Alexej Grudiev BERF Acknoledgements Rolf for useful information about present design R Wegner and E JensenCLICnote945 2012 Present CDR design Present CDR design Full beam loading in linear vg tapered structure ID: 409344
Download Presentation The PPT/PDF document "New RF design of CLIC DB AS" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
New RF design of CLIC DB AS
Alexej Grudiev, BE-RFSlide2
Acknoledgements
Rolf for useful information about present design (R. Wegner and E. Jensen,CLIC-note-945, 2012)Slide3
Present CDR designSlide4
Present CDR designSlide5
Full beam loading in linear vg tapered structure
Q
0
=
const
, R’=
const
but v
g = vg0(1+az)
Full beam loading condition for a≠-2
α
:
Efficiency in steady-state: Slide6
Full beam loading in Constant impedance TWS
Efficiency in steady-state:
Full beam loading condition:
In order to maximum efficiency means maximum relative beam loading parameter at the full beam loading condition: => Maximum
R’*vg*Q0 Slide7
Transfer function Pin -> V
i.e. PRST-AB
14
, 052001 (2011)
Const
impedance (CI):
-
Transfer function from the input gradient G0~sqrt(Pin) to voltage V
Linear tapered vg:
In addition, there is time of flight effect.
t_flight
~ 3-6 ns for 1-2 m long structure
t_f
’ =>
t_fill
-
t_flight
in FWS and
t_f
’
=>
t_fill
+
t_flight
in BWS
BWS enhance filtering effect of accelerating structureSlide8
Optimization of the 4-spoke cell: 120 degree, a=30mm
dzSpoke
=10mm; even spokes
Odd spokes
even spokes
dzSpoke
=10mm; odd spokes
eta2=R’*vg*Q0
dzSpoke
=15mm; even spokes
dzSpoke
=h-gap~27mm; even spokes
dzSpoke
=15mm; odd spokes;
P0=15MW ->
eta = 98.6
%;
Ls
= 2.8m;
tfill
=102ns; G0=2.5MV/m
RRSpWall
20mm
25mmSlide9
Optimization of the 4-spoke cell:
150
degree
dzSpoke
=15mm; odd spokes
dzSpoke
=25mm; odd spokesSlide10
Optimization of the 4-spoke cell: 9
0
degree
dzSpoke
=10mm; odd spokesSlide11
Optimization of the 4-spoke cell, odd spokes:
120 degree, a = 25mm
Odd spokes
dzSpoke
=15mm;
dxSpoke
=30mm ;
dRdt
=10mm
;
odd spokes;
P0=15MW -> eta = 98.7%; Ls
= 2.5m; tfill=102ns; G0=2.83MV/m
dzSpoke
=15mm;
dxSpoke
=40mm ;
dRdt
=10mm
;
odd spokes;
dzSpoke
=15mm;
dxSpoke
=30mm;
dRdt
=12mm; odd spokes; Slide12
Optimization of the 4-spoke cell, odd spokes:
120 degree, a = 20mm
Odd spokes
dzSpoke
=15mm;; odd spokes;
P0=15MW -> eta = 98.7
%;
Ls
= 2.34m;
tfill
=99ns; G0=3MV/m
dzSpoke
=19mm;; odd spokes; Slide13
Cell and CI structure parameters summary table
dphi
[
deg
]
a [mm]
Q
R’/Q [
Ω
/m]
vg/c [%]η_CI[%], @15MW
Ls [m]t_fill [ns]
12030145651846
9.2398.62.8102
15030157221581
4.59
90
30
13790
1088
~0
120
25
15602
2085
8.18
98.7
2.5
102
120
20
15821
2300
7.9
98.7
2.34
99
This can be improved by changing the cell shape.Slide14
Geometry of the BWSSlide15
Tapering vg
Structure with
a=30mm;
R
’/Q0 = 1846 Ohm/m; Q0 = 14565
P0=15MW; => G0=2.5MV/m
Const
Imp.; vg/c = 9.23%; solid lines
eta = 98.6
%; Ls = 2.8m; tfill=102ns; Linear vg/c from 9.23% to
2.9%. Dash lineseta = 98.4%; Ls = 2.2m; tfill=134ns; Const Imp.; vg/c =
5.6%; dash-dotted lineseta = 98.1%; Ls = 2.2m;
tfill=131ns; Slide16
Making filling time 245 ns
Structure with
a=30mm;
R
’/Q0 = 1846 Ohm/m; Q0 = 14565
P0=15MW; =>
Const
Imp.; vg/c = 1.54%; solid lines
eta = 96.5
%; G0=6.1MV/mLs = 1.13m; tfill=246ns; Linear vg/c from 2.6% to 0.908%. Dash lines
eta = 97.0%; Ls = 1.18m; tfill=244ns; Slide17
Material: why not Aluminium
CLIC DB Acc. Structure
with a=30mm;
vg/c = 9.23%; R
’/Q0 = 1846 Ohm/m;
Q0Cu
= 14565
P0=15MW; => G0=2.5MV/m
Influence of conductivity on efficiency:
conductivity
= 100 %IACS (Cu)
-> eta = 98.6%conductivity = 60 %IACS (Al) -> eta = 98.1%
conductivity = 40 %IACS (6061-T6) -> eta = 97.7%
Issues: Multipactor due to high secondary emission yield of Al. More studies are needed to address this issue both simulations and high power testing.RF and vacuum wise tight assembly. Prototyping is necessary.
Material used: Al 6082-T6, conductivity: 35MS/m = 60%IACSSlide18
Advantages of high vg BWS
High
Rf
-to-Beam efficiency 99%
High group velocity -> less tolerance -> no tuning -> lower cost
V
ariation of group velocity by magnetic coupling hole size is independent on the aperture radius -> max(R’) and max(vg) at the same timeSlide19
Questions
Aperture: the smaller the better for RF, lower limit comes from beam dynamics. Is a=30 mm acceptable? Avni is looking into it.
Filling time
t_fill
? How critical is to have factor 10 noise reduction at f=1/
t_fill
=4MHz ? Is factor 2 or 4 instead acceptable?
Const
impedance versus tapered?
Tapering can be done but , for the same filling time, it will reduce the lowest vg -> tighter tolerancesIncrease nose level at f_n=n/
t_fillArgument for the tapering aredetuning of the HOMs but with the bunch spacing of 2 buckets it has small effect. We must rely on the strong damping anyway.IF higher average gradient od longer filling time is needed reduction of vg linearly along the structure provides slightly higher efficiency compared to overall reduction of vg in CI structure
Gradient? The higher is the gradient the lower is the linac cost per MV, IF real estate gradient is limited by the structure and not by the RF power source layout. It seems that 2 times shorter DB linac still can be done by rearranging the RF power sources along the
linac. To be discussed…Slide20
Next steps
Minimum aperture limit to be defined by our beam dynamics experts
For this minimum aperture a detailed RF design to be done to maximize vg, R’ and Q0
If higher gradient needed vg can be tapered down or/and reduced
Design HOM damping (same type as for CDR)
If Aluminium then
Multipactor
studies which is probably main limitation for Al cavity
Prototypes of a few cells for fabrication/assembly studies
Full or half length prototype to be tested at a L-band test facility