Tom Peterson -- Cryomodules 650 MHz cryomodule design status Tom Peterson and Yuriy Orlov 10 Jan 2011

Tom Peterson -- Cryomodules Approach CW cryomodules with as much as 25 W per cavity at 2 K and tight constraints on cavity frequency present some different problems from TESLA/ILC cryomodules Let’s look at the requirements, consider what other labs have already done, and select best features for our own design

Tom Peterson -- Cryomodules Design team 650 MHz cryomodules –Camille Ginsburg, Yuriy Orlov, and Prashant Khare are leading and organizing the effort with me Cavities, input couplers, magnets, magnet current leads, tuners, instrumentation, 325 MHz cryomodules, microphonics, etc. –Many other people –First conclusion from my visits to other labs -- we have a strong and diverse SRF team

Tom Peterson -- Cryomodules Evolving requirements Writing a “functional specification” is in progress –Actually, listing requirements is a better description But we do not yet have a functional specification –We know many requirements –Others are still evolving For example, MAWP is a fundamental parameter. Pipe sizes were initially analyzed for a 2.5 bar cold MAWP. 4.0 bar now looks achievable, so these analyses are being revisited. Similarly, pressure stability requirements depend, among other things, on cavity df/dP, which is improving with new cavity designs

Tom Peterson -- Cryomodules Draft of a near-term task list - 1 RRCAT, Oct 2010, revised 650 CM requirements and functional specification (Fermilab) –Continue to collect list of features and requirements –Draft functional specification Refine and update the 650 MHz cryomodule/cryogenic flow schematic (Fermilab) –Illustrate the case of single stand-alone 650 MHz cryomodules –Perhaps include other features like 4.5 K phase separation Consider a BESSY-type scheme with an open-ended cryomodule, cold magnets, but single cryomodule liquid management length, and the option of variable string lengths. –Other ongoing refinements and considerations

Tom Peterson -- Cryomodules Draft of a near-term task list - 2 Flow rate analyses and pipe sizing (Fermilab and RRCAT) –Agree on a maximum credible air inlet port size –Review upset condition venting pipe sizing requirements –Review steady-state pipe sizing requirements Continue review of cryomodule operational experience at other labs, especially with respect to CW operation and pressure stability (Fermilab and RRCAT) –Correlate this experience with design features –Visit CW cryomodule labs: Cornell, JLab, BESSY (Berlin) –I was at Cornell in Nov for ERL review. Jlab Dec.) –Berlin/DESY visit week of 21 Feb., before TTC in Milan including RRCAT colleagues (Avinash and Prashant will meet us)

Tom Peterson -- Cryomodules Draft of a near-term task list - 3 Refine pressure change buffering considerations (Fermilab and RRCAT) –Understand the rate of return to pressure-temperature equilibrium following a vapor pressure change over a 2 K bath –Refine assumptions for heat pulse or pressure pulse drivers –Reconsider liquid and vapor volume advantages

Tom Peterson -- Cryomodules Draft of a near-term task list - 4 Begin a 3-D CAD model of a closed-ended 650 MHz TESLA-type cryomodule (RRCAT and Fermilab) –Cross section like presently conceived at RRCAT and Fermilab (Yuriy Orlov) –Bayonet connection design detail and heat exchanger size (not details) can be provided by Fermilab –Minimal end length including warm-cold transition. End concepts may be taken from SNS, from Fermilab 325 MHz cryomodule model, from end boxes, and from other sources –New thermal contraction constraints and fixed points –Cavity support scheme and position stability Reconcile results (slot lengths, drift spaces) with beam dynamics people (Fermilab) –Consider warm magnets between beta = 0.9 cryomodules

Tom Peterson -- Cryomodules Meetings every other week 650 MHz joint cavity/cryomodule meetings –Camille, Chuck, Yuriy O., Mike F., others –Avinash, Prashant, Jishnu, others –A details meeting Tuesday mornings, 6:30 AM at Fermilab, 6 PM at RRCAT –Last meeting was Tuesday, Jan 4 –Next meeting Tuesday, Jan 18

