Definition of the interfaces cavity/cryomodule (sub-task of "SPL Cavities") V.Parma, CERN/TE-MSC EuCARD-SRF Annual Review 2010, Cockroft institute 7-9 April 2010 With contributions from: SPL WG3 members, and in particular P.Coelho Moreira (EuCARD funded CERN fellow), and Th.Renaglia, O.Capatina, U.Wagner (CERN)

Outline SPL architecture and parameters Layouts, integration studies and machine sectorisation (vacuum+cryogenics) Cavity interfaces – Mechanical – Cryogenic Summary

The SPL study at CERN R&D study for a 5 GeV and multi MW high power beam, the HP-SPL Major interest for non-LHC physics: ISOLDEII/EURISOL/Fixed Target/Neutrino Factory Length: ~500 m Ejection to Eurisol High  cryomodules 12 x 8  =1 cavities Medium  cryomodule High  cryomodules Ejection 10 x 6  =0.65 cavities 5 x 8  =1 cavities 6 x 8  =1 cavities TT6 to ISOLDE Debunchers To PS2 High  cryomodules From Linac4 0 m0.16 GeV110 m0.73 GeV 186 m1.4 GeV~300 m2.5 GeV Option 1Option 2 Energy (GeV)2.5 or 52.5 and 5 Beam power (MW)3 MW (2.5 GeV) or 6 MW (5 GeV) 4 MW (2.5 GeV) and 4 MW (5 GeV) Rep. frequency (Hz)50 Protons/pulse (x )1.52 (2.5 GeV) + 1 (5 GeV) Av. Pulse current (mA)2040 Pulse duration (ms) (2.5 GeV) (5 GeV) HP-SPL beam characteristics ~500 m5 GeV

Planned supplies for SPL prototype cryo-module InstituteResponsible person Description of contribution CEA – Saclay (F)S. Chel 1.Design & construction of 2 β=1 cavities (EuCARD task ) 2.Design & construction of 2 helium vessels for β=1 cavities (French in-kind contribution) 3.Design & construction of cryostat assembly tools (French in- kind contribution) 4.Supply of 8 tuners (French in-kind contribution) CNRS - IPN – Orsay (F)P. Duthil 1.Design and construction of prototype cryomodule cryostat (French in-kind contribution) Stony Brook/BNL/AES team I.Ben-zvi (Under DOE grant) 1.Designing, building and testing of 1 β=1 SPL cavity. CERNO.Capatina 1.8 β=1 cavities 2.7 He tanks in titanium 3.1 He tank in st.steel CERNE.Montesinos 1.10 RF couplers (testing in CEA, French in-kind contribution)

Courtesy of F.Gerigk, CERN SPL parameters

Courtesy of F.Gerigk, CERN New design parameters for HP-SPL cryogenics ( preliminary)

Cavity/cryomodule interface: 3 relevant workshops Workshop on Mechanical Issues SPL cavities/cryomodules – CERN, 30 Sept. 2009, – 15 participants from CEA, IPN Orsay, FNAL, Ruprecht Univ.Heidelberg, and CERN Workshop on cryogenic and vacuum sectorization of the SPL – CERN, 9-10 Nov.2009, – 22 participants from CEA, IPN Orsay, FNAL, ORNL, JLAB, SLAC, DESY, ESS, and CERN Review of SPL RF power couplers – CERN, March 2010, – 24 participants from CEA, IPN Orsay, BNL, FNAL, ANL, JLAB, DESY, Lancaster Univ., Uppsala Univ., and CERN

Continuous cryostat "Compact" version (gain on interconnections):  5 Gev version (HP): SPL length = 485 m Segmented layout (warm quads, with separate cryoline):  5 Gev version (HP): SPL length = 535 m Cryogenic and vacuum sectoriz.of the SPL Preferred for SPL

segmented cryostat String of (or single) cryo-modules Technical Service Module (TSM) Cold-Warm Transition (CWT) Cryostat is “segmented”: strings of (or single) cryo-modules, 2 CWT each Warm beam zones Cryogenic Distributio Line (CDL) needed Individual insulation vacuum on every string of cryo-module (Vacuum Barriers, w.r.t. CDL) Cryogenic Distribution Line (CDL) Insulation vacuum barrier Warm beam vacuum gate valve

