Agenda
(Meeting 3/6/03)
J.W.
1. Testing structures for December 2002
High power test status: H90VG3N and H60VG3N (Chris).
3. Testing structures for May 2003
SW20a375F x 2 All stacks bonding completed.
Coupler subassemblies will be completed tomorrow.
Assemblies will be done by end of next weeks.
H60VG3S18 Universal coupler under assembly.
Assembly and QC two weeks after SW sections.
FXB structures (John)
4. Design and Study Issues:
•
2003 Structure R&D schedule.
•
H75VG4S18 coupler design (Chris, FNAL)
•
Design of H60VG3S17. (Zhenghai)
•
High band wakefield analysis for H60VG3 (Roger)
•
Pulse heating for different materials (Valery)
•
AOB.
Planning for Next Structures
Toward NLC/JLC READY Structures
Oct.
Nov.
Install
Install
Dec.
Jan.
Feb.
March
April
May
June
July
Aug.
3/05/03
Sept.
Oct.
Nov.
H90VG3N (0.18, 150°, no slots) 92 MW
FXB-002 (0.18, 150°, no slots) 73 MW
Install
KEK/SLAC
FNAL
H60VG3N (0.18, 150°, 6 slotted) 73 MW
SW20a375 x 2
Install
H60VG3S18 (0.18, 150°, slots, no HOM loads) 78 MW KEK/SLAC
(R1)
FXB-003 (0.18, 150°, no slots) 73 MW
FNAL
Assemble
H75VG4S18
Fabricate FXB 004-006
(H60VG3N18).
Complete design and
place cell order.
KEK/SLAC
Install
Delivery/Install?
Fabricate cells for
H75VG4S18.
Dec.
KEK/SLAC
Install H75VG4S18 (0.18, 150°,
KEK/SLAC
slots, no HOM loads) 86 MW
(R1)
Install CERNW (tungsten iris) 100 MW
CERN
Install FXB 004 (one structure) 73 MW
FNAL
H60VG3R17 (0.17, 150°, no slots) 67 MW
Complete H60VG3S17 design with
slots and HOM loads.
Schedule to complete?
SLAC
Planning for Next Structures
Toward NLC/JLC READY Structures
Oct.
Nov.
Dec.
Jan.
Feb.
March
April
May
June
July
Aug.
3/05/03
Sept.
Oct.
Nov.
Fabricate FXC 001-004 (H60VG3S17).
With HOM loads.
Fabricate HDDS 001-004 (H60VG3S17 or H75VG3S18).
With HOM loads.
Order H90VG5R cells for one
and two thirds structures.
H90VG5R-Moly (0.18, 150°, no slots)
Schedule??
??? H90VG5R-CLG (0.18, 150°, constant loaded gradient, no slots)
Dec.
FNAL
KEK/
SLAC
SLAC
SLAC
Matching of input coupler of
H75VG4
Output Coupler Match
R = ~0.015
|Ez| (relative)
|Es| (relative)
~16% enhancement
distance along axis (mm)
distance along surface (mm)
H75VG4S18 Output Match
~16% enhancement
b67 increased 3.8 µm
move iris back ~λ/2:
~1.28”
~6.4% enhancement
use thinner matching iris:
b67 decreased 9 µm
better
Final Output Match
R = 0.0074
HFSS simulation of the Input and Output
couplers for FXC structure (H75VG4S18)
Ivan Gonin, Nikolay Solyak
Fermilab
Procedure of couplers optimization
Creation of
Solid Model
Choice of
parameters for
optimization
C++ program for R
minimization (Kroll)
No
Yes
Mathcad program for
results analysis
FINISH
Solid model of INPUT coupler
H ~ 3.05E+5 A/m
T ~ 25º
Magnetic field in area of rounding at slot and b
Es max ~ 135Mv/m
iris of matching
cell
Electric field in input coupler and on surface
5.46
5.44
a match
5.42
8.26
8.28
8.30
8.32
8.34
8.36
8.38
8.40
a coupler
Final design
Path of reflection minimization
8.42
Average reflection in structure ~0.008, phase0 ~149º
1
1
0.8
0.6
f ( x)
15000 0.4
0.2
0.