Tom Peterson -- Cryomodules Experience at other labs In the past year (often by invitation to reviews), other Fermilab people and I have visited –SNS –Cornell –Triumf –MSU –Jlab –RRCAT A visit to FZB in Berlin and to DESY is planned for late February All of these labs have taken somewhat different approaches to cryomodule design depending on their specific requirements and experience

Tom Peterson -- Cryomodules Some highlights SNS –Stand-alone cryomodules, warm magnets and warm beam instrumentation between them, “space frame” support structure (no large vapor return pipe), open bath helium vessel (no separate 2-phase pipe), vacuum vessel provides pressure containment compliance

Tom Peterson -- Cryomodules SNS vs TTF cryomodule SNS (like CEBAF): self-contained vacuum vessel TTF: vacuum vessel string. End boxes and bellows would become part of vacuum/pressure closure

Tom Peterson -- Cryomodules ERL injector cryomodule

Tom Peterson -- Cryomodules ERL cryomodule features Figure 1 from CRYOGENIC HEAT LOAD OF THE CORNELL ERL MAIN LINAC CRYOMODULE, by E. Chojnacki, E. Smith, R. Ehrlich, V. Veshcherevich and S. Chapman, Cornell University, Ithaca, NY, U.S.A. Published in Proceedings of SRF2009, Berlin, Germany

Tom Peterson -- Cryomodules ERL cryomodule features TESLA-style support structure -- dressed cavities hang from gas return pipe (GRP), but –Titanium GRP –No invar rod, no rollers –6 cavities per CM, 9.8 m total CM length –HOM absorbers at K between cavities –GRP split with bellows at center, 4 support posts –Helium vessels pinned to GRP –Some flexibility in the input coupler –De-magnetized carbon-steel shell for magnetic shielding (this is like TTF) –2-phase pipe closed at each CM end, JT valve on each CM (like BESSY design) –String rolls into vacuum vessel on rails

Tom Peterson -- Cryomodules ERL linac cavity features 1.3 GHz, 7 cell E = 16.2 MV/m Active length of 0.8 m CW operation Q0 = 2 x K ~10 W/cavity End lever (Saclay-type) tuner –Even though the injector CM has blade tuners

Tom Peterson -- Cryomodules

ISAC tunnel

Tom Peterson -- Cryomodules FRIB tunnel mockup

Tom Peterson -- Cryomodules Jlab space frame

Tom Peterson -- Cryomodules Visit to BESSY planned in February

Tom Peterson -- Cryomodules SSR1 CM concept

Tom Peterson -- Cryomodules General arrangements under consideration Segmentation level and cavity support structure –BESSY/HZB (and Cornell ERL) liquid managed separately for each CM, 2-phase pipe closed at each end, but otherwise a string, TESLA style piping and supports –Three options for segmentation at the individual cryomodule level Completely close a TESLA style CM at each end Eliminate 300 mm pipe -- space frame support Eliminate 300 mm pipe -- support posts and frame (325 MHz concept from Tom Nicol) Helium vessel –Closed, TESLA-style –Open, Jlab/SNS style

Tom Peterson -- Cryomodules 650 MHz cryo schematic

Tom Peterson -- Cryomodules 650 MHz Cryomodule (concept) 650MHz Cryomodule Concept, Y. Orlov Page 25 Heat exchanger Vacuum vessel relief valve 48” Vacuum Vessel Vacuum vessel adjusting Supports MC & MC Ports Vacuum Vessel End-Plates

Tom Peterson -- Cryomodules 650 MHz Cryomodule (concept) 650MHz Cryomodule Concept, Y. Orlov Page 26 Control valves Bayonet connection Heat exchanger Vent line with check valve

Tom Peterson -- Cryomodules 650 MHz Cryomodule layout (concept) 650MHz Cryomodule Concept, Y. Orlov Page 27 Vacuum vessel Cold mass 650 MHz Cavity assembly Support post ( 2 for each cavity) Heat Exchanger pipe 2- phase He pipe (Ti)

