To: Distribution 7 Oct 96

From: Martin Nordby

Subject: Minutes of the IR Engineering and Physics Meeting of 4 Oct 96

Hard-Copy Distribution:

Bob Bell	41	David Kirkby	95
Lou Bertolini	LLNL L-287	Jim Krebs	41
Gordon Bowden	26	Harvey Lynch	41
Pat Burchat	95	Tom Mattison	17
David Coward	95	James Osborn	LBL B71J
Scott Debarger	17	Andy Ringwall	17
Hobey DeStaebler	17	John Seeman	17
Jonathan Dorfan	17	Knut Skarpaas VIII	18
Stan Ecklund	17	Mike Sullivan	17
Alex Grillo	95	Uli Wienands	17
John Hodgson	12	Mike Zisman	LBL B71J
Hank Hsieh	LBL B71J	Pran Kaul	22
David Humphries	LBL 46-161	Orrin Fackler	LLNL L-291
Roy Kerth	LBL 50-340	Lew Keller	41

Electronic Distribution:

Curt Belser	Rick Iverson	Jeff Richman	Jack Tanabe
Catherine Carr	Nadine Kurita	Natalie Roe	Rick Wilkins
David Coupal	Georges London	Ross Schlueter	Fran Younger
Fred Goozen	Joseph Rasonn	Joe Stieber

Q2 Shielding Plug Update

Lew Keller reported on progress in the analysis of the Q2 shielding plug. First, these are the inputs and assumptions used for the most recent runs:

--Mirror plates were added to both ends of Q2. This is assumed to be 0.635 cm thick, and stood off the core by 3.81 cm. One model has it attached magnetically to the core around its perimeter. Outside radius = 7.62 cm,. inside radius = 4.826 cm.

--Q2 length = 0.61 m (compared with older models, which assumed 0.5 m).

--Q2 pole tip radius = 4.8 cm (this is based on latest design; older models used a smaller value).

--Put 3 cm x 45° chamfers on the ends of the Q2 core (This is more realistic, but smaller than older models).

--The cone plate (conical part of IFR Doors) has been thickened to its correct thickness of 65 mm (older models used 50 mm thick, which was incorrect).

--The as-measured B-H curve for the 20 mm plate from KHI is slightly lower than for the 30, 50, and 100 mm plate. The new curve is used for all 20 mm plate in the model, while the curve measured for the thicker plate is used for all thicker plates in the model. KEK Venus steel properties are used for the Q2 Plug.

--Model the "Big Bore" Plug design, where the plug is made axisymmetric by machining out the smallest circumscribing cone, located on the detector axis, which clears the PEP-II stay-clear + 5 mm.

--Use 3 bucking coils with fixed currents, to make comparison easier:

1. Small-Finger Coil: almost centered on detector axis, inside Q2/4/5 Raft, and spaced 1 cm off in-board end of Q2 mirror plate. NI = 5900 A-T.

2. Large-Finger Coil: mounted on steel of third finger in Plug. NI = 618 A-T.

3. Q4 Global Coil: located on the outside radius of the Plug, where it bolts to the door. NI = 20,000 A-T.

A new criteria was used to judge the results, based on calculations shown by Stan Ecklund two weeks ago. PHI = Bz(avg) * (Bore Area) is calculated at the inner edge of each mirror plate (IR-Z = 3.307 m and 3.453 m). The maximum value of PHI = 0.5 G-m^2, which roughly corresponds to Bz(avg) = 70 G, based on what Stan showed.

Run 1:

Used a mirror plate, not connected to Q2, plus all input described above. Results:

PHI(front) = -0.49 G-m^2

PHI(back) = -0.35 G-m^2

Bmod = 461 G in Q2 steel

Bmod = 75 G in Q4

The same run, but with no mirror plate, shows that Bz is changing quickly with z near the front of Q2. Thus, the choice of z-location for integration can significantly affect results. However, the mirror plate provides a simple landmark for integration, and helps form a magnetic "pin-hole." Whatever Bz actually enters through the mirror plate will, indeed, produce octupole in Q2, so integrating at the in-board face of the mirror plate should reflect reality.

Run 2:

The mirror plate outside radius was increased to 15 cm, and it was magnetically mounted to the outside radius of Q2.

PHI(front) = -0.31 G-m^2

PHI(back) = -0.15 G-m^2

Bmod = 512 G in Q2 steel

Bmod = unknown in Q4

int[Br*(2*pi*radius)*dz] = 0.17 G-m^2, which agrees well with the difference between PHI from front to back.

This new criteria suggests that the Big Bore design does, indeed, do the job. Using PHI as the criteria better reflects what really affects the machine performance (i.e.: octupole in Q2), while using the old criteria of Bmod < 100 G appears to be far too conservative, and does not control the important variables.

Next Step:

--Questions now are focusing on the actual effect of the stray field on the asymmetric Q2 geometry. James Osborn will run a 3-D model of Q2, putting in solenoid fields on- and off-axis to the Q2 centerline to look at harmonics in both beam passages.

