IR Design Status and Update

1. IR Design Status and Update M. Sullivan For M. Boscolo, K. Bertsche, E. Paoloni, S. Bettoni, P. Raimondi, M. Biagini, P. Vobly, A. Novokhatski, S. Weathersby, et al. SuperB Workshop XIII Elba, Italy June 1-4, 2010

2. Outline • IR baseline design • CDR2 (white paper) baseline • Features • Layout • SR backgrounds • Energy changes • New designs using Vobly’s Panofsky quads • Three new designs • Vanadium Permendur • Holmium • Air Core “Italian” • SR calculations • Problems • Ideas • Summary

3. Present Parameters

4. Parameters used in the IR Design Parameter HER LER Energy (GeV) 6.70 4.18 Current (A) 1.89 2.45 Beta X* (mm 26 32 Beta Y* (mm 0.253 0.205 Emittance X (nm-rad) 2.00 2.46 Emittance Y (pm-rad) 5.0 6.15 Sigma X (m) 7.21 8.87 Sigma Y (nm) 36 36 Crossing angle (mrad) +/- 33

5. General IR Design Features • Crossing angle is +/- 33 mrads • Cryostat has a complete warm bore • Both QD0 and QF1 are super-conducting • PM in front of QD0 • Soft upstream bend magnets • Further reduces SR power in IP area • BSC to 30 sigmas in X and 100 sigmas in Y (7 sigmas fully coupled)

6. The Present Baseline Design

7. Larger view

8. SR backgrounds • No photons strike the physics window • We trace the beam out to 20 X and 45 Y • The physics window is defined as +/-4 cm for a 1 cm radius beam pipe • Photons from particles at high beam sigmas presently strike within 5-6 cm downstream of the IP • However, highest rate on the detector beam pipe comes from a little farther away

9. Beam Tail Distribution These tail distributions are more conservative than those used for PEP-II. The SuperB beam lifetime is shorter by about a factor of 10 so the tail distributions can be higher. But we will probably collimate at lower beam sigmas than shown here.

10. New beam tails from Touschek Manuela has been working on getting a beam tail distribution from BGB and Touschek. This is from the LER at -2 m from the IP. This will be a first as far as getting a real distribution for the beam tails for SR. More on this on Thursday.

11. On the airplane comparison Comparison shows that the guessed at X distribution is not so unreasonable for  > 10. However, it looks like we need to add more particles in the 4-10 range.

12. SR photon hits/crossing LER HER 748 215 1600 5300 4.4E4 1E4 1.3E6 7.5E5 1.1E5 1.8E7

13. SR photon hits/crossing on the detector beam pipe from various surfaces LER HER 0.24 0.07 10 13 111 13 8 9 968 105 Backscattering SA and absorption rate (3% reflected)

14. Energy Changes • For the QD0 and QF1 magnets we need to keep the ratio of the magnetic field strengths constant in order to maintain good field quality • We want the * values to remain constant to maintain luminosity • We need to match to the rest of the ring • No changes to the permanent magnets • Solutions found by iteration • Solutions found for all Upsilon resonances

16. Energy Changes • The 2S and the 3S LER energies would have very little polarization • It should be straightforward to develop a procedure to perform an energy scan • To go to the Tau-charm region (Ecm ~4 GeV) we will need to remove most if not all of the permanent magnets • With the air-core super quads we would need to approximately preserve the energy asymmetry • We might be able to get more creative by using the PMs to change the actual beam energies

17. Solenoid compensation • We have found out from our colleagues at KEK that we should pay much more attention to the fringe field of the detector solenoid • The radial part of the field causes emittance growth • This also means that we want to minimize the fringing fields of the solenoids • Catia et. al. have studied this for DANE and KLOE • Kirk has a presentation on this issue after this talk

18. Super-ferric QD0 and QF1 • Pavel Vobly from BINP has come up with a new idea for QD0 (mentioned previously) • Use Panofsky style quadrupoles with Vanadium Permendur iron yokes • This new idea has some added constraints but it is still attractive because it is easier to manufacture since the precision of the iron determines the quality of the magnet

