Performance in the vertical test of the 832 nine-cell 1.3 ... · divided between Ettore Zanon S.p.A., and RI Research Instruments GmbH. Production included both mechanical fabrication

Performance in the vertical test of the 832 nine-cell 13 GHz cavitiesfor the European X-ray Free Electron Laser

D Reschke V Gubarev J Schaffran L Steder N Walker and M WenskatDeutsches Elektronen-Synchrotron DESY Notkestrasse 85 22607 Hamburg Germany

L MonacoINFN MilanondashLASA via Fratelli Cervi 201 20090 Segrate Milano Italy

(Received 2 December 2016 published 24 April 2017)

The successful production and associated vertical testing of over 800 superconducting 13 GHzaccelerating cavities for the European X-ray Free Electron Laser (XFEL) represents the culmination ofover 20 years of superconducting radio-frequency RampD The cavity production took place at two industrialvendors under the shared responsibility of INFNMilanondashLASA and DESY Average vertical testing rates atDESYexceeded 10 cavities per week peaking at up to 15 cavities per week The cavities sent for cryomoduleassembly at Commissariat agrave lrsquoeacutenergie atomique (CEA) Saclay achieved an average maximum gradient ofapproximately 33 MV=m reducing tosim30 MV=mwhen the operational specifications on quality factor (Q)and field emission were included (the so-called usable gradient) Only 16 of the cavities required anadditional surface retreatment to recover their low performance (usable gradient less than 20 MV=m) Thesecavities were predominantly limited by excessive field emission for which a simple high pressurewater rinse(HPR)was sufficient Approximately 16of the cavities also received an additionalHPR eg due tovacuumproblems before or during the tests or other reasons but these were not directly related to gradientperformance The in-depth statistical analyses presented in this report have revealed several features of theseries produced cavities

DOI 101103PhysRevAccelBeams20042004

I INTRODUCTION

The successful construction of the 175 GeV EuropeanX-ray Free Electron Laser (XFEL) [1] represents theculmination of over 20 years of superconducting radio-frequency (SRF) RampD The core high-gradient 13 GHznine-cell niobium cavity and cryomodule technology wasoriginally developed by an international collaborationcoordinated by DESY for TeV-Energy SuperconductingLinear Accelerator (TESLA)mdasha 500 GeV center-of-masselectron-positron linear collider and integrated x-ray FELformally proposed in 2003 [2] The primary goal of thisdevelopment was the reduction in cost per MVover the thenexisting SRF accelerators by a factor of 20 This wassuccessfully achieved by an increase in gradient by over afactor of 5 to an average of approximately 30 MV=m theremaining factor of 4 accruing from the integration of eightcavities into a single cryogenic cryomodule A total of 102cryomodules containing 816 cavities have been producedof which 97 have been installed in the SRF linacs of the

European XFEL This represents by far the largest deploy-ment of TESLA technology to datePreparations for the industrial cavity production began in

2006 The foreseen production and testing rates of eightcavities per week represented a significant step up from anyprevious series production TESLA-type elliptical niobiumcavities had already been successfully mechanically fab-ricated by industry during the RampD phase but the final anddelicate surface chemical treatments required for highperformance had always been made by the labs For theEuropean XFEL series production the cavities werecompletely constructed by industry including mechanicalfabrication and chemical surface preparation and weredelivered to DESY ready for cold vertical testing Thisrequired significant technology transfer to industry the twovendors contracted for the cavity production [Ettore ZanonSpa (EZ) Italy and RI Research Instruments GmbH (RI)Germany] had to invest in the construction and commis-sioning of significant new infrastructure in order to meetthe stringent fabrication requirements and productionscheduleCavity delivery began in December 2012 ramping up to

full production rate in October 2013 and then continueduntil the end of 2015 [3] The niobium material wasprocured by DESY and supplied to the vendors [4]A key production strategy was the so-called ldquobuild toprintrdquo concept whereby the vendors where required toextensively document each step of the production adhering

Corresponding authordetlefreschkedesyde

Published by the American Physical Society under the terms ofthe Creative Commons Attribution 40 International licenseFurther distribution of this work must maintain attribution tothe author(s) and the published articlersquos title journal citationand DOI

PHYSICAL REVIEW ACCELERATORS AND BEAMS 20 042004 (2017)

2469-9888=17=20(4)=042004(31) 042004-1 Published by the American Physical Society

to well-defined procedures and tolerances Providing thecavities were fabricated to these specifications DESYaccepted the cavity without any final cold rf performanceguarantee As a result DESY took the responsibility of anyremedial (postproduction) measures necessary to recoverpoor-performing cavitiesThe resulting average performance of sim30 MV=m with

an unloaded Q value Q0 ge 1010 of such a large industrialseries production represents an unprecedented success forthis SRF technologySuch a large scale production can only be successfully

completed by the work of a large team of experts in variousfields The responsibility for cavity production was sharedbetween INFN MilanondashLASA and DESY and includedabout 25 people in the core team Cold vertical testingof individual cavities as well as cryomodule tests wereperformed at DESY in the Accelerator Module Test Facility(AMTF) [5] by a team of 30 technicians engineers andphysicists from the Henryk Niewodniczański Institute ofNuclear Physics Polish Academy of Sciences (IFJ-PAN)Krakow supported by DESY expertsThis report focuses on the detailed analysis of the

resulting performance of the series production cavitiesThe main body of the report is separated into the followingfour sectionsSection II Cavity Production provides a brief overview

of themechanical fabrication and chemical surface treatmentof the cavities by industry and introduces somekey conceptsand terminology important for the remainder of the reportSection III Vertical Acceptance Test details the setup

and procedures for the cold vertical testing of the cavitiesin AMTF This section introduces the overall workflow andin particular the definition of the usable gradient andprovides details of the careful procedures and qualitychecksmdashquality assurance (QA) and qauality control(QC)mdashof the cavities during this testing period Theobserved nonconformities during QC and cold verticaltesting are also briefly discussedSection IV Production Performance Statistics provides a

comprehensive set of statistical analyses which quantify theoverall cavity production in terms of gradient Q0 andfield-emission performance Performance limitations andthe impact of the surface retreatments performed atDESYon (predominantly) low-performing cavities are alsoquantified and performance trends during the production

discussed Finally the results from the European XFELseries production are compared to existing historical dataSection V Detailed Cavity Performance Studies covers

the following special-topic studies differences betweenvendor performance impact of the infrastructure at both thevendors and at DESY studies on high pressure water rinse(HPR) application impact of special procedures such astransport helium-tank welding repair of the two-phasehelium pipe studies of process and field-emission onsetduring the vertical test and an estimate of the residualsurface resistance based on additional 18 K measurementsFinally Sec VI provides a summary and some conclud-

ing remarks

II CAVITY PRODUCTION

The cavity design for the European XFEL acceleratorcavity is based on the ldquoTESLA cavity designrdquo [6] as asuperconductingnine-cell cavitymadeof solid niobiumwiththeπ-mode frequencyof theTM010-passband at 1300MHzMinor modifications compared to the original design havebeen applied for a simpler fabrication and cost reduction [7]

A Cavity fabrication and material

Series production of the 13 GHz cavities was equallydivided between Ettore Zanon SpA and RI ResearchInstruments GmbH Production included both mechanicalfabrication and surface preparation [38] together with therequired extensive quality assurance quality managementand associated documentation [9] Details about theniobium and niobium-titanium material used can be foundin [4] The Nb material for the accelerating cells wasdelivered in differing amounts by three companies (TokyoDenkai1 Ningxia OTIC2 and SE Plansee3) The materialfrom each company was distributed equally to the twocavity vendors The rf measurements for quality assuranceduring the cavity production are described in [10]Table I provides a summary of the cavity production

numbers A total of 852 cavities were produced by thevendors (429 and 423 cavities by RI and EZ respectively)

TABLE I Total number of cavities produced by vendors

RI EZ Total Comment

Infrastructure setup and commissioning 8 8 16 4 modified to series cavitiesILC HiGrade 12 12 24 For QC and RampD 8 modified to series cavitiesSeries 400 400 800Rejected and replaced 6 2 8 Replaced at vendor cost included in total 800Additional new orders 3 1 4Totals 429 423 852

1Tokyo Denkai Co Ltd2Ningxia Orient Tantalum Industry Co Ltdhttpwwwnnieccom

3SE Plansee httpswwwplanseecom

D RESCHKE et al PHYS REV ACCEL BEAMS 20 042004 (2017)

042004-2

This number comprises the original order of 840 cavitiestotal plus an additional eight cavities as replacements forrejected cavities (at the vendorsrsquo expense) together with anadditional order of four cavities Of these cavities 831have vertical tests which are included in the analysis(see Sec IV) 804 XFEL series cavities (401 by EZ403 by RI) were delivered complete with helium tank(Fig 1) ready for vertical testing at DESY in AMTF Eachvendor also produced an additional 12 cavities without ahelium tank for the ILC-HiGrade program [1112] whichwere used as a quality control tool as well as for furtherRampD For 8 of these 24 cavities a subsequent assembly ofthe He tank was made In addition 4 of the additional 16cavities used for infrastructure setup and commissioning atthe vendors [3] have since been fitted with a He tank for usein the assembly of the 102 cryomodulesThe concept of the delivery ldquoready for vertical testingrdquo

included transportation under vacuum conditions Allcavities were fully equipped with their Higher-OrderMode (HOM) antennas pick-up probe a high-Q inputcoupler antenna with a fixed coupling and two beam-tubeflanges The high-Q input coupler was only required for thevertical test and was subsequently removed prior to moduleassembly while the remaining HOM antennas and pick-upprobe remained in placeAs noted in the introduction both vendors were required

to follow exactly well-defined specifications for themechanical fabrication and surface treatments but no coldrf performance guarantee was required [3] An importantrisk DESY accepted was writing and enforcing the preciseand complete specification A key result of this approachwas that DESY ultimately accepted the risk of low-performance cavities together with the responsibility forany necessary mitigation (ie surface retreatment seeSec IV D) providing the cavities were shown to havefulfilled the build to print specifications

B Surface preparation

The surface preparation at both vendors (Fig 2) [3813ndash16]started with a bulk electropolishing (EP EZ 140 μm RI110 μm) followed by 800 degC annealing but for the final

surfacepolishing twoalternative recipeswereusedEZapplieda final chemical surface removal (ldquoflash BCPrdquo where BCPstands for buffered chemical polishing) of 10 μm while RIapplieda finalEPof40 μm(ldquofinalEPrdquo)Both finalpreparationsteps included extensive rinsing with ultrapure water at apressure of sim100 bar in an ISO 4 cleanroom (HPR) particlefree assemblyof all flanges under ISO4cleanroomconditionswell-defined vacuum requirements and a 120 degC bakingAll cavity transportsmdashfrom thevendor toDESYaswell as

from DESY to CEA Saclaymdashwere made by truck with thecavity orientated horizontally in specially designed transportboxes for minimizing mechanical shocks to the cavity [17]The cavities were transported under vacuum without activemonitoring The vacuum status was part of the incominginspection at both DESYand CEA Saclay Data loggers forshocks were used during the transport from DESY to CEASaclay for one cavity of each batch and on the truck itselfFor those cavities requiring an additional surface treat-

ment the surface reprocessing (either at DESY or at thevendors) was strictly performed under the same cleanlinessrequirements as for the initial surface preparation (seeSecs III A IV D and V F) The choice of the retreatmentprocedure applied depended on the characteristics of theobserved limitation in the cold vertical test and was decidedon a cavity-by-cavity basis Three main retreatment pro-cedures were appliedHPR After the cold vertical test the cavity first had its

outer surface cleaned before being brought into the ISO4cleanroom where it was then vented with ultrapureparticle-free nitrogen before undergoing six standardHPR of 2 h each For the HPR only the beam tube flangewith the angle valve was removed and all other flanges andantennas were kept on the cavity The beam tube flangewith the angle valvewas reassembled after HPR The cavitywas then pumped leak checked (including residual gasanalysis) and transported back to the vertical test standBCP After cleaning venting and disassembly the

cavity received a 10 μm chemical surface removal (BCP)followed by several rinsing steps and one standard 2-hHPR After reassembly and leak check the equipped cavityreceived six standard HPR of 2 h each The cavity was thenleak checked (including residual gas analysis) received a120 degC baking outside of the cleanroom and was trans-ported to the vertical test standGrinding Precondition for grinding was an optical

inspection at OBACHT (optical bench for automated cavityinspection with high resolution on short time scales)[1819] performed after the vertical test in order to identifythe surface defects4 to be ground Local grinding only tookplace at EZ [20] Afterwards the residual artefacts from thegrinding were removed by a 20 μm BCP and the surfacequality was restored by a 20 μm EP (or more depending on

FIG 1 3D model of the series XFEL cavity equipped fordelivery to DESY

4Defect here includes all mechanical anomalies including forexample inclusions scratches and sputters

PERFORMANCE IN THE VERTICAL TEST hellip PHYS REV ACCEL BEAMS 20 042004 (2017)

042004-3

the ground defect) followed by water and ethanol rinsingThe removal of the defects was confirmed by performinga further optical inspection at EZ The frequency andfield flatness were then checked and if necessary correctedto the required values Finally BCP was applied asdescribed above

III VERTICAL ACCEPTANCE TEST

To verify conformity with the specification for manu-facture and the criteria for module assembly several testprocedures were defined which were applied betweenarrival at DESY of the cavities from the manufacturer upto the preparation for shipment to the cryomodule assembly

plant at CEA Saclay The complete vertical acceptance testsequence for cavities is divided into three main partsincoming inspection cold vertical test and outgoinginspection

A Overview of vertical acceptance test workflow

The vertical acceptance test in the AMTF [521] wason one hand used as a tool for quality control of thecavity production and on the other hand to facilitate sortingof cavities of similar performance into cryomodules Anoverview of the workflow is shown in Fig 3 A seriescavity with no nonconformity (NC) and with acceptablerf performance (see Sec III B) followed the standard

FIG 2 Overview of the surface treatment at RI (final EP scheme left) and EZ (final flash BCP scheme right)


