Inj. And Ext. of the Main ring(MR)  Dr. T. Koseki (this morning)

1. Experience with RCS Injection and Extraction SystemsPranab SAHAfor the RCS of J-PARCAugust 27, 200842nd ICFAAdvanced Beam Dynamics Workshopon High-Intensity, High-Brightness Hadron Beams( HB2008 )( Nashville, Tennessee, USA. August 25-30, 2008 ) Inj. And Ext. of the Main ring(MR)  Dr. T. Koseki (this morning)

2. Introduction and Outline Injection is one of the key issues of RCS and has impact on the whole system, where an well extracted beam reflects the overall performance. In order to mitigate the space-charge effect, the beam density in the transverse plane is controlled by utilizing the painting injection in RCS •General Injection system, Extraction System •Beam commissioning of Injection and Extraction Comparison with design stage parameters, System performance •Painting injection study Reconstruction of the phase space footprint using (1) BPM pairs with a turn-by-turn mode, (2) Using a tune BPM spectrum •Beam loss at the injection and Extraction Comparison with design stage estimation Near future study concerning beam loss related to injection •Summary

3. J-PARC Facility RCS of J-PARCLayout and key parameters Accelerators at J-PARC facility: ☆ 400 MeV (181 MeV at present ) LINAC ☆ 3 GeV 25 Hz Rapid Cycling Synchrotron (RCS) ☆ 50 GeV Main Ring Synchrotron (MR) Design parameters of RCS Circumference: 348.333 m Cell Structure: (6 DOFO arc + 3 DOFO insertion) x 3 Nominal tune: (6.68, 6.27) Natural chromaticity: (-8.5, -8.8) Injection energy: 0.4 GeV(*) Extraction energy: 3.0 GeV Beam power at ext.: 1 MW(*) Particle per pulse: 8.3 x 1013(*) Avg. output current: 333 mA(*) Repetition rate: 25 Hz 3 GeV RCS * Day1 inj energy: 0.181 GeV Output Goal: 0.3 ~ 0.6 MW (other related numbers are also thus different)

4. RCS Injection and Extraction Systems BPM BPM Inj. - H0 dump line components Magnets: Shift bump: 4 (SB1~4) Hori. Paint bump: 4(PB1~4) Vert. Paint Bump: 2 (VPB1~2) Septum:4(ISEP1~2, DSEP1~2) DC Str. magnets: 6 Inj. 4(2H, 2V), Dump 2(1H, 1V) Pulse Str. magnets: 2 (PSTR1~2) Quad: 3 (QFL, QDL , Dump-Q ) Monitors: MWPM (Multi-wire Profile Monitor):7(6) BPM (Beam position monitor): 4(2) K-BPM, I-BPM(Big-BPM1,2) ---------------------------------------------------------------------------- 1st Foil , 2nd Foil, 3rd Foil, etc.. Ext. area components: DC kickers: 2 (KICK-A,B) Pulse Kickers: 8 (KICK-A1~3, KICK-B1~5) QUAD: 2 (QFL, QDL) Septum: 3 (SEP1~3) Monitors: BPM: 2 (in the ring) Injection beam control @ foil X  w/ Isep or SB X’  w/ Isep Y, Y’  Inj. STR Monitors: MWPM 3,4(5) Both beam profile as well as the beam position can be measured by the MWPMs (MWPM1) Injection – H0 dump line