--Magnetic testing results have been finished by John Seeman and Zach Wolf. John will report on this next week

Raft Support Cones

Scott Debarger reported on progress on the Q2/4/5 Raft in-board structure. Due to a layout error in the Raft file, the current Q2 design from James Osborn did not fit in the Raft design as shown last week. However, in a rare instance where two wrongs turn out to make a right, Scott and Catherine Carr re-laid out the cones to fit the new Q2, better center the Raft on the new Q2 location, and to make room for the "small-finger solenoid." The resulting new transition is only slightly larger than the original size, and should actually reduce the additional cut-out needed in the third finger to make room for the new bucking solenoid. As Scott mentioned last week, this also reduces the asymmetry in the third finger cutout, which will reduce the entire cutout diameter, since the symmetric "Big Bore" design is now the design of choice.

Also, in response to some recent editorializing on the stiffness of the open cone, Scott investigated the stiffness of the cone, as a function of the angle which the partial cone subtended. The original half-cone subtended 180°, and had a vertical stiffness which was only 9% of that of a closed cone. Laterally, its stiffness was 50% of a closed cone.

As part of an overall stiffening campaign, Scott increased the size of the cone to wrap above beamline and subtend 220°. This just left room for the septum chamber to be pulled out, but no room for access to the sides of the chamber. With this configuration, the vertical stiffness increased to 21% of a full-cone (2.5 times the original stiffness), while the lateral stiffness increased to 71% of a full-cone.

Vertically, this increased stiffness helps a lot. However, Scott still feels that a gusset will be needed. Horizontally, a gusset would be difficult, at best, since it would cut into the shielding plug fingers. However, as Scott's analysis showed, the design is 3 times stiffer and, as Martin Nordby pointed out, the vibration specification on horizontal vibration is 6 times looser horizontally than vertically.

Next Step:

--Complete FEA analysis of Raft for normal, and EQ scenarios (Scott Debarger).

IR-2 Pier Design

Pran Kaul of Plant Engineering reported on the conceptual design of the Piers which reach into the IR-2 hall at either end of the detector. These support the Q2/4/5 Rafts and out-board machine components in the hall. The Piers consist of two thick concrete columns with a concrete beam across the top. The beam is 63 inches wide and 14 inches thick, with a 3/8 inch thick steel plate on top. This serves as the mounting surface for all machine components. Walkways mount off each side of the Pier, but are remove-able.

The Piers are very over-built for the expected load capacity, since the size is dictated more be the size and location of machine elements. Because of this, two transverse tunnels have been put in each Pier. This both saves concrete and provides more access for walking and possible future cables or services.

The Piers are mounted to the concrete slab floor with epoxied dowels. These take only shear, since no moment-bearing connection is needed.

Pier Layout

Martin Nordby reported on how this Pier design fit into the overall PEP-II plans for the IR hall. The Piers first provide a support point for the Q2/4/5 Rafts. On the Forward end, a second support is added by mounting a cantilever off the front of the Pier. On the backward end, supports for the remove-able portion of rail for the Q2 Plug removal are mounted on the front face of the Pier.

On the side of the Pier facing the center of PEP-II, the walls of the Piers will be used to mount machine services. First, two cable trays will extend out the tunnel mouth, then drop vertically to the Pier. These trays were moved farther away from the detector, based uncertainties about how much space would be needed for detector access. Long-haul water-cooled D.C. magnet cables drop out of these trays and meet up with short-run air-cooled cables coming from the magnets on the Rafts at a junction box mounted on the Pier wall. This provides a clean connection for all magnets cables, while still maximizing access to the detector ends.

On the in-board end of the Pier, still on the side wall, two junction boxes for water connections will be mounted. One provides routing and flow measurements for the high-pressure water to the copper magnet coils. The second water box serves as the distribution and temperature-control center for the chilled LCW used inside the Support Tube. Here, re-heaters will control water temperature for multiple parallel cooling circuits. Insulated lines will run from here to the Raft, then into the detector.

Pier Fit wrt DIRC and BaBar

Jim Krebs checked the fit of the Pier design presented by Pran Kaul with respect to the detector. According to Jim's 2-D layout of the detector, there is 125 mm of clearance between the Pier and the SOB doors as they open on their compound hinge. The actual clearance should be larger, since the SOB doors are spherical, and the point of closest approach is actually at a skew vertical angle.

The Piers are also low enough that the Drift Chamber can be removed from BaBar on-beamline, if needed.

The platforms on either side of the Piers may need some future modification. Jim has yet to look at the Forward Calorimeter extraction gantry, but the gantry base may need vertical support members located right through the current platform location. Also, Jim is looking at adding a mezzanine-level platform in the entire northeast corner of the radiation area for racks and heat exchangers. This mezzanine would replace one of the Pier platforms. Any modifications to the platforms for BaBar roll-on is fine, but the junction boxes on the inside of the Pier should not be moved.

Next Steps:

--Layout F-Cal extraction gantry (Jim Krebs)

--Layout junction boxes (Martin Nordby)

These minutes, and agenda for future meetings, are available on the Web at:

http://www.slac.stanford.edu/accel/pepii/near-ir/home.html