1. Heat load due to e-cloud in the HL-LHC triplets G. Iadarola, G. Rumolo 19th HiLumi WP2 Task Leader Meeting - 18 October 2013 Many thanks to: H.Bartosik, R. De Maria, R. Tomas, L. Tavian

2. Outline Present inner triplets o Layout and optics o PyECLOUD simulations o Heat load measurements and estimate of the SEY Inner triplets for the HL-LHC upgrade o Layout and optics o Heat load estimates from PyECLOUD simulations

3. Outline Present inner triplets o Layout and optics o PyECLOUD simulations o Heat load measurements and estimate of the SEY Inner triplets for the HL-LHC upgrade o Layout and optics o Heat load estimates from PyECLOUD simulations

4. B2L1A2L1 3L1 1L1 Machine layout and optics near IP1 IP1, 4 TeV, β* = 0.6 m IP1 1R1 A2R1B2R1 3R1

5. Machine layout and optics near IP1 IP1, 4 TeV, β* = 0.6 m IP1 1R1 A2R1B2R1 3R1

6. Machine layout and optics near IP1 IP1, 4 TeV, β* = 0.6 m IP1 1R1 A2R1B2R1 3R1

7. Machine layout and optics near IP1 IP1, 4 TeV, β* = 0.6 m IP1 1R1 A2R1B2R1 3R1

8. Machine layout and optics near IP1 IP1, 4 TeV, β* = 0.6 m IP1 1R1 A2R1B2R1 3R1

9. The longitudinal direction The delay between two following bunch passages changes along the machine elements Locations of long range encounters spaced by half the bunch spacing 7.5 m Beam 2 Beam 1 IP1 25 ns spac.

10. The longitudinal direction The delay between two following bunch passages changes along the machine elements Locations of long range encounters spaced by half the bunch spacing 7.5 m Beam 2 Beam 1 IP1 25 ns spac.

11. Outline Present inner triplets o Layout and optics o PyECLOUD simulations o Heat load measurements and estimate of the SEY Inner triplets for the HL-LHC upgrade o Layout and optics o Heat load estimates from PyECLOUD simulations

12. PyECLOUD simulations for the inner triplets Quite a challenging simulation scenario: Two beams with: o Different transverse position (off center) o Different transverse shape o Arbitrarily delayed with respect to each other (no real bunch spacing)

13. PyECLOUD simulations for the inner triplets Quite a challenging simulation scenario: Two beams with: o Different transverse position (off center) o Different transverse shape o Arbitrarily delayed with respect to each other (no real bunch spacing) Quadrupolar magnetic field Non elliptical chamber

14. PyECLOUD simulations for the inner triplets Quite a challenging simulation scenario: Two beams with: o Different transverse position (off center) o Different transverse shape o Arbitrarily delayed with respect to each other (no real bunch spacing) Quadrupolar magnetic field Non elliptical chamber Required the introduction of new features in PyECLOUD

15. PyECLOUD simulations for the inner triplets Quite a challenging simulation scenario: Two beams with: o Different transverse position (off center) o Different transverse shape o Arbitrarily delayed with respect to each other (no real bunch spacing) Quadrupolar magnetic field Non elliptical chamber Required the introduction of new features in PyECLOUD Beam properties (delay, position, size) changing along the structure: Simulated 10 slices per magnet SEY = 1.0, 1.1, …, 2.0 50 ns and 25 ns

16. PyECLOUD simulations for the inner triplets Quite a challenging simulation scenario: Two beams with: o Different transverse position (off center) o Different transverse shape o Arbitrarily delayed with respect to each other (no real bunch spacing) Quadrupolar magnetic field Non elliptical chamber Required the introduction of new features in PyECLOUD Beam properties (delay, position, size) changing along the structure: Simulated 10 slices per magnet SEY = 1.0, 1.1, …, 2.0 50 ns and 25 ns ~1000 simulations, ~48 h for two bunch trains, ~120 GB of sim. data

17. Simulated scenarios 8 72 8 8 25 ns beam – 1.15 x 10 11 ppb 388 72 8 8 38 4 36 4 4 50 ns beam – 1.65 x 10 11 ppb 194 36 4 4 19 Other beam parameters: Beam energy: 4 TeV Transverse emittance: 2.4 μm Bunch length: 1.3 ns Optics with β* = 0.6 m

18. Few snapshots of the electron distribution (50 ns spacing) Beam 1 A look to the EC buildup

19. The presence of two beams enhances the multipacting efficiency, especially far enough from the locations of the long range encounters A look to the EC buildup

20. The presence of two beams enhances the multipacting efficiency, especially far enough from the locations of the long range encounters A look to the EC buildup