5. 3rd foil QFL QDL 2nd foil MWPM3 To beam dump MWPM5 MWPM4 H+ x ISEP1,2 H- H- s H0 H0 PB1,2 PB3,4 1st Foil Circulating beam SB1 SB2 SB3 SB4 Injection scheme in RCS —Paint inj (216.p.mm.mrad) (Hori.:131mm(90+41), -6.5mrad) (Vert: 0mm, -3.2mrad) —Center inj. (Hori:90mm,0mrad) (Vert.: 0mm, 0mrad) Design twiss parameters at the foil w/ nominal tune Inj. beam Circulating beam ax-1.452 1.550 bx(m) 11.138 11.275 ay-0.400 -1.589 by(m) 10.998 11.062 hx(m) 0.00 0.00 h’x0.00 0.00 Paint area for MLF target: 216p.mm.mrad Paint area for MR inj: 144p.mm.mrad (@ 400 MeV Linac) To change painting area in the horizontal plane, X’ of the inj beam will be change by two pulse steering magnets(PSTR1,2) Measured response matrix with ISEP1,2 and MWPM3,4,5  Used for controlling X and X’ of the inj beam at the foil paint inj parameters in the hori. Plane Measured response matrix with IVSTR1,2 and MWPM3,4,5  Used for controlling Y and Y’ of the inj beam at the foil. For painting in the vertical plane, measured response with VPB1,2 and MWPM3,4 was used

6. Linac Energy: 181MeV Peak current: 5mA * Run#9 data Beam commissioning of Injection and Extraction Inj.–H0 dump line orbit looking from RCS central orbit IBPM MWPM2 BigBpm1 MWPM3 MWPM4 MWPM5 BigBpm2 MWPM6 MWPM7 Cal: 418.6 226.0 214.7 133.787.70132.2 213.5 481.0 3416 Meas: 412.0 228.8 214.0* 133.487.70132.3 219.0* 491.1 3435 Comparison of Magnet Parameters(K0) Cal(mrad) Used(mrad) Diff(%) ISEP1 107.0 110.0 +2.8 ISEP2 160.4 173.8 +8.35 DSEP1 146.4 144.0 -1.66 DSEP2 311.4 317.4 +1.93 Measured response matrix near the present orbit was exactly similar with the model cal Efficiently used for the beam commissioning in both the center and paint injection. • Present center inj orbit is very much different than the • design painting orbit, while 2 Pulse STR magnets in addition • to septum using in the design stage are not available. • The beam orbit passes very off-center through ISEP2, where • no field measurement through this orbit couldn’t reach • May be the main reason!! w/ paint orbit (100p), the difference becomes: +5.3%  Gets better but not sufficient  under study Concerning extraction, no sufficient data but the overall agreement is satisfactory Unlike, injection area, there are lack of monitors in the extraction section

7. Stability of the injection and extraction beam (systemperformance) Inj. Beam jitter @ last few BPMs in the L3BT line 76X sx=70mm 76Y sy=82mm Stability of the ext. kicker magnets: ~ 0.2%  Dx ~0.1mm at the exit of ext. septum  negligible 79(K-BPM)X sx=159mm 79(K-BPM)Y sy=112mm Finally, the ext. orbit stability was measured as follows: •H •V Several hours data Upstream 4 BPMs were noisy! Position deviation(s)[mm] 0.3mm In RCS, the fluctuation of the Shift bumpflat top was within± 1%, while all other DC magnets both in the injection and the extraction line were just stable! 3NBT BPM ID#

8. medium pulse ~ Peak current Macro pulse ~ Vertical plane Correlated painting Anti-correlated painting Painting injection study: Run# 16, 17 Painting process in RCS: schematic view Painting injection major study items: 1. Horizontal Painting for 100p, 150p and 200p painting area. 2. Vertical Painting for 100p painting area [both with i) constant flat top and ii) with decay pattern ] 3. Horizontal and vertical together for 100p with decay pattern Horizontal plane Beam conditions Linac peak current: 30mA One intermediate pulse inj RCS DC circulating mode RF chopping: 560ns w/ SB(90mm) inj beam 6p.mm.mrad w/ PB(41mm,-6.5mrad) (500ms) 1st foil location One intermediate pulse inj Pulse length=560ns

