ILC Main Linac Simulation with
Survey/Alignment model
model,
update 200805
200805 K.Kubo
Recent Progress
• Bug
g in the simulation code SLEPT was fixed
– Setting of initial vertical angle was not correct for
reduced energy beam in DFS (Not fatal. Only affected
results
lt with
ith very smallll emittance
itt
growth.)
th )
• New survey-alignment model was included
– Survey (long range) + Random alignment (local)
– Primary references (every 2.5 km) are used for
“parabola correction” of accumulated error Æ see
later slides
• Look at sensitivity to each error
• Simulation
Si l ti with
ith suggested
t d error sett b
by LiCAS group
Alignment procedure
Every 2.5 km, primary references,
? using GPS? Random error.
Survey from one primary reference to the next.
Every about 5~50 m, mark reference point
Girders, cryomodules, etc. are aligned w.r.t. the reference.
Applied
to
tracking
simulation
i l ti
Not yet
applied to
simulation
Step by step survey:
Random Walk + systematic angle error
random offset
Parameters:
random angle + systematic angle:
with respect to the previous step
lstep : length of one step
a y : random offset/step
aθ : random angle
g error/stepp
θO : systematic angle error
Systematic offset error
was treated as “initial
g error. See later
angle”
slides.
Correction of accumulated error in Random Walk
using
g primary
y reference
primary reference -1
primary reference -2
design
g line
correct accumulated error
random walk from 1
Offset proportional to distance from ‘1’
This simple linear correction makes kinks at primary references and
is not a good choice. We chose parabola correction here.
(Correction is proportional to square of distance.)
Correction of accumulated survey line
error using
i primary
i
references
f
• Linear correction
– Correction proportional to distance from the
start point.
– Causes kinks at primary reference.
• Parabola correction: We have chosen this
– Correction proportional to square of distance
from the start point.
p
– No kinks.
• Other methods (?)
Example: Comparison of correction of accumulated error
Offset errror of com
mponents ((m)
Spacing of primary references: 2500 m, Error of primary reference: 0
Step length of survey (random walk): 50 m
Offset error /step, ay = 0, Angle error/step, aθ = 1 μrad
0.005
0
-0.005
No correction
Linear correction
P b l correction
Parabola
i
-0 01
-0.01
-0.015
0
1 10
3
2 10
3
3 10
kink ?
3
s (m)
4 10
3
5 10
3
6 10
3
Survey line to component alignment
alignment,
Alignment model w.r.t. reference points
(
(example)
l )
use several points to make a line
least square fit
reference points
offset
girder/cryomodule/magnet
tilt angle
Sensitivity to each error parameter
off survey line
li
• Step
p length
g of random walk is chosen as 50 m
• Set one error for each simulation run
–
–
–
–
–
Random angle/step, aθ
Random offset/step, ay
Error of primary references (every 2.5 km)
Systematic angle error θ0 (error of vertical curvature)
Error of initial angle
• DFS (Dispersion Free Steering) is performed
• Looked at final “linear η corrected” emittance
Sensitivity to Random Angle and Random
Off t / step
Offset
t after
ft DFS
Average of 40 random seeds
Angle, no other error
Offset, no other error
-8
3 10
-8
3 10
No correction
Linear Correction
Parabola corection
-8
2.8 10
η corrected
d γε (m)
η correctd γγε (m)
2.8 10
No correction
N
ti
Linear correction
Parabola correction
-8
-8
2.6 10
-8
2.4 10
-8
2.6 10
-8
2.4 10
-8
-8
2.2 10
-8
2 10
2 2 10
2.2
-8
2 10
0
0.5
1
a (μrad)
θ
1.5
2
0
5
10
a μm)
y
15
20
Sensitivity to Offset error of Primary
R f
References
after
ft DFS
Average of 40 random seeds
-8
2.5 10
Linear correction
Parabola correction
-8
y
η correctted γε (m)
2.4 10
-8
2.3 10
-8
2.2 10
-8
2.1 10
2 10
-8
0
10
20
30
40
50
60
Error of Primary References (mm)
70
Systematic angle error
η corrrected γε (m)
There should be no effect in case of Parabola correction.