042004-4

workflow (white marked boxes) starting from ldquoincominginspectionrdquo to ldquorf test (VT)rdquo and on to the ldquooutgoinginspectionrdquo Upon review each vertical test (VT) receivedone of several possible ldquodecisionsrdquo (see Fig 3) with thedecision ldquosend to Saclayrdquo (point 5 in Fig 3) the cavity wasaccepted for cryomodule string assembly and shipped toCEA Saclay ldquoILC HiGraderdquo cavities without NC wereinternally used for further QA and RampD purposes after thevertical test was complete (Fig 3 point 4)In the case of NC at key points in the workflow different

retest or retreatment cycles (or both) were applied asfollowsA NC during the incoming inspection (Sec III C 1) first

underwent an expert review and was then either rejectedand sent back to the vendor (Fig 3 point 3) or repaired atDESY depending on the reason for the NCHandling of vacuum and rf NC during the cold vertical

test itself (Fig 3 point 2) are described in Sec III DAn unacceptably low usable gradient (see Sec III B)

in the vertical test triggered a retreatment cycle (Fig 3point 1) As there was no performance guarantee requiredof the vendor performance-driven retreatments were ingeneral made at DESY5 The choice of retreatment processwas decided by an expert review and depended on thecritical limitation of the rf test as well as on the history ofthe cavity In most cases a cycle of HPR was applied as thefirst retreatment (see Secs II B and IV D)The special case of cavities being returned from CEA

Saclay is briefly discussed in Sec III EThe results of each incoming inspection and outgoing

inspection were documented in a report stored in the DESYEngineering Data Management System (EDMS) [922]Awell-defined data set of relevant rf parameters from each

valid cold vertical test was stored in the European XFELcavity database (see Sec III G)

B Definition of usable gradient and acceptance criteria

Although all cavities were tested to their maximumachievable gradient (Emax) of greater importance foraccelerator operation was the usable gradient (Eusable)which took Q0 as well as field-emission performance intoaccount It was defined as the lowest value of (i) quenchgradient (quench limited) (ii) gradient at which Q0 dropsbelow 1010 (Q0 limited) and (iii) gradient at which either ofthe x-ray detectors exceeds the threshold (field-emissionlimited)For the field-emission limit the acceptable x-ray thresh-

olds were set to 001 mGy=min and 012 mGy=min for thetop and bottom detector (see Fig 4) respectively Thethreshold 001 mGy=min was based on experience from thecavity testing for free-electron laser in Hamburg (FLASH)The higher limit for the lower detector was a geometricaleffect A bottom detector was implemented for the first timeat DESY in order not to miss downwards directed radiationand to get a more accurate value for the usable gradient Forabout 9 of the cold vertical tests the bottom detectorlimited the usable gradientAt the beginning of production the criterion for

acceptance for module assembly was specified asEusable ge 26 MV=m chosen to give a margin of sim10compared to the required average design operationalgradient (236 MV=m at Q0 ge 1010) Based on an analysisof about 270 cavities tested up to May 2014 including thenecessary retreatments and retests the acceptance thresholdwas reduced to Eusable ge 20 MV=m in order to optimizethe number of vertical tests while still maintaining anaverage module gradient of ge236 MV=mCavities with Eusable lt 20 MV=m were considered for

further processing or re-treatment The exact nature of the

FIG 3 Workflow of the vertical acceptance test A simplified list of ldquodecisionsrdquo (applied after each vertical test) is given in the graybox and their related branch points are indicated by the red numbered circles

5For 17 cavities the vendor agreed to accept responsibility forthe retreatment


042004-5

handling of low-performance cavities was judged on acase-by-case basis (expert review see Sec III A)

C Process and implementation

1 Quality control checks

An important tool for quality control for all EuropeanXFEL cavities was the incoming inspection which was thefirst step after the cavity was received at DESY from thevendor and was used to detect nonconformities fromthe specification Immediately after arrival a brief visualinspection of the cavity within the transport box (Sec II B)was made to look for obvious damage followed by themechanical part of the incoming inspection after removingthe cavity from its transport box Each cavity was checkedfor completeness correct assembly of all components aswell as for mechanical NC such as significant scratchesdents and other irregularities More than 500 cavitiesreceived an additional dimension check of the mechanicallongitudinal axis which was compared to the data providedby the vendorsIn order to detect possible deformations of the cavity

shape during transportation the frequency spectrum of theTM010 passband was compared to the last one measured atthe vendor before shipping In addition the electricalintegrity of the rf antennas was checkedAfter passing these checks the cavity was installed

into a test insert [23] ready for the cold vertical test (seeSec III C 2) After mechanical installation into the insertthe cavity was connected to a pumping unit and a leakcheck and residual gas analysis (RGA) were performed (thevacuum part of the incoming inspection see Sec III C 5)

All observations were recorded in the so-called ldquoincominginspection reportrdquoWhen the cavity failed to pass one or more of the criteria

of the incoming inspection and an ad hoc repair was notpossible it was sent back to the vendor for repair Table IIgives the total number of cavities which were returned tothe vendors from the 832 cavities that received a regularincoming inspection Approximately 6 of all cavitiesfailed to pass incoming inspection mostly in the firstmonths of the series production The main reasons forfailing incoming inspection were wrong torque on theaccessories screws inconsistent data for the mechanicalaxis or vacuum problems including visible contaminations(ldquofibersrdquo) on the inside of the cavity angle valve Anadditional 11 of the cavities were sent back after the coldvertical test The reasons included cavities with vacuumproblems due to poor assembly by the vendor (bad sealing)or cavities with surface irregularities observed on the innersurface [1819]Finally after the vertical test and before being shipped

for cryomodule assembly an outgoing inspection wasperformed It included a detailed visual outside inspectionas well as additional measurements of the TM010 passbandspectrum and a check of the HOM rejection filter tuning(The vacuum integrity had already been addressed duringthe leak check and RGA made directly after the coldvertical test see Sec III C 5) The results were recorded inan outgoing inspection report If the cavity passed theoutgoing inspection it was released for cryomoduleassembly and shipment to CEA Saclay

2 Vertical test stands and rf procedures

The vertical acceptance tests of the series as well as theHiGrade cavities took place in the AMTF at DESY [2425]In order to achieve the desired testing rate of at least eight toten cavities per week the cold vertical tests were madeusing two independent test systems (labeled XATC1 andXATC2) each consisting of an independent bath cryostat(Fig 4) and rf test stand Each test cryostat accepts oneldquoinsertrdquo which supports up to four cavities greatly increas-ing the efficiency of cool-down and warm-up cycles Sixinserts were available to fulfill the required testing rate(see Fig 5)

FIG 4 Section diagram of the vertical test cryostat Note thatonly three of the four cavities are visible

TABLE II Statistics of cavity returns (before and after the firstcold vertical test) The total of 142 returns corresponds to 129individual cavities (13 cavities were returned to the vendor twice)

CompanyReturns to

company (total)Returns aftervertical test

Failed incominginspection

RI 79 55 24EZ 63 36 27Total 142 91 51


042004-6

As all cavities were tested with assembled HOMfeedthroughs the cold vertical tests used a ldquolong pulserdquomode with typically 5ndash20 s rf-on followed by an rf-off timeof approximately 50 s in order to protect the HOMfeedthroughs against overheating during the Q0ethETHORN mea-surements described below A true continuous wave (CW)operation would have increased the heat load on thefeedthroughs by a factor of 100 compared to the nominalEuropean XFEL rf operation mode In addition the damp-ing properties of the HOM notch filters were such that onlyQ0ethETHORN measurements for the π mode of the TM010passband were possibleAll cavities were equipped with fixed rf input antennas

(high Q antenna) For 2 K operation the expected unloadedQ value (Q0) of the cavities was about 2 times 1010 The designvalue of the external input Q value (Qext) was 8 times 109 foran effective coupling at higher gradients in order to allow rfprocessing in case of field emission (see Sec IV B 3) Theexperimentally observed Qext typically varied between5 times 109 and 9 times 109 while the cavity probe showed atypical coupling factor of lt001 The external Q values ofboth HOM couplers were adjusted by tuning of the HOMnotch filters before each vertical test [26]The cold vertical tests followed a standardized and

automated procedure which included the measurementof the Q0 value versus the accelerating gradient (E) at 2 Kfor the fundamental π mode as well as the frequencies ofall modes of the TM010 passband6 For each point of theQ0ethETHORN curve x rays were measured inside the concreteshielding above and below the cryostat (Fig 4) For eachvertical test at least two power rises were performed andcompared with respect to the rf parameters and cavity

performance Only after achieving two consecutive repeat-able Q0ethETHORN curves was the vertical test considered com-plete For each power rise the operator attributed a limitingreason for the maximum achieved gradient according toone of the following categories breakdown breakdownwith field emission forward power limited forward powerlimited with field emission or quenching caused by heatingof the HOM couplerIn order to avoid frequent damage of the rf input cables

the maximum rf input power was limited to about 200 WBeyond this no general administrative limit was appliedbut the surrounding of the cryostats was under permanentradiation protection surveillance Radio-frequency cablesand connections were checked for defects before each testby time domain reflectivity measurementIn addition to the Q0ethETHORN curves many cavities had the

HOM frequencies and Q0 values of the TE111 TM110and TM011 modes measured [27]For the series vertical tests in AMTF no additional

diagnostics like T-mapping second sound or fluxgatemagnetometer were available A few cavities obtainedadditional vertical tests for diagnostics purposes in adifferent test facility [192829]The vertical test infrastructure started operation in

February 2013 and achieved full operation in October2013 with a stable average testing rate of approximately 40vertical tests per month (Sec IVA) The last verticalacceptance test was finished in March 2016

3 Ambient magnetic field in the test cryostats

The cryogenic losses for superconducting rf acceleratorsdepend linearly on the Q0 value of the resonators Onesignificant influence on the Q0 (more specifically the rfsurface resistance) is the ambient magnetic field Hence a

FIG 5 Left vertical test inserts ready for test Right 3D view drawing of a test insert supporting four equipped cavities

6A few cavities were additionally measured at 18 K


042004-7

suppression of the ambient magnetic field below a value of1ndash2 μT is necessary (see Chap 342 in [30] [3132])The magnetic field in the concrete pit for the two

vertical cryostats was approximately 40 μT Simulationsshowed that two magnetic shields were necessary adouble-walled warm magnetic shield around the cryostats(300 K) and an additional smaller one inside of thecryostat at 2 K All shields were cylindrical with a closedbottom (see Fig 4) Unfortunately there was no spaceavailable to use the more effective geometry of a smallerdouble-walled cold magnetic shield The outer warmshields cover the complete cryostat except the topThey were constructed from μ-metal which has itsmaximum permeability close to room temperature Thecold magnetic shield placed inside of the cryostat wasmade out of Cryopermtrade a material with a higherpermeability at cryogenic temperatures [33] Its heightabove the top of the cavities was limited to 270 mm dueto the position of helium supply and return pipes Allshields obtained a heat treatment (up to 1100 degC) aftermanufacturing (cutting drilling and welding) to bringthe materials back to their initial magnetic properties[34] Measurements of the magnetic field inside of thecryostat were made at room temperature and are givenin Fig 6

4 Cool-down procedures

The cool-down procedures at the AMTF vertical teststands were heavily automated (programmable logic con-troller based) and only required operator supervision[2135] Cernoxtrade temperature sensors on one of the insertswere used to commission the procedures and monitor themechanical stress on the inserts [23] Carbon temperaturesensors (TVO) [36] glued to the outside of the cryostatswere used during regular operation During series verticaltesting neither the inserts nor the individual cavities wereequipped with temperature sensorsThe cool down from 300 to 100 K took approximately

12 h which represents an average cool-down rate of5 mK=s The cryostat remained at 100 K for 4ndash6 h toenable a test for hydrogen Q disease (see Chap 341 in[30]) after which the cavity was further cooled to 4 K Thecool-down rate across Tc was of the order of 200 mK=s atwhich point a maximal longitudinal temperature gradientalong the cavities of about 25 K was typically observedThe final cool down to 2 K was performed manually

by reducing the vapor pressure of the helium bath tosim30 mbar achieving an average cool-down rate of05 mK=s

5 Vacuum systems

Fundamental to the cold-vertical-test philosophy for theseries production was the concept that the cavities remainedunder vacuum from the vendor to the cryomodule stringassembly at CEA Saclay However it was consideredbeneficial to allow active pumping on the cavity vacuumduring vertical testing to provide continuous monitoring ofthe vacuum status and in particular to quickly identify leakswhen they occurred during the tests This required thecavity to be openly connected to a local pumping systemduring the test which needed to be performed in a carefullycontrolled fashion so as not to compromise the vacuum ofthe cavity The fundamental requirements with respect toleak rate and RGA are described in [37]Figure 7 shows the vacuum schematic of the insert used

in the tests The vacuum for each of the four cavities waseffectively kept independent so that if there was a vacuumissue with one cavity this cavity could be actively pumpedand the remaining three could still be tested The fourvalves labeled V41ndashV44 were the all-metal angle valvesthat were delivered with the equipped cavity from thevendor (one per cavity) The vertical test insert had its ownset of all-metal angle valves (labeled V31ndashV34) located atthe top plate of the test insert which are connected to thecorresponding valves on the cavities via pipes with bellowsThis arrangement allows the insert vacuum system to beindividually monitored and pumped to the required vacuumconditions at both warm and cryogenic conditions Thevacuum-related procedure for a vertical test was as follows1 The four cavities (as delivered from the vendor

ie under vacuum with their angle valves closed) were

FIG 6 Magnetic field distribution inside the cryostats XATC1(top) and XATC2 (bottom) of the bottom 2-m vertical section ofthe cryostat The vertical extent of the cavities is indicated in gray


042004-8

mounted into the insert in the preparation area and con-nected to the insert vacuum system via the four bellowsThe insert vacuum was then pumped (up to the cavity anglevalves) using the local pump station in the preparation area(via the main valve labeled V1thorn V2 in Fig 7) and thevacuum status monitored including RGA In general allfour pipes were pumped in parallel and a common RGAmade to assure the insert vacuum had achieved the requiredspecification7