9. SB(620ms) PB(400ms) beam 500ms time 500ms SB(620ms) PB(500ms) t0 t1 t2 t3 t4 t5 time t0 t1 t2 t3 t4 t5 X’ SB(620ms) PB(500ms) X t0 t1 t2 t3 t4 t5 time Painting injection study: Run#16, 17 The top of the paint bump first fixed with a constant pattern by measuring directly the phase space coordinates at the foil using MWPM 3&4 The decay pattern was then made using following functions: Hori: Imax[1-Sqrt(t/T)] Vert: Imax[Sqrt(t/T)] Pattern for 150p and 200p were scaled from 100p and was perfect !! 100ms Hori. paint bump pattern@constant flat top inj(i) Hori. paint bump @ decay pattern(ii) Pattern moved t0~t5 w/ 100 ms step Vert. paint bump @ decay pattern(ii) Pattern moved t0~t5 w/ 100 ms step Schematic view for study with decay pattern

10. X_BPM X_foil = MnM X’_BPM X’_foil Painting injection study cont’d( Reconstruction of the phase space footprint ) Method: 1 BPM with single pass mode Using turn-by-turn data of BPM pair Two BPM pairs in drift space 1st pair: BPM id # 1101 and 1102 2nd pair: BPM id # 1902 &2001 Method: 2 Using spectrum of a Tune BPM measuring a betatron response matrix and by detecting the real and imaginary part of the betatron sideband peak gives the phase space coordinates at the injection point ( Detail : H. Harada, to be published in Ph.D. thesis ) Tune BPM Re [Xe(wx)] xe = R(wx) Im [Xe(wx)] x’e BPM: 1101 &1102 (1101 pair) BPM: 1902 &2001 (1902 pair) R(wx) = Measured Response matrix A(real) B(real) = A(imag) B(imag) Measured optics is reproduced well by model and the transfer matrices used here was calculated by the model ! Re [Xe(wx)] and Im [Xe(wx )]: Real and imaginary part of the betatron sideband peak Xe and x’e: reconstructed phase space coordinate at the foil X_BPM and X’: phase space coordinate of the beam center measured with BPM pair after subtracting COD Mn is the one turn transfer matrix M is the transfer matrix from foil to the BPM X_foil and X’_foil： reconstructed phase space coordinate at the foil

11. Injection beam Closed orbit Injection bump Checking of both methods first( matching of the injection and the closed orbit ) Single pass BPM: X[mm] X’[mrad] Y[mm] Y’[mrad] 1101 pair -2.28 0.69 -2.73 -1.05 1902 pair -2.20 0.71 -3.32 -1.12 Tune BPM -1.81 0.65 -1.10 -0.70 Both methods gave quite close and accurate results !! Just a single trial !! Quite accurate !! Horizontal Vertical Checked by IPM (Ionization profile monitor) Mountain plot of the beam profile Well minimized Betatron oscillation !! Then, starting with 100p painting, at first balance of 4 horizontal paint bump magnets were adjusted by measuring and correcting the COD during the paint bump on(0~1msec). Parameters for 150p and 200p were then just scaled. Vertical direction:  By measuring directly Y and Y’ of the inj beam at the foil with MWPM3,4

12. Circulating beam (200p.mm.mrad) Injection beam for (1) 100p, (2) 150p, (3) 200p painting X’(rad) •1101 pair result •1902 pair result 1 2 3 X[m] Measured Twiss parameters @ foil Inj. Beam: Circulating beam ax=-0.3302 ax=1.7144 ay=-0.23088 ay=-1.8712 bx=3.2207m bx=12.9575m by=4.8071m by=12.3682m ex,ey 3.5p.mm.mrad (3s) Painting study resultswith constant paint bump pattern For 100 p.mm.mrad painting ~ ~ X’(mrad) 94 125 X(mm) -4.4 Normalized phase space plot of the beam center measured by BPM turn-by-turn data Plotted reversely Dx[mm] Dx’[mrad] For 100p target : -31.1 4.4 Results: Singlepass 1101 pair: -30.6 4.69 Singlepass 1902 pair: -30.1 4.52 TuneBPM(h=1&downer): -29.9 4.90 TuneBPM(h=1&downer): -29.1 4.66 Measurement w/MWPM: -31.2 4.61 X’n • 100p •150p •200p Xn

