Microwave-Based High-Gradient
Structures and Linacs
Sami Tantawi
SLAC
4/12/2020
1
Acknowledgment
•
The work being presented is due to the efforts of
– V. Dolgashev, L. Laurent, F. Wang, J. Wang, C. Adolphsen, D. Yeremian, J.
Lewandowski, C. Nantista, J. Eichner, C. Yoneda, C. Pearson, A. Hayes, D.
Martin, R. Ruth, S. Pei, Z. Li
SLAC
– T. Higo and Y. Higashi, et. al., KEK
– W. Wuensch et. al., CERN
– R. Temkin, et. al., MIT
– W. Gai, et. al, ANL
– J. Norem, ANL
– G. Nusinovich et. al., University of Maryland
– S. Gold, NRL
– Bruno Spataro, INFN Frscati
4/12/2020
2
Overview
• Review of experimental data on high gradient
structure
• Efficiency and standing wave accelerator
structure designs
• A 1 TeV collider
– Improved efficiency
– Possible parameter set
4/12/2020
3
Research and Development Plan
4/12/2020
Experimental Studies
•
Basic Physics Experimental Studies
– Single and Multiple Cell Accelerator Structures (with major KEK and CERN contributions)
• single cell traveling-wave accelerator structures (Needs ASTA)
• single-cell standing-wave accelerator structures (Performed at Klystron Test Lab)
– Waveguide structures (Needs ASTA)
– Pulsed heating experiments (Performed at the Klystron Test Lab, also with major KEK and CERN
contributions)
•
Full Accelerator Structure Testing (Performed at NLCTA, with CERN contributions)
Can only be done at NLCTA at SLAC
4/12/2020
5
BreakdownProbability 1 pulse m
Breakdown Probability for a Standing Wave
Accelerator Structure with a/l=0.22
Flat Top Pulse width
0.1
0.001
Typical gradient
(loaded) as a
function of time.
600ns
10 5
150ns
Pulsed Temp Rise
10 7
60
80
100
120
140
Accelerating Gradient MV m
4/12/2020
6
Breakdown Probability for a Standing Wave
Accelerator Structure with a/l=0.14
BreakdownProbability 1 pulse m
0.01
Flat Top Pulse width
0.001
300ns
10 4
10 5
200ns
10 6
150ns
10 7
100
110
120
130
140
150
160
170
Accelerating Gradient MV m
4/12/2020
7
Breakdown Probability for a Standing
Wave Accelerator Structure with a/l=0.1
BreakdownProbability1 pulsem
0.1
Flat Top Pulse width
0.01
600ns
0.001
10 4
400ns
10 5
200ns
10 6
100
120
140
160
180
200
Accelerating Gradient MV m
4/12/2020
8
BreakdownProbability 1 pulse m
Breakdown Probability for a Standing Wave
Accelerator Structure with different
a/l=0.21
0.1
0.01
0.001
10 4
10 5
80
100
120
140
160
180
200
Accelerating Gradient MV m
4/12/2020
9
BreakdownProbability 1 pulse m
Breakdown Probability for a Standing Wave
Accelerator Structure with different a/l
1
0.1
0.01
0.001
40
50
60
70
80
90
Peak Temprature Rise C
4/12/2020
10
4/12/2020
Page 11
4/12/2020
Page 12
BreakdownProbability 1 pulse m
Breakdown Probability for a Standing Wave
Accelerator Structure with a/l=0.22
Flat Top Pulse width
1
0.1
600ns
0.01
400ns
0.001
10 4
200ns
10 5
90
100
110
120
130
140
150
Gradient MV m
4/12/2020
13
BreakdownProbability 1 pulse m
Breakdown Probability for a Standing Wave
Accelerator Structure with a/l=0.22, Pulse
length=200ns
0.1
0.01
CuCr
Cu
0.001
10 4
CuZr
Hard Cu
10 5
90
100
110
120
130
140
150
Gradient MV m
We have not tested yet structure with small apertures and different material
“Stay
Tuned” This is coming very soon
4/12/2020
14
Field Profile over the Iris
Es/Ea
2.0
Distance
Those two peaks are
correlated for structures
with different t and a/l
1.5
(Zo Es Hs)1/2/Es
1.0
Hs/Es
0.5
0.0
0.0
4/12/2020
0.2
0.4
0.6
Distance cm
0.8
1.0
1.2
15
Peak H and Peak ExH
Peak
ZoHsEs
Ea
1.6
1.5
1.4
1.3
1.2
1.1
4/12/2020
1.2
1.3
1.4
Peak Hs Ea
1.5
1.6
16
Full Accelerator structure testing ( the T18
Research andstructure)
Development Plan
Frequency.