Tom Peterson -- Cryomodules 650 MHz Cryomodule sections (concept) 650MHz Cryomodule Concept, Y. Orlov Page 28 Heat exchanger Cryo feed snout 80K shielding Cold mass tray Tray supports

Tom Peterson -- Cryomodules 650 MHz Cavity string on the strong back (XFEL cavity style) 650MHz Cryomodule Concept, Y. Orlov Page Cavity with 2-phase pipe Warm up cool down pipe Strong back (Tray) with supports He reservoir with level sensor 4K circuit (to MC) 70K circuit (shield)

Tom Peterson -- Cryomodules 650 MHz Cavity string on the strong back (SNS cavity style) 650MHz Cryomodule Concept, Y. Orlov Page 30 U-turn connection to heat exchanger

Tom Peterson -- Cryomodules 650 MHz Cold Mass assembly 650MHz Cryomodule Concept, Y. Orlov Page 31 80K shield top Cold mass configuration, before inserting in to Vacuum vessel

Tom Peterson -- Cryomodules CM650 TESLA STYLE- STANDALONE Number of cavities-8 Cavity Style- Tesla Vessel pipe OD- 48 inch MC Port Cold mass supports: Fix Sliding Heat exchanger Beam pipe Cryomodole Endplate

Tom Peterson -- Cryomodules CM650 TESLA STYLE- STANDALONE. (Main Section) Heat exchanger 300 mm pipe 80K shield 650 cavity MC port Cryo Feed Vacuum vessel He 2-phase pipe

Tom Peterson -- Cryomodules CM650 TESLA STYLE- STANDALONE. COLD MASS. 1.80K shield (supply, return pipes) 2.4K pipes (supply, return) 3.4.4K phase separator and 2-phase He thermal intercept 4.Warm Up cool down pipe 5. He reservoir with level sensor

Tom Peterson -- Cryomodules CM650 TESLA STYLE- STANDALONE. CAVITY STRING & 300mm PIPE 300mm pipe (St. steel) He 2-Phase pipe (Ti) Temp. shrinkage Compensator He reservoir with level sensor Warm up cool down pipe

Tom Peterson -- Cryomodules CM650 TESLA STYLE. T4CM-Bessy Number of cavities-8 Cavity Style- Tesla Vessel pipe OD- 48 inch

Tom Peterson -- Cryomodules CM650 TESLA STYLE. T4CM-Bessy

Tom Peterson -- Cryomodules CM650 TESLA STYLE. T4CM-Bessy COLD MASS

Tom Peterson -- Cryomodules CM650 TESLA STYLE. T4CM- Bessy. CAVITY STRING & GAS RETURN PIPE

Tom Peterson -- Cryomodules

SCRF Cavity supported on HGR pipe Information required on  Magnet package  Tuner details  Power Coupler B. Glimpses of 3-D Model (contd…)  The model incorporates a modified Cavity support system.  2K helium supply line includes a bellow in vertical configuration

Tom Peterson -- Cryomodules Thermal Shield with dressed Cavity  80K- Thermal shield  5K-Thermal shield is partial.(Upper Part only).  Thermal shield 80K shield.  Thermal shield 5K shield is partial. B. Glimpses of 3-D Model (contd…)

Tom Peterson -- Cryomodules Design considerations Cooling arrangement for integration into cryo system Pipe sizes for steady-state and emergency venting Pressure stability factors –Liquid volume, vapor volume, liquid-vapor surface area as buffers for pressure change Evaporation or condensation rates with pressure change Updated heat load tabulations (using recent info from Slava Yakovlev) Options for handling 4.5 K (or perhaps 5 K - 8 K) thermal intercept flow Alignment and support stability Thermal contraction and fixed points with closed ends Etc.

