2. LIU-ABT systems: SPS, W. Weterings M.J. Barnes, L. Ducimetière, N. Magnin, W. Weterings Acknowledgements: A Adraktas, M. Beck, G. Bellotto, S. Bouleghlimat, G. Bregliozzi, P. Burkel, E. Carlier, D. Comte, E. Chapochnikova, T. Fowler, H. Day, C. Pasquino, P.C. Pinto, G. Rumolo, B. Salvant, Y. Sillanoli, J. A. Ferreira Somoza, M. Taborelli, C. Zannini 20th November 2015 20/11/2015 SPS LIU-ABT Systems Review: MKE2

3. Content Present MKE configurations Expected heating issues for HL-LHC type beams MKE4 reconfiguration during YETS 2015-16 Possible mass-upgrade of MKE magnets Budget and Spending Profile Manpower and Resources Vacuum – aC? E-field simulations 20/11/2015 SPS LIU-ABT Systems Review: MKE3

4. Present MKE6 Configuration 20/11/2015 SPS LIU-ABT Systems Review: MKE4 MKE6: 3 magnets, 1 PFN & 1 short-circuit. 33.1kV PFN nominal. 616316163461637 L10L8S3 Serigraphed 8C11 BEAM PFN Short Circuit Magnet #1 Magnet #2Magnet #3 Enamelled Flange Serigraphy: HV Plates Ferrite Serigraphy (200mm long “fingers”)

5. MKE4 Configuration 20/11/2015 SPS LIU-ABT Systems Review: MKE5 MKE4 from YETS 2015-16: 4 magnets, 1PFN & 1 short-circuit. 30.7kV PFN nominal. Enamelled Flange BEAM PFN Short Circuit Magnet #1 Magnet #2 Magnet #3 Enamelled Flange Magnet #4 41631416344163741651 L1L9S6S7 Serigraphed 8C11 New Serigraphy 8C11 Serigraphed 8C11 Magnets in locations 41651/41654 to be exchanged/replaced by beam-pipe. o Clipper (initially) unused – installed but cold and no trigger; o Magnet S7 acceptance tested by VSC, & HV conditioning is successfully completed. Retain enamel flanges either side of MKE4 magnets; No damping resistor in bellows. As per pre-YETS MKE4 @ present: 5 magnets, each with a PFN & TMR. 51.2kV PFN nominal. 4163141634416374165141654 L1L9S6S4L5 Serigraphed 8C11 Magnet #1 Magnet #2Magnet #3Magnet #4Magnet #5 Serigraphed 8C11 Enamelled Flange BEAM Note: L5 had vacuum leak on tank flange PFN TMR

6. Predicted power loss due to serigraphy peak as a function of frequency of the peak (σ=20 cm): 200 mm Courtesy of C. Zannini ~70% of total ~60% of total 180 mm 200 mm  50 MHz 180 mm  56 MHz Motivation to reduce length of serigraphy (new S7): 1)200mm serigraphy: 1650 W, with cooling water temperature of 28˚C  T=123˚C: close to Curie point of ferrite. Theory: power expected to half with 180mm serigraphy. 2)To test modified serigraphy on a single MKE before mass upgrade. Power deposition based on lab measurements: 1)Short-circuiting magnet increases (~16%) power deposition; 2)For 25 ns beam, 180 mm serigraphy gives ~8% lower power deposition than 200 mm. 1150W  ~95˚C ferrite. 3)For 50 ns beam, 180 mm serigraphy gives ~9% higher power deposition than 200 mm.1460W  ~112˚C ferrite. Notes: frequency of resonant peak is ~15% higher in measurements than in predictions, e.g. without capacitors or load: hence very important to check relative dissipation in MKE. Measured Real Impedance: Expected Heating Issue for HL-LHC type beams 20/11/2015 SPS LIU-ABT Systems Review: MKE6 Theoretical beam induced power deposition for: 25ns (50ns) beam and 1.1x10 11 (1.6x10 11 ) ppb  320W (230W). For HL-LHC, 25ns (50ns) beam and 2.5x10 11 (3.6x10 11 ) ppb  1650W (1150W). Courtesy of H. Day

7. MKE: Measured water temperature An average of Up and Down temperatures, for all MKE magnets in a system, show that MKE6 magnets (short-circuit) have ~16% higher temperature rise than the MKE4 magnets (terminated) – which is consistent with the impedance measurements in the lab. ΔT=15.2˚C ΔT=13.0˚C 20/11/2015 SPS LIU-ABT Systems Review: MKE7 Post YETS: Closely monitor absolute and relative, to other MKEs, temperatures, to determine effectiveness of new serigraphy.