Mechanical Issues SPL cavities/cryomodules (Workshop CERN 30 Sep. 2009) Goals: Share experience amongst labs (CEA, IPN Orsay, FNAL, and CERN) Select materials for the SPL helium vessels and ancillary components: – Helium vessel (Ti vs. St.steel), flanges, bellows, magnetic shielding – Interface to cryogenic piping (flanged? Bi-metallic transitions?) Design considerations for the helium vessel: – Tunability, mechanical stiffness – Manufacturing and assembly of cavity in helium vessel – Magnetic shielding Rationale for choices: – Technical requirements – Availability of technology (or needed R&D) – Material compliance with safety regulations (pressure vessels) – Cost (base materials, manufacturing, assembly, ancillaries…)

11 General: – The SPL-study, as far as the cavities and cryo-modules are concerned, should distinguish between the demonstrator (or prototype) and the project. – Priority is given to the development of cavities, so the demonstrator, scheduled in a relatively short time from now should be built using at maximum the existing experience, recipes and technical solutions, i.e. the CEA-Saclay/XFEL design as baseline, unless they turn out to be not adaptable to, or not needed for the SPL project. Layout issues – Outer conductor of coaxial power coupler in contact with lHe (for heat removal & stability reasons) – No dedicated ports for antenna like HOM coupler including notch filter (→ to be confirmed by simulations); but a pickup antenna port is required that may be used in addition as beam position monitor and HOM damper, if → damping of HOM in bellows turns out as insufficient – Position of pickup probes still to be defined (now defined) – Position of power coupler to be defined (see later) – Tuner must follow the → differential contraction of cavity/He-tank (advantage Ti tank) Retained issues - 1

12 Materials issues: – Use Ti for the demonstrator (proven technology) – Interface to SS cryogenic piping requires Ti/SS transitions – SS vs. Ti for tank: → Comprehensive study required, including cost issues (SS as machine option?) SS avoids use of bi-metallic transitions (Ti/SS) Ti tank: alleviated stroke for tuner – Magnetic shielding with Cryoperm®, external to He vessel → to be refined for different permeabilities The second annealing after manufacture may be performed in vacuum (without H 2 residual gas) Should the magnetic shield be actively cooled by lHe (IPN experience)? 2 nd shield at room temperature needed? Design/regulations issues: Nb and Ti are non-conventional materials for pressure vessels, so special qualification and quality plan (i.e. cost) have to be followed; certification procedures for non-conventional materials exist under EN Design pressure for cavity & He tank. Minimize for the cavity (plastic def.limits) but increase for cryogenics (valve tightness, overpress. in case of accidental venting): Tentatively: 1.3 bar (300 K); 2 bar at low temperature Retained issues - 2

Cavity/He vessel/coupler assy Vacuum vessel interface Double-walled-tube Flanged assembly to coupler (RF/leaktight Cu gasket) Ti He vessel Pick-up port

Cryogenic scheme (preliminary)

Cryogenic interface Double-walled-tube Diameter/positions still pending (depend on dynamic heat loads and coupler position )

Transversal position tolerance of cavities inside cryostat BUDGET OF TOLERANCE StepSub-stepTolerances (3σ)Total envelopes Cryo-module assembly Cavity and He vessel assembly± 0.1 mm (TBD) Positioning of the cavity w.r.t. beam axis ± 0.5 mm Supporting system assembly± 0.2 mm (TBD) Vacuum vessel construction± 0.2 mm (TBD) Transport and handling (± 0.5 g any direction) N.A.± 0.1 mm (TBD) Stability of the cavity w.r.t. beam axis ± 0.3 mm Testing/operation Vacuum pumping ± 0.2 mm (TBD) Cool-down RF tests Warm-up Thermal cycles Construction precision Long-term stability

Fixed support Sliding support Inertia beam Invar longitudinal positioner External supports (jacks) RF coupler (with bellows) Possible supporting schemes “standard” supporting scheme Two-support preferrable  isostatic (=well defined forces on supports)...but is cavity straightness enough?? If not...

