MAD X User's Guide User
User Manual:
Open the PDF directly: View PDF .
Page Count: 247 [warning: Documents this large are best viewed by clicking the View PDF Link!]
- I Control
- Conventions
- Command Format
- Program Flow Statements
- General Control Statements
- File Handling Statements
- Table Handling Statements
- Beam Handling Statements
- Sequence Editor
- II Elements, Beamlines and Sequences
- Definition of Elements
- Element Types
- Marker
- Drift Space
- Bending Magnet
- Dipole edge
- Quadrupole
- Sextupole
- Octupole
- General Thin Multipole
- Solenoid
- Nonlinear Lens with Elliptic Potential
- Closed Orbit Corrector
- Transverse Kicker
- RF Cavity
- Thin Radio-Frequency Multipole
- Crab Cavity
- Electrostatic Separator
- Beam Position Monitor
- Instrument and Placeholder
- Collimator
- Beam-beam Interaction
- Arbitrary Matrix Element
- Rotation around the vertical axis
- Rotation around the longitudinal axis
- Coordinate translation
- Change of reference system
- Range and Class Selection Format
- Beam Lines
- Sequences
- III Input and Output
- IV MAD-X Modules
- SURVEY
- Twiss Module
- Matching Module
- EMIT: Equilibrium emittances
- Physical Aperture
- Slicing a sequence into thin lenses
- Error Definitions
- Orbit Correction
- SODD: Second Order Detuning and Distortion
- Touschek Lifetime and Scattering Rates
- Intra-Beam Scattering
- Particle Tracking
- V PTC Commands
- VI Trailing Material
EUROPEAN LABORATORY FOR PARTICLE PHYSICS
Preliminary draft
The MAD-X Program
(Methodical Accelerator Design)
Version 5.02.09
User’s Reference Manual
Hans Grote
Frank Schmidt
Laurent Deniau
Ghislain Roy (editor)
Abstract
MAD-X is a general-purpose tool for charged-particle optics design and studies
in alternating-gradient accelerators and beam lines. It can handle medium size to
very large accelerators and solves various problems on such machines.
MAD-X is the successor of MAD-8 and was specifically adapted to the needs of
the design of the LHC. The PTC library of E. Forest is also embedded in MAD-X
as an addition to better support small and low energy accelerators. A large part
of the present document is based on the MAD-8 documentation originally written
and published by F.C. Iselin.
This documentation is updated regularly as corrections, improvements and
additions are made to the program. It is also available online on the MAD-X website.
Comments and corrections from readers are most welcome. They may be sent
to the email address: mad@cern.ch
Geneva, Switzerland
April 6, 2016
![](asset-2.png)
Copyright Notice
CERN
EUROPEAN ORGANISATION FOR NUCLEAR RESEARCH
Program name: MAD-X --- Methodical Accelerator Design
CERN program library entry: T5001
Authors or contacts: Laurent Deniau
Beams Department
CERN
CH-1211 GENEVA 23
SWITZERLAND
Copyright CERN, Geneva 1990 - Copyright and any other appropriate legal
protection of this computer program and associated documentation reserved
in all countries of the world.
Organisations collaborating with CERN may receive this program and
documentation freely and without charge.
CERN undertakes no obligation for the maintenance of this program, nor
responsibility for its correctness, and accepts no liability whatsoever
resulting from its use.
Program and documentation are provided solely for the use of the
organisation to which they are distributed.
This program may not be copied or otherwise distributed without
permission. This message must be retained on this and any other
authorised copies.
The material cannot be sold. CERN should be given credit in all
references.
Contents
I Control 10
1 Conventions 11
1.1 Reference System .................................. 11
1.2 Closed Orbit .................................... 11
1.3 Global Reference System .............................. 12
1.4 Local Reference Systems .............................. 14
1.4.1 Reference System for Straight Beam Elements ............. 14
1.4.2 Reference System for Bending Magnets ................. 14
1.5 Sign Conventions for Magnetic Fields ....................... 15
1.6 Generalisation to normal and skew components ................. 17
1.7 Variables ...................................... 18
1.7.1 Canonical Variables Describing Orbits .................. 18
1.7.2 Normalised Variables and other Derived Quantities ........... 18
1.7.3 Linear Lattice Functions (Optical Functions) .............. 19
1.7.4 Chromatic Functions ............................ 20
1.7.5 Variables in the SUMM Table ....................... 21
1.7.6 Variables in the TRACK Table ...................... 22
1.8 Physical Units ................................... 22
2 Command Format 24
2.1 Input Statements .................................. 24
2.2 Comments ...................................... 24
2.3 Commands ..................................... 24
2.4 Keywords ...................................... 25
2.5 Attribute Types .................................. 25
2.6 Variable Declarations ................................ 26
2.7 Identifiers or Labels ................................ 26
2.8 Command Attributes ................................ 26
2.9 Name or String Attributes ............................. 27
2.10 Logical Attributes ................................. 27
2.11 Integer Attributes ................................. 28
2.12 Real Attributes ................................... 28
2.13 Real Expressions .................................. 28
2.13.1 Operators in Arithmetic Expressions ................... 29
2.13.2 Operands in Arithmetic Expressions ................... 31
2.13.3 Expressions and Random Values ..................... 33
2.14 Logical Expressions ................................. 34
2.15 Variable Names ................................... 34
2.16 Regular Expressions ................................ 34
2.17 Relations between Variable Parameters ...................... 35
3 Program Flow Statements 37
3.1 IF...ELSEIF...ELSE ................................ 37
1
![](asset-4.png)
2CONTENTS
3.2 WHILE ....................................... 38
3.3 MACRO ....................................... 38
4 General Control Statements 40
4.1 EXIT, QUIT, STOP ................................ 40
4.2 HELP ........................................ 40
4.3 SHOW ........................................ 40
4.4 VALUE ....................................... 40
4.5 OPTION ...................................... 41
4.6 EXEC ........................................ 41
4.7 SET ......................................... 42
4.8 SYSTEM ...................................... 43
4.9 TITLE ........................................ 43
4.10 USE ......................................... 44
4.11 SELECT ...................................... 44
5 File Handling Statements 47
5.1 ASSIGN ....................................... 47
5.2 CALL ........................................ 47
5.3 RETURN ...................................... 47
5.4 PRINT ....................................... 47
5.5 PRINTF ....................................... 47
5.6 RENAMEFILE ................................... 48
5.7 COPYFILE ..................................... 48
5.8 REMOVEFILE ................................... 49
6 Table Handling Statements 50
6.1 CREATE ...................................... 50
6.2 DELETE ...................................... 50
6.3 READTABLE ................................... 50
6.4 READMYTABLE ................................. 50
6.5 TABSTRING .................................... 51
6.6 WRITE ....................................... 51
6.7 SETVARS ...................................... 51
6.8 SETVARS LIN ................................... 51
6.9 FILL ......................................... 53
6.10 SHRINK ....................................... 53
7 Beam Handling Statements 55
7.1 BEAM ........................................ 55
7.2 RESBEAM ..................................... 58
7.3 Referring to BEAM data attributes ........................ 58
7.4 BV FLAG ...................................... 59
8 Sequence Editor 61
8.1 SEQEDIT ...................................... 61
8.2 FLATTEN ..................................... 61
CONTENTS 3
8.3 CYCLE ....................................... 62
8.4 REFLECT ..................................... 62
8.5 INSTALL ...................................... 62
8.6 MOVE ........................................ 63
8.7 REMOVE ...................................... 63
8.8 REPLACE ..................................... 64
8.9 EXTRACT ..................................... 64
8.10 ENDEDIT ...................................... 64
8.11 SAVE ........................................ 65
8.12 DUMPSEQU .................................... 66
II Elements, Beamlines and Sequences 67
9 Definition of Elements 68
9.1 Element Input Format ............................... 68
9.2 Editing Element Definitions ............................ 68
9.3 Element Classes ................................... 69
10 Element Types 71
10.1 Marker ........................................ 71
10.2 Drift Space ..................................... 71
10.3 Bending Magnet .................................. 71
10.4 Dipole edge ..................................... 74
10.5 Quadrupole ..................................... 75
10.6 Sextupole ...................................... 75
10.7 Octupole ....................................... 76
10.8 General Thin Multipole .............................. 77
10.9 Solenoid ....................................... 78
10.10Nonlinear Lens with Elliptic Potential ...................... 79
10.11Closed Orbit Corrector ............................... 81
10.12Transverse Kicker .................................. 82
10.13RF Cavity ...................................... 83
10.14Thin Radio-Frequency Multipole ......................... 84
10.15Crab Cavity ..................................... 85
10.16Electrostatic Separator ............................... 87
10.17Beam Position Monitor .............................. 87
10.18Instrument and Placeholder ............................ 88
10.19Collimator ...................................... 88
10.20Beam-beam Interaction .............................. 89
10.21Arbitrary Matrix Element ............................. 93
10.22Rotation around the vertical axis ......................... 93
10.23Rotation around the longitudinal axis ...................... 94
10.24Coordinate translation ............................... 95
10.25Change of reference system ............................ 95
11 Range and Class Selection Format 96
4CONTENTS
11.1 RANGE ....................................... 96
11.2 CLASS ....................................... 96
12 Beam Lines 98
12.1 Simple Beam Lines ................................. 98
12.2 Nested Beam Lines ................................. 99
12.3 Reflection and Repetition ............................. 99
12.4 Replaceable Arguments .............................. 100
12.5 Limits of Construction of Beam Lines ...................... 101
13 Sequences 102
III Input and Output 106
14 TFS File Format 107
14.1 Descriptor Lines .................................. 107
14.2 Column Formats .................................. 107
14.3 Twiss TFS file header ............................... 107
15 Conversion to SixTrack 108
16 SXF file format 110
16.1 SXFWRITE ..................................... 110
16.2 SXFREAD ..................................... 110
17 Plotting data 111
17.1 PLOT ........................................ 111
17.2 SETPLOT ..................................... 114
17.3 RESPLOT ..................................... 115
17.4 First example for plots of tracking data ..................... 115
17.5 Second example for plots of tracking data .................... 116
17.6 MAD-X PLUGINS ................................. 118
17.7 RPLOT ....................................... 119
IV MAD-X Modules 121
18 SURVEY 122
19 Twiss Module 123
19.1 Twiss Parameters for a Period .......................... 125
19.2 Initial Values from a Periodic Line ........................ 126
19.3 Given Initial Values ................................ 126
19.4 SAVEBETA ..................................... 127
19.5 BETA0: Set Initial Lattice Parameters ...................... 128
19.6 Sectormap output ................................. 129
19.7 Beam Threader ................................... 130
![](asset-7.png)
CONTENTS 5
19.8 Closed Orbit Guess ................................. 131
20 Matching Module 133
20.1 MATCH . . . ENDMATCH ............................. 134
20.2 Cell Matching .................................... 134
20.3 Insertion Matching ................................. 134
20.4 More than one active sequence .......................... 136
20.5 SLOW attribute .................................. 136
20.6 Useful TWISS attributes .............................. 136
20.7 VARY ........................................ 137
20.8 CONSTRAINT ................................... 138
20.9 User Defined Matching Constraints ........................ 139
20.10GLOBAL ...................................... 139
20.11WEIGHT, GWEIGHT ............................... 140
20.12Matching Methods ................................. 140
20.12.1 LMDIF: Fast Gradient Minimisation ................... 141
20.12.2 MIGRAD: Gradient Minimisation .................... 141
20.12.3 SIMPLEX: Simplex Minimisation ..................... 142
20.12.4 JACOBIAN: Newton Minimisation .................... 142
20.13USE MACRO .................................... 143
20.13.1 Initiating the Matching Module with USE MACRO ............ 143
20.13.2 VARY statements ............................. 144
20.13.3 Macro definitions .............................. 144
20.13.4 Examples .................................. 144
20.14Matching Examples ................................ 145
21 EMIT: Equilibrium emittances 147
22 Physical Aperture 148
22.1 Aperture definition ................................. 148
22.2 Aperture tolerance definition ........................... 151
22.3 Aperture offset definition ............................. 151
22.4 APERTURE .................................... 151
22.5 Not simply connex beam pipe profiles ...................... 153
22.6 Trueprofile file syntax ............................... 154
22.7 Offsetelem file syntax ................................ 156
22.8 Aperture command example ............................ 158
23 Slicing a sequence into thin lenses 160
23.1 MAKETHIN .................................... 160
23.2 Controlling the number of slices .......................... 161
23.3 Choice of options for dipoles ............................ 162
23.4 Additional information ............................... 162
24 Error Definitions 164
24.1 EALIGN: Alignment Errors ............................ 164
24.2 EFCOMP: Field Errors .............................. 167
![](asset-8.png)
6CONTENTS
24.3 EOPTION: Set Options for Error Definition ................... 169
24.4 EPRINT: List Machine Imperfections ...................... 169
24.5 ESAVE: Writing errors to a file .......................... 170
24.6 SETERR: Reading errors from a table or file .................. 170
25 Orbit Correction 172
25.1 CORRECT ..................................... 172
25.2 USEMONITOR, USEKICK ............................ 176
25.3 CSAVE ....................................... 177
25.4 SETCORR ..................................... 177
25.5 COPTION ..................................... 177
26 SODD: Second Order Detuning and Distortion 178
26.1 DETUNE ...................................... 179
26.2 DISTORT1 ..................................... 179
26.3 DISTORT2 ..................................... 180
27 Touschek Lifetime and Scattering Rates 181
28 Intra-Beam Scattering 184
29 Particle Tracking 187
29.1 Introduction to MAD-X Tracking Modules ..................... 187
29.2 Overview of Thin-Lens Tracking ......................... 187
29.3 TRACK ....................................... 189
29.4 START ....................................... 191
29.5 OBSERVE ..................................... 192
29.6 RUN ......................................... 192
29.7 DYNAP ....................................... 192
29.8 ENDTRACK .................................... 194
29.9 Space Charge .................................... 194
VPTC Commands 195
30 PTC Set-up Parameters 196
30.1 Command Synopsis ................................. 196
30.2 PTC CREATE UNIVERSE ............................ 197
30.3 PTC CREATE LAYOUT ............................. 197
30.4 PTC MOVE TO LAYOUT ............................ 199
30.5 PTC READ ERRORS ............................... 199
30.6 PTC ALIGN .................................... 200
30.7 PTC END ...................................... 200
30.8 Additional Options for Physical Elements .................... 200
31 Thick-Lens Tracking Module 202
31.1 Synopsis ....................................... 202
31.2 PTC START .................................... 202
![](asset-9.png)
CONTENTS 7
31.3 PTC OBSERVE .................................. 203
31.4 PTC TRACK .................................... 204
31.5 PTC TRACKLINE ................................. 206
31.6 PTC TRACK END ................................ 207
31.7 Choice of options .................................. 208
32 Ripken Optics Parameters 209
32.1 Introduction ..................................... 209
32.2 PTC TWISS .................................... 210
32.3 Periodic Solution .................................. 213
32.4 Evaluation of Twiss parameters inside magnets ................. 213
32.5 Solution with Initial Conditions .......................... 214
32.5.1 Initial Values from the Given Twiss Parameters ............. 214
32.5.2 Initial Values from a Map-Table ..................... 215
32.5.3 Initial Values from a Map-File ...................... 215
32.5.4 Initial Values from a Given Matrix .................... 215
32.5.5 Initial Values from Twiss Parameters via BETA0-block ........ 215
33 Non-Linear Machine Parameters 217
33.1 SELECT PTC NORMAL ............................. 217
33.2 PTC NORMAL ................................... 218
34 MAD-X-PTC Auxiliaries 220
34.1 PTC SETSWITCH ................................. 221
34.2 PTC KNOB ..................................... 222
34.3 PTC SETKNOBVALUE .............................. 223
34.4 PTC VARYKNOBS (Under Construction) .................... 223
34.5 PTC PRINTPARAMETRIC ........................... 225
34.6 PTC EPLACEMENT ............................... 226
34.7 PTC PRINTFRAMES ............................... 227
34.8 PTC SELECT ................................... 227
34.9 PTC SELECT MOMENT ............................. 228
34.10PTC MOMENTS .................................. 229
34.11PTC DUMPMAPS ................................. 230
34.12PTC SETCAVITIES ................................ 230
VI Trailing Material 232
35 Known Differences to Other Programs 233
35.1 Definitions in MAD-8 ................................ 233
35.2 Treatment of Energy Error in TWISS ...................... 233
36 MAD-X pitfalls 234
37 Contributors to MAD-X 237
37.1 MAD team ..................................... 237
37.2 Module keepers and contributors ......................... 237
CONTENTS 9
List of Tables
1.1 Physical Units used by MAD-X ........................... 23
2.1 Predefined Symbolic Constants in MAD-X ..................... 32
4.1 Flags available to OPTION command ....................... 42
7.1 Default Beam Data ................................. 58
14.1 Column Formats used in TFS Tables ....................... 107
20.1 Default Matching Weights ............................. 141
22.1 Predefined aperture types ............................. 149
23.1 Best choice of options in MAKETHIN ........................ 162
List of Figures
1.1 Local Reference System .............................. 11
1.2 Global Reference System .............................. 12
1.3 Reference System for Straight Beam Elements .................. 14
1.4 Reference System for a Rectangular Bending Magnet .............. 15
1.5 Reference System for a Sector Bending Magnet ................. 16
10.1 Contour plot of the scalar potential ........................ 80
10.2 Trapezoidal shape of radial density for beam-beam lens. ............ 91
10.3 Hollow parabolic shape of radial density for beam-beam lens. ......... 92
22.1 Definition of aperture tolerances ......................... 151
22.2 Determination of maximum halo size ....................... 153
22.3 Not connex beam pipe profile: problem ..................... 154
22.4 Not connex beam pipe profile: solution ...................... 155
22.5 Illustration of effect of OFFSETELEM ........................ 157
22.6 Example of plot showing aperture limits and n1 ................. 159
24.1 Alignment errors in the (x, s)-plane ........................ 165
24.2 Alignment errors in the (x, y)-plane ........................ 165
24.3 Alignment errors in the (y, s)-plane ........................ 166
24.4 Readout errors in a monitor ............................ 166
Part I
Control
10
![](asset-d.png)
Chapter 1. Conventions
1.1 Reference System
The accelerator and/or beam line to be studied is described as a sequence of beam elements
placed sequentially along a reference orbit. The reference orbit is the path of a charged
particle having the central design momentum of the accelerator through idealised magnets
with no fringe fields (see Figure 1.1).
y
x
z
ρ
ρ
centre of
curvature
actual
orbit
d~r
s
reference
orbit
Figure 1.1: Local Reference System
The reference orbit consists of a series of straight line segments and circular arcs. It is defined
under the assumption that all elements are perfectly aligned. The accompanying tripod of
the reference orbit spans a local curvilinear right handed coordinate system (x,y,s) The local
s-axis is the tangent to the reference orbit. The two other axes are perpendicular to the
reference orbit and are labelled x(in the bend plane) and y(perpendicular to the bend
plane).
1.2 Closed Orbit
Due to various errors like misalignment errors, field errors, fringe fields etc., the closed orbit
does not coincide with the reference orbit. The closed orbit also changes with the momentum
error. The closed orbit is described with respect to the reference orbit, using the local reference
system (x, y, s). It is evaluated including any nonlinear effects.
11
![](asset-e.png)
12 CHAPTER 1. CONVENTIONS
MAD-X also computes the betatron and synchrotron oscillations with respect to the closed
orbit. Results are given in the local (x, y, s)-system defined by the reference orbit.
1.3 Global Reference System
The global reference orbit of the accelerator is uniquely defined by the sequence of physical
elements. The local reference system (x, y, s) may thus be referred to a global Cartesian
coordinate system (X, Y, Z ) (see Figure 1.2).
The positions between beam elements are indexed with i= 0, . . . n. The local reference system
(xi,yi,si) at position i, i.e. the displacement and direction of the reference orbit with respect
to the system (X, Y, Z ) are defined by three displacements (Xi,Yi,Zi) and three angles (θi,
φi,ψi)
Z
X
Yx
ys
Zi
Xi
Yi
ψi
φi
θi
reference
orbit
intersection of
xy and ZX planes
intersection of
xs and ZX
planes
intersection of
ys and ZX
planes
projection of s
onto ZX-plane
Figure 1.2: Global Reference System showing the global Cartesian system (X, Y, Z) in black
and the local reference system (x, y, s) in red after translation (Xi, Yi, Zi) and rotation
(θi, φi, ψi). The projections of the local reference system axes onto the horizontal ZX plane
of the Cartesian system are figured with blue dashed lines. The intersections of planes ys,xy
and xs of the local reference system with the horizontal ZX plane of the Cartesian system
are figured in green dashed lines.
The above quantities are defined more precisely as follows:
XDisplacement of the local origin in X-direction.
YDisplacement of the local origin in Y-direction.
1.3. GLOBAL REFERENCE SYSTEM 13
ZDisplacement of the local origin in Z-direction.
THETA θis the angle of rotation (azimuth) about the global Y-axis, between the
global Z-axis and the projection of the reference orbit onto the (Z,X)-plane.
A positive angle THETA forms a right-hand screw with the Y-axis.
PHI φis the elevation angle, i.e. the angle between the reference orbit and its pro-
jection onto the (Z,X)-plane. A positive angle PHI corresponds to increasing
Y.
If only horizontal bends are present, the reference orbit remains in the (Z,
X)-plane and PHI is always zero.
PSI ψis the roll angle about the local s-axis, i.e. the angle between the line defined
by the intersection of the (x, y)-plane and (Z, X )-plane on one hand, and the
local x-axis on the other hand. A positive angle PSI forms a right-hand screw
with the s-axis.
The angles (θ,φ,ψ) are not the Euler angles. The reference orbit starts at the origin and
points by default in the direction of the positive Z-axis. The initial local axes (x,y,s) coincide
with the global axes (X,Y,Z) in this order. The initial values (X0,Y0,Z0,θ0,φ0,ψ0) are
therefore all zero unless the user specifies different initial conditions.
Internally the displacement is described by a vector Vand the orientation by a unitary matrix
W. The column vectors of Ware the unit vectors spanning the local coordinate axes in the
order (x, y, s). Vand Whave the values:
V=
X
Y
Z
, W = Θ Φ Ψ (1.1)
where
Θ =
cos θ0 sin θ
0 1 0
−sin θ0 cos θ
,Φ =
1 0 0
0 cos φsin φ
0−sin φcos φ
,Ψ =
cos ψ−sin ψ0
sin ψcos ψ0
0 0 1
(1.2)
The reference orbit should be closed, and it should not be twisted. This means that the
displacement of the local reference system must be periodic with the revolution frequency of
the accelerator, while the position angles must be periodic (modulo 2π) with the revolution
frequency. If ψis not periodic (modulo 2π), coupling effects are introduced. When advancing
through a beam element, MAD-X computes Viand Wiby the recurrence relations
Vi=Wi−1Ri+Vi−1, Wi=Wi−1Si(1.3)
The vector Riis the displacement and the matrix Siis the rotation of the local reference
system at the exit of the element iwith respect to the entrance of the same element. The
values of Riand Siare listed below for different physical element types.
![](asset-10.png)
14 CHAPTER 1. CONVENTIONS
1.4 Local Reference Systems
1.4.1 Reference System for Straight Beam Elements
In straight elements the local reference system is simply translated by the length of the element
along the local s-axis. This is true for Drift spaces,Quadrupoles,Sextupoles,Octupoles,
Solenoids,Crab cavities,RF cavities,Electrostatic separators,Closed orbit correctors and
Beam position monitors.
The corresponding R,Sare
R=
0
0
L
, S =
100
010
001
.(1.4)
A rotation of the element about the S-axis has no effect on Rand S, since the rotations of
the reference system before and after the element cancel.
y1y2s
x1x2
L
Figure 1.3: Reference System for Straight Beam Elements
1.4.2 Reference System for Bending Magnets
Bending magnets have a curved reference orbit. For both rectangular and sector bending
magnets, the Rand Sare expressed as function the bend angle α:
R=
ρ(cos α−1)
0
ρsin α
, S =
cos α0−sin α
0 1 0
sin α0 cos α
(1.5)
A positive bend angle represents a bend to the right, i.e. towards negative xvalues. For
sector bending magnets, the bend radius is given by ρ, and for rectangular bending magnets
it has the value ρ=L/(2 sin(α/2)). If the magnet is rotated about the s-axis by an angle ψ,
Rand Sare transformed by
R=T R, S =T ST −1(1.6)
where Tis the orthogonal rotation matrix
T=
cos ψ−sin ψ0
sin ψcos ψ0
0 0 1
(1.7)
The special value ψ=π/2 represents a bend down.
![](asset-11.png)
1.5. SIGN CONVENTIONS FOR MAGNETIC FIELDS 15
y1y2
s1s2
ρ
x1
ρ
x2
α
x
L
e1e2
Figure 1.4: Reference System for a Rectangular Bending Magnet; the signs of pole-face rota-
tions are positive as shown.
1.5 Sign Conventions for Magnetic Fields
The MAD-X program uses the following Taylor expansion for the field on the mid-plane y= 0,
described in [1] (Note the factorial in the denominator):
By(x, 0) =
∞
X
n=0
Bnxn
n!(1.8)
The field coefficients have the following meaning:
B0Dipole field, with a positive value in the positive ydirection; a positive field
bends a positively charged particle to the right.
B1Quadrupole coefficient B1= (∂By/∂x); a positive value corresponds to hori-
zontal focussing of a positively charged particle.
B2Sextupole coefficient B2= (∂2By/∂x2).
B3Octupole coefficient B3= (∂3By/∂x3).
. . . etc.
Using this expansion and the curvature hof the reference orbit, the longitudinal component
![](asset-12.png)
16 CHAPTER 1. CONVENTIONS
y1y2
s1s2
ρ
x1
ρ
x2
α
r
L
e1e2
Figure 1.5: Reference System for a Sector Bending Magnet; the signs of pole-face rotations
are positive as shown.
of the vector potential to order 4 is:
As= + B0x−hx2
2(1 + hx)
+B11
2(x2−y2)−h
6x3+h2
24(4x4−y4) + ···
+B21
6(x3−3xy2)−h
24(x4−y4) + ···
+B31
24(x4−6x2y2+y4)···
+···
(1.9)
![](asset-13.png)
1.6. GENERALISATION TO NORMAL AND SKEW COMPONENTS 17
Taking ~
B=∇ × ~
Ain curvilinear coordinates, the field components can be computed as
Bx(x, y) = + B1y+h2
6y3+···
+B2xy −h
6y3+···
+B31
6(3x2y−y3) + ···
+···
By(x, y) = + B0
+B1x−h
2y2+h2
2xy2+···
+B21
2(x2−y2)−h
2xy2+···
+B31
6(x3−3xy2) + ···
+···
(1.10)
It can be easily verified that both ∇ × ~
Band ∇.~
Bare zero to the order of the B3term.
Introducing the magnetic rigidity Bρ =ps/q as the momentum of the particle divided by its
charge, the multipole coefficients are computed as
Kn=qBn/ps=Bn/Bρ (1.11)
1.6 Generalisation to normal and skew components
The previous section assumed an expansion at the mid-plane (y= 0), symmetry around the
mid-plane and considered only the vertical component of the field.
An extension using complex notation for the position (x+iy) and the field is given as
By+iBx=
∞
X
n=0
(bn+ian)(x+iy)n
nn−1(1.12)
By introducing the normal and skew multipole coefficients KN and KS at order nas
KNn=q bn/ps=bn/Bρ (1.13)
and
KSn=q an/ps=an/Bρ (1.14)
the kicks received in each plane can be expressed as the summation over all components
∆Px−i∆Py=
∞
X
n=0 −(KNn+iKSn)(x+iy)n
n!(1.15)
Remark: need to add references to the (an, bn) field conventions in the magnet world.
18 CHAPTER 1. CONVENTIONS
1.7 Variables
For each variable listed in this section, the physical units are given between square brackets,
where [1] denotes a dimensionless variable.
1.7.1 Canonical Variables Describing Orbits
MAD-X uses the following canonical variables to describe the motion of particles:
XHorizontal position xof the (closed) orbit, referred to the ideal orbit [m].
PX Horizontal canonical momentum pxof the (closed) orbit referred to the ideal
orbit, divided by the reference momentum: PX = px/p0, [1].
YVertical position yof the (closed) orbit, referred to the ideal orbit [m].
PY Vertical canonical momentum pyof the (closed) orbit referred to the ideal
orbit, divided by the reference momentum: PY = py/p0, [1].
TVelocity of light times the negative time difference with respect to the reference
particle: T = −ct, [m]. A positive T means that the particle arrives ahead of
the reference particle.
PT Energy error, divided by the reference momentum times the velocity of light:
PT = ∆E/psc, [1]. This value is only non-zero when synchrotron motion is
present. It describes the deviation of the particle from the orbit of a particle
with the momentum error DELTAP.
DELTAP Difference between the reference momentum and the design momen-
tum, divided by the design momentum: DELTAP = ∆p/p0, [1]. This
quantity is used to normalize all element strengths.
The independent variable is:
SArc length salong the reference orbit, [m].
In the limit of fully relativistic particles (γ1, v=c,pc =E), the variables T, PT used here
agree with the longitudinal variables used in [2]. This means that T becomes the negative
path length difference, while PT becomes the fractional momentum error. The reference
momentum psmust be constant in order to keep the system canonical.
1.7.2 Normalised Variables and other Derived Quantities
XN The normalised horizontal displacement, [sqrt(m)]
xn=Re(ET
1S Z)
PXN The normalised horizontal transverse momentum, [sqrt(m)]
pxn =Im(ET
1S Z)
WX The horizontal Courant-Snyder invariant, [m]
WX = x2
n+p2
xn
1.7. VARIABLES 19
PHIX The horizontal phase, [1]
φx=−arctan(pxn/xn)/2π
YN The normalised vertical displacement, [sqrt(m)]
yn=Re(ET
2S Z)
PYN The normalised vertical transverse momentum, [sqrt(m)]
pyn =Im(ET
2S Z)
WY The vertical Courant-Snyder invariant, [m]
WY = y2
n+p2
yn
PHIY The vertical phase, [1]
φy=−arctan(pyn/yn)/2π
TN The normalised longitudinal displacement, [sqrt(m)]
tn=Re(ET
3S Z)
PTN The normalised longitudinal transverse momentum, [sqrt(m)]
ptn =Im(ET
3S Z)
WT The longitudinal invariant, [m]
WT = t2
n+p2
tn
PHIT The longitudinal phase, [1]
φt=−atan(ptn/tn)/2π
In the above formulas the vectors Eiare the three complex eigenvectors, Zis the phase space
vector, and the matrix Sis the “symplectic unit matrix”:
Z=
x
px
y
py
t
pt
, S =
010000
−100000
000100
0 0 −1 0 0 0
000001
0000−1 0
(1.16)
1.7.3 Linear Lattice Functions (Optical Functions)
Several MAD-X commands refer to linear lattice functions or optical functions.
Because MAD-X uses the canonical momenta (px,py) instead of the slopes (x0,y0), the defini-
tions of the linear lattice functions differ slightly from those in Courant and Snyder[3].
Notice that in MAD-X, PT substitutes DELTAP as longitudinal variable. Dispersive and chro-
matic functions are hence derivatives with respect to PT. And since PT=BETA*DELTAP,
where BETA is the relativistic Lorentz factor, those functions given by MAD-X must be mul-
tiplied by BETA a number of time equal to the order of the derivative to find the functions
given in the literature.
The linear lattice functions are known to MAD-X under the following names:
![](asset-16.png)
20 CHAPTER 1. CONVENTIONS
BETX Amplitude function βx, [m].
ALFX Correlation function αx=−1
2(∂βx/∂s), [1]
MUX Phase function µx=Rds/βx, [2π]
DX Dispersion of x:Dx= (∂x/∂pt), [m]
DPX Dispersion of px:Dpx = (∂px/∂pt)/ps, [1]
BETY Amplitude function βy, [m]
ALFY Correlation function αy=−1
2(∂βy/∂s), [1]
MUY Phase function µy=Rds/βy, [2π]
DY Dispersion of y:Dy= (∂y/∂pt), [m]
DPY Dispersion of py:Dpy = (∂py/∂pt)/ps, [1]
R11, R12, R21, R22 : Coupling Matrix
1.7.4 Chromatic Functions
Several MAD-X commands refer to the chromatic functions.
Because MAD-X uses the canonical momenta (px,py) instead of the slopes (x0,y0), the defini-
tions of the chromatic functions differ slightly from those in [4].
Notice also that in MAD-X, PT substitutes DELTAP as longitudinal variable. Dispersive and
chromatic functions are hence derivatives with respect to PT. Since PT=BETA*DELTAP,
where BETA is the relativistic Lorentz factor, those functions given by MAD-X must be mul-
tiplied by BETA a number of times equal to the order of the derivative to find the functions
given in the literature.
The chromatic functions are known to MAD-X under the following names:
WX Chromatic amplitude function Wx=pa2
x+b2
x, [1], where
bx=1
βx
∂βx
∂pt
, ax=∂αx
∂pt−αx
βx
∂βx
∂pt
PHIX Chromatic phase function Φx= arctan(ax/bx), [2π]
DMUX Chromatic derivative of phase function: DMUX = (∂µx/∂pt), [2π]
DDX Chromatic derivative of dispersion Dx:DDX =1
2(∂2x/∂p2
t), [m]
DDPX Chromatic derivative of dispersion Dpx :DDP X =1
2(∂2px/∂p2
t)/ps, [1]
WY Chromatic amplitude function Wy=qa2
y+b2
y, [1], where
by=1
βy
∂βy
∂pt
, ay=∂αy
∂pt−αy
βy
∂βy
∂pt
PHIY Chromatic phase function Φy= arctan(ay/by), [2π]
![](asset-17.png)
1.7. VARIABLES 21
DMUY Chromatic derivative of phase function: DMUY = (∂µy/∂pt), [2π]
DDY Chromatic derivative of dispersion Dy:DDY =1
2(∂2y/∂p2
t), [m]
DDPY Chromatic derivative of dispersion Dpy :DDP Y =1
2(∂2py/∂p2
t)/ps, [1]
1.7.5 Variables in the SUMM Table
After a successful TWISS command a summary table, with name SUMM, is created which
contains the following variables:
LENGTH The length of the machine, [m].
ORBIT5 The T (= ct, [m]) component of the closed orbit.
ALFA The momentum compaction factor αc, [1].
GAMMATR The transition energy γtr, [1].
Q1 The horizontal tune Q1[1].
DQ1 The horizontal chromaticity dq1=∂Q1/∂pt, [1]
BETXMAX The largest horizontal βx, [m].
DXMAX The maximum of the absolute horizontal dispersion Dx, [m].
DXRMS The r.m.s. of the horizontal dispersion Dx, [m].
XCOMAX The maximum of the absolute horizontal closed orbit deviation [m].
XRMS The r.m.s. of the horizontal closed orbit deviation [m].
Q2 The vertical tune Q2[1].
DQ2 The vertical chromaticity dq2=∂Q2/∂pt, [1]
BETYMAX The largest vertical βy, [m].
DYMAX The maximum of the absolute vertical dispersion Dy, [m].
DYRMS The r.m.s. of the vertical dispersion Dy, [m].
YCOMAX The maximum of the absolute vertical closed orbit deviation [m].
YCORMS The r.m.s. of the vertical closed orbit deviation [m].
DELTAP Momentum difference, divided by the reference momentum [1].
DELTAP = ∆p/p0
SYNCH 1 First synchrotron radiation integral
SYNCH 2 Second synchrotron radiation integral
SYNCH 3 Third synchrotron radiation integral
SYNCH 4 Fourth synchrotron radiation integral
SYNCH 5 Fifth synchrotron radiation integral
22 CHAPTER 1. CONVENTIONS
Notice that in MAD-X, PT substitutes DELTAP as longitudinal variable. Dispersive and chro-
matic functions are hence derivatives with respect to PT. And since PT=BETA*DELTAP,
where BETA is the relativistic Lorentz factor, those functions given by MAD-X must be mul-
tiplied by BETA a number of time equal to the order of the derivative to find the functions
given in the literature.
1.7.6 Variables in the TRACK Table
The command RUN writes tables with the following variables:
XHorizontal position xof the orbit, referred to the ideal orbit [m].
PX Horizontal canonical momentum pxof the orbit referred to the ideal orbit,
divided by the reference momentum.
YVertical position yof the orbit, referred to the ideal orbit [m].
PY Vertical canonical momentum pyof the orbit referred to the ideal orbit, divided
by the reference momentum.
TVelocity of light times the negative time difference with respect to the reference
particle, T=−c∆t, [m]. A positive T means that the particle arrives ahead of
the reference particle.
PT Energy difference, divided by the reference momentum times the velocity of
light, [1].
When tracking Lyapunov companions, the TRACK table defines the following dependent
expressions:
DISTANCE the relative Lyapunov distance between the two particles.
LYAPUNOV the estimated Lyapunov Exponent.
LOGDIST the natural logarithm of the relative distance.
LOGTURNS the natural logarithm of the turn number.
1.8 Physical Units
MAD-X uses units derived from the “Syst`eme International” (SI). These units are summarised
in the Units table.
![](asset-19.png)
1.8. PHYSICAL UNITS 23
Table 1.1: Physical Units used by MAD-X
Quantity Unit
Length m (metres)
Angle rad (radians)
Quadrupole coefficient m−2
Multipole coefficient, 2n poles m−n
Electric voltage MV (megavolts)
Electric field strength MV/m
Frequency MHz (megahertz)
Phase angles 2π
Particle energy GeV
Particle mass GeV/c2
Particle momentum GeV/c
Beam current A (amp`eres)
Particle charge e (elementary charges)
Impedance MΩ (Megohms)
Emittance π∗10−3m.rad
RF power MW (megawatts)
Higher order mode loss factor V/pc
![](asset-1a.png)
Chapter 2. Command Format
2.1 Input Statements
Input for MAD-X follows in broad lines the MAD-9 format, i.e. free format with commas (,) as
separators, although blanks may be used as separators outside {...}enclosures.
Blank input lines do not affect program execution.
The input is not case sensitive except for strings enclosed in double quotes (” ”).
The input file consists of a sequence of statements. A statement may occupy any number
of input lines. Several statements may be placed on the same line. A statement must be
terminated by a semicolon (;).
Several statements may be grouped into a block enclosed inside a {...}enclosure. In this case
the terminating semmicolon can be omitted.
if (a <3) {a=b*b; b=2*b+4; };
or
if (a <3) {a=b*b; b=2*b+4; }
are both valid.
2.2 Comments
When an exclamation mark (!) or double forward slash (//) is found in the input line, the
remaining characters until the end of the line are treated as a comment and are skipped.
A comment spreading over multiple lines starts with a string /* and ends with a string */.
2.3 Commands
The general format for a command is
label: keyword {,attribute};
where the { } are not part of the command and the items enclosed in { } can be omitted or
repeated any number of times.
A command contains three parts:
label A label is required for a definition statement. A label gives a name to the
stored command.
keyword A keyword identifies the action desired.
24
2.4. KEYWORDS 25
attributes The attributes are normally entered in the form
attribute-name = attribute-value
and serve to define data for the command, where:
attribute-name selects the attribute, and
attribute-value provides its value.
If a value is to be assigned to an attribute, the attribute-name is mandatory.
Whenever an attribute is not explicitly given a value, the default attribute-value specified
in the command dictionary is assumed.
2.4 Keywords
A keyword begins with a letter and consists of letters and digits.
The MAD-X keywords are protected; attempting to use a protected keyword as a label results
in a fatal error.
The keyword VERSION has been introduced since version 5.02.07. It contains the version
number of the MAD-X release quoted as a decimal. For example:
X:> value VERSION;
version = 50207 ;
This allows testing of the version number of the running MAD-X process. Note that any version
prior to version 5.02.07 reports the value version = 0;
2.5 Attribute Types
The command attributes can have one of the following types:
•String attribute,
•Logical attribute,
•Integer attribute,
•Real attribute,
•Expression,
•Range selection,
Any integer or real attribute can be replaced by a real expression; expressions are normally de-
ferred (see deferred expression), except in the definition of constants where they are evaluated
immediately.
![](asset-1c.png)
26 CHAPTER 2. COMMAND FORMAT
When a command has a label,MAD-X keeps this command in memory. This allows repeated
execution of the same command by simply entering EXEC label;. This construct may be
nested.
2.6 Variable Declarations
In the following, ”=” means that the variable on the left receives the current value of the
expression on the right, but does not depend on it any further, whereas ”:=” makes the
variable on the left depend on the expression on the right, i.e. every time the expression
changes, the variable is re-evaluated, except for ”const” variables.
var = expression;
var := expression;
real var = expression; // identical
real var := expression; // to above
int var = expression; // truncated if expression is real
int var := expression;
const var = expression;
const var := expression;
const real var = expression; // identical
const real var := expression; // to above
const int var = expression; // truncated if expression is real
const int var := expression;
2.7 Identifiers or Labels
A label begins with a letter, followed by up to fifteen letters, digits, decimal points (.), or
underscores ( ). Characters beyond the sixteenth are dropped, but should be avoided, and
the resulting sequence must be unique.
A label may refer to a keyword, an element, a beam-line, a sequence, etc.
The MAD-X keywords are protected; using one of them as a label results in a fatal error.
2.8 Command Attributes
•A name or string attribute refers to an object, or a string.
•A logical attribute selects or deselects an option.
•An integer attribute defines a value stored as integer data.
•A real attribute defines a value stored as double precision data.
•A real expression defines a datum for a command, it may be varied in matching. An
expression is built of a combination of operator and operand.
2.9. NAME OR STRING ATTRIBUTES 27
•A constraint specifies a matching constraint.
•A variable name selects a variable to be matched.
2.9 Name or String Attributes
A name or string attribute often selects one of a set of options:
USE, PERIOD=lhc; // expand the LHC sequence
It may also refer to a user-defined object:
TWISS, FILE=optics; // specifies the name of the OPTICS output file
It may also define a string:
TITLE, "LHC version 6.2";
The case of letters is only significant if a string is enclosed in quotes, otherwise all characters
are converted to lowercase at reading.
On the other hand, strings that do not contain blanks do not need to be enclosed in quotes.
Example:
CALL, FILE = "my.file";
CALL, FILE = my.file;
CALL, FILE = MY.FILE;
CALL, FILE = "MY.FILE";
CALL, FILE = ’MY.FILE’;
In the first three cases, MAD-X will try to read a file named my.file, in the last two it will try
to read the file named MY.FILE.
A string attribute makes alphanumeric information available, e.g. a title or a file name. It
can contain any characters, enclosed in single (’) or double (”) quotes, except for quotes of
the type used as its delimiter.
Examples:
TITLE, ’This is a title for the program run "test"’;
SYSTEM, "ln -fns some-lengthy-directory-name local-link";
2.10 Logical Attributes
Many commands in MAD-X require the setting of logical values (flags) to represent the on/off
state of an option. A logical value ”flag” can be set in two ways:
flag | flag = true
It can be reset by:
28 CHAPTER 2. COMMAND FORMAT
-flag | flag = false
Example:
OPTION, -ECHO; // switch off copying the input to the standard output
The default for a logical flag is normally false, but can be found e.g. for options by the
command
HELP, option;
2.11 Integer Attributes
An integer attribute usually denotes a count. Example:
myline: LINE = ( -3 * (a,b,c) );
In this case, a litteral integer is requested; however, in the following
rfc: RFCAVITY, HARMON = 12345;
or
rfc: RFCAVITY, HARMON = num;
”num” may be an integer variable, a real variable, or an expression. In the two latter cases,
the value is truncated.
2.12 Real Attributes
Most attributes are of type REAL and are treated internally as double precision values. They
may be set by integer values, real values, or expressions.
Example:
ddd: drift, l = 1.2345;
dddd: drift, l = ddd->l-0.3;
2.13 Real Expressions
To facilitate the definition of interdependent quantities, any real value and integer value can
be entered as an arithmetic expression. When a value used in an expression is redefined by the
user or changed in a matching process, the expression is reevaluated. Expression definitions
may be entered in any order. MAD-X evaluates them in the correct order before it performs
any computation. At evaluation time all operands used must have values assigned.
An expression is built from a combination of operator and operand, and it may contain
random generators.
2.13. REAL EXPRESSIONS 29
2.13.1 Operators in Arithmetic Expressions
An expression can be formed using the following operators:
Arithmetic operators
+Addition,
-Subtraction,
*Multiplication,
/Division,
^Exponentiation.
Ordinary Functions
SQRT(x) square root,
LOG(x) natural logarithm,
LOG10(x) logarithm base 10,
EXP(x) exponential,
SIN(x) trigonometric sine,
COS(x) trigonometric cosine,
TAN(x) trigonometric tangent,
ASIN(x) arc sine,
ACOS(x) arc cosine,
ATAN(x) arc tangent,
SINH(x) hyperbolic sine,
COSH(x) hyperbolic cosine,
TANH(x) hyperbolic tangent,
SINC(x) cardinal sine function,
ABS(x) absolute value,
ERF(x) Gauss error,
ERFC(x) complementary error,
FLOOR(x) floor, largest previous integer,
CEIL(x) ceiling, smallest next integer,
ROUND(x) round, closest integer;
![](asset-20.png)
30 CHAPTER 2. COMMAND FORMAT
Random Number Generators
RANF() random number, uniformly distributed in [0,1],
GAUSS() random number, gaussian distribution with unit standard deviation,
TGAUSS(x) random number, gaussian distribution with unit standard deviation, truncated
at x standard deviations;
in the above cases, ”x” can be any expression, i.e. contain other functions, variable or constant
expressions. To initialize the MAD-X random generator use the EOPTION command.
