pcm test: preliminary results seminar... · 2017-01-18 · – main and hom couplers test results...

pCM test: Preliminary results

N.Solyak (on behalf LCLS-II team)

Outline

• pCM chronology• Minimum acceptance tests

– MP, dark current, FE and radiation processing– Cavity processing, gradient and heat load measurements– CM Unit Testing– Main and HOM Couplers test results

• Extended Tests in Progress• Summary

1/17/2017"LCLS-II pCM test" at PIP-II meeting, January 17, 20172

3"LCLS-II pCM test" at PIP-II meeting, January 17, 2017

LCLS‐II Cryomodule design

Based on XFEL design with many modifications to meet LCLS‐II reqs

Baseline design: • 35 – 1.3GHz CMs (+1)• 2 – 3.9 GHz CMs (+1)

Energy upgrade +4 1.3GHz CMs

2 pCM are completed and under tests, 2 ‐ assembly

4

CM details

"LCLS-II pCM test" at PIP-II meeting, January 17, 2017

pCM Chronology

– pCM delivered to CMTF ‐ 20 July– Installed into Cave – 24 July– Warm conditioning (1‐7) – 25 August– Cool down and first cold operation (Min Tests) – Sept.

• pCM started cavity testing at 1.9K briefly under cold compressor mode.

• pCM cavity testing was completed at 2.1 K under warm pumping mode (Kinney pump) in October 2016

– Cryo shutdown for upgrade – November (for 2K operation)– Cool down to 2K and start Extended tests ‐ end December


E. Harms

6

Line A, B and G Configuration


7

Cryogenic: Operating experience


The CMTF Cryogenic system has two methods to achieve 2K operation:• Cryoplant Cold and Warm Vacuum Compressor (2K Refrig. mode, 500W capacity).

Advs: JT heat exchanger in plant; Disadv: Min. CC flow, Less forgiving, • Cryoplant in Liquefier Mode with Kinney Vacuum Pump

Advs: More forgiving/flexible, Better turndown; Disadv: No JT HX high T

pCM Cryogenic Operating History– 2K Plant: Long recovery from CC trips (learning curve); Single cavity tests at 1.93K– 4.5K Plant and Kinney Vacuum Pumping

• Could only achieve 2.1K due to pressure drop and high flows resulting from the lack of a JT HX• Single cavity tests, coupler tests were performed in this mode

– Shutdown (Nov,2016):• Addressed Kinney header ∆P, added LN2 shield cooling to DB, changed CC drives, changed one control

valve, added thermal intercept to CM JT valve, added flow meter to Kinney discharge• Current mode at 2K. Some issues still open

– Apparent thermoacoustic oscillation (TAO) in CM JT valve• Added thermal intercept, investigating valve stem wiper

– High flow rates due to high JT valve inlet temperature (no JT HX and TAO)• Addressing TAO, investigating JT HX possibilities

– Liquid level stability in CM was challenging Jay Theilacker


Coupler and cavity diagnostics

Cavities 1, 4, 5 and 8 have additional diagnostics inside HV‐ 4 T‐sensors, cell 1&5‐ 2 magnetic fluxgates

Magnetic shielding• 9 T‐sensors outside HV• Fluxgates 1,2,5,7,8,

Minimum Acceptance Criteria - Production


✔

✔✔

✔

✔

✔

✔

✔✔✔

✔

✔

✔

✔✔

✔

✔

✔

Beamline vacuum ~5E‐10, unbaked spools

pCM + FeedCap + EndCap ~13 W Static 2K64 W 2K dynamic of 128 MV

Also specs for: • Q0 average > 2.5e10• Main coupler (FPC)

‐ QL tuning range: 1.e7‐6.e7 • HOM coupler: 1W leak at 16MV/m• Frequency tuner

slow/fast tuning range: 450kHz/500Hz

MP in HOM: heating response on HOM-can & HOM-feedthru


HOM can

HOM feedthru

HOM overheating due to MP (pulse)

Cav#1 Cav#5

• Sign of MP in HOM –overheating HOM can

•MP in cavities #1 & #5 @15MV/m (dwn HOM)

•After processing no MP in cav#1. Gradient improved to 17.2 MV/m, (19.6 MV/m in pulse mode).

• In cav#5 gradient improved (MP still there)

•HOM‐can sensors are more sensitive to MP than HOM‐feedthrough sensors

MP in HOM of cavity#1 (zoom)


Gradient

RF power

HOM can Temp.