2Once the requiredvacuumofle10minus7 mbarwas achievedin the insert connections the cavity angle valves wereopened one by one According to [37] for each individualcavity the leak rate had to be below 10minus10 mbar l=s at apressure inside the cavity of le10minus7 mbar (typically after30 min of pumping) and the RGA requirements had to besatisfied After each vacuum check the cavity angle valvewas closed before the next cavity angle valve was openedAfter a successful checkof all cavities all cavity anglevalveswere reopened3 The main valves V1 V2 and the insert angle valves

(V31ndashV34) were then closed and the entire insert pluscavities disconnected from the pumping station in thepreparation area The loaded insert was then transportedacross to one of the two vertical test cryostats where it wasreconnected to the test-stand pumping station After thespecified vacuum of le10minus7 mbar was achieved on thepump-side vacuum system all valves were opened again toallow a common RGA measurement after which thevacuum gauges (FRG 1-4) were switched on When thepressure was below the requisite 10minus7 mbar the insertangle valves were closed

4 During the entire thermal cycle the pressure wasconstantly monitored If the pressure increased in a cavitythe insert angle valve connected to that cavity was openedfor active pumping If despite active pumping the pressureexceeded 10minus2 mbar the cool-down cycle had to beaborted5 After warm up to room temperature and after closing

all insert angle valves and the main valves the insert wasdisconnected from the test-stand pump station and trans-ported back to the preparation area6 A final common RGA and leak check was then made

for all cavities after which the cavity angle valves whereclosed the insert vacuum lines vented and the cavitiesdisassembled from the insertAll vacuum connections were made in a particle-free

environment using local clean rooms (ISO 5) to minimizethe risk of contamination or leaks and an oil- and particle-free state for all the vacuum parts was mandatory [37]The design of the insert vacuum system is based on

DN40 pipes which provide sufficient pumping speed forthe cavities As noted above with the exception of the needto actively pump a cavity the insert angle valves V31ndashV34 were closed during the test and subsequent warm upAs a safety measure against overpressure during the warmup (ensuing from a cold liquid helium leak during the test)four burst discs were located on the top of the insertbetween the warm insert valves and cold cavity valves

6 Measurement errors

A bottoms-up analytical estimate based on the usual rftest equations ([3839] see Chap 8 in [40]) results in anuncertainty of independent rf measurements (test to test) ofabout sim10 for E and up to sim20 for Q0 This takes intoaccount the reproducibility of cable connections errors ofthe test devices (eg power meter frequency counter andoscilloscope) as well as the finite directivity of the direc-tional couplers Within a single cold vertical test and eachQ0ethETHORN curve the observed measurement scatter is muchsmaller (about 1 and 3 for E and Q0 respectively) andonly depends on the errors (eg resolution and drift effects)of the test devicesAn additional statistical error analysis has been estimated

from the distributions of the cable calibration parametersand the external Q (Qext) based on about 800 cold verticaltests [41] The resulting total rms error from these effects is66 for Q0 measurement and 33 for the gradientmeasurement These errors are well within the analyticalestimates above and as they do not include all effects themore conservative values were generally takenThe estimated uncertainty in the usable gradient (see

Sec III B) was in general larger since its determination wasbased on interpolation of either the field-emission (x ray) orthe Q0 curves during the power rise In the event that thelocal derivative of either of these values with respect to themeasured gradient was small the error in the respective

FIG 7 The vacuum system of the vertical insert V1=V2 mainvalves V31ndashV34 all-metal angle valves placed on the top of theinsert (room temperature) V41ndashV44 all-metal angle valve onthe cavities (2 K) FRG 1ndashFRG 4 full range gaugethorn burst discI1=I21-I24 bellows

7For the first tests individual RGA were made on each of thefour pipes until it became clear that the system operated well andthe time-consuming individual testing was no longer necessary


042004-9

value would naturally transform into a larger uncertainty inusable gradient This was particularly true of Q0-limitedcavities where the slope around the Q0 frac14 1010 thresholdwas often relatively low

D Nonconformities during vertical test

The cold vertical tests are not without risk and non-conformities did occur The main NC were either vacuumproblems (cold leaks leaks or damage on accessories orvacuum problems with the test setup) or problems occur-ring during the rf measurements Both cases will beconsidered separately below

1 Vacuum problems

Leak tightness of the superconducting rf cavities undercryogenic conditions was an essential requirement forassembly into a cryomodule Hence a leak check wasdone before and after the cold vertical test During the coldvertical test the vacuum pressure was observed andrecorded If a deviation in the pressure to the standardbehavior was detected during the cool-down or warm-upcycle the test was declared as nonconforming If the leakcould be localized the defect component was repaired orexchanged If not the cavity was then reconnected to thevacuum system and tested again in order to exclude thatthe problem arose from the vacuum connection between thecavity and the insert (see Sec III C 5) If this test wassuccessful (ie no leak seen) then the cavity was releasedfor the next step In the case where the second test stillindicated the presence of a leak it was generally concludedthat the leak was due to the cavity itself and all components(antennas and flanges) were then disassembled and reas-sembled under cleanroom conditions followed by a fullHPR cycle after which a further vertical test was performedto check the performance (Sec IV D)Table III gives a breakdown of all vacuum NC during

the vertical acceptance tests No Q0ethETHORN measurementswere possible for 86 vertical tests due to a vacuum NCin one or more of the four cavities in the same insert (seeSec III C 5) All cavities could be successfully recovered

and eventually fulfilled the vacuum requirements accordingto the specifications

2 Radio-frequency-related NC

In 24 cold vertical tests a problem with the rf measure-ment itself was observed and it was decided to repeat thetest to verify the performance of the cavity At thebeginning of the series cavity testing several minor hard-and software problems with the new infrastructure atAMTF occurred this included saturated analog-to-digitalconverterrsquos (solved by correct adjustment of rf attenuators)and phase locking problems (solved by an additional phaseswitch) Over the 4 years of cold rf testing a few broken rfinput power cables required a retest In some cases anunusual coupling parameter value or some other strange rfbehavior was observed which could not be clarifiedimmediately during the cold vertical test In these casesan investigation took place after warm up and the coldvertical test was repeated

E Cavities returned from CEA Saclay

A total of 52 cavities were returned to DESYafter beingshipped to the cryomodule assembly facility at CEA SaclayThe reasons were due to either nonconformities duringstring assembly (for example incorrect venting or pumpingprocedures or coupler or gate valve assembly errors)nonconformities identified during the incoming inspectionat CEA Saclay (vacuum leaks defective valves etc) orcavities recalled by DESY (additional testing repairs ortransport tests) Most (but not all) of these cavities requiredretreatment (HPR) and cold vertical testing All the cavitiessuccessfully achieved the usable gradient requirements andwere subsequently returned to CEA Saclay The over 60vertical tests associated with these special cases are notincluded in the analysis presented in Sec IV

F Categorization of tests

To facilitate detailed statistical analysis of the over 1200vertical test results a series of flags were developed tocategorize each individual test based on the reason whythat test was performed The vertical tests were broadlydivided into the following categories the primary accep-tance test performed on receipt of the cavity from thevendors (referred to as ldquoas receivedrdquo tests) a vertical testafter an additional surface treatment (ldquoretreatmentrdquo)repeated tests (ldquoretestsrdquo) in general due to a technicalissue with the test itself and a relatively small fraction oftests (ldquootherrdquo) which were associated with commissioningof infrastructure or were performed on cavities beforeintegration into their helium tanks at the request of thevendors (referred to as ldquopreliminary acceptance testsrdquo)Where applicable the tests were further categorized bywhere the associated action (eg retreatment) occurred ieeither at DESY or at the vendor Finally to facilitate more

TABLE III Number of vacuum NC during vertical acceptancetest

NC during incoming inspectionContamination (bad RGAvacuum) 7Leaky component (flange antenna) 15Particle contamination 16Operator error 4Nonconformity during cold vertical testPressure increase during cold vertical test 76(a) with all components changed 12(b) with second test accepted 36(c) with leaky identified component (flange antenna) 28Problems with vacuum system 14


042004-10

detailed-level analysis the retreatment and retest categorieswere further subdivided into the exact reasons for the testFor the tests flagged as retreatment related the type of

retreatment was also recorded eg high-pressure rinse only(HPR) buffered chemical polishing (BCP) followed byHPR and a 120 degC bakeThe analysis results presented in Sec IV are extensively

based on the use of the flags to filter the cavity test resultsHowever to simplify the analysis only the primary level ofcategorization was used (as received retreatment retest orother) with the notable exception of retreatment where twoanalysis-level subcategories were defined retreatment(performance) and retreatment (other) The distinction isimportant since it allows us to cleanly separate out thosevertical test results which are associated with a retreatmentapplied in order to increase the rf performance of a cavity(recovery) from retreatments which were necessary due tosome NC with the vertical test itself (for example a vacuumleak during the test see Sec III D for details)Tests flagged as ldquoas receivedrdquo are particularly note-

worthy Ideally all cavities should have had an ldquoasreceivedrdquo test as the first test after acceptance of the cavityfrom the vendor However approximately 10 of thecavities do not have an ldquoas receivedrdquo test due to the factthat these cavities had a nonstandard production historyand in particular already underwent an additional surfaceretreatment before the first vertical test The set of ldquoasreceivedrdquo tests is considered to give the best indication ofthe production performance for cavities having undergoneonly the specified production process See Sec IV for moredetails and also Sec III C for information concerning NCduring incoming inspectionA complete summary table of the database flags is given

in the Appendix

G Data storage and analysis tools

1 Database for AMTF processes

The AMTF data base [42] (based on Oracletrade RDBMS8)was created to document (i) the progress and status of workperformed on cavity and accelerator modules including atime stamp and the name of the responsible operator (ii) theresults of rf tests in warm and cold conditions (iii) thecheck and release of key rf data of each vertical acceptancetest by the responsible expert as an additional QC beforetransferring them to the European XFEL cavity database(iv) the usage of infrastructure at AMTF (v) the ldquodecisionrdquomade after the vertical acceptance test defining the follow-ing workflow of a cavity (see Fig 3) and (vi) the physicallocation of cavities and cryomodulesThe AMTF database was a mandatory and effective tool

for organizing the daily workflow in AMTF handlingmultiple cavity and cryomodule activities in parallel

2 The European XFEL cavity database

The so-called ldquoEuropean XFEL cavity databaserdquo[43ndash46] is also based on the Oracle RDBMS platformand contains selected information about the fabricationtreatment and testing of cavities high-power input cou-plers and cryomodulesFor the cavity fabrication key data were transferred to

the database mainly by an interface to the DESY EDMSsystem but also by direct data loading especially for rfmeasurement data At AMTF the TM010 passbandfrequencies the characteristic data of the vertical accep-tance test (Sec III C 2) as well as HOM frequencies ofdedicated modes were stored Data loading from the AMTFdatabase to the European XFEL database was doneautomatically based on flagging new data for loadingIn addition for the vertical acceptance tests the

European XFEL database included a special graphical userinterface-based editor to allow an expert to manually add ormodify the following database entries (i) a flag tocategorize the reason of each individual test (Sec III F)(ii) description of the cavity treatment before the test(iii) correction of the x-ray thresholds with respect to theusable gradient for each power run of the test (iv) correctionof the measured Q0 parameter (v) general comment foreach test and (vi) cavity retreatment procedure descriptionas a standardized abbreviationThis information together with the calculation of the

ldquousable gradientrdquo Q0eth4 MV=mTHORN Q0eth236 MV=mTHORN foreach power run were extensively used for the analysispresented in Sec IV A significant effort was undertaken toensure the consistency of the database data by manualcrosschecking with the raw dataA subset of the vertical acceptance test data can be

accessed by a graphical user interface providing forexample individual Q0ethETHORN curves listing key rf dataand comparing TM010 passband frequencies of eitherindividual or several selected cavities The direct accessto all data requires Oracle SQL which allows the furtheranalysis with tools like Mathematicatrade as applied in thispublication

IV PRODUCTION PERFORMANCE STATISTICS

In this section the results of the vertical tests and theeffectiveness of the surface retreatments will be discussedHere the focus is on the overall statistical figures of merit(mean rms yield etc) of the cavities produced and thedominant performance limitations Section V will deal withmore specific specialized studies

A Cold vertical test statistics

All of the 831 cavities used in the analysis (Table IV)received one or more cold vertical tests The total numberof cold vertical tests was approximately 1340 of which1200 have been used in this analysis The remaining 108Relational Database Management System


042004-11

of the tests were either ldquopreliminary acceptancerdquo tests ofcavities requested by the vendors (before being mountedinto the helium tank) tests rejected due to technical issueswith the tests themselves special tests used for infra-structure commissioning or finally tests associated with the52 cavities that were returned from the cryomoduleassembly facility at CEA Saclay due to some NC duringthe string assembly (see Sec III E)Ignoring rejected and excluded tests the average number

of vertical tests per cavity was approximately 148 (145and 152 for RI and EZ respectively) This includes allnecessary additional testing due to surface re-treatment ofcavities or for other test-related reasonsFigure 8 shows the number of cold vertical tests per

calendar month in the AMTF The average total testing rateis indicated by the dashed line (taken from 1 Januray 2013to 31 October 2015) and is approximately 414 tests permonth corresponding to roughly 104 tests per weekHowever much higher testing rates were achieved as canbe seen from the figure corresponding to peaks of up to 15tests per week

Figure 9 shows the fraction of the 831 cavities used in theanalysis which have exactly 1 2 etc cold vertical testsrecorded in the databaseThe reasons that a cavity required an additional cold

vertical test can be loosely divided into two categoriesperformance related ie the usable gradient of the cavitywas considered unacceptable and the cavity was sent foradditional surface treatment and a subsequent cold verticaltest or nonperformance related were the cavity wasretested due to some technical issue with the cavity orthe test itself (see Secs III D and III E) Figure 10 shows abreakdown of the reasons for cavities having a second coldvertical test as an examplePerformance related tests are those labeled ldquoretreatment at

DESY (performance)rdquo and ldquoretreatment at vendor (perfor-mance)rdquo the latter representing cavities that were sent backto either RI or EZ for rework ldquoRetreatment at DESY(other)rdquo and ldquoretreatment at vendor (other)rdquo representnonperformance-related tests where some additional surfaceretreatment was still required (generally HPR) for example

TABLE IV Total number of cavities used in the analysis (see Sec II for more details) The total number of 831 differs from the numbergiven in Sec II A (832) due to the excluded HiGrade cavity