Table Access Functions
TABLE(x,z) accesses value of the named table column ”z” of table ”x”; example: ta-
ble(summ,q1) returns the hor. tune of the Twiss summary table ”summ”.
TABLE(x,y,z) accesses value of the named table column ”z” for element ”y” (first table row
with that name) of table ”x”; example: table(twiss,mb.12,betx) returns the
beta x at element mb.12 from the Twiss table ”twiss”. When the element is
called with its proper name, as in the example above, the value is returned
at the first occurrence of the element of this name. If the value is needed at
a occurrence number: NNN, then ”[NNN]” has to be appended to the name,
e.g. in the above example one obtains the betx of the 23th occurrence of the
element ”mb.12” by changing the example to: ”table(twiss,mb.12[23],betx)”.
Mind you that the old, but little known, form: ”table(twiss,mb.12->23,betx)”
continues to work. Lastly, if NNN exceeds the maximum occurrence number
the return value is arbitrarily small.
Beware:
•Unnamed Drifts are not included in the table name database, so as to speed up the
search for ”real” elements. Therefore, those unnamed drifts cannot be found. However,
named drifts can be searched for.
•Due to the importance of finding elements in the table for a proper functioning of the
MAD-X runs, the programs throws a ”fatal error” if an element cannot be found in the
table.
There is a second option of this function with 3 entries
•table(x,z,N row): accesses the value of the named table column ”z” at the ”N row”
number of rows of table ”x” (row numbers start at 1); example: table(twiss,betx,370)
returns the beta x at row number ”370” of the Twiss table ”twiss”. The return value
is zero if ”N row” is outside of the allowed range.
Note that ”N row” has to be a number and not a variable. However, the Macro facility in
MAD-X allows the use of a variable instead.
A typical example could look like this, in fact the square root of betx (user defined variable
myvar) is added to the twiss table:
![](asset-21.png)
2.13. REAL EXPRESSIONS 31
myvar := sqrt(table(twiss,betx));
select, flag=twiss, column=name, s, myvar, betx;
twiss, file;
Or another arbitrary test case of adding k1l taken from the Twiss table:
Define macro:
mymacro(xx,yy,zz): macro = {myval = table(xx,yy,zz);};
Use macro in loop:
i = 0;
incval = 0;
while (i < 100) {
value, i;
exec, mymacro(twiss,k1l,$i);
incval = incval + myval;
value, i, myval, incval;
i=i+1;
}
2.13.2 Operands in Arithmetic Expressions
An expression may contain the following operands:
Literal Constants
Numerical values are entered like FORTRAN constants. Real values are accepted in INTE-
GER or REAL format. The use of a decimal exponent, marked by the letter D or E, is
permitted.
Examples:
1, 10.35, 5E3, 314.1592E-2
Symbolic constants
MAD-X recognizes some mathematical and physical constants. Their names must not be used
for user-defined labels.
Additional symbolic constants may be defined to simplify their repeated use in statements
and expressions.
CONST name=constant-expression;
defines a real constant with the name given. An existing symbolic constant can be redefined,
but it cannot change in a matching procedure.
Example:
![](asset-22.png)
32 CHAPTER 2. COMMAND FORMAT
CONST IN = 0.0254;
A number of predefined symbolic constants exist in MAD-X and can be used in expressions.
The values of physical constants are regularly updated to reflect the values published by the
Particle Data Group [5] The values published in 2014 ([5]) have not changed with respect to
the values published in 2012 ([6]).
Table 2.1: Predefined Symbolic Constants in MAD-X
MAD-X name symbol value unit
PI π4 * atan(1) 1
TWOPI 2π2 * PI 1
DEGRAD 180/π 180 / PI deg/rad
RADDEG π/180 PI / 180 rad/deg
E e exp(1) 1
EMASS me0.510998928 ×10−3GeV
PMASS mp0.938272046 GeV
NMASS u0.931494061 GeV
MUMASS mµ0.1056583715 GeV
CLIGHT c2.99792458 ×108m/s
QELECT e1.602176565 ×10−19 A.s
HBAR ¯h6.58211928 ×10−25 MeV.s
ERAD re2.8179403267 ×10−15 m
PRAD re(me/mp) ERAD*EMASS/PMASS m
Note that the NMASS constant in MAD-X is the unified atomic mass unit and not the neutron
mass.
Parameter labels
Often a set of numerical values depends on a common variable parameter. Such a parameter
must be defined as a global parameter. A global parameter always has a current value;
however, this value may be re-evaluated or not, depending on the parameter definition:
x = a;
xis set to the current value of aand not changed, even if achanges. This makes assignments
such as
x=x+1;
perfectly valid (this replaces the deprecated SET instruction).
The definition of the deferred expression
x := a;
assign the current value of ato xevery time xis used, i.e. it is re-evaluated using the latest
value of a; therefore,
x := x + 1;
![](asset-23.png)
2.13. REAL EXPRESSIONS 33
results in an infinite loop (!) when xis used, which results in abnormnal termination of MAD-X.
Of course, the following definitions are equivalent:
x = 0.1;
x := 0.1;
When such a parameter is used in an expression, MAD-X uses the current value of the parameter
if the expression is deferred:
Example:
x := 1.0;
d1: drift, l = x;
d2: drift, l := 2.0 - x;
When the value of xis changed, the length of the drift d1 remains unchanged, while the
length of the drift d2 is recalculated.
Element or Command Attributes
In arithmetic expressions the attributes of physical elements or commands can occur as
operands. They are named respectively by
element-name->attribute-name
command-name->attribute-name
Values are assigned to attributes in element definitions or commands.
Example:
D1: DRIFT, L= 1.0;
D2: DRIFT, L= 2.0 - D1->L;
D1->Lrefers to the length Lof the drift space D1.
2.13.3 Expressions and Random Values
The definition of random machine imperfections requires evaluation of expressions containing
random functions. These are evaluated like any other expression when the expression is used,
i.e. only once if a ”=” assignment refers to it, or every time if the assignment is ”:=”; this
latter case is used by the error generation routines.
Example:
a := 3*ranf();
Every time ais used, it gets a random value assigned from a uniform distribution between 0
and 3.
error: ealign, range, dx:=sigma*gauss();
![](asset-24.png)
34 CHAPTER 2. COMMAND FORMAT
All elements in range are assigned independent random displacements sampled from a Gaus-
sian distribution with standard deviation sigma.
2.14 Logical Expressions
In matching it is desired to specify equality constraints, as well as lower and upper limits for
a quantity. MAD-X accepts the following forms of constraints:
name = expression ! equality constraint
name < expression ! upper limit
name > expression ! lower limit
name < expression, name > expression ! both upper and lower limit
! for the same name
2.15 Variable Names
A variable name can have one of the formats:
parameter-name
element-name->attribute-name
command-name->attribute-name
beam%sequence-name->attribute-name
table(table-name,..,..)
The first format refers to the value of the global parameter parameter-name.
The second and third formats refer to the real attribute attribute-name of the element
element-name, or the command command-name.
The fourth format is specific to beams belonging to a particular sequence (for details see
sequences and beams).
The fifth format allows extraction of variables from existing tables, as specified in table access.
2.16 Regular Expressions
Some commands allow selection of items via ”regular expression” strings. Such a pattern
string must be enclosed in single or double quotes. MAD-X follows regexp (Unix regular
expression patterns) for matching. The following features are implemented:
A ”search string” below is the string containing the pattern, a ”target string” is the string
being searched for a possible match with the pattern.
"^" at the start of the search string: Match following search string at the start of
the target string; otherwise the search string can start anywhere in the target
string. To search for a genuine ”ˆ” anywhere, use ”\ˆ”.
![](asset-25.png)
2.17. RELATIONS BETWEEN VARIABLE PARAMETERS 35
"$" at the end of the search string: Match preceding search string at the end of
the target string; otherwise the search string can end anywhere in the target
string. To search for a genuine ”$” anywhere, use ”\$”.
"." stands for an arbitrary character; to search for a genuine ”.”, use ”\.”
"[xyz]" stands for one character belonging to the string contained in brackets (example:
”[abc]” means one of a, b, c).
"[a-ex-z]" stands for ranges of characters (example: ”[a-zA-Z]” means any letter).
"[^xyz]" (i.e. a ”ˆ” as first character in a square bracket) stands for exclusion of all
characters in the list, i.e. ”[ˆa-z]” means ”any character but a lower case
letter”.
"*" allows zero or more repetitions of the preceding character, either specified
directly, or from a list. (examples: ”a*” means zero or more occurrences of
”a”, ”[A-Z]*” means zero or more upper-case letters).
"\c" (e.g. ”\.”) removes the special meaning of character c.
All other characters stand for themselves.
Example:
SELECT, FLAG=twiss, PATTERN="^d..$" ;
SELECT, FLAG=twiss, PATTERN="^k.*qd.*\.r1$" ;
The first command selects all elements whose names have exactly three characters and begin
with the letter ”D”. The second command selects elements beginning with the letter ”K”,
containing the string ”QD”, and ending with the string ”.R1”. The two occurrences of ”.*”
each stand for an arbitrary number (including zero) of any character, and the occurrence ”\.”
stands for a literal period.
2.17 Relations between Variable Parameters
A relation is established between variables by one of two statements
parameter-name = expression;
parameter-name := expression;
The first form evaluates the expression on the right immediately and assigns its value to the
parameter. It is an immediate assignment.
The second form assigns the value by evaluating the expression on the right every time the
parameter is actually used. It is a deferred assignment.
This mechanism holds as well for element parameters that can be defined with either imme-
diate or deferred assignments.
Attention! If you want to modify e.g. the strength of a quadrupole later (e.g. in a match, or
by entering a new value for a parameter on which it depends) then the defition has to be
36 CHAPTER 2. COMMAND FORMAT
QD: QUADRUPOLE, K1 := ak1;
and not
QD: QUADRUPOLE, K1 = ak1;
In the latter case, K1 will be set to the current value of ak1 at the time of declaration, and
will not change when ak1 later changes.
Parameters that have not yet been defined at time of evaluation have a zero value.
Example:
gev = 100;
BEAM, ENERGY=gev;
The parameter on the left may appear on the right as well in the computer science form of
assignments:
x = x+1;
increases the value of x by 1.
Successive definitions are allowed in the first form of relations or immediate assignments:
a = b + 2;
b = a * b;
But circular definitions in the second form of relations, or deferred assignments, are forbidden:
a := b + 2;
b := a * b;
results in an error.
![](asset-27.png)
Chapter 3. Program Flow Statements
3.1 IF...ELSEIF...ELSE
IF (logical expression) {statements; }
ELSEIF (logical expression) {statements; }
ELSE {statements; }
where ”logical expression” is one of
expr1 oper expr2
expr11 oper1 expr12 && expr21 oper2 expr22
expr11 oper1 expr12 || expr21 oper2 expr22
and oper is one of
== equal
<> not equal
<less than
>greater than
<= less than or equal
>= greater than or equal
The expressions are arithmetic expressions of type real. The statements in the curly brackets
are executed if the logical expression is true.
ELSEIF constructs are only possible (in any number) behind an IF, or another ELSEIF; the
branch is executed if ”logical expression” is true, and if none of the preceding IF or ELSEIF
logical conditions was true.
ELSE construct is only possible once behind an IF, or an ELSEIF; the branch is executed if
”logical expression” is true, and if none of the preceding IF or ELSEIF logical conditions was
true.
Warning:
Because IF ... ELSEIF ... ELSE constructs are a MAD-X special feature and not part of
a full language, MAD-X does not deal gracefully with other special constructs such as MACRO
or LINE when they are placed inside IF ... ELSEIF ... ELSE statements: this can lead
to silent and/or catastrophic errors and is due to the fact that MACRO and LINE constructs
contain, either explicitly or implicitly, a closing curly bracket that unbalances the IF ...
ELSEIF ... ELSE statements.
However it is possible to nest IF ... ELSEIF ... ELSE constructs to at least six levels
deep.
37
![](asset-28.png)
38 CHAPTER 3. PROGRAM FLOW STATEMENTS
3.2 WHILE
WHILE (logical condition) {statements; }
executes the statements in curly brackets while the logical expression is true.
Warning:
Because WHILE constructs are a MAD-X special feature and not part of a full language, MAD-X
does not deal gracefully with other special constructs such as MACRO or LINE when they are
placed inside WHILE statement blocks: this can lead to silent and/or catastrophic errors and
is due to the fact that MACRO and LINE constructs contain, either explicitly or implicitly, a
closing curly bracket that unbalances the WHILE statements.
However it is possible to nest WHILE statements to at least six levels deep.
Example giving the value of the first ten factorials:
n = 1; m = 1;
while (n <= 10) {
m = m * n; value, m;
n = n + 1;
};
3.3 MACRO
The MACRO construct allows the execution of a group of statements via a single command.
Optionally the MACRO construct takes arguments.
label: MACRO = {statements; };
label(arg1, ...,argn): MACRO = {statements; };
The first form allows the execution of the defined group of statements via a single command,
EXEC, label;
that executes the statements defined between curly brackets exactly once. The EXEC command
can then be issued any number of times.
The second form allows to replace strings anywhere inside the statements in curly brackets by
other strings, or integer numbers prior to execution. This is a powerful construct and should
be handled with care.
Simple example:
simple(xx,yy): MACRO = {xx = yy*yy + xx; VALUE, xx;};
a = 3; b = 5;
EXEC, simple(a,b);
yields
3.3. MACRO 39
a = 28 ;
Passing arguments
In the following example we use the fact that a ”$” in front of an argument means that the
truncated integer value of this argument is used for replacement, rather than the argument
string itself.
tricky(xx,yy,zz): MACRO = {mzz.yy: xx, l = 1.yy, kzz = k.yy;};
n=0;
WHILE (n < 3) {
n = n+1;
EXEC, tricky(quadrupole, $n, 1);
EXEC, tricky(sextupole, $n, 2);
};
Because MACRO statements are a MAD-X construct and not part of a full language, MAD-X
allows only one level of inclusion of another IF ... ELSEIF ... ELSE,WHILE or MACRO
statements.
Macros cannot be called with number arguments but always with string arguments. In case
numerical values should be passed to a MACRO in an EXEC statement, one can conveniently use
variables names:
n1=99; n2=219;
EXEC, thismacro($n1, $n2);
instead of
EXEC, thismacro($99, $129); ! fails...
![](asset-2a.png)
Chapter 4. General Control Statements
MAD-X consists of a core program, and modules for specific tasks such as twiss parameter
calculation,matching,thin lens tracking,etc.
The statements listed here are those executed by the program core. They deal with the I/O,
element and sequence declaration, sequence manipulation, statement flow control (e.g. IF,
WHILE), MACRO declaration, saving sequences onto files in MAD-X or MAD-8 format, etc.
4.1 EXIT, QUIT, STOP
Any of these three commands ends the execution of MAD-X:
EXIT;
QUIT;
STOP;
4.2 HELP
The HELP command prints all parameters, and their defaults values, for the statement given;
this includes basic element types.
HELP, statement name;
4.3 SHOW
The SHOW command prints the command (typically beam,beam%sequ, or an element name),
with the actual value of all its parameters.
SHOW, command;
4.4 VALUE
The VALUE command evaluates the current value of all listed expressions, constants or vari-
ables, and prints the result in the form of MAD-X statements on the assigned output file.
VALUE, expression{, expression};
Example:
40
![](asset-2b.png)
4.5. OPTION 41
a = clight/1000.;
value, a, pmass, exp(sqrt(2));
results in
a = 299792.458 ;
pmass = 0.938272046 ;
exp(sqrt(2)) = 4.113250379 ;
4.5 OPTION
The OPTION commands sets the logical value of a number of flags that control the behavior
of MAD-X.
OPTION, flag=logical;
Because all attributes of OPTION are logical flags, the following two statements are identical:
OPTION, flag = true;
OPTION, flag;
And the following two statements are also identical:
OPTION, flag = false;
OPTION, -flag;
Several flags can be set in a single OPTION command, e.g.
OPTION, ECHO, WARN=true, -INFO, VERBOSE=false;
The available flags, their default values and their effect on MAD-X when they are set to TRUE
are listed in table 4.1.
The option RBARC is implemented for backwards compatibility with MAD-8 up to version
8.23.06 included; in this version, the RBEND length was just taken as the arc length of an
SBEND with inclined pole faces, contrary to the MAD-8 manual.
4.6 EXEC
Each statement may be preceded by a label, when parsed and executed the statement is then
also stored and can be executed again with
EXEC, label;
This mechanism can be invoked any number of times, and the executed statement may be
calling another EXEC, etc.
tw: TWISS, FILE, SAVE; ! first execution of TWISS
...
![](asset-2c.png)
42 CHAPTER 4. GENERAL CONTROL STATEMENTS
Table 4.1: Flags available to OPTION command
FLAG default effect if TRUE
ECHO true echoes the input on the standard output file
WARN true issues warning statements
INFO true issues information statements
DEBUG false issues debugging information
ECHOMACRO false issues macro expansion printout for debugging
VERBOSE false issues additional printout in makethin
TRACE false prints the system time after each command
VERIFY false issues a warning if an undefined variable is used
TELL false prints the current value of all options
RESET false resets all options to their defaults
NO FATAL STOP false Prevents madx from stopping in case of a fatal error
Use at your own risk !
RBARC true converts the RBEND straight length into the arc length
THIN FOC true enables the 1/ρ2focusing of thin dipoles
BBORBIT false the closed orbit is modified by beam-beam kicks
SYMPL false all element matrices are symplectified in Twiss
TWISS PRINT true controls whether the twiss command produces output
THREADER false enables the threader for closed orbit finding in Twiss
(see Twiss module)
EXEC, tw; ! second execution of the same TWISS command
Note however, that the main usage of this MAD-X construct is the execution of a MACRO.
4.7 SET
The SET command is used in two forms:
SET, FORMAT=string {, string};
SET, SEQUENCE=string;
The first form of the SET command defines the formats for the output precision that MAD-X
uses with the SAVE,SHOW,VALUE and TABLE commands. The formats can be given in any
order and stay valid until replaced.
The formats follow the C convention and must be included in double quotes. The allowed
formats are
ndfor integers with a field-width = n,
n.mfor n.mgor n.mefor floats with field-width = nand precision = m,
nsfor strings with a field-width = n.
The default is ”right adjusted”, a ”-” changes it to ”left adjusted”.
Example:
![](asset-2d.png)
4.8. SYSTEM 43
SET, FORMAT="12d", "-18.5e", "25s";
The default formats are "10d", "18.10g" and "-18s".
Example:
set, format="22.14e";
changes the current floating point format to 22.14e; the other formats remain unchanged.
set, format="s","d","g";
sets all formats to automatic adjustment according to C conventions.
The second form of the SET command allows to select the current sequence without the USE
command, which would bring back to a bare lattice without errors. The command only
works if the chosen sequence has been previously activated with a USE command, otherwise a
warning is issued and MAD-X continues with the unmodified current sequence. This command
is particularly useful for commands that do not have the sequence as an argument like EMIT
or IBS.
4.8 SYSTEM
SYSTEM, "string";
transfers the quoted string to the operating system for execution. The quotes are stripped
and no check of the return status is performed by MAD-X.
Example:
SYSTEM, "ln -s /afs/cern.ch/user/j/joe/input shortname";
makes a local link to an AFS directory with the name shortname on a UNIX system.
Attention: Although this command is very convenient, it is clearly not portable across sys-
tems and it should probably be avoided if one intends to share MAD-X scripts with collaborators
working on other platforms.
4.9 TITLE
TITLE, "string";
defines a string that is inserted as title in various table outputs and plot results.
![](asset-2e.png)
44 CHAPTER 4. GENERAL CONTROL STATEMENTS
4.10 USE
MAD-X operates on beamlines that must be loaded and expanded in memory before other
commands can be invoked. The USE command allows this loading and expansion.
USE, SEQUENCE=sequence name, PERIOD=sequence name,
RANGE=range,
SURVEY=logical;
The attributes to the USE command are:
SEQUENCE name of the sequence to be loaded and expanded.
PERIOD name of the sequence to be loaded and expanded.
PERIOD is an alias to SEQUENCE that was kept for backwards compatibility with
MAD-8 and only one of them should be specified in a USE statement.
RANGE specifies a range. restriction so that only the specified part of the named
sequence is loaded and expanded.
SURVEY option to plug the survey data into the sequence elements nodes on the first
pass (see SURVEY).
Note that reloading a sequence with the USE command reloads a bare sequence and that any
ERROR or orbit correction previously assigned or associated to the sequence are discarded. A
mechanism to select a sequence without this side effect of the USE command is provided with
the SET, SEQUENCE=... command.
4.11 SELECT
Some MAD-X commands can perform specific operations on selected elements or ranges of
elements and can produce specific output for selected elements or ranges of elements.
The selection is made through the SELECT command and applies to subsequent commands.
SELECT, FLAG=string, RANGE=string, CLASS=string, PATTERN=string,
SEQUENCE=string, FULL=logical, CLEAR=logical,
COLUMN=string{,string}, SLICE=integer, THICK=logical;
The attributes to the SELECT command are:
FLAG determines the applicability of the SELECT statement and the attribute value
can be one of the following:
SEQEDIT selection of elements for the SEQEDIT module.
ERROR selection of elements for the error assignment module.
MAKETHIN selection of elements for the MAKETHIN command that converts
the sequence into one with thin elements.
![](asset-2f.png)
4.11. SELECT 45
SECTORMAP selection of elements for the SECTORMAP output file from the
TWISS module.
SAVE selection of elements for the SAVE command.
tablename is a table name such as twiss,track etc., and the rows and
columns to be written are selected.
RANGE the range of elements to be selected as defined in section 11.1 on range selection.
CLASS the class of elements to be selected as defined in section 11.2 on class selection.
PATTERN the regular expression pattern for the element names to be selected as defined
in section 2.16 on selection via regular expressions.
SEQUENCE the name of a sequence to which the selection is applied.
FULL a logical falg to select ALL positions in the sequence for the named flag.
For the flag TWISS, this attribute includes all standard columns for a TWISS
table, and therefore the following two statements are equivalent:
SELECT, FLAG=twiss, COLUMN= name, s, betx, ..., var1;
SELECT, FLAG=twiss, FULL, COLUMN= var1;
FULL=true is the default for the MAKETHIN flag and for tables: e.g. SELECT,
FLAG=makethin; is equivalent to SELECT, FLAG=makethin, FULL;
CLEAR deselects ALL positions in the sequence for the flag ”name”. This is the default
for ERROR and SEQEDIT flags.
e.g. SELECT, FLAG=error; is equivalent to SELECT, FLAG=error, CLEAR;
COLUMN is only valid for tables and takes as attribute value a list of columns to be
written into the TFS file. The special ” name” argument refers to the actual
name of the element.
SLICE is the number of slices into which the selected elements have to be cut and is
only used by MAKETHIN. (Default = 1).
THICK is a logical flag to indicate whether the selected elements are treated as thick
elements by the MAKETHIN command.
This only applies up to now to QUADRUPOLEs and BENDs for which thick maps
have been explicitely derived.
Composition of SELECT statements:
The selection criteria provided on a single SELECT statement are logically ANDed, i.e. selected
elements have to fulfill all provided criteria in the single SELECT statement.
The selection criteria on different SELECT statements are logically ORed, i.e. selected elements
have to fulfill any of the selection criteria provided by the different SELECT statements.
All selections for a given flag remain valid until a SELECT statement with the CLEAR argument
is specified for the same flag.
![](asset-30.png)
46 CHAPTER 4. GENERAL CONTROL STATEMENTS
Note that because of these composition rules, it is considered good practice to start by clearing
the selection for a given flag before making a new selection, eg:
SELECT, FLAG=twiss, CLEAR;
SELECT, FLAG=twiss, CLASS=MQ;
SELECT, FLAG=twiss, RANGE=MQ[5]/MQ[7];
...
Examples:
SELECT, FLAG = ERROR, CLASS = quadrupole, RANGE = mb[1]/mb[5];
SELECT, FLAG = ERROR, PATTERN = "^mqw.*";
selects all quadrupoles in the range mb[1] to mb[5], as well as all elements (in the whole
sequence) with name starting with ”mqw”, for treatment by the ERROR module.
SELECT, FLAG=SAVE, CLASS=variable, PATTERN="abc.*";
SAVE, FILE=mysave;
saves all variables (and sequences) containing ”abc” in their name, but does not save elements
with names containing ”abc” since the class ”variable” does not exist.
sig1 := sqrt(beam->ex*table(twiss,betx));
SELECT, FLAG=twiss, COLUMN= name, s, betx, ..., sig1; ! or equivalently
SELECT, FLAG=twiss, FULL, COLUMN= sig1; ! default columns + new
writes the current value of “sig1” into the TWISS table each time a new line is added; Note
that the values from the same (current) line can be are accessed by the variable “sig1”. The
PLOT command also accepts the new variable in the table.
![](asset-31.png)
Chapter 5. File Handling Statements
Note that the filenames given as attribute values in File Handling statements must be explicit
names and should not contain wildcard characters since the filename strings are not passed
to the underlying Operating System for evaluation.
5.1 ASSIGN
ASSIGN, ECHO="filename", TRUNCATE;
where filename is either the name of an output file, or the string terminal. This allows
switching the echo stream to a file or back to the terminal, but only for the commands VALUE,
SHOW, and PRINT. Allows easy composition of future MAD-X input files.
TRUNCATE specifies whether the file is truncated when opened (ignored for terminal).
5.2 CALL
CALL, FILE="filename";
where filename is the name of an input file. The named file is then read until a RETURN
statement is encountered, or until the End Of File; The file being ”called” may in turn contain
any number of CALL statements itself, and so on to any depth.
5.3 RETURN
RETURN;
ends the reading from a ”called” file. If encountered in the standard input file, it ends the
program execution.
5.4 PRINT
PRINT, TEXT="string";
prints the quoted text string to the current output file (see ASSIGN above). The text can be
edited with the help of a macro statement.
5.5 PRINTF
PRINTF, TEXT="string", VALUE= expr,expr;
47
![](asset-32.png)
48 CHAPTER 5. FILE HANDLING STATEMENTS
prints the numerical values specified in the VALUE field to the current output file, with for-
matting according to the format string provided in the TEXT field.
The string format can take numeric C or MAD-X format specifiers for double real values.
Integer and string formats are not supported but can be approximated with the %g format in
the case of integers, and via MACRO statements, which perform string substitution, themselves
containing a PRINT statement.
The maximum number of values that can be printed in one statement is limited to 100.
If the number of format specifiers given in the string is higher than the number of values in
the value field, undefined values are printed where they are not explicitly provided.
If the number of format specifiers given in the string is lower than the number of values in
the value field, the values that do not correspond to a format specifier are ignored.
Example:
a = 1.2; b = 3.4/0.3; c := 0.8*a/b;
PRINTF,TEXT="String with floats a=%f, b=%.3g, text and MAD float c=%F;",
VALUE = a,b,c;
PRINTF,TEXT="More specifiers than values: %f, %.3g, %F", VALUE = a,b;
PRINTF,TEXT="More values than specifiers: %f, %.3g", VALUE = a,b,c;
produces the following output:
String with floats a=1.200000, b=11.3, text and MAD float c= 0.08470588235;
More specifiers than values: 1.200000, 11.3, 6.953222976e-310
More values than specifiers: 1.200000, 11.3
Note that PRINTF, like PRINT, produces output that cannot be read back by MAD-X. For output
that can be read back by MAD-X, use the command VALUE or TFS tables.
Note also that a percent sign (%) can be printed using the format text="%%".
5.6 RENAMEFILE
RENAMEFILE, FILE="filename", TO="new filename";
renames the file filename to new filename on disk.
This command is more portable than an equivalent SYSTEM call:
SYSTEM("mv filename new filename"); ! Unix specific
5.7 COPYFILE
COPYFILE, FILE="filename", TO="new filename", APPEND=logical;
![](asset-33.png)
5.8. REMOVEFILE 49
copies the file filename to the file new filename on disk.
The attribute APPEND=true causes COPYFILE to append the content of filename at the end
of the file new filename.
The default value APPEND=false causes the replacement of the content of new filename with
the content of filename.
COPYFILE, APPEND=true... is more portable than an equivalent SYSTEM call:
SYSTEM("copy /y filename new filename"); ! Windows specific
5.8 REMOVEFILE
REMOVEFILE, FILE="filename";
removes the file filename from disk.
It is more portable than an equivalent SYSTEM call:
SYSTEM("rm filename"); ! Unix specific
![](asset-34.png)
Chapter 6. Table Handling Statements
6.1 CREATE
CREATE, TABLE=tabname, COLUMN= var{, var} {, name};
creates a table with the specified variables as columns. The table created is initially empty
and can be subsequently filled, and eventually written to file in TFS format.
The special variable name attribute name (name preceded by underscore) adds the element
name to the table at the specified column.
6.2 DELETE
DELETE, SEQUENCE=seqname, TABLE=tabname;
deletes a sequence with name seqname or a table with name tabname from memory. The
sequence deletion is done without influence on other sequences that may have elements that
werein common with the deleted sequence.
6.3 READTABLE
READTABLE, FILE="filename";
reads the TFS file filename containing a MAD-X table and loads the table into memory with the
name specified in the information section of the TFS file. The table can then be manipulated
as any other table, i.e. its values can be accessed, its data can be plotted or changed, and it
can be written out again.
6.4 READMYTABLE
READMYTABLE, FILE="filename", TABLE=tabname;
reads a TFS file filename containing a MAD-X table and loads the table into memory with the
name tabname. The table can then be manipulated as any other table, i.e. its values can be
accessed, its data can be plotted or changed, and it can be written out again.
An internal name for the table can be freely assigned, while for the command READTABLE the
table name is taken from the information section of the table itself. This feature allows to
store multiple tables of the same type in memory without overwriting existing ones.
50
![](asset-35.png)
6.5. TABSTRING 51
6.5 TABSTRING
Note: this is not a command and should appear in the variables section
TABSTRING(arg1,arg2,arg3)
The ”string function” TABSTRING(arg1,arg2,arg3) with exactly three arguments; arg1 is a
table name (string), arg2 is a column name (string), arg3 is a row number (integer), count
starts at 0. The function can be used in any context where a string appears; in case there is
no match, it returns ” void ”.
6.6 WRITE
WRITE, TABLE=tabname, FILE="filename";
writes the table ”tabname” onto the file ”filename”; only the rows and columns of a preceding
SELECT, FLAG=table,...; are written. If no SELECT has been issued for this table, only the
header is written to file. If the FILE argument is omitted, the table is written to standard
output.
6.7 SETVARS
The SETVARS command sets the variables with values extracted from the row of a table.
SETVARS, TABLE=tabname, ROW=integer;
The attributes of SETVARS are:
TABLE the name of the table. (Default: none)
ROW the row number containing the values. (Default: -1)
Negative ROW values are allowed and count the row numbers from the last row, allowing access
to the table in reverse order of rows: ROW = -1 accesses the last row of the table, ROW = -2
accesses the penultimate (one before last) row, etc. . .
Trying to access the table forward beyond the last row, i.e. ROW strictly greater than nrow
the number of rows in the table, or trying to access the table backwards before the first row,
i.e. ROW strictly lower than -nrow, or trying to access the illegal ROW=0, all result in a “row
out of bound” message and no variable values are returned and set.
6.8 SETVARS LIN
The SETVARS LIN command sets the variables with values calculated by linear interpolation,
or extrapolation, between two rows of a table.
![](asset-36.png)
52 CHAPTER 6. TABLE HANDLING STATEMENTS
SETVARS LIN, TABLE=tabname,
ROW1=integer, ROW2=integer, PARAM=string;
The attributes of SETVARS LIN are:
TABLE the name of the table. (Default: none)
ROW1 a first row number with values for interpolation. (Default: 0)
ROW2 a second row number with values for interpolation. (Default: 0)
PARAM a string containing the linear interpolation factor or the name of a variable or
expression containing the interpolation factor. If the resulting value of PARAM
is outside the [0,1] interval, the result is a linear extrapolation.
(Default: ”interp”, itself defaulting to a value of 0.0 when evaluated)
SETVARS LIN sets the variables with values calculated through the following formula that
MAD-X constructs internally as a deferred expression which is immediately evaluated:
value := value(row1)*(1-param) + value(row2)*param;
Both the expression and the value of the expression are available to the user through respec-
tively the commands SHOW and VALUE.
When the values are represented as strings, e.g. the name or keyword of elements, the resulting
value is the string in ROW1.
Negative ROWivalues are allowed and count the row numbers from the last row, allowing
access to the table in reverse order of rows: ROWi= -1 accesses the last row of the table,
ROWi= -2 accesses the penultimate (one before last) row, etc. . .
Trying to access the table forward beyond the last row, i.e. ROWistrictly greater than nrow
the number of rows in the table, or trying to access the table backwards before the first row,
i.e. ROWistrictly lower than -nrow, or trying to access the illegal ROWi=0, all result in a “row
out of bound” message and the expression is not constructed or evaluated.
Example:
! extracts the position of the centre of each element from a standard
! TWISS table giving positions at end of elements:
len = table(twiss,tablelength);
interpolate = 0.5;
i = 2;
WHILE (i < len) {
SETVARS LIN, TABLE=twiss, ROW1=i-1, ROW2=i, PARAM=interpolate;
! now variables are interpolated at the center of the elements.
! in particular S holds the position of the center of the element.
SHOW, s; VALUE, s;
...
i=i+1;};
![](asset-37.png)
6.9. FILL 53
6.9 FILL
The FILL command fills a row of a table with the current values of all declared column
variables of the table.
FILL, TABLE=tabname, ROW=integer;
The FILL command takes two arguments:
TABLE is the name of the table to be filled. The table must have been created before-
hand. The table can then be written to file in TFS format.
ROW is the row number to be filled with the current values of all column variables.
ROW=0, or ROW=nrow + 1, where nrow is the current number of rows in the
table, causes FILL to add a row at the end of the table and fill it with the
current values of all column variables.
(Default: 0)
Negative ROW values are allowed and count the row numbers from the last row, allowing access
to the table in reverse order of rows: ROW = -1 accesses the last row of the table, ROW = -2
accesses the penultimate (one before last) row, etc. . .
Trying to access the table forward beyond the last row, i.e. ROW strictly greater than nrow
+ 1, where nrow is the number of rows in the table, or trying to access the table backwards
before the first row, i.e. ROW strictly lower than -nrow, both result in a “row out of bound”
message and no values are filled in the table.
Reminder: One can get access to the current number of rows in a table using the variable
TABLE(tablenanme, TABLELENGTH)
6.10 SHRINK
The SHRINK command removes a number of rows at the end of a table.
SHRINK, TABLE=tabname, ROW=integer;
The SHRINK command takes two arguments:
TABLE is the name of the table from which rows should be removed. The table must
have been previously created and filled or read from file with READTABLE or
READMYTABLE.
ROW is the number of the last row to be kept in the table. All rows beyond the
given row number are removed.
Negative values are allowed and count the row numbers from the last row,
allowing access to the table in reverse order of rows: ROW = -1 removes the
last row of the table, ROW = -2 removes the last two rows of the table, etc. . .
(Default: -1)
54 CHAPTER 6. TABLE HANDLING STATEMENTS
Trying to access the table forward beyond the last row, i.e. ROW strictly greater than nrow,
where nrow is the number of rows in the table, or trying to access the table backwards before
the first row, i.e. ROW strictly lower than -nrow, both result in a “row out of bound” message
and no values are filled in the table.
![](asset-39.png)
Chapter 7. Beam Handling Statements
Many commands in MAD-X require the prior setting of various quantities related to the beam
in the machine. Therefore, MAD-X will stop with a fatal error if an attempt is made to expand
(USE) a sequence for which no BEAM command has been issued before.
7.1 BEAM
The quantities are entered by a BEAM command:
BEAM, PARTICLE=string, MASS=real, CHARGE=real,
ENERGY=real, PC=real, GAMMA=real, BETA=real, BRHO=real,
EX=real, EXN=real, EY=real, EYN=real,
ET=real, SIGT=real, SIGE=real,
KBUNCH=integer, NPART=real, BCURRENT=real,
BUNCHED=logical, RADIATE=logical, BV=integer,
SEQUENCE=string;
The attributes of the BEAM command are:
PARTICLE The name of particles in the beam. Default=POSITRON
MAD-X knows the restmass and the charge for the following particles:
POSITRON The particles are positrons (MASS=me, CHARGE=1)
ELECTRON The particles are electrons (MASS=me, CHARGE=-1)
PROTON The particles are protons (MASS=mp, CHARGE=1)
ANTIPROTON The particles are anti-protons (MASS=mp, CHARGE=-1)
POSMUON The particles are positive muons (MASS=mµ, CHARGE=1)
NEGMUON The particles are negative muons (MASS=mµ, CHARGE=-1)
ION The particles are simple generic ions (MASS=u, CHARGE=1)
MASS the restmass of the particles in the beam in GeV.
(Default=me≈0.511 10−3GeV).
Note that a zero mass particle is not allowed in MAD-X.
CHARGE the electrical charge of the particles in the beam in units of qp, the proton
charge. (Default=1)
Note that a zero charge particle is not allowed in MAD-X.
The order of precedence for arguments is: particle->(mass+charge)
If the particle name given is recognized in the list above, the restmass and charge are set
directly by MAD-X, and the MASS and CHARGE arguments provided in the BEAM command are
simply ignored. For other particles, and in particular for ions, any combination of name, mass
55
![](asset-3a.png)
56 CHAPTER 7. BEAM HANDLING STATEMENTS
and charge can be entered independently.
ENERGY Total energy per particle in GeV.
If given, it must be greater than the particle restmass. (Default=1 GeV)
PC Particle momentum times the speed of light, in GeV.
If given, it must be greater than zero.
GAMMA Relativistic factor, i.e. ratio between total energy and rest energy of the par-
ticles: GAMMA = ENERGY/MASS = E/m0c2.
GAMMA must be greater than one.
BETA Ratio between the speed of the particle and the speed of light: BETA=v/c.
BETA must be strictly less than one.
BRHO Magnetic rigidity of the particles in T.m.
BRHO=P/abs(q)= PC / ( abs(CHARGE) * c * 1.e-9).
The order of precedence for arguments is: energy->pc->gamma->beta->brho
Note that if the restmass is changed after the energy has been set, ie in separate BEAM com-
mands, the energy is left unchanged and the momentum PC and relativistic factor GAMMA are
recalculated.
EX The horizontal emittance x(default: 1 m).
EY The vertical emittance y(default: 1 m).
ET The longitudinal emittance t(default: 1 m).
EXN The normalised horizontal emittance [m]: xn =pγ2−1x=βγ x
EYN The normalised vertical emittance [m]: yn =pγ2−1y=βγ y
SIGT The bunch length c σtin [m].
SIGE The relative energy spread σE/E in [1].
The order of precedence for arguments is: ex->exn,ey->eyn.
Note that up to version 5.02.04 the definition of normalised emittance used in MAD-X was
referring to the so-called 2-sigma geometric emittance: n= 4pγ2−1= 4βγ This
definition was different from the definition usually found in literature and used for example
in the APERTURE module.
The standard one sigma definition is now used across all MAD-X modules.
Certain commands compute the synchrotron tune Qstaking into account the settings of RF
cavities. If Qsis non-zero, the relative energy spread and the bunch length are calculated
![](asset-3b.png)
7.1. BEAM 57
with
σE/p0c=st
2πQs
η C (7.1)
c σt=st
η C
2πQs
(7.2)
where Cis the machine circumference, and
η= 1/γ2−1/γ2
t(7.3)
KBUNCH The number of particle bunches in the machine (default: 1).
NPART The number of particles per bunch (default: 0).
BCURRENT The bunch current (default: 0 A).
BUNCHED A logical flag. If set, the beam is treated as bunched whenever this makes
sense.
RADIATE A logical flag. If set, synchrotron radiation is considered in all dipole magnets.
BV an integer specifying the direction of the particle movement in a beam line;
either +1 (default), or -1. For a detailed explanation see the section below on
bv flag.
SEQUENCE attaches the defined beam to the named sequence; if the name is omitted, the
BEAM command refers to the default beam which is always present. Sequences
without attached beam use this default beam. When updating a beam with a
corresponding sequence name, tye sequence name must always be mentioned.
Order and Precedence:
Internally the BEAM command processes the parameters in the following order and with the
following precedence (left to right):
particle -> (mass+charge)
energy -> pc -> gamma -> beta -> brho
ex -> exn
ey -> eyn
current -> npart
Warning: BEAM updates, i.e. it replaces attributes explicitly mentioned, may calculate
other attributes according to the precedence rules given, but does NOT return attributes not
specified to default values! In order to reset to reset BEAM attributes to their default values,
use the RESBEAM command.
Additional variables:
Some MAD-X modules may also compute and store data into a beam data block. These at-
tributes may NOT be set directly through the BEAM command. The corresponding variables
are:
CIRC total length or circumference of the machine [m]
![](asset-3c.png)
58 CHAPTER 7. BEAM HANDLING STATEMENTS
FREQ0 revolution frequency [Hz]
DTBYDS ???
DELTAP momentum deviation
ALFA momentum compaction factor
U0 radiation loss per turn [GeV]
QS synchrotron tune [1]
ARAD classical particle radius [m]
PDAMP damping partition numbers; Default is 1,1,2
N1MIN minimum available aperture, set by the APERTURE module
7.2 RESBEAM
The RESBEAM command resets the default values of the beam belonging to the specified se-
quence, or of the default beam if no sequence is given.
RESBEAM, SEQUENCE=string;
The only argument to RESBEAM is a string for the sequence name. If the sequence name is
omitted, the default beam is reset.
Table 7.1: Default Beam Data
Attribute Value Unit
PARTICLE POSITRON
ENERGY 1 GeV
EX 1 rad.m
EY 1 rad.m
ET 1 rad.m
KBUNCH 1
NPART 0
BCURRENT 0 A
BUNCHED TRUE
RADIATE FALSE
7.3 Referring to BEAM data attributes
Expressions may refer to data in the beam data block using the notation
BEAM->attribute-name
or
![](asset-3d.png)
7.4. BV FLAG 59
BEAM%sequence-name->attribute-name.
This notation refers to the value of attribute-name found in the default BEAM, respectively
the beam belonging to the sequence sequence-name. This can be used for receiving or using
values, e.g.
value, beam%lhcb2->bv;
but also for storing values in the beam bank, e.g.
beam->charge=-1;
Note however that this does NOT trigger an update of dependent variables and you are
strongly advised against setting BEAM parameters with this method.
The current values in the BEAM bank can be obtained by the command
SHOW, BEAM;
or to obtain the data for a beam linked to a named sequence:
SHOW, BEAM%sequence-name;
Example:
! select electrons, set energy and emittances
BEAM, PARTICLE=ELECTRON, ENERGY=50, EX=1.E-6, EY=1.E-8, SIGE=1.E-3;
...
! turn on synchrotron radiation
BEAM, RADIATE;
...
! reset all values to their defaults,
! ie positrons, energy = 1GeV, default emittances, no radiation...
RESBEAM;
...
! select new emittances
BEAM, EX=2.E-5, EY=3.E-7, SIGE=4.E-3;
7.4 BV FLAG
When reversing the direction (~
V) of a particle in a magnetic field ( ~
B) while keeping its charge
constant, the resulting force ~
V×~
Bchanges sign. This is equivalent to flipping the field, but
not the direction.
For practical reasons the properties of all elements of the LHC are defined in the MAD-X input
as if they apply to a clockwise proton beam (”LHC beam 1”). This allows a single definition
for elements traversed by both beams. Their effects on a beam with identical particle charge
but running in the opposite direction (”LHC beam 2”) must then be reversed inside the
program.
60 CHAPTER 7. BEAM HANDLING STATEMENTS
In MAD-X this may be taken into account by setting the value of the BV attribute in the BEAM
commands. In the case of LHC beam 1 (clockwise) and beam 2 (counter-clockwise), that are
both treated in MAD-X as clockwise proton beams, the BEAM commands must look as follows:
BEAM, SEQUENCE=lhcb1, PARTICLE=proton, PC=450, BV = +1;
BEAM, SEQUENCE=lhcb2, PARTICLE=proton, PC=450, BV = -1;
![](asset-3f.png)
Chapter 8. Sequence Editor
With the help of a few commands, an existing sequence may be modified in many ways: in
the case of a circular machine, the starting point of the sequence can be moved to another
place; the order of elements can be inverted; elements can be inserted one by one, or as a
whole group with one single command; single elements, or classes thereof can be removed;
elements can be replaced by others; finally, the sequence can be ”flattened”, i.e. all inserted
sequences are replaced by their actual elements, such that a flattened sequence contains only
elements.
It is good practice to add a FLATTEN statement at the end of a SEQEDIT operation to ensure
a fully operational sequence. This is particularly useful for the SAVE command to properly
save shared sequences and to write them out in MAD-8 format.
8.1 SEQEDIT
MAD-X provides an environment for the edition of sequences that is invoked with the command:
SEQEDIT, SEQUENCE=string;
The only attribute is the name of the sequence to be edited.
The editing is performed on the sequence as provided by the user and before it is expanded
with the USE command. At the end of sequence edition, the resulting sequence must be
expanded through the USE command as necessary.
8.2 FLATTEN
Sequences can be built from elements but also sub-sequences resulting in a nested structure
(see chapter 13 on sequence definition). The positioning of elements within a sequence can
also be specified with values or expressions, and by reference to other elements.
MAD-X provides a command to resolve these dependencies and transform a complex sequence
into a simple list of elements with positioning values referring to the start of the sequence,
discarding the user-specified expressions for the positioning.
This command takes no argument:
FLATTEN;
If the sequence being edited contains sub-sequences, FLATTEN recursively includes all sub-
sequences until the sequence is only composed of a simple list of elements.
FLATTEN also resolves the positioning of each element within the sequence to a single value
with reference to the start of the sequence and updates the AT attribute of that element while
also discarding the user-specified expression if present.