0
100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10
100
180
π
⋅ arg( a k+ b k) ⋅ e
cj ⋅ k ⋅θs
 −78
x
0
0
Amplitude of electric field on axis in input coupler
5
4
3
2
1
0
1
2
3
4
5
5
(phase-phase0) along the structure
1
2
3
4
k
5
6
Cell number
Solid model of OUTPUT coupler
Surface electric field in first cell before
output coupler cell is ~5% bigger
Electric field in output coupler and on surface
Magnetic field on surface of output coupler, H~3.5e5, T~32º
Average reflection in structure ~0.02, phase0 ~148º
1
0.8
f ( x)
0.6
22000 0.4
0.2
0
100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10
5
0
x
Amplitude of electric field on axis in output coupler
10
8
6
4
2
180
cj
⋅
k
⋅
θs
 −98 0
⋅ arg( a k+ b k) ⋅ e
π
2
4
6
8
10
(Phase-phase0) along the structure
1
2
3
4
k
5
Cell number
6
NLC - The Next Linear Collider Project
H60vg3 Structures
h60vg3a17
Tfill
Tau
P
(70MV/m)
Max
Es/Ea
DLWG
111
0.559
64.7
2.2
ELLIP
117
0.604
66.1
2.0
ZL 2/27/2003
NLC - The Next Linear Collider Project
H60VG3A17 Elliptical - Dipole Dispersion
ZL 2/27/2003
NLC - The Next Linear Collider Project
BNS Damping Parameters (P.T., T.R.)
14
1.4
0.19
0.18
0.17
0.16
0.15
1.2
12
10
Emittance Growth, nm
RMS energy Spread, %
1
0.8
0.6
1σy betatron oscillation
8
6
4
0.4
2
0.2
0
0.19
0.18
0.17
0.16
0.15
0
0
1000
2000
3000
4000
s position, m
5000
6000
7000
0
1000
2000
3000
4000
s position, m
5000
6000
7000
30
0.19
φ1,
degrees
13
E1,
GeV
30
φ2,
degrees
-1
E2,
GeV
191
φ3,
degrees
-30
0.18
12
30
0
165
-30
0.17
14
30
0
162
-30
0.16
16
30
1
142
-30
0.15
19
30
4
113
-30
25
End−linac emittance growth, nm
a/λ=
0.19
0.18
0.17
0.16
0.15
20
15
10
5
0
0
1000
2000
3000
4000
position of misaligned quad, m
5000
6000
ZL 2/27/2003
NLC - The Next Linear Collider Project
RF Requirement With BNS
• Beam off crest for BNS
• Large phase in field at the end due to off phase
beam loading
• Need to Phase SLED to obtain a steady state field,
so all bunches see the same phase distribution
along structure
• Real part of field contributes to beam loading
compensation and acceleration
• Imaginary part of field contributes to energy slope
• Need larger amplitude at the beginning of ramp to
keep the real part the same, which depends on the
phase ramping
• Does not need extra power from klystron, utilize
otherwise the “waste” part during the ramp
• No efficiency loss!
• Ok to use average GRF and Gb for Gload calculation
ZL 2/27/2003
NLC - The Next Linear Collider Project
Acceleration And BNS Voltages
Just like in the bunch
compressor
ZL 2/27/2003
Next Linear Collider
Higher Band Wakefield Analysis for H60VG3
* 55 cell structure, ~60cm in length,150 deg phase adv/per cell
*Use the geometry as found from the energy dipersion beam loading calc.
*Take all a,b,t (thickness is detuned) and run mode matching
code on complete structure => mode impedance.
*Fit all the resonance in the impedance to lorentzians
=> synchronous freq and the area under each resonance
gives the kick factor.
*Obtain wake up to 35GHz.
How important are the higher band wakes?
~6% of the total wake is in the 3rd band.
Roger M. Jones, (Structures Meeting, SLAC, March 6, 2003)
1 (of 6), March 6, 2003 1:31 pm, T:\RMJ\Common Work + Home Files\FrameMaker\HDS_3_03.fm
é ωs
j ù
K n Exp ê j n (1 +
)]ú
2Q n û
n =1
ë c
N
å
(2.1)
where for the nth mode, Kn is the transverse kick factor,
ωn/2π is the synchronous frequency and Qn is the quality
factor of the mode. A modal expansion similar to eq.