8. MKE4-Reconfiguration Overview ECA4 plan – unused hardware Generator 4 kept as operational spare. 20/11/2015 SPS LIU-ABT Systems Review: MKE8

9. 20/11/2015 MKE4: modifications during YETS 2015-2016 14/12/2015:  Replace 1 magnet (L5 – vacuum leak on tank) by 159mm diameter beam pipe;  Replace 1 magnet (S4) with spare magnet (S7) with 180 mm long serigraphy;  Reclose vacuum ready for pump-down. After 14/12/2015: - Two weeks of vacuum pumping in LSS4 during Christmas break; - Generators and control systems simplification in ECA4 carried out during W51 2015 and W2 2016, in parallel with LSS4 activities; - In January, one week of HV DC magnet conditioning in LSS4 (3 DC cond. units exist, 2 in production) - Installation of magnet capacitors, entry boxes and cabling; - One week for control and protection systems commissioning in ECA4; - Three weeks of HV pulsed magnet conditioning; - Tests with beam, e.g. scan of MKE4 waveform, steering of transfer lines, etc… SPS LIU-ABT Systems Review: MKE9

10. 20/11/2015 SPS LIU-ABT Systems Review: MKE10 25 ns Scrubbing (W25, 2015): Ecloud Ecloud is present in the MKEs with 25ns beam. Possible future development: coating the, aperture side, surface of the ferrites with Cr 2 O 3 or amorphous Carbon (aC) could eliminate Ecloud. Simulations indicate that this also reduces the Electric field adjacent to the serigraphy..

11. 20/11/2015 Modification of Serigraphy and possible aC Coating SPS LIU-ABT Systems Review: MKE11 Serigraphy: length reduced, from 200 mm to 180 mm, by scraping with a knife and finishing with emery cloth. This process avoids the need to purchase new ferrite (estimated cost of ferrite: 45 kCHF/magnet). aC Coating: check influence of cooling ferrite to below Curie temperature in a magnetic field; aC trial on a piece of ferrite required to check for issues of adhesion; aC coating of parts from the machines: OK if they are in the μS/hr range; Initial simulations indicate a significant reduction of electric field, adjacent to serigraphy, by applying a thin layer of aC – but instantaneous power dissipation??

12. LS 4LS 5Run 5 LS2 starting in 2018 (July)=> 18 months + 3 months BC LS3LHC: starting in 2023 =>30 months + 3 months BC Injectors: in 2024=>13 months + 3 months BC LHC schedule beyond LS1 Beam commissioning (BC) Technical stop (TS) Shutdown Physics (Extended) Year End Technical Stop: (E)YETS With the extra L-type spare (removed from MKE4 during reconfiguration)  2L+1S spares, i.e.: 3 MKEs, e.g. at MKE4, can be swapped during EYETS 2016-2017; 3 MKEs, e.g. at MKE6, can be swapped during YETS 2017-2018. 20/11/2015 SPS LIU-ABT Systems Review: MKE12 Run 2 Run 3

13. If experience with the 180 mm serigraphy shows all other MKE magnets should be modified: 3 remaining magnets @ MKE4 [2L, 1S] or MKE6 [2L, 1S] could potentially be exchanged in an EYETS or YETS. Total of 9 or 10 MKEs to be upgraded: Following spending profile is: with new serigraphy, large seals for tank (~2 kCHF/magnet – also includes rough estimate for aC coating of ferrite (~5 kCHF/magnet); without new ferrite (assuming that serigraphy can be modified (shortened) by scraping off end of existing 200 mm serigraphy and using emery cloth) or valves. Estimated spending per year in kCHF: Year20162017201820192020Total 45 45 + 80? 15 150 + 80? Spending Profile for Rebuilding MKEs Cooling option in purple. 20/11/2015 SPS LIU-ABT Systems Review: MKE13 Year 201520162017201820192020 PV 1963525 18 EV 390 AC 1164 Metrics per year in kCHF:

14. Manpower/Resources After YETS 2015-16 there will be 3 spare MKE magnets [2L, 1S]. For rebuilding, testing etc. 6 (9 inc. spares) magnets and installing 6 magnets: ~ 20/29 person months with manual conditioning; ~ 15/22 person months with automatic conditioning Up to 3 person months for new automatic conditioning system for lab (+ some hardware)  good investment. No detailed manpower planning available yet - conceivable that additional manpower is needed. Assembly area for magnet and generator will be 867 (within controlled zone). HV-Test cage for system testing is needed 2016 - 2019. For chiller system: 6 person months for design and installation. 20/11/2015 SPS LIU-ABT Systems Review: MKE14

15. Conclusions Estimates presented for reducing length of serigraphy for all installed MKEs & 3 spares; Consider trial of coating of aperture surface of ferrites e.g. with aC; A chiller would give additional safety margin (especially useful if bunch length < 20 cm); Replacing ferrite with a higher Curie Temperature version – expensive and probably not required; Vacuum valves – expensive therefore not included. 20/11/2015 SPS LIU-ABT Systems Review: MKE15