61
![](asset-40.png)
62 CHAPTER 8. SEQUENCE EDITOR
The FLATTEN command is recommended at the beginning of sequence edition as well as at
the very end as in
SEQEDIT, SEQUENCE=name;
FLATTEN;
...commands to edit the named sequence...
FLATTEN;
ENDEDIT;
8.3 CYCLE
CYCLE, START=string;
This makes the sequence start at the location given as attribute value of the START attribute.
The element named by the START attribute must be a marker.
In case there is a shared sequence in the used sequence, the command FLATTEN should be be
used before the command CYCLE.
Example:
FLATTEN;
CYCLE, START=place;
Note that the FLATTEN command inserts another marker before the start location, with a
name composed of the name of the sequence being edited, followed by the start location
name and the string ” P ”.
8.4 REFLECT
REFLECT;
This inverts the order of element in the sequence, starting from the last element.
If there are shared sequences inside the USEd sequence, the command FLATTEN must be used
before the command REFLECT. Alternatively each shared sequence must first be reflected.
Example:
FLATTEN;
REFLECT;
8.5 INSTALL
INSTALL, ELEMENT=string, CLASS=string,
AT=real, FROM={string|SELECTED};
where the parameters have the following meaning:
ELEMENT name of the (new) element to be inserted (mandatory)
![](asset-41.png)
8.6. MOVE 63
CLASS class of the new element to be inserted (mandatory)
AT position where the element is to be inserted; if no ”from” is given,this is relative
to the start of the sequence. If ”from” is given, it is relative to the position
specified there.
FROM either a place (i.e. the name(+occurrence count) of an element already existing
in the sequence, e.g. mb[15], or mq.a..i1..4 etc.; or the string SELECTED; in the
latter case an element of the type specified will be inserted behind all elements
in the sequence that are currently selected by one or several SELECT commands
of the type
SELECT, FLAG=seqedit, CLASS=.., PATTERN=.., RANGE=..;
Note: No element definition can occur inside a SEQEDIT ... ENDEDIT block.
8.6 MOVE
MOVE, ELEMENT={string|SELECTED}, BY=real, TO=real, FROM=string;
ELEMENT name of the existing element to be moved, or ”SELECTED”, in which case
all elements from existing SELECT commands will be moved; in the latter case,
the BY attribute must be given.
BY distance by which the element(s) is/are to be moved; no TO or FROM attributes
should be given in this case.
TO position to which the element has to be moved; if no FROM attribute is given,
the position is relative to the start of the sequence; otherwise, it is relative to
the location given in the FROM argument
FROM place in the sequence with respect to which the element is to be positioned.
8.7 REMOVE
REMOVE, ELEMENT={string|SELECTED};
ELEMENT name of existing element(s) to be removed.
The special case ELEMENT = SELECTED removes all elements previously selected
by SELECT commands
Note: MAD-X expects to find some special markers in a beam line and it is therefore dangerous
to remove all markers from a sequence! In particular the START=... marker and markers
added by a CYCLE command must never be removed from a sequence.
![](asset-42.png)
64 CHAPTER 8. SEQUENCE EDITOR
8.8 REPLACE
REPLACE, ELEMENT={string|SELECTED}, BY=string;
The parameters are defined as:
ELEMENT names the elements to be replaced.
The special case ELEMENT = SELECTED replaces all elements previously selected
by SELECT commands
BY names the elements replacing the elements selected for replacement.
8.9 EXTRACT
A new sequence can be extracted as a subset of an existing sequence. The extracted sequence
is given a new name and can be USEd as any user defined sequence.
EXTRACT, SEQUENCE=string, FROM=string, TO=string,
NEWNAME=string;
The parameters are defined as:
SEQUENCE the name of the existing sequence from which the new sequence is extracted.
FROM the name of an element in the sequence that becomes the first element of the
extracted sequence.
TO the name of an element in the sequence that becomes the last element of the
extracted sequence.
NEWNAME the name of the extracted sequence.
The extracted sequence is declared as SHAREd and can therefore be combined e.g. into the
cycled original sequence.
Note: the element given by the FROM attribute must be located, in the existing sequence,
before the element given by the TO attribute, or MAD-X fails with a fatal error. In the case of
circular sequences, this can be ensured by performing a CYCLE of the original sequence with
START pointing to the same element given in the FROM attribute of the EXTRACT command.
8.10 ENDEDIT
The sequence editing environment is closed with the command
ENDEDIT;
The nodes in the sequence are finally renumbered according to their occurrence which might
have changed during editing.
![](asset-43.png)
8.11. SAVE 65
8.11 SAVE
The SAVE command saves a sequence to a specified file together with all relevant information.
SAVE, SEQUENCE=string,string, FILE=filename,
BEAM=logical, BARE=logical, MAD8=logical,
NOEXPR=logical, NEWNAME=string;
The parameters are defined as:
SEQUENCE lists the sequences to be saved, separated by commas. This attribute is optional
and when omitted, all known sequences are saved.
However, because of internal inconsistencies that can result in spurious entries
in the output file, the user is strongly advised to always provide explicitly the
names of sequences to be saved.
FILE the filename of the output file. (Default: ”save”)
BEAM an optional flag to specify that all beams linked to the specified sequences are
saved at the top of the output file.
BARE an optional flag to save only the sequence without the element definitions nor
beam information. This allows to re-read in a sequence with might otherwise
create a stop of the program. This is particularly useful to turn a line into a
sequence in order to further edit it with SEQEDIT.
MAD8 an optional flag to request that the sequences should be saved using MAD-8
input format.
NOEXPR an optional flag to save values of expressions instead of the expressions them-
selves: the expressions in commands and variables are expanded and evaluated
before saving. This option must be used with care because the exported values
were not deeply checked and the code that writes variables and commands is
widely spread in the internal structure.
This option does not apply only for the saving of sequences in MAD-8 format.
NEWNAME provides a name for the saved sequence, overriding the original name. (see
EXTRACT above)
Any number of SELECT, FLAG=save, ... commands may precede the SAVE command. In
that case, the names of elements, variables, and sequences must match the pattern(s) if given,
and in addition the elements must be of the class(es) specified.
The precision of the output of the SAVE command depends on the defined output precision
for MAD-X, which can be adjusted with the SET, FORMAT... command.
Note that with BARE=false the saved sequence may contain redundant efinitions of elements,
i.e. the same element is defined in the declaration of elements in the form label: type...
and in the sequence itself in the form label: type, at=.... This is flagged by MAD-X as
implicit element redefinition and is ignored but a warning is issued.
Example:
![](asset-44.png)
66 CHAPTER 8. SEQUENCE EDITOR
tl3: LINE = ( ldl6, qtl301, mqn, qtl301, ldl7, qtl302,
mqn, qtl302, ldl8, ison);
dltl3: LINE=(delay, tl3);
Use, period=dltl3;
Save, sequence=dltl3, file=t1, bare; // only sequence is saved
Call, file=t1; // sequence is read in and is now a "real" sequence
// if the two preceding lines are suppressed, seqedit will print a warning
// and else do nothing
Use, period=dltl3;
Twiss, save, betx=bxa, alfx=alfxa, bety=bya, alfy=alfya;
Plot, vaxis=betx, bety, haxis=s, colour:=100;
SEQEDIT, SEQUENCE=dltl3;
remove,element=cx.bhe0330;
remove,element=cd.bhe0330;
ENDEDIT;
Use, period=dltl3;
Twiss, save, betx=bxa, alfx=alfxa, bety=bya, alfy=alfya;
8.12 DUMPSEQU
DUMPSEQU, SEQUENCE=string, LEVEL=integer;
This command is actually more of a debug statement, but it may come handy at certain
occasions. The argument of the SEQUENCE attribute is the name of an already expanded (i.e.
USEd) sequence. The amount of detail in the output is controlled by the LEVEL argument:
= 0 : print only the cumulative node length = sequence length
>0 : print all node (element) names except drifts
>2 : print all nodes with their attached parameters
>3 : print all nodes, and their elements with all parameters
Part II
Elements, Beamlines and Sequences
67
![](asset-46.png)
Chapter 9. Definition of Elements
9.1 Element Input Format
All physical elements are defined by statements of the form
label: keyword {,attribute};
where
label is a name to be given to the element.
keyword is an element type keyword.
attribute normally – with exception for multipoles – takes one of the two forms:
attribute-name = attribute-value
attribute-name := attribute-value
attribute-name selects the attribute, as defined for the element type key-
word.
attribute-value provides a value to the attribute name. The value may
be specified by an expression.
The ”=” assigns the value on the right to the attribute at the time of definition,
regardless of any further change of the right hand side; the ”:=” re-evaluates
the expression at the right every time the attribute is being used. For constant
right hand sides, ”=” and ”:=” are of course equivalent.
Omitted attributes are assigned a default value.
Example:
QF: QUADRUPOLE, L=1.8, K1=0.015832;
A special format is used for a multipole:
M: MULTIPOLE, KN= kn0, kn1, kn2, ..., knmax,
KS= ks0, ks1, ks2, ..., ksmax;
where KN and KS give the integrated normal and skew strengths, respectively. The commas
are mandatory, each strength can be an expression; their position defines the order.
Example:
M: MULTIPOLE, KN=0,0,0,myoct*lrad, KS=0,0,0,0,-1.e-5;
defines a multipole with a normal octupole component and a skew decapole component.
9.2 Editing Element Definitions
An element definition can be changed in two ways:
68
9.3. ELEMENT CLASSES 69
•Entering a new definition: The element will be replaced in the main beam line
expansion.
•Entering the element name together with new attributes: The element will be
updated in place, and the new attribute values will replace the old ones.
This example shows two ways to change the strength of a quadrupole:
QF: QUADRUPOLE, L = 1, K1 = 0.01; ! Original definition of QF
QF: QUADRUPOLE, L = 2, K1 = 0.02; ! Replace whole definition of QF
QF, K1 = 0.03; ! Replace value of K1 for QF
When the type of the element remains the same, replacement of an attribute is the more
efficient way.
Element definitions can be edited freely. The changes do not affect already defined objects
which belong to its element class (see below).
9.3 Element Classes
The concept of element classes solves the problem of addressing instances of elements in the
accelerator in a convenient manner.
It will first be explained by an example. All the quadrupoles in the accelerator form a class
QUADRUPOLE. Let us define three subclasses for the focussing quadrupoles, the defocussing
quadrupoles, and the skewed quadrupoles:
MQF: QUADRUPOLE, L = LQM, K1 = KQD; ! Focussing quadrupoles
MQD: QUADRUPOLE, L = LQM, K1 = KQF; ! Defocussing quadrupoles
MQT: QUADRUPOLE, L = LQT; ! Skewed quadrupoles
These classes can be thought of as new keywords which define elements with specified default
attributes. We now use these classes to define the real quadrupoles:
QD1: MQD; ! Defocussing quadrupoles
QD2: MQD;
QD3: MQD;
...
QF1: MQF; ! Focussing quadrupoles
QF2: MQF;
QF3: MQF;
...
QT1: MQT, K1S = KQT1; ! Skewed quadrupoles
QT2: MQT, K1S = KQT2;
...
70 CHAPTER 9. DEFINITION OF ELEMENTS
These quadrupoles inherit from their class all attributes that are not explicitly specified at
time of declaration. This allows to build up a hierarchy of objects with a rather economic
input structure.
The full power of the class concept is revealed when object classes are used to select instances
of elements for various purposes. Example:
SELECT, FLAG=twiss, CLASS = QUADRUPOLE; ! Select all quadrupoles for the
! Twiss TFS file
More formally, for each element keyword MAD-X maintains a class of elements with the same
name. A defined element becomes itself a class which can be used to define new objects, which
will become members of this class. A new object inherits all attributes from its class; but its
definition may override some of those values by new ones. All class and object names can be
used in range selections, providing a powerful mechanism to specify positions for matching
constraints and printing.
![](asset-49.png)
Chapter 10. Element Types
10.1 Marker
label: MARKER;
The simplest element which can occur in a beam line is a MARKER. It has zero length and has
no effect on the beam, but it allows one to identify a position in the beam line, for example
to apply a matching constraint.
Example:
M27: MARKER;
10.2 Drift Space
label: DRIFT, L=real;
A drift space has one real attribute:
LThe drift length (Default: 0 m)
Example:
DR1: DRIFT, L=1.5;
DR2: DRIFT, L=DR1->L;
The length of DR2 is always equal to the length of DR1.
The straight reference system for a drift space is a Cartesian coordinate system.
10.3 Bending Magnet
Two different type keywords are recognised for bending magnets, they are distinguished only
by the reference system used:
SBEND is a sector bending magnet.
The planes of the pole faces intersect at the centre of curvature of the curved
sbend reference system.
RBEND is a rectangular bending magnet.
The pole faces are parallel. The reference system is the curved rbend reference
system.
Bendig magnets are defined by the statements:
71
![](asset-4a.png)
72 CHAPTER 10. ELEMENT TYPES
label: SBEND, L=real, ANGLE=real, TILT=real,
K0=real, K1=real, K2=real,
E1=real, E2=real, FINT=real, FINTX=real,
HGAP=real, H1=real, H2=real,
THICK=logical;
label: RBEND, L=real, ANGLE=real, TILT=real,
K0=real, K1=real, K2=real,
E1=real, E2=real, FINT=real, FINTX=real,
HGAP=real, H1=real, H2=real,
THICK=logical,
ADD ANGLE=’array’;
Bending magnets have the following attributes:
LThe length of the magnet (default: 0 m).
For sector bends the declared length is the arc length of the reference orbit.
For rectangular bends the declared length is normally the length of a straight
line joining the entry and exit points, as in the Figure.
Internally MAD-X only uses the arc length of the reference orbit for both bend
types.
In order to define RBEND’s with a declared length equal to the arc
length of the reference orbit, the option RBARC must be previously
set to FALSE in MAD-X with Option, RBARC=false;
ANGLE The bend angle (default: 0 rad).
A positive bend angle represents a bend to the right, i.e. towards negative x
values.
ADD ANGLE An array of (maximum 5) bending angles for multiple passes. See ADD PASS
option of the SEQUENCE command. This is only allowed for RBEND elements
and is ignored for SBEND elements.
TILT The roll angle about the longitudinal axis (default: 0 rad, i.e. a horizontal
bend).
A positive angle represents a clockwise rotation.
An attribute TILT=pi/2 turns a horizontal into a vertical bend, and a positive
ANGLE then denotes a downwards deflection.
K1 The quadrupole coefficient (Default: 0 m−2)
K1= (1/Bρ)(∂By/∂x).
A positive quadrupole strength implies horizontal focussing of particles, irre-
spective of their charge.
E1 The rotation angle for the entrance pole face (Default: 0 rad).
E2 The rotation angle for the exit pole face (Default: 0 rad).
The pole face rotation angles are referred to the magnet model for rectangular
bend and sector bend respectively.
![](asset-4b.png)
10.3. BENDING MAGNET 73
E1 and E2 are positive if they reduce the length of the side of the bend that is
further away from the centre of curvature.
E1 and E2 must be specified as positive to give an SBEND parallel faces, i.e.
turning it into an RBEND.E1 and E2 must be negative to give an RBEND faces
whose planes intersect at the centrer of curvature, ie turning it into an SBEND.
FINT The fringe field integral at entrance and exit of the bend. (Default: 0).
FINTX If defined and positive, the fringe field integral at the exit of the element,
overriding FINT for the exit. (Default: =FINT)
This allows to set different fringe field integrals at entrance (FINT) and exit
(FINTX) of the element.
HGAP The half gap of the magnet (default: 0 m).
K2 The sextupole coefficient. (Default: 0 m−3)
K2= (1/Bρ)(∂2By/∂x2).
H1 The curvature of the entrance pole face. (Default: 0 m−1).
H2 The curvature of the exit pole face. (Default: 0 m−1)
A positive pole face curvature induces a negative sextupole component; i.e. for
positive H1 and H2 the centres of curvature of the pole faces are placed inside
the magnet.
K0 OBSOLETE
Please take note that K0and K0sare left in the data base but are no longer
used for the MAP of the bends, instead ANGLE and TILT are used exclusively.
However, specifying K0 is required in order to be able to assign relative field
errors to a bending magnet because K0 is used for the normalization instead
of the ANGLE. (see EFCOMP).
With K0= (1/Bρ)By, one gets K0 = ANGLE / arclength.
THICK If this logical flag is set to true the bending magnet is tracked through as a
thick-element, instead of being converted into thin-lenses.
(Default: false)
Note: Additional attributes can be given to bending magnets; They are useful for PTC and
are defined in 30.8.
Fringe Fields:
The quantities FINT and HGAP specify the finite extent of the fringe fields as defined in SLAC-
75 [1]:
FINT = Z∞
−∞
By(s)(B0−By(s))
g·B2
0
ds, g = 2 ·HGAP.(10.1)
The default values of zero corresponds to the hard-edge approximation, i.e. a rectangular
field distribution. For other approximations, enter the correct value of the half gap, and one
of the following values for FINT:
![](asset-4c.png)
74 CHAPTER 10. ELEMENT TYPES
Linear Field drop-off 1/6
Clamped ”Rogowski” fringing field 0.4
Unclamped ”Rogowski” fringing field 0.7
”Square-edged” non-saturating magnet 0.45
Entering the keyword FINT alone sets the integral to 0.5, which is a reasonable average of the
above values.
Note also that the possibility to specify both FINT and FINTX allows one to set different values
at entrance and exit of a bend element.
This can be particularly useful to set the fringe field integral to zero on one side only, e.g.
when slicing a dipole.
Examples:
BR: RBEND, L=5., ANGLE=+0.001; ! Deflection to the right
BD: RBEND, L=5., ANGLE=+0.001, TILT= pi/2; ! Deflection down
BL: RBEND, L=5., ANGLE=+0.001, TILT= pi; ! Deflection to the left
BU: RBEND, L=5., ANGLE=+0.001, TILT=-pi/2; ! Deflection up
10.4 Dipole edge
A thin element describing the edge focusing of a dipole has been introduced in order to make
it possible to track trajectories in the presence of dipoles with pole face angles. Only linear
terms are considered since the higher order terms would make the tracking non-symplectic.
The transformation of the machine elements into thin lenses leaves dipole edge (DIPEDGE)
elements untouched and splits correctly the SBEND’s.
It does not make sense to use a DIPEDGE alone. It can be specified at the entrance and the
exit of a SBEND. A dipole edge element is defined by the command:
label: DIPEDGE, H=real, E1=real, FINT=real,
HGAP=real, TILT=real;
ADIPEDGE has zero length and five attributes.
HIs angle/length or 1/rho (default: 0 m−1- for the default the dipedge element
has no effect). (must be equal to that of the associated SBEND)
E1 The rotation angle for the pole face. The sign convention is as for a SBEND
bending magnet. Note that it is different for an entrance and an exit. (default:
0 rad).
FINT field integral as for SBEND sector bend. Note that each DIPEDGE has its own
FINT, so that specifying FINTX is no longer necessary.
HGAP half gap height of the associated SBEND bending magnet.
TILT The roll angle about the longitudinal axis (default: 0 rad, i.e. a horizontal
bend). A positive angle represents a clockwise rotation.
![](asset-4d.png)
10.5. QUADRUPOLE 75
10.5 Quadrupole
label: QUADRUPOLE, L=real, K1=real, K1S=real, TILT=real,
THICK=logical;
AQUADRUPOLE has five attributes:
LThe quadrupole length (default: 0 m).
K1 The normal quadrupole coefficient: K1= 1/(Bρ)(∂By/∂x).
The default is 0 m−2. A positive normal quadrupole strength implies horizontal
focussing, irrespective of the charge of the particles.
K1S The skew quadrupole coefficient K1s= 1/(2Bρ)(∂Bx/∂x −∂By/∂y)
where (x,y) is now a coordinate system rotated by −45◦around s with respect
to the normal one. The default is 0 m−2. A positive skew quadrupole strength
implies defocussing, irrespective of the charge of the particles, in the (x,s) plane
rotated by 45◦around s(particles in this plane have x=y>0).
TILT The roll angle about the longitudinal axis (default: 0 rad, i.e. a normal
quadrupole). A positive angle represents a clockwise rotation. A TILT=pi/4
turns a positive normal quadrupole into a negative skew quadrupole.
Please note that contrary to MAD-8 one has to specify the desired
TILT angle, otherwise it is taken as 0 rad. This was needed to avoid
the confusion in MAD-8 about the actual meaning of the TILT attribute
for various elements.
THICK If this logical flag is set to true the quadrupole is tracked through as a thick-
element, instead of being converted into thin-lenses.
(Default: false)
Note also that K1or K1scan be considered as the normal or skew quadrupole
components of the magnet on the bench, while the TILT attribute can be consid-
ered as a tilt alignment error in the machine. In fact, a positive K1with a TILT
= 0 is equivalent to a positive K1swith TILT = +π/4
Example:
QF: QUADRUPOLE, L=1.5, K1=0.001, THICK=true;
Note: Additional attributes can be given to quadrupoles; They are useful for PTC and are
defined in 30.8.
The straight reference system for a quadrupole is a Cartesian coordinate system.
10.6 Sextupole
label: SEXTUPOLE, L=real, K2=real, K2S=real, TILT=real;
ASEXTUPOLE has four real attributes:
![](asset-4e.png)
76 CHAPTER 10. ELEMENT TYPES
LThe sextupole length (default: 0 m).
K2 The normal sextupole coefficient K2=1
Bρ (∂2By/∂x2).
(default: 0 m−3).
K2S The skew sextupole coefficient K2S=1
Bρ (∂2Bx/∂x2)
where (x,y) is now a coordinate system rotated by −30◦around s with respect
to the normal one. (default: 0 m−3). A positive skew sextupole strength
implies defocussing (!) irrespective of the charge of the particles, in the (x,s)
plane rotated by 30◦around s (particles in this plane have x > 0, y > 0).
TILT The roll angle about the longitudinal axis (default: 0 rad, i.e. a normal sex-
tupole). A positive angle represents a clockwise rotation. A TILT = pi/6
turns a positive normal sextupole into a negative skew sextupole.
Please note that contrary to MAD-8 one has to specify the desired TILT
angle, otherwise it is taken as 0 rad. This was needed to avoid the
confusion in MAD-8 about the actual meaning of the TILT attribute
for various elements.
Note also that K2or K2scan be considered as the normal or skew sextupole com-
ponents of the magnet on the bench, while the TILT attribute can be considered
as an tilt alignment error in the machine. In fact, a positive K2with a TILT = 0
is equivalent to a positive K2swith positive TILT =π/6.
Example:
S: SEXTUPOLE, L=0.4, K2=0.00134;
Note: Additional attributes can be given to sextupoles; They are useful for PTC and are
defined in 30.8.
The straight reference system for a sextupole is a Cartesian coordinate system.
10.7 Octupole
label: OCTUPOLE, L=real, K3=real, K3S=real, TILT=real;
An OCTUPOLE has four real attributes:
LThe octupole length (default: 0 m).
K3 The normal octupole coefficient K3=1
Bρ (∂3By/∂x3)
(default: 0 m−4).
K3S The skew octupole coefficient K3S=1
2Bρ (∂3Bx/∂x3−∂3By/∂y3)
where (x,y) is now a coordinate system rotated by -22.5oaround s with respect
to the normal one. (default: 0 m−4). A positive skew octupole strength implies
defocussing (!) irrespective of the charge of the particles, in the (x,s) plane
rotated by 22.5oaround s (particles in this plane have x >0, y >0).
![](asset-4f.png)
10.8. GENERAL THIN MULTIPOLE 77
TILT The roll angle about the longitudinal axis (default: 0 rad, i.e. a normal oc-
tupole). A positive angle represents a clockwise rotation. A TILT=pi/8 turns
a positive normal octupole into a negative skew octupole.
Please note that contrary to MAD-8 one has to specify the desired
TILT angle, otherwise it is taken as 0 rad. This was needed to
avoid the confusion in MAD-8 about the actual meaning of the TILT
attribute for various elements.
Note also that K3or K3Scan be considered as the normal or skew quadrupole
components of the magnet on the bench, while the TILT attribute can be con-
sidered as an tilt alignment error in the machine. In fact, a positive K3with a
TILT=0 is equivalent to a positive K3Swith positive TILT=+pi/8.
Example:
O3: OCTUPOLE, L=0.3, K3=0.543;
Note: Additional attributes can be given to octupoles; They are useful for PTC and are
defined in 30.8.
The straight reference system for a octupole is a Cartesian coordinate system. Octupoles are
normally treated as thin lenses, except when tracking by Lie-algebraic methods.
10.8 General Thin Multipole
AMULTIPOLE is a thin-lens magnet of arbitrary order, including a dipole component.
label: MULTIPOLE, LRAD=real, TILT=real,
KNL={real, ...}, KSL={real, ...};
LRAD A fictitious length, originally only used to compute synchrotron radiation ef-
fects.
A non-zero LRAD in conjunction with OPTION, THIN FOC=true takes into ac-
count the weak focussing of bending magnets.
TILT The roll angle about the longitudinal axis (default: 0 rad). A positive angle
represents a clockwise rotation of the multipole element. The roll angle affects
all components.
KNL An array of the integrated normal multipole coefficients, starting from order
zero and up to the maximum order (currently 20). The parameters are posi-
tional in the array, therefore zeros must be filled in for components that do
not exist.
The coefficient of rank iin the array corresponds to the integrated strength
KiL=Bi.L/(Bρ) where the strength is given by equation 1.11. Hence the
first argument of the array, the argument for a normal multipole of order zero,
K0L=B0.L/(Bρ) is equal to the normal horizontal rotation angle of the thin
dipole.
![](asset-50.png)
78 CHAPTER 10. ELEMENT TYPES
KSL An array of the integrated skew multipole coefficients, starting from order zero
and up to the maximum order (currently 20). The parameters are positional
in the array, therefore zeros must be filled in for components that do not exist.
Hence the first argument of the array, the argument for a skew multipole of
order zero, K0L=B0.L/(Bρ) is equal to the skew or vertical rotation angle
of the thin dipole.
Both KNL and KSL may be specified for the same multipole.
Contrary to MAD-8 the desired TILT angle must be explicitly specified, and defaults otherwise
to 0 rad. The roll angle specified with TILT is global to all multipolar components. Hence
the KNL and KSL components can be considered as the normal or skew multipole components
of the magnet as measured on the bench, while the TILT attribute can be considered as an
alignment error as measured in the machine.
A multipole with no dipole component has no effect on the reference orbit, i.e. the reference
system at its exit is the same as at its entrance. If it includes a dipole component, it has
the same effect on the reference orbit as a dipole with zero length, total deflection angle and
tilt defined by:
angle =pKNL(0)2+KSL(0)2
tilt = arctan(KSL(0)/KNL(0)) (10.2)
Note that the global TILT attribute of the MULTIPOLE is then added to the intrinsic tilt
calculated from KNL and KSL.
Examples:
A thin-lens sextupole:
ms: MULTIPOLE, KNL={0, 0, k2l};
A thin-lens skew octupole:
mso: MULTIPOLE, KSL={0, 0, 0, k3sl};
A thin-lens multipole with a normal octupole component and a skew decapole component:
mod: MULTIPOLE, KNL={0,0,0,myoct*lrad}, KSL={0,0,0,0,-1.e-5};
A thin-lens dipole bending to the right and down for a total angle of 2 milliradians and a tilt
of π/4 can be equivalently defined as:
hvbend: MULTIPOLE, KNL={1.414e-3}, KSL={1.414e-3};
hvbend: MULTIPOLE, KNL={2.e-3}, TILT= pi/4;
hvbend: MULTIPOLE, KSL={2.e-3}, TILT=-pi/4;
10.9 Solenoid
Solenoids can be defined in two forms, a thick and a thin version:
label: SOLENOID, L=real, KS=real; ! thick version
label: SOLENOID, L=0, KS=real, KSI=real; ! thin version
![](asset-51.png)
10.10. NONLINEAR LENS WITH ELLIPTIC POTENTIAL 79
ASOLENOID has three real attributes:
LThe length of the solenoid (default: 0 m)
KS The solenoid strength Ks=B0/Bρ (default: 0 rad/m). For positive KS and
positive particle charge, the solenoid field points in the direction of increasing
s.
KSI The solenoid integrated strength KsL(default: 0 rad). This additional at-
tribute is needed only when using the thin solenoid, where L= 0.
Example:
SOLO: SOLENOID, L = 2., KS = 0.001;
THINSOLO: SOLENOID, L = 0, KS = 0.001, KSI = 0.002;
Note: Additional attributes can be given to solenoids. They are useful for PTC and are defined
in 30.8. In particular multipole coefficients KNL and KSL can also be specified for solenoids.
They have no effect in MAD-X proper but are used in PTC for solenoid with multipoles.
The straight reference system for a solenoid is a Cartesian coordinate system.
10.10 Nonlinear Lens with Elliptic Potential
label: NLLENS, KNLL=real, CNLL=real;
The NLLENS element represents a thin nonlinear lens with the potential of ’Elliptic’ type as
specified in [7]. The lens is used to create fully integrable 2D nonlinear accelerator lattice
with very large nonlinear tune spread/shift. The NLLENS element is recognized by the thin
tracking module. The quadrupole term of the potential is included in the transport map and,
consequently, affects the calculation of tunes and Twiss functions.
KNLL The integrated strength of lens (m). The strength is parametrized so that the
quadrupole term of the multipole expansion is k1=2*KNLL/CNLL^2.
CNLL The dimensional parameter of lens (m). The singularities of the potential are
located at X=-CNLL,+CNLL and Y=0.
The scalar potential function of the element is given by
U(x, y) = k
c
ξpξ2−1 acoshξ+ηp1−η2(acosη−π/2)
ξ2−η2(10.3)
where k= KNLL, c= CNLL and
ξ=p(x+c)2+y2+p(x−c)2+y2
2c, η =p(x+c)2+y2−p(x−c)2+y2
2c,(10.4)
Figure below shows the contour plot of the scalar potential:
![](asset-52.png)
80 CHAPTER 10. ELEMENT TYPES
Figure 10.1: Contour plot of the scalar potential
The multipole expansion of the scalar potential is
U(x, y) = k·Re (x+iy
c2
+2
3x+iy
c4
+8
15 x+iy
c6
+16
35 x+iy
c8
+···)
(10.5)
Note that this expansion is only valid inside the r=c circle on the x,y plane.
In order to create integrable optics, one needs to shape the potential along z axis according to
the beta-function. Below is an example nonlinear section representing the necessary nonlinear
field with 20 thin lenses:
mu0 = 0.3; ! phase advance over straight section
l0 = 2.0; ! length of the straight section
nn = 20; ! number of nonlinear elements
tn = 0.45; ! strength of nonlinear lens
cn = 0.01; ! dimentional parameter of nonlinear lens
musect = mu0 + 0.5;
f0 = l0/4.0*(1.0+1.0/tan(pi*mu0)^2);
betae = l0/sqrt(1.0-(1.0-l0/2.0/f0)^2);
alfae = l0/2.0/f0/sqrt(1.0-(1.0-l0/2.0/f0)^2);
betas = l0*(1-l0/4.0/f0)/sqrt(1.0-(1.0-l0/2.0/f0)^2);
value, f0, betae, alfae, betas;
ncreate(ii,kk,cc): macro = {n.ii: nllens, knll=kk, cnll=cc;};
i=0;
while(i < nn)
![](asset-53.png)
10.11. CLOSED ORBIT CORRECTOR 81
{
i = i+1;
sn = l0/nn*(i-0.5);
bn = l0*(1-sn*(l0-sn)/l0/f0)/sqrt(1.0-(1.0-l0/2.0/f0)^2);
knn = l0*tn*cn^2/nn/bn;
cnn = cn*sqrt(bn);
exec, ncreate($i,knn,cnn);
value, i, bn, cnn, knn;
};
10.11 Closed Orbit Corrector
Three types of magnetic closed orbit correctors are available:
HKICKER a corrector for the horizontal plane,
VKICKER a corrector for the vertical plane,
KICKER a corrector for both planes.
label: HKICKER,L=real, KICK=real, TILT=real;
label: VKICKER,L=real, KICK=real, TILT=real;
label: KICKER, L=real, HKICK=real, VKICK=real, TILT=real;
The type KICKER should not be used when an orbit corrector kicks only in one
plane.
The attributes have the following meaning:
LThe length of the closed orbit corrector (default: 0 m).
KICK The momentum change δP X =δpx/p0or δP Y =δpy/p0for respectively
horizontal or vertical correctors. (default: 0).
HKICK The horizontal momentum change δP X =δpx/p0for a corrector acting in
both planes (default: 0).
VKICK The vertical momentum change δP Y =δpy/p0for a corrector acting in both
planes (default: 0).
TILT The roll angle about the longitudinal axis (default: 0 rad). A positive angle
represents a clockwise rotation of the kicker.
A positive kick increases pxor pyrespectively. This means that a positive horizontal kick
bends to the left, i.e. to positive xwhich is opposite of what is true for bends.
The deviation angle θof the particle trajectory is related to the momentum change through
sin θ=δP =δp/p0.
It should be noted that the kick values assigned to an orbit corrector like above are not
overwritten by an orbit correction using the CORRECT command. Instead the kicks computed
by an orbit correction and the assigned values are added when the correctors are used.
82 CHAPTER 10. ELEMENT TYPES
Examples:
HK1: HKICKER, KICK = 0.001;
VK3: VKICKER, KICK = 0.0005;
VK4: VKICKER, KICK := AVK4;
KHV1: KICKER, HKICK = 0.001, VKICK = 0.0005;
KHV2: KICKER, HKICK := AKHV2H, VKICK := AKHV2V;
The assignment in the form of a deferred expression has the advantage that the values can
be assigned and/or modified at any time (and matched!).
The straight reference system for an orbit corrector is a Cartesian coordinate system.
Please note that there is a new feature introduced by Stefan Sorge from GSI. Here his decrip-
tion:
The elements KICKER, HKICKER, and VKICKER can also be used as magnetic exciters providing
sinusoidal momentum kicks. The usage in this case is:
xykick: KICKER, SINKICK=integer, SINPEAK=real, SINTUNE=real, SINPHASE=real;
xkick : HKICKER,SINKICK=integer, SINPEAK=real, SINTUNE=real, SINPHASE=real;
ykick : VKICKER,SINKICK=integer, SINPEAK=real, SINTUNE=real, SINPHASE=real;
where a sinusoidal momentum kick dpz as a function of the revolution number ngiven by
dpz(n)=SINPEAK * sin(2*PI*SINTUNE*n + SINPHASE), pz=px,py
is provided.
The KICKER element generates synchronous kicks in both horizontal and vertical planes.
HKICKER generates only a horizontal kick, and VKICKER generates only a vertical kick.
The variables are
SINKICK must be set to 1 to switch on the sinusoidal signal, default: 0.
SINPEAK amplitude of the bending angle (rad); default: 0 rad.
SINTUNE frequency of the signal times the revolution frequency. Hence, the phase per
revolution is 2*PI*SINTUNE; default: 0.
SINPHASE initial phase; default: 0 rad.
The momentum kick of a kicker has only a single frequency. An element having a finite
bandwidth can approximately created by defining thin kickers with all amplitudes SINPEAK,
frequencies SINTUNE, and initial phases SINPHASE desired and putting them at the same
position sin the accelerator.
10.12 Transverse Kicker
The type TKICKER should be used to create horizontal, vertical or combined transverse mag-
netic kickers physically equivalent to elements of type KICKER, but not used by the COR-
RECT command of the closed orbit correction module.
Examples of elements that may use the type TKICKER:
![](asset-55.png)
10.13. RF CAVITY 83
•Fast kickers for injection, dump and tune
•Magnetic septa towards beam dump
•Dampers of transverse beam oscillations
•Undulator and Wiggler magnets
For further information on element type TKICKER and its attributes, look at the documentation
of the orbit corrector type KICKER.
10.13 RF Cavity
label: RFCAVITY, L=real, VOLT=real, LAG=real,
FREQ=real, HARMON=integer,
N BESSEL=integer, NO CAVITY TOTALPATH=logical;
An RFCAVITY has eight real attributes and one integer attribute:
LThe length of the cavity (DEFAULT: 0 m)
VOLT The peak electrical RF voltage (DEFAULT: 0 MV). The effect of the cavity is
delta(E) = VOLT * sin(2π* (LAG - HARMON * f0t)).
LAG The phase lag [2π] (DEFAULT: 0).
FREQ The frequency [MHz] (no DEFAULT).
Note that if the RF frequency is not given, it is computed from the
harmonic number and the revolution frequency f0. For accelerating
structures this makes no sense, and the input of the frequency is
mandatory.
HARMON The harmonic number h(no DEFAULT). This attribute is only used if the
frequency is not given.
Caveats:
•Please take note, that the following MAD-8 attributes: BETRF, PG, SHUNT and TFILL
are currently not implemented in MAD-X.
•Important Note: The TWISS command is 4D only. As a consequence the TWISS parame-
ters in the plane of non-zero dispersion may not close as expected. Therefore, it is best
to perform TWISS in 4D only, i.e. with cavities switched off. If 6D is needed one has to
use the PTC TWISS command.
The RFCAVITY can also have attributes that only become active in PTC:
N BESSEL (DEFAULT: 0):
Transverse focussing effects are typically ignored in the cavity in MAD-X or
even PTC. This effect is being calculated to order n bessel, with n bessel=0
disregarding this effect and with a correct treatment when n bessel goes to
infinity.
![](asset-56.png)
84 CHAPTER 10. ELEMENT TYPES
NO CAVITY TOTALPATH (Default: false):
flag to select whether the transit time factor in the cavity is to be considered
(NO CAVITY TOTALPATH = false) or if the particle is kept on the crest of RF
voltage (NO CAVITY TOTALPATH = true).
A cavity requires the particle ENERGY and the particle CHARGE to be set by a BEAM command
before any calculations are performed.
Example:
BEAM, PARTICLE=ELECTRON, ENERGY=50.0;
CAVITY: RFCAVITY, L=10.0, VOLT=150.0, LAG=0.0, HARMON=31320;
The straight reference system for a cavity is a Cartesian coordinate system.
10.14 Thin Radio-Frequency Multipole
label: RFMULTIPOLE, VOLT=real, LAG=real,
FREQ=real, HARMON=integer,
LRAD=real, TILT=real,
KNL={real, ...}, KSL={real, ...},
PNL={real, ...}, PSL={real, ...};
ARFMULTIPOLE is a thin-lens element which exhibits the properties of an RF-cavity and of a
magnet of arbitrary order oscillating at a certain frequency.
The effect of the cavity is delta(E) = VOLT * sin(2π* (LAG - HARMON * f0 t)).
VOLT The peak RF voltage (DEFAULT: 0 MV).
LAG The phase lag [2π] (DEFAULT: 0)
FREQ The frequency [MHz] (no DEFAULT).
Note that if the RF frequency is not given, it is computed from the harmonic
number and the revolution frequency f0 as before. However, for accelerating
structures this makes no sense, and the frequency is mandatory.
HARMON The harmonic number h (no DEFAULT). Only if the frequency is not given.
LRAD A fictitious length, which was originally just used to compute synchrotron
radiation effects. A non-zero LRAD in conjunction with OPTION, thin foc =
true; takes into account the weak focussing of bending magnets.
TILT The roll angle about the longitudinal axis (default: 0 rad). A positive angle
represents a clockwise rotation of the multipole element.
Please note that contrary to MAD-8 one has to specify the desired
TILT angle, otherwise it is taken as 0 rad. We believe that the MAD-8
concept of having individual TILT values for each component and on
top with default values led to considerable confusion and allowed
for excessive and unphysical freedom. Instead, in MAD-X the KNL,
![](asset-57.png)
10.15. CRAB CAVITY 85
KSL components can be considered as the normal or skew multipole
components of the magnet on the bench, while the TILT attribute
can be considered as an tilt alignment error in the machine.
KNL An array of the integrated normal rfmultipole coefficients, starting from order
zero and up to the maximum order (currently 20). The parameters are po-
sitional in the array, therefore leading zeros must be filled in for components
that do not exist.
KSL An array of the integrated skew rfmultipole coefficients, starting from order
zero and up to the maximum order (currently 20). The parameters are po-
sitional in the array, therefore leading zeros must be filled in for components
that do not exist.
PNL The phase for each normal rfmultipole coefficients from order zero to the max-
imum; the parameters are positional, therefore leading zeros must be filled in
for components that do not exist.
PSL The phase for each skew rfmultipole coefficients from order zero to the maxi-
mum; the parameters are positional, therefore leading zeros must be filled in
for components that do not exist.
Example:
MS: RFMULTIPOLE, KSL={0, 0, 0, k3sl};
Both KNL and KSL may be specified for the same multipole.
ARFMULTIPOLE requires the particle ENERGY and the particle CHARGE to be set with a BEAM
command before any calculation is performed.
ARFMULTIPOLE with no dipole component has no effect on the reference orbit, i.e. the
reference system at the exit is the same as at the entrance. If the RFMULTIPOLE includes a
dipole component, it has the same effect on the reference orbit as a thin MULTIPOLE with
equivalent parameters.
10.15 Crab Cavity
label: CRABCAVITY, L=real, VOLT=real, LAG=real,
FREQ=real, HARMON=integer,
RV1=integer, RV2=integer,
RV3=integer, RV4=integer,
RPH1=integer, RPH2=integer, LAGF=real ;
ACRABCAVITY has five attributes describing its steady state and seven attributes to describe
dynamic behaviour:
LThe length of the cavity (default: 0 m)
VOLT The peak RF voltage (default: 0 MV).
![](asset-58.png)
86 CHAPTER 10. ELEMENT TYPES
LAG The initial phase lag [2π] (default: 0).
FREQ The RF frequency [MHz] (no default).
Note that if the RF frequency is not given, it is computed from the
harmonic number and the revolution frequency f0. For deflecting
structures this makes no sense, and the frequency is mandatory.
HARMON The harmonic number h(no default).
Only if the frequency is not given.
The other attributes describe the time evolution of a CRABCAVITY behaviour:
RV1 Number of initial turns with zero voltage (default: 0).
RV2 Number of turns to ramp voltage from zero to nominal value (default: 0).
RV3 Number of turns with nominal voltage (default: 0).
RV4 Number of turns to ramp voltage from nominal value to zero (default: 0).
LAGF Value of the final crab RF phase lag [2π] (default: 0).
RPH1 Number of initial turns with nominal phase (default: 0).
RPH2 Number of turns to ramp phase [2π] from nominal to specified value
(default: 0).
Caveats:
•Please take note, that the following MAD-8 attributes: BETRF, PG, SHUNT and TFILL
are currently not implemented in MAD-X!
•Note that crab cavities are only implemented for tracking purposes.
TWISS ignores any effect of the crab cavity.
•Important Note: The TWISS command is 4D only. As a consequence the TWISS parame-
ters in the plane of non-zero dispersion may not close as expected. Therefore, it is best
to perform TWISS in 4D only, i.e. with cavities switched off. If 6D is needed one has to
use the PTC TWISS command.
Before any calculation is performed with a CRABCAVITY, the particle ENERGY and the particle
CHARGE must be set with the BEAM command.
The effect of a CRABCAVITY on particle coordinates during tracking is
δpx=VOLT ∗sin(PHI −OMEGA ∗t)
δE=−VOLT ∗OMEGA ∗x∗cos(PHI −OMEGA ∗t)
where PHI =2π∗(LAG −HARMON ∗f0t),
and OMEGA =2π∗FREQ/c
Example:
![](asset-59.png)
10.16. ELECTROSTATIC SEPARATOR 87
BEAM, PARTICLE=PROTON, ENERGY=7000.0;
CAVITY: CRABCAVITY, L=10.0, VOLT=5.0, LAG=0.0, FREQ=400,
RV1=0, RV2=50, RV3=1000, RV4=50,
RPH1=100, RPH2=500, LAGF=0.125;
The straight reference system for a cavity is a Cartesian coordinate system.
10.16 Electrostatic Separator
label: ELSEPARATOR, L=real, EX=real, EY=real, TILT=real;
An ELSEPARATOR element has four real attributes:
LThe length of the separator (default: 0 m).
EX The horizontal electric field strength (default: 0 MV/m). A positive field
increases pxfor positive particles.
EY The vertical electric field strength (default: 0 MV/m). A positive field in-
creases pyfor positive particles.
TILT The roll angle about the longitudinal axis (default: 0 rad). A positive angle
represents a clockwise of the electrostatic separator.
An electrostatic separator requires the particle ENERGY and the particle CHARGE to be set by
aBEAM command before any calculation is performed.
Example:
BEAM, PARTICLE=positron, ENERGY=50.0;
SEP: ELSEPARATOR, L=5.0, EY=0.5;
The straight reference system for a separator is a Cartesian coordinate system.
10.17 Beam Position Monitor
A beam monitor has no specific effect on the beam and behaves like a drift space. In addition
it serves to record the beam position for closed orbit correction.
Three different types of beam position monitors are recognised:
HMONITOR Monitor for the horizontal beam position,
VMONITOR Monitor for the vertical beam position,
MONITOR Monitor for both horizontal and vertical beam position.
label: HMONITOR, L=real;
label: VMONITOR, L=real;
label: MONITOR, L=real;
![](asset-5a.png)
88 CHAPTER 10. ELEMENT TYPES
A beam position monitor has one real attribute:
LThe length of the monitor (default: 0 m).
Examples:
MH: HMONITOR, L = 1;
MV: VMONITOR;
The straight reference system for a monitor is a Cartesian coordinate system.
10.18 Instrument and Placeholder
An instrument or a placeholder has no specific effect on the beam and behaves like a drift
space. An instrument is different from beam a position monitor and is not used for closed
orbit correction.
