Q slope analysis of global data and new techniques for Q slope studies. Anna Grassellino TRIUMF, University of Pennsylvania

Transcription

1 Q slope analysis of global data and new techniques for Q slope studies Anna Grassellino TRIUMF, University of Pennsylvania

2 Outline Q slopes: the high beta definifon A well defined problem? The low beta world Trends for low beta and comparison with high beta New technique for Q slopes studies: TRIUMF

3 Q slope: the high beta definifon

4 A well defined problem? High beta We define three different problems based on slope trends. Is this the correct approach? Global or localized effect? MagneFc field effect? Electric? Some test (CiovaF Jlab, Eremeev Cornell) show peak magnefc fields responsible for HFQS, what about Medium Field Q slope?

5 A well defined problem? High beta Analysis of correlafon between the slope in the medium and high field regions performed on the BCP not baked cavity Cornell data (G. Eremeev, PhD thesis) homogeneity of MFQS is concluded Spread in individual dt(h) curves observed in Fermilab EP baked cavity in the medium field region spafal nonhomogeneity is observed, especially in the high electric field region Need for more thermometry/cutout studies At TRIUMF with musr

6 Q slope in the low beta world Do definifons of high beta/low beta Q slope coincide? What is Q slope at low, medium and high field regimes in low beta cavifes? 1.E+11 MB10 cavity, 1.E+10 Qo 1.E+09 1.E+08 MB10 Praxede before 4K: MB10 Praxede a<er baking (48h@110 C) Eacc (MV/m)

7 Q slope analysis: approaches Model surface resistance and use the model to find figng values for different parameters: beher a posteriori (based on experimental evidence) Fit and look for trends in: temperature, treatments, frequency: it might allow to find correlafon and draw conclusions

8 Trend analysis LF MF HFQS, peak magnefc field range <20mT, 20 60mT, above 60mT Low beta cavifes analyzed include ~50 cavifes: QWR TRIUMF ISAC2 phase 1 (106MHz) and 2(141MHz), SPIRAL2(88MHz), MSU(80.5MHz), SPOKE ANL(345MHz), LANL(350MHz), ORSAY (352MHz) High RRR, 2 3mm walls, standard treatments include BCP (80 200microns), HPR, EP for ANL MFQS: QuadraFc and linear fit:

9 On G/Rs DefiniFon of quality factor: But magnefc field is NOT constant over cavity surface and IF AND ONLY IF Rs(H) = const it can be simplified to If the goal is giving a rough esfmate of the avg surface resistance then OK But if we try to understand the field dependence of Rs, it s meaningless to first assume Rs does not depend on H and then look for the (strong) H dependence.

10 Case study: low beta quarter wave 88 MHz Incorrect procedure Correct procedure by A.Romanenko (FNAL) RaFo of the correct gamma to the G/Rs gamma for this geometry: 1.76

11 G/Rs summary Comparison within one fixed geometry between different treatments qualitafve trend G/Rs values can be used Across different geometries G/Rs is incorrect, only numerical surface integrafon should be used to extract Gammas and other Rs parameters CorrecFon factors (GammaReal/GammaG_Rs) Low beta quarter wave 1.76 High beta ellipfcal 1.27

12 Low field Q slope Low field Q increase is never observed in low beta cavifes Steeper (than mf) slope below mt Effect more pronounced ater baking 1.00E E E E+09 Q 6.00E E E E Bpeak (mt) SPIRAL Gilia SPIRAL Verena "SPIRAL Praxaede" "SPIRAL Praxaede baked" "SPIRAL Gilia baked"

13 40 35 Medium field Q slope Average gamma=11, range 1 30 With G/Rs correcfon: avg gamma ~ 19, range Gamma

14 BCP, unbaked, 4.2K Average gamma = 14.43, with correcfon

15 40 CorrelaFon with residual resistance? gamma E E E E E E E E E E E 08 Residual resistance (Ohm)

16 2K vs 4.2K IS2 15 IS2 15* SP Pr* SP Pr*2K MSU 1 MSU 1 2K LANL2K LANL4K ANLspoke1.9 ANLspoke4.2K 4.2K 2K 4.2K 2K ISAC2 cav e e 7 Spiral Praxaede MSU LANL SS ANL TS

17 120C bake effect on MFQS 120C bake always improves significantly MFQS in SPIRAL cavifes Preliminary results of studies at TRIUMF (D.Longuevergne, B.Laxdal, V.Zvyagintsev) on ISAC2 cavifes show also improvement of MFQS with 120C bake 1.E+11 1.E+10 Qo 1.E+09 1.E+08 1.E+07 MB10 cavity, Eacc (MV/m) Gamma before Gamma a<er Rlin before SPIRAL Rlin a<er ISAC e e 7

18 BCP vs EP Lower MFQS with EP than BCP in TRIUMF 106 MHz QWR Appearance of linear component ater EP ANL spoke resonators also show lower gammas Qo 1.00E E E E+07 Qo vs Ea QoBCP QoEP2 Q7W ISAC-II specifications: Ea=6MV/m P=7W Ea, MV/m ISAC2 cav11 EP ISAC2 cav11 BCP Gamma Rlinear e