RI EZ Total Comment

Infrastructure setup and commissioning 2 2 4 Converted to series cavities for European XFELILC HiGrade 12 11 23 One early cavity excluded due to identified fabrication errorSeries 403 401 804 Includes new orders and replaced cavitiesTotal 417 414 831

FIG 8 Stacked bar chart showing the number of cold verticaltests by calendar month for RI (orange) and EZ (blue) cavitiesbased on test date The dashed line is the total average test ratetaken from 1 October 2013 to 31 November 2015 which isconsidered the peak testing period (after ramp up and before rampdown)

FIG 10 Reasons for cavities requiring a second cold verticaltest Note that the percent figures indicate the fraction of the totalnumber of cavities (831) and not the fraction of the chart itself(corresponding to the 28 of cavities with more than one test inFig 9)

FIG 9 Relative breakdown of the 831 cavities showing thefraction of cavities which have exactly 1 2 3 or 4 or more testsin the database


042004-12

in the case of a vacuum leak during the cold vertical testldquoRetestrdquo indicates a simple repeat of the cold vertical testusually due to some rf problems or check of cavity after the2-phase pipe repair or some transport issueFinally we consider the breakdown of the first cold

vertical test in the database and the important concept of theldquoas receivedrdquo test (see Sec III F) All cavities received atleast one cold vertical test at DESY of which approx-imately 90 were categorized as ldquoas receivedrdquo Theremaining 10 of cavities had either undergone an addi-tional retreatment at the vendor or in a few cases at DESYbefore this first cold vertical test in general due to problemswith incoming inspection as outlined in Sec III CFigure 11 shows the breakdown of the first cold verticaltest in the database for the 831 cavities

B ldquoAs receivedrdquo yields

In the following the primary statistics for the ldquoasreceivedrdquo performance of the European XFEL cavityproduction are presented As explained above this corre-sponds to approximately 90 of all the cavities in theanalysis However cold vertical tests categorized as ldquoasreceivedrdquo were considered the best indication of the resultsof the nominal production process since this excluded

cavities which required additional treatment due to iden-tified nonconformities or other nonperformance relatedissues before being cold vertical tested at DESYFigure 12 shows the distribution and yield curves for the

maximum and usable gradients achieved in the ldquoasreceivedrdquo cold vertical testsThe average maximum and usable ldquoas receivedrdquo gradient

is 314 MV=m and 277 MV=m respectively with acorresponding rms spread of 68 MV=m (22) and72 MV=m (26) Thus approximately 4 MV=m (12)is lost from the maximum measured performance due toeither above-threshold field emission or low-Q0 behaviorFor the production the more important figure of merit is theusable gradient yield above the performance acceptancecriteria for European XFEL cryomodule assembly origi-nally set to 26 MV=m and then later dropped to 20 MV=mTable V summarizes the key figures of merit for both themaximum and the usable gradientIt is important to note that only approximately 50 of

the cavities were limited by quench (breakdown) of whichsim12 also showed some field emission Of the remainingcavities approximately 22 (35 if field-emitting cavitiesare included) were rf-power limited in the cold vertical testsuggesting that their true physical quench limit is higherHowever the rf-power-limited tests were predominantlycaused by a low Q0 as a result of high-field Q sloperesulting in Q0 values that were below the acceptablethreshold Figure 13 shows the maximum achieved gradientand the associated Q0 for all ldquoas receivedrdquo cold verticaltests broken down by test-limit reason The low Q0 valuesof the rf-power-limited cavities (orange) can be clearlyseen The remaining 16 of cavities were limited in the testdue to other technical reasons including excessive heatingof the HOM coupler9 or in a few cases technical limitscaused by problems of the rf test standOf more interest for European XFEL is the breakdown of

the limiting criteria for the usable gradient as shown inFig 14 The large fraction (51) of cavities limited by onlyQ0 (no FE)mdashie with no field emissionmdashreflects thesimilar fraction of cavities having achieved gradientsgreater than 30 MV=m where the high-field Q slope

TABLE V Key statistics for ldquoas receivedrdquo cold vertical tests(cavities)

Maximum Usable

Average gradient MV=m 314 277Rms MV=m 68 72Median (50) MV=m 325 287Yield ge 20 MV=m 92 86Yield ge 26 MV=m 85 66

FIG 11 Breakdown of the categories for the first cold verticaltest in the database of the 831 cavities

FIG 12 The distributions and yield curves for the maximumachieved gradient (blue) and the usable gradient (orange) for allcavities ldquoas receivedrdquo Yield is defined as those cavities whichhave a gradient greater than or equal to the specified value (Thedarker color represents the overlap of the two histograms)

9The HOM-coupler heating could also be associated withsignificant field emission but this was not separately categorized


042004-13

was the predominant limiting effect The other half of thecavities was limited by breakdown (20) or field emission(29) where for the latter the 9 of tests limited by Q0

indicated in Fig 14 are included since these tests alsoshowed strong indications of field emission which isassumed to be the underlying cause of the observed lowQ0Figure 15 shows the distribution of usable gradient in the

ldquoas receivedrdquo cold vertical tests with cavities limited byfield emission separated out (including the 9 Q0-limitedcases indicated in Fig 14) Although the non-FE limiteddistribution does show a long tail down to low gradientsthe dominant limiting factor below approximately24 MV=m is field emission while above it is mainly eitherbreakdown or Q0 Figure 16 shows the non-FE distribution(as shown in Fig 15) separated into breakdown (BD) and

Q0 Here the dominance of the Q0 limit is evident above20 MV=m The low-gradient tail is mostly BD althoughseveral poor-performing cavities showed Q0 below 1010 atgradients less than 10 MV=mFigure 17 shows the distribution of Q0 measured at low

gradient (4 MV=m) generally field emission free and at236 MV=m (nominal design gradient) The majority ofmeasurements exceed the requirement of 1010 except for asmall fraction at 236 MV=m (sim6) The average perfor-mance is 21 times 1010 and 13 times 1010 for 4 MV=m and236 MV=m respectively clearly indicating the drop inQ0 performance with higher gradientAlthough there was predominantly no field emission

observed in the cavities at 4 MV=m this is not true for

FIG 15 Distribution of the usable gradient in the ldquoas receivedrdquocold vertical tests separated out by field-emission limited(FEthorn Q0) and nonfield-emission limited [BDthorn Q0 (No FE)]

FIG 13 Scatter plot showing the maximum gradient andassociated Q0 achieved for the ldquoas receivedrdquo cold vertical testsbroken down by test limit reason The dashed line indicates theideal 200 W forward power limit Nearly all power-limitedcavities are below the Q0 acceptance criterion of 1010 (BDbreakdown BDthorn FE breakdown with significant field emis-sion PWR forward power limited PWRthorn FE forward powerand significant field emission other eg HOM coupler heatingsee text for more details)

FIG 14 Breakdown of the limiting criteria for the usablegradient for ldquoas receivedrdquo cold vertical tests Those cavities withthe usable gradient technically limited by the Q0 threshold (1010)have been further divided into those with no field emission [Q0(No FE)] and those with field emission (Q0)

FIG 16 Distribution of usable gradient in the ldquoas receivedrdquocold vertical tests for cavities showing no field emissionseparated into breakdown (BD) and Q0(No FE) limited


042004-14

236 MV=m (nominal design gradient) as indicated inFig 15 Figure 18 shows the Q0 results at 236 MV=mseparated into those cold vertical tests showing fieldemission and no field emission For the cavities showingfield emission it is not possible to determine if the Q0 islimited by that field emission or represents the fundamen-tal Q0 of the cavity (comparable to the no-field emissioncase) However below 1010 field emission is likely thedominant mechanism

C Production performance trends

In the previous section the key ldquoas receivedrdquo perfor-mance statistics for the complete cavity production waspresented In this section the history of the performanceover the approximately three-year production period isdiscussed Again the ldquoas receivedrdquo cold vertical testperformance will be presented as a function of the deliverydate to DESY10

Figure 19 shows the production history binned bymonth for the ldquoas receivedrdquo usable gradient as a box-whisker chart The two horizontal red dashed linesindicate the average for the first and second half ofproduction An overall improvement in the averagegradient can be seen in the second halfThe overall improvement can be further seen in the

improvement in yield above 20 MV=m (leading to anoverall reduction in the number of retreatments andsubsequent vertical tests) Figure 20 shows the numberof cavities with ldquoas receivedrdquo usable gradient in the ranges

lt20 MV=m 20ndash26 MV=m and 26ndash28 MV=m Cavitieswith usable gradients ge28 MV=m are not explicitly shownbut can be inferred The data is calculated for quarter-periodproduction A marked reduction in the number of cavitieswith usable gradient below 20 MV=m can be clearly seen

FIG 17 Q0 distributions for the ldquoas receivedrdquo cold verticaltests measured at 4 MV=m (generally field emission free) and236 MV=m (Note that 15 of the cavities did not reach236 MV=m in the ldquoas receivedrdquo cold vertical test and thereforedo not appear in the respective histogram)

FIG 18 ldquoAs receivedrdquo Q0 measured at 236 MV=m separatedinto cavities with and without field emission

FIG 19 ldquoAs receivedrdquo usable gradient performance by deliverymonth The box-whisker symbol represents the gradient distri-bution for the respective calendar month The dark blue joineddots indicate the monthly average gradient The number ofcavities delivered in each month is given at the base of thechart The red dashed horizontal lines indicate the averageperformance over the first and second halves of the productionindicating improved average performance during the latter(The last bin contains all remaining cavities arriving at DESYafter 1 November 2015)

10This differs slightly from the definition used in [3] whereshipment date from the vendor was used


042004-15

in the second half of production leading to an overallrejection rate of approximately sim10 as compared tosim20 during the first production periodFinally for completeness Fig 21 shows the monthly ldquoas

receivedrdquo production statistics for Q0 measured at agradient of 4 MV=m The average low-field Q0 remainedrelatively constant over the entire production period atapproximately 21 times 1010 and unlike the usable gradientno discernible improvement in the second half of produc-tion was seen

D Surface retreatment

As already noted an important aspect of the EuropeanXFEL procurement contracts was that there were norequirements specified for vertical acceptance test per-formance the so-called ldquobuild to printrdquo vendor specifi-cations [3] meant that it was the responsibility of DESYto provide remedial measures to improve poor perfor-mance cavitiesAs discussed in Sec III B the original acceptance

threshold for the usable gradient was set to 26 MV=m(European XFEL nominal design gradient plus 10) butthis was then relaxed to 20 MV=m inMay 2014 essentiallyto reduce the number of additional vertical tests associatedwith retreatments while keeping the required averagemodule gradientA total of 313 retreatments11 were performed for 237 of

the 831 cavities (sim29) included in the analysis (anaverage of 13 retreatments per cavity) Of these retreat-ments roughly two-thirds were performed in the DESYinfrastructure while the remaining third was performed atthe vendors12 However not all of these retreatments weredue to the performance criteria discussed above and inSec III B Approximately 49 of the retreatments (44 atDESY 5 at vendor) were a direct result of nonacceptablerf performance in the vertical test corresponding toapproximately 16 of the cavities The remaining 51of the retreatments (22 at DESY 29 at vendor alsocorresponding to sim16 of the cavities) were necessary as aresult of some nonconformance with either the cavity(Sec III C 1) or the test itself (Sec III D) eg a vacuumleak which subsequently required an additional surfacetreatment (in the majority of cases an additional HPR)About 3 of the cavities got surface retreatments due toboth performance criteria and nonconformance of thecavity or the testFor 16 of the cavities which required retreatment due

to performance limitation the exact choice of retreatmentapplied depended on the details of the observed perfor-mance in the cold vertical test and was decided on acavity-by-cavity basis Of these a standard HPR cyclewas by far the most applied (sim86 of the cases)approximately 9 underwent BCP followed by HPRand 120 degC bake while in a few cases (sim4) mechanicalgrinding [20] was required (in general after an opticalinspection [1819] see Sec V F) followed by furthersurface treatment as necessaryIn general a cavity whose performance in the cold

vertical test was considered unacceptable (ie the usablegradient was below 20 MV=m) was first retreated with

FIG 20 The fractional numbers of cavities having ldquoasreceivedrdquo usable gradient in the indicated ranges for quarterlyproduction A clear improvement (reduction) of the number ofcavities with gradients less than 20 MV=m can be seen in thesecond half of production

FIG 21 ldquoAs receivedrdquo low-field Q0 performance (measured at4 MV=m) by delivery month (For an explanation of the plot seeFig 19 caption) The two red dashed horizontal lines indicatingthe average performance over the first and second halves of theproduction are indistinguishable indicating no average improve-ment in the latter half (The last bin contains all remaining cavitiesarriving at DESY after 1 November 2015)

11Not including retreatments associated with cavities returnedfrom CEA Saclay due to an NC during cryomodule stringassembly

12A small fraction (sim3) were treated at both DESY and thevendors


042004-16

HPR This proved to be very effective in increasingthe usable gradient in particular for those cavities whoseperformance was field-emission limited Figure 22shows the breakdown of the reasons for the first retreat-ment at DESY (16 of the total number of cavities inthe analysis see Fig 10) clearly showing that fieldemission dominates the reason for the performance-drivenretreatmentsIn general cavities were sent for BCP (Sec II B) if

after an additional HPR their usable gradient was stillbelow 20 MV=m Several of the cavities which repeat-edly resulted in an unacceptably low usable gradientwere sent for optical inspection in OBACHT in order toidentify the reason of the poor performance and sub-sequent remedial action was decided depending on theresultFigure 23 shows the fraction of cavities undergoing

subsequent retreatments at both DESY and the vendorsApproximately 15 of cavities required an initial perfor-mance-driven retreatment after the first vertical test

dropping to sim4 for a second retreatment Less than3 of the cavities required a third retreatmentFigure 24 shows the usable gradient before and after

retreatment for both HPR and BCP retreatments Figure 25shows the impact of just the HPR on the treated-cavityusable gradient distribution and yield The effectiveness ofHPR is clearly visible in both figures For cavities with

FIG 22 Breakdown of the reasons for the first cavity retreat-ment at DESY Field emission low Q and quench are perfor-mance related (total of 66) while leak and other arenonperformance driven (Total chart represents 16 of thecavities)