19. Pictures from Vobly’s paper

20. The quads can be on axis with the beams

21. Super-ferric QD0 Vobly had a 2 T limit but we need 10% headroom for any above 4S energy scan • Constraints • Maximum field of no more than 1.8 T at the pole tips (we assume this is the same as the half width) • Equal magnetic field strengths in each twin quad • Square apertures • Might be able to relax these a little • If we have room between the windings to add Fe then we can have some magnetic field difference • Might be able to make the apertures taller than they are wide – means the windings get more difficult • For now assume constraints are there and then see what we can do

22. Permanent Magnets • Upon embarking on the task of looking at the Super-Ferric design we realized we could significantly improve the IR design by improving the permanent magnet performance by doing the following: • Give up some vertical aperture in order to go back to circular magnet designs (~1.4 stronger field) • Open up the crossing angle 10% to get more space for permanent magnet material • Add a couple of permanent magnet slices in front of the septum (shared magnets but close to the IP and hence minimal beam bending)

23. Permanent Magnets (2) • Moved some of the slices previously used on the HER to the LER in order to get more vertical focusing to the LER • We now have more equal vertical beta maximums • The beam pipe inside the magnets is 1 mm smaller in radius • 6 mm from 7 mm • The magnetic slices are now only 1 cm long and are perpendicular to the beam line instead of the detector axis • Better packing and better magnetic field performance for each beam

24. Permanent Magnets (3) • With a 6 mm inside radius beam pipe that is 1 mm thick and allowing for 0.5mm of space between magnet material and beam pipe we arrive at a 7.5 mm inside radius for the magnet material • The chosen remnant field of 13.4 kG is conservative. Some materials can reach 14-14.5 kG. All materials are Neodymium-Iron-Boron. Higher temperature dependence. • This gives us some headroom for packing fraction losses between magnetic blocks, etc. • There are two shared quad slices on either side of the IP in fairly close (0.17-0.21 m) • These magnets bend the beams slightly out in X increasing the beam separation for the other magnets • LER beam 1.864 mrad • HER beam 1.164 mrad

25. Details of the permanent magnet slices • Z from IP Len. R1 R2 G • Name Beam m cm mm mm T/cm • QDSA both 0.17 2 13 28 1.076 • QDSB both 0.19 2 14 30 0.994 • QDPA LER 0.30 1 7.5 12.5 1.392 • QDPB LER 0.31 1 7.5 13.0 1.473 • QDPC LER 0.32 1 7.5 13.5 1.547 • QDPD LER 0.33 1 7.5 14.0 1.616 • QDPE LER 0.34 1 7.5 14.5 1.680 • QDPF LER 0.35 1 7.5 15.0 1.740 • QDPG LER 0.36 1 7.5 15.5 1.796 • QPDH drift 0.37 1 • QDPI HER 0.38 1 7.5 16.5 1.899 • QDPJ HER 0.39 1 7.5 17.0 1.945 • QDPK HER 0.40 1 7.5 17.5 1.989 • QDPL HER 0.41 1 7.5 18.0 2.030 • QDPM HER 0.42 1 7.5 18.5 2.070 • QDPN HER 0.43 1 7.5 19.0 2.107 • QDPO HER 0.44 1 7.5 19.5 2.142 • QDPP HER 0.45 1 7.5 20.0 2.175 • QDPQ HER 0.46 1 7.5 20.5 2.207 • QPDR HER 0.47 1 7.5 21.0 2.238 • QDPS HER 0.48 1 7.5 21.5 2.266

26. Vanadium Permendur Design • We use the above redesigned permanent magnet slices • QD0 face is 55 cm from the IP. If we move in closer the field strength gets too high. In addition, we lose space for the stronger PM slices • We start by setting the LER side of QD0 and QF1 • We impose the beta function match requirements for the LER (* and the match point at 16.17 m) and we also try to get the maximum field close to 1.8 T • We keep the L* value constant but are allowed to change the separation and the lengths of QD0 and QF1 • These set the QD0 and QF1 strengths for the HER • Add another smaller defocusing quad to the HER behind QD0 to complete the vertical focusing for the HER • Also add another smaller focusing quad behind QF1 to complete the horizontal focusing of the HER • This design has much greater flexibility as far as beam energies are concerned