(2.1) is also found in [8]. The kick factor is evaluated as:
2
Kn =
| ò E z Exp[ jωn s / c]dz |
L
(2.2)
vgn
ωn a
U n L(1 −
)
c
c
Here, an is the radius of the nth iris, L the periodic length
of the cell, Ez is the on-axis electric field and Un is the
energy stored per cell in a mode. The kick factor also
depends on the group velocity vgn [9] and provided the
synchronous phase is close to π then the group velocity
dependence will be a negligible correction and it can be
ignored.
For all DDS structures is has indeed been
found to be a small correction. However, for the SW
and new high phase advance structures this is no longer
2
n
50
40
1st Band
30
20
10
3rd Band
2nd Band
15
20
6th Band
4th Band 5th Band
25
30
Frequency (GHz)
35
Figure 1. Kick factors in a travelling wave accelerator:
DS1 (Detuned Structure). The complete set of 206 kick
factors is obtained by interpolation from the calculation of
kick factors for 5 cells (shown with dots). The largest
kick factors are all concentrated in the first band. The
third and sixth bands, although they are almost an order of
magnitude smaller than the first, also affect the beam
dynamics in the linac. All of these three bands must be
detuned.
the case. We calculated the kick factors and wakefields
for a representative TW detuned structure known as DS1
and for a 8 SW structures each of which consists of 15
cells and they are both detuned with a 10% bandwidth.
The 8 different SW structures effectively make up a 120
cells structure. We require 8 structures as the detuning
provided by 15 cells in one structure alone is insufficient.
The results of these calculations, performed with HFSS
and GdfidL [10], are shown in Figs 1 and 2 respectively.
For the standing wave structure the kick factors are no
longer linearly dependent on the synchronous frequency
15
12.5
1st Band
10
2nd Band
3rd Band
7.5
5
2.5
4th Band
5th Band
0
6th Band
15 17.5 20 22.5 25 27.5 30 32.5
Frequency (GHz)
Figure 2. Kick factors in standing wave accelerator
design SW1. The complete set of 120 kick factors is
obtained by interpolation from the calculation of kick
factors for 5 cells (shown with dots). The first three
bands kicks are of similar order of magnitude and they all
must be damped and detuned. The 4th and 5th bands are
an order of magnitude smaller than the first three but they
also must considered in a full analysis of the beam
dynamics as they also contribute to BBU instability.
Wakefield HVêpCêmmêmL
Kick factor HVêpCêmmêmL
4
and they are not concentrated in only the first band. The
wakefield that results from each of these bands for the
travelling wave detuned structure and the standing wave
structure is shown in Fig 3 and 4 respectively. In this
calculation we have not included the effects of the finite
group velocity of the dipole mode.
Kick FactorHVêpCêmmêmL
W(s) = 2
100
1
10
1
3
6
4
5
2
0.1
0.01
0
0.5
1
1.5
2
s (m )
Figure 3. Individual bands, ranging from the first to the
sixth of the envelope of the wakefield corresponding to
the kick factors of the travelling wave structure given in
Fig 1. The dots are positioned at the location of each
individual bunch (spaced by 42cm). Four out of a total
191 trailing bunches are shown
Beam dynamics studies [11] indicate that the wakefield
must be below unity in order that the BBU instability not
be an issue. The wake at the position of the bunches is
shown in Fig 3 for the travelling wave structure and it is
clear that the wake remains below unity at these locations
and thus BBU is unlikely to be a problem. The 3rd and
6th bands have significant kick factors compared to the
first band and these modes were detuned by enforcing and
1
2
10
1
Kick Factor VpCmmm
100
25
P.
A.
20
120
135
150
15
(a)
The arrows are in the
direction of increasing
P.A. (Phase Advance).
The P.A. is indicated
in degrees on the first
band
1st Band
165
170
10
180
P.A
5
P.A.
nd
2 Band
3rd Band
3
15
20
25
30
Frequency GHz
5
35
40
4
0.1
6
0.01
0
0.5
1
1.5
2
s (m)
Figure 4. Individual bands of the envelope of the
wakefield corresponding to the kick factors of the
standing wave structure given in Fig 2.
Kick Factor VpCmmm
Wakefield HVêpCêmmêmL
Erf variation to the iris thickness of all cells.
The
wakefield for the SW structure shown in Fig 4 reveals
that the first three band are all equally important and
consequently they all must be carefully damped. The
wake at the position of the first trailing bunch is below
unity for all bands apart from the third band. The third
band requires additional detuning in order to accelerate
the rate of decay of the wake.
