Study The Integro-Differential Equation in A Lorentzian Energy Spread Case

Eng. & Technology, Vol.25, Suppl. of No.
2, 2007
Study the Integro-Differential Equation in a Lorentzian

Energy Spread Case
Dr. R. A. Khamis*, H. M. Daoud**, M. Z. Al-Faiz** & M. I. Sanduk**

Received on: 30/5/2005
Accepted on: 26/11/2006
Abstract
The fundamentals of free electron laser (FEL) theory are now well–established
and can provide a sophisticated description of experiments over a wide range
of parameters. While new technology is being developed for systems working
from 1mm to 10nm wavelengths, the theory remains the same. In this work,
the final consequences of Lorentzian energy distribution in FEL amplifier will
be found by solving of Integro-differential equation using symbolic toolbox
services in MATLAB software v. 6.5 and the relation between several
parameters will be delimited and plotted in order to extract and present the
complex relations between the FEL parameters.
Keyword: Evolutionary programming and Symbolic Analysis, Mathematical

Modeling, FEL Amplifier Process
‫ﺍﻟﺨﻼﺼﺔ‬
‫ ﺘﺼـﻑ ﺍﻟﺘﺠـﺎﺭﺏ‬‫ﻜِﻨﹸﻬـﺎ ﺃَﻥ‬‫ﻤ‬‫ ﺃﺼﺒﺤﺕ ﺍﻻﻥ ﻤﻌﺭﻓﺔ ﻭﻴ‬‫ﺭ‬‫ﺍﻥ ﺃﺴﺎﺴﻴﺎﺕ ﻨﻅﺭﻴﺔِ ﻟﻴﺯﺭِ ﺍﻷﻟﻜﺘﺭﻭﻥ ﺍﻟﺤ‬
ُ‫ﻤل‬‫ﺭ ﻟﻸﻨﻅﻤﺔِ ﺍﻟﺘﻲ ﺘﹶﻌ‬‫ ﺒﻴﻨﻤﺎ ﺍﻟﺘﻘﻨﻴﺎﺕ ﺍﻟﺤﺩﻴﺜﺔ ﻗﺩ ﺒﺩﺍﺕ ﺘﺘﹸﻁﻭ‬.ِ‫ﺍﻟﻤﻌﻘﹼﺩﻩ ﻟـﺘﺸﻜﻴﻠﺔ ﻭﺍﺴﻌﺔ ﻤﻥ ﺍﻟﻌﻭﺍﻤل‬
.‫ﻘﺕ ﺍﻟﻨﻅﺭﻴﺔﹸ ﺒﺩﻭﻥ ﺘﻐﻴﻴـﺭ‬‫ﺒ‬‫ ﻭﻤﻊ ﺫﻟﻙ ﻴ‬، ‫ ﻤﻠﻴﻤﺘﺭِ ﻭﺍﺤﺩ ﺇﻟﻰ ﻋﺸﺭﺓ ﻨﺎﻨﻭﻤﺘﺭ ﻤﻥ ﺍﻟﻁﻭلِ ﻤﻭﺠﺔ‬‫ﻤِﻥ‬
‫ـﺘﹶﻭﺠِﺩ‬‫ ﺴ‬‫ﺭ‬‫ ﺍﻟﻨﺘﺎﺌﺞ ﺍﻟﻨﻬﺎﺌﻴﺔ ﻟﺘﻭﺯﻴﻊِ ﻁﺎﻗﺔِ ﻟﻭﺭﻨﺯﻴﺔ ﻓﻲ ﻤﻀﺨﹼﻡِ ﻟﻴﺯﺭِ ﺍﻷﻟﻜﺘﺭﻭﻥ ﺍﻟﺤ‬،ِ‫ﻓﻲ ﻫﺫﺍ ﺍﻟﻌﻤل‬
‫ ﺍﻟﺘﻔﺎﻀﻠﻴﺔِ ﺤﻴـﺙ ﺍﹶﺴـﺘﻌﻤﻠﺕ ﺨﺼـﺎﺌﺹ ﻭﻤﻴـﺯﺍﺕ ﻟﻠﻐـﺔ ﺍﻟﺒﺭﻤﺠـﺔ‬- ‫لّ ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻟﺘﻜﺎﻤﻠﻴﺔ‬‫ﺒﺤ‬
‫ﺓ‬‫ ﻋﻼﻭﺓ ﻋﻠﻰ ﺫﻟﻙ ﺘﻡ ﻭﺼﻑ ﺍﻟﻌﻼﻗـﺔِ ﺒـﻴﻥ ﻋِـﺩ‬.‫( ﻓﻲ ﻫﺫﺍ ﺍﻟﺒﺤﺙ‬MATLAB software v. 6)
.‫ﺭ‬‫ﻋﻭﺍﻤل ﻷﺴﺘﺨﻼﺹ ﻭﺘﻤﺜﻴل ﺍﻟﻌﻼﻗﺎﺕﹶ ﺍﻟﻤﺭﻜﹼﺒﺔﹶ ﺒﻴﻥ ﻋﺩﺩ ﻤﻥ ﻋﻭﺍﻤلِ ﻟﻴﺯﺭِ ﺍﻷﻟﻜﺘﺭﻭﻥ ﺍﻟﺤ‬
1 Introduction To close the problem, the field

