RF structure design

KT high-gradient medical project kick-off30.05.2013

Alberto DegiovanniTERA Foundation - EPFL

A. Degiovanni 2

Outline

• Introduction– RF cavities constraints for hadrontherapy

• Backward travelling wave cell design and optimization for high gradient operations– Nose cone study– Tapering

• Comparison of different structure designs– SW SCL design– backward TW

• Preliminary studies for linac design• Conclusions

30/05/2013

A. Degiovanni 4

T1 T2 T3 T4 T5 T6 T7 T8 T9

Linac layout and BDR requirements

• Quasi-periodic PMQ FODO lattice sets a limit to the length of each structure and determines the group velocity range.

• The cells in each structure (tank) have the same length, while from one tank to the next, the cell length increases:

β tapering in the range 0.22-0.60• Trade-off between transverse acceptance and RF efficiency:

bore aperture = 5 mm• Max BDR: 1 BD per treatment session (~ 5 min) on the whole

linac length (~ 10 m). BDR ~ 10-6 bpp/m

...

30/05/2013

A. Degiovanni 5

NOVEL DESIGN FOR HIGH GRADIENT OPERATION

30/05/2013

A. Degiovanni 6

Proposal for bTW design for hadrontherapy

with: Sc < 4 MW/mm2

tTERA = 2500 nstCLIC = 200 nsBDRTERA = BDRCLIC = 10-6 bpp/m

.515

constBDRtS pulsec

DESIGN GOAL and CONSTRAINTS

Ea:= E0T ≥ 50 MV/m

Sc/Ea2 < 7 10-4 A/V

Proposed by A. Grudiev

P_0P_load

P_wall

z

Lvg_in ~ 0.4% cvg_out ~ 0.2% c

filling time ~ 0.3 µs

30/05/2013

8 holes (22.5 deg sweep) – radius scan

doubling the number of holes will double vgwhile keeping Sc_holealmost constant

normalized Sc in the coupling hole [10-4 Ω-1]

cone A gap Rc cs_h ac_Diam vg R'/Q Sc/Ea^2_slot

deg mm mm mm mm ‰ Ohm/m 10-4 V/A

25 5.2 2 28 72.331 0.45 8111 2.00

25 5.2 2.5 28 72.172 1.04 8127 2.46

25 5.2 3 28 71.927 2.04 8147 3.01

25 5.2 3.5 28 71.564 3.54 8177 3.57

25 5.2 4 28 71.093 5.63 8215 4.40

ROI nose_Sc/Ea^2

ROI group velocity

vg [10-3 c] ~ (r[mm])3.65

A. Degiovanni 8

Nose geometry optimization

• Scan on:– Nose cone angle– Gap– Nose cone radius(*)– Phase advance (120°-150°)– coupling hole radius

(vg = 4 ‰ and 2 ‰ )

• Optima:– Minimum of the quantity:

22a

c

a ES

EP

Q

RE

Sv a

c

g

'

2

* based also on results of the SCL optimization

nose radii

bore radius

half gap

septumnose angle

30/05/2013

A. Degiovanni 9

angle scan – 120 deg

g5 a25 g5 a55 g5 a75

30/05/2013

Optimization plots

Optimization plots

R’/Q [Ω/m]

R’ (or ZTT) [MΩ/m]

Optimization plots - fields

Optimization plots

vg [10-3 c]

Sc/Ea2 [10-3 Ω-1]

R’/Q [Ω/m]

22a

c

a ES

EP

Q

RE

Sv a

c

g

'

2

Optimization plot – 120 deg – gap = 5.5 mm

min {Max {Xnose,Xslot}}

Optimization plot – 120 deg – gap = 5.5 mm

g4 a25 g4 a55

gap and angle scan – 120 deg

g5 a25

g6 a25 g6 a55

g5 a55

g4 a75

g5 a75

g6 a75

120° - 16 holes – nose 1 -2 mm – gap and angle scan

g 5.5 mmA 65 deg

g 5.5 mmA 65 deg

120° - 16 holes – nose 1 -2 mm – gap and angle scan

g 5.5 mmA 65 deg

g 5.5 mmA 65 deg

150° - 16 holes – nose 1 -2 mm – gap and angle scan

g 7.0 mmA 55 deg

g 7.0 mmA 55 deg

150° - 16 holes – nose 1 -2 mm – gap and angle scan

g 7.0 mmA 55 deg

g 7.0 mmA 55 deg

A. Degiovanni 21

COMPARISON BETWEEN TW AND SW STRUCTURES

30/05/2013

A. Degiovanni 22

Geometry of LIBO structure

Comparison between TW structure and SCL

PUT NEW PLOT

Tapered structures:the coupling holes are smaller along the structure

30/05/2013

A. Degiovanni 23

Comparison of E-field in TW and SWπ/2 phase advance2/3 π phase advance

30/05/2013

A. Degiovanni 24

+ simpler mechanically+ less material and brazing needed (lower number of cells)+ tuning is easier for TW+ shorter filling time+ no bridge couplers