Tom Peterson -- Cryomodules Our plan Continue analyses, modeling, and reviews of existing designs through Feb Following visits to HZB, DESY, and TTC meeting (Feb 21 - Mar 3), down-select a more specific design approach –Goal is to have a specific 650 MHz cryomodule design proposal for discussion before the Project X Collaboration meeting (April 11) –Also complete (draft) specifications and fundamental CM parameter lists in this timeframe

Tom Peterson -- Cryomodules Backup slides –Present thoughts –Heat loads –Pipe sizes –Current leads –Heat exchanger –Pressure stability

Tom Peterson -- Cryomodules My opinion Very high heat flux (200 W per CM) and relatively short linac (not large quantity production nor several km long strings) ==> stand- alone cryomodules Stand-alone CM ==> “300 mm” pipe is unnecessary for helium flow –Possibly even risky due to uneven cooling Not need 300 mm pipe ==> different support structure (space frame or posts) High heat loads and tight pressure stability ==> large liquid-vapor surface area for liquid-vapor equilibrium –Acts as thermal/pressure buffer with heat load and pressure changes Linac is short enough that total helium inventory not an issue ==> open helium vessel I come to an SNS-style or our 325 MHz style cryomodule

Tom Peterson -- Cryomodules However Helium vessel style (open vs. closed) is independent of support style (hung from 300 mm pipe or not) For the stand-alone CW cryomodule, a closed TESLA-type helium vessel may be favored by –Tuner design –Input coupler design –Reduced pressure sensitivity Then it looks like Jlab 12-GeV upgrade CM or again a version of our 325 MHz CM design

Tom Peterson -- Cryomodules Cryomodule Pipe Sizing Criteria Heat transport from cavity to 2-phase pipe –1 Watt/sq.cm. is a conservative rule for a vertical pipe (less heat flux with horizontal lengths) Two phase pipe size –5 meters/sec vapor “speed limit” over liquid –Not smaller than nozzle from helium vessel Gas return pipe (also serves as the support pipe in TESLA-style CM) –Pressure drop < 10% of total pressure in normal operation –Support structure considerations Loss of vacuum venting P < cold MAWP at cavity –Path includes nozzle from helium vessel, 2-phase pipe, may include gas return pipe, and any external vent lines 48

Tom Peterson -- Cryomodules Stand-alone CM pipe sizes 49

PX Briefing to OHEP 50 Reference design: cw linac Prelim design (preferred) SSR0SSR1SSR2β=0.6β= MHz MeV 650 MHz GeV SectionFreqEnergy (MeV)Cav/mag/CMType SSR0 (  G =0.11) /26/1SSR, solenoid SSR1 (  G =0.22) /18/ 2SSR, solenoid SSR2 (  G =0.4) /24/ 4SSR, solenoid LB 650 (  G =0.61) /24/ 45-cell elliptical, doublet HB 650 (  G =0.9) /34/185-cell elliptical, doublet 180 elliptical cavities From Project X Briefing to DOE/Office of High Energy Physics November 16, 2010, by Sergei Nagaitsev Consider stand-alone cryomodules -- Tom P.

Tom Peterson -- Cryomodules Air inflow heat flux limit Atmospheric air flowing into a vacuum via a round hole –~23 grams/sec per cm2 hole size Heat deposition by air condensing on cold surface –~470 J/g, so 10.8 kW per cm2 hole size Helium heat input per gram ejected for typical ( bar) pressures –~13 J/g Helium mass flow per air inlet area –~830 grams/sec helium per cm2 hole size 51

Tom Peterson -- CryomodulesProject X CW Cryomodules, Tom Peterson For a 3-inch (76 mm) diameter opening, air flow becomes the limiting factor in heat deposition after a few cryomodules. But for one cryomodule, total cavity surface area determines the flow rate inch port

Tom Peterson -- Cryomodules Current lead design Magnets in sub- atmospheric 2 K helium require conductively cooled current leads. 4 leads at 50 A, 2 leads at 200 A Thermal intercepts at 80 K and 5 K Tom Nicol, Valeri Poloubotko, Sergey Cheban