The problem of electromagnetic wave equations and equations of motion
amplification in the undulator refers to should be solved simultaneously. In
a class of self-consistent problems [1]. principle, modern super-computers
It can be separated into two parts: allow one to perform direct simulation
Solution of the dynamical problem of the FEL process. The results of
and Solution of the electrodynamics' such simulations depend on a large
problem. number of problem parameters. They
provide the possibility of obtaining a
Applied Science Dept., UOT., Baghdad, IRAQ drraad2001@yahoo.com
Opto-Electronics & Laser Eng. Dept., Nahrain University, Baghdad, Al-Jadrea, IRAQ.
scientific@issambg.com
291
Eng. & Technology, Vol.25, Suppl. of No.2, 2007 Study the Integro-differential Equation in
a Lorentzian Energy Spread Case
numerical answer for a specific set of 2 Theory

input data. The description of the Free Electron
A deeper insight into FEL physics can Laser given here is based on Kim [1]
be obtained only by introducing some and Reiche [2]. A more detailed
simplifying assumptions about the description can be found in [3] and [4]
properties of the electron beam and of and the references therein.
the electromagnetic field. The treatment in this search will be
Theoretical investigation of the free considered the case of an electron
electron laser should be performed in beam with Lorentzian energy
two stages. In the beginning one distribution as [2]:
should study the general properties of 1 ΛT
the FEL, namely the ideal mechanism F (ε − ε o ) = (1)
π (ε − ε o ) 2 + Λ2T
of amplification. At the next stage
different complications can be where:
introduced into the FEL model F (ε − ε o ) is the distribution function
allowing it to extend the number of of the canonical momentum.
additional effects influencing the Λ T is the normalized energy spread
operation of the FEL amplifier. These
parameter.
factors can be divided into two
ε is the energy, and
groups. The space charge effects and
diffraction effects are fundamental. ε o is the nominal energy of an
On the other hand, there are a lot of electron.
other factors such as energy spread The corresponding expression for the
effects, non-ideality of the undulator reduced distribution function F̂ has
field, etc. the principal difference of
the form (see Figure (1));
the fundamental effects and all the
others is that the fundamental effects 1 Λ ˆ
Fˆ ( Pˆ ) = T
(2)
depend on the same physical π Pˆ 2 + Λ ˆ2
T
parameters as the ideal FEL where
amplification mechanism itself.
The one-dimensional model is an
important one from the Pˆ = (ε − ε o ) / ( ρ ε o ) ,
methodological point of view. The
following assumptions are used in the is the normalized energy deviation.
one-dimensional, steady-state model with a normalized energy spread
of the FEL amplifier for linear region: Λ
ˆ = Λ /(ε ρ )
T T o
- The electron beam has Gaussian
and
density distribution in the direction
perpendicular to the undulator axis;
- The electron moves along identical ΛT = (∆ ε )2 ,
trajectories parallel to the undulator
axis; which is at least an approximation for
- The amplified wave is a the more realistic Gaussian
monochromatic plane wave; distribution.
- The electron beam is infinitely long.
292
cubic equation (in normalized form)