Fixed support Sliding support Inertia beam Invar longitudinal positioner External supports (jacks) RF coupler (with bellows) Possible supporting schemes “standard” supporting scheme...add 3rd support  becomes hyperstatic (= forces depend on mech. coupling vessel/inertia beam)

“standard” supporting Max. sag minimized at a/L=0.20. To limit self-weight sag below 0.2 mm, requires an inertia beam equivalent to a 10-mm thick ~700-mm diam. tube (almost vac.vessel size)  A 3rd central support is mandatory

External supports (jacks) RF coupler + longitudinal positioner + vertical support Intercavity support structure Possible supporting schemes Alternative: coupler supporting scheme...the coupler is also a supporting/aligning element

Need inter-cavity support structure? l - If sag small enough - If strenght OK - isostatic - couple cavities - hyperstatic No Yes Equivalent sketch Layout inter-cavity guides

Static analysis (self-weight) Von Mises stress (Pa) Total displacement (m), amplification of 1100x Von Mises stress (Pa) Total displacement (m), amplification of 1100x RF coupler in cantilever (Courtesy of P. Coelho Moreira) Max.displacement: 0.04 mm Max.displacement: 11 mm RF coupler + intercavity guides Max.stress: 650 MPa Max.stress: 20 MPa

Coupler position: top or bottom...? Contras: Interferes with bi-phase tube  move sideways Waveguides/coupler more exposed to personnel/handling (damage, breaking window?)... Pros: Easier connection of waveguides Easier access (needed?)... Contras: Space needs for waveguides under cryostat If coupler not a support (bellows)  support on top, i.e. centered tube not possible... Pros: Centered bi-phase tube  symmetry Waveguides/coupler protected... There is no clear preference by the cavity/coupler experts, so the choice can be made based on cryomodule and machine integration needs  An exhaustive comparison still has to be made to make a final decision

RF coupler assembly constraints Note: drwgs for information only (concept not final) Single-window RF coupler  needs assembly to cavities in clean room Defines minimum diameter of “pipeline” type vessel: – Lenght of double-walled tube – Integration of thermal shield

Comparing solutions A) Coupler supporting scheme B) “standard” supporting scheme ProsContras Design simplicityVacuum vessel: - Stiffness (thickness, stiffeners) - Dim. stability - Precision machining - Cost Single cavity adjustment at assy Positioning stability (thermal, weld relieving...) Inter-cavity guiding Mid cavities guiding sufficient? ProsContras Cavities mechanical isolation from external perturbations: dim. changes (thermal, weld relieving), vibrations... Design complexity Vacuum vessel simplicity: -Reduced machining precision - reduced thickness Central support needed (?) Complex cavity adjustment at assy

Actively cooled RF coupler tube Vacuum vessel Heater Helium gas cooling the double wall 4.5 K 300 K No cooling T profile  21W to 2K Cooling (42 mg/sec) T profile  0.1 W to 2K SPL coupler double walled tube, active cooling to limit static heat loads Connected at one end to cavity at 2K, other end at RT (vessel) Requires elec. Heater to keep T > dew point (when RF power off) (Analysis: O.Capatina, Th.Renaglia) Massflow mgram/sec PowerONOFFONOFFONOFFONOFFONOFF Temp. gas out 286 K277 K283 K273 K271 K242 K255 K205 K232 K180 K Q thermal load to 2K 2.4 W0.1 W1.7 W0.1 W0.4 W0.1 W Q heater19 W32 W21 W34 W29 W38 W39 W41 W46 W44 W LL 0.1 mm ( )mm 0.05 mm ( ) ~ 0 mm ( )  Yields a certain degree of position uncertainty (<0.1 mm?)