13. SB(620ms) 500ms PB t0 t1 t2 t3 t4 t5 100ms time Target Single pass BPM 1101 pair Single pass BPM 1902 pair PB magnet off at t5 (1101 pair) PB magnet off at t5 (1902 pair) Measured by MWPM3-4 Footprint with Horizontal paint bump decay pattern Overlay all together Well reconstructed the phase space Footprint by the BPM single data. At t =t5, the pattern was a bit strange and was confirmed by the measurement with PB off at t=t5. Gets little disagreement with larger amplitudes  Under Inspection !! 200p Hori. paint bump @ decay pattern Pattern moved t0~t5 w/ 100 ms step 200p 150p 100p

14. SB(620ms) PB(500ms) t0 t1 t2 t3 t4 t5 time Vert. paint bump @ decay pattern Pattern moved t0~t5 w/ 100 ms step Target Single pass BPM 1101 pair Single pass BPM 1902 pair Measured by MWPM3-4 Footprint with vertical decay pattern(100p)Horizontal and vertical together(100p) Vertical only 100p Horizontal w/ HV together 100p (correlated) In the vertical plane, only with 100p painting study was performed. Reconstructed footprint in the vertical plane found a bit zigzag but finally gets to the expected position. A correlated painting study with 100p was also performed and the reconstructed results were expected Basic painting study in Run#16 was done very successfully. BPM with single pass mode was found to very effective and accurate. ( Detail with method 2 under analysis ) Paint bump patterns made in this study was used for the painting with high power run Vertical w/ HV together 100p (correlated)

15. Beam loss @ Injection and Extraction areas Residual level measured after Run#14 ( w/ high current operation ) Run ended: 2/25, 3:55 Measured: 2/25, 13:30 Extraction wide area Clean !! Except H0 branch, there is no significant beam loss in the inj and ext areas Losses in the arc sections  During RF study • H0 dump branch • Not understood yet clearly !! • H0? ( if 2nd foil not working well ! ) • Long tail of the inj. Beam + scattered? • (beam loss signal even in the 1/3 mode ! ) • …………………………. • Foil system broken no study done so far • To be done in the next run (foil system recovered) Injection branch Mainly because of the present center inj orbit Very close to the aperture limit!!

16. Major sources and quantity of the beam • loss in the injection area: • ( Considering inj beam power of 18kW • [0.3MW@RCS extraction] ) • 1. Lorentz striping loss of • the injection(H-) beam: << 1W • Inj beam line clean!! • 2. H0 excited states losses: < 2W • 1st foil to the bump region clean!! • 3. Nuclear scattering together with multiple • coulomb scattering at the charge-exchange foil: •  Major source among all • Total loss: 38W(~0.2%) • P.K. Saha, PAC07 (GEANT + SIMPSONS) • Simulation with Real painting process (Space charge: off) • Foil size: 32mm x 36mm (design) • and with ideal bump systems • Average foil hits: 24 Beam loss in the RCS injection area (Estimation vs. reality ) Mostly may be due to the 3rd source: Average foil hit with present foil and bump systems: ~ 200  8 times higher than the design  Checked with simulation qrms(hand cal)=~1mrad !  large !! Detail study to separate foil scattering loss with recovered foil and systems to be done in the next run Survival ratio: 97% DCCT (beam survival ) 5.0x1012 ppb (Peak: 25 mA, Macro: 0.12 ms, Chopping: 560 ns) 4.9x1012 ppb (Peak: 5.8 mA, Macro: 0.5 ms, Chopping: 560 ns) Time[s]

17. Summary and future plan The injection and extraction systems of RCS are found to be performed well during the beam commissioning so far. Except one case, design stage parameters are found to be consistent with the beam commissioning parameters, where the relative values were exactly similar each other  Beam commissioning proceeded quite smoothly for any operation Painting study utilizing one intermediate pulse injection was successfully performed by measuring accurately the phase space footprint..  Phase space control can be done in the painting inj. Estimation of the beam loss in the design stage was found to be consistent. Detail study with nuclear scattering loss is planned to perform in the next run.