21. Heat load density along the triplet - 50 ns Locations of long range encounters Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly 1R1 A2R1B2R13R1

22. Heat load density along the triplet - 50 ns 1R1 A2R1B2R13R1 Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly The presence of the two beams leads to an important lowering of the multipacting threshold w.r.t. the single beam case (dashed)

23. Heat load density along the triplet - 50 ns 1R1 A2R1B2R13R1 Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly The presence of the two beams leads to an important lowering of the multipacting threshold w.r.t. the single beam case (dashed)

24. Heat load density along the triplet - 50 ns 1R1 A2R1B2R13R1 Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly The presence of the two beams leads to an important lowering of the multipacting threshold w.r.t. the single beam case (dashed)

25. Heat load density along the triplet - 50 ns 1R1 A2R1B2R13R1 Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly The presence of the two beams leads to an important lowering of the multipacting threshold w.r.t. the single beam case (dashed)

26. Heat load density along the triplet - 25 ns Locations of long range encounters Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly 1R1 A2R1B2R13R1

27. Heat load density along the triplet - 25 ns Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly As in the other quardupoles in the LHC, the multipacting threshold is already quite low even for a single beam with 25 ns spacing 1R1 A2R1B2R13R1

28. Heat load density along the triplet - 25 ns Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly The presence of the two beams leads to an important lowering of the multipacting threshold w.r.t. the single beam case (dashed) 1R1 A2R1B2R13R1

29. Heat load density along the triplet - 25 ns Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly The presence of the two beams leads to an important lowering of the multipacting threshold w.r.t. the single beam case (dashed) 1R1 A2R1B2R13R1

30. Heat load density along the triplet - 25 ns Remarks: EC much weaker close to long range encounters Modules with the same beam screen behave very similarly The presence of the two beams leads to an important lowering of the multipacting threshold w.r.t. the single beam case (dashed) 1R1 A2R1B2R13R1

31. Average heat load density – 50 ns vs. 25 ns Remarks: Above threshold values are very similar  enhancement from hybrid spacings much stronger on the 50 ns than on 25 ns For each triplet we get:

32. Outline Present inner triplets o Layout and optics o PyECLOUD simulations o Heat load measurements and estimate of the SEY Inner triplets for the HL-LHC upgrade o Layout and optics o Heat load estimates from PyECLOUD simulations

33. Machine observations – 50 ns No big change during ramp and squeeze Data provided by L. Tavian

34. Machine observations – 50 ns No big change during ramp and squeeze

35. Machine observations – 50 ns Heat load much higher than in two-aperture quadrupoles Data provided by L. Tavian

36. Machine observations – 25 ns (4 TeV) Data provided by L. Tavian No big change during ramp and squeeze

37. Machine observations – 25 ns (4 TeV) No big change during ramp and squeeze

38. Machine observations – 25 ns (4 TeV) Data provided by L. Tavian Heat load similar to two-aperture quadrupoles

39. Comparison against simulations Fill 3429 Fill 3286 Fill 3405 50 ns, 4 TeV 25 ns, 450 GeV 25 ns, 4 TeV 1.2 < SEY max < 1.3 is compatible with the observations

40. Crosscheck with single beam MDs For the estimated SEY value we do not expect nor observe EC in the triplets with only one beam with 50 ns spacing circulating in the LHC

41. Crosscheck with single beam MDs For the estimated SEY value we do not expect nor observe EC in the triplets with only one beam with 50 ns spacing circulating in the LHC

42. Outline Present inner triplets o Layout and optics o PyECLOUD simulations o Heat load measurements and estimate of the SEY Inner triplets for the HL-LHC upgrade o Layout and optics o Heat load estimates from PyECLOUD simulations

43. Optics, layout and apertures Q1 (A/B) Q2 (A/B) Q3 (A/B)D1 Q1 Q2 (A/B) Q3 D1 Present HiLumi IP1, 4 TeV, β* = 0.6 m IP1, 7 TeV, β* = 0.15 m

44. Optics, layout and apertures Q1 (A/B) Q2 (A/B) Q3 (A/B)D1 Q1 Q2 (A/B) Q3 D1 For the moment, only quadrupoles have been simulated IP1, 7 TeV, β* = 0.15 m Thanks to R. De Maria

45. Optics, layout and apertures Q1 (A/B) Q2 (A/B) Q3 (A/B)D1 Q1 Q2 (A/B) Q3 D1 Present HiLumi IP1, 4 TeV, β* = 0.6 m IP1, 7 TeV, β* = 0.15 m

46. Optics, layout and apertures Q1 (A/B) Q2 (A/B) Q3 (A/B)D1 Q1 Q2 (A/B) Q3 D1 Present HiLumi IP1, 4 TeV, β* = 0.6 m IP1, 7 TeV, β* = 0.15 m