3. GENERAL PROPERTIES
REGARDING BAND PARTITIONING
OF KICK FACTORS
35
30
25
20
15
120
135
P. 150
A.
(b)
1st Band
165
170
10
180
5
nd
P.A.
2 Band
15
Kick Factor VpCmmm
In order to assess the behaviour of band partitioning as
a function of synchronous frequency we used the Fortran
code Transvrs [12] driven with by a Mathematica input to
the data set to calculate the kick factors and synchronous
frequencies. The results of this calculation are shown in
Fig 5 for a/λ given by 0.229 (a), 0.19 (b) and 0.161 (c), in
which the kick factors are calculated for structures with a
phase advance ranging from 120 to 180 degrees. The
general trend for the first three bands is quite clear,
namely, rather independently of the iris dimension, the
second and third dipole bands are enhanced at the expense
of the first band as the phase advance per cell increases
from the initial value of 2π/3. The effect of finite group
velocity on the kick factor has been left to a later
publication as until recently the code was unable to
incorporate this effect. However, inclusion of the finite
group velocity does not modify our general conclusions
on the partitioning of modes.
In conclusion, the present NLC design limits a/λ ~0.18
and thus the first three bands of the SW structure will be
required to be damped and detuned. For the 5π/6 structure
only the first dipole bands must be damped and the cell
frequencies detuned together with the third band which
will be required to be detuned and moderately damped or
not damped at all. Further studies are in progress on
assessing the damping requirements of the 3rd dipole band
in the 5π/6 structure.
40
3rd Band
P.A.
20
25
30
Frequency GHz
35
40
35
40
60
50
40
120
135
P. 150
A.
30
20
10
165
(c)
st
1 Band
170
P.A.
180
.
P.A
3rd Band
2nd Band
15
20
25
30
Frequency GHz
Figure 5. Kn as a function of synchronous frequency for
several irises radi: (a) 6mm, (b) 5 and (c) 4.23 mm.
4. REFERENCES
[1] R.M. Jones et al, PAC99, also SLAC-PUB-8103
[2] R.M. Jones et al, EPAC96, also SLAC-PUB-7187
[3] C. Adolphsen et al, 1997, SLAC-PUB-7519
[4] J.W. Wang et al, TH464 these proceedings
[5] T. Shintake, PAC99, also KEK-PREPRINT-99-11
[6] K.LF. Bane & R. Gluckstern, Part.Accel.42:123
-169,1993
[7] R.M. Jones et al, LINAC96, also SLAC-PUB-7287
[8] K.A. Thompson et al, Part.Accel.47:65-109,1994
[9] A. Millich and L. Thorndahl, 1999 CLIC-NOTE-366
[10] W. Bruns, PAC97, also TET-Note 97/07
[11] R.M. Jones, TH465 these proceedings
[12] B. Zotter & K. Bane, 1980, CERN ISR-TH/80-25
Next Linear Collider
400
350
1st Band
Real Impedance of H60VG3
11
300
10
9
8
200
7
150
6
5
100
3rd
50
4th
0
2nd
15
10
20
30
40
50
60
5th
200
20
25
Freq. GHz
30
35
150
Imaginary Impedance
100
50
ImZ
ReZ
250
0
50
100
150
15
20
25
Freq. GHz
Roger M. Jones, (Structures Meeting, SLAC, March 6, 2003)
2 (of 6), March 6, 2003 1:31 pm, T:\RMJ\Common Work + Home Files\FrameMaker\HDS_3_03.fm
30
35
Next Linear Collider
300
60
First Band Re{Z}
250
Kick Factor
Weighted
Density Func:
Kdn/df
50
Kdndf VpCmmmGHz
ReZ
200
150
100
50
0
14
14.5
15
Freq. GHz
100
15.5
40
30
20
10
16
0
First Band Wakefield Envelope
Wake VpCmmm
10
14
14.5
(Q~1000 for all modes)
15
Freq. GHz
1
0.1
0.01
10
20
30
40
s m
50
60
70
80
Roger M. Jones, (Structures Meeting, SLAC, March 6, 2003)
3 (of 6), March 6, 2003 1:31 pm, T:\RMJ\Common Work + Home Files\FrameMaker\HDS_3_03.fm
15.5
16
Next Linear Collider
10
0.02
Kick Factor VpCmmm
Second Band Impedance
6
4
2nd Band Kick Factors
0.015
0.01
0.005
2
100
0
16.5
17
17.5
18
Freq. GHz.