18. Discussion on several questions by the committee Does the system perform as expected? Ans: System performance satisfactory so far! No remarkable trouble concerning inj and ext. Bump system working fine. Problem remains with Bump falling time and hope soon gets better. Foil system gets recovered Did the simulations/calculations performed during the design stage accurately predict the actual performance? Ans: Yes. Inj and ext. beam orbit even with very different parameters goes through expectation. The beam profiles are also quite similar. The relative parameters of all magnets are exactly similar to the simulation/calculation. Needs more experimental data What are the major limitations in performance? Were they known in the design stage? Ans: Foil system was one big limitation of the injection commissioning. But it’s going to be fine from the next run. So far not face any other big limitation and would be clear with long time (and high power) operation. No BPM in the extraction section is one small limitation concerning extraction. If someone were to begin now designing the same type of system for a similar machine, what is the one piece of advice that you would give them? Ans: Please try to have enough monitors, especially in the complicated areas like injection and extraction. Keep (check for) enough space for not to push nearby magnets/elements nor make thinner the shield at the last moment to install later elements/magnets

19. Extra slides backup: 1 Typical optics of the injection and extraction Inj - H0 dump line(~30m) bxby √b(√m) DQ ISEP1 QFL SB End of H0 dump ISEP2 QDL DSEP1 DSEP2 End of L3BT line S(m) Ext - 3NBT dump line(~70m) bxby KA KB Ent. of Ext. unit ESEP Ent. of H0 dump

20. u cosq -sinq x = v sinq cosq y s2u cos2q sin2qs2x = s2v sin2q cos2qs2y backup:2 Measured beam profile by MWPM MWPM4: Measured profile: U plane MWPM4: Measured profile: V plane sy=1.63mm sx=1.35mm The mean and the width of the measured profiles in the U and V planes were transformed to the X and Y plane as follows: q is the wire inclination: 17.7 deg.

21. Inj beam PSTR1 ISEP1 PSTR2 IBPM MWPM2 Dx ISEP2 RCS Central Orbit backup:3 Reason for discrepancy with ISEP2 Inj orbit with different patterns Paint inj[x,x'@foil=131mm,6.45mrad] (Knobs: Isep1,2) Center inj[x,x'@foil=100mm,0.0mrad] (Knobs:Isep1,2 Pstr1,2) Center inj[x,x'@foil=90mm,0.0mrad] (Knobs:Isep1,2 Pstr1,2 ) Center inj[x,x‘@foil=90mm,0.0mrad] (Knobs: Isep1,2) • Day1 center orbit is much different than the design • center orbit as well as paint orbit. As PSTR1,2 are • not used, orbit @ ISEP2 passes very off-center.. • [ (x-plane) < -110mm from the paint orbit @ISEP2 ent.] • and thus ISEP2 gets very Strong!! • Field measurement could not cover • for the day1 center orbit region !! • (Field quality may be slightly different • than the central region? ) • To be estimated by field calculation Day1 Center inj Orbit 111mm!

22. backup: 4 Hints from ISEP1 response data Discrepency of ISEP2 with Paint inj orbit gets better !! Paint inj orbit for 100pmm.mrad Isep1  3.2% ISEP2  5.0% Agreement better as compared to the comparison with center inj(page#8) but still more to go!!  Under analysis Call ISEP1 Day1 parameter:110mrad Cal –Meas pos @ I-BPM=12mm Measured Both are linear but inclination is different Response of ISEP1 with I-BPM ISEP1 design parameter(paint inj):152mrad Cal –Measured position@I-BPM=-1.46 mm Real Data: difference=-1.58mm  consistent !! Response of ISEP2 with MWPM2 Could check only with Dq of ± 5% agreement was good • Difference gets smaller as q • gets higher. • Minimum(~0) with design parameter(paint inj). • May be one key source which Makes the difference!! Cal Measured Cal – measured position @ Ibpm Day1used para :174mrad