27. Vanadium Permendur Design

28. VP design details • PM as described above • Magnet QD0 QD0H QF1 QF1H • IP face (m) 0.55 0.90 1.25 1.70 • Length (m) 0.30 0.15 0.40 0.25 • G (T/cm) 0.938 0.707 0.407 0.381 • Aperture (mm) 33 (40) 49 (58)* 75 77 • Max. Field (T) 1.69 (1.88) 1.73(2.05) 1.53 1.47 • *increased aperture for increased magnetic offset for SR

29. Air-Core Italian Design • We replace the shared QD0 and QF1 parts of the VP design with the air-core design • Also place QD0 and QF1 parallel to the detector axis • Then we have the same field strengths and the LER and HER pieces are the same strength as the VP design

30. Air core “Italian” QD0, QF1

31. AC design details • PM as described above • Magnet QD0 QD0H QF1 QF1H • IP face (m) 0.55 0.90 1.25 1.70 • Length (m) 0.30 0.15 0.40 0.25 • G (T/cm) 0.938 0.707 0.407 0.381 • Aperture (mm) 36 49 (67)* 77 77 • Max. Field (T) 1.69 1.73(2.37) 1.57 1.47 • *increased aperture for increased magnetic offset for SR

32. Latest New Idea • We have discovered there are several rare earth metals that have very high magnetization curves • Holmium • Dysprosium • Gadolinium • Holmium has the highest magnetic moment of any element and is reputed to have a magnetization curve up to 4 T (Vanadium Permendur is about 2.4 T) • One of the reasons these metals are not used is that they only become ferromagnetic at temperatures well below room temperature (except for Gadolinium) • Curie temperatures • Ho is 20 K • Dy is 85 K • Ga is 289 K

33. Some properties of these metals* • Den. Young’s Shear Bulk Poisson Vickers Brinell Cost • Elem. g/cc Mod. Mod. Mod. Ratio Hard. Hard. $/kg • Ho 8.80 64.8 26.3 40.2 0.231 481 746 1000 • Dy 8.55 61.4 24.7 40.5 0.247 540 500 120 • Ga 7.90 54.8 21.8 37.9 0.259 570 --- <120 • Fe 7.87 211 82 170 0.29 608 590 0.4 (scrap) • Pb 11.35 16 5.6 46 0.44 --- 38.3 2 • Sn 7.31 50 18 58 0.36 --- 51 18 • Cu 8.96 120 48 140 0.34 369 874 15 • Ni 8.90 200 76 180 0.31 638 700 18 • Al 2.70 70 26 76 0.35 167 245 21 • Au 19.30 120 27 180 0.44 216 --- 34,000 • Zn 7.13 108 43 70 0.25 --- 412 2 • Ag 10.50 83 30 100 0.37 251 25 530 • *Wikipedia, Metalprices.com and VWR Sargent Welch These elements appear to be somewhere between Tin and Aluminum in hardness and strength with a density of Ni or Cu

34. Holmium Design • Set maximum field at 3.2 T which means 2.9 T max to allow for headroom to scan above the 4S • Shorten and bring the magnets closer together to lower beta maximums • Make apertures smaller when possible which allows us to increase the field strength

35. Holmium design details • PMs as described above • Magnet QD0 QD0H QF1 QF1H • IP face (m) 0.55 0.80 1.15 1.45 • Length (m) 0.20 0.10 0.25 0.15 • G (T/cm) 1.494 1.147 0.727 0.727 • Aperture (mm) 34 (39) 41 (49) 67 68 • Max. Field (T) 2.54(2.91) 2.35(2.81) 2.44 2.47