Tom Peterson -- Cryomodules Heat exchanger design 4.5 K - 2 K heat exchanger for 10 grams/sec (~200 W at 2 K, designed by Tom Peterson) in the TTF feedbox assembly 2 K supply flow precooled by 2 K pumped return flow improves 2 K capacity by ~40%. Heat exchanger must be vertically oriented and higher than the liquid in the cryomodule Arkadiy Klebaner, Sergey Cheban, Ben Hansen working on heat exchanger design for Project X

Tom Peterson -- Cryomodules 650 MHz assumptions 2.0 K helium cooling (~30 mbar) 1.5 uncertainty factor for flow estimation and pipe sizing Cavity MAWP = 2.0 bar warm, 4.0 bar cold Heat loads on following slides – From "CAVITY.dat" from Slava Yakovlev (28 Dec 2010) 55

Tom Peterson -- Cryomodules 2 K heat loads 56

Tom Peterson -- Cryomodules 5 K heat loads 57

Tom Peterson -- Cryomodules 70 K heat loads 58

Tom Peterson -- Cryomodules Cornell proposes an Energy Recirculating Linac (ERL) CLASSE = Cornell Laboratory for Accelerator-based Sciences and Education CHESS = Cornell High Energy Synchrotron Source

Tom Peterson -- Cryomodules HOM absorbers 7 absorbers per CM, between the 6 cavities and at each end 200 W per absorber Take heat at ~100 K Mechanical and thermal design still not resolved and difficult

Tom Peterson -- CryomodulesProject X CW Cryomodules, Tom Peterson

Tom Peterson -- Cryomodules Further considerations Support structure –Stiffness of pipe if used as support backbone –Or other support structure options Emergency venting scenarios drive pipe sizes and influence segmentation –Cold MAWP may be low for 650 MHz, driving up pipe sizes and/or reducing spacing between relief vent ports –Trade-off of pipe size with vent spacing requires further work –Thermal shield pipe may also require frequent venting 5 K has a large surface area for large heat flux 70 K starts at a high pressure Liquid management length needs further work –May want to subdivide strings for liquid management due to large specific liquid flow rate per cavity 62

Tom Peterson -- Cryomodules Liquid management length The 2 K to 4.5 K heat exchanger needs to be divided (not one large heat exchanger) in order to be a practical size, which means distributing multiple heat exchangers in the tunnel. 2 K to 4.5 K heat exchanger size which fits in the Project X tunnel will be roughly grams/sec (about Watts of 2 K heat) With 650 MHz and 1.3 GHz CW heat loads of nearly 200 W per cryomodule, this implies liquid management lengths of one or two cryomodules as limited by JT heat exchanger practical size limits 63

Tom Peterson -- Cryomodules From “Controlling Cavity Detuning for Project X” by Ruben Carcagno, et. al. SNS –Long term pressure regulation to better than 100 ubar (Fabio Casagrande) Regulates flow and maintains constant heat load using bath heaters Distributes He at 4K with a cold box on each cryomodule CEBAF –25 ubar when everything is ‘quiet’ Also regulates flow and uses heaters to maintain constant load Distributes He at 2K –Transients of up to several 1-2 mbar about once a day Identified a control system problem not a cryogenic problem HoBiCaT (BESSY) –15-30 ubar steady state –Large non-gaussian tails in microphonics distribution ~17  fluctuation every one or two hours Origin not completely understood but cryogenic likely a large component Cornell –Heat leak at dead end pipe in cryo system creates gas bubbles which ‘pop’ about once per second

Tom Peterson -- Cryomodules65 Pressure stability Consider short duration pulses (~ a few seconds or less) for which the vapor and liquid are not in equilibrium. –Treat the vapor volume as a closed volume of ideal gas –Heat or vapor flow change results in a direct change of stored gas –Pressure changes in proportion to net mass flow in or out –Pressure changes inversely with total vapor volume –Using total mass flow, if a TTF cryomodule with TTF-sized pipes and 16 W total heat (dynamic plus static), pressure change due to total heat flow into a closed volume is dP/dt =0.023 mbar/sec SSR0 dP/dt = 0.10 mbar/sec SSR1 dP/dt= 0.14 mbar/sec SSR2 dP/dt = 0.33 mbar/sec 650 MHz CW dP/dt = mbar/sec