Two different types of instruments are recognised:
INSTRUMENT A place holder for any type of beam instrumentation. Optically it behaves like
a drift space; it returns no beam observation. It represent a class of elements
which is completely independent from drifts and monitors.
PLACEHOLDER A place holder for any type of element. Internally it is equivalent to an
INSTRUMENT: optically it behaves as a drift space, it returns no beam obser-
vation. It represent a class of elements which is completely independent from
drifts and monitors.
label: INSTRUMENT, L=real;
label: PLACEHOLDER, L=real;
An instrument or placeholder has one real attribute:
LThe length of the instrument (default: 0 m).
The straight reference system for an instrument is a Cartesian coordinate system.
10.19 Collimator
ACOLLIMATOR has no specific effect on beam optics and behaves like a drift space.
label: COLLIMATOR, L=real,
APERTYPE=string, APERTURE={values},
APER OFFSET={values}, APER TOL={values};
ACOLLIMATOR has one specific real attribute:
LThe collimator length (default: 0 m).
![](asset-5b.png)
10.20. BEAM-BEAM INTERACTION 89
Additionally, like any other element, except DRIFT space, a COLLIMATOR can have specific
aperture related attributes as defined in the related section Defining aperture in MAD-X
During tracking in MAD-X, particle loss is checked at the entrance of the element by comparing
the particle coordinates and the defined aperture, provided that the APERTURE flag is true
in the TRACK command, and that the APERTYPE attribute value of the element is one of the
predefined types. An aperture model defined in an external file (APERTYPE=filename) is not
used to check particle loss during tracking.
Example:
COLLIM: COLLIMATOR, L=0.5, APERTYPE=ellipse, APERTURE=0.01,0.005;
The straight reference system for a collimator is a Cartesian coordinate system.
NOTE: A collimator can be displaced transversally in order to model an asymmetric col-
limator by means of the APER OFFSET attributes; During tracking particle losses are then
reported with coordinates with respect to the displaced reference system, not with respect
to the surrounding beam line.
Other collimator elements have been inherited from MAD-8 and still exist in MAD-X for back-
ward compatibility. ECOLLIMATOR (elliptic aperture collimator) and RCOLLIMATOR (rectangular
aperture collimator) behave both as drift spaces in MAD-X They are declared with
label: ECOLLIMATOR, L=real, XSIZE=real, YSIZE=real;
label: RCOLLIMATOR, L=real, XSIZE=real, YSIZE=real;
Either type has several real attributes:
LThe collimator length (default: 0 m).
XSIZE The horizontal half-aperture (default: unlimited).
OBSOLETE : parsed and stored but not used.
YSIZE The vertical half-aperture (default: unlimited).
OBSOLETE : parsed and stored but not used.
Users are STRONGLY advised to replaced all instances of RCOLLIMATOR and ECOLLIMATOR
in input files with appropriate COLLIMATOR elements. The RCOLLIMATOR and ECOLLIMATOR
elements are only kept for the time being for backward compatibility and will be removed in
the near future.
Note also that the XSIZE and YSIZE parameters can be declared but are simply ignored both
in the APERTURE command an in tracking.
10.20 Beam-beam Interaction
The BEAMBEAM element may be inserted in a beam line to simulate a beam-beam inter-
action point:
![](asset-5c.png)
90 CHAPTER 10. ELEMENT TYPES
label: BEAMBEAM, CHARGE=real,
XMA=real, YMA=real,
SIGX=real, SIGY=real, WIDTH=real,
BBSHAPE=integer, BBDIR=integer;
The beam-beam interaction is represented by a four-dimensional interaction with a thin el-
ement, i.e. horizontal and vertical non-linear kicks. The code for this element has been
contributed by J.M. Veuillen (1987) and extended by S. Sorge (2007).
CHARGE The charge of particles in the opposite beam, in elementary charges. (default:
1).
In order to describe interactions between beams containing the same particles
and having a charge different from 1, one should set the CHARGE explicitly in
the BEAM command as well as in the BEAMBEAM element.
XMA The horizontal displacement of the opposite beam with respect to the ideal
orbit (default: 0 m).
YMA The vertical displacement of the opposite beam with respect to the ideal orbit
(default: 0 m).
SIGX The horizontal extent of the opposite beam (default: 1 m). Meaning depends
on parameter BBSHAPE.
SIGY The vertical extent of the opposite beam (default: 1 m). Meaning depends on
parameter BBSHAPE.
WIDTH The extent of the edge region, relative to the SIGX, SIGY parameters. The ab-
solute value is given by WIDTH*SIGX and WIDTH*SIGY in horizontal and vertical
directions respectively.
BBSHAPE A parameter to select the radial density shape of the opposite beam (default:
1)
•BBSHAPE=1: Gaussian shape (default). SIGX, SIGY are the standard de-
viations in horizontal and vertical directions. WIDTH is ignored.
•BBSHAPE=2: trapezoidal shape (see fig.10.2). SIGX, SIGY are the half
widths of density profile, i.e. distance from the centre to half edge region
with linear decrease of density in horizontal and vertical directions.
Only circular opposite beam is possible, i.e. in the calculations SIGX’ =
SIGY’ = (SIGX+SIGY)/2 is used, if SIGX and SIGY have different values.
WIDTH denotes the full width of the edge region in units of SIGX (or
SIGX’ and SIGY’, respectively, if SIGX and SIGY are not equal), i.e. if
WIDTH=0.01 and SIGX=5mm, the edge region has a full width of 0.05 mm.
Condition: WIDTH <2.0.
•BBSHAPE=3: hollow-parabolic shape (see fig.10.3). SIGX, SIGY are the
distances from the centre to the maximum of the parabolic density profile
in horizontal and vertical directions.
Only circular opposite beam is possible, i.e. in the calculations SIGX’ =
SIGY’ = (SIGX+SIGY)/2 is used, if SIGX and SIGY have different values.
![](asset-5d.png)
10.20. BEAM-BEAM INTERACTION 91
Figure 10.2: Trapezoidal shape of radial density for beam-beam lens.
WIDTH denotes the full width at half maximum of the parabolic density
profile in units of SIGX, or SIGX’ and SIGY’, respectively, if SIGX and
SIGY are not equal. Condition: WIDTH <SQRT(2.0)
The restriction to circular opposite beams in the cases BBSHAPE=2,3 appears
to be sufficient, because such beam profiles are more important for the descrip-
tion of the interaction between the particle beam and an electron beam of an
electron cooler, which are usually circular.
BBDIR A parameter to select the direction of motion of the opposite beam relative to
the beam being studied. It determines the sign of the Lorentz force between
both beams (default: -1):
•BBDIR=-1: Beams move in opposite directions as in a collider. The
Lorentz force enhances the beam-beam interaction.
•BBDIR=0: Opposite beam does not move. The Lorentz force is neglected
•BBDIR=1: Beams move in the same direction as in an electron cooler. The
Lorentz force reduces the beam-beam interaction.
Note:
•The particles in the beam considered may have a momentum deviation
![](asset-5e.png)
92 CHAPTER 10. ELEMENT TYPES
Figure 10.3: Hollow parabolic shape of radial density for beam-beam lens.
given by DELTAP defined in the TRACK command.
•The opposite beam is assumed to have the velocity according to the unper-
turbed energy of the particles in the beam considered. Only the direction
of motion can be chosen.
•In the case of motion in the opposite direction (BBDIR=-1), the time of in-
teraction between the beams is given by tau=length/(2*beta*clight),
where length is the length of a bunch in the opposite beam. In the case of
motion in the same direction (BBDIR=1) as in an electron cooler, this time
is given by tau=length/(beta*c light), where length is the length of
the cooler. Note that the factor 1/2 is inserted only for BBDIR=-1 to
calculate correct results.
A beam-beam element requires the particle ENERGY and the particle CHARGE, as well as the
number of particles per bunch (NPART) to be set by a BEAM command before any calculation
is performed.
Example of a four-dimensional beam-beam element definition in Collider regime:
beam, particle = positron, npart = 1.e12, energy = 50.0;
bb: beambeam, sigx = 1.e-3, sigy = 5.e-4, charge = 1.;
![](asset-5f.png)
10.21. ARBITRARY MATRIX ELEMENT 93
Example of a four-dimensional beam-beam element definition in an Electron cooler:
gamma0 = 1.032; ! relativistic factor
beta0 = sqrt(1.0-1.0/gamm0/gamma0);
i e = 0.2; ! electron current
re cool = 0.01; ! electron beam radius
l cool = 5.0; ! cooling length
nelect = i e*l cool/beta0/clight/qelect; ! electron number in e-cooler
beam, particle=antiproton, gamma=gamma0, npart=nelect;
bb ecool: beambeam, sigx=re cool, sigy=re cool, bbshape=2,
width=0.01, charge=-1, bbdir=1;
10.21 Arbitrary Matrix Element
label: MATRIX, TYPE=string, L=real,
KICK1=real,..., KICK6=real,
RM11=real, ..., RM66=real,
TM111=real, ..., TM666=real;
The MATRIX element allows the definition of an arbitrary transfer matrix. It has four real
array attributes:
LLength of the element, which may be zero.
KICKi Defines the kick of the element acting on the six phase space coordinates.
RMik Defines the linear transfer matrix (6*6) of the element.
TMikl Defines the second-order terms (6*6*6) of the element.
Data values that are not explicitly entered are taken from the identity transformation for the
RMik matrix elements, and taken as zero for the KICKi kick factors and the TMikl second order
terms. In the thin-lens tracking module a non-zero length for an arbitrary matrix is accepted,
however no non-zero second order terms are allowed to avoid non symplectic tracking runs.
In the latter case the tracking run is aborted.
10.22 Rotation around the vertical axis
The element YROTATION rotates the straight reference system about the vertical (y) axis.
label: YROTATION, ANGLE=real;
YROTATION has no effect on the beam, but it causes the beam to be referred to the new coor-
dinate system
![](asset-60.png)
94 CHAPTER 10. ELEMENT TYPES
x2=x1cos θ−s1sin θ
s2=x1sin θ+s1cos θ(10.6)
It has one real attribute:
ANGLE The rotation angle θ(default: 0 rad).
A positive angle means that the new reference system is rotated clockwise about the local
y-axis with respect to the old system.
Important note:
The rotation angle θmust be small,i.e. it must be at most comparable to the transverse
angles of the orbit.
Example:
KINK: YROTATION, ANGLE = 0.0001;
10.23 Rotation around the longitudinal axis
The element SROTATION rotates the straight reference system about the longitudinal (s) axis.
label: SROTATION, ANGLE=real;
SROTATION has no effect on the beam, but it causes the beam to be referred to the new coor-
dinate system
x2=x1cos ψ−y1sin ψ
y2=x1sin ψ+y1cos ψ(10.7)
It has one real attribute:
ANGLE The rotation angle ψ(default: 0 rad)
A positive angle means that the new reference system is rotated clockwise about the s-axis
with respect to the old system.
Example:
ROLL1: SROTATION, ANGLE= PI/2.;
ROLL2: SROTATION, ANGLE= -PI/2.;
HBEND: SBEND, L=6.0, ANGLE=0.01;
VBEND: LINE=(ROLL1,HBEND,ROLL2);
The VBEND definition above is a way to represent a bend down in the vertical plane, it could
be defined more simply by
VBEND: SBEND, L=6.0, K0S=0.01/6;
![](asset-61.png)
10.24. COORDINATE TRANSLATION 95
10.24 Coordinate translation
The element TRANSLATION changes the reference system by applying a translation of the
reference system.
label: TRANSLATION, X=real, Y=real, T=real,
PX=real,PY=real,PT=real;
TRANSLATION has no effect on the beam, but it causes the beam to be referred to the new
coordinate system translated by the vector given.
10.25 Change of reference system
The element CHANGEREF changes the reference system by applying both translations and
rotations.
label: CHANGEREF, PATCH ANG=real, real, real,
PATCH TRANS=real, real, real;
CHANGEREF has no effect on the beam, but it causes the beam to be referred to the new
coordinate system where both translations and rotations are applied.
Chapter 11. Range and Class Selection Format
11.1 RANGE
A range can be defined starting at a given element and ending at another element, both
elements included. Two forms exist:
RANGE=position;
RANGE=position1/position2;
In the first case, only one element is selected; in the second case, one or several elements are
selected. NOTE: position1 must not be behind position2 in the sequence.
”position” is composed of the element name and an optional occurrence count in the sequence:
mq.ir5.l6..1 ! no occurrence count given
mb[17] ! occurrence count given
There are two predefined indices in MAD-X:
#S The start of the beam line expanded by USE,
#E The end of the beam line expanded by USE.
If a range is selected in a USE statement:
USE, PERIOD=lhcb1, RANGE=ir1/ir5;
then the #S and #E indices refer to the start and end of the range expanded by the USE
statement.
Examples for ranges:
.., RANGE=#S; ! first element
.., RANGE=#S/#E; ! full expansion range
.., RANGE=mb[5]/#E; ! from mb 5 to end
.., RANGE=mq.ir5.l6..1; ! first occurrence of element mq.ir5.l6..1
11.2 CLASS
The single name of a class (no occurrence counts). A class is the name of an element (or basic
type) from which other elements have been derived.
Example:
MQ: quadrupole; ! makes the element MQ
Q1: MQ; ! makes the element Q1 from class MQ
Q2: MQ; ! makes the element Q2 from class MQ
Q1..A: Q1; ! makes an element from class Q1
Q2..B: Q2; ! makes an element from class Q2
96
11.2. CLASS 97
Subsequent selection with criterion CLASS="MQ" will actually select Q1, Q2, Q1..A, and
Q2..B in this example.
![](asset-64.png)
Chapter 12. Beam Lines
The accelerator to be studied is known to MAD-X either as a sequence of physical elements
called sequence, or as a hierarchically structured list of elements called a beam line. A beam
line is built from simpler beam lines whose definitions can be nested to any level. A powerful
syntax allows to repeat, to reflect, or to replace pieces of beam lines. However, internally
MAD-X knows only sequences, and lines as shown below are converted into flat sequences with
the same name when they are expanded. Consequently, only sequences can be SAVEd onto
a file.
Formally a beam line is defined by a LINE command:
label( arg ,arg ): LINE=( member ,member );
Label gives a name to the beam line for later reference.
The formal argument list (arg{,arg}) is optional (see below). Each ”member” may be one of
the following:
•Element label,
•Beam line label,
•Sub-line, enclosed in parentheses,
•Formal argument name,
•Replacement list label.
Beam lines may be nested to any level (see 12.2)
Warning:
MAD-X has been developed using sequences. Machine description using beamlines was kept
in MAD-X to a minimal extent to keep backward compatibility but also because it is often
easier to start from beamlines when designing an accelerator ab initio. But since MAD-X works
only with sequences internally, there may exist some inconveniences when only beamlines are
defined by the user. It is therefore recommended to convert beamlines into sequences (e.g.
by means of the SAVE command) as soon as possible lines in the design phase and to use only
sequences for a finalised machine.
Attempting to do sequence edition on a sequence that has been expanded in memory from a
beam line specified by the user is strongly discouraged.
12.1 Simple Beam Lines
The simplest beam line consists of single elements:
label: LINE= ( member {,member});
Example:
98
12.2. NESTED BEAM LINES 99
exmp1: LINE=(a,b,c,d,a,d);
USE, PERIOD=exmp1;
The USE command causes MAD-X to load and expand the specified beamline. This in particular
loads the beamline as a SEQUENCE representation in memory. All subsequent calculations are
done on this sequence in memory.
12.2 Nested Beam Lines
Instead of referring to an element, a beam line member can refer to another beam line defined
in a separate command. This provides a shorthand notation for sub-lines or nested beam
lines which can occur several times in a beam line. Lines and sub-lines can be entered in any
order, but when a line is expanded, all its sub-lines must be known.
Example:
top: LINE=(a,b,S,b,a,S,a,b);
S: LINE=(c,d,e);
USE, PERIOD=top;
produces the following expansion steps:
1. Replace sub-line S:
(a, b, (c,d,e), b, a, (c,d,e), a, b)
2. Omit parentheses:
a, b, c, d, e, b, a, c, d, e, a, b
12.3 Reflection and Repetition
An unsigned repetition count and an asterisk (multiplication sign) indicate repetition of a
beam line member. A repetition count can be applied to sub-lines as well as elements.
A minus prefix causes reflection, i.e. all elements in the sub-line are taken in reverse order.
Sub-lines of reflected lines are also reflected. Physical elements are not reflected head-to-tail,
hence a negative repetition count for a single element is treated as positive.
If both reflection and repetition are desired, the minus sign must precede the repetition count.
Example:
R: LINE= (g, h);
S: LINE= (c, R, d);
top: LINE= (2*S, 2*(e,f), -S, -(a,b), -2*c);
USE, PERIOD=top;
produces the following expansion steps:
1. Replace sub-line S:
![](asset-66.png)
100 CHAPTER 12. BEAM LINES
( (c,R,d), (c,R,d), (e,f), (e,f), (d,-R,c), (b,a), c, c)
2. Replace sub-line R:
( (c, (g,h), d), (c, (g,h), d), (e,f), (e,f), (d, (h,g), c), (b,a), c, c)
3. Omit parentheses:
c, g, h, d, c, g, h, d, e, f, e, f, d, h, g, c, b, a, c, c
Note that the inner sub-line Ris reflected together with the outer sub-line S.
Note also that -2*c at the end of the top line is equivalent to 2*c since single elements are not
reflected, or 2*(c) which would first promote cas a sub-line of a single element, or -2*(c)
since reverting a sub-line of a single element gives the same single element.
12.4 Replaceable Arguments
A beam line definition may contain a formal argument list, consisting of labels separated by
commas and enclosed in parentheses.
label( arg {,arg}): LINE= ( member {,member});
where member can be one of the arg arguments.
A beam line defined with arguments can be expanded for different values of its arguments.
The arguments can be beam lines or element names. The number of actual arguments given
at time of use must agree with the number of formal arguments defined at declaration time.
All occurrences of a formal argument on the right-hand side of the line definition are replaced
by the corresponding actual argument.
Example:
fodo(Q1,Q2): LINE=(Q1, d, Q2, d);
top: LINE = ( fodo(qf12, qd23), fodo(qd34,qf45) );
USE, PERIOD=top;
expands to
qf12, d, qd23, d, qd34, d, qf45, d
Example showing that arguments can also be beam lines:
s: LINE = (a,b,c);
l(x,y): LINE = (d,x,e,3*y);
l4f: LINE = (4*f);
lm2s: LINE = (-2*s);
res: LINE = l(l4f,lm2s);
Proceeding step by step, this example generates the expansion
d, f, f, f, f, e, c, b, a, c, b, a, c, b, a, c, b, a, c, b, a, c, b, a
12.5. LIMITS OF CONSTRUCTION OF BEAM LINES 101
12.5 Limits of Construction of Beam Lines
Since LINEs are in fact deprecated there are some limits to how they can be constructed.
Below is a running MAD-X example which shows a collection of valid (marked OK) and invalid
cases (marked WRONG).
!----------------------------------------------------------------------
beam, PARTICLE=electron, energy=1;
qf: QUADRUPOLE, L:=1,K1:=1;
qd: QUADRUPOLE, L:=1,K1:=-1;
d: DRIFT, l=1;
m: MARKER;
rpl(a,b): LINE=(a,b);
sl: LINE=(qf,d,qd);
test0: LINE=(rpl(sl,sl)); !OK
test1: LINE=(rpl((sl),(sl))); !OK
test2: LINE=(rpl((sl),(-sl))); !OK
test3: LINE=(sl,-sl); !OK
test4: LINE=(rpl((3*sl),(3*sl))); ! WRONG
test5: LINE=(3*sl,3*sl); !OK
test6: LINE=(rpl((3*sl),(-3*sl))); ! WRONG
test7: LINE=(3*sl,-3*sl); !OK
use, period=test0; twiss,BETX=1,bety=1;
use, period=test1; twiss,BETX=1,bety=1;
use, period=test2; twiss,BETX=1,bety=1;
use, period=test3; twiss,BETX=1,bety=1;
use, period=test4; twiss,BETX=1,bety=1;
use, period=test5; twiss,BETX=1,bety=1;
use, period=test6; twiss,BETX=1,bety=1;
use, period=test7; twiss,BETX=1,bety=1;
!----------------------------------------------------------------------
![](asset-68.png)
Chapter 13. Sequences
MAD-X accepts two forms of an accelerator definition: sequences and beam lines (See 12).
However, the sequence definition is the only one used internally by MAD-X.
A sequence is declared with the following statements:
name: SEQUENCE, L=real, REFER=string, REFPOS=string,
ADD PASS=integer, NEXT SEQU=string;
...
label: class, AT=real, FROM=string {, add. attributes};
...
class, AT=real, FROM=string {, add. attributes};
...
seqname, AT=real, FROM=string;
...
ENDSEQUENCE;
The first statement declares a sequence, giving it a name in the form of a label, and providing
some key parameters:
L(Default: 0)
the total length of the sequence in meters.
REFER = keyword in {entry, centre, exit}(Default: centre)
specifies which part of the element is taken as the reference point at which the
position along the beamline is given.
REFPOS (Default: none)
specifies the name of an element in this sequence to be used as the reference
point for the insertion of this sequence in another sequence.
ADD PASS (Default: 0)
specifies a number of additional passes (max. 5) through the structure; in case
of an RBEND the angle will be overwritten in survey using the i-th component
(1 ≤i≤add pass ≤5) of its array of angles attribute (see RBEND).
NEXT SEQU (Default: none)
specifies the name of a sequence to be concatenated to the end of this sequence.
The ENDSEQUENCE statement terminates the declaration of a sequence.
Inside the SEQUENCE ...ENDSEQUENCE bracketing keywords, several types of statements may
appear:
•An element declaration with label:
label: class, AT=real, FROM=string {, attributes};
an element declaration as usual, with additional AT attribute giving the relative element
position, and an optional FROM attribute;
102
![](asset-69.png)
103
CAUTION: an existing definition for an element with the same name (label) will be
replaced, however, defining the same element twice inside the same sequence results
in a fatal error, since a unique object (this element) would be placed at two different
positions.
•An element declaration from class:
class, AT=real, FROM=string {, attributes};
a class name causing an instance of that class to be created with specified attributes,
with additional AT attribute giving the relative element position, and an optional FROM
attribute;
For uses inside ranges, instances of the same class can then be accessed with an occur-
rence count.
•A sub-sequence name:
seqname, AT=real, FROM=string ;
a sequence name with additional AT attribute giving the relative element position, and
an optional FROM attribute;
Depending upon the REFER attribute of the current (containing) sequence, the entry,
centre, or exit of the inserted sequence is placed at the position specified.
HOWEVER, if the inserted sequence has a REFPOS attribute containing the name of an
element in the inserted sequence, the inserted sequence is placed such that the element
pointed to by REFPOS is at the location specified in the current sequence.
The additional arguments that can be given in the declaration of sequence elements or sub-
sequences are:
AT mandatory argument giving the location at which the reference point (entry,
centre or exit) of the element is to be placed in the sequence.
The value is absolute with the zero reference at the start of the sequence,
unless a FROM argument is specified.
FROM optional argument giving the name of an element of the same sequence.
The location given for the current element or sequence in the AT argument is
then taken as relative to the position of the center of the element given by
the FROM argument.
Only simple elements should be used in FROM; specifying a sequence as the
FROM reference can lead to erroneous sequence expansion at best.
When the sequence is expanded in a USE command, MAD-X generates the drift spaces between
elements according to the following rules:
1. When the distance between the exit of the previous element and the entrance of the
next element is positive and greater than a threshold of 1µm an explicit drit is generated
with its own name, unless an already existing drift space with the same length (within
10−12mtolerance) can be re-used.
2. When the absolute value of the distance between the exit of the previous element and
the entrance of the next element is less than a given tolerance of 1µm, no drift space is
![](asset-6a.png)
104 CHAPTER 13. SEQUENCES
created and the elements are considered as contiguous.
Note that in very specific cases this can cause very small errors to accumulate
and the actual length of the sequence can vary slightly from the declared
length. (see second example below)
3. When the distance between the exit of the previous element and the entrance of the next
element is negative and less than −1µm, the elements are considered to be overlapping
and MAD-X terminates with a “negative drift length” fatal error.
When the sequence is expanded in a USE command, MAD-X also checks that the calculated
sequence length is positive and that the lengths of all elements in the sequence are positive
or zero. A negative sequence length or negative element length stops the MAD-X run with a
fatal error.
For efficiency reasons MAD-X imposes an important restriction on element lengths and
positions: once a sequence is expanded, the element positions and lengths are considered as
fixed; in order to vary a position or element length, a re-expansion of the sequence becomes
necessary. The MATCH command contains a special flag VLENGTH to vary element lengths
during matching.
Example:
! Define element classes for a simple cell:
b: sbend, l = 35.09, angle = 0.011306116;
qf: quadrupole, l = 1.6, k1 = -0.02268553;
qd: quadrupole, l = 1.6, k1 = 0.022683642;
sf: sextupole, l = 0.4, k2 = -0.13129;
sd: sextupole, l = 0.76, k2 = 0.26328;
! define the cell as a sequence:
sequ: sequence, l = 79;
b1: b, at = 19.115;
sf1: sf, at = 37.42;
qf1: qf, at = 38.70;
b2: b, at = 58.255, angle = b1->angle;
sd1: sd, at = 76.74;
qd1: qd, at = 78.20;
endm: marker, at = 79.0;
endsequence;
Example of very small drift space being ignored during sequence expansion:
QTEST: QUADRUPOLE, L=1.000001;
TEST: SEQUENCE, REFER=centre, L=2.;
![](asset-6b.png)
105
QTEST, AT=1.5;
ENDSEQUENCE;
USE, SEQUENCE=TEST;
SURVEY, FILE=’test’;
The above sequence will expand to a total length of 2.0000005m, half a micron longer than
the claimed length of 2m, but will not fail.
Part III
Input and Output
106
![](asset-6d.png)
Chapter 14. TFS File Format
TFS[8] is a an acronym for the “Table File System”. TFS files have been used in the LEP
control system. The MAD-X program knows only coded TFS files. The TFS format has been
chosen for all table output of MAD-X.TFS formatted tables can be read back into MAD-X, and
may then be further processed.
14.1 Descriptor Lines
MAD-X writes the following descriptors in all tables:
•COMMENT: The current title string from the most recent TITLE command.
•ORIGIN: The version of MAD-X used.
•DATE: The date of the MAD-X run.
•TIME: The wall clock time of the MAD-X run.
•TYPE: The type of the table: e.g. TWISS
Additional descriptors exist in the Twiss table, as well as the Track tables.
14.2 Column Formats
The column formats used are listed below:
Table 14.1: Column Formats used in TFS Tables
C format Meaning C format
%hd Short integer (%8d)
%le Long float (%-18.10g)
%ks String of length k (”\”-18s\””)
Control lines begin with the TFS control character, followed by a blank. Data lines begin
with two blanks. Columns are also separated by one blank character. The column width is
chosen such as to accommodate the largest of the column name and the width of data values
of the column.
14.3 Twiss TFS file header
MAD-X gives access to parameters from TWISS and other tables using the table access com-
mands.
107
![](asset-6e.png)
Chapter 15. Conversion to SixTrack
SixTrack[9][10] is a separate beam optics code that is often used for long term tracking of
particles, e.g. for dynamic aperture studies, because of its speed and controllability.
However the input files are notoriously difficult to produce by hand. This command may be
used to generate SixTrack input files from a sequence loaded in MAD-X.
SIXTRACK, CAVALL=logical,
MULT AUTO OFF=logical, MAX MULT ORD=integer,
SPLIT=logical, APERTURE=logical, RADIUS=real;
The parameters are defined as:
CAVALL (optional flag) This puts a cavity element (SixTrack identifier 12) with Volt,
Harmonic Number and Lag attributes at each location in the machine. Since
for large hadron machines the cavities are typically all located at one particular
spot in the machine and since many cavities slow down the tracking simulations
considerably all cavities are lumped into one and located at the first appearance
of a cavity. This default is enforced by omitting this flag.
MULT AUTO OFF (optional flag, default = FALSE) If TRUE, this module does not process zero
value multipoles. Moreover, multipoles are prepared by SIXTRACK (output file
fc.3) to be treated up to the order as specified with MAX MULT ORD.
MAX MULT ORD (optional parameter, default = 11) Process up to this order for MULT AUTO OFF
= TRUE.
SPLIT (optional flag) OBSOLETE. This splits all the elements in two. This is for
backward compatibilty only. The user should now use the MAKETHIN command
instead.
APERTURE (optional flag) flag to convert the apertures from MAD-X to SixTrack so that
SixTrack can track with apertures defined. The aperture data is found in file
fc.3.aper.
RADIUS (optional, default value is 1 m). This sets the reference radius for the magnets.
This argument is optional but should normally be set.
Important Notes:
•The files contain all information concerning optics, field errors and misalignments.
Hence these should all be set and a
TWISS, SAVE;
command should always be issued before calling the SIXTRACK command.
•The BV flag is presently ignored by SIXTRACK.
•SixTrack and the MAD-X command SIXTRACK are presently set up for names of a max-
imum of 16 characters!!!!! Therefore, it is mandatory to respect this limit for MAD-X
names.
108
109
The SIXTRACK command always produces at least one output file:
•fc.2 - the basic structure of the lattice.
It may also produce any or all of the following files, depending on the sequence and command
attributes:
•fc.3 - multipole mask(s).
•fc.3.aux - various beam parameters.
•fc.3.aper - aperture element data (units are mm and degrees).
•fc.8 - misalignments and tilts.
•fc.16 - field errors and/or combined multipole kicks.
•fc.34 - various optics parameters at various locations
This file is not needed by SixTrack but may be used as input to the program SODD[11].)
For a full description of these files see the SixTrack website[10], the SixTrack user manual[9];
and for information on running SixTrack see the SixTrack run environment description[12].
![](asset-70.png)
Chapter 16. SXF file format
An SXF[13] lattice description is an ASCII listing that contains one named, “flat”, ordered list
of elements, delimited as {. . . }, with one entry for each element. The list resembles a MAD-X
“sequence” describing the entire machine. The syntax is supposed to be adapted for ease of
reading by human beings and for ease of parsing by LEX and YACC.
16.1 SXFWRITE
The command
SXFWRITE, FILE=filename;
writes the current sequence with all alignment and field errors in SXF format onto the file
specified. This then represents one ”instance” of the sequence, where all parameters are
given by numbers rather than expressions; the file can be read by other programs to get a
complete picture of the sequence.
16.2 SXFREAD
The command
SXFREAD, FILE=filename;
reads the file ”filename” in SXF format, stores the sequence away and loads the sequence in
memory through the USE mechanism in order to keep the existing errors.
It is therefore possible to write a lattice complete with errors to a named file and reload it
later in a different MAD-X job:
! define sequence MYSEQU
USE, mysequ;
! add alignment errors and field errors
SXFWRITE, FILE = file;
STOP;
and later:
SXFREAD, FILE = file;
! sequence mysequ is now reloaded and active, complete with errors.
TWISS;
...
110
![](asset-71.png)
Chapter 17. Plotting data
Values contained in MAD-X tables can be plotted directly in MAD-X in the form column versus
column, with up to four differently scaled vertical axes and up to 10 different variables in
total for all vertical axes.
If the horizontal axis is the position ”s” of the elements in a sequence, a symbolic represen-
tation of the beamline can also be displayed at the top of the plot.
In certain conditions true interpolation of optical functions and parameters inside the element
is available through internal slicing of the element and a call to the TWISS module for each
slice.
The basic plot attributes, such as line thickness, annotation size, and PostScript format and
interpolation can be set with the SETPLOT command and reset with the RESPLOT command.
Note also that for various reasons a sequence must be defined before the PLOT command can
be invoked.
17.1 PLOT
PLOT, HAXIS= string, HMIN=real, HMAX=real,
VAXIS= string1, ..., stringn,
VAXIS1= string1, ..., stringn,
VAXIS2= string1, ..., stringn,
VAXIS3= string1, ..., stringn,
VAXIS4= string1, ..., stringn,
VMIN=reals, VMAX=reals,
TITLE=string,
BARS=integer, STYLE=integer,
COLOUR=integer, SYMBOL=integer,
INTERPOLATE=logical, ZERO SUPPR=logical,
NOVERSION=logical, NOLINE=logical, NOTITLE=logical,
MARKER PLOT=logical, RANGE PLOT=logical, RANGE=range,
TABLE=tabname, PARTICLE=particle1,particle2,...,particlen,
MULTIPLE=logical, FILE=file name start, TRACKFILE=basename,
PTC=logical, PTC TABLE=tabname;
where the parameters have the following meaning:
HAXIS name of the horizontal variable
HMIN, HMAX lower and upper edge for horizontal axis. Both values must be provided.
VAXIS one or several variables from the table to be plotted against a single vertical
axis. If more than 10 variables are specified, the plot is not produced.
111
![](asset-72.png)
112 CHAPTER 17. PLOTTING DATA
VAXISione or several variables from the table to be plotted against vertical axis num-
ber i. There is a maximum of 4 vertical axes.
If the number of variables given for a single vaxisiwould push the total number
of variables beyond the maximum of 10, the variables given for this vaxisi, as
well as those for subsequent vaxes, are ignored but the plot is produced for the
variables accumulated so far.
Important: VAXIS and VAXIS1..4 are exclusive in their application! if VAXIS
is given, VAXIS1..4 will be simply ignored.
VMIN, VMAX lower and upper edge(s) for vertical axis or axes, up to four numbers separated
by commas.
Note that both vmin and vmax must be given for an axis to be effective.
TITLE plot title string; if absent, the last overall title is used; if no such overall title
as well, the sequence name is used.
BARS 0 (default) or 1 - with bars=1, all data points are connected to the horizontal
axis with vertical bars.
STYLE 1 (default), 2, 3, or 4: line style for connecting the successive data points,
respectively solid, dashed, dotted, and dot-dashed; the special value style=100
uses the four styles in turn for successive curves in the same plot. With style=0
successive data points are not connected.
COLOUR 1 (default), 2, 3, 4, or 5: colour for the symbols and lines, respectively black,
red, green, blue, and magenta; The special value colour=100 uses the five
colours in turn for successive symbols and lines.
SYMBOL 0 (default), 1, 2, 3, 4, or 5: The symbols to be plotted at data points, respec-
tively none, dot (”.”), plus (”+”), star (”*”), circle (”o”), and cross (”x”). The
symbol size may have to be adapted through the SETPLOT command (see
below).
Note that if symbol and style are both zero, style is forced to its default value
(style=1) otherwise the plot would be invisible.
INTERPOLATE logical, default=false. The data points are normally connected by straight
lines; if INTERPOLATE is specified, the following on-momentum Twiss parame-
ters are interpolated with calls to the Twiss module inside each element: beta,
sqrt(beta), alfa, phase advance, orbit, angle, dispersion and its first derivative,
for both planes.
For all other variables splines are used to smooth the curves.
Note that setting this option is ineffective if the INTERPOLATE option of the
SETPLOT command has been set to true; in this case all plots will be interpo-
lated.
Note also that because the INTERPOLATE option causes TWISS to be called
internally with a range, any range that might have been defined in a previous
USE command is lost. In this case the USE statement must be reiterated with
the range, which in turn could cause assigned errors to be lost.
ZERO SUPPR To be documented (default = false)
![](asset-73.png)
17.1. PLOT 113
NOVERSION logical, default=false. If NOVERSION=true, the information concerning the
version of MAD-X and the date of the run are suppressed from the title. This
option frees additional space available for the user specified title of the plot.
NOLINE logical, default=false. If sis the horizontal variable, then a symbolic rep-
resentation of the beamline is plotted above the plot, except for tables that
have been read back into MAD-X. In case the horizontal scale is too large and
the density of beamline elements is too high, this may result in hardly legible
representation and a thick black block in the worst case. The NOLINE=true
option suppresses this symbolic representation of the beamline.
NOTITLE logical, default=false. If true, suppresses the title line, including the informa-
tion on the version and date.
MARKER PLOT logical, default=false. If true, plotting is done also at the location of marker
elements. This is only useful for the plotting of non-continuous functions like
the ”N1” from the aperture module. Beware that the PS file might become
very large if this flag is invoked.
RANGE PLOT logical, default=false. Allows the specification of a plotting range for the user
defined horizontal axis.
RANGE horizontal plot range given by elements.
TABLE name of the table from which data is plotted (default: twiss).
If the first part of the name of the table is ”track”, the data to be plotted are
taken from the tracking file(s) generated by a previous TRACK command for
each requested particle as defined by the particle attribute. If the required file
has not been generated by a preceding TRACK command, no plot is done for
that particle.
The plot is generated through the GNUPLOT program and is available in the
format specified by the SETPLOT command.
The preceding TRACK command should contain the attribute DUMP and
may contain the attribute ONETABLE. The tracking plots are appended to
the file file name.ps where file name can be specified via the attribute file-
name=file name. Note that the plots are appended to this file and the file is
not overwritten.
The PLOT command uses the following tracking output files depending on the
name of the table.
With the attribute table=trackone, the data file is assumed to have been gen-
erated with the ONETABLE=true attribute of the TRACK command, and
the file name has the following format: basisone where the basis for the file
name is defined by the attribute trackfile=basis (default=track).
With the attribute table=trackxxx where xxx is any string other than ”one”,
the data files are assumed to have been generated with the ONETABLE=false
attribute of the TRACK command, and the file names have the following for-
mat: basis.obs0001.p00i where the basis for the file name is defined by the
attribute trackfile=basis (default=track), the observation point fixed is to 1
and the particle number iis given by the attribute particle=i.
![](asset-74.png)
114 CHAPTER 17. PLOTTING DATA
PARTICLE one or several numbers associated to the tracked particles for which the spec-
ified plot has to be displayed.
MULTIPLE logical, default=false. If true all the curves generated for each tracked particle
are put on one plot. Otherwise there will be one plot for each particle.
FILE basename for the Postscript file(s). Only the first occurrence of such a name
will be used. Default is ”madx” or ”madx track” if the table attribute is
track. Depending on the format (.ps or .eps, see below) the plots will either
all be written into one file file name.ps, or one per plot into file name01.eps,
file name02.eps, etc.
Note: in the case of several PLOT commands in a single MAD-X job, the first
FILE argument determines the basename and other FILE arguments in subse-
quent PLOT commands are ignored.
TRACKFILE basename of the files containing tracking data for each particle (default: track)
PTC logical, default=false. If set true, the data to be plotted are taken from the
table defined by the attribute ptc table which is expected to be generated pre-
viously by the ptc package. The data belong to the column identified by one
of the names set in the definition of the ptc twiss table. Interpolation is not
available and the attribute interpolate has no effect.
This option is recommended when plotting data obtained from PTC TWISS since
there is no mechanism for PLOT to physically interpolate the optical functions
beyond using splines with no mechanism for constraints with derivatives. In
most cases the INTERPOLATE option with data obtained with PTC TWISS pro-
duces unphysical data representation.
PTC TABLE name of the ptc twiss table from which data is plotted (default: ptc twiss)
17.2 SETPLOT
SETPLOT, POST= integer, FONT= integer, LWIDTH= real,
XSIZE= real, YSIZE= real,
ASCALE= real, LSCALE= real, SSCALE= real, RSCALE= real,
INTERPOLATE= logical;
where the parameters have the following meaning:
POST default = 1. If =1, makes one PostScript file (.ps) with all plots; if =2, makes
one Encapsulated PostSscript file (.eps) per plot.
FONT there are two defaults: 1 for screen plotting: this uses characters made from
polygons; -1 for PostScript files; this is Times-Italic. There are various fonts
available for positive and negative integers, best to be tried out, since they
will look different on different systems anyway. GhostView will show strange
vertical axis annotations, but the printed versions are normally OK.
LWIDTH default = 1. Allows the user to set the curve line width. Depends on the
system as well, so to be tried out.
![](asset-75.png)
17.3. RESPLOT 115
XSIZE bounding box size for PostScript, default=27 cm.
YSIZE bounding box size for PostScript, default=19 cm.
ASCALE annotation character height scale factor, default=1.
LSCALE axis label character height scale factor, default=1.
SSCALE curve symbol (see above) scale factor, default=1.
RSCALE axis text character height scale factor, default=1.
INTERPOLATE (default=false) The data points are normally connected by straight lines; if
INTERPOLATE is specified, the following on-momentum Twiss parameters are in-
terpolated with calls to the Twiss module inside each element: beta, sqrt(beta),
alfa, phase advance, orbit, angle, dispersion and its first derivative, for both
planes.
For all other variables splines are used to smooth the curves.
Note that if INTERPOLATE=true, all subsequent PLOT commands use interpo-
lation, irrespective of the setting of their INTERPOLATE attribute;
If INTERPOLATE=false (default), all subsequent PLOT commands respect the
setting of their INTERPOLATE attribute.
17.3 RESPLOT
RESPLOT;
resets all defaults for the SETPLOT command.
17.4 First example for plots of tracking data
The following MAD-X code sample defines the tracking of four particles with the generation of
a single file with name basisone holding the tracking data for all four particles.
// track particles
track, file=basis, dump, onetable;
start, x= 2e-3, px=0, y= 2e-3, py=0;
start, x= 4e-3, px=0, y= 4e-3, py=0;
start, x= 6e-3, px=0, y= 6e-3, py=0;
start, x= 8e-3, px=0, y= 8e-3, py=0;
run,turns=1024;
endtrack;
The following sample code defines the plotting of the x-px and y-py phase space coordinates
for all four particles. It takes into account the fact that all coordinates are in a single file
![](asset-76.png)
116 CHAPTER 17. PLOTTING DATA
with table=trackone and defines the filename where tracking data is to be found (basisone)
with trackfile=basis.
// plot trajectories
setplot, post=1;
title, "FODO phase-space test";
plot, file=plot, table=trackone, trackfile=basis, noversion, multiple,
haxis=x, vaxis=px, particle=1,2,3,4;
plot, file=plot, table=trackone, trackfile=basis, noversion, multiple,
haxis=y, vaxis=py, particle=1,2,3,4;
With each plot command a temporary file gnu plot.gp containing GNUPLOT instructions is
generated. The file generated by the first plot command reads:
set terminal postscript color
set pointsize 0.48
set output ’tmpplot.ps’
set title "FODO phase-space test"
set xlabel ’x’
set ylabel ’px’
plot ’basisone’ using 3:($1==1 ? $4 : NaN) title ’particle 1’ with points pointtype 1 , \
’basisone’ using 3:($1==2 ? $4 : NaN) title ’particle 2’ with points pointtype 2 , \
’basisone’ using 3:($1==3 ? $4 : NaN) title ’particle 3’ with points pointtype 3 , \
’basisone’ using 3:($1==4 ? $4 : NaN) title ’particle 4’ with points pointtype 4
MAD-X then calls GNUPLOT as a subprocess to execute this file, which generates the file
tmpplot.ps. The file tmpplot.ps is then appended to the file plot.ps determined by the
attribute file=plot. The files gnu plot.gp and tmpplot.ps are then discarded.
The same process is repeated for the second plot command, resulting in a growing file plot.ps.
17.5 Second example for plots of tracking data
The following MAD-X code sample defines the tracking of four particles with the generation of
individual files with name basis.obs0001.p000i with i=1..4 holding the tracking data for each
of the four particles.
// track particles
track, file=basis, dump;
start, x= 2e-3, px=0, y= 2e-3, py=0;
start, x= 4e-3, px=0, y= 4e-3, py=0;
start, x= 6e-3, px=0, y= 6e-3, py=0;
start, x= 8e-3, px=0, y= 8e-3, py=0;
![](asset-77.png)
17.5. SECOND EXAMPLE FOR PLOTS OF TRACKING DATA 117
run,turns=1024;
endtrack;
The following sample code defines the plotting of the x-px and y-py phase space coordinates
for all four particles with the data of all four particles on a single plot. It takes into account the
fact that coordinates for all four particles are in separate files with table=trackfodo and defines
the filename where tracking data is to be found (basis.obs0001.p000i) with trackfile=basis.
// plot trajectories
setplot, post=1;
title, "FODO phase-space test";
plot, file=plot, table=trackfodo, trackfile=basis, noversion, multiple,
haxis=x, vaxis=px, particle=1,2,3,4;
plot, file=plot, table=trackfodo, trackfile=basis, noversion, multiple,
haxis=y, vaxis=py, particle=1,2,3,4;
With each plot command a temporary file gnu plot.gp containing GNUPLOT instruction is
generated. The file generated by the first plot command reads:
set terminal postscript color
set pointsize 0.48
set output ’tmpplot.ps’
set title "FODO phase-space test"
set xlabel ’x’
set ylabel ’px’
plot ’basis.obs0001.p0001’ using 3:4 title ’particle 1’ with points pointtype 1 , \
’basis.obs0001.p0002’ using 3:4 title ’particle 2’ with points pointtype 2 , \
’basis.obs0001.p0003’ using 3:4 title ’particle 3’ with points pointtype 3 , \
’basis.obs0001.p0004’ using 5:4 title ’particle 4’ with points pointtype 4
MAD-X then calls GNUPLOT as a subprocess to execute this file, which generates the file
tmpplot.ps. The file tmpplot.ps is then appended to the file plot.ps determined by the
attribute file=plot. The files gnu plot.gp and tmpplot.ps are then discarded.
The same process is repeated for the second plot command, resulting in a growing file plot.ps.
![](asset-78.png)
118 CHAPTER 17. PLOTTING DATA
17.6 MAD-X PLUGINS
MAD-X provides a plug-in mechanism for functionality extensions. Plug-in technique does not
require linking at build time. The job is done at the time plug-in is loaded by the dynamic
linker. In order to use any plug-in, the plugin support must be compiled in MAD-X. At the top
of every MAD-X makefile there is variable PLUGIN SUPPORT that must be set to ”YES”.