11.424GHz
Cells
18+input+output
Filling Time
36ns
a_in/a_out
4.06/2.66 mm
vg_in/vg_out
2.61/1.02 (%c)
S11
0.035
S21
0.8
Phase
120Deg
Average Unloaded Gradient
over the full structure
55.5MW100MV/m
Field
Amplitude
Eacc _ out Eacc _ in ~ 1.5
•Structure designed by CERN based on all empirical
laws developed experimentally through our previous
work
•Cells Built at KEK
•Structure was bonded and processed at SLAC
Cumulated
•Structure was also tested at SLAC
Phase Change
4/12/2020
120°
Page 17
RF Processing of the T18 Structure
RF BKD Rate Gradient Dependence for 230ns Pulse at Different
Conditioning Time
RF BKD Rate Pulse Width Dependence at Different Conditioning
Time
-4
10
-4
10
After 500hrs RF
Condition
-5
10
After 900hrs RF
Condition
-6
10
After 1200hrs RF
Condition
-7
10
95
100
105
110
Unloaded Gradient: MV/m
115
G=108MV/m
BKD Rate: 1/pulse/m
BKD Rate: 1/pulse/m
After 250hrs RF
Condition
10
10
-5
G=108MV/m
-6
G=110MV/m
10
-7
100
150
200
RF Flat Top Pulse Width: ns
This performance may be good enough for 100MV/m structure for a warm collider, however, it does not yet
contain all necessary features such as wakefield damping. Future traveling wave structure designs will also
have better efficiency
4/12/2020
Page 18
2007-09 CERN/CLIC Design Structures Tested at
NLCTA
(Yellow = Quad Cell Geometry, Green = Disk Cell Geometry)
In Beamline
Structure
Note
Performance
11/06 – 2/07
C11vg5Q16
First X-band Quad
- Irises Slotted
Poor: 57 MV/m, 150 ns, 2e-5 BDR – grew whiskers
on cell walls
2/08 - 4/08
C11vg5Q16
Redux
Refurbished
Initially good (105 MV/m, 50 ns,1e-5 BDR) but
one cell degraded
4/07 – 10/07
C11vg5Q16-Mo
Molybdenum Version of
Above
Poor: 60 MV/m, 70 ns, 1e-6 BDR
10/08 – 12/08
TD18vg2.6_Quad
No Iris Slots but WG
Damping
Very Poor: would not process above 50 MV/m,
90ns – gas spike after BD
4/08 – 7/08
T18vg2.6-Disk
Cells by KEK, Assembled
at SLAC
Good: 105 MV/m, 230 ns at LC BDR spec of 5e7/pulse/m but hot cell developed
7/08 – 10/08
T18vg2.6-Disk
Powered from Downstream
End
Good: 163 MV/m, 80 ns, 2e-5 BDR in last cell,
consistent with forward operation
12/08 – 2/09
T18vg2.6-Disk
CERN
CERN Built, Operate in
Vac Can
Very Poor: very gassy with soft breakdowns at 60
MV/m, 70 ns
4/12/2020
Page 19
SQRT(EmaxHmaxZ0)/Ea
Emax/Ea
HmaxZ0/Ea
4/12/2020
20
T18 Surface Field Parameters in
Comparison with SW Structures
Emax/Ea
SW
SW a/l=0.22
SW a/l=0.14
SW a/l=0.1
SQRT(EmaxHmaxZ0)/Ea
4/12/2020
HmaxZ0/Ea
21
So, What does all this imply for a 1 TeV collider parameters?