- small wall thickness- material properties change during brazing- Dissipated power is higher (half power goes to the load) Recirculation loop (power for TW 10-20% higher than SW)

PROs and CONs of bTW compared to standard SCL design

waveguide

accelerating cavities

coupling cavities

I. Syratchev30/05/2013

A. Degiovanni 25

Preliminary design for high gradient bTW linac

1. Independent rotary joints 2. -3 dB recirculation3. Small RF load compared to TW

T2T1

load

load

MKs~15-16 MW

klystron

2

3

T2T1

load

load

MKs

1

2

3

2 x 7.5 MW klystrons

30/05/2013

A. Degiovanni 26

Global optimization of bTW linac• Energy range: 60-230 MeV• 16 structures (tanks) – Total length: 5.9 m• 16x7.5 MW klystrons – 8 modulators• Total peak power needed: 206 MW• Peak power with recirculators: 114 MW

– Effective filling time increases to 2.1-3 μs

• Total average power from MKs: 150 kW

• Energy range: 65-230 MeV• 16 structures (tanks) – Total length: 5.5 m• 16x7.5 MW klystrons – 8 modulators• Total peak power needed: 260 MW• Peak power with recirculators: 120 MW

– Effective filling time increases to 1.8-2.5 μs

• Total average power from MKs: 150 kW

Gain ~ 2.2

Gain ~ 1.8

30/05/2013

A. Degiovanni 27

Fast active energy and intensity modulation:RF pulses and beam pulses

Klystron RF pulse

RF power into the tanks

Proton pulses from source

5.0 μs

2.5 μsActive energy modulation

8 ms

Active intensity modulation

time

I

P

P

30/05/2013

A. Degiovanni 28

Summary

• Optimization of TW structures for high gradient operations has been performed for 120° and 150° phase advance.

• The RF design of the input and output coupler is now ongoing.• The optimization of the whole linac layout has started recently and needs

some iterations, but looks promising• The design and test of the novel bTW structures is boosting the TULIP

project!

T

H

A

NK

Y

O

U

30/05/2013

BACK-UP slides

A. Degiovanni 30

TULIP-CLIC-bTW – beta=0.3798 (W=76 MeV)

Rc

csl_h

DIAM/2

GAP/2

LENGTH

position

R_coupling

angle

angle

R_coupling

positionRcorner

Racetrack slot

Rectangular slot

Circular coupling holes

30/05/2013

A. Degiovanni 31

SUMMARY 120 deg 150 deg SCL base SCL – HG

wall thickness (mm) 1.5 1.5 3.0 3.0

gap (mm) 5.5 7.0 5.1 9.5

nose cone angle (deg) 65 55 25 55

length (mm) 189.9 189.9 189.9 189.9

ncell 15 12 10 10

Ea_avg (MV/m) 25 25 25 25

Sc_nose (MW/mm2) 0.149 0.185 0.486 0.188

t_pulse (ns) flat 2500 2500 2500 2500

expected BDR (at given Ea and t_pulse) (bpp/m) based on Sc limit 1.1 E-22 2.9 E-21 5.7 E-15 3.7 E-21

max Ea (for BDR of 10-6 bpp/m) (MV/m) 85.2 76.3 47.1 75.7

Pin (MW) (w/o recirculation) 2.70 5.19 2.49 5.10 1.75 2.26

Pout (MW) (w/o recirculation) - 2.90 - 3.02 - -

Q0 (first/last) 6482/6721 7088/7545 8291 8250

vg (first/last) [%c] 0.421/0.226 0.404/0.236 - -

R’/Q (first/last) [Ohm/m] 7872/7847 7835/7794 8406 6355

time constant (ns) 320 340 440 440field rise time (time to reach 99% field)

(ns) (w/o recirculation) 750 204 800 204 1050 105030/05/2013