16. THANK YOU FOR YOUR ATTENTION!

17. Predicted power loss due to serigraphy peak as a function of frequency of the peak (σ=20 cm): 200 mm Courtesy of C. Zannini ~70% of total ~60% of total 180 mm 200 mm  50 MHz 180 mm  56 MHz Motivation to reduce length of serigraphy: 1)200mm serigraphy: 1650 W, with cooling water temperature of 28˚C  T=123˚C: close to Curie point of ferrite. 2)To test modified serigraphy on a single MKE before mass upgrade. Notes: frequency of the peak is higher in measurements than in predictions, e.g. without capacitors or load: a)new (180mm long) serigraphy: 49 MHz pred., 56 MHz meas. b)old (200mm long) serigraphy: 44 MHz pred., 50 MHz meas. Conclusions re power deposition from above measurements: 1)Short-circuiting magnet can significantly (~16%) increase power deposition; 2)For 25 ns beam, new (180 mm long) serigraphy is ~8% lower power deposition than 200mm long. 1150W  ~95˚C ferrite. 3)For 50 ns beam, old (200 mm long) serigraphy is ~9% lower power deposition than 180mm long.1460W  ~112˚C ferrite. Measured Real Impedance: Expected Heating Issue 20/11/2015 SPS LIU-ABT Systems Review: MKE 17 Courtesy of C. Zannini

18. Ferrite Temperature  MKEs are currently cooled with demineralised water (20-28˚C).  Water cooling allows the MKE to operate with ~twice the beam induced power deposition (ref: AB-Note-2004-005 BT (Rev.2)).  With the expected beam induced power deposition of 1.65kW/magnet for HL-LHC (with the present serigraphy), the estimated (actual) ferrite temperature is:  ~125˚C for 25ns beam (corresponds to the Curie temperature);  ~105˚C for 50ns beam (uncomfortably close to the Curie temperature).  Mixed (12˚C) or chilled (6˚C) water would give a reasonable reduction (up to ~20˚C) in ferrite temperature.  Back of the envelope calculations indicate that increasing the emissivity of the inside of the MKE tanks (as per the MKIs) could result in a significant radiated power (15-25%), further improving cooling of the ferrite – but current tanks could be radioactive, thus difficult to treat.  Samples of CMD10 ferrite have been obtained for evaluation for possible future use in the MKIs: CMD10 has a Curie temperature of ~250˚C. The CMD10 samples have been tested and Oked by the vacuum team. CMD5005/CMD10/CN20, or a mix of some of these, could give a significant thermal operating margin, for the MKEs, for HL-LHC beam. Equivalent to 8C11 CMD10 also has a saturation flux-density (Bs) greater than CMD5005 but has a lower u’ (acceptable for MKIs, but needs to be checked for MKEs) 20/11/2015 SPS LIU-ABT Systems Review: MKE18

19. Ferrite Power Deposition 20/11/2015 SPS LIU-ABT Systems Review: MKE19 For given properties of the ferrite (permittivity and permeability [which are frequency dependent]), the length/overlap of the serigraphy determines the resonant frequency.  Reducing the length of serigraphy from 200 mm to ~180 mm is expected to increase the resonant frequency from 44 MHz to 49 MHz, close to an optimum for 25 ns and 50 ns beams.  The total power deposition is expected to be reduced from ~1650 W to ~850 W for 25 ns beam, and ~1150 W to ~820 W for 50 ns beam: 850 W  80˚C estimated with water @ 28˚C.  Measurements (on a single cell or MKE magnet?) are required to confirm the predictions and hence the optimum length of serigraphy. Measurements on an S-type magnet would leave an L–type spare. 230 mm (length of ferrite block) 200 mm length (presently) Predicted power loss due to serigraphy peak as a function of frequency of the peak (σ=20 cm): 200 mm Courtesy of Carlo Zannini ~70% of total ~60% of total

20. Budget breakdown (updated from Nov. 2014) The following assumes that MKE4 is kept as per present (5 terminated MKEs): 2 less MKEs are required for a short-circuit configuration.  Total budget for new serigraphy: ~30 kCHF/10 mags (including labour for removing end (~20mm) of present serigraphy, cleaning);  Total manpower (FSU) for assembly and installation: ~50 kCHF/10 mags (6 person months);  Modify lifting device to reduce radiation exposure of personnel: ~5 kCHF;  Total budget for new seals for vacuum tanks: ~20 kCHF/10 mags ;  Estimated total budget for aC: ~50kCHF/10 mags ;  Total budget for chiller system: ~80 kCHF (to be spent ~2018, depending on effectiveness of new serigraphy and/or desired thermal margin).  Detailed breakdowns to be done.  Total budget for new ferrite type is ~450 kCHF (expensive and not necessary if optimum serigraphy length reduces power deposition to ~850 W, as predicted).  Total budget for 2 valves per tank is ~500 kCHF (expensive for gain !). 4/11/2014 LIU-SPS ABT review20

21. 20/11/2015 SPS LIU-ABT Systems Review: MKE21 Cost/Manpower: Chiller Chiller: Chiller (variable down to 6˚C) installed in tunnel, with a closed secondary cooling circuit; Disadvantage Condensation issues on magnet feedthroughs to consider (at low water temperatures); Advantages Magnets do not need to be rebuilt. Cost: approximate estimate of 40 kCHF per each of MKE4 and MKE6, including stainless- steel piping; Manpower: 6 person months for design and installation (parallel cooling circuits per extraction, where to put chiller, ….?) If optimized serigraphy is used, a chiller gives additional operating margin for ferrite heating (especially important if σ < 20 cm).