Status on HL-LHC heating B. Salvant for the impedance team Many thanks to: Gianluigi Arduini, David Belohrad, Riccardo de Maria, Stephane Fartoukh, Giovanni Iadarola, Thibaut Lefevre, Elias Metral, Nicolas Mounet, Giovanni Rumolo, Ralph Steinhagen, Simon White, Carlo Zannini. HL-LHC WP2 task leader meeting - October 19 th 2013
Status on HL-LHC heating. B. Salvant for the impedance team Many thanks to: Gianluigi Arduini, David Belohrad , Riccardo de Maria, Stephane Fartoukh, Giovanni Iadarola, Thibaut Lefevre, Elias Metral, Nicolas Mounet, Giovanni Rumolo, Ralph Steinhagen , Simon White, Carlo Zannini. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Status on HL-LHC heating
B. Salvant for the impedance team
Many thanks to: Gianluigi Arduini, David Belohrad, Riccardo de Maria, Stephane Fartoukh,
Giovanni Iadarola, Thibaut Lefevre, Elias Metral, Nicolas Mounet, Giovanni Rumolo, Ralph Steinhagen, Simon White, Carlo Zannini.
HL-LHC WP2 task leader meeting - October 19th 2013
Agenda
• Beam spectra for 800 MHz?• Scaling to PIC1, PIC2, US1, US2• Potential issues for hardware (difficult to
predict)
bunch flattening of the LHC beam at 7 TeV(ESME Simulations)
Chandra Bhat 3
Vrf(400MHz)=16MV
E vs t
Line charge Distribution
Energy Distribution
E vs t
Line charge Distribution
Energy Distribution
Vrf(400MHz)=16MV +Vrf(800MHz)=8.5MV
Normal Bunch Flattened BunchMountain Range
RMS Bunch Length vs Time
RMS Energy Spread vs Time
2.5 eVs
z=7.5cm
E=3.2GeVrms=0.72GeV
lb=41cm
E=2.6GeVrms=0.6GeV
C. Bhat 2009
Bunch with 800 MHz
Function proposed by Elena:Rho=exp(-t4/(2*sig4)) with sig=0.45 ns
Function proposed by Stephane:Rho=(1+erf(4*(1-abs(t)/0.5e-9)))/2
With fitted parameters to the curve obtained in ESME
Flat bunches (C. Bhat, 2009) quotedby F. Zimmermann and S. Fartoukh
Vrf(400MHz)=16MV +Vrf(800MHz)=8.5MV
Bunch in 400 MHz (coarse fit from Chandra)
Power spectra
Flat bunches excite less power below 1.1 GHz, but more above
Flat bunches not far from single RF situation in terms of heating for very broadband impedances (constant over frequency). Effect will depend on the spectrum of each device.
Agenda
• Beam spectra for 800 MHz?• Scaling to PIC1, PIC2, US1, US2• Potential issues for hardware (difficult to
predict)
Increase in heat load only from intensity increase
*Narrow band is a worst case scenario assuming that the resonance stands exactly at a multiple of 40 MHz
Increase in heat load only from intensity increase (table)
Current beam screens• Expected from theory, accounting for the weld on the side (+44%, see PhD of
Andrea Mostacci and Carlo Zannini) and magnetoresistance (for instance in PhD of Nicolas Mounet, pessimistic for quadrupoles), accounting for factor 2 in addition (could be worst case for 2 beams in same aperture, pessimistic). Note: optics aperture chosen instead of mechanical aperture (more pessimistic).
For the arcs, cooling power is 200 W per half cell (i.e. 3800 mW/m). Is that enough margin for synchrotron radiation and electron cloud?
Could also be limiting for standalones and triplets (if cooling power is 250 W 8300 mW/m).
Half gap 18.4 mm (arc) 290 509 165 386 409 733 1040 1060 1350
Half gap 24 mm(inner triplets Q2 and Q3) 445 781 253 592 628 1120 1590 1630 2060
Half gap 18.95 mm (inner triplets Q1) 563 989 321 750 795 1420 2020 2060 2610
New triplet beam screens• Expected from round pipe theory as in talk of Nicolas Mounet on new triplets
(01/07/13), accounting for the weld on the side (+10% to 25% estimated by Carlo Zannini depending on the position of the weld) and magnetoresistance (pessimistic for quadrupoles), accounting for factor 2 in addition (could be worst case for 2 beams in same aperture, pessimistic), not yet accounting for the change of impedance linked to the transverse position inside the triplets (currently studied by Carlo and GIovanni).