18.5
19
19.5
17
20
18
19
Frequency GHz
20
10
Wake VpCmmm
ReZ VpCmmmGHz
8
2nd Band Wakefield Envelope
(two orders of magnitude smaller than 1st)
1
(Q~4000 imposed on all modes)
0.1
0.01
10
20
30
40
s m
50
60
70
Roger M. Jones, (Structures Meeting, SLAC, March 6, 2003)
4 (of 6), March 6, 2003 1:31 pm, T:\RMJ\Common Work + Home Files\FrameMaker\HDS_3_03.fm
80
21
22
Next Linear Collider
Third Band Synchronous
Freqs.
50
Third Band Impedance
40
30
Cell Num.
ReZ VpCmmmGHz
40
30
20
20
10
10
0
0
23
100
23
24
25
Freq. GHz.
3
26
25
Frequency GHz
Q~4000 imposed
0.1
0.01
27
Kick Factor
Weighted Density
Function
Third Band Wake Env. (~6 of first band at s=0)
1
26
27
2.5
KdndfVpCmmmGHz
Wake VpCmmm
10
24
2
1.5
1
0.5
0
10
20
30
40
s m
50
60
23
70
80
24
25
Frequency GHz
Roger M. Jones, (Structures Meeting, SLAC, March 6, 2003)
5 (of 6), March 6, 2003 1:31 pm, T:\RMJ\Common Work + Home Files\FrameMaker\HDS_3_03.fm
26
27
Next Linear Collider
Real Impedance for
4th and 5th Bands
30
20
Kick Factors
0.1
Kick Factor VpCmmm
ReZ VpCmmmGHz
40
0.08
0.06
0.04
10
0.02
100
0
0
28
30
28
34
30
32
34
Frequency GHz
Freq. GHz.
10
Wake VpCmmm
32
Wake Envelope
1
(Q~4000 imposed)
0.1
Wake of 3rd band at s=0 is ~6%
that of the 1st band.
Combined wake of the 4th and 5 th
bands is ~3% of the 1st band
0.01
10
20
30
40
s m
50
60
70
80
Roger M. Jones, (Structures Meeting, SLAC, March 6, 2003)
6 (of 6), March 6, 2003 1:31 pm, T:\RMJ\Common Work + Home Files\FrameMaker\HDS_3_03.fm
Pulsed heating of TW structures
made of different materials
V.A.Dolgashev, 6 March 03
Pulsed temperature rise for different materials for
magnetic field 0.19 MA/m, and pulse width 400 ns
1 .10
3
320 deg. C
Temperature [deg.C]
100
39 deg .C
11 deg. C
10
1
0.1
0.01
0
10
20
30
Depth [um]
40
50
copper
molybdenym
stainless steel
V. Dolgashev, 6 March 03
Comparison of surface electric and magnetic fields for first
cell of T53VG3 and H60VG3 for 100 MW of input power
250
E [MV/m], H*Zo [MA/m]
200
150
100
50
0
0
2
4
6
s [mm]
8
10
E, T53VG3
H, T53VG3
E, H60VG3
H, H60VG3
V. Dolgashev, 26 February 03
250
1000
200
200
800
150
100
50
0
E [MV/m], T [deg. C]
250
E [MV/m], T [deg. C]
E [MV/m], T [deg. C]
Comparison of surface electric and pulsed temperature rise
for first cell of T53VG3 and H60VG3 for 100 MW of
input power
150
100
50
0
1
2
3
4
5 6 7
s [mm]
8
9 10 11
E, T53VG3
T, T53VG3
E, H60VG3
T, H60VG3
Copper, T yield ~25 deg. C
0
600
400
200
0
1
2
3
4
E, T53VG3
T, T53VG3
E, H60VG3
T, H60VG3
5 6 7
s [mm]
8
9
10 11
0
0
1
2
3
4
5 6 7
s [mm]
8
9 10 11
E, T53VG3
T, T53VG3
E, H60VG3
T, H60VG3
Molybdenum, T yield ~ 260 deg. C SS304, T yield ~41? deg. C
V. Dolgashev, 6 March 03