23. backup: 5 1st Foil 2nd Foil 3rd Foil SB1 SB4 K0~1mrad SB2 SB3 MWPM4 MWPM5 MWPM3 Response of Shift bump and paint bump magnets • Run#9 : Unit: (mm, mrad ) • SB MWPM3 MWPM4 MWPM5 • OFF 180.74 x 180.62 • 90 138.8 98.0 139.4 • Bump height: • (180.74+180.62)/ 2 – 98.0 = 82.7mm • Using SB off data: (X, X’)@QFL_END • becomes = (180.8mm, -0.03mrad) • Simulation using upper X,X’ and with SB=83mm, • 138.7 98.0 139.4 • SB response: 83/90 = 0.922 Gradient = 0.93 MWPM4 DX[mm] SB set DDx[mm] Unit:( mm, mrad ) 100p.mm.mrad150p.mm.mrad Target: ( 31.1, -4.4 ) ( 39.9, -5.65 ) Run16: ( 27.6, -4.0 ) ( 34.8, -5.08 ) Run17:( w/ Current = + 10% ) ( 31.2, -4.6 ) ( 39.3, -5.9 ) Circulating beam 150p.mm.mrad Circulating beam 100p.mm.mrad X’[rad] X’[rad] Injection beam 3.5p.mm.mrad Vertical paint bump: No sufficient data but +10% already added and gave expected results Just may be a missing factor (10) in the analysis of field measurement data!! Under inspection !! Expected Run16 meas. Run17 meas. X[m] X[m] Response of hori. Paint bump magnets

24. backup: 6 Ext. Kickers flat top measurement Without Timing Correction With Timing Correction • Required flatness = 2 % in the time length of 840 nsec • The trigger timing of the each kicker was adjusted in order • to cancel out the peaks and troughs of the flattop. • The flatness of 2 % was achieved in the time length of 850 nsec!

25. backup: 7 Results with constant flat top cont’d( for 150p, 200p painting ) 150p Target: Dx = -39.86 mm Dx’= 5.65 mrad Result from single pass BPM 1101 pair: -35.1 mm, 5.83 mrad 1902 pair: -36.6 mm, 5.92 mrad Measuring with MWPM3,4 Dx = -39.30 mm Dx’ = 5.92 mrad w/ inj beam 150p paint orbit but PB off: 1101 pair: 37.2 mm, -6.13 mrad 1902 pair: 36.3 mm, -5.90 mrad 200p Target: Dx = -47.02 mm Dx’= 6.67 mrad Result from single pass BPM 1101 pair: -42.5 mm, 6.85 mrad 1902 pair: -42.8 mm, 6.81 mrad Measuring with MWPM3,4 Dx = -48.05 mm Dx’ = 7.076 mrad Not measured w/ inj beam 200p paint orbit but PB off

26. backup: 8 Beam profile with IPM@painting inection 150p painting Beam: 5mA, 500ms, 56ns bx @ IPM=8.27m h~4m

27. backup: 9 入射軸ずれによるビーム運動 The transfer matrix M foil→TuneBPM from foil to monitor is The matrix for n turns is The position and derivative after n turns if there is injection error are

28. backup: 10 入射点の変位に伴うベータトロン振動の応答 x0 : initial position, x0’ : initial derivative Betatron response matrix can be defined as α, β: Twiss parameter for transverse plane Real and Imagine part of the detected betatron sideband peak

29. backup: 11 応答行列の測定と入射座標の同定 Response Matrix is We can identify the injection error (xe,xe’) from this RM and the measured real & imagine part of betatron peak

30. backup: 12 Bump pattern@paint inj study

31. Profile Comparison @ IPM : RCS 1/3 Mode, Run#9 Data backup: 13 MWPM4:U MWPM4:X MWPM4:Y Reproduce for Simulation input MWPM4:V

32. backup: 14 Profile Comparison Result @ IPM Input @MWPM4 Output @IPM-V Output @IPM-H IPM-H IPM-V Measured Data