Tom Peterson -- Cryomodules66 Pressure stability Consider pressure pulses for which the 2 K liquid and vapor are always in complete thermal equilibrium. –Liquid volume becomes a factor (liquid helps to buffer a step change in heat load due to a significant heat capacity with small temperature change) J/g per mK, so J/g for 0.1 mbar pressure change –The process is the addition of heat or transfer of mass (to or from) a closed 2- phase system –Am looking into the equilibrium (liquid-vapor phase equilibrium) rate of change of pressure with incremental change of heat and flow Various ratios of liquid and vapor Various rates of heat addition Rate of return to vapor-liquid equilibrium is limited by heat transport through the “chimney” as helium condenses (adding heat) or vaporizes (carrying away heat) –Surface area of liquid is a factor to help damp pressure oscillations

Tom Peterson -- Cryomodules67 Pressure stability -- role of liquid surface Consider a 0.1 mbar step change in pressure due to vapor flow –Corresponding temperature change is 1 mK –So surface temperature of the liquid essentially immediately changes by 0.1 mK and heat transport proceeds as vapor evaporates or condenses. –The nominal maximum of 1.4 W/cm 2 corresponding to liquid head reaches to a depth of 0.7 cm for 0.1 mbar. For greater depth, heat transport with 1 mK delta-T is reduced in proportion to length. For a 5 cm chimney height, this becomes 0.2 W/cm 2 for the 1 mbar pressure change. –Corresponding surface mass flux (23.4 J/g latent heat) is g/s per cm 2. –The TESLA 55 mm chimney (24 cm 2 ) can then buffer a vapor flow rate change of up to 24 x = 0.2 gr/sec with a 0.1 mbar pressure change. For the closed 650 MHz dressed cavity, a 110 mm chimney can buffer a 0.1 mbar pressure change with up to 0.8 gr/sec vapor-liquid exchange rate. For an open 650 MHz dressed cavity, the maximum vapor exchange rate with 0.1 mbar delta-P is about 47 gr/sec. Conclusion: there is a significant pressure stability advantage for the large 2.0 K liquid-vapor surface area.

Tom Peterson -- Cryomodules Conclusions Work on cryomodule requirements and configurations is in progress. CW cryomodule piping requirements may be quite different from TESLA/ILC –325 MHz cryomodule 2-phase pipe sizes are overwhelmingly determined by venting requirements –650 MHz and 1.3 GHz CW liquid management lengths may be only one or a few cryomodules Thermal shield piping sizes may be like TTF or smaller Division of 325 MHz cryomodules into shorter sections for emergency venting looks necessary Division of CW 650 MHz and 1.3 GHz cryomodules into short strings or individual cryomodules for liquid management may be necessary A large liquid-vapor surface area is an advantage for pressure stability – This may be an important consideration if we use SNS and Jlab pressure stability experience to set our expectations for Project X 68

Tom Peterson -- Cryomodules References “Project-X, CW Linac (ICD-2+) Lattice Design,” by Nikolay Solyak (presentation to a Project X meeting, March 16, 2010) “Notes about the Limits of Heat Transport from a TESLA Helium Vessel with a Nearly Closed Saturated Bath of Helium II,” by Tom Peterson, TESLA report #94-18 (June, 1994). "Latest Developments on He II Co-current Two-phase Flow Studies," by B. Rousset, A. Gauthier, L. Grimaud, and R. van Weelderen, in Advances in Cryogenic Engineering, Vol 43B (1997 Cryogenic Engineering Conference). "Optical Investigations of He II Two Phase Flow," by E. di Muoio, et. al., in Advances in Cryogenic Engineering, Vol 47B (2001 Cryogenic Engineering Conference). Dynamic RF heat loads from "CAVITY.dat”, Slava Yakovlev (28 Dec 2010) Current lead heat from "Conduction-cooled 60 A Resistive Current Leads for LHC Dipole Correctors", by A. Ballarino and other studies 69