FIG 23 Fraction of cavities undergoing 1 2 and 3 retreat-ments (after the first cold vertical test) broken down by category(Note ldquo1strdquo does not include the retreatments done before the firstcold vertical test)

FIG 24 Usable gradient before and after a retreatment for thespecified retreatments Those points above the dashed diagonalrepresent improvement

FIG 25 Improvement in the usable gradient distributionafter the application of HPR To emphasize the effectivenessof HPR on low-performing cavities only those cavities forwhich the initial (ldquobeforerdquo) gradient was less than 20 MV=mare shown


042004-17

usable gradients below 20 MV=m (Fig 25) sim70 arerecovered to above 20 MV=m with just over half beingabove 26 MV=m The resulting average is 255 MV=mA similar improvement can be seen for the application ofBCP (although the statistics are much lower)The application of HPR also makes a small but

statistically significant improvement in the low-field Q0

(see Fig 26) It is interesting to note that this cannot bedirectly associated with field emission at these low fieldsand the result remains somewhat surprising One specula-tive possibility is that some fraction of the metallicconducting particulates assumed responsible for fieldemission at higher gradients are located on surface areaswith high magnetic field and can therefore influence theQ0 at lower fields A rough estimate suggests a smallfraction of a square millimeter of normal conductingmaterial would be sufficient to create the observed differ-ence in the average Q0Finally Fig 27 shows the breakdown in limiting

criterion of the usable gradient before and after retreatment(for those cavities having undergone retreatment) Thereduction in field emission is clearly evidentSeventeen cavities underwent HPR retreatment after they

were found to be limited by ldquobreakdownrdquo between 10 and

27 MV=m with no or low field emission in the initial testMost of these retreatments were done as a result of a NC ofeither the cavity or the test itself As a result the retreatmentaction included a partial or complete reassembly of theflanges before the final HPR treatment The averagemaximum gradient and low-gradient Q0 (measured at4 MV=m) of the initial test was eth176 51THORN MV=m andeth23 04THORN times 1010 respectively After the HPR retreatmentthe average maximum gradient was eth20 78THORN MV=mThe increase of gradient and gradient rms spread wasmainly caused by two cavities showing an improvement of14 and 21 MV=m respectively The Q0 value at 4 MV=mafter retreatment was eth25 03THORN times 1010 Neither resultrepresents a significant difference It can be concludedthat a HPR retreatment applied on breakdown-limitedcavities did not result in a significant improvement ofthe performance

E Final performance

Figure 28 and Table VI show the final (accepted)gradient distributions and statistics respectively forcavities used for string assembly which includes thosecavities having undergone one or more retreatments Theyield at 20 MV=m is not 100 as might be expected

FIG 27 Breakdown of limiting criteria for the usable gradient of cavities before (left) and after (right) retreatment Field emission isclearly significantly reduced

FIG 28 Final maximum and usable gradient distributions ofcavities accepted for string assembly (including retreatments)

FIG 26 Impact of HPR on the Q0 measured at 4 MV=m Theaverage increases from 21 times 1010 to 24 times 1010 (sim14)


042004-18

since 3 of the cavities were accepted with lowergradients13

Figure 29 and Table VII show the usable gradientdistributions and statistics respectively for the first andfinal (accepted) cold vertical tests (the latter correspondingto the usable gradient distribution in Fig 28) The impact ofselective retreatment of low-performance cavities is clearlyevident with the average usable gradient being increasedby approximately 2 MV=m and the yield at 20 MV=mbeing increased from 86 to 97 The rms is also reducedfrom 72 to 51 MV=m The breakdown of the limitingcriterion for the final accepted usable gradients is given inFig 30 where comparison with Fig 14 in Sec IV B clearlyshows that the overall field emission has been reducedFinally Fig 31 shows the distribution ofQ0 as measured

at 236 MV=m for the first and final (accepted) cold verticaltest As expected from the Q0 criterion for the usablegradient definition (ge1010) the low value tail below 1010 isreduced but not completely removed these cavities arepredominantly those with usable gradients between 20and 26 MV=m

F Performance comparison

In this section the gradient performance of the EuropeanXFEL cavities will be compared to available historicalTESLA-type cavity dataIn preparation for the European XFEL cavity production

about 45 nine-cell cavities produced for the FLASHaccelerator at DESY [47] have been surface treated testedand analyzed [48] following the two final surface treatmentstrategies adopted for the European XFEL series cavities[3] The mechanical production of the FLASH cavities tookplace at the same vendors EZ and RI (formerly AccelInstruments) as for the European XFEL cavity productionbut in partially different fabrication facilities The mainbulk EP (initial surface removal of 110 or 140 μm depend-ing on the choice of final surface treatment see Sec II B)was performed at either Accel Instruments DESY orHenkel Lohnpoliertechnik14 The final surface treatmentstook place exclusively at DESY The vertical tests have

been performed in the ldquoHall 3rdquo test facility which was notequipped with the second x-ray detector at the bottom ofthe cryostat as is the case for the vertical test setup inAMTF Most of the tests have been performed withoutHOM feedthroughs in CW mode compared to the ldquolongpulserdquo mode used in AMTFIn general the main conclusions described in [48] have

been confirmed by the European XFEL cavity productionFirst both surface treatment schemes final EP and

flash BCP have proven successful for series cavityproduction It is notable that the different schemes havebeen applied to cavities of both vendors in [48] (seeTable VIII)Second all average gradients ldquoas receivedrdquo during

European XFEL series production exceed the FLASHproduction by 2ndash45 MV=m clearly demonstrating thesuccessful knowledge transfer to industry and a stablewell-established industrial production cycle The loweraverage gradient for EZ cavities is caused by the recipe(flash BCP) and by some cavities showing an early quench(see Sec VA)Third the application of HPR as an effective retreatment

in the case of field emission (see Sec IV D) could beconvincingly confirmedFinally in [48] no difference for the gradient yield

ldquowith He tankrdquo and ldquowithout He tankrdquo was demonstratedAgain this was confirmed by the European XFEL series

TABLE VI Key statistics for accepted cavities (includingretreatments)

Maximum Usable

Average gradient MV=m 330 298Rms MV=m 48 51Yield ge 20 MV=m 98 97Yield ge 26 MV=m 94 79

FIG 29 Comparison of the usable gradient distributions for thefirst and final (accepted) test

TABLE VII Comparison of the key statistics for the first andfinal (accepted) tests

First test Accepted


13The gradient of these cavities could either not be improved orretreatment was not attempted due to schedule constraints

14For the European XFEL series production no cavities under-went bulk EP at Henkel Lohnpoliertechnik


042004-19

production but a difference in the quality factor at4 MV=m was observed (see Sec V F)Another interesting comparison is with the maximum-

gradient yield data published in the technical design reportfor the International Linear Collider (ILC) project [49]which included 52 cavities tests from 2006 to 2012performed at Cornell University DESY FermilabNational Accelerator Laboratory (FNAL) ThomasJefferson National Accelerator Laboratory (TJNAL) andHigh Energy Accelerator Research Organisation (KEK) In

this case we can directly compare the RI cavity productionfor European XFEL as the ILC base line preparation recipe(first pass) is mainly identical to the European XFEL finalEP treatment (adopted by RI) Figure 32 shows a clearimprovement for ldquoas receivedrdquo European XFEL seriesproduction compared to the first-pass ILC data for the28 MV=m maximum gradient yield which can be inter-preted as the benefit of an established industrial seriescavity production However the yield at 35 MV=m maxi-mum gradient remains essentially unchanged (The conceptof ldquousable gradientrdquo was not formally defined for the ILCand therefore no direct comparison is possible)

V DETAILED CAVITY PERFORMANCE STUDIES

A Vendor differences and impact of final surfacetreatment recipe

In Sec IVonly the total cavity production was consideredand no distinction between the two vendors was made Inthis section first the differences of the cavity performancebetween the two vendors is analyzed while in the second

FIG 30 Breakdown of usable gradient limiting criterion forcavities accepted for string assembly (including retreatments)

FIG 31 Comparison of the distribution of Q0 measured at236 MV=m for the first and final (accepted) tests In both casesvery few cavities were below the nominal specification of 1010

TABLE VIII Comparison of the maximum and usable gradients (average rms in MV=m) for the European XFEL series productionand FLASH cavities

European XFEL production FLASH production

As received Tests Max gradient Usable gradient Tests Max gradient Usable gradient

Total 743 314 68 277 72 43 267 6 237 62EZ (flash BCP) 368 298 66 263 68 12 262 41 243 38EZ (final EP) 13 238 63 204 51RI (flash BCP) 9 289 43 253 66RI (final EP) 375 330 66 290 74 9 303 73 257 84

FIG 32 Comparison of yield data for the first-pass maximumgradient taken from the ILC technical design report and theEuropean XFEL ldquoas receivedrdquo final EP treatment The EuropeanXFEL data points were manually added to the original figuretaken from [49] using the same error calculation formulasNumbers in parentheses refer to cavity sample size


042004-20

part an analysis of the impact of the choice of final surfacetreatment (flash BCP versus final EP) is presented

1 ldquoAs receivedrdquo vendor dependence

In general there was a small but statistically significantdifference in the average cavity performance between RIand EZ cavities Figures 33 and 34 show the vendorperformance comparison for ldquoas receivedrdquo maximum andusable gradient respectivelyIn both cases cavities manufactured by RI showed an

average gradient approximately 10 higher than EZ withthe average maximum (usable) gradient for RI and EZbeing 330 MV=m (290 MV=m) and 298 MV=m(264 MV=m) respectively The dominant reason wasassumed to be the application of final EP by RI as thelast surface preparation step as opposed to the flash BCPused by EZ (Sec II B) which is known to affect the so-called high-field Q slope typically seen in gradients abovesim25 MV=m (see Chap 5 in [40]) This is discussed in more

detail in the next section The average EZ performance wasalso further slightly reduced compared to RI by the highernumber of low performing cavities (below 20 MV=m)which is likely due to differing fabrication techniquesbetween the two vendorsFigure 35 shows the low-field Q0 (measured at

4 MV=m) ldquoas receivedrdquo distributions for both vendorsThe mean values are 21 times 1010 and 22 times 1010 for RI andEZ respectively a small but statistically significant highervalue for EZ cavities by sim5

2 Flash BCP versus final EP surface treatment

It is well known that BCP tends to produce a relativelylarge high-field Q slope compared to EP (Chap 5 in[40]) For the cavity production the application of flashBCP by EZ as the final surface treatment manifests itselfin a lower Q0 performance at high gradients effectivelyreducing their usable gradient (due to the Q0 ge 1010

acceptance threshold) as well as their maximum achiev-able gradient (due to the 200 W power limit in the coldvertical test)In order to study this effect statistically for the produc-

tion cavities in an unbiased fashion a subset of cavitieswith no known ldquononconformancerdquo to the standard produc-tion procedure was used Typically rejected cavities werethose having clearly identified surface defects or someother nonconformance during fabrication In addition thefinal ldquoacceptedrdquo test was taken (as opposed to the ldquoasreceivedrdquo test see Sec IV E) in order to reduce theinfluence of field emission Figure 36 plots for each vendorthe recordedQ0 value at the maximum achieved gradient inthe vertical testThe clustering of EZ cavities (orange points) at lowerQ0

values on the 200 W constant forward power (dashed gray)line is an indication of the stronger Q0 slope (in particularwhen compared to the predominantly higher Q0 at the

FIG 33 Distribution and yield of the ldquoas receivedrdquo maximumgradient separated by vendor

FIG 34 Distribution and yield of the ldquoas receivedrdquo usablegradient separated by vendor

FIG 35 Distribution and yield of the ldquoas receivedrdquo low-gradient Q0 (measured at 4 MV=m) separated by vendor


042004-21

lower gradients) By comparison the RI data (blue points)do not show such a strong dependency and in general go tohigher gradients before reaching the power limit Note thatdata points above the dashed line are in general limited byquenchIn order to further quantify the difference in high-field Q

slope between EZ and RI cavities the following simplead hoc figure of merit (s) was employed

s frac14 Q0eth25 MV=mTHORN minusQ0ethEmaxTHORNEmax minus 25 MV=m

where Emax is the maximum field achieved during thepower-rise measurement Figure 37 shows the resultingdistributions for RI and EZ cavities The strong Q slope forflash BCP EZ cavities is clearly evident The average valueof s is ethminus 41 01THORN times 108=MVmminus1 and ethminus 77 02THORN times108=MVmminus1 for RI and EZ cavities respectively(mean σ=

ffiffiffi

np

)Although Fig 37 clearly indicates the larger high-fieldQ

slope for EZ (flash BCP) cavities on average it is importantto note that there are some RI cavities (final EP) that showequally large Q slope as can be seen by the long tail in theblue histogram Equally there are examples of EZ cavitiesshowing no or little high-field Q slope More detailedanalysis is merited requiring more sophisticated fits basedon physical models to further quantify this behavior

B Impact of niobium material

The niobium material for the accelerating cells of thecavities was delivered by Tokyo Denkai Ningxia OTIC andSE Plansee The final performance for 816 cavities acceptedfor string assembly was analyzed with respect to the materialvendor (Fig 38) There was no observed difference betweencavities fabricated from Tokyo Denkai or Ningxia OTICmaterial (the dominant number of cavities) However the174 cavities manufactured from SE Plansee material showeda small but statistically significant higher average value forboth the maximum gradient (Emax) and Q0eth4 MV=mTHORN Asall delivered material batches from all three material vendorsfulfilled the specifications no obvious explanation can begiven The observed behavior will be the topic of a furtherdetailed analysis

C Impact of infrastructure

In the AMTF hall two test cryostats (XATC 1=2) wereavailable both capable of taking one of six vertical testinserts each carrying at most four cavities (see Sec III C 2)Figure 39 shows the top view of the test cryostats indicating

FIG 36 Q0 at the maximum achieved gradient for ldquostandardrdquoaccepted cavities separated by vendor The gray dashed curverepresents the 200 W power limit in the cold vertical test

FIG 37 The distribution of the high-gradient (ge25 MV=m)linear Q slope expressed by the figure of merit (s) of themeasured Q0ethETHORN curves separated by vendor