Radiation

MP starts at ~15MV/m

HOM can T quickly goes up above threshold gradient

Processed in short pulses (seconds) until T< 20K

After processing no MP up to max gradient

Radiation reduced after processing,Exist w/o MP

Before processing processing After

processing

Field Emission and Dark Current / X-ray in cavity

12

X‐ray (mrem/hr)

RF power (W)

Dark CurrentPeak114nA

Peak20.5 nA

Peak34nA

• Cavity 2 & 3 tested, but not plotted here. No FE/ MP @16MV/m

•DC on cav#4 (pk 1&3) Processed away.

• X‐ray activity on cavities 5, 8; no DC

~16 MV/m

C7 C5 C6 C8 C1 C5

History plot


FE: Dark Current / X-ray on cav#4 before/after processing


FE processing

Cavity #4 processing

Dark current1nA/div

RF power

16MV/m

X‐ray

Pulse

mod

e

14

MP Processing (plot for CAV7)


• CAV2 was limited at 16.8 MV/m due to MP (15.6 MV/m FE onset).RF pulsed processing improved the gradient to 19 MV/m in pulsed mode. Quench after 19 MV/m (as VTS). Ran 18.8 MV/m stably.

• CAV3: Processing has little effect (~2 hrs). Gradient stable at 17 MV/m.

• CAV4: Processing 30min , 40ms, 2Hz (radiation). Gradient >20 MV/m, parked to 19.2 MV/m CW, small FE at this gradient.

• CAV6: limited 16.3 MV/m due to MP (14.5 MV/mFE onset). RF pulsed processing improved the gradient to >17 MV/m (same onset).

• CAV7: quenched at 17.4 MV/m, after pulsed processing 40 ms 2Hz, cavity went to 20 MV/m without pulsed quench. Work at 19.4MV/m (below 50 mR/h)

Eacc

Eacc

15

Cavity Limitations

Summary as of 2016‐10‐31 at 2.1K

Cavity Name

Eacc @<50mR/h [MV/m]

Max Gradient [MV/m] Gradient limitation

FE onset [MV/m] NOTE

1 TB9AES021 18.2 21.2 quench (cryo, admin limit) 14.6 Multipacting improved, <15min2 TB9AES019 18.8 19.0 quench 15.6 Multipacting improved, <15min

3 TB9AES026 17.2 17.3 Quench (Multipacting, cryo, no x‐ray) 17.3No improvement, ~2h, limited by cryogenics

4 TB9AES024 20.5 21.0 Quench (Multipacting, admin limit) 21.0 Multipacting improved, <30 min5 TB9AES028 14.2 14.9 Field Emission 13.96 TB9AES016 16.9 17.1 Quench (Multipacting, cryo) 14.5 Multipacting improved, <15 min7 TB9AES022 19.4 20.0 Quench (Multipacting, admin limit) 12.7 Multipacting improved, <15 min

8 TB9AES027 17.5 20.0 Quench (Multipacting, cryo) 20.0 Multipacting improved, <10 min,limited by vacuum interlock

TOTAL VOLTAGE 148.1 156.2 Average Eacc @FE onset = 16.2

CAV3, CAV6 and CAV8 all exceeds specification. Additional RF processing at 2.0K may further improve cavity gradients.


16

MP issue. Summary and lessons learned

Multipacting Determination• Consistent gradient limit (almost all cavities)• No or very low radiation (except CAV5 FE)• Quench field can be conditioned• No electron signals in FPC• Two cavities experienced HOM can heating

- One processed away (CAV1), One still present (CAV5 with high FE)• VTS experience was similarPossible Causes of MP in pCM• Non baked beam line components that release H2• Long period (5-6 month) of cavity filled with N2• Coupler outgassing after exposed to air brieflyMultipacting Mitigation• RF pulse conditioning (Successful in FNAL pCM)• Warm up and pump longer time • No coupler conditioning


17

Gradient and Q0 after MP Processing – in progress

*Usable Gradient: demonstrated to stably run CW, FE < 50 mR/h, no dark current

**Fast cooldown from 45K, >40 g/sec, extrapolated from 2.11K

***Also limited by administrative limit ~20MV/m


Genfa/Elvin


JLAB result

19

Field Emission

0.1

1

10

100

1000

0 5 10 15 20 25

Rad [m

R/h]

Eacc [MV/m]

CAV1 Rad#2

CAV2 Rad#N

CAV3 No Rad

CAV4 No Rad

CAV5 Rad#6

CAV6 Rad#6

CAV7 Rad#6

CAV8 No Rad

Average Eacc@FE‐onset >16 MV/m


Arrays of rad. detectors along the CM and ends.