Then, the appropriate C interface is compiled in and MAD-X is linked dynamically. Plug-ins
must be accessible to the dynamic linker. The default location where plug-ins are searched
is $HOME/.madx/plugins. Otherwise, the directory containing the library must be either
listed in
1. the configuration file of the dynamic linker (on SLC3 it is /etc/ld.so.conf)
2. LD LIBRARY PATH environment variable
Existing plug-ins
1. RPLOT
Example
PROGRAMMERS MANUAL
The interface is not yet fully defined. The documentation apears at the moment it happens.
![](asset-79.png)
17.7. RPLOT 119
17.7 RPLOT
RPLOT is a MAD-X plug-in that privides additional functionality using ROOT. It contains
several tools.
RVIEWER plotting tool that handles the results in paramremtric form
What makes it different from the standard PLOT module of MAD-X is that it is also able
to deal with the parmateric results. RPLOT provides graphical user interface that allows
to choose which functions shall be drawn, set its ranges and adjust all the details of the
plot formatting. Of course, the result is immendiately visible on the screen, in contrary
to the standard plot tool that is able to work solely in the batch mode. The user can
choose several formats to save his plot, including postscript, gif, pdf, root macro and
many others.
RVIEWER is able to draw the lattice functions
1. along the layout
2. at given position in function of one or two knobs
It provides a convienient way to set the knob values. As the value is set, the plotted
functions are immediately drawn for the new value.
In order to run RVIEWER simpy issue ”rviewer;” command
RTRACKSTORE enables storage of the tracking data in ROOT NTuple/Tree format
Ntuple and its modern extension called Tree are formats designed for storing particle
tracking data. It is proven to provide the fastest data writing and reading thanks to
column wise I/O operations. It is commonly used for data storage by HEP experiments.
Additionally, ROOT provides automatical ZIP data compression that is transparent for
the user algorithms. Morover, ROOT provides wide set of very comfortable tools for
advanced analysis and plotting of the data stored in Trees.
Addtionally, we plan to extend RVIEWER functionality that would provide intuitive
graphical user interface to most commonly used features in particle tracking in ac-
celerators. Thanks to that, the user is not forced to learn how to use the ROOT
package.
Currently the feature is enabled only for tracking using the PTC TRACKLINE command,
however, it will be extended to other tracking modes.
Installation
Prerequisite: ROOT must be installed before compilation and whenever the user wants to
use the plug-in. See explanations on ROOT webpage.
To install RPLOT
1. Unpack the archive, it will create directory rplot
tar xvzf rplot-X.XX.tgz
120 CHAPTER 17. PLOTTING DATA
2. Change to rplot directory
cd rplot
3. Type
make install
Examples
SYNOPSIS
RVIEWER;
PROGRAMMERS MANUAL
To be continued...
Part IV
MAD-X Modules
121
![](asset-7c.png)
Chapter 18. SURVEY
The SURVEY command computes the geometrical layout, i.e. the coordinates of all machine
elements in a global reference system. These coordinates can be used for installation. In
order to produce coordinates in a particular system, the initial coordinates and angles can be
specified.
SURVEY, SEQUENCE=string, FILE= string,
X0=real, Y0=real, Z0=real,
THETA0=real, PHI0=real, PSI0=real;
The SURVEY command has the following attributes:
SEQUENCE the name of sequence to be surveyed. By default the last sequence expanded
with the USE command will be surveyed.
FILE the name of external file to which the results are written.
(Default: survey)
X0, Y0, Z0 the initial horizontal, vertical and longitudinal coordinates in meters.
(Default: 0.0, 0.0, 0.0)
THETA0, PHI0, PSI0 the initial horizontal and vertical angles and transverse tilt in radians.
(Default: 0.0, 0.0, 0.0)
The computation results are written on the internal table named ”survey”. Results can also
be written on an external file. Each line contains the global coordinates of an element taken
at the end of the element.
The computation takes into account the length of each element, as well as the rotation angles
defined for SBEND,RBEND, thin MULTIPOLE and thin RFMULTIPOLE elements exclusively. Rota-
tion angles introduced via the KNL,KSL mechanism for other elements are ignored by SURVEY,
other MAD-X commands, as well as PTC commands.
Example: average LHC ring with CERN coordinates.
REAL CONST R0 = 1.0; ! to obtain the average ring
USE, SEQUENCE=lhcb1;
SURVEY, X0=-2202.21027, Y0=2359.00656, Z0=2710.63882,
THETA0=-4.315508007, PHI0=0.0124279564, PSI0=-0.0065309236,
FILE=survey.lhcb1;
WARNING :
In the case a machine geometry is constructed with thick lenses, the circumference changes
when the structure is converted into thin lenses via the MAKETHIN command. This is an
unavoidable feature of MAKETHIN. ONLY the structure with thick lenses must be used for
practical purposes.
122
![](asset-7d.png)
Chapter 19. Twiss Module
The TWISS command calculates the linear lattice functions [3], and optionally the chromatic
functions. The coupled functions are calculated in the sense of Edwards and Teng[14]. For
the uncoupled cases they reduce to the C and S functions.
TWISS, SEQUENCE=seqname, LINE=linename, RANGE=range,
DELTAP=real{,real}||initial:final:step,
CHROM=logical,
CENTRE=logical, TOLERANCE=real,
FILE=filename,
TABLE=tabname, NOTABLE=logical,
RMATRIX=logical, SECTORMAP=logical,
SECTORTABLE=tabname, SECTORFILE=filename,
KEEPORBIT=name, USEORBIT=name,
COUPLE=logical,
RIPKEN=logical;
The attributes of the TWISS command are:
SEQUENCE the name of a valid sequence for which the calculation of optical functions
should be performed.
SEQUENCE and LINE are mutually exclusive.
(Default: sequence or beam line defined in the latest USE command)
LINE the name of a valid beamline for which the calculation of optical functions
should be performed.
SEQUENCE and LINE are mutually exclusive.
(Default: sequence or beam line defined in the latest USE command)
RANGE (Default: #S/#E)
The TWISS calculation is restricted to the specified range. See RANGE.
DELTAP =real{,real}or initial:final:step (Default: 0.0)
The relative energy error DELTAP may be entered in one of the two forms above.
The first form lists several numbers, which may be general expressions, sepa-
rated by commas. The second form specifies an initial value, a final value, and
a step, which must be constant expressions, separated by colons.
For example, DELTAP=0.001 defines a single value, DELTAP=0.001,0.005 de-
fines two values and DELTAP=0.001:0.007:0.002 defines four values.
CHROM a logical flag to trigger computation of the chromatic functions as well as the
radiation synchrotron integrals.
Please note that this option is needed for a proper calculation of the chromatic-
ities in the presence of coupling!
Please note that this option also changes the way that the chromaticities are
calculated: The chromaticities are normally calculated from the analysis of
the first and second order matrices. With CHROM, the chromaticities are re-
calculated by explicitely calculating the tunes for the case of the specified mo-
123
124 CHAPTER 19. TWISS MODULE
mentum deviation DELTAP and for the case of a momentum deviation equal to
DELTAP+1.e-6. The tune differences divided by 1.e-6 yield the chromaticities.
CENTRE a logical flag to enforce the calculation of the linear lattice functions at the
center of the element instead of the end of the element. The values in the
tables and in the output files are affected by this flag.
(Default: false)
Since the lattice functions are calculated inside the element the closed orbit
coordinates in the output also include the misalignment of the element.
TOLERANCE the maximum closed orbit error, for all six orbit components, that can be toler-
ated during the closed orbit search. The value given in the TWISS command is
only valid for the current calculation; the COGUESS command allows to change
the default value for all subsequent closed orbit search calculations.
(Default: 1.e-6)
FILE causes MAD-X to write a TFS Twiss table to the file specified. (Default: “twiss”)
The columns of the table can be selected using the SELECT command with the
FLAG=twiss attribute.
TABLE the name of the table where linear lattice functions as well as chromatic func-
tions are stored. (Default: “twiss”)
Note: If the TABLE option is given the selection of column names via the
SELECT command is ignored. Hence if both TABLE and FILE options are given,
the table written to file is the full twiss table, containing all elements as rows
and all known Twiss parameters as columns.
NOTABLE logical flag to prevent the creation of the internal twiss table. Consequently,
no output file is created either.
(Default: false)
RMATRIX If this flag is used the the one-turn map at the location of every element
is calculated and prepared for storage in the twiss table. Using the SELECT
command and using the column RE, RE11 ...RE16 ...RE61 ...RE66 these
components will be added to the twiss table, i.e. with "COLUMN, RE" and
"COLUMN, REij" one gets all or the component ”ij” respectively.
SECTORMAP a logical flag to initiate the calculation of a sector map.
SECTORTABLE the name of the table containing the SECTORMAP values. The elements (lines)
and parameters (columns) of the table can be tailored using the SELECT com-
mand as specified in SECTORMAP
(Default: sectortable)
SECTORFILE the name of the file to which the SECTORMAP is written. The format of the
output can be adjusted as specified in SECTORMAP
(Default: ”sectormap”)
KEEPORBIT The keeporbit attribute (with an optional name, keeporbit=”name”) stores
the orbit under this name at the start, and at all monitors.
USEORBIT The useorbit attribute (with an optional name, useorbit=”name”) uses the
![](asset-7f.png)
19.1. TWISS PARAMETERS FOR A PERIOD 125
start value provided for the closed orbit search; the values at the monitors are
used by the threader.
COUPLE (obsolete) This MAD-8 option can no longer be set since TWISS in MAD-X is
always calculated in coupled mode. MAD-X computes the coupled functions in
the sense of Edwards and Teng [14]. For the uncoupled cases they reduce to
the C and S functions.
Twiss calculation is 4D only! : The TWISS command calculates an approxi-
mate 6D closed orbit when the accelerator structure includes an active cavity.
However, the calculation of the Twiss parameters are 4D only. This may result
in apparently non-closure of the beta values in the plane with non-zero disper-
sion. The full 6D Twiss parameters can be calculated with the PTC TWISS
command.
RIPKEN This flag calculates the Ripken-Mais Twiss parameters (beta11, beta12,
beta21, beta22, alfa11, alfa12, alfa21, alfa22, gamma11, gamma12,
gamma21 and gamma22) using the parameters betx, bety, alfx, alfy,
gamax, gamay, R11, R12, R21 and R22 as input.
The tables are suitable for PLOT.
After a successful TWISS run MAD-X creates a table of summary parameters named ”SUMM”
which includes tunes, chromaticities, etc. versus the selected values of DELTAP. Please note
that the CHROM attribute is needed for a proper calculation of the chromaticities in the presence
of coupling!
Notice also that in MAD-X,DELTAP is converted to PT, which is used as longitudinal variable.
Dispersive and chromatic functions are hence derivatives with respect to PT. (see summ table).
These summary parameters can later be accessed via the table access functions using the
”SUMM” table.
19.1 Twiss Parameters for a Period
The simplest form of the TWISS command is
TWISS;
which calculates the periodic solution for the last beamline or sequence declared in a USE
statement, and with zero DELTAP. Chromatic functions are not calculated. Standard tables
(”TWISS” and ”SUMM”) are created in memory but no file is written to disk.
The slightly more elaborate version
TWISS, DELTAP=real{,real}, CHROM, TABLE=tabname;
computes the periodic solution, including chromatic functions, for the last beam line or se-
quence declared in a USE statement, for all values of DELTAP entered (or for DELTAP=0, if
![](asset-80.png)
126 CHAPTER 19. TWISS MODULE
none is entered). The tables ”tabname” and ”SUMM” are created in memory and no file is
written to disk.
Example:
USE, period=OCT;
TWISS, DELTAP=0.001, CHROM;
computes the periodic solution for the linear lattice and chromatic functions for the beam
line OCT and for DELTAP=0.001.
19.2 Initial Values from a Periodic Line
It is possible to track the lattice functions starting with the periodic solution for another
beam line. If this is desired the TWISS command takes the form
TWISS, DELTAP=real{,real}, LINE=beamline,
MUX=real, MUY=real,
TABLE=tabname;
No other attributes should appear in the command. For each value of DELTAP, MAD-X first
searches for the periodic solution for the beam line mentioned in LINE=beamline: The result
is used as an initial condition for the lattice function tracking.
Example:
CELL: LINE=(...);
INSERT: LINE=(...);
USE, period=INSERT;
TWISS, LINE=CELL, DELTAP=0.0:0.003:0.001, CHROM, FILE;
For four values of DELTAP the following happens: First MAD-X finds the periodic solution for
the beam line CELL: Then it uses this solution as initial conditions for tracking the lattice
functions of the beamline CELL: Output is also written on the file TWISS:
If any of the beam lines was defined with formal arguments, actual arguments must be pro-
vided:
CELL(SF,SD): LINE=(...);
INSERT(X): LINE=(...);
USE, period=INSERT;
TWISS, LINE=CELL(SF1,SD1);
19.3 Given Initial Values
Initial values for linear lattice functions and chromatic functions may also be numerical. Initial
values can be specified on the TWISS command:
![](asset-81.png)
19.4. SAVEBETA 127
TWISS, BETX=real, ALFX=real, MUX=real,
BETY=real, ALFY=real, MUY=real,
DX=real, DPX=real, DY=real, DPY=real,
X=real, PX=real, Y=real, PY=real,
T=real, PT=real,
WX=real, PHIX=real, DMUX=real,
WY=real, PHIY=real, DMUY=real,
DDX=real, DDY=real, DDPX=real, DDPY=real,
R11=real, R12=real, R21=real, R22=real, !coupling matrix
TABLE=tabname,
TOLERANCE=real,
DELTAP=real:real:real;
All initial values for linear lattice functions and chromatic functions are permitted, but BETX
and BETY are required. Moreover, a BETA0 block can be added as filled by the SAVEBETA
command (see below). The lattice parameters are taken from this block, but are also over-
written by lattice parameters explicitly decalred on the command line. As entered, the initial
conditions cannot depend on DELTAP, and can thus be correct only for one such value.
19.4 SAVEBETA
Initial lattice parameters can be saved and transfered for later commands, in particular for
TWISS or the match module, with the SAVEBETA command sequence.
SAVEBETA, LABEL=string, PLACE=string, SEQUENCE=sequencename;
marks a location given by attribute PLACE in an expanded sequence sequence name; at the
next TWISS command execution, a BETA0 block is saved at that location with the label given
by the attribute LABEL. This is done only once; in order to get a new BETA0 block at the same
location in a subsequent TWISS command, the SAVEBETA command must be repeated. The
content of the BETA0 block can then be used in other commands, e.g. TWISS and MATCH.
Example (after sequence expansion):
SAVEBETA, LABEL=sb1, PLACE=mb[5], SEQUENCE=fivecell;
TWISS;
SHOW, sb1;
saves and then shows the BETA0 block parameters at the end (!) of the fifth element of type
mb in the sequence.
Parameters in tables can also be accessed using the table access functions.
USE, period=...;
SAVEBETA, LABEL=name, PLACE=place, SEQUENCE=s_name;
TWISS,...;
When reaching the PLACE in the sequence s name during execution of TWISS,MAD-X saves a
BETA0 block with the LABEL name: This block is filled with the values of all lattice parameters
![](asset-82.png)
128 CHAPTER 19. TWISS MODULE
in place.
Example 1:
USE, period=CELL;
SAVEBETA, LABEL=END, PLACE=#E, SEQUENCE=CELL;
TWISS;
USE, period=INSERT;
TWISS, BETA0=END;
This first example calculates the periodic solution of the line CELL, and then tracks lattice
parameters through INSERT, using all end conditions (including orbit) in CELL at the start of
INSERT.
Example 2:
USE, period=CELL;
SAVEBETA, LABEL=END, PLACE=#E, SEQUENCE=CELL;
TWISS;
USE, period=INSERT;
TWISS, BETX=END->BETY, BETY=END->BETX;
This is similar to the first example,but the beta functions are interchanged (overwritten).
19.5 BETA0: Set Initial Lattice Parameters
Initial lattice parameters can be set ’manually’ for later commands, in particular for TWISS
or the MATCH module, by using the BETA0 command attached to a label.
label: BETA0, BETX=real, ALFX=real, MUX=real,
BETY=real, ALFY=real, MUY=real,
{etc for linear and chromatic lattice functions};
ABETA0 block can be used as a whole with all values declared, as a block with overriden
values explicitly, or by extracting single values as shown in the three examples below:
Example of BETA0 block used as a whole in TWISS:
initial: BETA0, BETX=10., ALFX=0.0, MUX=0.0,
BETY=10., ALFY=0.0, MUY=0.0,
DX=1., DPX=0.0;
TWISS, BETA0=initial;
Example of BETA0 block used as a whole but with overriden values in the TWISS command:
initial: BETA0, BETX=10., ALFX=0.0, MUX=0.0,
BETY=10., ALFY=0.0, MUY=0.0,
DX=1., DPX=0.0;
TWISS, BETA0=initial, ALFX=-0.1, ALFY=0.1;
Example of using BETA0 block by extracting single values in the TWISS command:
![](asset-83.png)
19.6. SECTORMAP OUTPUT 129
initial: BETA0, BETX=10., ALFX=0.0, MUX=0.0,
BETY=10., ALFY=0.0, MUY=0.0,
DX=1., DPX=0.0;
TWISS, BETX=initial->BETX, BETY=initial->BETY;
19.6 Sectormap output
The flag SECTORMAP of the TWISS command (together with an element selection via SELECT,
FLAG=sectormap,...) causes a file ”sectormap” to be written.
For each user-selected element, it contains the user-selected coefficients of the kick vector K
(6 values), of the first-order map R(6 ×6 values) and of the second-order map T(6 ×6×6
values)
The sector file is the output of a standard TFS table, which means that the set of columns
of interest may be selected through a SELECT command as in the following example:
SELECT, FLAG=my sect table, COLUMN=name, pos, k1, r11, r66, t111;
The sectormap file contains for each selected element, one element per line, the set of chosen
K, R, and T matrix coefficients:
@ NAME %13s ”MY SECT TABLE”
@ TYPE %09s ”SECTORMAP”
@ TITLE %08s ”no-title”
@ ORIGIN %19s ”MAD-X 3.04.62 Linux”
@ DATE %08s ”18/12/08”
@ TIME %08s ”10.33.58”
* NAME POS K1 R11 R66 T111
$ %s %le %le %le %le %le
”FIVECELL$START” 0 0 1 1 0
”SEQSTART” 0 0 1 1 0
”QF.1” 3.1 -1.305314637e-05 1.042224745 1 0
”DRIFT 0” 3.265 7.451656548e-21 1 1 0
”MSCBH” 4.365 -1.686090613e-15 0.9999972755 1 0.006004411526
”CBH.1” 4.365 0 1 1 0
”DRIFT 1” 5.519992305 -6.675347543e-21 1 1 0
”MB” 19.72000769 2.566889547e-18 1.000000091 1 -4.135903063e-25
”DRIFT 2” 21.17999231 -1.757758802e-20 1 1 0
”MB” 35.38000769 2.822705549e-18 1.000000091 1 -4.135903063e-25
”DRIFT 2” 36.83999231 2.480880093e-20 1 1 0
”MB” 51.04000769 3.006954115e-18 1.000000091 1 -4.135903063e-25
”DRIFT 3” 52.21 -4.886652187e-20 1 1 0
... ... ... ... ... ...
... ... ... ... ... ...
... ... ... ... ... ...
Of course, the SELECT statement can be combined with additional options to filter-out the
list of elements, such as in the following statement, which for instance only retains drift-type
elements:
SELECT, FLAG=my sect table, CLASS=drift,
COLUMN=name, pos, k1, r11, r66, t111;
130 CHAPTER 19. TWISS MODULE
Kcoefficients range: K1... K6
Rcoefficients range:
R11 ... R61
R12 ... R62
... ... ...
R16 ... R66
Tcoefficients range:
T111 ... T611
T121 ... T621
... ... ...
T161 ... T661
T112 ... T612
... ... ...
T166 ... T666
In the above notation, Rij stands for ”effect on the i-th coordinate of the j-th coordinate in
phase-space”, whereas Tijk stands for ”combined effect on the i-th coordinate of both the
j-th and k-th coordinates in phase-space” along the lattice.
The maps are the accumulated maps between the selected elements. They contain both the
alignment, and field errors present. Together with the starting value of the closed orbit (which
can be obtained from the standard twiss file) this allows the user to track particles over larger
sectors, rather than element per element. A typical usage therefore lies in the interface to
other programs, such as TRAIN.
19.7 Beam Threader
In a machine with magnetic and alignment errors it can happen that the beam does not
circulate and the closed orbit cannot be established and measured. This can also happen in
MAD-X and the closed orbit finder does not converge.
The THREADER simulates beam steering through such a machine with repeated measurement
of trajectory over a certain number of monitors and correction of the trajectory with upstream
correctors.
When enabled, threading is executed whenever a trajectory or closed orbit search is carried
out by the TWISS module.
The threader is controlled as an option. The following MAD-X command enables the threader:
OPTION, THREADER;
and the threader can be disabled with
OPTION, -THREADER;
During the trajectory search in TWISS, or the first turn of the orbit search for a closed machine,
the threader calculates at each monitor the difference between the present trajectory reading
and a reference value. If the difference exceeds a threshold (see below), the threader searches
backwards for the first corrector that will efficiently cancel the difference and calculates the
corresponding kick. The trajectory is then recalculated starting again from that corrector
![](asset-85.png)
19.8. CLOSED ORBIT GUESS 131
and progressing forward. The calculated kicks are added to already existing kicks. If Twiss
is searching for a closed orbit which involves tracking this trajectory over many turns, the
threader is only active during the first turn.
The reference value for the trajectory difference is zero by default but can also be obtained
from a previous orbit calculation if the current TWISS command has the USEORBIT flag enabled
and a previous TWISS command had the KEEPORBIT flag enabled. This allows for example to
thread the beam into a machine with orbit bumps present.
The threshold values for triggering the threader correction can be set with the command
THREADER, VECTOR={xmax, ymax, att};
where
xmax, ymax threshold orbit excursions beyond which the threader is applied.
Defaults: 0.005, 0.005
att attenuation factor for the kicks applied by the threader
Default: 1.0
The attenuation factor defines the fraction of the calculated kick that is actually applied by
the threader. An attenuation factor of 0.5 will apply 50% of the calculated kicks.
19.8 Closed Orbit Guess
In order to help the initial finding of the closed orbit by the TWISS module, it is possible to
specify an initial guess for the coordinates of the fixed point at the start of the lattice.
COGUESS, X=real, PX=real, Y=real, PY=real, T=real, PT=real,
TOLERANCE=real,
CLEAR=logical;
The COGUESS command has the following attributes:
X, PX, Y, PY, T, PT each parameter specified defines a first guess for all future closed orbit
searches in case they are different from zero.
TOLERANCE sets the required convergence precision in the closed orbit search.
(Default: 1.e-6)
CLEAR a flag to reset the tolerance to its default value and to cancel the effect of a
previous COGUESS in defining a first guess for subsequent closed orbit searches.
(Default: false)
Example
COGUESS, X=1.e-3;
...
TWISS;
132 CHAPTER 19. TWISS MODULE
...
COGUESS, Y=-2.e-3;
...
TWISS;
...
COGUESS, CLEAR;
...
TWISS;
...
![](asset-87.png)
Chapter 20. Matching Module
Before a match operation at least one sequence must be selected by means of a USE com-
mand. Matching is initiated by the MATCH command. The matching module can act on more
than one sequence simultaneously by specifying more than one sequence when initiating the
matching mode. From this command to the corresponding ENDMATCH command MAD-X accepts
the matching commands listed below. For a mathematical description of the minimisation
procedures see [15].
In particular one may do the following:
•Define the sequence(s) the matching module will work on
•Set initial conditions for transfer line matching
•Define constraints
•Define the parameters to be varied
•Match by different methods.
The matching commands are described in detail below. Some other commands can also be
issued during matching.
•Enter and Leave Matching Mode
–MATCH: Initiate Matching Mode
–ENDMATCH: Leave Matching Mode
•Define Variable Parameter
–VARY: Set Parameter to Vary
•Define Constraint
–CONSTRAINT: Simple Constraint
–CONSTRAINT: User Defined Variables
–WEIGHT: Matching Weights
–GLOBAL: Global Constraints
–GWEIGHT: Weights for Global Constraints
•Matching Methods
–LMDIF: Fast Gradient Minimisation
–MIGRAD: Gradient Minimisation
–SIMPLEX: Simplex Minimisation
–JACOBIAN: Newton Minimisation
•Expression Matching with USE MACRO
133
![](asset-88.png)
134 CHAPTER 20. MATCHING MODULE
20.1 MATCH . . . ENDMATCH
The MATCH command is used for matching either a periodic cell or an insertion to another
part of the machine.
Both matching modes are initiated by the MATCH command using one of several forms outlined
below.
The ENDMATCH command terminates the matching section and deletes all tables related to the
matching run.
ENDMATCH;
20.2 Cell Matching
The matching of a periodic cell is initiated by a MATCH command of the form:
MATCH, SEQUENCE=’name1’, ’name2’, ..., ’name-n’,
SLOW=logical;
In this first mode the matching routine is initiated only with one attribute specifying the
sequence(s) the matching module will work on. In this matching mode the periodicity of the
optics functions is enforced at the beginning and end of the selected range.
MATCH, SEQUENCE=’name1’, ’name2’, ...,’name-n’;
20.3 Insertion Matching
The matching of an insertion to another part of the machine is initiated by a MATCH command
taking one of the following forms:
MATCH, SEQUENCE= ’name1’, ’name2’, ..., ’name-n’,
BETA0= ’beta01’, ’beta02’, ..., ’beta0n’,
SLOW=logical;
or
MATCH, SEQUENCE=’seqname’,
BETX=real, ALFX=real, MUX=real,
BETY=real, ALFY=real, MUY=real,
X=real, PX=real, Y=real, PY=real,
DX=real, DY=real, DPX=real, DPY=real,
DELTAP=real, SLOW=logical;
In the second mode, called insertion matching, the matching routine is initiated with two at-
tributes: one specifying the sequence(s) the matching module will work on and one specifying
the initial conditions of the optic functions for each sequence. In this case the initial values
are assumed as exact.
20.3. INSERTION MATCHING 135
Specification of Initial Values: The initial values of the optical functions for the insertion
matching can either be specified in form of a SAVEBETA command or by explicitly stating the
individual optic function values after the MATCH command. The two options can be entered
as
MATCH, SEQUENCE= ’name1’, ’name2’,.., ’name-n’,
BETA0= ’beta01’, ’beta02’,..., ’beta0n’;
or
MATCH, SEQUENCE=’seqname’,
BETX=real, ALFX=real, MUX=real,
BETY=real, ALFY=real, MUY=real,
X=real, PX=real, Y=real, PY=real,
DX=real, DY=real, DPX=real, DPY=real,
DELTAP=real;
Example 1:
CELL: SEQUENCE=(...) ;
INSERT: SEQUENCE=(...) ;
USE, PERIOD=cell;
SAVEBETA, LABEL=bini, PLACE=#e;
TWISS, SEQUENCE=cell;
USE, PERIOD=insert;
MATCH, SEQUENCE=insert, BETA0=bini;
CONSTRAINT, SEQUENCE=insert, RANGE=#e, MUX=9.345, MUY=9.876;
This matches the sequence ’INSERT’ with initial conditions to a new phase advance. The
initial conditions are given by the periodic solution for the sequence CELL1.
Example 2:
USE, PERIOD=insert;
MATCH, SEQUENCE=insert;
CONSTRAINT, SEQUENCE=insert, RANGE=#e, MUX=9.345, MUY=9.876;
This matches the beam line ’INSERT’ with periodic boundary conditions to a new phase
advance.
The initial conditions can also be transmitted by a combination of a SAVEBETA command and
explicit optic function specifications:
USE, cell1;
SAVEBETA, LABEL=bini, PLACE=#E;
TWISS, SEQUENCE=cell1;
USE, PERIOD=line1;
MATCH, SEQUENCE=line1, BETA0=bini, MUX=1.234, MUY=4.567;
This example transmits all values of the SAVEBETA array ’bini’ as initial values to the MATCH
command and overrides the initial phase values by the given values.
136 CHAPTER 20. MATCHING MODULE
An additional CONSTRAINT may be imposed in other places, i.e. intermediate or end values of
the optics functions at the transition point.
20.4 More than one active sequence
The matching module can act on more than one sequence simultaneously by specifying more
than one sequence after the MATCH command:
MATCH, SEQUENCE=line1, CELL1, BETA0=bini1, bini2;
This example initiates the matching mode for the ’LINE1’ and the ’CELL1’ sequence. The
Twiss functions of the two sequences are calculated with fixed initial conditions.
The SAVEBETA array ’bini1’ is used for calculating the optics functions of sequence ’LINE1’ and
the SAVEBETA array ’bini2’ for calculating the optics functions of sequence ’CELL1’. Without
the initial conditions the matching module will work in the CELL mode.
20.5 SLOW attribute
The SLOW logical flag enforces the old and slow matching procedure which allows to use the
special columns mvar1, ..., mvar4, if they are added to the twiss table.
Recently a number of parameter, like RE56, have been added to the list of parameters that
can be matched.
Nevertheless, some parameters might only be available when using the SLOW attribute.
20.6 Useful TWISS attributes
Some of the attributes of the TWISS command can be used in the MATCH command and are
transmitted to the TWISS command when it is internally invoked during the matching process.
The main TWISS attributes that can be used also in the MATCH command are:
RMATRIX If this flag is used the one-turn map at the location of every element is calcu-
lated and prepared for storage in the TWISS table.
Target values for the matrix elements at certain positions in the sequence can
be specified with the help of the CONSTRAINT command and the keywords:
RE, RE11...RE16...RE61...RE66, where REij refers to the ”ij” matrix
component.
CHROM This logical flag sets the matching process to transmit the CHROM attribute to
the TWISS command when it is invoked, enforcing the calculation of chromatic
functions and synchrotron radiation integrals, and the alternative calculation
of chromaticities as documented in TWISS.
![](asset-8b.png)
20.7. VARY 137
If this flag is used the chromatic functions at the location of every element are
calculated and prepared for storage in the TWISS table.
Target values for the chromatic functions at certain positions in the sequence
can be specified with the help of the CONSTRAINT command and the repective
keywords WX, PHIX, WY, PHIY,... for the chromatic functions.
Examples:
MATCH, RMATRIX, SEQUENCE=’name’, BETA0=’beta-block-name’;
CONSTRAINT, SEQUENCE=insert, RANGE=#e, RE11=-2.808, RE22=2.748;
VARY, NAME=kqf, STEP=1.0e-6;
VARY, NAME=kqd, STEP=1.0e-6;
This matches the sequence ’name’ with initial conditions to new values for the matrix elements
RE11 and RE22 by varying the strength of the main quadrupole circuits.
20.7 VARY
A parameter to be varied is specified by the command
VARY, NAME=variable, STEP=real, LOWER=real, UPPER=real
SLOPE=integer, OPT=real;
It has six attributes:
NAME The name of the parameter or attribute to be varied,
STEP The approximate initial step size for varying the parameter. If the step is not
entered, MAD-X tries to find a reasonable step, but this may not always work.
LOWER Lower limit for the parameter (optional),
UPPER Upper limit for the parameter (optional).
SLOPE allowed change rate (optional, available only using JACOBIAN routine). Limit
the parameter to increase (SLOPE=1) or decrease (SLOPE=-1) only.
OPT optimal value for the parameter (optional, available only using JACOBIAN
routine).
Examples:
VARY, NAME=PAR1, STEP=1.0E-4; ! vary global parameter PAR1
VARY, NAME=QL11->K1, STEP=1.0E-6; ! vary attribute K1 of the QL11
VARY, NAME=Q15->K1, STEP=0.0001, LOWER=0.0, UPPER=0.08; ! vary with limits
If the upper limit is smaller than the lower limit, the two limits are interchanged. If the
current value is outside the range defined by the limits, it is brought back to range.
If a parameter comes outside the limits during the matching process the matching module
resets the parameter to a value inside the limits and informs the user with a message. If such
a ’rescaling’ occurs more than 20 times the matching process terminates. The user should
138 CHAPTER 20. MATCHING MODULE
either eliminate the corresponding parameters from the list of varied parameters or change
the corresponding upper and lower limits before restarting the matching process.
After a matching operation all varied attributes retain their value after the last successful
matching iteration. Using JACOBIAN routine, STRATEGY=3, in case the number of parameters
is greater than the number of constraint, if a parameter comes outside the limits, it is excluded
automatically from the set of variables and a new solution is searched.
20.8 CONSTRAINT
Simple constraints are imposed by the CONSTRAINT command. The CONSTRAINT command
has three attributes:
•the SEQUENCE entry specifies the sequence for which the constraint applies.
•the RANGE entry specifies the position where the constraint must be satisfied. The
RANGE can either be the name of a single element in the sequence or a range between
two elements. In the later case the two element names must be separated by a ’/’:
RANGE=name1/name2
•the optics functions to be constrained.
The optic functions can be constrained in four different ways:
1. lower limit: BETX >value
2. upper limit: BETX <value
3. lower and upper limits: BETX <value1, BETX >value2
4. target value: BETX=value
In case one element is affected by more than one constraint command the last CONSTRAINT
will be chosen. For example, one can specify the maximum acceptable beta function over a
range of the sequence and specify the target beta function for one element that lies inside this
range. In this case one must first specify the constraint that affects the whole range and then
the constraint for the single element. This way the constraint of the target value overrides
the previous constraint on the upper limit for the selected element.
For example, the following constraint statements limit the maximum horizontal beta function
between ’marker1’ and ’marker2’ to 200 meter and require a horizontal beta function of 100
meter at element ’name1’:
CONSTRAINT, SEQUENCE=seqname, RANGE=’marker1’/’marker2’, BETX<200.0;
CONSTRAINT, SEQUENCE=seqname, RANGE=’name1’/’marker2’, BETX=100.0;
When the two constraint statements are interchanged, and supposing that name1 is an element
in the range marker1/marker2, the horizontal beta function at element ’name1’ will only be
limited to less than 200 meter and NOT constrained to 100 meter!
The CONSTRAINTS can either be specified with explicit values for the constraints of the
optic functions or via a pre-calculated SAVEBETA module. The first options has the form:
![](asset-8d.png)
20.9. USER DEFINED MATCHING CONSTRAINTS 139
CONSTRAINT, SEQUENCE=seqname, RANGE=position,
BETX=real, ALFX=real, MUX=real,
BETY=real, ALFY=real, MUY=real,
X=real, PX=real, Y=real, PY=real,
DX=real, DY=real, DPX=real, DPY=real;
Here all linear lattice functions (BETX, BETY, ALFX, ALFY, MUX, MUY, DX, DY, DPX,
DPY) or chromatic lattice functions (WX, XY, PHIX, PHIY, DMUX, DUMY, DDX, DDY,
DDPX, DDPY) are constrained at the selected range to the corresponding values.
The second form of the CONSTRAINT command has the form
CONSTRAINT, SEQUENCE=seqname, RANGE=position,
BETA0=beta0-name, MUX=real, MUY=real;
Here all of (BETX, BETY, ALFX, ALFY, MUX, MUY, DX, DY, DPX, DPY) are constrained in
the selected points to the corresponding values of a pre-calculated SAVEBETA module. In the
above example the phases (MUX, MUY) are overridden by the numerical values specified via
MUX=real and MUY=real. Normally RANGE refers to a single position.
20.9 User Defined Matching Constraints
In addition to the nominal TWISS variables, the user can define a limited set of ’user-defined’
variables in the constraint statement. This allows, for example, the matching of the nor-
malized dispersion or the mechanical aperture. The MATCH module allows four user defined
variables called: MVAR1, MVAR2, MVAR3 and MVAR4. The variables can be defined according
to the general variable declaration rules of deferred expressions. For example, in order to
match the normalized dispersion at a certain location in the sequence one would first define
a variable:
MVAR1 := table(twiss,dx)/sqrt(table(twiss,betx));
After that the user has to select the variable for output in the TWISS statement (see details
in the TWISS module and SELECT statement):
SELECT, FLAG=twiss, CLEAR;
SELECT, FLAG=twiss, COLUMN=keyword, name, s, betx, dx, mvar1;
TWISS, SEQUENCE=seqname, FILE=twissfile;
The variable can now be referenced like any other TWISS variable in the constraint command:
CONSTRAINT, SEQUENCE=seqname, RANGE=range, MVAR1=targetvalue;
20.10 GLOBAL
In addition to conventional matching constraints that specify the optics functions at a certain
position in the sequence the user can also constrain global optics parameters such as, for ex-
![](asset-8e.png)
140 CHAPTER 20. MATCHING MODULE
ample, the overall machine tune and the machine chromaticity. Such global optics parameters
can be constraint via the GLOBAL command, having the following syntax:
GLOBAL, SEQUENCE=seqname,
Q1=real, DQ1=real,
Q2=real, DQ2=real;
All attributes are optional and have the following meaning:
SEQUENCE the name of the sequence on which to operate the matching.
Q1, Q2 enable a correction of tunes in presence of magnetic imperfections or misalign-
ments
DQ1, DQ2 enable a correction of chromaticities in presence of magnetic imperfections or
misalignments
20.11 WEIGHT, GWEIGHT
The matching procedures try to fulfil the constraints in a least square sense. A penalty
function is constructed which is the sum of the squares of all errors, each multiplied by the
specified weight. The larger the weight, the more important a constraint becomes. The
weights are taken from a table of current values. These are initially set to default values
listed in the table below, and may be reset at any time to different values. Values set in this
way remain valid until changed again.
The WEIGHT command changes the weights for subsequent constraints:
WEIGHT, BETX=real, ALFX=real, MUX=real,
BETY=real, ALFY=real, MUY=real,
X=real, PX=real, Y=real, PY=real,
DX=real, DPX=real, DY=real, DPY=real;
The weights are entered with the same name as the linear lattice functions and orbit coordinate
to which the weight applies. Frequently the matching weights serve to select a restricted set
of functions to be matched.
Matching weights for global matching constraints can be set by the GWEIGHT command, having
attributes identical to those of GLOBAL.
GWEIGHT, Q1=real, DQ1=real,
Q2=real, DQ2=real;
Default values for matching weights are given in the table below.
20.12 Matching Methods
MAD-X currently supports four different matching algorithms each associated to a command
with its own attributes.
![](asset-8f.png)
20.12. MATCHING METHODS 141
Table 20.1: Default Matching Weights
NAME WEIGHT NAME WEIGHT NAME WEIGHT
BETX 1.0 ALFX 10.0 MUX 10.0
BETY 1.0 ALFY 10.0 MUY 10.0
X 10.0 PX 100.0
Y 10.0 PY 100.0
T 10.0 PT 100.0
DX 10.0 DPX 100.0
DY 10.0 DPY 100.0
WX 1.0 PHIX 1.0 DMUX 1.0
WY 1.0 PHIY 1.0 DMUY 1.0
DDX 1.0 DDPX 1.0
DDY 1.0 DDPY 1.0
MVARi10.0
Q1 10.0 DQ1 1.0
Q2 10.0 DQ2 1.0
20.12.1 LMDIF: Fast Gradient Minimisation
The LMDIF command minimises the sum of squares of the constraint functions using their
numerical derivatives. It is the fastest minimisation method available in MAD-X.
LMDIF, CALLS=integer, TOLERANCE=real;
The command has two attributes:
CALLS The maximum number of calls to the penalty function. (Default: 1000)
TOLERANCE The desired tolerance for the minimum. (Default: 1E-6)
20.12.2 MIGRAD: Gradient Minimisation
The MIGRAD command minimises the penalty function using the numerical derivatives of the
sum of squares.
MIGRAD, CALLS=integer, TOLERANCE=real, STRATEGY=integer;
The command has three attributes:
CALLS the maximum number of calls to the penalty function. (Default: 1000)
TOLERANCE the desired tolerance for the minimum. (Default: 1E-6)
STRATEGY specifies the strategy to be used. (Default: 1)
Details are given in [15].
![](asset-90.png)
142 CHAPTER 20. MATCHING MODULE
20.12.3 SIMPLEX: Simplex Minimisation
The SIMPLEX command minimises the penalty function by the simplex method. Details are
given in [15].
SIMPLEX, CALLS=integer, TOLERANCE=real;
The command has two attributes:
CALLS The maximum number of calls to the penalty function. (Default: 1000)
TOLERANCE The desired tolerance for the minimum. (Default: 1E-6)
20.12.4 JACOBIAN: Newton Minimisation
The JACOBIAN command minimises the penalty function calculating the Jacobian and solving
the linear problem. A QR or LQ decomposition is performed when the system is over or
under-determined. Before starting the matching routine two optional transformations (COOL
and RANDOM) are performed.
JACOBIAN, CALLS=integer, TOLERANCE=real, REPEAT=integer,
STRATEGY=integer, COOL=real, BALANCE=real,
RANDOM=real;
The command has the folowing attributes:
CALLS The maximum number of calls to the penalty function. (Default: 30)
TOLERANCE The desired tolerance for the minimum. (Default: 1E-6)
REPEAT The number of calls of the JACOBIAN routine. (Default: 1)
BISEC Selects the maximum number of iteration used to determine the step length
which reduces the penalty function during the main iteration. A large number
(i.e. 6) reduces the probability to diverge from the solution, but increases the
probability of being trapped in a local minimum.
STRATEGY A code for the strategy to be used. (Default: 3)
If STRATEGY=1 the routine resets the values of the variables which exceeds
the limits. If STRATEGY=2 the routine print the Jacobian and exit without
matching. If STRATEGY=3 the routine disables the variables which exceeds
the limits keeping however the number of variables greater or equal to the
number of the constraints.
COOL, BALANCE The factors which specify the following transformation:
if "balance" >=0
newval = (1-cool)*oldval + &
cool*((1-balance)*maxval + balance*minval)
else
newval = (1-cool)*oldval + cool*optval
![](asset-91.png)
20.13. USE MACRO 143
where newval is the new value after the transformation, oldval is the previous
value, maxval, minval, optval are the maximum, minimum and optimal
value of the variable specified in the VARY command.
RANDOM The factors which specify the following transformation:
newval = ( 1 + random * rand() ) * oldval
where newval is the new value after the transformation, oldval is the previous
value, rand() is a stochastic variable with a uniform (-0.5,0.5) distribution.
20.13 USE MACRO
It is possible to match user defined expressions with the USE MACRO keyword.
The general input structure for a MATCH command is the following:
MATCH,USE_MACRO;
... VARY statements ...
USE_MACRO, NAME=macro1;
or
macro1: MACRO={ ... madx statements};
CONSTRAINT, expr= "lhs1 < | = | > rhs1";
CONSTRAINT, expr= "lhs2 < | = | > rhs2";
... CONSTRAINT statements ...
MACRO 2 definition
... CONSTRAINT statements ...
MACRO n definition
... CONSTRAINT statements ...
... METHODS statements ...
ENDMATCH;
The algorithm for evaluating the penalty function is the following:
•execute the first macro,
•evaluate and compute the difference between the left hand side (lhs) and the right hand
side (rhs) of the first set of expressions,
•in case of other macros, evaluates in order the macro and the expressions
•the set of differences are minimized by the selected method using the variables defined
in the VARY statements.
20.13.1 Initiating the Matching Module with USE MACRO
With:
MATCH, USE MACRO;
![](asset-92.png)
144 CHAPTER 20. MATCHING MODULE
the MATCH command can be used for matching any expression which can be defined through
expression. It requires a slightly different syntax.
20.13.2 VARY statements
In the USE MACRO mode the VARY statement follows the same rules of the other modes explained
in the section Define Variable Parameter
20.13.3 Macro definitions
The macro to be used in the matching routine can be defined in two ways:
•using USE MACRO statement:
USE MACRO, NAME=macro1;
defining a new macro on the fly using the usual syntax for MACROs.
After a macro definition a set of constraints should be defined, with the following syntax for
the CONSTRAINT command:
CONSTRAINT, expr = "lhs = rhs";
CONSTRAINT, expr = "lhs < rhs";
CONSTRAINT, expr = "lhs > rhs";
where ”lhs” and ”rhs” are well defined MAD-X expressions. Another set of macro and con-
straints can be defined afterwards.
20.13.4 Examples
The USE MACRO mode can emulate a matching script which uses the normal syntax.
Normal syntax:
MATCH,SEQUENCE=LHCB1,LHCB2;
VARY, NAME=KSF.B1, STEP=0.00001;
VARY, NAME=KSD.B1, STEP=0.00001;
VARY, NAME=KSF.B2, STEP=0.00001;
VARY, NAME=KSD.B2, STEP=0.00001;
GLOBAL,SEQUENCE=LHCB1,DQ1=QPRIME;
GLOBAL,SEQUENCE=LHCB1,DQ2=QPRIME;
GLOBAL,SEQUENCE=LHCB2,DQ1=QPRIME;
GLOBAL,SEQUENCE=LHCB2,DQ2=QPRIME;
LMDIF, CALLS=10, TOLERANCE=1.0E-21;
ENDMATCH;
USE MACRO syntax:
20.14. MATCHING EXAMPLES 145
MATCH,USE_MACRO;
VARY, NAME=KSF.B1, STEP=0.00001;
VARY, NAME=KSD.B1, STEP=0.00001;
VARY, NAME=KSF.B2, STEP=0.00001;
VARY, NAME=KSD.B2, STEP=0.00001;
M1: MACRO={ TWISS,SEQUENCE=LHCB1; };
CONSTRAINT, EXPR= TABLE(SUMM,DQ1)=QPRIME;
CONSTRAINT, EXPR= TABLE(SUMM,DQ2)=QPRIME;
M2: MACRO={ TWISS,SEQUENCE=LHCB2; };
CONSTRAINT, EXPR= TABLE(SUMM,DQ1)=QPRIME;
CONSTRAINT, EXPR= TABLE(SUMM,DQ2)=QPRIME;
LMDIF, CALLS=10, TOLERANCE=1.0E-21;
ENDMATCH;
20.14 Matching Examples
All matching examples and the related files for executing the MAD-X sample jobs can be found
in the examples directory (http://cern.ch/madx/madX/examples) of the MAD-X webiste.
