Pros and Cons of the 200 MHz System G. Iadarola, K. Li, E. Metral, G. Rumolo Thanks to: L. Medina, J. Esteban Mueller, B. Salvant, E. Shaposhnikova, R. Tomas Joint LARP CM26/Hi-Lumi Meeting at SLAC

Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/20162/35

Context & Outline Context: Operating the LHC with an additional 200 MHz RF system has been proposed in the past as option providing some advantages compared to the present HL-LHC baseline. The option also comes with some disadvantages that need to be carefully evaluated. In this talk we will focus on: Potential reduction of electron cloud activity as one of the advantages Lowering of the TMCI thresholds as one of the most critical performance limitations Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 Outline: 1.Motivation 2.Benefits – e-cloud 3.Limitations – TMCI, thresholds for 400 vs. 200 MHz 4.Open questions and future plans 3/37

Some examples – not all covered here Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 Advantages: Potentially acceleration of higher intensity and longer bunches with improved capture efficiency Minimization of electron cloud effects, intra- beam scattering and heating The possibility of luminosity leveling with bunch length See e.g. R. Calaga: “A 200 MHz SC-RF System for the HL-LHC”, Proceedings of the IPAC 2016 Drawbacks: Co-existence with 400 MHz crab-cavities, in particular, potential DA restrictions – this is being studied, see e.g.: 4/37

Some examples – not all covered here Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 Advantages: Potentially acceleration of higher intensity and longer bunches with improved capture efficiency Minimization of electron cloud effects, intra- beam scattering and heating The possibility of luminosity leveling with bunch length See e.g. R. Calaga: “A 200 MHz SC-RF System for the HL-LHC”, Proceedings of the IPAC 2016 Drawbacks: Co-existence with 400 MHz crab-cavities, in particular, potential DA restrictions – this is being studied, see e.g.: K. Ohmi, “Beambeam effect for collision with Large Piwinski angle scheme and high frequency crab cavity in LHC”: R. Tomas, “HL-LHC alternative scenarios, parameters and lay-out”: /attachments/605732/833603/SLIDES.pdf /attachments/605732/833603/SLIDES.pdf Y. Luo, title: "Beam-beam simulation with crab cavities for ERHIC“: /dateposted-public/ /dateposted-public/ 5/37

Some examples – not all covered here Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 Advantages: Potentially acceleration of higher intensity and longer bunches with improved capture efficiency Minimization of electron cloud effects, intra- beam scattering and heating The possibility of luminosity leveling with bunch length See e.g. R. Calaga: “A 200 MHz SC-RF System for the HL-LHC”, Proceedings of the IPAC 2016 Drawbacks: Co-existence with 400 MHz crab-cavities, in particular, potential DA restrictions – this is being studied, see e.g.: K. Ohmi, “Beambeam effect for collision with Large Piwinski angle scheme and high frequency crab cavity in LHC”: R. Tomas, “HL-LHC alternative scenarios, parameters and lay-out”: /attachments/605732/833603/SLIDES.pdf /attachments/605732/833603/SLIDES.pdf Y. Luo, title: "Beam-beam simulation with crab cavities for ERHIC“: /dateposted-public/ /dateposted-public/ Pile-up distribution and machine protection for different crab cavity failure scenarios Longitudinal stability, for or reference, see: E. Shaposhnikova et al., “Possible Beam Parameters in Double RF Operation of the CERN LHC”, Proceedings of the IPAC /35

Part 1 Potential benefits for e-cloud

e-cloud dependence on bunch length – arc dipoles Heat load decreasing with bunch length for all values of SEY Significant change of the multipacting threshold Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/20168/37

e-cloud dependence on bunch length – arc quadrupoles Heat load decreasing with bunch length only for large values of SEY Multipacting threshold decreasing for longer bunches Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/20169/35

e-cloud dependence on bunch length – half cell In the region where heat loads from e-cloud are worrying, longer bunches are beneficial Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201610/37

e-cloud dependence on bunch intensity Heat load scales non-monotonically with bunch intensity  Start of fill might not be the most critical moment Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201611/37