20

Cryomodule Unit Testing (in progress)

All cavities powered to nominal gradient (#5 @12MV/m)All 3 magnet coils powered to 10A‐nominalRan continuously for ~45minLow radiation < 34 mR/h (not phase synchronized)No detectable dark current

Vacc = 128 MV (spec)


(10A spec)

Endurance test (10hrs) expected later this week

21

Unit heat load test at 16MV/m @2K (preliminary)


HeaterAll cavities @16MV/m, except cav5 ‐ 12 MV/m (Vacc = 128MV)

Heat calibration at 50WΔM = 5 [g/s];

Dynamic heat load ~70 W (~9W/cav) <Q0> ~ 3.e10

Next steps– long unit test (>10hrs)

22

Heat Load Measurement - Static

Measured at 4K; Total ~ 13 W.( pCM + FeedCap + EndCap)Will be repeated at 2K. Courtesy of R. Wang


Mass flow

He Liquid level

JT‐valve


Spec < 5mG

24

Fast Cool Down


>40 g/s 32 g/s

Cavitydelta T(K)

delta T(K)

CAV1 3.8 3CAV4 5.6 2CAV5 8.3 5CAV8 4.7 2.5

Fast cool down in a cryomodule demonstrated

pCM Test Review January 11, 2017

Slow cooldown to be tested

25

HOM and FP performance at FNAL pCM

AUTHOR - CM PRR Sep.13-14 2016

Cavity HOM1 Qext

HOM2 Qext

Field Probe Qext

HOM1 Power @16MV/m (W)

HOM2 Power @16MV/m (W)

TB9AES021 1.6e13 5.0e13 2.04e12 0.017 0.005TB9AES019 3.8e13 1.1e12 8.26e11 0.007 0.248TB9AES026 3.0e12 6.5e12 7.48e11 0.091 0.042TB9AES024 4.5e13 2.7e12 7.40e11 0.006 0.101TB9AES028 5.0e11 2.4e12 4.61e11 0.545 0.114TB9AES016 8.1e11 2.6e12 1.07e12 0.336 0.105TB9AES022 1.8e13 1.3e12 5.48e11 0.015 0.210TB9AES027 6.8e11 1.4e12 1.60e12 0.401 0.195

Specification: HOM Qext > 2.7e11 and 1 W maximumAll have < 1W @ 16MV/m At 20 MV/m, P_HOM ~ 1.56 W specification.

Tuner side Coupler side

HOM performance: Effect of RF losses in HOM cable

1/18/2017Presenter | Presentation Title26

•HOM power heat cable heating of HOM antenna & FT antenna quench•Expected HOM power from 0.3mA beam < 5W, system designed for 10 W.•Cav#8 HOM1 cable (coupler side) powered at 2GHz up to 5W, full reflection Ptot=10W

HOM FT

HOM1 can

Beam tube

Gradient 16MV/m

Test Summary•No antenna/FT overheating

•FT static T=5.3K (<50mW from cable)

•RF power in cable 5.1W; ΔTFT< 0.9K. ‐ Same ΔTFT for 16MV/m

•Temp of HOM can & beampipe is not sensitive to cable heating; ΔT<0.04K

•At 16MV/m: ΔTHOMcan~0.3K; ΔTbeampipe~ 0.4K (FPC power & Eacc)

Main coupler tests in pCM


• Room temp. power processing (couplers #1 thru #7): 2.5 to 4 kW• Coupler #8 was processed at 4kW with cavity OFF‐resonance (2.1K) • All couplers are tested at cold, P= 3.5 kW & 3.2 kW (~12‐20 hrs)

‐ QL =1.e7 (tuned to low Qext for low cavity gradient )‐ Cavity gradient 10‐12 MV/m‐ All couplers in pCM are TTF3 modified (different from production)

• Coupler #4 was retested at 2K (13hrs). From other couplers data are taken from cavity Q0 measurement runs (0.9 kW and 1.6kW)


Interlock=200K Interlock=250K

Extended test of the main coupler #4 (2K)


• Power 3.2kW• Duration ~13hrs

• TCT100 = 208 K

• Consistent with 2.1K result

RF power =3.2kW

Gradient=11.510 MV/m

TCT100

Summary of coupler heating

• Different thermal intercept configuration in HTS

Coupler#4

50K 5K


pCM

pCM

pCM

No air cooling in pCM

HOM coupler: static and dynamic heating


Conclusion from pCMmeasurements and simulation:

• Static heat load to 2K < 0.15 W‐ FPC ~ 50mW‐ HOM+FP cables < 100mW

• Dynamic heatload (2K) < 1.2 W‐ bellow < 150mW @16MV/m‐ FPC @3.5kW < 1W

HOM‐can HOM‐can

HOM‐feedthruHOM‐feedthru

Acnetname Location cavity Note

#1 #2 #3 #4 #5 #6 #7 #8CT116 CavBeamTube (coupler) 2.51 2.38 2.29 2.38 2.63 2.32 2.38 2.81CT121 HOMcan(coupler) 2.41 2.34 2.26 2.31 2.30 2.25 2.31 2.87CT120 HOMfeedthru (coupler) 6.43 5.35 5.02 5.29 4.97 5.28 5.90 5.30 Heatedby2.5mcable

CT117 CavBeamTube (tuner) 5.01 5.47 2.85 2.84 10.73 4.64 2.71 2.35CT119 HOMcan (tuner) 2.68 2.24 2.31 2.24 2.83 2.38 2.20 2.22CT118 HOMfeedthru (Tuner) 4.30 4.59 4.43 4.05 3.96 4.04 4.26 4.35 Heatedby3mcable

Static temperatures (no RF power)

FPC&HOM Couplers 2K Heat Load (CTS/COMSOL analysis of pCM)

34 1/18/2017Andrei Lunin | LCLS-II CM01 Cavity Thermal Analysis

T‐map (QFPC = 1.2 W)

Temperature Profile along the 5K Tube

QFPC = 1.2 W

QFPC = 0.8 W

QFPC = 0.4 W

QFPC = 0.2 W

QFPC = 0.05 W

QFPC = 0.1 W

Static heat load

Dynamic heat load

P=55 mW@10MV/m

A.LuninTotal 2K Heat Loads from FPC & HOM Couplers in the LCLS‐II pCM(based on T‐meas: FPC/HOM/BP)

35

Summary and Lessons Learned

Procurement CM Final Review, May 12-14, 2015

• Coupler warm conditioning may causes cavity surface contamination (hydrogen) MP

• Coupler has acceptable 2K heat load (<1W @ 3.5kW), even at relatively high temperature of the 50K/5K intercepts (~200K/20K). Spec 150K @50K

• small Q0 degradation due to FPC overheating (very preliminary, to be tested in pCM extended tests).

• FPC thermal intercept design is improved in production CMs, incl. JLAB pCM. (next slide)

• Antenna heating changes QL ~7 % (16MV/m) at ~10 hrs timescale

Thermal intercept improvements in production CM (vs. pCM)


• Use XFEL FPC design (vs.TTF3 in pCM). Better AL can ‐flange thermal contact

• Thermal strap is connected close to 50K pipe (vs. 50K shield‐see thermal analysis) Better thermal contact between braid and

copper extension (flat surfaces) JLAB pCM already has this feature

pCM: Thermal analysis – R.Fischer/ANL

zoom

5K th

ermal

Intercep

t50K thermal

intercep

t

pCM

Production CM

120K110K

55K150K

Other systems

– Magnet consists of conduction-cooled Quadrupole and Horizontal & Vertical correctors

• All coils able to achieve 20 A without quenching• Deguassing ramps demonstrated


– CMTS1 commissioning in parallel with first tests– Other measurements/testing

• Tuners• Low Level RF • Thermometry• Interlocks

FNAL pCM Vibration, microphonics (microphonics team)


• RMS vibration levels are up to 100 Hz or more (spec 10Hz peak)

• significant variation in time, amplitude, frequency and space.

• Some frequency lines correlate well with output massflow. Cryogenic noise is most likely candidate.

JLAB data is coming. Changes in design is under discussion (PRR; Test results workshop)

JLAB first microphonics measurements (T.Powers)

Peak to peak value is 96 Hz, rms value is 10.4 Hz‐rms

Preliminary result: CAV3 has 70 Hz pp and CAV4 has 20 Hz pp"LCLS-II pCM test" at PIP-II meeting, January 17, 2017

39

Summary

– pCM testing at Fermilab has advanced from minimum acceptance to extended testing• Cryomodule exceeds total voltage specification• Cryomodule exceeds heatload specification• Magnet specs are demonstrated• Radiation acceptable, no dark current measurement until

microphonics is suppressed to below 10Hz.– Great progress in ability to test CM’s at CMTS1

• more to do• likely take 2 more test cycles to reach stride

– Schedule not out of line with original estimates – Staffing, test tools require additional thought and discussion– Once both pCM’s (FNAL and JLAB) are tested it would be

beneficial to regroup and compare both results and processes



• Thanks all LCLS-II team for excellent job and good results.

• Many slides are borrowed from FNAL team presentation at LCLS-II pCMtest result workshop, (JLAB, January 11-13, 2017. Thank you) and Cryomodule PRR.


Thank you for your attention

pcm test: preliminary results seminar... · 2017-01-18 · – main and hom couplers test results...

Documents