To: Distribution 4 Sep 96

From: Martin Nordby

Subject: Minutes of the IR Engineering and Physics Meeting of 30 Aug 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
David Humphries	LBL 46-161	Yunhai Cai	26
Roy Kerth	LBL 50-340

Electronic Distribution:

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

BSC and Magnet Positions

Mike Sullivan reported on the final configuration for the B1-Q5 magnet positions. The "old" lattice had tight BSC and beam separations at the inboard ends of Q4 and Q5. With tweaking on B1 and Q1, this had gotten worse.

To correct for this, Mike moved Q2 further off the center of the LEB, producing more steering and, hence, more separation. This also widened the LEB BSC to 3 mm wider than the beampipe. By widening the beampipe to 96 mm half-width, and trimming down the wall thickness on the septum size, the new BSC envelopes and positions are accommodated. This leaves only a 1.5 mm overlap at Q5, which drops to 0.5 mm overlap with solenoid on. The decision was to eat this overlap in the BSC.

Martin Donald has run this configuration for the both the HER and LER, and agrees that it is O.K.

Conclusion: This new lattice is accepted.

Next Steps:

--Release lattice and BSC's as is (Mike Sullivan, et al)

--Develop design for modified (wider) LEB chambers (Lou Bertolini).

--Look into thinning down the Q5 magnetic support to buy back the lost 1.5 mm. Build the part as-is, but look into modifying it later, if possible (John Seeman, Mike Sullivan, Uli Wienands, Matt Kendall).

Tolerance of Q2 to Stray Solenoid Fields

Yunhai Cai reported on tracking studies which included a skew octupole harmonic in the Forward Q2 (a.k.a.: Q2R). With no added skew octupole, but with standard magnet tolerances for all magnets (including Q2), and with the solenoid on, the dynamic aperture is all right. The dynamic aperture is defined as 10 sigma. This is the specification for all of PEP-II, determined from past experience with other machines. A tighter dynamic aperture makes the machine difficult to tune, and although the dynamic aperture is acceptable with the solenoid on, the horizontal component is tight.

For the C.D.R., the Q2 field quality tolerance was set at 2 x 10^-4 of the main quad field for each harmonic. To simulate the effect of the skew octupole, Yunhai added this harmonic at multiples of the C.D.R. value. At 3X CDR (6 x 10^-4) skew quad harmonic, the dynamic aperture worsened, especially in the horizontal plane, with some of the seeds infringing on the 10 sigma region. At 10X CDR, all of the seeds were marginal or infringing, and at 30X CDR, the dynamic aperture was collapsing in both planes.

To investigate whether the skew octupoles of the two Q2 magnets cancels (since that part of the skew octupole generated by a solenoid would be of opposite phase), Yunhai put in equal and opposite skew octupoles in the two magnets. At 10X CDR, one of the seeds was lost, while the others looked somewhat better than for the single-magnet study. This showed that the errors only partially cancel, if the magnitude of the harmonic is the same. However, the backward Q2 sees much less field than the forward, so cancellation should not be counted on.

In conclusion, a specification of 6 x 10^-4 (3X CDR value) appears to be a reasonable specification for the skew octupole. This includes skew oct. due both to the solenoid and magnet construction errors.

John Seeman went on to translate this new harmonic spec to actual gauss values for the stray solenoid field. In June, he and Zach Wolf had performed tests on a HER arc quad to investigate this effect, and John showed these results, along with correction factors to account for discrepancies between the tests and the detector magnet models. In June, based on the HER quad tests, John had set the allowable solenoid stray field at 125 G. Below is a list of scaling factors which modify this number:

--Reference radii for the quad test and magnetic model were different: 1.1X

--Quoted field values for the solenoid test were at the center of the solenoid, not the front of the quad: 0.5X

--Mirror plate reduce field, so overall field can be somewhat higher: 1.25X

--Cancellation of both Q2's not factored in: 1.25X

--The original 125G assumed that half of the octupole harmonic was due to natural harmonics in the as-built magnet. However, Yunhai's tracking analysis included both: 2X

This produces a new spec of 107 or 214 G, depending on whether the last factor is added. A few issues were raised concerning the magnitude of some of these scaling numbers, so John will re-check them.

