Minutes of the IR Engineering and Physics Meeting of 23 Feb 96 Solenoid Field Harmonics Mike Sullivan reported on progress in developing an algorithm to find the harmonics of the solenoid field from field-map information produced by Orrin Fackler. The following is a summary, with details shown in the transparencies: First, knowing B(z) for r = 0 cm, and assuming a perfect axially-symmetric field, allows the z-field to be expanded for any arbitrary B(r,z). Thus, if the z-field can be found along the magnet axis, the field in the (r, z) volume of interest can be found. To find an expression for B(z) at a given point, Mike fits a 7th-order polynomial to 40 points surrounding a given z-location. The values for these points are the output from Orrin's analysis. This produces coefficients for a harmonic expansion for a given (r, z) as a function of z. These can then be used to find multipoles at the beam orbit locations. Since the both beams are actually rotated slightly from the axis of the detector, the field of interest is actually the component of the solenoid field parallel to the beam axis. This has been ignored, with an expected error of 1%. Real data from Orrin shows that the steps due to discretization of the finite- element mesh introduce noise at about the third derivative of the fit curves. This produces errors in the higher harmonics. A finer-meshed discretization would reduce this error. Alternatively, if the radial field were used as input, the numerical inaccuracies may be reduced, making the algorithm less sensitive to the discretization and round-off errors of the input. Mike will look into this. Q2 Septum Magnet Design Fran Younger discussed an alternative Q2 septum electro-magnet design, with a reduced cross-section and longer bore length. The new design includes four identical septum coils, to reduce harmonics in the quad field, and a modified shape to the pole tip hyperbolas to reduce the n = 6 and 10 harmonics. Also, trim coils are used to eliminate the dipole and sextupole harmonics. The magnet iron was lengthened to 66 cm long, with 72 cm long coils. This essentially fills up the space between the back of the Q2 septum housing and the front of a single-ring P.M. skew quad. By lengthening the magnet, the current density of the coils can be reduced by 20-25%, and the amount of iron in the cross-section reduced. The CDR magnet measured 410 mm on a side, while this one is 322 mm high by 358 mm wide, which is comparable to the P.M. Q2 design (v 4). Because all four coils are septum coils, they all will require precise positioning, to better than 0.005". There are 2 flow paths per coil, with a velocity of 18 ft/sec, temp rise of 20¡C, and a pressure drop of 50 psi. The current density is 54 A/mm^2. The velocity could be reduced by increasing the number of parallel circuits, or increasing the allowed temp rise. Fran felt like the cross-section could be reduce 1-2 cm more in either dimension, if needed. The next step is to double-check the BSC values for both beams, and the Luminosity Monitor cone stay-clear. The design can then be updated. Also, John Seeman volunteered to find and map the stray field in a Collin's quad, and Fran will look into running a 3-D model of the end of the magnet to investigate stray end-fields in the HEB beampipe. Q1 Field Map Stan Ecklund updated his report from last week on the applied de-magnetizing force on both the quad and dipole blocks in Q1. He included both sets of blocks in the simulation, as well as a linearly-varying external radial field (from the solenoid), and used a "real" P.M. material B-H curve (Shin-Etsu R26HS) He found theat the peak de-magnetizing force in the quad blocks is 13 kOe, and in the dipole blocks it is 12 kOe. The effect of the external quad field on the dipole blocks turns out to be negligible, largely because of the radial gap between the blocks for the mid-ring support. The de-magnetizing forces are slightly less than originally expected, but the material spec will not be changed. This gives us a slightly greater margin of comfort. Trip to UGIMAG Andy Ringwall and David Humphries visited UGIMAG, in Valparaiso, Indiana earlier in the week. They are one of two US suppliers of Sm2Co17 P.M. material. Below are notes from their trip: Trip Report: Visit to UGIMAG Andy Ringwall Background Largest US producer of permanent magnets French parent, two US plants Valparaiso, IN: Nd-Fe-B(primary), Sm-Co(Incor 26) Boonton, NJ: Sm-Co(Recoma 28) Parent bought Valparaiso plant in early '80's, invested