can be written as equation (5 ):
Fˆ (Pˆ )
((λ − i Λ
ˆ + Cˆ ) 2 − Λ
T
ˆ 2 ) λ + 1 = 0 , (5)
p
The general solution of above

equation can be written as follow:
13 2ˆ 2
λ1 = α − 6ϑ + i Λ T − Ĉ (6a)
6 3 3
1 3 2ˆ 2
Λ̂ T λ 2,3 = − α + 3ϑ + i Λ T − Ĉ
P̂ 12 3 3
Figure (1)
Distribution function of the reduced 3  13 
±i  α + 6ϑ  (6b)
canonical momentum Fˆ (Pˆ ) versus reduced 2 6 
canonical momentum P̂ and energy spread
parameter Λ̂ T . where;
Λ
ˆ 2 + i 2 Ĉ Λˆ − Ĉ 2 − 3 Λ̂ 2
ϑ =
T T p
Substituting the Lorentzian
distribution function into the final 9 3
α
form of integro-differential equation
α = i 8 Λˆ 3T − 24 Λˆ 2T Ĉ - i 24 Λˆ T Ĉ 2 + 8 Ĉ 3 + i 72 Λˆ T Λ̂ 2
solution [3], we find that D̂ is given
by the expression: [
− 72 Λ̂ 2p Ĉ − 108 + 12 81− i 48 Λ
ˆ 3 Ĉ Λ̂ 2
T p
ˆ + i Cˆ ) −2 + i 48 Λˆ T Ĉ 3 Λ̂ 2p − i 48 Λˆ T Ĉ Λ̂ 4p −12 Ĉ 3
Dˆ = i ( λ + Λ T (3)
−12 Λ
ˆ 4 Λ̂ 2 + 24 Λˆ 2 Λ̂ 4 +12 Ĉ 4 Λ̂ 2 − 24 Ĉ 2 Λ̂ 4
T p T p p p
And a lengthily calculation yields the − i12 Λˆ 3T + 36 Λˆ 2T Ĉ +108 Ĉ Λ̂ 2p + 72 Λˆ 2T Ĉ 2 Λ̂ 2p

cubic dispersion equation [2]:
+ i 36 Λ
ˆ Ĉ 2 + i108 Λ
ˆ Λ̂ 2 +12 Λ̂ 6 ]1/2
((λ − i Λ
ˆ + Cˆ ) 2 − Λ
ˆ 2 ) λ + 1 = 0 (4) T T p p
T p
This equation is similar to equation The three roots are not symmetrically
for a mono-energetic beam but with located in the complex plane and there
complex coefficients [5]. might be more than one growing and
decaying modes.
3 Results and Discussion In Figure (2) the largest growth rate of
For the realization, it can be found the all three solution is plotted depending
general solution of integro-differential on different settings of the energy
equation for a Lorentzian Energy
Spread case using the Laplace spread Λ̂T . With increasing energy
transform technique [6] and spread the growth rate for a fixed
MATLAB symbolic toolbox [7,8]. value of Ĉ < 0 is reduced almost in a
The same result can be extract using linear way. The steep edge of the gain
another technique, which the resultant
curved at Cˆ ≈ 2 is smeared out
293
yielding a slightly increased growth Due to the FEL interaction more