•Simple Periodic Cell
Match a simple cell to given phase advances: FIVE-CELL
•Simple Periodic Cell
Match the matrix elements of the linear transfer matrix at the end of a sequence 5
periodic cells: RMATRIX
•Transfer line with initial conditions
Match a sequence of 5 periodic cells with initial conditions to given beta-functions at
the end of the sequence: Transfer line
•Global tune matching in a sequence of 5 periodic cells
Match the global tune of a sequence of 5 periodic cells: Global tune
•Global tune matching for the LHC
Match the global tune for beam1 of the LHC: Global tune for the LHC
•Global chromaticity matching for the LHC
Match the global chromaticity for beam1 of the LHC: Global chromaticity for the LHC
•Global chromaticity matching for both beams of the LHC
Match the global chromaticity for beam1 and beam2 of the LHC: Global chromaticity
for both beams of the LHC
•IR8 insertion matching for beam1 of the LHC
Match the insertion IR8 with initial conditions to given values of the optics functions
at the IP and the end of the insertion: IR8 insertion matching for beam1 of the LHC
•IR8 insertion matching for beam1 of the LHC with upper limits on the optics functions
Match the insertion IR8 with initial conditions to given values of the optics functions
at the IP and the end of the insertion while limiting the maximum acceptable beta
![](asset-94.png)
146 CHAPTER 20. MATCHING MODULE
functions over the whole insertion: IR8 insertion matching for beam1 of the LHC with
upper limits for all beta functions inside the insertion
•Simultaneous orbit matching at IP8 for beam1 and beam2 of the LHC
Match simultaneously the orbit of beam1 and beam of the LHC at IP8 with initial
conditions to the same given values at the IP: Orbit matching at IP8 for beam1 and
beam2 of the LHC
•IR8 beta squeeze for beam1 using JACOBIAN matching routine
Try to find a beta squeeze for IR8 starting from 10 meters. Beta squeeze for IR8
•Matching first and second order chromaticity of the LHC using USE MACRO option.
Match simultaneously the first and second order chromaticity by defining macros which
compute them using the TWISS command or PTC. Second order chromaticity
•Matching s position using VLENGTH flag.
match the positions of elements and the total sequence length for a simple sample
sequence. s position matching
•Matching s position using USE MACRO.
match the positions of elements and the total sequence length for a simple sample
sequence using USE MACRO. s position matching
![](asset-95.png)
Chapter 21. EMIT: Equilibrium emittances
EMIT calculates the equilibrium emittances:
EMIT, DELTAP=real, TOL=real;
The attributes for the EMIT command are:
DELTAP the average energy error.
EMIT adjusts the RF frequencies in order to obtain this average energy error:
the revolution frequency f0is determined for a fictitious particle with constant
momentum error DELTAP =δs=δ(E)/psctravelling along the design orbit.
The RF frequencies are then set to fRF =hf0.
TOL The tolerance attribute is for the distinction between static and dynamic cases:
if for the eigenvalues of the one-turn matrix, |e val 5|<tol and |e val 6|<tol,
then the longitudinal motion is not considered, otherwise it is.
(Default: 1.000001)
If the machine contains at least one RF cavity, and if synchrotron radiation is enabled with
BEAM, RADIATE=true;, the EMIT command computes the equilibrium emittances and other
electron beam parameters using the method in [16].
In this calculation the effects of quadrupoles, sextupoles and octupoles along the closed orbit
are also considered. Thin multipoles are used only if they have a fictitious length LRAD
different from zero.
If the machine does not contain any RF cavity, if synchrotron radiation is turned off (BEAM,
RADIATE=false;, or if the longitudinal motion is not stable, EMIT only computes the param-
eters that are not related to radiation and does not update the BEAM values.
If synchrotron radiation is enabled (BEAM, RADIATE=true;, and the DELTAP attribute is zero,
and the longitudinal motion is stable, EMIT calculates and updates the following values for the
BEAM attached to the current sequence: both geometric and normalized transverse emittances,
longitudinal emittance and beam sizes (σEand σt), damping partition numbers, energy loss
per turn and synchrotron tune.
Example:
RFC: RFCAVITY, HARMON..., VOLT=...;
...
BEAM, ENERGY = 100.0, RADIATE=true;
EMIT, DELTAP = 0.01;
Remark:
This module assumes nearly constant lattice functions inside elements. This assumption
works for many machines, like LEP but it fails when the lattice functions largely vary inside
single elements. In the later case it is advised to slice the elements.
147
![](asset-96.png)
Chapter 22. Physical Aperture
Physical apertures can be defined and associated to most elements in MAD-X.
The APERTURE command calculates the beam stay clear values (n1 values).
During tracking the particle excursion can be checked against the available aperture, and the
particle lost if it falls outside the defined aperture.
22.1 Aperture definition
The aperture for a particular element or class of elements can be set in MAD-X at the time of
definition or instantiation of the element or class.
Note that instantiation can also happen at the time of declaration of an element in a sequence.
The aperture can be specified for any element or class of elements, with the exception of drift
spaces.
The definition of the aperture takes the following form and parameters:
..., APERTYPE=string, APERTURE={values},
APER OFFSET={values}, APER TOL={values},...
where the four specific attributes are attributes of an element declaration or instantiation.
The minimum aperture definition uses the following two attributes:
APERTYPE defines the aperture type from a set of preselected types, or from a file if a
filename is provided as argument. The preselected types are:
CIRCLE
RECTANGLE
ELLIPSE
RECTCIRCLE a superposition of a CIRCLE and a RECTANGLE
LHCSCREEN an alias for RECTCIRCLE
RECTELLIPSE a superposition or intersection of an ELLIPSE and a RECTANGLE
RACETRACK the union of an ELLIPSE and a RECTANGLE forming a RACETRACK
shape with four quarters of ellipse, one per quadrant, connected
by straight lines
OCTAGON a simply connex octagon
If a filename is provided as argument, the file must contain a list of x and y
coordinates, one pair per line, outlining the complete aperture shape, i.e. no
symmetry is assumed. This option is only supported by the APERTURE module
148
![](asset-97.png)
22.1. APERTURE DEFINITION 149
and an aperture thus defined is ignored in the TRACK module. When a filename
is provided as argument, the APERTURE attribute is ignored.
APERTURE is an array of values, the number and meaning of which depends on the
APERTYPE.
Table 22.1: Predefined aperture types
APERTYPE # of values meaning of values
CIRCLE 1 radius of circle
RECTANGLE 2 half width and half height of rectangle
ELLIPSE 2 horizontal and vertical semi-axes of ellipse
RECTCIRCLE or LHCSCREEN 3 half width and half height of rectangle, radius
of circle
RECTELLIPSE 4 half width and half height of rectangle, hori-
zontal and vertical semi-axes of ellipse
RACETRACK 4 half width and half height along main axes,
and horizontal and vertical semi-axes of el-
lipse for the rounding of the corners
OCTAGON 4 half width and half height along main axes,
two angles sustaining the cut corner in the
first quadrant, given in radians and in order
of increasing values.
Examples
The following statements are equivalent in setting an elliptic aperture using the variables APH
and APV:
..., APERTYPE=ellipse, APERTURE={APH, APV};
..., APERTYPE=rectellipse, APERTURE={APH, APV, APH, APV};
..., APERTYPE=racetrack, APERTURE={APH, APV, APH, APV};
Similarly the following statements are equivalent in setting a rectangular aperture using the
variables APX and APY, and APR:=SQRT(APX**2+APY**2):
..., APERTYPE=rectangle, APERTURE={APX,APY};
..., APERTYPE=rectellipse, APERTURE={APX, APY, APR, APR};
..., APERTYPE=racetrack, APERTURE={APX, APY, 0.0, 0.0};
When the aperture is described in a file, the only parameter to be provided is a filename given
as value to the APERTYPE attribute:
MB: SBEND, L=l.MB, APERTYPE="MB-aper.txt";
where "MB-aper.txt" contains the list of x-y coordinates defining the aperture shape:
x0 y0
xi yi
![](asset-98.png)
150 CHAPTER 22. PHYSICAL APERTURE
...
xn yn
Notes
•There are inconsistencies in the parameter definition for the different aperture types.
This is historical and is kept for backwards compatibility. Pay some attention to the
parameters you introduce!
•The RECTELLIPSE and RACETRACK shapes are fundamental in the sense that both can
properly model the other CIRCLE,ELLIPSE and RECTANGLE shapes, and that RECTCIRCLE
and LHCSCREEN shapes are a particular of RECTELLIPSE
•It is considered good practice to always set the first two APERTURE parameters for
predefined shapes to be the maximum horizontal and vertical extents. This proves
useful when trying to plot horizontal and vertical beam stay-clear. For example, it is
better to use
..., APERTYPE=rectellipse, APERTURE=APX, APY, APX, APY;
rather than
..., APERTYPE=rectellipse, APERTURE=100, 100, APX, APY;
•When MAKETHIN is called all thin slices inherit the aperture of their respective original
thick lens version.
•When the SIXTRACK command is called (see the SixTrack converter module) the aper-
tures are ignored by default. To convert the apertures as well the APERTURE flag has to
be set in C6T.
•Aperture parameters are like all parameters and are inherited by derived elements. Like
other parameters they can also be overridden by the derived elements if necessary.
•The APERTYPE and the APERTUREs themselves can be conveniently added to the TWISS
(chapter 19) table by using the SELECT command (section 4.11). For example the
command:
Select, flag=twiss, clear;
Select, flag=twiss, column=name,s,betx,alfx,mux,bety,alfy,muy,
apertype,aper 1,aper 2;
and a subsequent TWISS command put the aperture information together with the spec-
ified TWISS parameters into the TWISS table.
•The minimum beam stay-clear value (n1 value) computed along the beamline is printed
to standard out and is also available as an output variable N1MIN of the BEAM command:
Value, beam->n1min;
nmin = beam->n1min;
![](asset-99.png)
22.2. APERTURE TOLERANCE DEFINITION 151
22.2 Aperture tolerance definition
A parameter closely connected to the aperture is the sum of the mechanical and alignment
tolerances. The mechanical tolerance is the maximal error margin of errors in the element
body which causes a decrease of aperture, and the alignment tolerance is a misalignment of
the element in the accelerator, which also causes a decrease of aperture. The tolerance is
given in the transverse plane as a racetrack, like in picture 22.1 below.
Figure 22.1: Definition of aperture tolerances
A tolerance can be assigned to each element in a MAD-X sequence as a vector:
APER TOL = {r, g, s};
Example:
MB: SBEND, L=l.MB, APER TOL={1.5, 1.1, 0};
22.3 Aperture offset definition
An aperture offset can be assigned to each element in a MAD-X sequence as a vector:
APER OFFSET = {real, real};
where the two real values are respectively the horizontal and vertical offsets of the aperture
inside the element.
The offsets are only used in the tracking module of MAD-X and are ignored by the APERTURE
command and by PTC TRACK.
22.4 APERTURE
The APERTURE module was developed specifically for the LHC.
Default parameter values are LHC values.
![](asset-9a.png)
152 CHAPTER 22. PHYSICAL APERTURE
The APERTURE module computes the n1 values for a piece of machine. Each element is sliced
into thick sub-elements at given intervals, and the available aperture is computed at the end
of each slice. The APERTURE module gets the geometric emittances from the values given
or calculated in the BEAM command. The computation is based on the last Twiss table
computed by the TWISS command. It is important to properly define a BEAM, and run TWISS
and APERTURE commands on the same period or sequence.
The APERTURE example below also shows how to plot the resulting n1 values.
The minimum n1 value is written to the last Twiss table, to allow for matching by aperture.
APERTURE, RANGE=range,
DQF=real, DPARX=real, DPARY=real,
BETAQFX=real, BBEAT=real, DP=real,
COR=real, NCO=integer,
HALO={real,real,real,real}, HALOFILE=filename,
INTERVAL=real, SPEC=real, NOTSIMPLE=logical,
TRUEPROFILE=filename, OFFSETELEM=filename,
FILE=filename;
where the parameters have the following meaning:
RANGE Range given by elements. Default = #s/#e
DQF Peak linear dispersion [m]. Default = 2.086
DPARX Fractional horizontal parasitic dispersion. Default = 0.273
DPARY Fractional vertical parasitic dispersion. Default = 0.273
BETAQFX Beta x in standard qf [m]. Default = 170.25
BBEAT Beta beating coefficient applying to beam size. Default = 1.1
DP Bucket edge at the current beam energy. Default = 0.0015
COR Maximum radial closed orbit uncertainty [m]. Default = 0.004
NCO Number of azimuth per quadrant for halo radial scan. Default = 5
HALO Halo parameters: {n, r, h, v}. n is the radius of the primary halo, r is the
radial part of the secondary halo, h and v is the horizontal and vertical cuts
in the secondary halo. Default = {6, 8.4, 7.3, 7.3}
HALOFILE Input file with halo polygon coordinates. Will suppress an eventual halo pa-
rameter. Default = none
INTERVAL Approximate length in meters between measurements. Actual value: nslice
= nodelength/interval, nslice is rounded down to closest integer, interval =
nodelength/nslice. Default = 1.0
SPEC Aperture spec, for plotting only. Gives the spec line in the plot. Default = 0.0
NOTSIMPLE Use only if one or more beam-screens in the range are considered not to be a
”simply connex”. Since all predefined MAD-X aperture types are simply connex,
![](asset-9b.png)
22.5. NOT SIMPLY CONNEX BEAM PIPE PROFILES 153
this is only possible if an input file with beam screen coordinates are given.
See below for a graphical example. Default = false.
TRUEPROFILE A file containing a list of magnets, and for each magnet a list of horizontal and
vertical deviations from the ideal magnet axis. These values may come from
measurements done on the magnet. See below for example. Default = none.
OFFSETELEM A file containing a reference point in the machine, and a list of magnets with
their offsets from this point described as a parabola. See below for example.
Default = none.
Note that the reference point should be within the range of elements
given for the offsets to be taken into account.
FILE Output file with aperture table. Default = none
22.5 Not simply connex beam pipe profiles
The algorithm for finding the largest possible halo is the following:
The distance from halo centre to the first apex (i = 0) in the halo is calculated (l i), and the
equation for a line going through these points is derived. This line is then compared with all
lines making the pipe polygon to find their respective intersection coordinates. The distance
h i between halo centre and intersection are then divided by l i, to find the maximal ratio of
enlargement, as seen in figure 22.2 below. This procedure is then repeated for all apexes i in
the halo polygon, and the smallest ratio of all apexes is the maximal enlargement ratio for
this halo to just touch the pipe at this particular longitudinal position.
Figure 22.2: Determination of maximum halo size
There is one complication to this solution; polygons which are not simple connexes. (Geo-
metrical definition of “simply connex”: A figure in which any two points can be connected
by a line segment, with all points on the segment inside the figure.) The figure 22.3 below
![](asset-9c.png)
154 CHAPTER 22. PHYSICAL APERTURE
shows what happens when a beam pipe polygon is not a simple connex. The halo is expanded
in such a way that it overlaps the external polygon in the area where the latter is dented
inwards.
Figure 22.3: Not connex beam pipe profile: problem
To make the module able to treat all sorts of polygons, the logical attribute NOTSIMPLE must
be specified. With this option activated, apexes are strategically added to the halo polygon
wherever the beam pipe polygon might have an inward dent. This is done by drawing a line
from halo centre to each apex on the pipe polygon. An apex with its coordinates on the
intersection point line-halo is added to a table of halo polygon apexes. The result is that the
halo polygon has a few “excessive” points on straight sections, but the algorithm used for
expansion will henceforth not miss a dent in the beam pipe. See figure 22.4. The use of the
notsimple option significantly increases computation time.
22.6 Trueprofile file syntax
This file contains magnet names, and their associated displacements of the axis for an arbitrary
number of S, where So is the start of the magnet and Sn the end. The interval between each S
must be regular, and X and Y must be given in meters. The magnet name must be identical
to how it appears in the sequence. The displacements are only valid for this particular
magnet, and cannot be assigned to a family of magnets. n1 is calculated for a number of
slices determinated by the number of Si.
Layout of file:
magnet.name1
![](asset-9d.png)
22.6. TRUEPROFILE FILE SYNTAX 155
Figure 22.4: Not connex beam pipe profile: solution
So X Y
Si X Y
Si X Y
Sn X Y
magnet.name2
So X Y
Si X Y
Si X Y
Sn X Y
etc.
Example of file:
mb.a14r1.b1
0 0.0002 0.000004
7.15 1.4e-5 0.3e-3
14.3 0.0000000032 4e-6
mq.22r1.b1
0 0.3e-5 1.332e-4
1.033 0.00034 0.3e-9
2.066 0 0.00e-2
3.1 4.232e-4 0.00000003
![](asset-9e.png)
156 CHAPTER 22. PHYSICAL APERTURE
22.7 Offsetelem file syntax
This file contains parameters describing how certain elements are actually located in space
with respect to a given reference element in the machine.
The basis for this file is a pair of coordinate systems, {s,x}and {s,y}with the origin at the
reference point. The units for s, x and y are meters.
The actual location of the magnetic axis of a given element can be described, in each plane,
as a parabola defined with three coefficients:
X act(s) = DDX OFF * s**2 + DX OFF * s + X OFF
Y act(s) = DDY OFF * s**2 + DY OFF * s + Y OFF
The reference position for the element, X ref(s) and Y ref(s), is calculated by MAD-X via an
internal call to the SURVEY module, taking the reference element as the origin.
The resulting offset, in each plane, which is taken into account in the aperture calculation, is
the difference between reference position and actual position:
X offset(s) = X ref(s) - X act(s)
Y offset(s) = Y ref(s) - Y act(s)
The offsets are only calculated for elements for which actual positions have been given through
the OFFSETELEM file mechanism.
The file must be given in TFS format according to the following template, with mandatory
header and any number of data lines, one per element.
@ NAME %06s "OFFSET"
@ TYPE %06s "OFFSET"
@ REFERENCE %10s "reference-element-name"
* NAME S_IP X_OFF DX_OFF DDX_OFF Y_OFF DY_OFF DDY_OFF
$ %s %le %le %le %le %le %le %le
"elementname" real real real real real real real
Note that the column S IP is actually not used, and the values are ignored, but the column
and values are parsed nevertheless and must be present. The absence of this column will
trigger an error.
Example:
@ NAME %06s "OFFSET"
@ TYPE %06s "OFFSET"
@ REFERENCE %10s "mq.12r1.b1"
* NAME S_IP X_OFF DX_OFF DDX_OFF Y_OFF DY_OFF DDY_OFF
$ %s %le %le %le %le %le %le %le
"mq.12r1.b1" 0.0 -3.0 -2.56545 0.0 0.0 -2.344366 0.0
"mcbxa.3r2" 0.0 -0.84 32.443355 0.3323 0.0 32.554363 0.2522
As an example we see in the picture below how the horizontal axes of elements m1 and m2 do
not coincide with the reference trajectory.
Note that prior to MAD-X version 4, the layout of the file was different and not formatted as
TFS file:
![](asset-9f.png)
22.7. OFFSETELEM FILE SYNTAX 157
Figure 22.5: Illustration of effect of OFFSETELEM
reference-element-name
elementname
DDX OFF DX OFF X OFF
DDY OFF DY OFF Y OFF
Example:
! Reference point
mq.12r1.b1
! List of elements and their displacement w.r.t. reference point.
mcbxa.3l2
0 -2.5654500 -3
0 -2.3443666 0
! The next element uses the same reference point.
! Elements offset w.r.t. another point must be given in another file,
! together with the new reference point.
mcbxa.3r2
0.3323 32.443355 -0.84
0.2522 32.554363 0.0
![](asset-a0.png)
158 CHAPTER 22. PHYSICAL APERTURE
22.8 Aperture command example
The aperture module needs a Twiss table to operate on. It is important not to USE another
period or sequence between the Twiss and aperture module calls, else aperture looses its table.
One can choose the ranges for Twiss and aperture freely, they need not be the same.
use, period=lhcb1;
select, flag=twiss,range=mb.a14r1.b1/mb.a17r1.b1,column=keyword,name,
parent,k0l,k1l,s,betx,bety,n1;
twiss, file=twiss.b1.data, betx=beta.ip1, bety=beta.ip1, x=+x.ip1,
y=+y.ip1, py=+py.ip1;
plot,haxis=s,vaxis=betx,bety,colour=100;
select, flag=aperture, column=name,n1,x,dy;
aperture, range=mb.b14r1.b1/mb.a17r1.b1, spec=5.235;
plot,table=aperture,noline,vmin=0,vmax=10,haxis=s,vaxis=n1,spec,
on_elem,style=100;
The SELECT command can be used to choose which columns to print in the output file.
Column names: name, n1, n1x m, n1y m, apertype, aper 1, aper 2, aper 3, aper 4,
rtol, xtol, ytol, s, betx, bety, dx, dy, x, y, on ap, on elem, spec
n1 is the maximum beam size in sigma, while n1x m and n1y m is the n1 values in si-units
in the x- and y-direction.
Note that specifying the apertype column automatically selects also the aper 1, aper 2,
aper 3 and aper 4 columns. The statement
SELECT, FLAG=aperture, COLUMN=apertype;
is equivalent to
SELECT, FLAG=aperture, COLUMN=apertype, aper 1, aper 2, aper 3, aper 4;
aper # means for all apertypes but racetrack:
aper 1 = half width rectangle
aper 2 = half heigth rectangle
aper 3 = half horizontal axis ellipse (or radius if circle)
aper 4 = half vertical axis ellipse
For racetrack, the aperture parameters will have the same meaning as the tolerances:
aper 1 and xtol = horizontal displacement of radial part
aper 2 and ytol = vertical displacement of radial part
aper 3 and rtol = radius
aper 4 = not used
ON ELEM indicates whether the node is an element or a drift, and ON AP whether it has a valid
aperture. The Twiss parameters are the interpolated values used for aperture computation.
When one wants to plot the aperture, the TABLE=aperture parameter is necessary. The
normal line of hardware symbols along the top is not compatible with the aperture table,
so it is best to include NOLINE. Plot instead the column ON ELEM along the VAXIS to have
![](asset-a1.png)
22.8. APERTURE COMMAND EXAMPLE 159
a simple picture of the hardware. SPEC can be used for giving a limit value for n1, to have
something to compare with on the plot. The following plot shows the n1, beta functions, and
the hardware symbolized by ON ELEM.
Figure 22.6: Example of plot showing aperture limits and n1
![](asset-a2.png)
Chapter 23. Slicing a sequence into thin lenses
This module takes a sequence with thick elements and creates another one with similar func-
tionality but composed of thin (zero length) element slices or simplified thick slices as required
by MAD-X tracking or conversion to SIXTRACK input format.
23.1 MAKETHIN
The slicing is performed by the command:
MAKETHIN, SEQUENCE=seqname, STYLE=string, MAKEDIPEDGE=logical;
The parameters are defined as:
SEQUENCE seqname is the name of the sequence to be processed to thin slices. The
sequence must be active, i.e. it should have been previously loaded with a
USE command. The sequence must use the default positioning of elements
(REFER=centre).
STYLE the slicing style to be used for all elements.
Available slicing styles are:
TEAPOT is the default slicing algorithm for all elements except collima-
tors and is described in [17]. TEAPOT is a clear improvement
over the SIMPLE algorithm.
SIMPLE produces slices of equal strength and length at equidistant po-
sitions.
COLLIM is the default slicing algorithm for collimators. If only one slice
is chosen it is placed in the middle of the old element. If two
slices are chosen they are placed at either end. Three slices or
more are treated as one slice.
HYBRID is the previous default algorithm that was active up to version
5.02.05 when the STYLE attribute was not given; the standard
slicing for all elements, except collimators, is TEAPOT if the num-
ber of slices is less or equal to four, and SIMPLE otherwise.
MAKEDIPEDGE is a logical flag to control the generation of DIPEDGE elements at the start
and/or end of bending magnets, to conserve edge focusing from pole face angles
E1, E2 or extra fields described by FINT, FINTX, in the process of slicing
bending magnets to thin multipole slices. Selection with THICK=true will
translate a complex thick RBEND or SBEND, including edge effects, to a simple
thick SBEND with edge focusing transferred to extra DIPEDGE elements.
(Default: true)
Example:
160
23.2. CONTROLLING THE NUMBER OF SLICES 161
! keep translated rbend as thick sbend
SELECT, FLAG=makethin, CLASS=rbend, THICK=true;
23.2 Controlling the number of slices
The number of slices can be set individually for elements or groups of elements using the
SELECT command
SELECT, FLAG=makethin,
RANGE=range, CLASS=class, PATTERN=pattern[,FULL][,CLEAR],
THICK=logical, SLICE=integer;
where the argument to the attribute SLICE stands for the number of slices for the selected
elements. The default is one slice and THICK=false for all elements, i.e. conversion of all
thick elements to a single thin slice positioned at the centre of the original thick element.
Note that THICK=true only applies to dipole or quadrupole magnetic elements and is ignored
otherwise.
MAKETHIN allows for thick quadrupole slicing with insertion of thin MULTIPOLE elements be-
tween thick slices. Positioning is done with markers between slices, here however with thick
slice quadrupole piece filling the whole length.
Examples:
! slice quadrupoles in three thick slices, insert 2 markers per quadrupole
SELECT, FLAG=makethin, CLASS=quadrupole, THICK=true, SLICE=3;
! thick slicing for quadrupoles named mqxa, insert one marker in the middle
SELECT, FLAG=makethin, PATTERN=mqxa\., THICK=true, SLICE=2;
Slicing can be turned off for certain elements or classes by specifying a number of slices <1.
Examples:
! turn off slicing for sextupoles
SELECT, FLAG=makethin, CLASS=sextupole, SLICE=0;
! keep elements unchanged with names starting by mbxw
SELECT, FLAG=makethin, PATTERN=mbxw\., SLICE=0;
This option allows to introduce slicing step by step and monitor the resulting changes in
optics parameters.
Keep in mind however that subsequent tracking generally requires full slicing, with possible
exception of dipole and quadrupole magnetic elements.
![](asset-a4.png)
162 CHAPTER 23. SLICING A SEQUENCE INTO THIN LENSES
23.3 Choice of options for dipoles
There are several options that affect the slicing of a sequence. Depending whether dipole
magnets (RBEND or SBEND) are kept as thick elements (See SELECT, FLAG=makethin) and
whether the MAKEDIPEDGE option of MAKETHIN is used to generate DIPEDGE elements around
these bending magnets, the resulting sequence is more adapted to optics calculation or track-
ing. The following table gives an indication of the best choice of options for dipole elements
Table 23.1: Best choice of options in MAKETHIN
THICK=false THICK=true
MAKEDIPEDGE=false backward compatibility thick optics calculations
MAKEDIPEDGE=true thin tracking thick tracking
The combination of THICK=false (dipoles converted to thin lenses) and MAKEDIPEDGE=false
(no DIPEDGE thin elements inserted around the dipoles) results in a lattice with Twiss param-
eters that can differ significantly from those of the original sequence. This is most obvious
with the tune or phase advance parameters that are generally differing by a significant amount
(see first hint below). This combination is however useful for backward compatibility with
previous versions of MAKETHIN before DIPEDGE elements were implemented.
The combination of THICK=true (dipoles are kept as thick elements and not converted to thin
lenses) and MAKEDIPEDGE=false (DIPEDGE thin elements are not inserted around the thick
dipoles to replace the edge effects that stay with the bends themselves) results in a lattice
with TWISS parameters very comparable to those of the original sequence. The dipoles are
still thick elements, eventually sliced, with proper edge effects up to second order, although
non-symplectic. This is the best combination for optics studies without particle tracking, e.g.
with TWISS.
The combination of THICK=false for dipoles and MAKEDIPEDGE=true transforms the original
dipoles into one or several thin slices without any edge effect, surrounded by a pair of DIPEDGE
elements at the original location of dipole entrance and exit. Note however that these DIPEDGE
elements only contain first order effects for the purpose of tracking.
The combination of THICK=true for dipoles and MAKEDIPEDGE=true conserves the dipoles as
thick element dipole bodies only while their associated edge effects are transferred to DIPEDGE
elements that are taken into account at first order only for symplectic tracking.
23.4 Additional information
The generated thin lens sequence has the following properties:
•The new sequence has the same name as the original. The original sequence is replaced
by the new one in memory. If the original sequence is needed for further processing in
MAD-X, it should be reloaded.
•The algorithm also processes any sub-sequence inserted in the main sequence. These
sub-sequences are also given the same names as the original ones.
![](asset-a5.png)
23.4. ADDITIONAL INFORMATION 163
•Any element transformed into a single thin lens element has the same name as the
original.
•If an element is sliced into more than one slice, the individual slices have the same
basename as the original element plus a suffix ..1,..2, etc. and a marker, with the
name of the original element, is placed at the location of the center of the original
element.
Hints
1. Compare the main optics parameters like tunes before and after slicing with MAKETHIN.
Rematch tunes and chromaticity as necessary after MAKETHIN.
2. In tests, turn off slicing for some of the main element classes to identify the main sources
of changes.
3. For sextupoles and octupoles, a single slice should always be sufficient.
4. Increase the number of slices for critical elements like mini-beta quadrupoles. Even
there, more than four slices should rarely be required.
5. In case of problems or doubts, consider to FLATTEN the sequence before slicing.
6. See the examples for makethin.
See also the presentations on the upgrade of the makethin module:
LCU makethin 2012 09 18.pdf, and
LCU makethin 2013 04 19.pdf.
TEAPOT is documented in IPAC’13 MOPWO027
![](asset-a6.png)
Chapter 24. Error Definitions
This chapter describes the commands which provide error assignment and output of errors
assigned to elements. It is possible to assign alignment errors and field errors to single beam
elements or to ranges or classes of beam elements.
Elements, classes or ranges of elements are selected by the SELECT command.
ATTENTION: since errors can only be assigned to machine elements, it is necessary to
FLATTEN a sequence if it includes other sequences.
Errors can be specified with both constant or random values.
Error definitions consist of four types of statements listed below. They may be entered after
having selected a beam line by means of a USE command.
WARNING: any further USE command will destroy the assigned errors. Use the ESAVE option
to save and reload errors.
24.1 EALIGN: Alignment Errors
Alignment errors are defined by the EALIGN command. The misalignments refer to the local
reference system for a perfectly aligned machine. Misalignments are defined as displacements
along the three coordinate axes, and rotations about the coordinate axes. Alignment errors
can be assigned to all beam elements except drift spaces. The effect of misalignments is
treated in a linear approximation.
Beam Position Monitors can be given readout errors as well as readout scaling errors in both
horizontal and vertical planes. Monitor readout and scaling errors are ignored for all elements
other than monitors.
Each new EALIGN statement replaces the misalignment errors for all elements in its range,
unless the logical ADD attribute of EOPTION has been specified.
Alignment errors are defined by the statement
SELECT, FLAG=ERROR, RANGE=range, CLASS=name, PATTERN=string;
EALIGN, DX=real, DY=real, DS=real,
DPHI=real, DTHETA=real, DPSI=real,
MREX=real, MREY=real,
MSCALX=real, MSCALY=real,
AREX=real, AREY=real;
for elements selected by the SELECT command.
The attributes are:
DX The misalignment in the x-direction for the entry of the beam element. (De-
fault: 0 m).
DX>0 displaces the element in the positive x-direction
164
![](asset-a7.png)
24.1. EALIGN: ALIGNMENT ERRORS 165
original
beam line s
y
x
original entrance
of the magnet DS
DX
DTHETA
Figure 24.1: Alignment errors in the (x, s)-plane
horizontal
plane
x×s
y
DPSI
ROT
Figure 24.2: Alignment errors in the (x, y)-plane
DY The misalignment in the y-direction for the entry of the beam element. (De-
fault: 0 m).
DY>0 displaces the element in the positive y-direction
DS The misalignment in the s-direction for the entry of the beam element. (De-
fault: 0 m).
DS>0 displaces the element in the positive s-direction
DPHI The rotation around the x-axis. (Default: 0 rad).
A positive angle gives a greater y-coordinate for the exit than for the entry.
DTHETA The rotation around the y-axis according to the right hand rule. (Default: 0 rad).
DPSI The rotation around the s-axis according to the right hand rule. (Default: 0 rad).
MREX The horizontal read error for a monitor. This is ignored if the element is not
a monitor.
If MREX>0 the reading for xis too high (default: 0 m).
MREY The vertical read error for a monitor. This is ignored if the element is not a
monitor.
If MREY>0, the reading for yis too high (default: 0 m).
![](asset-a8.png)
166 CHAPTER 24. ERROR DEFINITIONS
original
beam line s
×
x
y
original entrance
of the magnet DS
DY
DPHI
Figure 24.3: Alignment errors in the (y, s)-plane
MSCALX The relative horizontal scaling error for a monitor. This is ignored if the
element is not a monitor.
If MSCALX>0 the reading for xis too high (default: 0). A value of 0.5 implies
that the actual reading is multiplied by 1.5.
MSCALY The relative vertical scaling error for a monitor. This is ignored if the element
is not a monitor.
If MSCALY>0 the reading for yis too high (default: 0). A value of -0.3 implies
that the actual reading is multiplied by 0.7.
AREX The misalignment in the x-direction for the entry of an aperture limit (default:
0 m).
AREX>0 displaces the element in the positive x-direction
AREY The misalignment in the y-direction for the entry of an aperture limit (default:
0 m).
AREY>0 displaces the element in the positive y-direction
horizontal
plane
x×
true beam
position
s
y
beam position given
by the monitor
MREX
MREY
Figure 24.4: Readout errors in a monitor
Examples:
SELECT, FLAG = ERROR, CLASS = MQ;
![](asset-a9.png)
24.2. EFCOMP: FIELD ERRORS 167
EALIGN, DX = 0.002, DY = 0.0004*RANF(), DPHI = 0.0002*GAUSS();
Assigns alignment errors to all elements of class MQ.
SELECT, FLAG = ERROR, PATTERN = "QF.*";
EALIGN, DX = 0.001*TGAUSS(2.5), DY = 0.0001*RANF();
Assigns alignment errors to all elements starting with "QF".TGAUSS(2.5) specifies a Gaus-
sian distribution cut at 2.5 sigma.
24.2 EFCOMP: Field Errors
Field errors can be entered as relative or absolute errors. Different multipole components can
be specified with different kinds of errors (relative or absolute). Relations between absolute
and relative field errors are listed below.
In MAD-8 two commands were used for that purpose: EFIELD and EFCOMP. Only EFCOMP was
implemented in MAD-X since it provides the full functionality of EFIELD and there was no need
for duplication.
All field errors are specified as the integrated value RKds of the field components along the
magnet axis in m−i. There is no provision to specify a global relative excitation error affecting
all field components in a combined function magnet. Such an error may only be entered by
defining the same relative error for all field components.
Field errors can be specified for all magnetic elements by the statement
SELECT, FLAG=ERROR, RANGE=range, CLASS=name, PATTERN=string;
EFCOMP, ORDER=integer, RADIUS=real,
DKN= {dkn(0), dkn(1), dkn(2),...},
DKS= {dks(0), dks(1), dks(2),...},
DKNR= {dknr(0), dknr(1), dknr(2),...},
DKSR= {dksr(0), dksr(1), dksr(2),...};
for elements selected by the SELECT command.
Each new EFCOMP statement replaces the field errors for all elements in its range(s). Previous
field errors present in the range are discarded or incremented depending on the setting of ADD
logical attribute of the EOPTION command. EFCOMP defines the field errors in terms of relative
or absolute components.
The attributes are:
ORDER If relative errors are entered for multipoles, this defines the order of the base
component to which the relative errors refer. This reference strength kref
always refers to the normal component. In order to use a skew component as
the reference, the reference radius should be specified as a negative number.
(Default: 0)
Please note that this implies to specify k0to assign relative field errors to a
bending magnet since k0is used for the normalization and NOT the ANGLE.
168 CHAPTER 24. ERROR DEFINITIONS
RADIUS radius Rwhere dknr(i) or dksr(i) are specified for 0...i...20 (default 1 m). This
attribute is required if dknr(i) or dksr(i) are specified. If Ris negative, the
skew component is used for the reference strength.
DKN(i) Absolute error for the normal multipole strength with (2i+2) poles. (De-
fault: 0 m−i).
DKS(i) Absolute error for the skewed multipole strength with (2i+2) poles. (De-
fault: 0 m−i).
DKNR(i) Relative error for the normal multipole strength with (2i+2) poles. (De-
fault: 0 m−i).
DKSR(i) Relative error for the skewed multipole strength with (2i+2) poles. (De-
fault: 0 m−i).
Time memory effects:
The relative errors can be corrected for possible time memory effects. A correction term is
computed and added to the relative error.
The correction term is parametrized as a 3rd order polynomial in the reference strength kref
according to:
∆ = X(ci∗ki
ref )i= 0...3
The coefficients cifor the polynomial must be supplied in the command.
Two additional parameters and options are required:
HYSTER if it is set to 1 applies the correction term derived from the reference strength
and the coefficients.
HCOEFFN, HCOEFFS normal and skew coefficients for the computation of the correction term.
The four coefficients are specified in increasing order, starting with the 0th
order. Each group of four coefficients is valid for one order of the field errors.
Trailing zeros can be omitted, preceding zeros must be given.
Examples:
Example to assign relative errors to quadrupoles:
SELECT, FLAG=error, PATTERN="q.*";
EFCOMP, ORDER=1, RADIUS=0.010,
DKNR={0, 4e-1, 1e-1, 2e-3, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0},
DKSR={0, 4e-1, 1e-1, 2e-3, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
Example to add time memory effect to relative errors:
SELECT, FLAG=error, PATTER="^q.*"; [FIXME]
EFCOMP, ORDER=1, RADIUS=0.020, HYSTER=1,
DKNR={0, 1e-2, 2e-4, 4e-5, 1.e-5, 0, 0, 0, 0, 0,
![](asset-ab.png)
24.3. EOPTION: SET OPTIONS FOR ERROR DEFINITION 169
0, 0, 0, 0, 0, 0, 0, 0, 0, 0},
DKSR={0, 1e-2, 2e-4, 4e-5, 1.e-5, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0},
HCOEFFN={0.000, 0.000, 0.000, 0.000, ! coeff. mult. order 0
0.001, 0.000, 0.000, 0.000, ! coeff. mult. order 1
0.000, 0.000, 0.002, 0.000}; ! coeff. mult. order 2
See also: random values and deferred expressions.
24.3 EOPTION: Set Options for Error Definition
The error option command specifies different seeds for random values:
EOPTION, SEED=real, ADD=logical;
SEED Selects a particular sequence of random values. A SEED value is an integer in
the range [0...999999999] (default: 123456789). SEED alone continues with the
current sequence See also: Random Values.SEED may be an expression.
ADD If this logical flag is set, an EALIGN or EFCOMP causes the errors to be added
on top of existing ones. If it is not set, new errors overwrite any previous defi-
nitions. The default value is TRUE if it is omitted in the EOPTION command.
The default value is false if no EOPTION command is used.
Please note a recent modification: the default value for the ADD option is only
applied as long as the ADD option has not been set explicitly.
Once it was set with EOPTION, it is NOT reset to the default when the ADD
option is omitted in subsequent calls to EOPTION.
Example:
EOPTION, SEED = 987456321;
The random number generator for MAD-X is taken from [18].
24.4 EPRINT: List Machine Imperfections
This command prints a table of errors assigned to elements. The range for these elements
has to be specified. Field errors are printed as absolute errors, because all relative errors are
transformed to the corresponding absolute error at definition time. An error print is requested
by the statement
SELECT, FLAG=ERROR, RANGE=range, CLASS=name, PATTERN=string;
EPRINT;
and elements are now selected by the SELECT command.
A listing for ALL elements, i.e. not only the selected, can be obtained with the command
![](asset-ac.png)
170 CHAPTER 24. ERROR DEFINITIONS
EPRINT, FULL=TRUE;
In that case, the SELECT command has no effect.
24.5 ESAVE: Writing errors to a file
ESAVE, FILE=string;
This command saves a table of errors assigned to elements on a file, using a format which can
be read in again to obtain the same results. This allows dumping the errors and reloading
them after a new USE command. The range for these elements has to be specified. An error
save is requested by the statement
Example:
SELECT, FLAG=ERROR, RANGE=range, CLASS=name, PATTERN=string; ESAVE, FILE=err.file;
and elements selected by the SELECT command are saved to the file.
To save the errors of all elements to a file, one can use:
SELECT, FLAG = ERROR, FULL;
ESAVE, FILE = err.file;
Please note: in case of field errors, the absolute errors are saved and not relative
errors.
24.6 SETERR: Reading errors from a table or file
To assign errors from a file is not a priori straightforward. It may be required to re-assign
existing errors after a USE command was executed since the USE command deletes all errors
attached to a sequence).
Errors stored in the form of an internal table (errtab) can be directly attached to the appro-
priate positions in the sequence with the command:
SETERR, TABLE=errtab;
The table errtab can be generated internally or from an external file (errfile) with the generic
command READMYTABLE.
The command sequence:
READMYTABLE, file=errfile, table=errtab;
SETERR, TABLE=errtab;
reads the file errfile into the table errtab and the command SETERR attaches the errors to the
elements in the active sequence.
24.6. SETERR: READING ERRORS FROM A TABLE OR FILE 171
The file errfile can be produced by a preceding ESAVE command or any other utility. It should
follow the format of a file generated with ESAVE (see example program).
Please note:
1. To assign correctly the errors from the file to the elements in the sequence, all elements
must have individual names, otherwise an identification is not possible. Elements in the
file not identified in the active sequence are ignored.
2. Errors are assigned to ALL elements found in the file and the FLAG=ERROR is set. There-
fore the number of elements selected corresponding to a command like
SELECT, FLAG=ERROR,...;
can be different after the execution of SETERR.
![](asset-ae.png)
Chapter 25. Orbit Correction
This chapter describes the commands to correct a closed orbit or a trajectory. The initial
orbit or trajectory to be corrected can be obtained from an internal or external TFS table.
The purpose of this orbit module is to provide some basic tools to assess the performance of
an orbit correction system of a machine in the design phase.
Although some interface is available, it cannot and does not provide the full functionality
expected from a dedicated online orbit correction and steering program.
25.1 CORRECT
The CORRECT statement makes a complete closed orbit or trajectory correction using the
computed values at the monitors from the Twiss table.
The CORRECT command has the following format (some options are valid only for special
algorithms):
CORRECT, SEQUENCE=seqname,
FLAG=string, MODE=string, PLANE=string,
COND=integer, NCORR=integer,
SNGVAL=real, SNGCUT=real,
MONERROR=integer, MONON=real, MONSCALE=real,
CORRLIM=real, TWORING=logical, UNITS=real,
CORZERO=integer, ERROR=real,
ORBIT=table, MODEL=table, TARGET=table,
BEAM1TAB=table, BEAM2TAB=table,
EXTERN=logical,
NAME COL=string, X COL=string, Y COL=string,
CLIST=filename, MLIST=filename,
RESOUT=integer, TWISSUM=integer;
The command CORRECT is set up with defaults which should allow a reasonable correction for
most cases with a minimum of required options (see Example 1 below).
The orbit correction must always be preceded by TWISS commands which generate Twiss
tables. The most recent Twiss table is assumed to contain the optical parameters and the
distorted orbits.
The options used in the CORRECT command are:
FLAG can be ”ring” or ”line”, either a circular machine or a trajectory is corrected.
(Default: ring)
MODE defines the method to be used for corrections.
Available modes are LSQ, MICADO and SVD. The first performs a least squares
minimization using all available correctors. The mode SVD uses a Singular
172
25.1. CORRECT 173
Value Decomposition to compute a correction using all available correctors.
The latter can also be used to condition the response matrix for the modes
LSQ or MICADO (using COND=1). It is highly recommended to precede a LSQ
correction by a SVD conditioning (set COND=1).
The mode MICADO is a ”best kick” algorithm. Naive use or using it with a large
number of correctors (see option NCORR) can get unexpected results. To avoid
the creation of local bumps, it is recommended to precede a MICADO correction
by a SVD conditioning (set COND=1).
(Default: MICADO)
PLANE With PLANE=x, the orbit correction is performed for the horizontal plane; With
PLANE=y, the correction is performed for the vertical plane. made. (This differs
from the MAD-8 implementation).
(Default: X for horizontal plane)
COND When COND=1, a Singular Value Decomposition is performed and the response
matrix is conditioned to avoid linearly dependent correctors. This can be
used to avoid creation of artificial bumps during a LSQ or MICADO correction.
Please note: this option is not robust since it depends on parameters which
control the determination of singular values and redundant correctors. These
can be set with the attributes SNGVAL and SNGCUT. Both parameters depend
on the machine and may need adjustment. Default values are adjusted to large
machines and give ”reasonable” performance for smaller machines.
NCORR Defines the number of correctors to be used by the MICADO algorithm. If
NCORR=0 all available correctors are used. Only used by the MICADO algorithm.
(Default: 0 i.e. all available correctors)
SNGVAL Used to set the threshold for finding singular values with the COND attribute.
(Hint: smaller number finds fewer singular values).
Use with care !
(Default: 2.0)
SNGCUT Used to set the threshold for finding redundant correctors with the COND at-
tribute. (Hint: larger number finds fewer redundant correctors).
Use with extreme care !
(Default: 50.0)
MONERROR When MONERROR is 1, the MREX and MREY alignment errors on monitors assigned
by EALIGN are taken into account, otherwise they are ignored.