capital equipment $$ 400 workers at Valparaiso, 2 buildings, expanding Bread & Butter: Neo magnets for hard drives Some hard drive magnets final machined and magnetized at two Singapore plants Experience with Halbach quads and dipoles Receives casted Neo and Sm-Co mixtures from US, European, and Japanese vendors- processes to final magnet Sm-Co Processing Steps Receive castings mixed to Ugimag specifications Samples are chromatographically inspected, QC check Dissolve in Nitric acid and oxidize Compare optical spectrum to a standardized concentrations Check for Sm, Co, other RE metals Use different techniques to find N and O Electron microscopy available Crush to coarse powder(Hydrogen decrepetation) Eccentrically rotate drum in hydrogen environment at temperature Hydrogen helps break down casting(I think by a reduction reaction with RE) Jet mill to fine powder High pressure N2 Separate domain-sized particles and dust Store powder in inert Argon atmosphere until pressed: reactive metal + oxygen + large surface area = fire More chromatography, density & size characterization Blend mixture based on chromograph Check avg. size by pressure drop across sample Weigh out for pressing, ready to press Particle Size and Coercivity, Sm-Co Magnets Define a critical particle spherical radius, rc Pariticle radius, r, above rc allows easy domain wall motion(creation of new domains decrease total energy) Pariticle radius below rc difficult domain wall motion(creation of new domains increases total energy) SmCo5 is a nucleation-type magnet Domain wall motion is easy within a grain, rgrain > rc. The grain boundaries prevent cell wall motion and define the intrinsic coercivity. Grain size is important. Particle size is of less importance. Low saturation fields and lower coercivity Sm2Co17 is a pinning-type magnet Domain pinning occurs at the walls between particles or cells. It is the precipitated phase(SmCo5) at the cell wall that prevents domain wall motion and defines the intrinsic coercivity. Cell size(near rc) is important. Grain size is of less importance. Sintering and heat treatment forms the precipitate. So 2:17 is seperated by thin walls of 1:5 in the microstructure. High saturation fields and higher coercivities Sm-Co Processing Steps Select dye and punch Dye has flat for transverse orientation Punch is non-ferromagnetic(400 SST) for transverse orientation Dyes and punches are hand polished for fit Press and orient Ugimag custom builds solenoid coils and adapts to presses Die is centered in solenoidal field Compress 60% then pulse orient(about 10 kOe) Finish compression Demagnetize with decaying 60 Hz field Sinter and heat treat in vacuum furnace Contaminates will affect magnetic properties Sinter(some liquid phase present), quench, and age: Intrinsic coercivity is defined during quench: limits block size. [ed note: This is one of the key steps. The temperature produces partial melting of the material, which affects the angularity tolerance of the magnetization.] Pulse magnetize(40-45 kOe) samples from lot, run B-H curve in low reluctance circuit Grind blocks to shape Pulse magnetize blocks Thermally stabilize Helmholtz coil inspect, magnetic moment & angle Package and ship B-Factory I.R. Engineering D Humphries - LBNL Feb. 23, 1996 UGIMAG Visit - 2/20/96 UGIMAG Permanent Magnet Issues : UGIMAG, Inc. is a well supported high volume producer of Nd Fe B and samarium cobalt permanent magnet materials. They have modern production capabilities and good material characterization capabilities. They are motivated to work on relatively small scale research projects and are setting up parallel small scale production facilties for these efforts. They are currently producing insertion device magnets for both ESRF and STI. Production facilities : Transverse and axial die pressing machines w/ good control of field asymetries, 8 in. dia poles, 5 in. gap. Isostatic pressing to be avoided. Meierburger cutting machines for precision slicing, production grinding machines. Magnetic characterization instruments : Well designed Helmholz system w/ .2 deg meas. accuracy for orientations. determine gamma perpendicular. Specific to samarium cobalt: Recoma28 has good Br, Hc specs but is highly non-linear. Incor26 is likely candidate for Q2 : Br = 10.7 kG, Hc = -10.1 kOe, Hci = -21.0 kOe Mu rel is ~ 1.06, Mu perp. ~ 1.05 but has not specifically been measured These minutes, and agenda for future meetings, are available on the Web at: http://www.slac.stanford.edu/accel/pepii/near-ir/home.html