To correct for the skew octupole produced by stray solenoid field, octupole windings are being added to Q2. Fran Younger's initial analysis on these shows that, at 6X 107G octupole field, the n = 8 harmonic of the octupole windings goes out of spec. This sets an upper limit on the magnitude of harmonic which can be corrected by octupole windings.

Conclusion: set spec at 200 G on the face of Q2. We should be able to correct for up to 600 G with the octupole windings.

A discussion ensued about changing the forward Q2 plug to a cylindrical outside shape. This drastically reduces stray field in Q2 by shunting more flux out of the plug to the intermediate IFR door plates. However, this cuts down on solid angle coverage for the RPC. It is not clear to anyone how this impacts the physics. Also, this changes the loading on the superconducting solenoid. Orrin Fackler is looking at this.

Vertex Vacuum Chamber Design Update

Knut Skarpaas presented progress on the development of the beryllium Vertex Vacuum Chamber. The mechanical design has been changed significantly since the last flurry of work finished in August 1995. The focus of the changes was to minimize the risk of corrosion to the beryllium by making the parts and assembly easier to assemble and inspect. The new design moves the Be/St. Steel braze joint out of the water passage completely. This reduces the possibility of corrosion at the braze joint, where trapped flux could react with water to accelerate corrosion. This appeared to be a problem for both VXD2 and VXD3 beampipes for SLD, as well as other beampipes.

The outer shell has been thickened to 0.020 inches (from 0.016 inches) in the acceptance region, and thicker still outside the active area. This improves the mechanical stability of the tube.

On the backward end, the water inlet/outlets have been brought together into a stainless steel manifold. This protects the beryllium from erosion, provides a smoother transition from flow in the tubes to flow in the annular space between beryllium pipes, and routes water better, to prevent eddying. Also, bellows have been added to provide strain-relief for the in-/outlet tubes.

The vacuum tube is machined with built-in ribs and turn-around passages for water flow control. It will be made from low carbon-content Be (<0.1%), to minimize the risk of pitting at inclusions.

The vacuum tube contains the Be/St Steel joint. Just out-board of this joint is the GTAW weld to a stainless steel eyelet. Because of the proximity of the weld to the braze, cooling rings will be needed to prevent damage to the braze.

After brazing, the inner tube assembly is painted with epoxy to protect the Be from water. If gold plating is needed on the vacuum side of the tube, it must be sputtered on first.

Corrosion

Various sources claim that, with very pure de-ionized water, corrosion can essentially be eliminated. However, many other sources point to the importance of maintaining purity, since even trace amounts of impurities can bring on rapid corrosion. The main culprits are chloride, sulfate, and copper ions. This clearly points to needing a dedicated water system with continuous de-ionization/cleaning, and no copper in the system.

Another main source of corrosion is carbon in the Be metal. This can be the source of pitting, but can be reduced by using hot isostatic-pressed (HIP'd) material. Here, the material is sintered in a pressurized argon atmosphere, sealed in a bag, instead of being pressed into carbon dies.

Corrosion Protection

Passivation using a chromate coating is a standard protection method, but it does not last long.

Anodizing the surface produces a thin film around the metal, which is in compression, there-by minimizing the likelihood of cracking. However, this, also, breaks down with time.

The primary protection method is to paint exposed surfaces with an epoxy-based paint: Cytec BR-127. This is used often to protect aerospace aluminum parts, and has been used for Be, also. It contains chromium and strontium, which provide the protection, but which can also separate from the liquid suspension very easily. This has been used for the CESR II beryllium beampipe, and for SLD VXD2 and 3. The Vertex Vacuum Chamber will use this, but painted onto the two cylinders before assembly. This allows the paint job to be inspected for cracks or holes while it can still be repaired.

A final line of defense against corrosion is to use a sub-atmospheric water system, so any possible leak would vent air into the water system, and not water into the SVT volume. This would not help protect the vacuum, however.

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

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