rate for Ĉ >1.9 with increasing electrons in the center of the
spread. distribution gain energy than the
Beside the predicted reduction of the electrons lose in the tail. The width of
growth rate the gain curve becomes the bucket is reduced and electrons
may be detrapped if the center of the
more and more antisymmetric for Λ̂ T electron distribution is close to the
> 0.5. For an injection below the separatrix of the bucket. This principle
resonant energy ( Ĉ < 0), where all is repeated till the bucket completely
roots have positive imaginary parts, vanishes.
all modes are exponentially decaying. The final formula of field amplitude
The complete energy of the radiation
~
E ( zˆ) in case of Lorentian energy
field is transferred to the electron distribution can be written as [15]:
beam. ~ exp ( λ j zˆ)
E ( zˆ) = Eext ∑ (6)
Re Λ̂ j 1 − 2i ( λ + Λ
j
ˆ + i Cˆ ) λ2
T j
where
λ j are the roots of the cubic equation
1 (Eq. (5a,b)).
2
3
4
5
Ĉ
E/Eext
Figure (2)
Reduced field growth rate Re Λ̂ versus the
detuning parameter Ĉ , Curve 1: Λ
ˆ → 0,
T
Ĉ ẑ
Curve 2: Λ
ˆ → 0.25 , Curve 3:
T (a)
Λ
ˆ → 0.5 , Curve 4: Λ
T
ˆ → 1.5 , Curve 5:
T
Λ
ˆ → 2.5 at Λ
T
ˆ2 → 0 .
p
This can be understood regarding the

electron motion in the longitudinal
phase space. The bucket of the Ĉ
ponderomotive wave is filled by the
initial distribution in such a way that ẑ
the distribution thins out towards (b)
higher energy. This happens if the Figure (3)
mean energy of the distribution lies at Field gain E/Eext versus Ĉ and the reduced
the lower border of the separatrix and
the tail towards higher energy covers length of undulator ẑ . Here Λ
ˆ2 →
p 0 and
the bucket. Λ
ˆ 2T → 0 .
294
Figure (5)
Field gain E/Eext versus Ĉ and the reduced
length of undulator ẑ . Here Λ
ˆ2 →
p 0.5
E/Eext
and Λ
ˆ 2T → 0 .
ẑ
Ĉ
E/Eext
(a)
ẑ
Ĉ (a)
Ĉ
ẑ
(b)
Figure (4)
Ĉ
ˆ2 →
p 0.25
and Λ
ˆ 2T → 0 .
ẑ
(b)
Figure (6)
Ĉ
E/Eext
Field gain E/Eext versus and the reduced

ˆ2 →
p 1.5
and Λ
ˆ 2T → 0 .
ẑ
Ĉ (a)
E/Eext
Ĉ ẑ
Ĉ
ẑ (a)
(b)
295
Figure (8)
ˆ2 →
p 0.5
and Λ
ˆ2 → 1.
T
Ĉ
E/Eext
ẑ
(b)
Figure (7) ẑ
Field gain E/Eext versus Ĉ and the reduced Ĉ
(a)
ˆ2 →
p 2.5
and Λ
ˆ 2T → 0 .
Ĉ
E/Eext
ẑ
(b)
Figure (9)
ẑ length of undulator ẑ . Here Λ
ˆ2 →
p 1.5
Ĉ (a)
and Λ
ˆ2 → 1.
T
Ĉ
E/Eext
ẑ
ẑ Ĉ
(b) (a)
296
4 Conclusion
Figures (3) to (11) illustrate the
dependence of the field gain on the
detuning parameter Ĉ , space charge
parameter Λ
ˆ 2 and spread energy
p
Ĉ
parameter Λ
ˆ along the undulator
2
T
length.
According to our process in the
previous equations, the vectors shown
ẑ in Figures (3b) to (11b), and the
(b) directions of these vectors in Figures
Figure (10) (3b) to (5b) will be from the undulator
Field gain E/Eext versus Ĉ and the reduced entrance in the direction of the end of
undulator length where this
ˆ2 →
p 0.5
phenomena occur only with
and Λ
ˆ2 → 4.
T
neglecting the space charge
parameter Λ
ˆ2 .
p
When we enter the effects of the space
charge, the field gain equation (see
Figures (6) to (8)), the negative peak
E/Eext
in the field gain amplitude will be

