1D-FEL Without Approximations
description
Transcript of 1D-FEL Without Approximations
![Page 1: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/1.jpg)
1D-FEL Without Approximations
motivation, capabilities
1D theory 1D-solver for waves
implementation (without and with Lorentz transformation)
excitation of waves (single particle)
without self effects
one and few particles with self effects
mystery
![Page 2: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/2.jpg)
Motivation, CapabilitiesFEL codes use many approximations, as averaged equation of motion,local periodical approximation, EM-field calculation by paraxial approximation,one or several harmonics
these approximations are questionable for ultra-short bunches or buncheswith extreme energy modulation
it is easy to implement complete FEL effects in 1D
1D model can be used to verify approximations
capabilities of 1D model: ultra broadband (not split into harmonics, excitationand radiation)
no local periodic approximation
complete 1D field computation
seems possible: particle = macro particle
LT method can be tested
![Page 3: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/3.jpg)
1D TheoryEM Fields
txxzy
tyzx
EcJB
BE
,2
,
,,
c) longitudinal self field
a) undulator field
b) external wave
zzvA
qtzJ xx ,,
2
2
RLcB
RLE
y
x
tctVJL
tctUJR
t
t
,~
,~
1
1
tVLtctVL
tURtctUR
,~
,
,~
,
with
d) transverse self field
zzA
qtzzEz ,
no principle problem, but neglected
L and R are waves to theleft and to the right
![Page 4: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/4.jpg)
waves to the left and to the right
![Page 5: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/5.jpg)
Equation of Motion
f
v
p
r
dt
d
202
0 cmpc
m
ppvwith
BvEf 0q
horizontal plane, only x and z components:
Coupled Problem
tctUJ
tctVJ
tUR
tVLdt
d
,
,
,~
,~
1
1
f
v
p
r
zzvA
qtzJ xx ,, binning and smoothing (on equi. mesh)
PDE solver (f.i. RK4)
needs spatial resolution
needs time resolution
LR
LRLRqBv
BvEq
F
F
x
z
yx
yzx
z
x~~
~~~~
20
e,
e,e,0
external self
Implementation
![Page 6: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/6.jpg)
without Lorentz transformation
with Lorentz transformation
integration of PDE (by rk4) needs small time step for left wave
criterion: slip between source and wave < dz/c
dt < dz/c, number of time steps ~ undulator length / photon wavelength !!!
solution a: neglect L, 24
maxmax
R
L
solution b: the part of the left wave, seen by the bunch, is determined by near interaction; use tvzJtzJ xx ,,
solution c:
differenced between length scales are shrunk
huge external fields (from undulator)
it is possible, it is applicable even in 3D!
uyuy
uxB
c
B
E,
LTLT
,
,
1'
'
same magnitude of left and right wave
![Page 7: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/7.jpg)
intense fwd. wave !weak backward wave
cm 6w nm 775 nm, 773 nm, 77 w
Excitation of Waves (single particle)left and right wave
RL
![Page 8: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/8.jpg)
less power in fwd. direction !
higher harmonics:
2
2
12
2
2424
12
K
mKJ
K
mKJJJ
nm
nn
left and right wave with Lorentz transformation
mm 67.0w
L R
![Page 9: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/9.jpg)
Bu = 1Tu = 3 cmE = 500 MeVNu = 10K = 2.80w = 77 nm
before undulator
Example: Without Self Effects
zp
![Page 10: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/10.jpg)
start
same parameters,with Lorentz transformationframe = initial velocity
E/E0 = 978 = LT
zp
![Page 11: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/11.jpg)
same parameters,with Lorentz transformationframe = av. velocity start
av = 441 = LT
zp
![Page 12: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/12.jpg)
Why are left and right waves asymmetric?
L R
![Page 13: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/13.jpg)
One and Few Particles with Self effectsparameters as before
left wave out of window
pictures from solutions (a) and (b) cannot be distinguished by eye!
excitationof waves
longitudinal momentumleft and right waves
particle after undulator
L
R
zp
![Page 14: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/14.jpg)
parameters as before, with Lorentz transformation to frame = av. velocityparticle after undulator
zp
zp
L
R
![Page 15: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/15.jpg)
transformation back to lab-frame
parameters as before, with Lorentz transformation to frame = av. velocity
direct calculation in lab frame:
9600 time steps
2700 time steps
zpzp
![Page 16: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/16.jpg)
parameters as before, with Lorentz transformation to frame = av. velocitytwo particles, separated by one photon wavelength
LT^-1
particle 2
particle 1
-1
-2.8
left wave vs. z:
right wave vs. z:
zp
zp
L
R
![Page 17: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/17.jpg)
parameters as before, with Lorentz transformation to frame = av. velocitytwo particles, separated by half photon wavelength
left wave vs. z:
right wave vs. z:
particle 2
particle 1
LT^-1
R
L
zp
zp
![Page 18: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/18.jpg)
Mystery
In the frame “av. undulator velocity” the energy loss to both waves (left and right) is about equal. It seems the effect from both waves to the one-particle dynamic is similar.
In the rest frame the effect of the left wave seems negligible.
What happens if we neglect the left wave in the frame “av. undulator velocity”?
![Page 19: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/19.jpg)
example as before, with Lorentz transformation to frame = av. velocity
tctUJtUR
tVLdt
d
,
0
,~
,~
1
f
v
p
rno stimulation of left wave:
LT^-1L
R
zp zp
for comparison: complete calculation
L
R
zp zp
![Page 20: 1D-FEL Without Approximations](https://reader033.fdocuments.net/reader033/viewer/2022050819/56813c2a550346895da5a4f9/html5/thumbnails/20.jpg)
Why is L negligible in the frame “av. undulator velocity”?