FIG 38 Comparison of average maximum gradient (left) andaverage Q0 at 4 MV=m (right) of 814 cavities accepted for stringassembly separated by the vendors of the cell material The barsgive the 95 confidence level


042004-22

the numbered locations of the cavities which could not berotatedA statistical analysis was performed in order to exclude

any systematic differences between the test infrastructure(test cryostat insert number and cavity location in thecryostat) on the cavity performance notably the usablegradient (Eusable) and Q0eth4 MV=mTHORNWhile no significant difference could be seen between

the two test cryostat studies of the six individual insertswere found to produce variations in both the Eusable andQ0eth4 MV=mTHORN Figure 40 shows the example of the averageQ0eth4 MV=mTHORN for each of the test insertsInserts 1 and 6 showed statistically significant lower and

higher average values of Q0eth4 MV=mTHORN respectively sug-gesting a slight bias for both by approximately 6 Todate no possible cause for this difference has beenidentified Analysis of the historical usage of the insertsover the three-year testing period showed no correlationwith cavity performance trends The observed insertdependency cannot be explained by individual cables atthe inserts since each cable is calibrated for each test Theobserved fluctuations of the cable damping are understood

quantified and well below the magnitude of the observedeffect [41] One possible explanation of the Q0eth4 MV=mTHORNeffect is the detailed magnetic environment of each indi-vidual insert which still has to be checked especially undercryogenic conditions (see Sec III C 3 for a description ofroom temperature magnetic measurements)A similar study has been made with respect to the cavity

position in the cryostat Figure 41 shows the results forQ0eth4 MV=mTHORN In this case (and also in the usable gradient)a much clearer bias between positions 1 and 2 and positions3 and 4 can be seen (approximately 8 difference) Onepossible explanation lies in the test stand geometry withrespect to the rf connections The rf input power circuit wasfirst split in a so-called relay box housing three relays (twooutputs per relay) distributing power to positions 1=2 and3=4 after which the circuit was further split by twoadditional relays to each of the individual positions (seeFig 39) Though this is the only known systematicdifference with respect to positions 1=2 versus positions3=4 it seems unlikely to be the reason for the observed biassince each insert had its own independent relay box placedon the top plate requiring the cause to be systematic acrossall six inserts Moreover all typical rf checks and calibra-tions performed for the cold vertical tests showed nosuspicious behavior nevertheless a detailed hardware studyof the systematics will be made in the near future Inaddition the above mentioned additional checks of themagnetic environment will naturally include a possibledependence on the cavity positionThe typical observed maximum spread in the average

values is sim15 MV=m and sim04 times 1010 for Eusable andQ0eth4 MV=mTHORN respectively Practically speaking such smallsystematic shifts in the results are not particularly relevant forthe overall cavity productionNonetheless for the statisticallysignificant outliers discussed above it is important for futuremeasurements to identify the cause of the apparent systematicbias associated with the test infrastructure and further hard-ware tests and measurements are planned

FIG 39 Schematic top view of both test cryostats (XATC1=2)showing the cavity positions inside the stands and the rf inputpower distribution The orientation of the insert in the cryostatwas always the same it cannot be rotated The insert structure isnot visible

FIG 40 Mean values of the quality factor measured at4 MV=m [Q0eth4 MV=mTHORN] for all inserts The error bars representσ=

ffiffiffi

np

FIG 41 Mean values of the Q0eth4 MV=mTHORN distributions overthe four possible insert positions (cf Fig 39) The error barsrepresent σ=

ffiffiffi

np


042004-23

D High-pressure rinse analysis

A crucial part of cavity preparation was the HPR whereultrapure water with a pressure of 95 to 105 bar was used toclean the inner surface of the cavities This step was appliedseveral times and with different durations during theEuropean XFEL cavity fabrication according to therecipes for both surface preparation methods as describedin Sec II B The success of recovering low-performancecavities by an additional HPR only (see Sec IV D) raisedthe question if whether or not more HPR cycles duringproduction would systematically improve the cavityperformanceThe rf performance values Emax Eusable and

Q0eth4 MV=mTHORN were statistically analyzed based on whichof the two HPR stands were used at each of the vendors(infrastructure impact) and then by the number of HPRcycles applied (ie duration of HPR) at both DESYand thevendors Variations of the duration of HPR were caused bynonconformities during the process of cavity surfacepreparation (eg failed leak checks followed by reassemblyand re-HPR nonconformities during chemical treatmentsand nonconformities during an HPR cycle itself) or theincoming inspection (see Sec III C 1)For all 831 cavities no correlation with performance

could be identified with respect to the HPR stand or thenumber of cycles Restricting the analysis to higher-performance cavities with no field emission likewise showedno correlation We therefore conclude that variation inapplication of HPR beyond the standard specified procedurehad no influence on the performance of the cavitiesTogether with the success of HPR as a highly effectiveretreatment process especially for field emission limitedcavities we conclude that the source of field emission lies inthe handling steps following the final HPR cycle (ieassembly of the beam tube flange vacuum operations ortransport see Sec V F) Furthermore the analysis indicatesthat the specified standard number and duration of HPRcycles during the surface preparation is adequate

E Degradation and processing

ldquoProcessingrdquo and ldquoturn-onrdquo of field emission are well-known effects during vertical testing (see for exampleChap 122 in [40]) Turn-on of field emission results in adecrease of the Q value the maximum gradient or bothtogether with an enhanced radiation level If the resultingfield emission cannot be ldquoprocessedrdquo away by applyinghigher rf power it will lead to a permanent reduction(ldquodegradationrdquo) of the cavity performanceFor all vertical acceptance tests no administrative gradient

limitation was applied up to the maximum rf input power of200W (see Sec III C 2) Very few tests with strong radiationwere limited by the radiation protection system installedoutside of the cryostats Moreover no dedicated processingprocedures like helium processing or high peak powerprocessing (see Chaps 1351 and 137 in [40] respectively)

were applied During cold vertical tests the power risemeasurements [Q0ethETHORN curves] were repeated until two stableidentical curves were obtained Hence the comparison offirst and last power rise provides a handle to judge thebehavior of the cavity during testing The change in gradient(Δ) in the following analysis was defined as the usablegradient of last power rise minus the usable gradient of firstpower rise Thus a positive value represents processingduring the test cycle while a negative delta indicates aso-called degradation In the following only values of jΔj ge10 are considered relevant degradation or processingPower rises which were limited due to technical reasons

and not the cavity itself were excluded from the analysisFurthermore a few pathological cavities showing lowgradient behavior were also excluded This resulted in adata set of 1100 usable vertical tests for use in the statisticalanalysis An overview of the observed relative Δ versus theusable gradient measured in the first power rise can be seenin Fig 42Of the vertical tests shown in Fig 42 89 (81) showed

a degradation while 88 (80) showed a larger usablegradient in the last as compared to the first power riseTable IX gives a summary of all vertical tests and the

corresponding numbers for degradation and processing asdefined above The first data column shows the results forall tests while the second and third columns distinguishbetween cavities that achieved either below or above31 MV=m (the operational limit in the linac and in theAMTF cryomodule test stand [50]) in the first power risemdasha likely choice of administrative limit in the cold verticaltest should one have been applied While there is a smallloss of gradient for cavities going above 31 MV=m due toa few cases of degradation this can be practically ignored

FIG 42 Relative change of usable gradient (Δ) with respect tothe usable gradient as determined from the first power riseVertical tests with degradation or processing (jΔj gt 10) areshown outside the horizontal dashed lines Note the 1 MV=mbinning of the usable gradient causes overlaying data pointsMore than half of the vertical tests (573 out of 1101) showed nodifference in the usable gradient within the vertical test (repre-sented by the many overlaying dots at Δ frac14 0)


042004-24

It should also be noted that if the tests had been admin-istratively limited at 31 MV=m then the knowledge of thetrue performance limits of over one-third of the cavitieswould not have been known and the average performancereported in Sec IV would have been significantly less Thechoice not to impose an administrative limit in the verticaltest thus appears to be justified

F Special topics

1 Cavity performancemdashproduction related issues

Beyond the chemical surface preparation the cavityproduction process included a number of additional proce-dures which could have possibly influenced the cavityperformance Specifically four key procedures were identi-fied for further study titanium2-phase pipe repair [51] cavitytransport mechanical surface repairs (grinding) and heliumtank welding These four studies are briefly reported below

Impact of titanium 2-phase pipe repairmdashThe longitudinalwelded titanium 2-phase pipe is a part of the He tank andhas to fulfill the requirement of the pressure equipmentdirective as well as internal DESY standards As describedin detail in [51] pores discovered in the weld were out oftolerance and more than 750 cavities had to be repairedincluding some fraction that had already received a cold

vertical test In most cases the end parts of the 2-phase pipewere cut and replaced It is worth noting that the repair wasdone with all flanges sealed and the cavity under vacuumTo check the possible impact of the repair on cavity

performance the modification was first made on one cavitywhose performance was then remeasured No degradationwas observed and the repair process was then applied to theremaining cavities A retest of all repaired cavities was notpossible due to schedule constraints but a random selectionof the repaired cavities was tested to provide some elementof quality control To make an accurate statement onlycavities having had the 2-phase pipe repair and no otherkind of surface treatment in between the tests were takeninto account (a subset yielding a total of 20 cavities)Table X gives the statistics for Emax Eusable andQ0eth4 MV=mTHORN The deviations in the averaged gradientsbetween the measurements before and after repair are lessthan 2 and are well within the range of the accuracy of themeasurement An increase in field emission was also notobserved Hence based on these sampled statistics it isassumed that there was no significant impact on any of thecavities which underwent the repair

Impact of horizontal transport under vacuummdashThe dis-tributed sites for cavity surface treatment vertical accep-tance test and cryomodule string assembly made thereliable and cost-effective transport of the cavities ldquoreadyfor testrdquo (see Sec II B) an important part of cavityproduction Several investigations during the preparatoryphase were made to find a suitable shock absorbing systemfor the horizontal cavity transport in boxes [17] In thissection the impact of the long-distance cavity transports(vendors to DESY and DESY to CEA Saclay) by truck isdiscussed A total of 20 cavities with vertical tests beforeand after transport were identified (as above without anyadditional surface treatment) 12 cavities were transportedfrom DESY to and back from the module assembly site(CEA Saclay) and 8 cavities were sent between DESYandEZ The results are shown in Table XI The deviations in theaveraged gradientsmdashbefore and after shipmentmdashare less

TABLE IX Overview of degradation and processing statisticsfor all measured power-rise data (total) as well as those cavitiesachieving below or above 31 MV=m on the first power rise Theaverage Δ values were calculated over the corresponding numberof tests (Reported error is σ

ffiffiffi

np

)

Total le31 MV=m gt31 MV=m

Number of tests 1101 729 372Degradation 89 72 17Processing 88 86 2Average Δ (MV=m) minus01 008 013 011 minus055 008Degradation 81 98 46Processing 80 118 01

TABLE X Impact on the rf performance of 2-phase pipe repair [average (σ σ=ffiffiffi

np

)]

n frac14 20 Before repair After repair

Emax [MV=m] 353 (39 09) 347 (47 11)Eusable [MV=m] 308 (46 10) 305 (47 11)Q0eth4 MV=mTHORN 215 times 1010 (03 times 1010 006 times 1010) 217 times 1010 (03 times 1010 007 times 1010)

TABLE XI Impact on the rf performance of cavity transport [average (σ σ=ffiffiffi

np

)]

n frac14 20 Before shipment After shipment



042004-25

than 1 An increase in the field emission was also notobserved Based on these sampled results it was assumedthat transport had no influence on the average performanceof all series cavitiesImpact of grinding of surface defectsmdashInner surfacedefects were identified during cavity fabrication at thevendors as well as after cold vertical tests Repair by localgrinding took place at EZ [20] The quantities of affectedcavities are given in Table XIIThe cavity performance results are given in Table XIII

During cavity production different types of surface defectswere identified A detailed overview is given in [3] Most ofthese nonconformities were removed by a grinding pro-cedure as described in Sec II B Both Emax and Eusable areapproximately 3 MV=m higher than for all ldquoas receivedrdquoEZ cavities (29 MV=m and 264 MV=m respectively)The quality factor Q0eth4 MV=mTHORN is slightly increasedcompared to all ldquoas receivedrdquo EZ cavitiesIn the cases where the surface defect was identified

using OBACHT after a low-performance vertical test bothEmax and Eusable increased significantly by 9 MV=m and11 MV=m respectively indicating the effectiveness of therepair The results after grinding were in good agreementwith the average maximum and usable gradients for all ldquoasreceivedrdquo EZ cavities The Q0eth4 MV=mTHORN was unchangedwithin the standard error and is also slightly higher than theoverall average value of 21 times 1010As a remark it should be mentioned that in a few cases

applying the BCP retreatment process (see Sec II B)instead of grinding was sufficient to remove the surfacedefects On the other hand several cavities with defects onthe inner cavity surface were not treated since they fulfilledthe acceptance criteria (see Sec III B) Unfortunately a

small number of cavities with a specific type of surfacedefect called erosion could not be repaired These defectstypically on the welding seam were so deep that a repair bygrinding or BCP was not attemptedIn conclusion the above results indicate that mechanical

grinding was an extremely effective measure in recoveringpoor performing cavities with identified mechanicaldefects and that the results of grinding produced cavitieswith the same average performance as nongrindedcavities

Impact of He tank assemblymdashThe He tank welding processwas developed during the preparatory phase of theEuropean XFEL cavity production and included a clean-room compatible field profile measurement system (FMS)[5253] At that time and with limited statistics no differ-ence in the cavity performance ldquowithrdquo and ldquowithout Hetankrdquo was observed (see Sec IV F) Based on a bead-pull measurement the FMS allows the frequency andfield-profile monitoring during the various steps of theHe tank welding process while keeping the cavity closedand under clean conditions Nevertheless several additionalhandling and cleaning steps are necessary which hold thepotential risk of a contamination and hence performancedegradationTwo different analyses were made first a direct com-

parison ldquobeforerdquo versus ldquoafterrdquoHe tank assembly second astatistical comparison of the 23 HiGrade cavities withoutHe tank ldquoas receivedrdquo (see Sec II A) versus the 741 ldquoasreceivedrdquo series cavities The cavity performance results areshown in Table XIV As noted above a total of 23 cavitieshad two vertical tests with a standard tank assemblyprocedure in between and no surface treatment other thanthe HPR cycle applied after the tank welding procedure (ieafter the removal of the FMS) None of the rf parametersEmax Eusable or Q0eth4 MV=mTHORN showed any statisticallysignificant change Noteworthy is however the higher Qvalue in general for this subset of cavities (sim27 times 1010) ascompared to the overall average (214 times 1010) This maybe partially attributed to the fact that the sample includeseight HiGrade cavities subsequently equipped with a Hetank as well as seven cavities that had already undergoneretreatment