Estimation approach Evolution of the heat load during the fill is the result of several factors: Heat load sharing between dipoles and quadrupoles Dependence on bunch intensity (start of fill is not necessarily the critical point) Dependence on bunch length Contribution of synchrotron radiation and impedance Built python module to estimate evolution of the heat load during the fill: Simple formulas (same as for online displays) for impedance and synchrotron radiation Interpolation over a database of build-up simulations of the e-cloud contributions. Simulations available for: Dipoles and quadrupoles Intensity in [0.1, …, 2.5] x 1011 p/bunch Bunch length in [0.5, …, 3.5] ns SEY in [1.0, …, 2.0] Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201612/37

Considered two possible situations with respect to e-cloud e-cloud suppression in the dipoles is achieved: SEYdip = SEYquad = 1.30 e-cloud suppression in the dipoles is not achieved: SEYdip = SEYquad = 1.40 For each case we simulated a fill for the HL-LHC baseline (400 MHz) and two 200 MHz cases (fast recapture, long bunches kept along the fill) Considered scenarios Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201613/37

Scenario 1: e-cloud suppression achieved in dipoles SEY dip = SEY quad = 1.30 Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201614/37

e-cloud quad. e-cloud dip. Impedance Syn. radiation HL-LHC baseline (400 MHz) Assumption: SEY dip = SEY quad = 1.30 Bunch length is kept constant at 8 cm Synchrotron radiation and impedance take ~75 % of the available cooling capacity at the beginning of the fill No e-cloud in dipoles all along the fill e-cloud in the quadrupoles appears with the decrease in intensity Scenario 1 – e-cloud suppression achieved in dipoles Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201615/37

e-cloud quad. e-cloud dip. Impedance Syn. radiation 200 MHz option, fast recapture Assumption: SEY dip = SEY quad = 1.30 The picture does not change significantly No e-cloud in dipoles all along the fill Bunches are already short when the “critical” intensity is achieved Impedance contribution is reduced at the beginning of the fill Scenario 1 – e-cloud suppression achieved in dipoles Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201616/37

e-cloud quad. e-cloud dip. Impedance Syn. radiation 200 MHz option, constant bunch length Assumption: SEY dip = SEY quad = 1.30 The picture does not change significantly Scenario 1 – e-cloud suppression achieved in dipoles Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201617/37

Scenario 2: e-cloud suppression not achieved in dipoles SEY dip = SEY quad = 1.40 Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201618/37

e-cloud quad. e-cloud dip. Impedance Syn. radiation 256 fb -1 /year Scenario 2 – e-cloud suppression not achieved in dipoles Assumption: SEY dip = SEY quad = 1.40 e-cloud in dipoles all along the fill Heat load exceeds the available cooling capacity Maximum heat load reached later during the fill HL-LHC baseline (400 MHz) Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201619/37

e-cloud quad. e-cloud dip. Impedance Syn. radiation 256 fb -1 /year Assumption: SEY dip = SEY quad = 1.40 e-cloud in dipoles develops when bunches are shortened Heat load still exceeds the available cooling capacity Scenario 2 – e-cloud suppression not achieved in dipoles 200 MHz option, fast recapture Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201620/37

e-cloud quad. e-cloud dip. Impedance Syn. radiation Cryo limit 236 fb -1 /year (-8% w.r.t. 400 MHz) Assumption: SEY dip = SEY quad = 1.40 e-cloud in dipoles is suppressed Heat load within the cryogenics limit Margin to shorten the bunches towards the end of the fill to optimize performance 200 MHz option, constant bunch length Scenario 2 – e-cloud suppression not achieved in dipoles Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201621/37

e-cloud quad. e-cloud dip. Impedance Syn. radiation Cryo limit 243 fb -1 /year (-5% w.r.t. 400 MHz) 200 MHz option, “natural” shortening Assumption: SEY dip = SEY quad = 1.40 e-cloud in dipoles is suppressed Heat load within the cryogenics limit Margin to shorten the bunches towards the end of the fill to optimize performance Scenario 2 – e-cloud suppression not achieved in dipoles Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201622/37