observed along the detuning
parameter Ĉ axis.
The negative concavity values in the
ẑ field gain amplitude will be increases
Ĉ when we increase the effective value
(a)
of the space charge parameter Λ
ˆ2
p
(see Figures (6) to (11)). Also, the
universal distribution of the field gain
amplitude will be shifted in the
positive side of the detuning
Ĉ parameter values.
The rigorous results obtained in
reduced form furnish universal plots
for calculation the output
characteristics of the FEL amplifier in
the linear mode of operation. These
ẑ
(b) solutions serve as a reliable basis for
Figure (11) the development of numerical
methods.
The analysis of nonlinear processes
ˆ2 →
p 1.5 refers to problems solvable only
numerically by a computer. On the
and Λ
ˆ2 → 4.
T other hand, testing of the numerical
297
simulation codes would be difficult [6] Saldin E.L., Schneidmiller E.A.,

without the use of rigorous results of Yurkov M.V., Statistical Properties of
FEL amplifier linear theory as a the Radiation from SASE FEL
primary standard. Operating in the Linear Regime,
TESLA-FEL 1997-1, August 1997,
References
China, pp. 71-77
[1] Kim K. J., Three-Dimensional
[7] Press W. H., Teukolsky S. A.,
Analysis of Coherent Amplification
Vetterling W. T., Flannery B. P.,
and Self-Amplified Spontaneous
Numerical Recipes in C, Cambridge
Emission in Free Electron Lasers,
University Press, Cambridge, 1992.
Phys. Rev. Let., Vol. 57, 1986, pp.
[8] Group of MathWork, Symbolic
1871-1874.
Math Toolbox User’s Guide, 7th
[2] Reiche S., Numerical Studies for a
edition by The MathWorks, Inc.,
Single Pass High Gain Free-Electron
2002.
Laser, Tech. Rep. DESY thesis-00-
[6] Ayvazyan V., et al., A new
012, Deutsches Electronen
Powerful Source for Coherent VUV
Synchrotron, 2000.
Radiation: Demonstration of
[3] Saldin E., Schneidmiller E.,
Exponential Growth and Saturation at
Yurkov M., The Physics of Free
the TTF Free Electron Laser, Eur.
Electron Lasers, Springer, New York,
Phy. J., D20, 2002, pp.149-156.
1st ed., 1999.
[7] Markus H., Surface Roughness
[4] Turner S., CERN Accelerator
Wake Fields and Longitudinal Phase
School Synchrotron Radiation and
Space in a Linear Electron
Free Electron Lasers, Tech. Rep.
Accelerator, Tech. Rep. DESY thesis-
CERN 90-03, CERN, 1990.
00-020, Deutsches Electronen
[5] Dauod H. M., Al-Faiz M. Z.,
Synchrotron, 2002.
Sanduk M. I. and Khamis R. A.,
Extraction the General Solution of
Integro-Differential Equation in a
Mono-Energetic Electron Beam Case,
to be published.
298

Study The Integro-Differential Equation in A Lorentzian Energy Spread Case

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Study The Integro-Differential Equation in A Lorentzian Energy Spread Case

Uploaded by

Copyright:

Available Formats

Eng. & Technology, Vol.25, Suppl. of No.

Study the Integro-Differential Equation in a Lorentzian

Dr. R. A. Khamis*, H. M. Daoud, M. Z. Al-Faiz & M. I. Sanduk**

Keyword: Evolutionary programming and Symbolic Analysis, Mathematical

1 Introduction To close the problem, the field

numerical answer for a specific set of 2 Theory

cubic equation (in normalized form)

The general solution of above

And a lengthily calculation yields the − i12 Λˆ 3T + 36 Λˆ 2T Ĉ +108 Ĉ Λ̂ 2p + 72 Λˆ 2T Ĉ 2 Λ̂ 2p

yielding a slightly increased growth Due to the FEL interaction more

This can be understood regarding the

Field gain E/Eext versus and the reduced

in the field gain amplitude will be

simulation codes would be difficult [6] Saldin E.L., Schneidmiller E.A.,

You might also like