(Default: 0)
MONSCALE When MONSCALE is 1, the MSCALX and MSCALY scaling errors on monitors
assigned by EALIGN are taken into account, otherwise they are ignored.
(Default: 0)
MONON takes a real number between 0.0 and 1.0. It determines the number of available
monitors. If the command is given, each monitor is considered valid with a
probability MONON. In the average a fraction (1.0−MONON) of the monitors will
be disabled for the correction, i.e. they are considered not existing. This allows
![](asset-b0.png)
174 CHAPTER 25. ORBIT CORRECTION
to study the effect of missing monitors.
(Default: 1.0 i.e. 100 %)
CORRLIM A limit on the maximum corrector strength can be given and a WARNING is
issued if it is exceeded by one or more correctors. Please note: the strengths
computed by the correction algorithms are NOT limited, only a warning is
printed !
(Default: 1.0 mrad)
TWORING When true, the correction will be done on two rings at once. The only cor-
rection mode available in this case is MICADO. The attribute ORBIT=table is
ignored and BEAM1TAB=table and BEAM2TAB=table are used instead.
(Default: false)
UNITS when this parameter is set the value is a multiplier [TO BE COMPLETED]
Default unknown
CORZERO an integer value to specify whether corrector settings should be all reset to
zero before starting the orbit correction (CORZERO>0) or corrector settings
calculated by the orbit correction should be added to existing corrector settings
(CORZERO= 0, Default).
ERROR specifies the maximum RMS value, in meters, of the orbit to be reached by
the correction algorithm, e.g. ERROR=1.e-3 for a 1 mm RMS target value.
(Default: 1.e-5 m)
Normally the last active table provides the orbit to be corrected and the model for the
correction. This can be overwritten by the appropriate options. Optionally, these tables can
be given names like in TWISS, TABLE=name;. To use these named tables, one of the following
optional parameters must be used:
ORBIT When this parameter is given, the orbit to be corrected is taken from a named
table. The default is the last (named or unnamed) Twiss table.
MODEL When this parameter is given, the model for the correction is taken from a
named Twiss table. The default is the last (named or unnamed) Twiss table.
TARGET When this parameter is given, the correction is made to a named target orbit,
pre-computed with a TWISS command. Default is correction to the zero orbit.
EXTERN (default: false): When false, the ORBIT and TARGET table are assumed to be
computed by MAD-X with a previous TWISS command. When set to true, that
option allows to use twiss tables imported from an external file (with the
READMYTABLE command), for example to use measured BPM data. In that
case, the imported twiss table is allowed to contain coordinate data only at
the location of the monitors.
NAME COL The name of the column that contains the name of monitors in the tables.
(Default: ”name”)
X COL The name of the column that contains the Xposition readings of monitors in
the tables. (Default: ”X”)
![](asset-b1.png)
25.1. CORRECT 175
Y COL The name of the column that contains the Yposition readings of monitors in
the tables. (Default: ”Y”)
Example of use of CORRECT to reproduce a measured orbit:
! To have a refererence optical model
twiss, table=twiss_ref;
! The bpm.tsv is a reduced Twiss file containing only lines for the BPMs
readmytable, file="bpm.tsv", table="twiss_bpm";
! correct orbit using external measurements
correct, flag=ring, mode=micado, ncorr=5, cond=1 ,plane=x, extern,
model=twiss_ref, orbit=twiss_ref, target=twiss_bpm,
error=1.0e-21;
Two attributes affect the printing of tables and results:
CLIST the name of the file where corrector settings (in units of rad) before and after
correction are printed.
MLIST the name of the file where monitor readings (in units of m) before and after
correction are printed.
RESOUT This command outputs the results for all monitors and all correctors in a com-
puter readable format if its integer argument is larger than 0. The argument
is added to the output. Useful to analyze runs with loops to produce large
statistics.
ATTENTION: May produce gigantic outputs for large machines.
TWISSUM If the attribute value is larger than 0, CORRECT prints maximum orbit and r.m.s.
values for both planes, taken from the Twiss summary table, in computer
readable form. This allows to analyze orbits etc. at elements that are not
monitors or correctors. The argument is added to the output. This attributes
is used only to produce output: no correction is made, and all other attributes
are ignored.
Obsolete commands or options:
ITERATE, ITERMAX /* Done with loop feature in MAD commands */
THREADER, THRTOL, WRORBIT /* Not part of orbit correction module */
M1LIST, M2LIST /* Replaced by MLIST */
C1LIST, C2LIST /* Replaced by CLIST */
GETORBIT, PUTORBIT /* Replaced by generic TFS access */
GETKICK, PUTKICK /* Replaced by generic TFS access */
EXAMPLES:
for complete MAD input files see section on examples:
Example 1: correct orbit in horizontal plane, taken from most recent Twiss table, using
default algorithm (MICADO)
CORRECT, PLANE = x;
![](asset-b2.png)
176 CHAPTER 25. ORBIT CORRECTION
Example 2: no correction, only output of Twiss summary
CORRECT, TWISSUM = 1;
Example 3: correct orbit in horizontal plane, corrector and monitor output on table
CORRECT, PLANE = x, MODE = lsq, CLIST = corr.out, MLIST = mon.out;
Example 4: correct orbit in horizontal plane, use alignment and scaling errors, 15% of orbit
correctors faulty
CORRECT, PLANE = x, MONERROR = 1, MONSCALE = 1, MONON = 0.85;
Example 5: correct orbit in horizontal plane for a two-beam machine, using MICADO with
no SVD conditioning, zeroing correctors before the correction.
CORRECT, FLAG=ring,
TWORING, BEAM1TAB=twb1, BEAM2TAB=twb2,
MODE=micado, COND=0, NCORR=4, ERROR=1E-6, PLANE=x,
MLIST=’mx12.tab’, CLIST=’cx12.tab’, RESOUT=1, CORZERO=1;
25.2 USEMONITOR, USEKICK
To provide more flexibility with orbit correction two commands are provided:
USEMONITOR, STATUS=flag, SEQUENCE=sequence, RANGE=range,
CLASS=class, PATTERN=regex;
USEKICK, STATUS=flag, SEQUENCE=sequence, RANGE=range,
CLASS=class, PATTERN=regex;
The command USEMONITOR activates or deactivates a selection of beam position monitor. This
command affects elements of types MONITOR, HMONITOR, and VMONITOR.
The command USEKICK activates or deactivates a selection of orbit correctors. This command
affects elements of types KICKER, HKICKER and VKICKER.
Both commands have the same attributes:
STATUS If this flag is true (on), the selected elements are activated. Active orbit moni-
tor readings will be considered, and active correctors can change their strengths
in subsequent correction commands. Inactive elements will be ignored subse-
quently.
SEQUENCE The sequence can be specified, otherwise the currect sequence is used for this
operation.
RANGE, CLASS, PATTERN : The usual selection commands are used to identify the elements
for this operation.
Example:
![](asset-b3.png)
25.3. CSAVE 177
USE,... ! set working beam line
... ! define imperfections
USEKICK, STATUS = OFF, RANGE = ...; ! deactivate selected correctors
USEMONITOR, STATUS = OFF, RANGE = ...; ! deactivate selected monitors
CORRECT, NCORR = 32; ! uses different set of correctors
USEKICK, STATUS = OFF, RANGE = ...; ! deactivate different set of correctors
CORRECT, NCORR = 32; ! uses different set of correctors
25.3 CSAVE
This section is under construction, options presently only available in MADX development
version.
25.4 SETCORR
This section is under construction, options presently only available in MADX development
version.
25.5 COPTION
COPTION, SEED=integer, PRINT=integer;
In the orbit program monitors can be randomly disabled and the correct option command
specifies different seeds for random values:
SEED Selects a particular seed for the sequence of random values. The attribute
value is an integer in the range [0...999999999], or an expression that evaluates
to an integer in the same range (Default: 123456789). SEED alone with no
attribute value continues with the current sequence.
See also: Random values.
PRINT A flag taking integer values to control the printout.
In general: the higher the value the more printout is produced.
For PRINT=0 no output is produced.
The default value is 1 (Correction summary is given).
Example:
COPTION, SEED=987456321, PRINT=2;
Note that the random generator for MAD-X is taken from [18].
![](asset-b4.png)
Chapter 26. SODD: Second Order Detuning and
Distortion
The SODD command calculates the Second Order Detuning and Distortion, as described in
[19], on the beam line defined by the last USE command followed by a TWISS command.
The SODD command is based on the stand-alone program[11] with the same name, with
analytical computation extended to the second order distortion[20].
SODD, DETUNE=logical, DISTORT1=logical, DISTORT2=logical,
START STOP = start,stop,
MULTIPOLE ORDER RANGE = first,last,
NOPRINT=logical, PRINT ALL=logical, PRINT AT END=logical,
NOSIXTRACK=logical;
The attributes of the SODD command are:
DETUNE flag to trigger calculation of the detuning function terms in first and second
order in the strength of the multipoles. (Default: false)
DISTORT1 flag to trigger the calculation of the distortion function and the Hamiltonian
terms in first order in the strength of the multipoles. (Default: false)
DISTORT2 flag to trigger the calculation of the distortion function and Hamiltonian terms
in second order in the strength of the multipoles. (Default: false)
START STOP positions (reals in meters) along the beamline defining a longitudinal interval.
(Default: 0.0,0.0)
MULTIPOLE ORDER RANGE lowest and highest multipole order to be taken in account, given as
integers. (Default: 1,2)
NOPRINT flag to the effect that no file or internal table is created to hold the results. If
true, the attributes PRINT ALL or PRINT AT END have no effect. (Default: false)
PRINT ALL flag to generate files and internal tables with results for each mutipole. (De-
fault: false)
PRINT AT END flag to generate files and internal tables with results at the end of the longitu-
dinal interval. (Default: false)
NOSIXTRACK flag to signal that fc.34 shall not be generated internally by invoking the conver-
sion routine of sixtrack. The user should provide this file before the execution
of the SODD command. (Default: false)
Note that the first row of every file generated by SODD is a header containing the names of
the columns. This row is absent in the internal tables.
A more detailed description can be found in [21].
178
![](asset-b5.png)
26.1. DETUNE 179
26.1 DETUNE
The attribute DETUNE triggers the calculation of the detuning function terms in first and
second order in the strength of the multipoles.
If the logical attribute PRINT AT END is set to true, the following two files, and corresponding
tables, are created:
•”detune 1 end” contains five columns : multipole order, horizontal or vertical plane
coded as 1 or 2, horizontal or vertical detuning, order of horizontal invariant and order
of vertical invariant.
•”detune 2 end” contains five columns : first multipole order, second multipole order,
horizontal detuning, order of horizontal invariant and order of vertical invariant.
If the logical attribute PRINT ALL is set to true, the following two files, and corresponding
tables, are created :
•”detune 1 all” contains five columns : multipole order, horizontal or vertical plane
coded as 1 or 2, horizontal or vertical detuning, order of horizontal invariant and order
of vertical invariant.
•”detune 2 all” contains five columns : first multipole order, second multipole order,
horizontal detuning, order of horizontal invariant and order of vertical invariant.
26.2 DISTORT1
The attribute DISTORT1 triggers the calculation of the distortion function and the Hamiltonian
terms in first order in the strength of the multipoles.
If the logical attribute PRINT AT END is set to true, the following two files, and corresponding
tables are created:
•”distort 1 F end” contains eight columns : multipole order, cosine and sine part of
distortion, amplitude of distortion, j, k, l, m.
•”distort 1 H end” contains eight columns : multipole order, cosine and sine part of
Hamiltonian, amplitude of Hamiltonian, j, k, l, m.
If the logical attribute PRINT ALL is set to true, the following two files, and corresponding
tables, are created :
•”distort 1 F all” contains eleven columns : multipole order, appearance number in posi-
tion range, number of resonance, position, cosine and sine part of distortion, amplitude
of distortion, j, k, l, m.
•”distort 1 H all” contains eleven columns : multipole order, appearance number in
position range, number of resonance, position, cosine and sine part of Hamiltonian,
amplitude of Hamiltonian, j, k, l, m.
![](asset-b6.png)
180 CHAPTER 26. SODD: SECOND ORDER DETUNING AND DISTORTION
26.3 DISTORT2
The attribute DISTORT2 triggers the calculation of the distortion function and Hamiltonian
terms in second order in the strength of the multipoles.
If the attribute PRINT AT END is set to true, the following two files, and corresponding tables,
are created:
•”distort 2 F end” contains nine columns : first multipole order, second multipole order,
cosine and sine part of distortion, amplitude of distortion, j, k, l, m.
•”distort 2 H end” contains nine columns : first multipole order, second multipole order,
cosine and sine part of Hamiltonian, amplitude of Hamiltonian, j, k, l, m.
![](asset-b7.png)
Chapter 27. Touschek Lifetime and Scattering
Rates
The TOUSCHEK module computes the Touschek lifetime and the scattering rates around a
lepton or hadron storage ring, based on the formalism of Piwinski in [22] and his article on
Touschek lifetime in [23].
The syntax of the TOUSCHEK command is:
TOUSCHEK, TOLERANCE=real, FILE=filename;
The arguments have the following meaning:
TOLERANCE the tolerance for the numerical integrator DGAUSS. (Default: 1.e-7)
FILE The name of the output file (Default: ’touschek’)
TOUSCHEK should only be called after fully qualified BEAM command and a TWISS com-
mand. One or several cavities with rf voltages should be defined prior to calling TWISS and
TOUSCHEK.
Warning: Calling EMIT between the TWISS and TOUSCHEK commands leads to TOUSCHEK
using wrong beam parameters, even if the BEAM command is reiterated.
The momentum acceptance is taken from the bucket size taking into account the energy loss
per turn U0 from synchrotron radiation. The value of U0 is computed from the second
synchrotron radiation integral synch 2 in the TWISS summ table (synch 2 is calculated only
when the TWISS option ’chrom’ is invoked), using Eq. (3.61) in [24], which was generalized to
the case of several harmonic rf systems. If synch 2 is zero, not defined or not calculated, zero
energy loss is assumed. In the case of several rf systems with nonzero voltages, it is assumed
that the lowest frequency system defines the phase of the outer point on the separatrix
when calculating the momentum acceptance, and that all higher-harmonic systems are either
in phase or in anti-phase to the lowest frequency system. (Note: if a storage rings really
uses a different rf scheme, one would need to change the acceptance function in the routine
cavtousch0 for that ring.)
Example:
BEAM, PARTICLE = PROTON, ENERGY = 450, NPART = 1.15e11,
EX = 7.82E-9, EY = 7.82E-9, ET = 5.302e-5;
USE, PERIOD = LHCb1;
...
VRF = 400;
...
SELECT, FLAG = TWISS, CLEAR;
TWISS, CHROM, TABLE, FILE;
...
TOUSCHEK, FILE, TOLERANCE=1.e-8;
181
182 CHAPTER 27. TOUSCHEK LIFETIME AND SCATTERING RATES
The first command defines the beam parameters. It is essential that the longitudinal emittance
is set. The command USE selects the beam line or sequence. The next command assign a
value to the cavity rf voltage vrf (example name). The SELECT clear previous assignments
to the TWISS module, TWISS calculates and saves the values of all twiss parameters for all
elements in the ring; the TOUSCHEK command computes the Touschek lifetime and writes
it to the file ’touschek’ (default name).
The results are stored in the TOUSCHEK tables, and can be written to a file (with the
default name ’touschek’ in the example above), or values can be extracted from the table
using the value command as follows
VALUE, table(touschek,name), table(touschek,s), table(touschek,tli),
table(touschek,tliw), table(touschek,tlitot);
where ’name’ denotes the name of a beamline element, Sthe position of the center of the
element, TLI the instantaneous Touschek loss rate within the element, and TLIW the in-
stantaneous rate weighted by the length of the element divided by the circumference (its
contribution to the total loss rate), and TLITOT the accumulated loss rate adding the rates
over all beamline elements through the present position. The value of TLITOT at the end of
the beamline is the inverse of the Touschek lifetime in units of 1/s.
All results can also be printed to a file using the command
WRITE, TABLE=touschek, FILE;
The MAD-X Touschek module was developed by Catia Milardi and Frank Zimmermann.
The MAD-X Touschek module was partially rewritten in November 2013 by Ghislain Roy after
the discovery of a few bugs in the original code:
The first bug concerned a numerical instability in the computation of the B2 parameter as
listed in Eq. 34 in [22].
The initial alogorithm implemented the calculation of square root of the difference between
two expressions. It turned out that the numerical values of both expressions could sometimes
be very large and nearly equal.
Because of limited precision in floating point calculations, the difference could sometimes lead
to negative values and the square root returned an undefined value (NaN). The integrator
then failed to compute the integral and returned a value of zero, with the printing of a faintly
related message that too high accuracy was required for integrator DGAUSS. The algorithm
didn’t stop there and the end result was that the summation over all elements in the range
was wrong and the end results were also wrong.
This bug was eliminated by evaluation the first expression in equation 34 which calculates
directly the B2 factor by taking the square root of the sum of two squares, hence ensuring
that an instability of the same kind cannot happen.
Another problem was that in the original algorithm the inverse Touschek lifetime was cal-
culated by taking the average of the twiss parameters at both ends of the element as input.
The resulting set of parameters was no longer consistent, resulting also in poor calculation.
This has been changed by calculating the inverse Touschek lifetime at specific points, always
183
considering as input the Twiss parameters given by Twiss at a single location. This provides
at least very accurate results for the TLI parameters.
The integration over the length of the element is now done in different ways, depending
whether the preceding TWISS command calculated the Twiss parameters at the end of the
element or at the centre (CENTRE option of TWISS).
In the first case (calculation at the end of the element, CENTRE=false), the inverse Touschek
lifetime (TLI) is calculated at the end of each element. The weighted contribution of element
ito the total inverse Touschek lifetime is then given by
TLIW[i] = 0.5 * (TLI[i] + TLI[i-1]) * L[i] / CIRC
In the second case (calculation at the center of the element, CENTRE=true), the inverse
Touschek lifetime (TLI) is also calculated at the center of each element. The weighted con-
tribution of element ito the total inverse Touschek lifetime is then given by
TLIW[i] = TLI[i] * L[i] / CIRC
Another bug that was uncovered in the original algorithm was that the vertical dispersion
was wrongly taken into account and mostly ignored: DY and DPY were set uniformly to half
the initial values for lack of updating in the loop over elements.
The new algorithms have been inserted in MAD-X as of version 5.01.04, a development release
dated early december 2013.
![](asset-ba.png)
Chapter 28. Intra-Beam Scattering
The Intra-Beam Scattering command computes the contribution to emittance growth rates
due to Coulomb scattering of particles within relativistic beams. The algorithms in this mod-
ule have been derived from the formalism presented in 1982 by J.D. Bjorken and S.K. Mtingwa
[25], and are also using the expansion of M. Conte and M. Martini [26] developped in 1985,
generalized to the case of nonzero vertical dispersion.
The present implementation of the IBS module in MAD-X is described in a forthcoming note
[27].
The syntax of the IBS command is:
IBS, FILE=string;
The IBS command has one attribute:
FILE outputs the resulting ”ibs” table to the named file. (Default: ”ibs”)
The Bjorken-Mtingwa formalism takes into account the variation of the lattice parameters
(beta and dispersion functions) around the machine and consequently, the knowledge of the
optical functions along the machine is required: IBS should only be called after fully qualified
BEAM command and a TWISS command.
Warning: Calling EMIT between the TWISS and IBS commands leads to IBS using wrong
beam parameters, even if the BEAM command is reiterated.
The IBS module does not include a consistent treatment of linear betatron coupling.
The intra-beam scattering growth times are given by:
1
τi
=Ci×N
γxys
(i=x, y, s)
where Ciaccounts for some constants and the integrals for the scattering functions, N is
the number of particles in the bunch, γis the relativistic factor and iare the normalized
emittances in the horizontal, vertical and longitudinal plane respectively. These key beam
parameters must be specified through the BEAM command.
If the CENTRE=true option of TWISS was specified, the optical functions are calculated by
TWISS at the center of each element and IBS uses these values for the element. If by default
TWISS calculated the optical functions at the end of each element, IBS calculates the values
at the center of each element by performing a linear interpolation between the end values for
the previous element and the end values for the current element.
Input of the beam parameters:
A number of parameters have to be present in the BEAM command in order to run the IBS
module:
PARTICLE This is mandatory but MAD-X provides default value of PARTICLE=proton.
For ions, this parameter specifies only the name of the ions, and the MASS
184
![](asset-bb.png)
185
(approximated to the atomic unit number times the neutron mass NMASS) and
CHARGE must be provided as well.
NPART the number of particles (or number of ions).
ENERGY The definition of the energy (total, kinetic, total energy of the ions or energy
per nucleon) is a difficult one. In the present approach, the energy is the total
energy of the particle. For ions, the expected input is the proton equivalent
energy, i.e. the total energy a proton would have when circulating in the
defined machine. As an illustration, in the LHC, protons will be injected with
an energy of 450 GeV. Consequently, to evaluate the growth times for Lead
ions at injection in the LHC, one has to input ENERGY=450*charge.
An important check for the correctness of the input is the printed value of the
relativistic factor γ. The latter should correspond to:
γion =γproton ×charge
nucleon
emittances This part of the input is used to define the normalized horizontal, vertical and
longitudinal emittances. The required parameters are the physical transverse
emittances, EX and EY, and the longitudinal emittance ET.
The longitudinal emittance is defined as the product of the bunch length SIGT
times the relative energy spread SIGE, which are therefore required input.
If only the longitudinal emittance is defined, and SIGT and SIGE are omitted,
an active RF cavity is also necessary in the lattice to infer SIGT and SIGE.
Example of BEAM input:
A beam of fully stripped Lead ions at the LHC injection energy may be defined as follows for
IBS calculations:
nucleon = 208;
charge = 82;
BEAM, PARTICLE= lead, CHARGE= charge, MASS= nucleon*nmass,
ENERGY= 450*charge, NPART= 1.1E7, BUNCHED,
EX= 7.82E-9, EY= 7.82E-9, SIGE= 4.68E-4, SIGT= 0.115;
Resulting Table and File:
The IBS command produces a table ”ibs” containing the following data for each element of
the machine: element name, position, optical functions (beta, alfa, dispersion and derivative)
in both transverse planes, as well as the particular variables DELS, the length difference in
meters between consecutive elements, and TXI, TYI and TLI, the IBS growth times in the
two transverse and longitudinal planes.
This table can be accessed through the usual mechanisms. If the attribute FILE="file name"
is also given, MAD-X writes the table to the named file.
Features:
The average growth rates in [sec] are defined as variables called ibs.tx,ibs.ty,ibs.tl
for the horizontal, vertical and longitudinal growth times respectively. They are directly
accessible as variables after the IBS command, e.g.
186 CHAPTER 28. INTRA-BEAM SCATTERING
IBS;
Tx = ibs.tx;
defines a variable Tx which is the average horizontal growth rate in seconds.
Examples:
The two examples provided for the module Intra-Beam Scattering illustrate the commands
required to run the module. The two examples have been selected such as to highlight the
differences between a computation for protons and that for ions. Both examples compute the
IBS growth times at injection into the LHC.
The examples are located at http://madx.web.cern.ch/madx/madX/examples/ibs/.
![](asset-bd.png)
Chapter 29. Particle Tracking
29.1 Introduction to MAD-X Tracking Modules
A number of particles with given initial conditions can be tracked through a beam-line or a
ring. The particles can be tracked either for a single passage or for many turns.
While MAD-X keeps most of the functionality of MAD-8, the trajectory tracking in MAD-X is
considerably modified compared to MAD-8. The reason is that in MAD-8 the thick lens tracking
is inherently not symplectic, which implies that the phase space volume is not preserved during
the tracking, i.e. contrary to the real particle the tracked particle amplitude is either growing
or decreasing.
The non-symplectic tracking as in MAD-8 has been completely excluded from MAD-X by taking
out the thick lens part from the tracking modules. Instead two types of tracking modules
(both symplectic) are implemented into MAD-X.
The first part of this design decision is the thin-lens tracking module (THINTRACK) which tracks
symplecticly through drifts and kicks and by replacing the end effects by their symplectic
part in the form of an additional kick on either end of the element. This method requires a
preliminary conversion of a sequence with thick elements into one composed of thin elements
(see the MAKETHIN command).
The second part of this design decision is to produce a thick lens tracking module based on
the PTC code of E. Forest that allows a symplectic treatment of all accelerator elements giving
the user full control over the precision (number of steps and integration type) and exactness
(full or extended Hamiltonian) of the results.
The first PTC thick-lens tracking module is named PTC TRACK. It has the same features
as the thin-lens tracking code (thintrack) except that it treats thick-lenses in a symplectic
manner.
There is a second PTC tracking module called the line tracking module (PTC TRACKLINE). It
was developped for tracking particles in CLIC, with the specificities that it can deal with
beam-lines containing traveling-wave cavities and includes actual beam acceleration.
29.2 Overview of Thin-Lens Tracking
The thin-lens tracking module of MAD-X performs element per element tracking of one or
several particle trajectories in the last USEd sequence.
Only thin elements are allowed (apart from the element DRIFT), which guarantees the sym-
plecticity of the coordinate transformation. Any lattice can be converted into a ”thin element”
lattice by invoking the MAKETHIN command.
Several commands are actually required to complete a tracking run:
187
![](asset-be.png)
188 CHAPTER 29. PARTICLE TRACKING
TRACK, DELTAP=real, ONEPASS=logical, DAMP=logical;
QUANTUM=logical, DUMP=logical, APERTURE=logical,
ONETABLE=logical, RECLOSS=logical, FILE=filename,
UPDATE=logical;
...
START, X=real, PX=real, Y=real, PY=real, T=real, PT=real;
START, FX=real, PHIX=real, FY=real, PHIY=real,
FT=real, PHIT=real;
...
OBSERVE, PLACE=string;
...
RUN, TURNS=integer, MAXAPER=double array, FFILE=integer;
...
DYNAP, TURNS=real, FASTUNE=logical, LYAPUNOV=real,
MAXAPER=real array, ORBIT=logical;
...
ENDTRACK;
Inside the block TRACK-ENDTRACK a series of initial trajectory coordinates can be specified by
the START command (as many commands as trajectories). This will be usually done in a
WHILE-loop. Note that the coordinates are either canonical coordinates or action-angle
variables!
For usual tracking (single/multi-turn), all coordinates are specified with respect to the actual
closed orbit (possibly off-momentum, with magnet errors) and NOT with respect to the
reference orbit.
If the option ONEPASS is used, the coordinates are specified with respect to the reference orbit.
The name ONEPASS might be misleading: Still tracking can be single- or multi-turn!
The tracking is actually started with the RUN command, where the option TURNS defines for
how many turns the particles will be tracked in the given sequence.
If the option DUMP is used, the particle coordinates are written to files at each turn. The
output files are named automatically. The name given by the user is followed by .obsnnnn
(observation point), followed by .pnnnn (particle number).
Hence filenames look like track.obs0001.p0001.
Tracking creates a number of internal tables and can create files on disk: TRACKSUMM,
TRACKLOSS, and TRACKONE or TRACK.OBS$$$$.P$$$$ (depending on the attribute ONETABLE
of the RUN command).
These internal tables can be accessed via the TABLE-access functions.
Plotting of particle coordinates or other data in these tables is possible in MAD-X. Plotting
can also be done with external programs by using the files created by TRACK.
MAD-X also has the capability to treat space-charge during tracking runs. There is no space-
charge command per se but space charge is controlled through several options of MAD-X (see
OPTION) and specific attributes of the RUN command in this TRACK environment. A section
specific to space charge options and particularities appears below.
![](asset-bf.png)
29.3. TRACK 189
29.3 TRACK
The TRACK command initiates trajectory tracking by entering the thin-lens tracking module.
TRACK, DELTAP=real, ONEPASS=logical, DAMP=logical;
QUANTUM=logical, DAMP=logical, APERTURE=logical,
ONETABLE=logical, RECLOSS=logical, FILE=filename,
UPDATE=logical;
The attributes of the TRACK command are:
DELTAP relative momentum offset for reference closed orbit (switched off for ONEPASS)
Defining a non-zero DELTAP results in a change of the beam momentum/energy
without changing the magnetic properties in the sequence, which leads to an
off-momentum closed orbit different from the on-momentum reference orbit.
Particle coordinates are then given with respect to this new closed orbit, unless
the option ONEPASS=true is used!
(Default: 0.0)
ONEPASS flag to ensure that no closed orbit search is done, which also means that no sta-
bility test is done. This is always the case for transfer lines, but this option can
also be enabled for multi-turn tracking of a circular machine. ONEPASS=true
does NOT restrict tracking to a single turn.
With ONEPASS=true, the particle coordinates are specified with respect to the
reference orbit.
With ONEPASS=false, the closed orbit is calculated and the particle coordi-
nates are given with respect to the closed orbit coordinates.
The name of this attribute is misleading but was kept for backwards compat-
ibility.
(Default: false)
DAMP flag to introduce synchrotron damping (needs RF cavity and flag RADIATE in
the BEAM command).
(Default: false)
QUANTUM flag to introduce quantum excitation via random number generator and tables
for photon emission.
(Default: false)
DUMP flag to write the particle coordinates in files, whose names are generated auto-
matically.
(Default: false)
APERTURE a logical flag to trigger aperture check at the entrance of each element (except
DRIFTs). A particle is lost from the table of tracked particles if its position lies
outside the aperture of the current element at the entrance of this element.
(Default: false)
The APERTYPE and APERTURE information of each element in the sequence is
190 CHAPTER 29. PARTICLE TRACKING
used to assess the particle loss. However TRACK only takes into account the
predefined aperture types listed in table 22.1
Note that if no aperture information was specified for an element, the following
procedure still takes place:
→No aperture definition for element →Default apertype/aperture assigned
(currently this is APERTYPE=circle, APERTURE={0})
→If tracking with APERTURE is used and an element with APERTYPE=circle
AND APERTURE={0}is encountered, then the first value of the MAXAPER vec-
tor is assigned as the circle’s radius (no permanent assignment!). See option
MAXAPER for the default values.
⇒Hence even if no aperture information is specified by the user for certain
elements, default values will be used!
ONETABLE flag to write all particle coordinates in a single file instead of one file per
particle.
(Default: false)
RECLOSS flag to create in memory a table named ”trackloss” containing the coordinates
of lost particles.
(Default: false)
Traditionally, when a particle is lost on the aperture, this information is written
to stdout. To allow more flexible tracking studies, the coordinates of lost
particles and additional information can also be saved in a table in memory.
Usually one would save this table to a file using the WRITE command after the
tracking run has finished. The following information is available in the TFS
table ”trackloss”:
•Particle ID (number)
•Turn number
•Particle coordinates (x,px,y,py,t,pt)
•Longitudinal position in the machine (s)
•Beam energy
•Element name, where the particle is lost
FILE name for the track table. The default name is different depending on the value
of the ONETABLE attribute.
(Default: ”track” if ONETABLE=true, ”trackone” if ONETABLE=false)
UPDATE flag to trigger parameter update per turn.
(Default: false)
Specifying UPDATE=true gives access to the following additions:
tr$turni this special variable contains the turn number; it can be used
in expressions like KICK := SIN(tr$turni) and is updated at
each turn during tracking.
![](asset-c1.png)
29.4. START 191
tr$macro this special macro can be user-defined and is executed/updated
at each turn, during tracking. A macro structure is necessary
to provide for table access. e.g.
tr$macro(turn): macro={
commands that can depend on the turnnumber;
};
Remarks
IMPORTANT: If an RF cavity has a non-zero voltage, synchrotron oscillations are automat-
ically included. If tracking with constant momentum is desired, then the voltage of the RF
cavities has to be set to zero. If an RF cavity has a no zero voltage and DELTAP is non zero,
tracking is done with synchrotron oscillations around an off-momentum closed orbit.
29.4 START
After the TRACK command, initial trajectory coordinates must be provided for each trajectory
or particle to be tracked, with one START command per trajectory or particle.
The coordinates can be expressed as either canonical or action-angle coordinates.
START, X=real, PX=real, Y=real, PY=real, T=real, PT=real;
START, FX=real, PHIX=real, FY=real, PHIY=real,
FT=real, PHIT=real;
For the case of action-angle coordinates, the normalised amplitudes are expressed in number
of r.m.s. beam size FX,FY,FT(the actions being computed with the emittances given in
the BEAM command) in each mode plane. The phases are ΦX, ΦYand ΦTexpressed in
radian. In the uncoupled case, we have in the plane mode labelled z, and with Ezbeing the
r.m.s. emittance in that plane:
Z=Fz√Ezcos Φz, Pz=Fz√Ezsin Φz(29.1)
The attributes of the START command are:
X, PX, Y, PY, T, PT canonical coordinates.
FX, PHIX, FY, PHIY, FT, PHIT action-angle coordinates.
Remarks
For usual tracking (single/multi-turn), all coordinates are specified with respect to the actual
closed orbit (possibly off-momentum, with magnet errors) and NOT with respect to the
reference orbit.
If the option onepass of the TRACK is used, the coordinates are specified with respect to the
reference orbit.
![](asset-c2.png)
192 CHAPTER 29. PARTICLE TRACKING
29.5 OBSERVE
During the tracking process, particle coordinates at specific named locations along the ma-
chine can be printed to file(s). The declaration of an observation point is with the OBSERVE
command:
OBSERVE, PLACE=string;
The single attribute of OBSERVE is:
PLACE the name of the observation point.
Several OBSERVE commands can be given for the same tracking job, one per observation point.
If no OBSERVE command is given in a tracking job, but the DUMP option in the TRACK command
is used, the trajectory coordinates are still recorded and one observation point is provided at
the starting point of the sequence.
The output files are named automatically. The name given by the user (attribute FILE of the
TRACK command) is followed by ”.obsnnnn”, where nnnn is the observation point number, and
followed by ”.pnnnn” wherer nnnn is now the particle number. Hence the default filename
for the first obseration point and first particle looks like track.obs0001.p0001.
29.6 RUN
The actual tracking is triggered by the RUN command.
RUN, TURNS=integer, MAXAPER=real array, FFILE=integer;
The RUN command has three attributes:
TURNS number of turns to be tracked.
MAXAPER upper limits for the six coordinates.
(Default: {0.1, 0.01, 0.1, 0.01, 1.0, 0.1}
The limits defined by the MAXAPER option are only being taken into account if
the APERTURE option of the TRACK command is used.
FFILE defines the turn periodicity for printing coordinates at observation points. (De-
fault: 1)
FFILE=n will print coordinates every n-th turn only.
29.7 DYNAP
The DYNAP command calculates tunes, tune footprints, smear and Lyapunov exponent from
tracking data. DYNAP can be called instead of RUN inside a TRACK command environment.
DYNAP, TURNS=integer, FASTUNE=logical, LYAPUNOV=real,
MAXAPER=real array, ORBIT=logical;
29.7. DYNAP 193
For each previously entered start command, DYNAP tracks two close-by particles over a selected
number of turns (minimum 64 and maximum 1024), from which it obtains the betatron tunes
with error, the action smear, and an estimate of the lyapunov exponent. Many such companion
particle-pairs can be tracked at the same time, which speeds up the calculation.
The smear is defined as 2×(wxymax −wxymin )/(wxymax +wxymin ), where the wxymin,max
refer to the minimum and maximum values of the sum of the transverse betatron invariants
wx +wy during the tracking.
The tunes are computed by using an FFT and formula (18) in reference [28] if the number
of turns is 64 or less, or formula (25) in the same reference if the number of turns is strictly
larger than 64.
DYNAP has the following attributes:
TURNS the number of turns to be tracked (Default: 64, minimum: 64 and maxi-
mum: 1024).
FASTUNE a logical flag to compute the tunes. (Default: false)
MAXAPER a vector of 6 real numbers defining the maximum aperture beyond which the
particle is considered to be lost.
(Default: {0.1, 0.01, 0.1, 0.01, 1.0, 0.1}
LYAPUNOV the initial distance which is added to the xcoordinate of the companion particle
of every particle declared with START commands. (Default: 1.e-7 m)
ORBIT A logical flag. If set, the flag orbit is true during the tracking and its initial-
ization (default: true). This flag should be set to be true, if normalized
coordinates are to be entered.
The first command defines the beam parameters. It is essential that the longitudinal emittance
ET is set. The command USE selects the beam line or sequence. The TRACK command activates
the tracking module, START enters the starting coordinates (more than one particle can be
defined), DYNAP finally tracks two nearby particles with an initial distance equal to the value
of the LYAPUNOV attribute for each START definition over TURNS revolutions, and ENDTRACK
terminates the execution of the tracking module.
The results are stored in the DYNAP and DYNAPTUNE tables, and can be obtained by the two
commands
VALUE, TABLE(dynap,smear);
VALUE, TABLE(dynaptune,tunx),
TABLE(dynaptune,tuny),
TABLE(dynaptune,dtune);
More generally, all results can be printed to a file, using the commands
WRITE, TABLE=dynap, FILE;
WRITE, TABLE=dynaptune, FILE;
The output file lyapunov.data lists the turn number and phase distance between the two
Lyapunov partners, respectively, allowing for visual inspection of chaoticity.
![](asset-c4.png)
194 CHAPTER 29. PARTICLE TRACKING
29.8 ENDTRACK
Tracking is terminated by the command ENDTRACK with no attributes.
ENDTRACK;
29.9 Space Charge
MAD-X can perform tracking using a frozen space charge model. This process is rather in-
volved and requires careful setting of several options and switches as well as the insertion of
space-charge kicks inserted within regular elements. The Space-Charge specifics of MAD-X are
documented in [29].
Part V
PTC Commands
195
![](asset-c6.png)
Chapter 30. PTC Set-up Parameters
The Polymorphic Tracking Code [30] of Etienne Forest is a kick code, allowing a symplectic
integration through all accelerator elements giving the user full control over the precision
(number of steps and integration type) and exactness (full or extended Hamiltonian) of the
results. The degree of exactness is determined by the user and the speed of his computer.
The main advantage is that the code is inherently based on the map formalism and provides
users with all associated tools.
The PTC code is actually a library that can be used in many different ways to create an actual
module that calculates some property of interest.
Attention: PTC exists inside of MAD-X as a library. MAD-X offers the interface to PTC,i.e.
the MAD-X input file is used as input for PTC. Internally, both PTC and MAD-X have their
own independent databases which are linked via the interface. With the PTC CREATE LAYOUT
command, only numerical values are transferred from the MAD-X data structures to the PTC
data structures. Any modification to the MAD-X data structure is unknown to PTC until the
next call to PTC CREATE LAYOUT. For example, a deferred expression of MAD-X is only evaluated
at the time of the PTC CREATE LAYOUT command and is ignored within PTC afterwards.
Several modules using the PTC code have been presently implemented in MAD-X. These
MAD-X-PTC modules[31] are executed by the following commands: PTC TWISS,PTC NORMAL,
PTC TRACK,PTC TRACK LINE.
To perform calculations with these MAD-X-PTC commands, the PTC environment must be ini-
tialized, handled and turned off by special commands within the MAD-X input script.
30.1 Command Synopsis
A typical set of commands to invoke PTC is given below:
PTC CREATE UNIVERSE, SECTOR NMUL MAX= integer, SECTOR NMUL= integer,
NTPSA= logical, SYMPRINT= logical;
PTC CREATE LAYOUT, TIME= logical, MODEL= integer,
METHOD= integer, NST= integer, EXACT= logical,
OFFSET DELTAP= double, ERRORS OUT= logical,
MAGNET NAME= string, RESPLIT= logical,
THIN= double, XBEND= double,
EVEN = logical;
...
PTC MOVE TO LAYOUT, INDEX= integer;
...
PTC READ ERRORS, OVERWRITE= logical;
...
PTC ALIGN;
...
196
![](asset-c7.png)
30.2. PTC CREATE UNIVERSE 197
PTC END;
30.2 PTC CREATE UNIVERSE
The PTC CREATE UNIVERSE command is required to set-up the PTC environment.
PTC CREATE UNIVERSE, SECTOR NMUL MAX=integer, SECTOR NMUL=integer,
NTPSA=logical, SYMPRINT=logical;
The attributes are:
SECTOR NMUL MAX a global variable in PTC needed for exact sector bends defining up to
which order Maxwell’s equation are solved (see [30] page 76-77). The value
of SECTOR NMUL MAX must not be smaller than SECTOR NMUL otherwise MAD-X
stops with an error.
(Default: 10)
SECTOR NMUL a global variable in PTC needed for exact sector bends defining up to which
order the multipole are included in solving Maxwell’s equation up to order
SECTOR NMUL MAX. Multipoles of order N with N >SECTOR NMUL and N ≤
SECTOR NMUL MAX are treated similar to SixTrack.
(Default: 10)
NTPSA invokes the Differential Algebra (DA) package written in C++ and kindly
provided by Lingyun Yang (lyyang@lbl.gov).
Etienne Forest has written the wrapper to allow the use of both the legendary
DA package written in Fortran by Martin Berz (default) and this new DA
package of Lingyun Yang. It is expected that this DA package will allow for
the efficient calculation of a large number of DA parameters.
(Default: false)
SYMPRINT a flag to enable the printing of the check of symplecticity. It is recommended
to leave this flag set to TRUE.
(Default: true)
30.3 PTC CREATE LAYOUT
The PTC CREATE LAYOUT command creates the PTC-layout according to the specified integra-
tion method and fills it with the current MAD-X sequence defined in the latest USE command.
PTC CREATE LAYOUT, TIME=logical, MODEL=integer, METHOD=integer,
NST=integer, EXACT=logical, OFFSET DELTAP=double,
ERRORS OUT=logical, MAGNET NAME=string,
RESPLIT=logical, THIN=double, XBEND=double,
EVEN=logical;
The attributes are:
![](asset-c8.png)
198 CHAPTER 30. PTC SET-UP PARAMETERS
TIME a logical flag to control which coordinate system is being used.
(Default= true)
This option changes the canonical coordinate system depending whether the
calculation is done in 5D or 6D:
5D if TIME is true, the fifth coordinate is PT,pt= ∆E/p0c
if TIME is false, the fifth coordinate is DELTAP,δp= ∆p/p0
6D if TIME is true, the MAD-X coordinate system {−ct,pt}is used.
if TIME is false, the second PTC coordinate system {-pathlength,
δp}is used.
Note: at small energy (β0<< 1), momentum-dependent variables like disper-
sion will depend strongly on the choice of the logical input variable ”time”. In
fact, the derivative ( ∂
∂δp) and ( ∂
∂pt) are different by the factor β0. One would
therefore typically choose the option ”time=false”, which sets the fifth variable
to the relative momentum deviation δp.
MODEL an integer to switch between models:
1 for ”Drift-Kick-Drift”; (Default value)
2 for ”Matrix-Kick-Matrix” and
3 for ”Delta-Matrix-Kick-Matrix” (SixTrack-code model).
METHOD the integration order: 2, 4, or 6 (See [30] Chapter K)
(Default: 2)
NST the number of integration steps. (Default: 1)
The body of each element is divided into NST equal slices and Forest-Yoshida
integration is carried out on each slice. For best results NST should increase
with strength and length of elements. The optimum NST value corresponds to
the value beyond which the studied properties no longer change. However, for
time consuming calculations the user may reduce NST. (See below the RESPLIT
option for automatic adjustment.)
This attribute sets the same NST value for all ”thick” elements (l > 0) of a
beam-line; however each individual element may also have its own NST value
defined independently (see below).
EXACT a logical flag to turn on calculations with an exact Hamiltonian, otherwise the
expanded Hamiltonian is used.
(Default: false)
OFFSET DELTAP [ Beware: Expert attribute! ] provides relative momentum deviation of
the reference particle (6D case ONLY). This option implies TOTALPATH=true.
(Default: 0.0)
ERRORS OUT a logical flag to write-out multipolar errors in EFCOMP table format.
(Default: false)
Two tables are created and filled: ”errors field” contains only field errors,
”errors total” contains also desired field components, which can include the
strength of correctors. The choice of magnets is defined by the MAGNET NAME
![](asset-c9.png)
30.4. PTC MOVE TO LAYOUT 199
attribute (see below). The tables can be written to file, and can be read back
via the ERRORS IN flag.
The ERRORS IN flag has precedence over this ERRORS OUT flag.
MAGNET NAME a string giving a simple selection for the names of magnet to be used for an
error write-out using the ERRORS OUT flag (see above). The errors are recorded
for all magnets with names starting with the exact string given here, which
would be equivalent to the ??? regular expression.
(Default: nil)
RESPLIT a logical flag to apply the PTC resplit procedure. This is meant to create an
”adaptive” setting of the METHOD and NST attributes according to the strengths
of quadrupoles (using the THIN attribute) and dipoles (using the XBEND at-
tribute). The EVEN attribute further controls the number of splits.
(Default: false)
THIN is the main RESPLIT attribute and is meant for splitting quadrupoles according
to their strength. The default value THIN=0.001 has shown in practice to work
well without costing too much with respect of performance.
XBEND is an optional RESPLIT attribute and is meant for splitting dipoles. A value
XBEND=0.001 is also advisable for dipoles.
(Default: -1.0 for no splitting)
EVEN a logical switch to ensure even number of splits when using the RESPLIT pro-
cedure of PTC, which is particularly useful when one attempts to calculate
PTC TWISS with the CENTER MAGNETS option, i.e. to calculate the TWISS pa-
rameters in the center of the element. Uneven number of splits is ensured with
EVEN=false. (Default: true)
30.4 PTC MOVE TO LAYOUT
Several PTC layouts can be created within a single PTC-”universe”. The layouts are auto-
matically numbered with sequential integers by the MAD-X code. The PTC MOVE TO LAYOUT
command is used to activate a specific layout, and the next PTC commands will be applied to
this active PTC layout until a new PTC layout is created or activated.