TABLE XII Cavities requiring repair of identified surfacedefects

Type of repairdefect Counts

Grinding during production 39Grinding after vertical test 13BCP after vertical test 6Erosion no repair 5

TABLE XIII Impact on the rf performance of surface mechanical grinding [average (σ σ=ffiffiffi

np

)]

Grinding during fabrication (n frac14 39)

Before grinding After grinding

Emax [MV=m] 323 (57 09)Eusable [MV=m] 296 (57 09)Q0 (4 MV=m) 230 times 1010 (04 times 1010 007 times 1010)After vertical test (n frac14 13)Emax [MV=m] 207 (50 14) 299 (80 23)Eusable [MV=m] 163 (26 08) 270 (76 22)Q0 (4 MV=m) 225 times 1010 (04 times 1010 010 times 1010) 238 times 1010 (04 times 1010 013 times 1010)


042004-26

To further confirm the benign effect of tank weldingobserved from the direct before and after analysis the ldquoasreceivedrdquo performance of 23 ILC HiGrade cavities (withouttank) and the series cavities (741 with tank) were sta-tistically compared While the average Emax is unchangedthe observed average difference in Q0eth4 MV=mTHORN issim02 times 1010 which would indicate a weak negative influ-ence of the He tank assembly on the low-gradient Q valueIn addition statistically weak differences between thevendors have been observed Despite these small effectswe conclude that the He tank assembly procedure had littleor no influence on the cavity performance

2 Estimate of residual surface resistance

For a better production characterization the Q0ethETHORNperformance of 24 cavities was measured at 18 K inaddition to the standard 2 K measurements made during thecold vertical test With only two measurement points thedeconvolution of the surface resistance into BCS (RBCS)and residual (Rres) parts (see Chap 31 in [30]) is notpossible Assuming a low BCS surface resistance ofeth8ndash10THORN nΩ at 2 K for a cavity after a 120 degC bake(confirmed by earlier measurements) and a residual resis-tance of about 4 nΩ [equivalent to RBCS (18 K)] a ratio of

the observed quality factors of about 15 is expected This isin good agreement with the observed ratio of 14 02found on the 24 analyzed cavities (Fig 43) Furthermorethe above assumptions fit well to the observed average ldquoasreceivedrdquo Q0eth4 MV=mTHORN of 21 times 1010 corresponding to asurface resistance of sim13 nΩ Together with five cavities(infrastructure commissioning and HiGrade) where theresidual resistance could be fitted to eth51 07THORN nΩ fromthe temperature dependent surface resistance measurementan average residual resistance of eth4ndash6THORN nΩ has beenestimated

3 Analysis of cool-down procedure

It has been reported in [5455] that the cool-downdynamics across the critical temperature Tc has a significantinfluence on the observed Q0 which is most likely due totrapped flux [56] We investigated if such a correlation couldbe observed during the European XFEL cavity productionAs mentioned in Sec III C 4 only one insert was

equipped with Cernox temperature sensors which alloweda direct measurement of the helium temperature profile atthe cavity positions This instrumentation was only used forcommissioning For regular operations TVO temperaturesensors were glued to the outside of the cryostat Thisinstrumentation was designed specifically for operation andcontrol of the cryogenic system only and cold vertical testswere performed in stable cryogenic conditions The analy-sis of the available data showed that a time-resolved cool-down rate across the cavities could not be deducedaccurately enough to make any statements about theinfluence of cool down [57]In passing we note that hydrogen Q disease was not

observed in any of the cold vertical tests within the givenmeasurement uncertainties

VI SUMMARY

The production of over 800 cavities for the EuropeanXFEL by industry and the associated cold vertical testing atDESY has been a remarkable success The total productionwas divided equally between E Zanon Spa (EZ) Italy andResearch Instruments GmbH (RI) Germany The cavities

TABLE XIV Impact on the rf performance of helium tank assembly [average (σ σ=ffiffiffi

np

)] (See text for details)

Direct comparison (n frac14 23)

Before tank assembly After tank assembly

Emax [MV=m] 318 (41 09) 325 (35 08)Eusable [MV=m] 297 (53 11) 304 (44 09)Q0 (4 MV=m) 269 times 1010 (04 times 1010 009 times 1010) 252 times 1010 (04 times 1010 008 times 1010)HiGrade comparison (ldquoas receivedrdquo)

HiGrade (w=o tank n frac14 23) Series (w=tank n frac14 741)Emax [MV=m] 314 (82 17) 314 (68 03)Eusable [MV=m] 295 (86 18) 277 (72 03)Q0 (4 MV=m) 238 times 1010 (04 times 1010 008 times 1010) 214 times 1010 (04 times 1010 001 times 1010)

FIG 43 Ratios of Q0 measured at 18 K and 2 K at differentaccelerating gradients The average ratio is 14 02


042004-27

sent for cryomodule assembly at CEA Saclay achieved anaverage maximum gradient of approximately 33 MV=mreducing to sim30 MV=m when the operational specificationson quality factor (Q0) and field emission were included (theso-called usable gradient) Only 16 of the cavities requiredan additional surface retreatment to recover their lowperformance (usable gradient less than 20 MV=m) Thesecavities were predominantly limited by excessive fieldemission for which a simple high pressure water rinsewas sufficient Approximately 16 of the cavities alsoreceived an additional HPR eg due to vacuum problemsbefore or during the tests or other reasons but these were notdirectly related to gradient performanceThe successful achievement of the overall high perfor-

mance and unprecedented production rate (a total of eightcavities per week) was due in part to a careful transfer toboth vendors of the decades of knowledge at both DESYand INFNMilano Both vendors needed to expand existinginfrastructure as well as develop new installations to meetthe stringent demands of the cavity fabrication and chemi-cal surface treatment at the required delivery rates Carefulautomated workflow extensive QC and documentationwas an essential part of the build to print process includingthe numerous inspections and measurements Furthermorethe process would not have been successful without theclose collaboration of industry and the lab personnel whichcontinued throughout the entire production periodThe team from IFJ-PAN Krakow successfully performed

over 1300 cold vertical tests of the cavities (includingrepeated tests) at the purpose built Accelerator Module TestFacility at DESY Average testing rates exceeded 10cavities per week peaking at up to 15 cavities per weekHeavy automation of the test procedures was essential toachieve the test rates as well as maintaining consistency ofthe tests themselves and the associated documentation anddata storageThe in-depth statistical analyses presented in this report

have revealed several features of the series producedcavities The overall production quality from the vendorsincreased during the second half of production Theapplication of an additional HPR proved extremely effec-tive for the low-gradient performing cavities (mostly field-emission limited) recovering nearly 80 of these cavitiesto gradients higher than 20 MV=m Although both vendorsexceeded the specifications cavities produced by RIperformed 10 better on average in gradient than thosefrom EZ which has been demonstrated to be predomi-nantly due to the fundamental difference in high-field Qslope between the final surface treatments applied (final EPand flash BCP respectively) In general the averageperformance of the European XFEL cavities has slightlyimproved on the historical TESLA-type cavity performance(FLASH cavities and ILC) taken over the last decadeDespite the success there remain questions and

ldquolessons learnedrdquo for future projects Although the

average performance is impressive the large spread inthe observed results (ranging from 10 to 40 MV=m) is stilla strong indication that the production could be improvedThe large spread (in particular the low-gradient tail)coupled with the high average was the driving reasonwhy the acceptance threshold for retreatment wasreduced from 26 to 20 MV=m early in the productionUnderstanding the relevant factors affecting the perfor-mance of individual cavities which essentially underwentthe same process will be an important part of futurestudies using the large legacy of data from the seriesproduction Furthermore systematic studies of the differ-ence of the test infrastructure have revealed small (lt10)but statistically significant bias in the cavity performanceresults These differences together with a better analysisof the overall uncertainties and measurement errorsrequire further studyFinally the European XFEL experience is already

proving to be of great benefit to future projects such asLinac Coherent Light Source II (LCLSII) [58] currentlyunder construction at SLAC California USA and theEuropean Spallation Source (ESS) [59] under constructionat Lund Sweden Looking further into the possible futurethe European XFEL has laid a strong foundation for thepossible construction of the ILC [49] in Japan Until suchtime that the ILC is built the European XFEL will likelyremain the largest deployment of the TESLA technologyfor some time to come

ACKNOWLEDGMENTS

The primary focus of this paper is the analysis of thecold vertical tests for the series cavity production for theEuropean XFEL The authors of this paper were primarilyresponsible for these analyses but would like to empha-size that they are a small part of the very large cavityproduction and testing team who are too many to nameindividually here This team comprised of DESY (notablythe groups responsible for niobium material QAQCcavity production QAQC test infrastructure supportincluding rf cryogenics and controls the clean roomteams responsible for retreatment and finally documen-tation and database maintenance) INFN MilanondashLASA(cavity production QAQC) IFJ-PAN Krakow (cavitytesting) and last but not least the two cavity vendors RIResearch Instruments GmbH and Ettore Zanon SpaWithout exception all worked together in an extremelyprofessional and productive manner for more than threeyears to bring this unprecedented production to such asuccessful conclusion The authors would like to expresstheir hope that in presenting the analyses of these teststhey have managed to do justice to the remarkable work ofso many dedicated people This work was supported bythe respective funding agencies of the contributing insti-tutes for details please see httpwwwxfeleu


042004-28

APPENDIX DATABASE ANALYSIS FLAGS AND THEIR DEFINITIONS

Table XV summarizes the flags used to categorize the vertical tests in the database For the retreatment secondarycategories NCout and NCin refer to external or internal nonconformities respectivelyTable XVI provides a mapping from the detailed flags to the summary flags used in the analysis presented in the main

body of the report

TABLE XV Cold vertical test analysis flags

Primary Secondary Explanation

As received Generally the first vertical test after delivery of a cavity having had no unusualtreatment or handling

Retest rf problem Test repeated for issues with the rf system or resultsNC at Saclay Test performed after a cavity was returned from the cryomodule assembly plant

at CEA Saclay due to a nonconformitywarm-up cycle Retests related to specific cool-down RampDother Miscellaneous (generally minor) reasons for not falling into one of the above categories

Retreatment quench Cavity was sent for retreatment due to low breakdown gradientlow Q Cavity was sent for retreatment due to low Q0 performanceFE Cavity was sent for retreatment due to excessive field emissionNCout fibers Cavity underwent retreatment as a result of the observation of fibers (incoming inspection)NCoutmechanical

Cavity underwent retreatment due to the observation of external mechanical defects(incoming inspection)

NCout 3DT Cavity underwent retreatment as a result of nonconformity with the mechanical axismeasurements (incoming inspection)

NCout vacuum Cavity underwent retreatment as a result of vacuum nonconformityNCin defect Cavity underwent retreatment as a result of an identified internal surface irregularity

(optical inspection)NCin defectgrinded

Cavity underwent retreatment as a result of an identified internal surfaceirregularity which underwent local grinding (optical inspection)

NC at Saclay Cavity underwent retreatment as a result of a nonconformity at the cryomoduleassembly plant at CEA

other Miscellaneous (generally minor) reasons for not falling into one of the above categories

Other preliminaryacceptance

Cavity tested without helium tank at the request of the vendor(prior to formal acceptance testing)

Tank integrationfor HiGrade

Test of selected HiGrade cavities after they were integrated into helium tanksfor cryomodule assembly

infrastructurecommissioning

Vertical tests of specific cavities used solely for testing and commissioningof infrastructure

TABLE XVI Summary flags used during analysis

Analysis category Vertical test flags

Retreatment (performance) RetreatmentquenchRetreatmentlow QRetreatmentFE

Retreatment (other) RetreatmentNCout fibersRetreatmentNCout mechanicalRetreatmentNCout 3DTRetreatmentNCout vacuumRetreatmentNCin defectRetreatmentNCin defect grindedRetreatmentNC at SaclayRetreatmentother


042004-29

[1] M Altarelli et al DESY Report No 2006ndash097 2006[2] F Richard J R Schneider D Trines and A Wagner

arXivhep-ph0106314[3] W Singer et al Production of superconducting 13-GHz

cavities for the European X-ray Free Electron Laser PhysRev ST Accel Beams 19 092001 (2016)

[4] W Singer X Singer A Brinkmann J Iversen AMatheisen A Navitski Y Tamashevich P Michelatoand L Monaco Superconducting cavity material for theEuropean XFEL Supercond Sci Technol 28 085014(2015)

[5] Y Bozhko et al in Advances in Cryogenic Engi-neering Transactions of the Cryogenic EngineeringConference-CEC (AIP Publishing Spokane 2012)Vol 57 pp 1100ndash1107

[6] B Aune et al Superconducting TESLA cavities PhysRev ST Accel Beams 3 092001 (2000)

[7] J Iversen R Bandelmann G Kreps W-D Moeller DProch J K Sekutowicz and W Singer in Proceedings ofthe 25th International Linear Accelerator ConferenceLINAC-2010 Tsukuba Japan (KEK Tsukuba Japan2010) p 794

[8] W Singer J Iversen A Matheisen H Weise and PMichelato in 16th International Conference on RF Super-conductivity (SRF2013) Paris France (JACoW Geneva2013) p 18

[9] J Iversen J A Dammann A Matheisenand N Steinhau-Kuehl in 17th International Conference on RF Super-conductivity (SRF2015) Whistler Canada (JACoWGeneva 2015) p 1154

[10] A Sulimov in 17th International Conference on RFSuperconductivity (SRF2015) Whistler Canada (JACoWGeneva 2015) p 955

