/home.html