19 Hydrogen role on low beta MFQS Test at TRIUMF, look for correlafon slope hydrogen Several QW cavifes tested ater fast cooldown and ater 1 2 hr at 100K (V.Zvyagintsev) Trend in slope addifonal resistance from Q disease test Also, the lowest gamma value (gamma~2)among analyzed low beta resonators (2K) is the degassed ANL 0.63 triple spoke However ANL TS 4.2K slope (gamma ~10) did not change significantly with degassing 1.0E+10 7W 1.0E+09 Qo 1.0E E Ea, MV/m "4.2K test, no q- disease" "after Q- disease test"

20 Comparisons MFQS low high beta: 2K, 4.2K From this analysis 2K MFQS range ~ gamma from 2 to 15 MFQS at 2K higher for low beta cavifes Slope at 4.2K average around gamma ~ 25 for low beta, which is slightly higher than what observed in high beta (~20)

21 Comparison MFQS low high beta: contradictory results 120C bake 1.E+11 1.E+10 Qo 1.E+09 1.E+08 1.E+07 MB10 cavity, Eacc (MV/m)

22 HF onset and slope frequency dependence 1.E+11 QWR B, beta E+10 Qo MB01 Gilia MB02 Erentrude MB03 Verena SPIRAL, ISAC2 1.E+09 1.E+08 MB03 Verena (baking) MB04 Colehe MB07 Pezenne Eacc (MV/m)

23 Cutout studies at TRIUMF: musr

24 Magnetic field distribution of a vortex lattice Very sensifve local probe of magnefsm, able to tell about magnefsm that is localized in certain regions of the sample, and how much of the sample contains it. Asymmetry spectrum plotted in a rotating reference frame A(t) Fourier transform

25 RF losses due to fluxoids: of interest for MF and HFQS Two mechanisms for fluxoids in Nb: 1. Trapped flux 2. PenetraFon at sites with lower Hp (<Hc1) Two mechanisms of dissipafon: 1. StaFonary normal region 2. OscillaFng fluxoid Pinned Depinned

26 Hypothesis to test: HFQS Steep losses above mT due to early flux penetrafon Is the surface entering an intermediate mixed state? Correlate hot spots cutout from cavifes with areas of higher density of islands in the mixed state

27 Samples to be used RF Side Outer Side 27

28 DescripFon of the first experiment Look for intermediate mixed state in hot spots cutout samples (~ cm 2 size, 3mm thick, interested only in RF side) Need for a local probe: musr LAMPF spectrometer Field range mt, Temperature range 1K 4.2K 5 samples: PrisFne Nb from vendor Hot/cold spot cutout from large grain cavity (before and ater bake) provided by Alexander Romanenko, Hasan Padamsee (Cornell) BeamFme approved: ~1 day per sample 12 shits, starfng Oct 27th

29 Hypothesis to test: MFQS Field dependence of penetrafon depth Field dependent losses due to increased volume where dissipafon occurs H( t) E(z,t ) R s µ 2 0ω 2 λ 4 Δn 0 k B Tp F Ermolov, Marchenko, Chizov, 1986 R s (µ 02 ω 2 λ 3 σ n Δ/T)exp(-Δ/T) ln Δ + C ω 0 exp Δ k B T

30 ! ab (Å) NbSe 2 T = 0.02 K T = 2.5 K T = 4.2 K " ab (Å) ! e /! n H/H c2 Freeze out thermal excitations of quasiparticle core states to reveal multiband vortices H (koe) 0.0

31 EffecFve MagneFc PenetraFon Depth: MagneFc Field Dependence! ab (Å) NbSe 2 YBa 2 Cu 3 O 6.95 V 3 Si V 3 Si fully gapped LuNi 2 B 2 C anisotropic gap YBa 2 Cu 3 O 6.95 d 2 x -y2 -wave gap NbSe 2 multiband LuNi 2 B 2 C H/H c2

32 Pure Vanadium (marginal type II) Laulajainen, Callaghan, Kaiser & Sonier PRB 74, (2006)

33 DescripFon of the second experiment Determine the field dependence of the effecfve penetrafon depth (and vortex core size) in the vortex and intermediate mixed states. Will do this at several temperatures to invesfgate the possibility of two SC gaps. Take advantage of musr unique sotware for measurements of the vortex lagce in a marginal type II TF musr, dilufon refrigerator PrisFne single crystal sample BeamFme approved: 12 shits

34 Conclusions HFQS: well defined problem one physical underlying mechanism unsolved musr experiment at TRIUMF to confirm or rule out role of early flux penetrafon MFQS: several contributors, both local and global To be found in both microscopic and macroscopic parameters Hydrogen plays a role at low beta, need for more degassing studies (planned at TRIUMF) Need for diagnosfc tools like thermometry and more cutout studies some planned at TRIUMF again with musr