TDI without coating (55 mm half gap) -> collision param.
60 67 219 283
TDI with 1 micron copper coating (55 mm half gap)-> collision param.
2.9 3.2 10.5 13.6
TDI without coating (3.8 mm half gap) -> injection param.
640 705 2330 2985
TDI with 1 micron copper coating (3.8 mm half gap)-> injection param.
32 35 116 150
Not relevant (as TCTP), as needs to be completely refurbished (cut in 5 different collimators with different materials and new cooling).Current cooling for TCP in the range of the needed cooling
Elements to be considered• Beam screens
– arcs, standalones, inner triplets• Collimators
– TDI, single beam collimators, TCTP, TCDQ• Kickers
No detailed update from TE/ABT for PIC/US parameters, other high priority activities for the moment. Will come soon , but should be in line with the general scaling: PIC1 39-67 W/m
PIC2 41-71 W/m US1 75-126 W/m US2 106-180 W/m ( could be limiting with the upgraded hardware)
Elements to be considered• Beam screens
– arcs, standalones, inner triplets• Collimators
– TDI, single beam collimators, TCTP, TCDQ• Kickers
Are these heat loads an issue for the stripline?Losses are mostly in the cables possibility to use attenuators , and need to change electronics. Signal can be perturbed with so much attenuation (loss of bandwidth).Already checked for 3.5e11 OK for the peak power
Wire scanner could be an issue due to the cavity.
BTV ok thanks to the shielding chamber.
WCM same issue as stripline
Current transformers new design under way (temperature probes installed but never used)
New BSRT design under fabrication (without ferrites)
Summary• Beneficial effect of studied flat bunches on heating is not evident. However, important
to point out that the MD in 2012 showed beneficial effect for several devices.
• Upgrades of electronics for instrumentation would be needed, and the MKI upgrade foreseen for LS1 may not be enough for US1 and especially US2.
• Current TDI is already not enough for current power loss. Needs upgrade of design and especially of the cooling.
• Impedance will take a significant portion of the cooling power for the arcs and the triplets (10 to 30%). Is there enough margin for the other contributions (ecloud and synchrotron radiation?)
• Effect of “real” filling scheme and bunch length for 25 ns could also be checked
General consideration on power loss for HL-LHC parameters (1/3)
• intensity per bunch increase Ploss Nb
2
for 25 ns: 1.15e11p/b to 2.2e11 p/b leads to a factor 3.7 for 50 ns: 1.6e11 p/b to 3.5e11 p/b leads to a factor 4.8
• Number of bunches 2808 for 25 ns and 1404 for 50 ns for broadband impedance, what matters is M*Nb
2
for narrow band impedance, what matters is (M*Nb)2
Factor from situation before LS1
Nominal 25 ns (2808*1.15)
Before LS150 ns (1374*1.6e11)
HL-LHC25 ns(2808*2.2e11)
HL-LHC50 ns(1404*3.5e11)
Broadband (M*Nb
2)*1.05 1 *3.9 *4.9
Narrow band (M*Nb)2
*2.1 1 *7.9 *5
Significant increase of power loss expected for HL-LHC intensities
1
2 22Re2p
revrevlongrevbloss pMfrumPowerspectpMfZfeMNP
General consideration on power loss for HL-LHC parameters (3/3)
• intensity per bunch increaseFactor from situation before LS1
Nominal 25 ns (2808*1.15)
Before LS150 ns (1374*1.6e11)
HL-LHC25 ns(2808*2.2e11)
HL-LHC50 ns(1404*3.5e11)
Broadband (M*Nb
2)*1.05 1 *3.9 *4.9
Narrow band (M*Nb)2
*2.1 1 *7.9 *5
Significant increase of power loss expected for HL-LHC parameters (factor 5 to 7)
• Accounting for decrease of bunch length (*1.4)Factor from situation before LS1