47. Optics, layout and apertures Q1 (A/B) Q2 (A/B) Q3 (A/B)D1 Q1 Q2 (A/B) Q3 D1 Present HiLumi IP1, 4 TeV, β* = 0.6 m IP1, 7 TeV, β* = 0.15 m

48. Optics, layout and apertures Q1 (A/B) Q2 (A/B) Q3 (A/B)D1 Q1 Q2 (A/B) Q3 D1 Present HiLumi IP1, 4 TeV, β* = 0.6 m IP1, 7 TeV, β* = 0.15 m

49. Optics, layout and apertures Q1 (A/B) Q2 (A/B) Q3 (A/B)D1 Q1 Q2 (A/B) Q3 D1 Present HiLumi IP1, 4 TeV, β* = 0.6 m IP1, 7 TeV, β* = 0.15 m

50. Optics, layout and apertures Q1 (A/B) Q2 (A/B) Q3 (A/B)D1 Q1 Q2 (A/B) Q3 D1 Present HiLumi IP1, 4 TeV, β* = 0.6 m IP1, 7 TeV, β* = 0.15 m

51. Outline Present inner triplets o Layout and optics o PyECLOUD simulations o Heat load measurements and estimate of the SEY Inner triplets for the HL-LHC upgrade o Layout and optics o Heat load estimates from PyECLOUD simulations

52. Simulated scenario 8 72 8 8 25 ns beam – 2.20 x 10 11 ppb 388 72 8 8 38 Other beam parameters: Beam energy: 7 TeV Transverse emittance: 2.5 μm Bunch length: 1.0 ns Optics with β* = 0.15 m

53. A look to the EC buildup – HiLumi triplets Few snapshots of the electron distribution  much wider stripes for HiLumi triplets HiLumi Present

54. Heat load along the triplet Present triplets HiLumi triplets Locations of long range encounters

55. Bunch intensity is larger but also chamber is wider  energy of multipacting electrons is quite similar Present tripletsHiLumi triplets Heat loads

56. Bunch intensity is larger but also chamber is wider  energy of multipacting electrons is quite similar  number of impacting electrons about x2 larger Heat loads Present tripletsHiLumi triplets

57. Bunch intensity is larger but also chamber is wider  energy of multipacting electrons is quite similar  number of impacting electrons about x2 larger  Total heat load about x2 larger Heat loads Present tripletsHiLumi triplets D1 and correctors not included! SEY max in the present triplets at the end of 2012

58. Summary and conclusions Simulations show a strong enhancement of the EC due to the presence of the two beams  extremely low multipacting thresholds (SEY th <1.1 for 25 ns beams) Comparing measured heat load against simulation we infer SEY max = ~1.3 for the present triplets (crosschecked in different beam conditions ) Simulations for the HiLumi scenario show x2 stronger heat load w.r.t. the present scenario (SEY max = ~1.3  HL = ~500 W – D1 and correctors to be added) Heat load could be strongly reduced by EC mitigation strategies: o Low SEY coating (but we need SEY max <1.1 to have full suppression) o Clearing electrode (EC localized in a region where there should be enough aperture to place a polarized plate/wire)

59. Thanks for your attention!

60. Machine layout and optics near IP1 - present Common beam pipe Separated beam pipes IP1 IP1, 4 TeV, β* = 0.6 m

61. Machine layout and optics near IP1 - HiLumi Common beam pipe Separated beam pipes IP1 IP1, 4 TeV, β* = 0.6 m

62. B2L1A2L1 3L1 1L1 IP1, 4 TeV, β* = 0.6 m IP1 1R1 A2R1B2R1 3R1 Machine layout and optics near IP1 - present

63. B2L1A2L1 3L1 1L1 IP1, 4 TeV, β* = 0.6 m IP1 1R1 (A/B) 2R1 (A/B) 3R1 (A/B) Machine layout and optics near IP1 - HiLumi

64. The longitudinal direction The delay between two following bunch passages changes along the machine elements Locations of long range encounters spaced by half the bunch spacing 15 m Beam 2 Beam 1 IP1 50 ns spac.

65. Average heat load density – 50 ns vs. 25 ns Remarks: Above threshold values are very similar  enhancement from hybrid spacings much stronger on the 50 ns than on 25 ns

66. A look to the EC buildup – HiLumi triplets Few snapshots of the electron distribution  much wider stripes

67. Few snapshots of the electron distribution Beam 1 A look to the EC buildup – present triplets

68. A look to the EC buildup – HiLumi triplets Few snapshots of the electron distribution  much wider stripes