•
High gradient requires high power density/unit length, quadratic with gradient.
Hence the two beam choice for the CLIC design
•
However, since one also accelerates in shorter length the total required RF power
increases linearly with gradient.
~
x
RI
;x  s 0
Ea
1 x
•
High gradient also implies reduced efficiency;
•
However, one can compensate by increasing the efficiency the accelerator
structure and RF sources.
–
Increasing the efficiency of the accelerator structure would reduce the power/unit length, and
it might allow for the use of a conventional RF unit
–
At the moment the best ( published ) design that respects high gradient constraints imply an RF
to beam efficiency of about 28%
–
Higher efficiency RF source and accelerator efficiency imply lower RF system cost
4/12/2020
22
Yet another Finite Element Code
Motivated by the desire
•to design codes to perform Large Signal Analysis for microwave tubes
(realistic analysis with short computational time for optimization)
•study surface fields for accelerators
•the need of a simple interface so that one could “play”
•A finite element code written completely in Mathematica was realized.
•To my surprise, it is running much faster than SuperFish or Superlance
•The code was used with a Genetic Global Optimization routine to
optimize the cavity shape under surface field constraints
Surface field penalty function
1.0
0.6
0.5
0.8
0.4
0.6
0.3
0.2
0.4
4/12/2020
0.1
23
0.2
0.4
0.6
0.8
1.0
1
2
3
4
Iris shaping for a structure with a/l=0.14
Shunt Impedance
Quality Factor
Peak Es/Ea
Peak Z0Hs/Ea
4/12/2020
83 MW/m
8561
2.33
1.23
Shunt Impedance 104 MW/m
Quality Factor
9778
Peak Es/Ea
2.41
Peak Z0Hs/Ea
1.12
• With 1 nC/bunch and bunch separation of 6
rf cycles the RF to beam efficiency~70%
24
• The power required/m~287 MW
Iris shaping for a structure with a/l=0.1
Shunt Impedance
Quality Factor
Peak Es/Ea
Peak Z0Hs/Ea
4/12/2020
102 MW/m
8645
2.3
1.09
Shunt Impedance 128 MW/m
Quality Factor
9655
Peak Es/Ea
2.5
Peak Z0Hs/Ea
1.04
• With 0.5 nC/bunch, bunch separation of 6
rf cycles, and loaded gradient of 100
MV/m the RF to beam efficiency~53%
25
• The power required/m~173 MW
Feeding and Wakefield damping of a set
of p-mode structures
•Each cell is fed through a directional coupler.
•The feeding port serve as the damping port
•The system has a four-fold symmetry, only one section is shown
4/12/2020
26
Parameter Options for a 1 TeV machine,
2 1034 cm-2 s-1
Option-1
Option-2
Frequency (GHz)
Option-3
11.424
Gradient (MV/m)
100
100
80
Power/meter (MW)
287
173
126
1
0.5
0.5
Charge/bunch (nC)
Bunch separation ( RF
cycles)
6
Number of bunches
300
600
600
<a/l>
0.14
0.1
0.1
Two beam
Two
Beam/Conventional
Unit
Conventional Unit
RF/Beam Efficiency(%)
73
53
60
Beam duration Pulse
length (ns)
158
315
315
Repetition Rate (Hz)
60
60
60
RF source
4/12/2020
27
Conclusion
• With recent advances on high gradient
accelerator structures, it is possible to think of
an “efficient” room temperature collider
• Reduced power levels/unit length may permit
designs based on conventional RF unit.
4/12/2020
28