Tutorial: Emittance Calculations in Trak

Tutorial: emittance calculations in Trak
Stanley Humphries, Copyright 2009
Field Precision
PO Box 13595, Albuquerque, NM 87192 U.S.A.
Telephone: +1-505-220-3975
Fax: +1-617-752-9077
E mail: [email protected]
Internet: http://www.fieldp.com
1
In this tutorial I’ll discuss the basic concepts of beam emittance and show
how the quantity is calculated for cylindrical beams in the two-dimensional
Trak program. I’ll begin by reviewing some basic concepts of emittance. A
comprehensive discussion is given in Chaps. 3 and 4 of my book Charged
Particle Beams. It’s available in PDF format on our Internet site at:
http//www.fieldp.com/cpb.html
A charged particle beam is a collection of electrons or ions where particles

have about the same energy and move in about the same direction. In a
perfect beam, the transverse velocities of particles are coherent so that they
could be focused to a point.
The quantity emittance quantifies the deviation from coherency. The
concept is effectively illustrated by a phase-space plot. Assume that the
average particle motion is in the z direction and consider transverse motion
in the x direction. We can characterize the transverse beam distribution at
location in z with a two-dimensional plot of position x and angle with respect
to the beam axis:
dx
x′ = . (1)
dz
Generally, the angle is small compared to a radian for beams that are useful
for applications.
The phase-space plot is a straight line for a perfect beam. The phase-
space distribution of an imperfect beam has non-zero area. In principle, the
distribution could have any shape. In practice, we often assume the shape
is elliptical because particles follow elliptical trajectories in phase space as
the beam moves along z in a focusing system with linear transverse forces.
The upright ellipse of Fig. 1 shows a beam at a waist point. The phase space
ellipses of diverging and converging beams are inclined with respect to the x
and x′ axes. If the ellipse of the figure encloses a uniform density of particle
points, then the emittance is defined as the area of the ellipse divided by π,
or
ǫx = x0 x′0 . (2)
If the value of x0 is specified in meters and x′o in radians, then the unit of
emittance is π-m-radian.
In the more general case, the distribution shape may not be elliptical,
the density may not be uniform and the beam may not be at a waist point.
Here, it is useful to use the RMS (root-mean-squared) emittance, given by
the equation:

2
ǫx = 4 (x − x)2 (x′ − x′ )2 − (x − x)(x′ − x′ ) , (3)
2
Figure 1: Elliptical phase-space distribution with x0 = 2.5 cm and x′0 = 0.05
radian.
where the overline symbol represents an average taken over the particle dis-
tribution. The equation applies to general distributions (e.g., converging and
diverging beams). The factor of 4.0 ensures that the equation gives the re-
sult of Eq. 2 when applied to a uniformly-filled, upright ellipse. Equation 3
is employed for emittance calculations in Trak and Gendist.
The example EMITCALC illustrates Trak emittance figures. An accel-
erating electric field and focusing magnetic field with cylindrical symme-
try may be included in the solution. The electric field volume defined by
EMITCALCE.MIN extends axially from z = 0.0 cm to z = 100.0 cm and radi-
ally from r = 0.0 cm to r = 10.0. Fixed potential boundaries with φ = 0.0
V at z = 0.0 cm and φ = 1.0 V at z = 100.0 cm give a uniform electric field
Ez = 10.0 V/m. The magnetic field calculation (shown in Fig. 2) extends ra-
dially to 20.0 cm. It represents a solenoid magnetic lens. The drive current of
1000 A gives a peak field Bz = 2.4967 × 10−3 tesla. It is important to under-
stand the role of model particles in cylindrical calculations with Trak. In an
electron beam simulation, the program tracks model particles as individual
electrons in Cartesian coordinates. An electron occupies a single value of az-
imuth at each axial position. On the other hand, a model particle is treated
as if it extended over 360o of azimuth for the assignment of space-charge and
current to calculate beam-generated fields. We must exercise some care in
interpreting emittance calculations.
3
Figure 2: Magnetic field solution EMITCALCB.
When Trak generates model particles at an emission surface, the con-

vention is place them along the x axis (θ = 0.0o ). To represent this choice,
we shall start the EMITCALC example with list input that defines a distribu-
tion with non-zero emittance such that particles have initial positions and
angles along x. The spreadsheet EMITCALC.XLS creates a distribution of 100
electrons with kinetic energy 250 keV distributed uniformly around the pe-
riphery of a phase-space ellipse with dimensions x0 = 2.5 cm and x′0 = 0.05
radian (Fig. 1). A uniformly-filled ellipse with this boundary would have an
emittance 1.25 × 10−3 π-m-rad (Eq. 2). We copy the particle information
in the final nine columns of the spreadsheet to the text file EMITCALC.PRT.
Table 1 shows the Trak input file. In the Fields section, we can use the nor-
malization factors in the EFile and BFile commands to turn off the electric
and/or magnetic field or to change the magnitude. The Record command
generates output PRT files at positions z = 25.0, 50.0 and 75.0 cm in addition
to the normal output EMITCALCOUT.PRT at the end of the solution volume
(z = 100.0 cm). The PartDist command write listing-file data that include
emittance calculations in x and y using Eq. 3.
We now have sufficient resources to begin calculations. In the first run, we
set both field normalization factors equal to 0.0. In this case the beam simply
drifts so that the output distribution is an expanding and tilted version of
the input (diverging beam). The listing generated by the PartDist command
contains the following entry:
Model particle RMS emittances with respect to the z-axis