[11] E Elsen in Proceedings of Linear Collider Workshop2011 (LCWS2011 Granada Spain (2011) (unpublished)

[12] Website of the ILC-Higrade project wwwilc‑higradeeu[13] A Matheisen Bvd Horst N Krupka M Schalwat M

Schmoekel A Schmidt W Singer P Michelato LMonaco and M Pekeler in 16th International Conferenceon RF Superconductivity (SRF2013) Paris France (JA-CoW Geneva 2013) p 201

[14] A Matheisen Bvd Horst N Krupka M Schalwat MSchmoekel A Schmidt W Singer P Michelato LMonaco and G Corniani in 16th InternationalConference on RF Superconductivity (SRF2013) ParisFrance (JACoW Geneva 2013) p 547

[15] S Aderhold S Chel E Elsen F Eozenou L Lilje and DReschke ILC HiGrade Report No 2010-005 2010 wwwilc‑higradeeu

[16] A Sulimov M Giaretta A Gresele and A Visentin inProceedings of LINAC Conference (LINAC2016) EastLansing Michigan USA (JACoW Geneva 2016) (to bepublished)

[17] A Schmidt A Matheisen R Bandelmann H Brueck andJ Schaffran in 15th International Conference on RFSuperconductivity (SRF2011) Chicago USA (JACoWGeneva 2011) p 504

[18] A Navitski S Aderhold U Cornett E Elsen G FalleyM Lemke J Schaffran L Steder H Steel and

Y Tamashevich Cavity inspection system OBACHT(unpublished)

[19] M Wenskat PhD thesis DESY Universitaumlt Hamburg2015

[20] G Massaro N Margano G Corniani A MatheisenA Navitski P Michelato and L Monaco in 17thInternational Conference on RF Superconductivity(SRF2015) Whistler Canada (JACoW Geneva 2015)p 368

[21] B Petersen in 6th International Workshop on CryogenicOperation Daresbury Great Britain (2014) (unpublished)

[22] J Iversen A Brinkmann J Dammann A PoerschmannW Singer and J-H Thie in 16th InternationalConference on RF Superconductivity (SRF2013) ParisFrance (JACoW Geneva 2013) p 183

[23] J Schaffran Y Bozhko P Petersen D Meissner MChorowski and J Polinski Design Parameters andCommissioning of Vertical Inserts used for testing theXFEL Superconducting Cavities AIP Conf Proc 1573223 (2014)

[24] D Reschke in 16th International Conference on RFSuperconductivity (SRF2013) Paris France (JACoWGeneva 2013) p 812

[25] K Krzysik et al in 16th International Conference on RFSuperconductivity (SRF2013) Paris France (JACoWGeneva 2013) p 191


[27] A Sulimov A Ermakov J-H Thie and A Gresele in17th International Conference on RF Superconductivity(SRF2015) Whistler Canada (JACoW Geneva 2015)p 1277

[28] Y Tamashevich E Elsen and A Navitski in 17thInternational Conference on RF Superconductivity(SRF2015) Whistler Canada (JACoW Geneva 2015)p 774

[29] R Laasch in TESLA Technology Collaboration MeetingHamburg Germany (2014) (unpublished)

[30] H Padamsee RF Superconductivity Science Technol-ogy and Applications (Wiley-VCH Verlag Weinheim2009)

[31] M Bolore B Bonin B Daillant O Delferriegraverre JFaure S Jaiumldane E Jacques H Safa and C ValletTESLA Report No 1994ndash23 httpflashdesydereports_publications

[32] H D Bruumlck and M Stolper TESLA Report No 1994ndash10httpflashdesydereports_publications

[33] Vacuumschmelze brochure PW-004 Cryopermreg 10(Vacuumschmelze GmbH Hanau 1992)

[34] Vacuumschmelze brochure PHT-001 WeichmagnetischeWerkstoffe und Halbzeuge (Vacuumschmelze GmbH amp CokG Hanau 2002)

[35] V Anashin L Belova Y Bozhko K Escherich BPetersen S Putselyk E Pyata T Schnautz JSwierblewski and A Zhirnov Experience with cryogenicoperation of Accelerator Module Test Facility duringtesting of one third of XFEL cryomodules IOP ConfSer Mat Sci Eng 101 012139 (2015)


042004-30

[36] V I Datskov and J G Weisend II Characteristics ofRussian carbon resistance (TVO) cryogenic thermometersCryogenics 34 425 (1994)

[37] U Hahn and K Zapfe Guidelines for UHV-Components atDESY Technical Specification Vacuum 0052008 httpsedmsdirectdesydeitemD00000001425601D12

[38] T Powers in 12th International Conference on RF Super-conductivity (SRF2005) New York (JACoW Geneva2005) p 40

[39] F He Uncertainty of data obtained in SRF cavity verticaltest arXiv13103900

[40] H Padamsee J Knobloch and T Hays RF Supercon-ductivity for Accelerators (John Wiley amp Sons IncNew York 1998)

[41] Y Yamamoto W-D Moeller and D Reschke in 7thInternational Particle Accelerator Conference (IPACrsquo16)Busan Korea (JACoW Geneva 2016) p 2128

[42] M Wiencek K Krzysik J Swierblewski and J Chodakin 17th International Conference on RF Superconductivity(SRF2015) Whistler Canada (JACoW Geneva 2015)p 910

[43] Website of the DESY European XFEL project grouphttpxfeldesydecavity_database

[44] P D Gall V Gubarev S Yasar A Sulimov and JIversen in 16th International Conference on RF Super-conductivity (SRF2013) Paris France (JACoW Geneva2013) p 205

[45] P D Gall et al in 17th International Conference on RFSuperconductivity (SRF2015) Whistler Canada (JACoWGeneva 2015) p 1168

[46] S Yasar P D Gall and V Gubarev in 17th InternationalConference on RF Superconductivity (SRF2015) WhistlerCanada (JACoW Geneva 2015) p 1166

[47] Website of the Free-electron laser FLASH at DESY flashdesyde

[48] D Reschke L Lilje and H Weise in 14th InternationalConference on RF Superconductivity (SRF2009) BerlinGermany (JACoW Geneva 2009) p 342

[49] C Adolphsen et al The International Linear ColliderTechnical Design ReportmdashVolume 3 I Accelerator RampDin the Technical Design Phase arXiv13066353

[50] K Kasprzak et al in Proceedings of the 6th InternationalParticle Accelerator Conference (IPAC2015) RichmondVirginia USA (JACoW Geneva 2015) p 2994

[51] M Schalwat A Matheisen A Daniel S Saegebarth PSchilling H Hintz K Jensch S Barbanotti and ASchmidt in 17th International Conference on RF Super-conductivity (SRF2015) Whistler Canada (JACoWGeneva 2015) p 1122

[52] A Schmidt A Matheisen H Kaiser W Menck and GWeichert in Proceedings of the 13th International Work-shop on RF Superconductivity (SRF2007) Beijing China(JACoW Geneva 2007) p 192

[53] A Schmidt G Kreps A Matheisen and H Weitkaemperin Proceedings of the 14th International Workshop onRF Superconductivity (SRF2009) Berlin Germany(JACoW Geneva 2009) p 665

[54] J-M Vogt O Kugeler and J Knobloch Impact of cool-down conditions at Tc on the superconducting RF cavityquality factor Phys Rev ST Accel Beams 16 102002(2013)

[55] A Romanenko A Grasselino O Melnychik and D ASergatskov Dependence of the residual surface resistanceof superconducting RF cavities on the cooling dynamicsaround Tc J Appl Phys 115 184903 (2014)

[56] T Kubo Flux trapping in superconducting acceleratingcavities during cooling down with a spatial temperaturegradient Prog Theor Exp Phys 2016 053G01(2016)

[57] M Wenskat and J Schaffran Analysis of the cool downrelated cavity performance of the European XFEL verticalacceptance tests arXiv160906503

[58] Website of the LCLS II project at SLACNational AcceleratorLaboratory httpsportalslacstanfordedusiteslcls_publiclcls_iipagesdefaultaspx

[59] Website of the European Spallation Source ESS httpseuropeanspallationsourcese


042004-31

to well-defined procedures and tolerances Providing thecavities were fabricated to these specifications DESYaccepted the cavity without any final cold rf performanceguarantee As a result DESY took the responsibility of anyremedial (postproduction) measures necessary to recoverpoor-performing cavitiesThe resulting average performance of sim30 MV=m with

an unloaded Q value Q0 ge 1010 of such a large industrialseries production represents an unprecedented success forthis SRF technologySuch a large scale production can only be successfully

completed by the work of a large team of experts in variousfields The responsibility for cavity production was sharedbetween INFN MilanondashLASA and DESY and includedabout 25 people in the core team Cold vertical testingof individual cavities as well as cryomodule tests wereperformed at DESY in the Accelerator Module Test Facility(AMTF) [5] by a team of 30 technicians engineers andphysicists from the Henryk Niewodniczański Institute ofNuclear Physics Polish Academy of Sciences (IFJ-PAN)Krakow supported by DESY expertsThis report focuses on the detailed analysis of the

resulting performance of the series production cavitiesThe main body of the report is separated into the followingfour sectionsSection II Cavity Production provides a brief overview

of themechanical fabrication and chemical surface treatmentof the cavities by industry and introduces somekey conceptsand terminology important for the remainder of the reportSection III Vertical Acceptance Test details the setup

and procedures for the cold vertical testing of the cavitiesin AMTF This section introduces the overall workflow andin particular the definition of the usable gradient andprovides details of the careful procedures and qualitychecksmdashquality assurance (QA) and qauality control(QC)mdashof the cavities during this testing period Theobserved nonconformities during QC and cold verticaltesting are also briefly discussedSection IV Production Performance Statistics provides a

comprehensive set of statistical analyses which quantify theoverall cavity production in terms of gradient Q0 andfield-emission performance Performance limitations andthe impact of the surface retreatments performed atDESYon (predominantly) low-performing cavities are alsoquantified and performance trends during the production

discussed Finally the results from the European XFELseries production are compared to existing historical dataSection V Detailed Cavity Performance Studies covers

the following special-topic studies differences betweenvendor performance impact of the infrastructure at both thevendors and at DESY studies on high pressure water rinse(HPR) application impact of special procedures such astransport helium-tank welding repair of the two-phasehelium pipe studies of process and field-emission onsetduring the vertical test and an estimate of the residualsurface resistance based on additional 18 K measurementsFinally Sec VI provides a summary and some conclud-

ing remarks

II CAVITY PRODUCTION

The cavity design for the European XFEL acceleratorcavity is based on the ldquoTESLA cavity designrdquo [6] as asuperconductingnine-cell cavitymadeof solid niobiumwiththeπ-mode frequencyof theTM010-passband at 1300MHzMinor modifications compared to the original design havebeen applied for a simpler fabrication and cost reduction [7]

A Cavity fabrication and material

Series production of the 13 GHz cavities was equallydivided between Ettore Zanon SpA and RI ResearchInstruments GmbH Production included both mechanicalfabrication and surface preparation [38] together with therequired extensive quality assurance quality managementand associated documentation [9] Details about theniobium and niobium-titanium material used can be foundin [4] The Nb material for the accelerating cells wasdelivered in differing amounts by three companies (TokyoDenkai1 Ningxia OTIC2 and SE Plansee3) The materialfrom each company was distributed equally to the twocavity vendors The rf measurements for quality assuranceduring the cavity production are described in [10]Table I provides a summary of the cavity production

numbers A total of 852 cavities were produced by thevendors (429 and 423 cavities by RI and EZ respectively)

TABLE I Total number of cavities produced by vendors

RI EZ Total Comment

Infrastructure setup and commissioning 8 8 16 4 modified to series cavitiesILC HiGrade 12 12 24 For QC and RampD 8 modified to series cavitiesSeries 400 400 800Rejected and replaced 6 2 8 Replaced at vendor cost included in total 800Additional new orders 3 1 4Totals 429 423 852

1Tokyo Denkai Co Ltd2Ningxia Orient Tantalum Industry Co Ltdhttpwwwnnieccom

3SE Plansee httpswwwplanseecom


042004-2















042004-3









042004-4

















042004-5














CompanyReturns to



RI 79 55 24EZ 63 36 27Total 142 91 51


042004-6














042004-7







5 Vacuum systems






042004-8
















042004-9




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31















042004-3









042004-4

















042004-5














CompanyReturns to



RI 79 55 24EZ 63 36 27Total 142 91 51


042004-6














042004-7







5 Vacuum systems






042004-8
















042004-9




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31









042004-4

















042004-5














CompanyReturns to



RI 79 55 24EZ 63 36 27Total 142 91 51


042004-6














042004-7







5 Vacuum systems






042004-8
















042004-9




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31

















042004-5














CompanyReturns to



RI 79 55 24EZ 63 36 27Total 142 91 51


042004-6














042004-7







5 Vacuum systems






042004-8
















042004-9




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31














CompanyReturns to



RI 79 55 24EZ 63 36 27Total 142 91 51


042004-6














042004-7







5 Vacuum systems






042004-8
















042004-9




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31














042004-7







5 Vacuum systems






042004-8
















042004-9




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31







5 Vacuum systems






042004-8
















042004-9




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31
















042004-9




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31




1 Vacuum problems













042004-10





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31





in the Appendix
















042004-11








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31








RI EZ Total Comment






042004-12











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31











Maximum Usable






042004-13












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31












042004-14












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31












042004-15














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31














042004-16











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31











042004-17







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31







E Final performance






042004-18














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31














Maximum Usable




First test Accepted





042004-19















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31















042004-20
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31
















042004-21





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31





ffiffiffi

np











042004-22










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31










ffiffiffi

np


ffiffiffi

np


042004-23














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31














042004-24


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31


F Special topics








ffiffiffi

np

)




np

)]




np

)]




042004-25













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31













np

)]





042004-26









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31









VI SUMMARY



np








042004-27








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31








ACKNOWLEDGMENTS



042004-28



body of the report

























042004-29






































042004-30


























042004-31



body of the report

























042004-29






































042004-30


























042004-31






































042004-30


























042004-31


























042004-31

Documents

Performance in the vertical test of the 832 nine-cell 1.3 ... · divided between Ettore Zanon S.p.A., and RI Research Instruments GmbH. Production included both mechanical fabrication