A simple tool has been developed to evaluated the evolution of the heat loads on the arc beam screen during a physics fill The baseline HL-LHC scenario has been compared with different 200 MHz operation modes 200 MHz becomes very interesting in the case where the e-cloud in the dipole magnets is not suppressed (SEY>=1.4) o In this case heat load exceeds the available capacity for the 400 MHz scenario o The increase in bunch length given by the 200 MHz RF, is sufficient to suppress the e-cloud in the dipoles and bring the heat load within the available capacity o Long bunches have to be kept at least for the first part of the fill o Bunch length evolution can be optimized to maximize luminosity within the allowed heat load (5% loss compared to more than 20% loss for 8b+4e)  needs setup o After long shutdowns, 200 MHz operation would mitigate the luminosity loss due the SEY recovery Summary and conclusions – part 1 Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201623/37

Part 2 Potential threats from TMCI

Parameters Some changes occurred compared to the past Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 S. White - 6th LHC crab cavity workshop Among the largest impact originates from the slightly lower transition gamma compared to LHC:  Increase in Qs – good (TMCI) Increase in bunch length – good (TMCI) 25/37

Parameters Some changes occurred compared to the past Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 Taken from: HL-LHC OPERATIONAL SCENARIOS (CERN-ACC-NOTE ) RF parameters from E. Shaposhnikova S. White - 6th LHC crab cavity workshop including double harmonic option i.e., 400 MHz (base) MHz (harmonic) vs. 200MHz (base) MHz (harmonic) Among the largest impact originates from the slightly lower transition gamma compared to LHC:  Increase in Qs – good (TMCI) Increase in bunch length – good (TMCI) 26/37

Earlier findings Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 S. White - 6th LHC crab cavity workshop 27/37

Earlier findings Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 S. White - 6th LHC crab cavity workshop From a pure and naive scaling we can assume an increase of the TMCI threshold by:  an increase by roughly 15% (~4.4e11 ppb) for the 400 MHz and roughly 25% (~3.3e11 ppb) for the 200 MHz options Additional changes: Refined impedance model: Allthemachine_HL-LHC_15cm_7TeV_5umMo+MoC_IP7_TCT5_B1 Modeling of non-linear synchrotron motion to take into account synchrotron tune spread 28/37

200 MHz vs 400 MHz TMCI Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/ MHz400 MHz 29/37

Bunch shortening mode for extrapolation Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/ MHz400 MHz 30/37

Thresholds summarized Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 Both the effect of Qs and the bunch length are included, here. 31/37

Thresholds normalized Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 Nonlinear synchrotron motion makes scaling and extrapolation to bunch lengths tricky… 32/37

Thresholds vs. Qs and bunch length Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201633/37

Thresholds vs. Qs and bunch length Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201634/37

Thresholds vs. Qs and bunch length Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 E.g.: bunch shortens and voltage constant, i.e. retaining low Qs Nevertheless, stabilization in the transverse plane expected from the head-on collisions similar to S. White, X. Buffat, N. Mounet, T. Pieloni, “Transverse mode coupling instability of colliding beams”: TAB TAB MHz option, “natural” shortening 35/37

Thresholds vs. Qs and bunch length Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/2016 E.g.: bunch shortens and voltage constant, i.e. retaining low Qs Will anyway be dominated rather by potential longitudinal stability issues, see: E. Shaposhnikova, “Possible Beam Parameters in Double RF Operation of the CERN LHC”, Proceedings of the IPAC 2016” 200 MHz option, longitudinal stability 36/37

Cross-check thresholds vs. bunch length Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/ e cm15 cm Set of simulations for some of the scenarios yields comparable results (scaling for 400 MHz was slightly overestimated) 37/37

Cross-check thresholds vs. bunch length Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/ e cm15 cm Set of simulations for some of the scenarios yields comparable results (scaling for 400 MHz was slightly overestimated) 38/37 HL-LHC 400 MHz baseline 200 MHz constant bunch length 200 MHz „natural“ shortening 200 MHz fast recapture

Summary and conclusions We computed the expected TMCI thresholds for 400 MHz and 200 MHz scenarios with updated parameters. We extrapolated the expected thresholds to estimate the impact of different bunch-length-“leveling” scenarios. We cross checked the extrapolation with simulation of reference cases. 200 MHz TMCI threshold, for current parameters at roughly 4e11 ppb (flat-top before collision). Still, 200 MHz gives lots of additional flexibility. With the previous limits becoming less critical this option may have considerable benefits. Joint LARP CM26/Hi-Lumi Meeting at SLAC - 05/20/201639/37