EpsiX (pi-m-rad): 2.5008E-03
EpsiY (pi-m-rad): 0.0000E+00
Note that the calculated value of ǫx is twice the product x0 x′0 because the
particles are distributed on the envelope of the ellipse rather than uniformly
distributed through the area.
An important question is how does the emittance of an azimuthally-
symmetric beam compare to the value when particles are confined to the
x axis? We can use GenDist to resolve the issue. Load EMITCALCIN.PRT
into the program and click the command Beam analysis in the Analysis
4
Table 1: Trak input file EMITCALC.TIN.
FIELDS
EFILE: EmitCalcE.EOU 0.0
BFILE: EmitCalcB.POU 6.0
DUNIT: 1.0000E+02
END
PARTICLES TRACK
PFILE: EmitCalcIn
RECORD UP Z 25.0 50.0 75.0
END
DIAGNOSTICS
PARTFILE: EmitCalcOut
PARTDIST
END
ENDFILE
menu. The resulting output listing show the results ǫx = 0.25008 π-cm-rad
and ǫy = 0.0 π-cm-rad. Next, use the Beam section tool to rotate and repli-
cate particle entries to create a full circular beam. Use the default axis of z,
pick the date type 2D and accept the default of Nθ = 60. Save the results
as EMITCALCIN COMPLETE.PRT. When loaded in GenDist, the distribution
gives the projection plot shown in Fig. 3. An application of the Beam analysis
command gives the results ǫx = 0.12506 π-cm-rad and ǫy = 0.12506 π-cm-
rad. As expected, the emittance values are equal in x and y. The effect of
including particles with positions and angles off the x axis was to reduce the
RMS emittance by a factor of 2.0.
To finalize our understanding of emittance values created by Trak, we
must include effects of rotation in solenoid magnetic fields that mix positions
and angles in x and y. We run a solution with the electric field normalization
factor set to 0.0 and the magnetic field factor set to 4.8. The resulting
magnetic field magnitude gives a rotation of about 90o passing through the
lens. Table 2 lists emittance values calculated with GenDist using the initial
and final PRT files as well as those generated by the Record command. The
sum of ǫx and ǫy is conserved to high accuracy and is unaffected by the
ordered rotational motion in the magnetic field.
The test calculations suggest the following procedure to find emittance for
an azimuthally-symmetric beam from the RMS values calculated for a dis-
tribution of model particles (where each particle occupies a single azimuth):
5
Figure 3: Azimuthally symmetric distribution created with the GenDist
Beam section tool.
Table 2: Emittance calculations for a 250 keV electron beam moving through
a magnetic lens with B0 = 0.0120 tesla.
z ǫx ǫy ǫx + ǫy
(cm) (π-cm-rad) (π-cm-rad) (π-cm-rad)
0.0 0.25008 0.00000 0.25008
25.0 0.24897 0.00113 0.25010
50.0 0.12262 0.12761 0.25023
75.0 0.00075 0.24981 0.25056
100.0 0.00010 0.25012 0.25022
6
1. Take the sum of ǫx and ǫy .
2. Divide the sum equally between x and y.
These operations are performed automatically by Trak in response to the

PartDist command for runs where the fields have cylindrical symmetry. Here
is the resulting listing for the example:
Cylindrical beam RMS emittances with respect to the z-axis

EpsiX (pi-m-rad): 1.2511E-03
EpsiY (pi-m-rad): 1.2511E-03
To conclude, we can use the test example to verify conservation of nor-

malized emittance:
ǫnx = βγǫx . (4)

In this case, we use a normalization factor of 0.0 for the magnetic field and
2.5 × 105 for the electric field. For this choice, the electron exit energy is
500 keV. The initial relativistic factors are βi = 0.7410 and γi = 1.4892 and
the final factors are βi = 0.8629 and γf = 1.9785. If normalized emittance is
conserved, then the final beam emittance should be
βi γi
ǫxf = ǫxi . (5)
βf γf
The value at z = 100.0 cm reported by the PartDist command agrees with
the theoretical prediction of 1.616 × 10−3 π-m-rad.

Tutorial: Emittance Calculations in Trak

Uploaded by

Copyright:

Available Formats

Tutorial: Emittance Calculations in Trak

Uploaded by

Document Information

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

Tutorial: Emittance Calculations in Trak

Uploaded by

Copyright:

Available Formats

Tutorial: emittance calculations in Trak

Stanley Humphries, Copyright 2009

A charged particle beam is a collection of electrons or ions where particles

When Trak generates model particles at an emission surface, the con-

Model particle RMS emittances with respect to the z-axis

2. Divide the sum equally between x and y.

These operations are performed automatically by Trak in response to the

Cylindrical beam RMS emittances with respect to the z-axis

To conclude, we can use the test example to verify conservation of nor-

ǫnx = βγǫx . (4)

You might also like