Summary SPL cryomodule studies are concentrated on the 8 β=1 cavity cryomodule for a High Power version (with high dynamic heat loads) The cryomodule is a “standalone” unit connected to a cryogenic distribution line The demonstrator cryomodule shall make use well settled technology for the cavities/helium vessel/tuner (in particular Ti tank) The coupler is a single-window design solution derived from the CERN/LHC cryomodules  assembly constraints Coupler could be used as cavity support and is an interesting option Vacuum vessel become of paramount importance for mechanical positioning/stability Guiding between cavities probably needed (relieve cantilever) In case of fall back to a “standard” supporting solution, the mechanical interface to the cavity needs to be reassessed Still to be settled (next steps): Decide on position of coupler (top or bottom) Finalize cryogenic piping interface (depends on dynamic heat loads and coupler position) Review of cavity supporting system and definition of mechanical interface Fix alignment referentials on cavity and helium vessel

Thank you for your attention! Questions??

Spare slides

SPS PS2 SPL Linac4 PS ISOLDE Layout injector complex

The SPL study Goals of the study ( ): Prepare a Conceptual Design Report with costing to present to CERN’s management for approval This is the low power SPL (LP-SPL) optimized for PS2 and LHC Length: ~430 m Medium  cryomodule High  cryomodules Ejection 10 x 6  =0.65 cavities 5 x 8  =1 cavities 13 x 8  =1 cavities TT6 to ISOLDE Debunchers To PS2 High  cryomodules From Linac4 0 m0.16 GeV110 m0.73 GeV 186 m1.4 GeV427 m4 GeV Kinetic energy (GeV)4 Beam power at 4 GeV (MW)0.16 Rep. period (s)0.6 Protons/pulse (x )1.5 Average pulse current (mA)20 Pulse duration (ms)1.2 LP-SPL beam characteristics

β =0.65 cryo-module in SPL layout

β =1 cryo-module in SPL layout

Coupler integration functionalities (non-exhaustive list) FunctionalityRequirementComment RT/atm. to 2K/vacuum penetrations -minimise static HL; - leak-tight penetration (o-ring); - mechanical decoupling from vacuum vessel (if not a support); - Optimise thermal design Isolation between coupler and beam vacua - leak tight window - vacuum gauge outside cryostat - Single ceramic window for HP RF Gas cooling of double-walled tube (5 K-300 K) - minimise static HL; - cryostat feedthrough of cooling line: - leak-tight - avoid atm. moisture (T>dew point) - externally-mounted manual valves (presetting); - Heater for control (Ofelia’s talk) Coupler maintenanceMaintenance-freeNo in-situ intervention Coupler as support (option under study) -support/position cavities; - coupler flange mechanically fixed to vacuum vessel; - fix longitudinal position of cavities; - mechanical support of thermal shield - impact of gas cooling on position (Ofelia’s talk); - Vacuum vessel as mechanical reference

Possible cryogenic feeding: Cryogenic Distribution Line Cryogenic Interconnect Box Cryo-module Units Cryogenic Distribution Line Cryo-plant Isolde extraction Eurisol extraction CIB 10 β= β=1 cryo-modules 17 β=1 cryo-modules + 2 demodulators 6 β=1 cryo-modules 1.7% tunnel slope

Cavities cooled with helium II at ~ 2.0 K (3.1 kPa), similar to XFEL – Cavities inside saturated bath (slightly sub-cooled due to hydrostatic head) – Profit from low Δp and good p-stability Cryogenics XFEL Effect of slope

Cryogenic issues (ongoing) High dynamic heat load: ~200 W/cryo-module (~25 W/m)! – Thermal break-down to be checked (q<q film boiling ?) – Keep (efficient) stratified flow in bi-phase pipe  R&D & testing High dynamic heat loads requires large design capacity (=diameter) of helium Gas Return Pipe (GRP): – Limit number of cryomodules in series – Large GRP: better in a cryo distribution line Pipe sizing: – Bi-phase pipe: – HL dependant (vapour mass flow). 200W  ~ 14 g/s per module, for v=1m/s  Φ 140 – GRP: – HL dependant (vapour mass flow). 200W – can be kept relatively small with a CDL – Coupler gas cooling – Thermal shield pipes

Mass loads on internal supports (preliminary) DescriptionLoadComment Cavity/he vessel/tuner~2.0 kNper assembly (x 8 per cryomodule) Piping/thermal shield/MLI0.8 kNequally shared on coupler tube Transport accelerations± 0.5 geach direction