Results in the case of No correction are shown for reference.
(Li
(Linear
correction
ti h
has h
huge sensitivity
iti it and
d nott shown
h
h
here))
2.7 10
-8
2.6 10
-8
2.5 10
-8
2.4 10
-8
2.3 10
-8
2.2 10
-8
2.1 10
-8
2 10
-8
No accumulated error correction
step
t length
l th 50 m
-60
-40
-20
0
20
40
60
Systematic angle error (nrad/m)
y=
(angle error / length ) 2
s
2
10 nrad/m causes 0.5 m offset at 10 km
Initial angle error
No effect in cases of Linear correction and No correction.
Only results of Parabola correction are shown.
Initial angle error is equivalent to systematic offset error.
To adjust orbit-alignment angle, offset of Primary References were
also set in simulations (assuming we can correct injection angle).
0.4
-8
2.5 10
Parabola correction of
accumulated error
Initial Angle Error 0.1 mrad
0.2
η ccorrected γε (m
m)
-8
y (m)
0
-0.2
-0.4
-0.6
No correction
Parabola correction
Used in simulation
-0.8
2.4 10
-8
2.3 10
-8
2.2 10
-8
2.1 10
-1
-8
2 10
0
2 10
3
3
4 10
3
6 10
s (m)
8 10
3
1 10
4
-100
-50
0
50
Initial Angle Error (μrad)
100
“Tolerances” in the case of Parabola correction
(looks the best among three cases)
cases),
Error causing 0.2 nm (1% of nominal) average emittance growth
R d
Random
walk
lk step
t llength
th 50
50m
Error
for1% Δε
Random angle
g / step
p
300 nrad
Random offset / step
3 um
Offset of Primary References
15 mm
Initial angle
15 urad
rad
Include all survey-misalignment
survey misalignment errors
• Survey parameters suggested by LiCAS Group (or Armin
Reinchold)
– Step Length: 25 m, Random angle: 60 nrad/step,
– Random offset: 5 mm/step, Systematic angle: 250 nrad/step,
– Primary reference: 10 mm
• Set “standard” local misalignment w.r.t. the survey line.
Error
Quad Offset
Cold Sections
With Respect To...
300 μm
Cryomodule
300 μrad
Gravity
RF Cavityy Offset
300 μ
μm
Cryomodule
y
RF Cavity Pitch
200 μrad
Cryomodule
BPM Offset (initial)
300 μm
Cryomodule
Cryomoduloe Offset
200 μm
Survey Line
Cryomodule Pitch
20 μrad
Survey Line
Quad roll
Example
p of misalignment
g
in ML
Step Length: 25 m, Random angle: 60 nrad/step,
Random offset: 5 μm/step, Systematic angle: 250 nrad/step,
Primaryy reference: 10 mm
(Suggested by
LiCAS Group)
+ “Standard” local misalignment
0.03
y off allll quad
d magnets
t
y (m)
0.02
0.01
0
Survey Line
Total
Total - Survey
-0.01
-0.02
0
3
2 10
3
4 10
3
6 10
s (m)
3
8 10
1 10
4
ML simulation with misalignment
Mean of emittance and standard deviation from 40 random seeds.
(initial emittance is 2E-8 m)
<Δγε> (m)
STD
Survey
0.053E-8
0.052E-8
Local misalignment
0.670E-8
0.581E-8
Survey + local
0.673E-8
0.591E-8
Suggested set of survey line errors by LiCAS group (?)
is good enough.
SUMMARY
• Main Linac simulations were performed
p
including a survey-alignment model
• Tolerances are estimated in the case the
random walk step length 50 m
m.
– They should depend on step length. Will be studied .
• Simulation with the error set suggested by
LiCAS group was performed.
f
d
– No significant emittance growth due to the survey
errors.
• The validity of the model and the parameters
should be checked.