Beam Dynamics for CSNS Linac. Jun Peng September. 18, 2012

Transcription

1 Beam Dynamics for CSNS Linac Jun Peng September. 18, 2012

2 Outline Beam loss study MEBT optimization DTL optimization Geometry Accelerating field and synchronous phase Focusing scheme End to end simulation Page September 17,

3 Introduction L-DumpB 50keV 3MeV 80MeV H - RFQ MEBT DTL LRBT L-DumpA Beam current 15mA(1 st phase), 30mA(2 nd phase) Macro-pulse width 420 s Repetition rate 25Hz Duty factor 1.05% Chopping rate 50% RCS September 17, 2012 Page

4 Beam loss study PARMILA with 2D space charge routine.the initial distribution at the exit of RFQ is from PARMTEQM The beam current is 15mA(1 st phase) 30mA (2 nd phase) Error Analysis For the quadrupoles: Transverse displacements 0.1mm, Rotations 3mrad Integrated field 1, For the accelerating field: RF amplitude 1 RF phase 1degree September 17, 2012 Page 4

5 Beam loss rate-1 st phase, I=15mA Beam loss rate-2 nd phase, I=30mA Tank1 Tank2 Tank3 Tank4 Tank1 Tank2 Tank3 Tank4 Beam loss 89% 100% 100% 100% Beam loss 39% 100% 100% 100% <1W/m <1W/m Probability Probability Page September 17,

6 Tank1 Tank2 Tank3 Tank4 FD FD FD FD Old design Rb=0.6cm Rb=1.3cm Rb=1.3cm Rb=1.3cm FFDD FFDD FFDD FFDD New design: Rb=0.8cm Rb=1.0cm Rb=1.0cm Rb=1.2cm September 17, 2012 Page

7 Beam size (cm) x_rms y_rms x_max y_max Bore radius Beam size along the DTL Element number Beam size (cm) x_rms y_rms x_max y_max Bore radius Beam size along the DTL September 17, 2012 Element number Page

8 100 RMS Emittance Growth Rate(% x y z Don t simulate RFQ+MEBT+DTL together, Don t consider beam emittance growth in the MEBT Element number RMS emittance growth along the DTL RMS Emittance Growth Rate(%) x y z Element number RMS emittance growth along the MEBT+DT

9 DTL in (from MEBT) DTL in (K-V) X(cm) Y(cm) Y(cm) W(MeV) X (rad) Y (rad) X (rad) X(cm) Y(cm) Y(cm) W(MeV) Y (rad) X(cm) (deg) X(cm) (deg) Page September 17,

10 MEBT optimization September 17, 2012 Page

11 MEBT layout Old design: the MEBT comprises of a chopper, two 324MHzbuncher cavities and eight quadrupole magnets New design: the MEBT comprises of two 324MHzbuncher cavities and ten quadrupole magnets September 17, 2012 Page

12 Old design New design The rms emittance growth in x (dot), y (square) and z (triangle) direction versus the element obtained by code PARMILA September 17, 2012 Page

13 DTL optimization September 17, 2012 Page

14 DTL cell geometry-old old design Beta Bore Radius (cm) Face angle (degree) Inner nose Radius (cm) Outer nose Radius (cm) Corner Radius (cm) Diamter of drift tube(cm) Cavity Diameter (cm) Flat length (cm) Tank Tank Tank3~ September 17, 2012 Page 14

15 DTL cell geometry-new design Beta Bore radius (cm) 0.8 Face angle (degree) Inner nose radius (cm) Outer nose radius (cm) Corner radius (cm) Diameter of Drift tube(cm) Cavity diameter (cm) Flat length (cm) Tank Tank Tank Tank1 Tank4 September 17, 2012 Page 15

16 4 Old design Accelerating field 4 New design E0(MV/m) Tank1 Tank2 Tank3 Tank4 E0(MV/m) Tank1 Tank2 Tank3 Tank NCell NCell In the 1 st tank, the E 0 is ramped from 2.2MV/m to 3.1MV/m over the first 24 cells, and then keeps constant. To simplify RF tuning and obtain high accelerating gradient, constant field E 0 of 2.86MV/m is chosen over the 1 st tank. No field ramp in other tanks. No field ramp in other tanks September 17, 2012 Page16

17 Synchronous phase Old design New design s (degree) Tank1 Tank2 Tank3 Tank4 s (degree) Tank1 Tank2 Tank3 Tank NCell NCell Multi partlcle simulation shows that, at the exit of MEBT, the beam RMS phase width is 5.58degree, the total phase width is 30.98degree September 17, 2012 Page 17 17

18 Old design Tank number total Output energy (MeV) Length (m) Number of cell Total RF power (MW) I=30mA New design Tank number total Output energy (MeV) Length (m) Number of cell Total RF power (MW) I=30mA September 17, 2012 Page 18

19 Beam stability requirements t0, z0 <90 o n / 2,for n=1, 3 3 0t 0l Equipartitioning require: k k z0 3 2 t0 nz nt 1 2 Zero-Current Phase Advance per Period for FD lattice Zero-Current Phase Advance per Period for FFDD lattice September 17, 2012 Page 19

20 Focusing scheme FD VS FFDD FD Lattice FFDD Lattice Disadvantage: high quadrupole gradient, small beam bore radius Advantage: lower quadrupole gradient, bigger beam bore radius Page September 17,

21 FD Lattice FFDD Lattice MEBT MEBT DTL-1 DTL-2 DTL-3,4 DTL-1 DTL-2,3 DTL-4 Advantage: strong focusing, small beam size, small amplitude of envelope oscillation Disadvantage: weaker focusing, bigger beam size, bigger amplitude of envelope oscillation Page September 17,

22 Error Analysis For the quadrupoles: Transverse displacements 0.1mm, Rotations 3mrad Integrated field 1, FD Lattice For the accelerating field: RF amplitude 1 RF phase 1degree FFDD Lattice Disadvantage: In DTL-1, 89% probability beam loss <1W/m. Advantage: In DTL-1, 98.1% probability beam loss <1W/m. In DTL-3, 99.5% probability beam loss<1w/m

23 End to end simulation September 17, 2012 Page

24 End to End Simulation 50keV 3MeV 80MeV H - RFQ MEBT DTL LRBT RCS Stripper The simulation is performed from the RFQ exit to the RCS injection: Simulation code: PARMILA with 2D space-charge charge routine Number of particles: Peak current: 15mA Initial distribution: PARMTEQM output Page September 17,

25 MEBT match

26 LRBT match

27 End to End Simulation (Phase ) MEBT DTL LRBT X(cm) Y(cm) Beam envelope

28 Beam bore radius/rms rms beam size MEBT DTL LRBT ratio>5.2 DTL-1: ratio>5 DTL-2,3,4: ratio>5.8 Ratio>5.7 Page September 17,

29 RMS emittance along Linac MEBT DTL LRBT

30 RMS emittance growth along Linac September 17, 2012 Page 3 0

31 Summary In the old design, the beam loss in the 1 st DTL tank was found serious while all errors applied. The reason to this problem was that we don t t consider halo formation and emittance growth in the MEBT while optimizating DTL geometric parameters. So end-to to-end simulation must be conducted before reaching a final design. Two transverse focusing lattices have been compared, namely FD and FFDD focusing lattice. FFDD lattice was finally chosen for its lower quadrupole gradient and smaller beam loss rate compared with FD one. End to end simulation has shown that the beam loss, emittance growth rate and the ratio of bore radius to rms beam size were acceptable along the linac,, and now we reached a final design. September 17, 2012 Page 31

32 Thank you! Page September 17,