Meeting 1/23/02 with David Beck, Tim Houck, Gary Ritchie, Derek Shuman, Dave Vanacek, Will Waldron, and Simon Yu present. Some additional topics added by D. Shuman. Any corrections or comments welcome -D. Shuman 1. Coil Winding(G. Ritchie): A. Excessive Coil Width Measurements of wound and stretched coils were found to show excessive end width, approx .040-.050 wider than nominal, as measured at the widest point across the bends. This may be OK, but could possibly result in end arcs which present excessive radial excursion from nominal when arcs are formed, possibly compromising voltage holdoff to ground. This excessive width is thought to be the result of insufficient development of tension in the end sections due to friction in the bends of the coil forming shoes. It may be helpful in future winding operations to use Teflon lubricated Delrin for shoes. G. Ritchie had found that stretching the coils a total of 5% (1 inch for a 20 inch coil length ) rather than 2.5% (1/2 inch for a 20 inch coil length) that the coil turn widths come out very close to nominal. D. Shuman felt that excessive coil stretching might compromise dielectric strength of the insulation, and might possible result in unacceptable increase in coil resistance. After this meeting, coil resistances for 2.5 and 5% stretched coils (same overall length) were measured with a precision microohmeter and found to correlate with the cross-sectional area decrease, which is 2.5%. The additional work hardening appears to have a negligible contribution to the resistance increase. The 2.5% resistance increase is acceptable from a heating standpoint. the coil resistances were measured at 10.945 and 11.223 milliohm for the 2.5% and 5% streched coils, respectively (one sample of each). These measurements are 22% higher than the rough original calculations, and when combined with the 30% longer pulse width, will result in a 60% higher temperature drop from coil to core (8 deg C vs. 5 deg C). This should be quite acceptable, as temperature drops of over 40C were not a problem for the elliptical pulsed quads, though these were not bonded to the cores. Stresses in the epoxy will be compressive in all directions due to restraint by the cooler core, so there should not be any problem with cracking. G. Ritchie has also performed tensile tests on conductor samples to the point of breakage. It appears an elongation of 30-40% is required to achieve breakage, and only in broken samples can the insulation be seen to separate, either from the copper, or to part at the break. No visible evidence of cracking can be seen, though viewing under a microscope should be done. These samples will be checked for excessive leakage current in the electrolyte solution, along with several coil samples stretched at the 2.5 and 5% level. Unstretched conductor samples will be tested for comparison. No visible separation in any of the coils can be seen, though inside corners shoud be checked under a microscope, if possible. A two person winding process was attempted with D. Shuman assisting G. Richie in stabilizing conductors during the hand forming of corners during winding. This reduced the width slightly but createdoverbent corners for some strange reason. It was more difficult to do, as well, and did not seem worth the effort. Assuming there are no problems with insulation, or when coil is checked in the flat plate dimension check fixture (to be built), the 5% coils can be used. If no similar problems are encountered with the 2.5% coils which show excessive end width, yet still fit the dimensional checking fixture, without escessive end excursion (.030 is OK on each side) then the 2.5% stretched coils should be used, since we have most of them made already. A 3.7% stretched coil (3/4" may also be tried to see if better dimensional control is possible with less stretch). B. Finished Coil Dimensions(G. Richie): Gary has made a go/no go length checking gauge of plastic to allow length measurement without scratching the insulation. A plate gauge having conductor slots and a rectangular pocket for the ends (essentially an unrolled section of the coil form) will be made for final dimensional check, to help assure that coils will fit the coil forms after arcing the ends. This is needed because the end arc machine and procedure have not been worked out, nor will coil forms be available for several weeks, minimum. Since we can easily finish coil winding soon, we need a quick way to verify that they are correct. C. Coil End Arcing Machine(G. Richie, W. Tiffany): This machine is still being fabricated. Combs remain to be made, but the machine should be ready for trial next week D. Coil Insulation Check (G. Ritchie, D Shuman) Stainless steel tanks have been located to perform both an electrolyte immersion and a rinse. Coil holders have been fabricated to suspend coils into the elctrolyte. In order to avoid electrochemical voltage generated between bare copper and the stainless steel, a copper sheet will be used as the ground for vlotage source. The hipotter will be used, with a separate voltmeter used to adjust voltages to the 100V level. At this voltage, even weak electrolytes of conductivity 1 mho/m should be sufficient to show milliamp currents from a micron sized copper barespot, though higher conductivities should be used, if feasible. 2. Coil Forms (D. Beck) A. Potting test specimens: These remain to be designed and fabricated B. Coil Form Mandrel These have been received from LLNL and are awiating final fabrication 3. Magnet Laminations(D. Vanecek): A. Purchase Order: This was awarded to Haig on a sole source order B. End Plates(D. Vanecek: A change to stainless steel was made, and this order will go to LBNL shops for competitive bidding, as well as to outside vendors, including Haig. 4. Potting Mold (D. Beck): A discussion between D. Beck, G. Ritchie, and D. Shuman will be held to help finalize the design. 5. Pulsers (W. Waldron): Capacitors and Power supplies should be received by the 3rd week of February. Experimenters will determine which racks will be used for the pulsers and various diagnostics. 6. Schedule (D. Vanecek): D. Vanecek estimates the first quad will be available for testing by the first of June; if everything sails along without a hitch, perhaps a month earlier. This sounds reasonable if not slightly optimistic, to D. Shuman 7. Beamtube (D. Shuman): Since resistivity variations in welded tube might introduce higher order multipole eddy currents, and buckling resistance for thin walled S.S. tubes of 1-1.5mm are not much higher than vacuum, it was decided to continue pursuing a thin walled composite SS/epoxy fiberglass tube. Carbon fiber is an option as its conductivity is relatively weak compared to SS, but need not be used unless stringent radial space requirements are present. We will pursue the FG wrapped SS tube (.016-.020 inch thick), which shows minimal eddy currents, and (being thicker) would be more damage resistant than a thin carbon reinforced tube. D. Shuman will do a strawman design for the layup, then work closely with potential vendors to make the tube. Current thinking is that the thin walled tube will not fit well on a mandrel but could be held round on the ends and lightly pressurized with air to hold it round against the wrapping tension from the filament winding process. Curing may need to be done in stages, to progressively stiffen the tube during the winding process, or pressure may need to be ramped as layers are built up. the first is less risky than the second. The mfr will likely have their own ideas of course. Any other ideas are welcome, of course.