36. Holmium Design

37. Beta function comparison with V12 baseline • V12 VP/AC Ho • LER x max 316 309 221 • HER x max 388 480 328 • LER y max 1562 1424 1300 • HER y max 1266 1208 1111

38. SR backgrounds for the Super-ferric QD0 Slide from previous workshop • SR backgrounds have not been checked yet • The outward bending of the beams from the shared quads makes the SR shielding harder • We have some natural inward bending from the QD0 magnets which we need to steer the QF1 radiation away from the central chamber • We may find that the bending from the shared quads causes too much trouble but we would like to keep the option open as it improves the beta functions • SR studies may force some iterations to the design

39. SR backgrounds • We have been struggling to get a solution for the SR backgrounds • So far, none of the new designs work as well as the previous baseline design (concentrating on the HER so far) • Why? It looks like a combination of things • Opening the crossing angle to 66 mrad from 60 mrad costs us 3 mrad • The shared quads costs 1.1 mrad for the HER • Sloping the QD0 to match the beam angle in order to minimize the aperture costs some free QD0 bending • Moving QF1 closer to QD0 costs leverage. Offsetting QD0 is less effective. • The QF1 beta X max is reduced by moving in closer but apparently the leverage loss is a bigger effect • Strengthening PM reduces the strength of QD0 forcing a bigger offset for the same effect • Net result is that even with a beam bend angle of 10 mrad in QD0 we are not able to steer all of the SR away from the detector beam pipe (previous baseline was successful with a 1.6 mrad bend in QD0)

40. SR backgrounds (2) • The Holmium solution looks the best and that may be due to the stronger magnets and to the lower beta X max • Even this solution needs a least an 8 mrad bend angle from the HER QD0 magnets • Air-core quads is second best but needs a 10 mrad bend and still doesn’t quite get all of the direct SR off of the beam pipe • A 10 mrad bend puts a lot (kWs) of SR power on the beam pipe just upstream of the detector beampipe • The all vanadium permendur Panofsky quad solution deposits ~5kW on 10 cm of beam pipe – a power level we can not handle • However, we do not want to give up all of the lattice improvements either

41. SR backgrounds (3) • What to do? • Suggest the following compromises: • Remove the shared quads • Reduce the crossing angle back to 60 mrad • This hurts a little because we gained a lot by increasing the strength of the PMs. We would still have most of the improvement. • Set a maximum bend angle of 3-4 mrad in the incoming QD0 magnets • We will want to have a zero bend angle in the outgoing QD0 and QF1 magnets • This favors the panofsky style quads for the outgoing beams • We presently scan out to 20  in x. We scanned out to 10  for PEP-II. Propose scanning out to 12-15  in x.

42. Baseline Summary • The present baseline design for the IR is the self-compensating air core dual quad QD0 and QF1 design • All the magnets inside the detector are either PM or SC • The beam pipes inside the cryostats are warm • We have a 30 BSC in X and 100-140 BSC in Y (7-10 fully coupled) • Synchrotron radiation backgrounds look ok, but needs more study • This is the White paper (CDR2) design • Radiative bhabha backgrounds should be close to minimal – nearly minimal beam bending

43. Super- ferric Summary • We have three designs that include super-ferric magnets • One design has all vanadium permendur magnets • A second design replaces the shared QD0 and QF1 magnets with the air core “Italian” design • The Third design uses Holmium • The simplicity of construction and the ability to decouple some of the magnetic elements make the idea very attractive

44. Super-ferric Summary (2) • The SR backgrounds are proving to be troublesome. Several weeks of work and no satisfactory solution yet… • We are pretty sure it is a combination of design changes that has caused this problem • We have some proposed compromises we want to try… • Work in progress…….

45. Conclusions • The flexibility of the IR design has been improved by re-optimizing the permanent magnets • We have more focusing in closer to the IP now • We have three designs for the Panofsky style magnets • Vanadium Permendur • Holmium • Air-core + vanadium permendur • The new designs greatly improve the lattice and energy flexibility of the IR design but we need to find a reasonable set of parameters that also satisfy the SR background demands • (superKEK has the same problem)