PTC MOVE TO LAYOUT, INDEX=integer;
The only attribute is:
INDEX is the numeric index of the PTC layout to be activated.
(Default: 1)
30.5 PTC READ ERRORS
The PTC READ ERRORS command reads any number of ”errors read” table through the
READMYTABLE mechanism.
![](asset-ca.png)
200 CHAPTER 30. PTC SET-UP PARAMETERS
PTC READ ERRORS, OVERWRITE=logical;
The only attribute is
OVERWRITE a flag to specify that the read-in errors overwrite previous errors instead of
adding the read-in errors to existing errors, ie multipole components already
present.
(Default: false)
Note:
Because of the way the table is read in memory, a warning will always be issued by default
in the form:
warning: string_from_table_row: row out of range: errors_read->name[1>=n+1<=n]
where nis the number of records read from the table. This warning has no consequence on
the errors read and the following calculation.
The warning is purely the result of the way that the reading loop is programmed with a break
based on the return value of the routine string from table row. But if string from table row
tries to read in a row (n+1) past the last row (n) of the table, it prints a warning before
returning a value that will effectively break the loop. Of course this will only happen if the
WARN option is true and this can be turned off with
OPTION, -WARN;
30.6 PTC ALIGN
The PTC ALIGN command is used to apply the MAD-X alignment errors to the current PTC
layout, and takes no attributes.
PTC ALIGN;
30.7 PTC END
The PTC END command turns off the PTC environment, which releases all memory and returns
control to the MAD-X world proper.
PTC END;
30.8 Additional Options for Physical Elements
For some of the MAD-X elements, additional attributes can be defined that are available to
PTC only. PTC also uses standard MAD-X attributes in a slightly different way.
SBEND | RBEND | QUADRUPOLE | SEXTUPOLE | OCTUPOLE | SOLENOID ,
L=real, ... , TILT=real, ... , NST=integer, ... ,
KNL={0, real, real,...}, KSL={0, real, real,...};
30.8. ADDITIONAL OPTIONS FOR PHYSICAL ELEMENTS 201
These attributes are:
Lthe length of the element.
PTC treats bending magnets (SBEND or RBEND) as MARKER if their length is equal
to zero.
NST gives a specific NST values for a particular ”thick” element (L > 0).
For example RF cavities are represented in MAD-X by a single kick, while PTC
splits the RF cavity into NST segments thereby taking into account properly the
transit-time effects of the cavity. Specifying explicitly NST=1 for RF cavity re-
produces in PTC the approximate results of MAD-X, ignoring transit time effects.
KNL, KSL The full range of normal and skew multipole components on the bench can
be specified for the following physical elements: sbend,rbend,quadrupole,
sextupole,octupole and solenoid.KNL and KSL multipole coefficients are
specified as the integrated value (RKds) of the field components along the
magnet axis. The multipole components in PTC are spread over the length of
thick elements. This is a considerable advantage of PTC input compared to
MAD-X which allows only thin multipoles.
KNL is an array representing the normal multipole coefficients.
(Default: 0 m−1)
KSL is an array representing the skew multipole coefficients.
(Default: 0 m−1)
To preserve the reference orbit of straight elements, the dipole components
for those elements are ignored and must be specified as zero: KNL(0)=0,
KSL(0)=0.
A full range of additional multipole field errors can be additionally specified
with the EFCOMP command. Errors are added to the above multipole fields on
the bench.
![](asset-cc.png)
Chapter 31. Thick-Lens Tracking Module
The PTC-TRACK module [32,31] is the symplectic thick-lens tracking facility in MAD-X. It is
based on PTC library [30] written by E. Forest. The commands of this module are described
below, optional parameters are denoted by square brackets ([]).
Prior to using this module the active beam line must be selected by means of a USE command.
The general PTC environment must also be initialized.
Examples
Several examples can be found on the web at http:cern.chfrsmad-X examplesptc track.
31.1 Synopsis
A typical tracking job in PTC requires a number of commands to be issued:
PTC CREATE UNIVERSE;
PTC CREATE LAYOUT, MODEL=integer, METHOD=integer, NST=integer, [EXACT];
...
PTC START, X=real, PX=real, Y=real, PY=real, T=real, PT=real;
PTC START, FX=real,PHIX=real, FY=real,PHIY=real, FT=real,PHIT=real;
...
PTC OBSERVE, PLACE=string;
...
PTC TRACK, ...;
...
PTC TRACKLINE, ...;
...
PTC TRACK END;
...
PTC END;
31.2 PTC START
To start particle tracking, a series of initial trajectory coordinates must be given with the
PTC START command; and as many commands as initial trajectories can be given.
PTC START commands must appear before the PTC TRACK command.
PTC START, X=real, PX=real, Y=real, PY=real, T=real, PT=real,
FX=real, PHIX=real, FY=real, PHIY=real, FT=real, PHIT=real;
The coordinates can be
X, PX, Y, PY, T, PT i.e. the standard canonical coordinates.
(Default: 0.0)
202
![](asset-cd.png)
31.3. PTC OBSERVE 203
FX, PHIX, FY, PHIY, FT, PHIT i.e. the action-angle coordinates which are expressed by
the normalized amplitude, Fzand the phase, Φzfor the z-th mode plane (z
= x, y, t). The actions are computed with the values of the emittances, Fz,
which must be specified in a preceding BEAM command. Fzare expressed in
number of r.m.s. beam sizes and Φzare expressed in radians.
(Default: 0.0)
Remarks
In the uncoupled case, the canonical and the action-angle variables are related with equations
z=Fz(Ez)1/2cos(Φz)pz=Fz(Ez)1/2sin(Φz) (31.1)
If both the canonical and the action-angle coordinates are given in the PTC START command,
they are summed after conversion of the action-angle coordinates to canonical coordinates.
The use of the action-angle coordinates requires the option CLOSED ORBIT in the PTC TRACK
command.
If the option CLOSED ORBIT in the PTC TRACK command is active (see above) all coordinates
are specified with respect to the actual closed orbit (possibly off-momentum with magnet
errors) and NOT with respect to the reference orbit. If the option CLOSED ORBIT is absent,
then coordinates are specified with respect to the reference orbit.
31.3 PTC OBSERVE
Besides the beginning of the beam-line, one can define an additional observation points along
the machine. Subsequent PTC TRACK command will then record the tracking data on all these
observation points.
PTC OBSERVE, PLACE=string;
The only attribute is
PLACE the name of observation point.
(Default: NULL)
Remarks
The first observation point at the beginning of the beam-line is marked as ”start”.
It is strongly recommended to specify markers as observation points.
The data at observation points other than ”start” can be produced in two different ways:
1. traditional element-by-element tracking. (See MAD-X thin tracking) which requires the
option ELEMENT BY ELEMENT of PTC TRACK to be active.
2. coordinate transformation from ”start” to the respective observation points using high-
order PTC transfer maps, which requires the option CLOSED ORBIT of PTC TRACK to be
active, and the options RADIATION and ELEMENT BY ELEMENT of PTC TRACK to be inac-
tive.
![](asset-ce.png)
204 CHAPTER 31. THICK-LENS TRACKING MODULE
31.4 PTC TRACK
The PTC TRACK command initiates trajectory tracking by entering the thick-lens tracking
module.
The tracking can be done element-by-element or ”turn-by-turn” with coordinate transforma-
tions over the whole turn.
Tracking is done in parallel, i.e. the coordinates of all particles are transformed through each
beam element, or over full turns.
A particle is lost if its trajectory is outside specified boundaries. In PTC, there is a continuous
check that the particle trajectories stay within the aperture limits.
The Normal Form calculation is controlled through options of the PTC TRACK command.
PTC TRACK, ICASE=integer, DELTAP=real, CLOSED ORBIT=logical,
ELEMENT BY ELEMENT=logical, TURNS=integer,
DUMP=logical, ONETABLE=logical, MAXAPER=real array,
NORM=integer, NORM OUT=logical,
FILE[=string], EXTENSION=string, FFILE=integer,
RADIATION=logical, RADIATION MODEL1=logical,
RADIATION ENERGY LOSS=logical,
RADIATION QUADR=logical,
BEAM ENVELOPE=logical, SPACE CHARGE=logical;
The attributes are:
ICASE the user-defined dimensionality of the phase-space (4, 5 or 6).
ICASE has higher priority over other options. In particular:
1. RF cavities with non-zero voltage are ignored for ICASE=4 or ICASE=5.
2. A non-zero DELTAP is ignored for ICASE=4 or ICASE=6.
However, if an RF cavity has voltage set to zero and ICASE=6 is specified, PTC
sets ICASE=4.
(Default: 4)
DELTAP the relative momentum offset for reference closed orbit (used for 5D case
ONLY).
DELTAP is ignored for ICASE=6, but the option OFFSET DELTAP of command
PTC CREATE LAYOUT may be used, if the reference particle should have a mo-
mentum offset.
(Default: 0.0)
CLOSED ORBIT a logical switch to activate the closed orbit calculation. This option must be
used for closed rings only. This option allows to activate the Normal Form
analysis, if required. With CLOSED ORBIT=false, the sequence is treated as a
transfer line.
(Default: false)
![](asset-cf.png)
31.4. PTC TRACK 205
ELEMENT BY ELEMENT a logical switch to activate the element-by-element tracking, from the
default turn-by-turn tracking.
(Default: false)
TURNS number of turns to be tracked.
(Default: 1)
DUMP a logical flag to enforce writing particle coordinates to formatted text files.
(Default: false)
ONETABLE a logical switch to write all particle coordinates to a single file instead of
separate files.
(Default: false)
MAXAPER an array defining upper limits for particle coordinates, essentially defining the
aperture to trigger particle loss.
(Default: {0.1, 0.01, 0.1, 0.01, 1.0, 0.1})
NORM NO order of the Normal Form.
NORM NO=1 makes the Normal Form linear (always true for MAD-X).
(Default: 1)
NORM OUT a logical switch to transform canonical variables to action-angle variables.
(Default: false)
FILE if FILE is omitted, no output is written to file.
if FILE is present, track tables are printed, optionally to files with name con-
structed from the base filename specified.
The actual name of the output file is constructed from tyhe baseline given with
FILE to which are appended the strings ”.obsnnnn” (where nnnn is the obser-
vation point index) and ”.pnnnn” (where nnnn is now the particle number),
unless the ONETABLE option is activated.
(Default: ”track”)
EXTENSION a string providing the filename extension for the track table files, e.g., txt,
doc...
(Default: nil)
FFILE defines the periodicity nof the printout: coordinates are printed every n turns.
(Default: 1)
RADIATION a logical flag to turn on the synchrotron radiation calculated by an internal
procedure of PTC.
The option RADIATION has precedence over RADIATION MODEL1 when both are
activated.
(Default: false)
RADIATION MODEL1 a logical flag to turn on the synchrotron radiation according to the method
given in [33]. This model simulates quantum excitation via a random number
generator and tables for photon emission. It can be used only with the option
ELEMENT BY ELEMENT.
(Default: false)
![](asset-d0.png)
206 CHAPTER 31. THICK-LENS TRACKING MODULE
RADIATION ENERGY LOSS a logical flag to add back the average energy loss thereby taking only
the quantum excitation into effect. It applies only when for RADIATION MODEL1
is active.
(Default: false)
RADIATION QUADR a logical flag to add the effect of synchrotron radiation in quadrupoles. It
supplements either model RADIATION or RADIATION MODEL1.
(Default: false)
BEAM ENVELOPE a logical switch to activate the calculation of the beam envelopes with PTC.
It requires the options RADIATION and ICASE=6.
(Default: false)
SPACE CHARGE [under construction]
a logical flag to activate the simulation of space charge forces between particles.
(Default: false)
31.5 PTC TRACKLINE
The PTC TRACKLINE command performs particle tracking that takes into account acceleration
in travelling wave cavities. It must be invoked in the scope of correctly initialized PTC
environment,i.e. after PTC CREATE UNIVERSE and PTC CREATE LAYOUT commands, and before
corresponding PTC END.
All tracks created with PTC START commands before PTC TRACKLINE command is issued are
tracked. Track parameters are dumped at every defined observation point (see PTC OBSERVE
command).
Please note that MAD-X always creates an observation point at the end of a sequence.
PTC TRACK LINE, TURNS=integer,
ONETABLE=logical, FILE=string, EXTENSION=string,
ROOTNTUPLE=logical,
EVERYSTEP=logical, TABLEALLSTEPS=logical,
GCS=logical;
The attributes are:
TURNS number of turns to be tracked. If the layout of the machine is not closed, this
value is forced to TURNS=1 by PTC
(Default: 1)
ONETABLE a logical switch to write all particle coordinates to a single file instead of
separate files.
(Default: false) [to be clarified]
FILE if FILE is omitted, no output is written to file.
if FILE is present, track tables are printed, optionally to files with name con-
structed from the base filename specified.
The actual name of the output file is constructed from tyhe baseline given with
![](asset-d1.png)
31.6. PTC TRACK END 207
FILE to which are appended the strings ”.obsnnnn” (where nnnn is the obser-
vation point index) and ”.pnnnn” (where nnnn is now the particle number),
unless the ONETABLE option is activated.
(Default: ”track”)
EXTENSION a string providing the filename extension for the track table files, e.g., txt,
doc...
(Default: nil)
ROOTNTUPLE a logical switch to store data to ROOT file as ntuple. Accessible only if RPLOT
plugin is available. i.e. only if MAD-X is dynamically linked and RPLOT plugin
is present.
(Default: false)
EVERYSTEP a logical switch to activate the recording of track parameters at every integra-
tion step. Normally tracking data are stored internally only at the end of each
element. EVERYSTEP provides the user with finer data points. It implies usage
of the so called node (thin) layout.
Track parameters are stored for each step in file ”thintracking ptc.txt”. Stor-
age of parameters in a table for each step might be very memory consuming.
To switch it off use option TABLEALLSTEPS.
Collective effects can be taken into account only using this mode (this feature
of PTC is not interfaced into MAD-X).
(Default: false)
TABLEALLSTEPS (to be completed)
(Default: false)
GCS a logical switch to store track parameters in Global Coordinate System - nor-
mally it starts at the entrance face of the first element.
(Default: false)
Plotting of track parameters (see PLOT command) is only possible if ONETABLE switch is set
to false (status as for Feb. 2006). This unfortunate solution is the legacy of the regular MAD-X
TRACK command, designed for circular machines where the user usually tracks a few particles
for many turns rather then many particles for one turn each.
Tracks that do not fit in the defined aperture for elements are immediately stopped.
Behavior of PTC calculations can be adapted with PTC SETSWITCH command and with ap-
propriate switches of PTC CREATE LAYOUT command.
31.6 PTC TRACK END
The PTC TRACK END command terminates the command lines related to the PTC TRACK mod-
ule.
PTC TRACK END;
![](asset-d2.png)
208 CHAPTER 31. THICK-LENS TRACKING MODULE
The initial and final canonical coordinates are collected in the internal table ”tracksumm”,
which can be written to file.
31.7 Choice of options
The following table facilitates the choice of the correct options for a number of typical tasks:
1. The tracking of a beam-line with default parameters.
2. Smilar to ”1.” but with element-by-element tracking and an output at observation
points.
3. Tracking in a closed ring with closed orbit search and the Normal Forms calculations.
Both canonical and action-angle input/output coordinates are possible.
Output at observation points is produced via PTC maps.
4. Similar to ”3.” except that output at observation points is created by element-by-element
tracking.
5. The ??? with PTC radiation.
Option case 1 case 2 case 3 case 4 case 5
CLOSED ORBIT - - + + +
ELEMENT BY ELEMENT - + - + -
PTC START, X, PX, ... ++++ +
PTC START, FX, PHIX, ... - - + + +
NORM NO - - >1>1>1
NORM OUT - - + - +
PTC OBSERVE - + + + -
RADIATION - - - - +
RADIATION MODEL1 - - - - -
RADIATION ENERGY LOSS - - - - -
RADIATION QUAD - - - - +/-
BEAM ENVELOPE - - - - -
SPACE CHARGE - - - - -
![](asset-d3.png)
Chapter 32. Ripken Optics Parameters
32.1 Introduction
The PTC TWISS module[31] of MAD-X is based on the PTC code and is supplementary to
the TWISS module of MAD-X. In PTC TWISS the Twiss parameters are calculated according to
the formalism of G. Ripken, developped in [34] and most accessible in [35].
PTC TWISS tracks a special representation of the beam in three degrees of freedom. It works
on the coupled lattice functions which are essentially the projections of the lattice functions
for the eigen-modes on the three planes.
PTC TWISS lists the projections of the ellipses of motion onto the three planes (x,px), (y,py),
(t,pt) expressed via Ripken’s parameters bk,j ,ak,j ,gk,j along with the phase advances mjin
selected positions, where index k= 1...3 refers to the plane (x,y, ...), and the index j= 1...3
denotes the eigen-mode.
The PTC TWISS command also calculates the dispersion values D1, ... ,D4.
In MAD-X commands and tables, these parameters are denoted as beta11, ..., beta33,
alfa11, ..., alfa33,gama11, ..., gama33,mu1, ..., mu3,disp1, ..., disp4, respec-
tively.
The Ripken parametrization can be transformed into the Edwards-Teng parametrization (used
in the module TWISS of MAD-X) using the formulae of Lebedev [36].
The parameters are noted as betx, bety, alfx, alfy and the coupling matrix: R11, R12,
R21 and R22. In absence of coupling, the following holds: betx = beta11,bety = beta22,
alfx = alfa11 and alfy = alfa22.
PTC TWISS can also compute the deltap/p-dependency of the Twiss parameters. The col-
umn names beta11p, ..., beta33p, alfa11p, ..., alfa33p, gama11p, ..., gama33p
denote the derivatives of the optics parameters with respect to deltap/p.
In order to evaluate the deltap/p-dependency of the Twiss parameters, the order (NO) of the
map must set to at least 2.
The derivatives of the dispersion with respect to deltap/p have column names: disp1p, ...,
disp4p. Second and third order derivatives have respective column names: disp1p2, ...,
disp4p2 for the second order, and disp1p3, ..., disp4p3 for the third order.
In addition, PTC computes the momentum compaction factor αcup to 1st order for ICASE=5,
or 3rd order (for ICASE=56). The values appear in the header of the PTC TWISS output file,
and a value of zero means the value has not been computed.
This feature is currently only available in the development version.
[To be checked]
For clarification: in the 4-D case, there is the following correspondence between MAD-X and
the Ripken’s notations: beta11 ≡βxI ,beta12 ≡βxII ,beta21 ≡βyI ,beta22 ≡βyII . In the
uncoupled 4-D case, beta11 is the same as the classical βx(betx) and beta22 is βy(bety),
209
![](asset-d4.png)
210 CHAPTER 32. RIPKEN OPTICS PARAMETERS
while beta12 and beta21 are zero. in the coupled case all betaNN are non-zero and beta11,
beta22 are distinctively different from βx,βy, respectively.
PTC TWISS also tracks the eigenvectors and prints them to Twiss table according to the SELECT
command with FLAG=ptc twiss. Either all 36 components or particular components of the
eigenvectors can be selected with EIGN or EIGNij, respectively (j= number of eigenvector,
i= number of coordinate {x,px,y,py,t,pt}).
For ring lattices, PTC TWISS computes momentum compaction, transition energy, as well as
other one-turn characteristics such as the tunes (Q1, Q2 and if ICASE=6 with cavity, Qs) and
chromaticities (for NO≥2).
Synopsis:
PTC CREATE UNIVERSE;
PTC CREATE LAYOUT, MODEL=integer, METHOD=integer, NST=integer, [EXACT];
...
SELECT, FLAG=ptc twiss, CLEAR;
SELECT, FLAG=ptc twiss, COLUMN=name, s,
beta11,...,beta33, alfa11,...,alfa33, gama11,...,gama33,
beta11p,...,beta33p, alfa11p,...,alfa33p, gama11p,...,gama33p,
mu1,...,mu3,
disp1,...,disp4, disp1p,...,disp4p,
disp1p2,...,disp4p2, disp1p3,...,disp4p3,
[eign], eign11,...,eign16,...,eign61,...,eign66;
...
PTC TWISS;
...
PTC END;
32.2 PTC TWISS
The PTC TWISS command causes computation of the Twiss parameters in Ripken’s style. It
operates on the working beam line defined in the latest USE command.
Applications for the PTC TWISS command are similar to the TWISS command of MAD-X. The
PTC TWISS can be applied to two basic tasks. It can calculate either a periodic solution or a
solution with initial conditions.
![](asset-d5.png)
32.2. PTC TWISS 211
PTC TWISS, ICASE=integer, DELTAP=real,
CLOSED ORBIT=logical, DELTAP DEPENDENCY=logical,
SLICE MAGNETS=logical,
RANGE=string, FILE[=filename], TABLE[=tabname],
INITIAL MATRIX TABLE=logical, INITIAL MATRIX MANUAL=logical,
INITIAL MAP MANUAL=logical, BETA0=string, MAPTABLE=logical,
IGNORE MAP ORBIT=logical, RING PARAMETERS=logical,
BETX=real, ALFX=real, MUX=real,
BETY=real, ALFY=real, MUY=real,
DX=real, DPX=real, DY=real, DPY=real,
X=real, PX=real, Y=real, PY=real, T=real, PT=real,
RE11=real, RE12=real, ..., RE16=real,
...
RE61=real, RE62=real, ..., RE66=real;
The attributes are:
ICASE the dimensionality of the phase-space (4, 5 or 6). (Default: 4)
Note that ICASE is internally set to 56 when attempting to set ICASE=6 with
no cavity, and ICASE is internally set to 4 when attempting to set ICASE=6
with an RF cavity with zero voltage.
NO the order of the map. (Default: 1)
For evaluating the derivatives of the Twiss parameter w.r.t. deltap/p, e.g. for
evaluating the chromaticities, the order must be at least equal to 2.
DELTAP relative momentum offset for reference closed orbit. (Default: 0.0)
CLOSED ORBIT a logical switch to trigger the closed orbit calculation (applies to periodic
solution ONLY).
(Default: false)
DELTAP DEPENDENCY a logical switch to trigger the computation of the Twiss and dispersion
derivatives w.r.t. deltap/p. Derivation formula assume that ICASE ≥5, so
that deltap/p is supplied as a parameter.
(Default: false)
SLICE MAGNETS a logical switch to activate the evaluation of Twiss parameters at each inte-
gration step inside magnets, in addition to the end face. The number of slices
is determined by the number of steps (NST) that can be separately defined
for each element, or otherwise set by NST parameter when creating the PTC
layout. Note that the orbit rms calculated in this mode counts as valid data
points both the end of the previous element and the entrance of the current el-
ement. Since the first integration node is always at the entrance of the magnet
(after position offset and fringe effects are calculated) which corresponds to
the same s position (and usually optical functions) as the end of the previous
element, the points at the interface between magnets are included twice in the
rms calculation.
(Default: false)
![](asset-d6.png)
212 CHAPTER 32. RIPKEN OPTICS PARAMETERS
CENTER MAGNETS a logical switch to activate the evaluation of Twiss parameters at the middle
of each magnet. This relies on internal slicing and ’integration nodes’ as deter-
mined by the number of steps (NST) selected when creating the PTC layout.
This number is assumed to be even otherwise the program issues a warning.
(Default: false)
FILE if the FILE attribute is omitted, no output is written to file.
If the FILE attribute name is present, the optional attribute value argument is
the name of the file for printing the PTC TWISS output. The default file name
is ”ptc twiss”.
(Default: false)
TABLE if the TABLE attribute is omitted, no output is written to an internal table.
If the TABLE attribute name is present, the optional attribute value argument
is the name of the internal table for PTC TWISS variables. The default table
name is ”ptc twiss”.
(Default: false)
SUMMARY FILE if the SUMMARY FILE attribute is omitted, no summary output is written to
file.
If the SUMMARY FILE attribute name is present, the optional attribute value
argument is the name of the file for printing the PTC TWISS SUMMARY table
output. The default file name is ”ptc twiss summary”.
(Default: false)
SUMMARY TABLE if the SUMMARY TABLE attribute is omitted, no summary output is written to
an internal table.
If the SUMMARY TABLE attribute name is present, the optional attribute value
argument is the name of the internal summary table for PTC TWISS SUMMARY
variables. The default table name is ”ptc twiss summary”.
(Default: false)
RANGE a string in RANGE format that specifies a segment of beam-line for the PTC TWISS
calculation.
(Default: #S/#E)
INITIAL MATRIX TABLE a logical flag to trigger the reading of the transfer map from table
named ”map table” created by a preceding PTC TWISS or PTC NORMAL com-
mand. The table can be also read before hand from files using a READTABLE
command.
(Default: false)
INITIAL MATRIX MANUAL a logical flag to trigger the use of the input variables RE11, ...,RE66
as the transfer matrix.
(Default: false)
RE11,..., RE66 values of the 6 ×6 transfer matrix.
(Default: 6 ×6 unit matrix)
INITIAL MAP MANUAL a logical flag to trigger the use of an input map stored beforehand in
file ”fort.18”, e.g. by a previous initial run of PTC NORMAL).
![](asset-d7.png)
32.3. PERIODIC SOLUTION 213
(Default: false)
IGNORE MAP ORBIT a logical flag to ignore the orbit in the map and use the closed orbit
instead if requested, or the orbit defined by the starting point specified with
X, PX, Y, P, T, DT parameters otherwise.
(Default: false)
BETA0 the name of a BETA0 block containing the Twiss parameters to be used as
input. When ICASE=6, this information must be complemented by supplying
a value for BETZ on the PTC TWISS command line.
(Default: beta0)
MAPTABLE a logical flag to save the one-turn-map to table ”map table”. The one-turn-
map can then be used as starting condition for a subsequent PTC TWISS, see
INITIAL MATRIX TABLE parameter above.
(Default: false)
BETX, ALFX, MUX, BETY, ALFY, MUY, DX, DPX, DY, DPY :
Edwards and Teng [14]Twiss and dispersion parameters: βx,y,αx,y,µx,y,Dx,y,
Dpx,py.
(Default: 0)
RING PARAMETERS a logical flag to force computation of ring parameters (γtr,αc, etc.).
(Default: false)
X, PX, Y, PY, T, PT the canonical coordinates of the initial orbit.
(Default: 0.0)
32.3 Periodic Solution
This is the simplest form of the PTC TWISS command, which computes the periodic solution
for a specified beam line. It may accept all basic attributes described in PTC TWISS above.
PTC TWISS, ICASE=integer, DELTAP=real, CLOSED ORBIT=logical,
RANGE=string, FILE[=string], TABLE[=string];
32.4 Evaluation of Twiss parameters inside magnets
This computes the periodic solution for a specified beam line and evaluates the Twiss pa-
rameters at each thin-slice (a.k.a ”integration-node”) inside magnets. The number of such
integration-nodes is given by the number of steps (NST) selected when creating the PTC layout.
All other basic attributes described in PTC TWISS above may be selected.
PTC TWISS, ICASE=integer, DELTAP=real, CLOSED ORBIT=logical,
RANGE=string, FILE[=string], TABLE[=string],
SLICE MAGNETS=logical;
![](asset-d8.png)
214 CHAPTER 32. RIPKEN OPTICS PARAMETERS
Example:
An example is found in the PTC TWISS Examples repository.
32.5 Solution with Initial Conditions
Initial conditions can be supplied in different ways. Naturally only one of the methods below
can be used at a time, and they can not be mixed. In this mode it is assumed that the
lattice is a line and no ring parameters are evaluated (their values are set to -1000000), unless
RING PARAMETERS=true, which forces computation of closed solution for the resulting map.
If a closed solution does not exist, PTC reports an error and exits.
The following logic is programmed in PTC to identify the source of initial conditions:
IF ( INITIAL MATRIX TABLE=true && (a map-table exists)) THEN
use initial values from a Map-Table
ELSEIF ( INITIAL MAP MANUAL=true ) THEN
use initial values from a Given Map File
ELSEIF ( INITIAL MATRIX MANUAL=true ) THEN
use initial values from a Given Matrix
ELSEIF ( BETA0 block is given ) THEN
use initial values from a BETA0 block
ELSE
use initial values from Given Twiss parameters
ENDIF
32.5.1 Initial Values from the Given Twiss Parameters
PTC TWISS calculates a solution with initial conditions given by the Twiss parameters, which
are explicitly typed as attributes to the command. This case is also limited to uncoupled
motion of the preceding ring or beam-line.
PTC TWISS, ICASE=integer, DELTAP=real,
RANGE=string, FILE[=string], TABLE[=string],
BETX=real, ALFX=real, MUX=real,
BETY=real, ALFY=real, MUY=real,
DX=real, DPX=real, DY=real, DPY=real,
X=real, PX=real, Y=real, PY=real, T=real, PT=real;
Example:
An example is found in the PTC TWISS Examples in the folder ”Example2”.
![](asset-d9.png)
32.5. SOLUTION WITH INITIAL CONDITIONS 215
32.5.2 Initial Values from a Map-Table
PTC TWISS calculates a solution with initial conditions given as a map-table of preceding ring
or beam-line. It requires the input option INITIAL MATRIX TABLE and an existing map-table
in memory, as generated by a preceding PTC NORMAL command.
PTC TWISS, ICASE=integer, DELTAP=real,
RANGE=string, FILE[=string], TABLE[=string],
INITIAL MATRIX TABLE;
Example:
An example is found in the PTC TWISS Examples in the folder ”Example3”.
32.5.3 Initial Values from a Map-File
PTC TWISS calculates a solution with initial conditions given as a map-file (fort.18) obtained
from a preceding ring or beam-line. It requires the input option INITIAL MAP MANUAL and an
existing map-file in file ”fort.18”, as generated by a preceding PTC NORMAL command.
PTC TWISS, ICASE=integer, DELTAP=real,
RANGE=string, FILE[=string], TABLE[=string],
INITIAL MAP MANUAL;
Example:
An example is found in the PTC TWISS Examples in the folder ”Example3”.
32.5.4 Initial Values from a Given Matrix
PTC TWISS calculates a solution with initial conditions given by a matrix explicitly given as
attribute to the command. It requires the option INITIAL MATRIX MANUAL.MAD-X expects a
symplectic 6x6 transfer matrix as input.
PTC TWISS, ICASE=integer, DELTAP=real,
RANGE=string, FILE[=string], TABLE[=string],
INITIAL MATRIX MANUAL,
RE11=real, RE12=real, ... , RE16=real,
...
RE61=real, RE62=real, ... , RE66=real;
Example:
An example is found in the PTC TWISS Examples in the folder ”Example4”.
32.5.5 Initial Values from Twiss Parameters via BETA0-block
PTC TWISS calculates a solution with initial conditions given by Twiss parameters, which are
transferred from the BETA0 block. The data in the the BETA0 block have to be filled by a
![](asset-da.png)
216 CHAPTER 32. RIPKEN OPTICS PARAMETERS
combination of the SAVEBETA and TWISS commands of MAD-X for a preceding ring or beam-line.
This case is limited to uncoupled motion of the preceding machine.
PTC TWISS, ICASE=integer, DELTAP=real,
RANGE=string, FILE[=string], TABLE[=string],
BETA0=string ;
Example:
An example is found in the PTC TWISS Examples in the folder ”Example1”.
![](asset-db.png)
Chapter 33. Non-Linear Machine Parameters
The PTC NORMAL module ([31] and [37]) of MAD-X is based on PTC code. This module takes
full advantage of the PTC Normal Form analysis which is a considerable upgrade of what was
available with the Lie Algebra technique used in MAD-8. It allows to calculate dispersions,
chromaticities, anharmonicities and Hamiltonian terms to very high order. In fact, the order
is only limited by the available computer memory and computing time.
The number of terms per order increases with some power law. The internal MAD-X tables are
not adequate to keep such large amounts of data. On the other hand, only a reduced set of
this data is actually needed by the user. Thus a much easier and flexible solution is to gather
user requirements with a series of SELECT PTC NORMAL command. A special MAD-X table is
dynamically built using those commands and is filled by a subsequent call to PTC NORMAL.
Another essential advantage of this table is that it is structured to facilitate exchange of
Normal Form (including Hamiltonian terms of high order) between MAD-X modules. The
immediate goal is to use this table to allow non-linear matching inside the present MAD-X
MATCHING module.
Synopsis:
PTC CREATE UNIVERSE;
PTC CREATE LAYOUT, MODEL=integer, METHOD=integer, NST=integer, [EXACT];
...
SELECT PTC NORMAL, DX, ..., GNFU;
...
PTC NORMAL;
WRITE, TABLE=normal results, FILE=normal results;
...
PTC END;
33.1 SELECT PTC NORMAL
The SELECT PTC NORMAL command selects the parameters to be calculated by a subsequent
PTC NORMAL command.
SELECT PTC NORMAL, DX=integer, DPX=integer, DY=integer, DPY=integer,
Q1=0, Q2=0, DQ1=integer, DQ2=integer,
ANHX=integerarray, ANHY=integerarray,
GNFU=integer,0,0, HAML=integer,0,0,
EIGN=integer,integer;
The attributes are:
DX, DPX, DY, DPY the dispersion paramaters specified as integer numbers specifying their
order: D(n)
x=∂(n)xco/∂δ(n)
p,D(n)
px =∂(n)pxco/∂δ(n)
p,D(n)
y=∂(n)yco/∂δ(n)
p,
D(n)
py =∂(n)pyco/∂δ(n)
p, where co is abbreviation of ”closed orbit”.
217
![](asset-dc.png)
218 CHAPTER 33. NON-LINEAR MACHINE PARAMETERS
Q1, Q2 horizontal and vertical tune paramaters are fixed to order 0: Q(0)
1,Q(0)
2
DQ1, QD2 the tune derivative parameters specified as integer numbers specifying their
order. ∂(n)Q1/∂δ(n)
p,∂(n)Q2/∂δ(n)
p
ANHX, ANHY the anharmonicities, each defined by three integer numbers: the order n1of 1,
the order n2of 2and the order n3of δp.∂(n1+n2+n3)Qz/(∂(n1)
1∂(n2)
2∂δ(n3)
p).
For example, ANHX=2,0,0 represents second order in 1:∂(2)Q1/∂(2)
1. And
ANHY=3,1,2 represents ∂(6)Q2/(∂(3)
1∂(1)
2∂δ(2)
p).
EIGN components of the eigenvectors at the end of the structure can be specified
by two integers: the eigenvector number and the coordinate coded in the list:
{x, px, y, py, t, pt}:
The pair n1,n2defines the n2-th component of the n1-th eigenvector.
GNFU The Generating Function can be specified by {n1,n2,n3}. The positive and
negative values of n−1 define the order of upright or skew resonances respec-
tively. The integers n2and n3are reserved for a future upgrade and must be
set to 0.
For example GNFU=-5,0,0 calculates all Generating Function terms for skew
decapoles. In the output table the cosine, sine and amplitude coefficients are
denoted as ”GNFC”, ”GNFS” and ”GNFA” respectively.
HAML the Hamiltonian terms can be specified by {n1,n2,n3}The positive and neg-
ative values of n1define the order of upright or skew resonances, respectively.
The integers n2and n3are reserved for a future upgrade and must be set to 0.
For example, HAML=3,0,0 calculates all Hamiltonian terms for upright sex-
tupoles. In the output table the cosine, sine and amplitude coefficients are
denoted as ”HAMC”, ”HAMS” and ”HAMA” respectively.
Caution: if more than one order of terms is selected only the lower one is correct because
higher orders contain ”cross terms” from the lower ones.
33.2 PTC NORMAL
The calculation of the parameters specified by the preceding SELECT PTC NORMAL commands
is initiated by the PTC NORMAL command, which operates on the working beam line defined
in the latest USE command.
PTC NORMAL, ICASE=integer, NORMAL=logical, CLOSED ORBIT=logical,
NO=integer, MAPTABLE=logical, DELTAP=double;
The attributes are:
ICASE user-defined dimensionality of the phase-space (4, 5 or 6).
(Default: 4)
NO the order of the map.
(Default: 1)
![](asset-dd.png)
33.2. PTC NORMAL 219
CLOSED ORBIT a logical switch to turn on the closed orbit calculation.
(Default: false)
DELTAP relative momentum offset for reference closed orbit.
(Default: 0.0)
MAPTABLE a logical flag to activate the storage of the map-table in memory.
MAPTABLE=true and NO=1 creates the one-turn matrix which can be used by
the next PTC TWISS command.
(Default: false)
NORMAL a logical flag to activate the calculation of the Normal Form.
(Default: false)
Example
The simple example is located on the Web-page for the PTC NORMAL example.
![](asset-de.png)
Chapter 34. MAD-X-PTC Auxiliaries
This chapter documents the interface between MAD-X and PTC and the auxiliary commands
available in the PTC library.
Available Commands
•PTC SETSWITCH
•PTC KNOB
•PTC SETKNOBVALUE
•MATCH WITHPTCKNOBS (Under Construction)
•PTC PRINTPARAMETRIC
•PTC EPLACEMENT
•PTC PRINTFRAMES
•PTC SELECT
•PTC SELECT MOMENT
•PTC MOMENTS (Under Construction)
•PTC DUMPMAPS
•PTC SETCAVITIES
220
![](asset-df.png)
34.1. PTC SETSWITCH 221
34.1 PTC SETSWITCH
Routine that sets the internal PTC switches.
PTC SETSWITCH, DEBUGLEVEL=integer,
MAXACCELERATION=logical,
EXACT MIS=logical,
TOTALPATH=logical,
RADIATION=logical,
FRINGE=logical,
TIME=logical;
Using this command the user can set switches of PTC and the MAD-X-PTC interface, adapting
this way the program behavior to his needs.
Command parameters and switches:
DEBUGLEVEL (Default: 1)
Sets the level of debugging printout: 0 prints none, 4 prints everything
MAXACCELERATION (Default: true)
Switch to set cavities phases so the reference orbit is always on the crest, i.e.
gains max energy
EXACT MIS (Default: .false.)
Switch ensures exact misalignment treatment.
TOTALPATH (Default: .false.)
If true, the 6th variable of PTC, i.e. 5th of MAD-X, is the total path.
If false it is deviation from the reference particle, which is normally the closed
orbit for closed layouts.
RADIATION (Default: false)
Sets the radiation switch/internal state of PTC.
FRINGE (Default: false)
Sets the fringe switch/internal state of PTC.
If true the influence of the fringe fields is evaluated for all elements.
Please note that currently fringe fields are always taken into account for some
elements (e.g. traveling wave cavities) even if this flag is set to false. The
detailed list of elements will be provided later, when the situation in this
matter will be definitely settled.
TIME (Default: true)
If true, Selects time of flight (cT to be precise) rather than path length as the
6th variable of PTC, i.e. 5th of MAD-X.
![](asset-e0.png)
222 CHAPTER 34. MAD-X-PTC AUXILIARIES
34.2 PTC KNOB
PTC KNOB, ELEMENTNAME=string,
KN=integer{, integer}, KS=integer{, integer},
EXACTMATCH=logical;
Sets knobs in PTC calculations. This is currently valid only in PTC TWISS;PTC NORMAL will
follow.
Knobs appear as additional parameters of the phase space. Twiss functions are then obtained
as functions of these additional parameters (Taylor series). Map elements may also be stored
as functions of knobs. The PTC SELECT command description shows how to request a given
element to be stored as a Taylor series.
The parametric results can also be:
1. written to a file with PTC PRINTPARAMETRIC.
2. plotted and studied using rviewer command (RPLOT plugin).
3. used to rapidly obtain approximate values of lattice functions for given values of knobs
(PTC SETKNOBVALUE). This feature is the foundation of a fast matching algorithm with
PTC.
Command parameters and switches:
ELEMENTNAME a string in range format (Default: NULL)
Specifies name of the element containing the knob(s) to be set.
KN,KS list of integers (Default: ???)
Defines which order
EXACTMATCH (Default: .true.)
Normally a knob is a property of a single element in a layout. The specified
name must match 1:1 to an element name. This is the case when exactmatch
is true.
Knobs might be also set to all family of elements. In such case the exactmatch
switch must be false. A given order field component of all the elements that
name starts with the name specified by the user become a single knob.
INITIAL ???
Example
dog leg chicane: Dipolar components of both rbends and dipolar and quadrupolar components
of the focusing quads set as knobs. Some first and second order map coefficients set to be
stored as parametric results. ptc twiss command is performed and the parametric results are
written to files in two formats.
dog leg chicane: Knob values are matched to get requested lattice functions.
![](asset-e1.png)
34.3. PTC SETKNOBVALUE 223
34.3 PTC SETKNOBVALUE
The PTC SETKNOBVALUE command sets a given knob value.
PTC SETKNOBVALUE, ELEMENTNAME=string,
KN=integer{,integer}, KS=integer{,integer},
VALUE=real;
All values in the twiss table used by the last PTC TWISS command and the columns specified
with PTC SELECT, PARAMETRIC=true; are reevaluated using the buffered parametric results.
The parameters of the command basically contain the fields that allow to identify uniquely
the knob and the value to be set.
Command parameters and switches:
ELEMENTNAME a string in range format that specifies the name of the element containing the
knob to be set.
(Default: none)
KN, KS are lists of integers that define the knob.
(Default: -1)
VALUE specifies the value to which the knob is to be set.
(Default: 0)
Example:
dog leg chicane: strength of dipole field component in quadrupoles is matched to obtain the
required R56 value.
34.4 PTC VARYKNOBS (Under Construction)
The PTC VARYKNOBS command allows matching with PTC knobs.
PTC VARYKNOB, INITIAL=string, ELEMENT=string,
KN=integer{,integer}, KS=integer{,integer},
EXACTMATCH=logical, TRUSTRANGE=real,
STEP=real, LOWER=real, UPPER=real;
where the attributes are
INITIAL
ELEMENT
KN, KS
EXACTMATCH
TRUSTRANGE defines the range over which the expansion is trusted
(Default: 0.1)
![](asset-e2.png)
224 CHAPTER 34. MAD-X-PTC AUXILIARIES
STEP
LOWER
UPPER
This matching procedure takes advantage of the parametric results that are accessible with
PTC. Namely, parameters occurring in the matching constrains are obtained as functions
(polynomials) of the matching variables. In other words, each variable is a knob in PTC
calculation. Evaluation of the polynomials is relatively fast comparing to the regular PTC
calculation which makes finding the minimum with the parametrized constraints very fast.
However, the algorithm is not faster in a general case:
1. The calculation time dramatically increases with the number of parameters and at
some point penalty rising from this overcomes the gain we get from the fast polynomial
evaluation.
2. A parametric result is an approximation that is valid only around the nominal parameter
values.
The algorithm is described below.
MATCH, use_ptcknobs=true;
...
PTC_VARYKNOB:
initial = [s, none] ,
element = [s, none] ,
kn = [i, -1],
ks = [i, -1],
exactmatch = [l, true, true],
trustrange = [r, 0.1],
step = [r, 0.0],
lower = [r, -1.e20],
upper = [r, 1.e20];
...
END_MATCH;
For user convenience the limits are specified in the MAD-X units (k1,k2, etc). This also ap-
plies to dipolar field where the user must specify limits of K0 = angle/path length. This
guarantees consistency in treatment of normal and skew dipole components.
Important: Note that inside the code skew magnets are represented only by normal compo-
nent and tilt, so the nominal skew component is always zero. Inside PTC tilt can not become a
knob, while skew component can. Remember about this fact when setting the limits of skew
components in the matching. When the final results are exported back to MAD-X, they are
converted back to the ”normal” state, so the nominal skew component is zero and tilt and
normal component are modified accordingly.
Example
dog leg chicane.
![](asset-e3.png)
34.5. PTC PRINTPARAMETRIC 225
Algorithm
1. Buffer the key commands (ptc varyknob,constraint,ptc setswitch,ptc twiss,
ptc normal, etc) appearing between match, useptcknobs=true; and any of matching
actions calls (migrad, lmdif, jacobian, etc)
2. When matching action appears,
(a) set ”The Current Variables Values” (TCVV) to zero
(b) perform THE LOOP, i.e. points 3-17
3. Prepare PTC environment (ptc createuniverse, ptc createlayout)
4. Set the user defined knobs (with ptc knob).
5. Set TCVV using ptc setfieldcomp command.
6. Run a PTC command (twiss or normal).
7. Run a runtime created script that performs a standard matching; all the user defined
knobs are variables of this matching.
8. Evaluate constraints expressions to get the matching function vector (I).
9. Add the matched values to TCVV.
10. End PTC session (run ptc end).
11. If the matched values are not close enough to zeroes then goto 3.
12. Prepare PTC environment (ptc createuniverse, ptc createlayout).
13. Set TCVV using ptc setfieldcomp command.
( — please note that knobs are not set in this case – )
14. Run a PTC command (twiss or normal).
15. Evaluate constraints expressions to get the matching function vector (II).
16. Evaluate a penalty function that compares matching function vectors (I) and (II).
See points 7 and 14.
17. If the matching function vectors are not similar to each other within requested precision
then goto 3.
18. Print TCVV, which are the matched values.
34.5 PTC PRINTPARAMETRIC
Editor’s Note: This command exists but is not documented.
The file PTC PrintParametric.html contains the same content as PTC Knob.html. I presume
the original file has been lost and overwritten...
![](asset-e4.png)
226 CHAPTER 34. MAD-X-PTC AUXILIARIES
34.6 PTC EPLACEMENT
Places a given element at required position and orientation. All rotations are made around
the front face of the element.
PTC EPLACEMENT, RANGE=string, REFFRAME=string,
X=real, Y=real, Z=real, PHI=real, THETA=real,
ONLYPOSITION=logical, ONLYORIENTATION=logical,
AUTOPLACEDOWNSTREAM=logical, SURVEYALL=logical;
Command parameters and switches:
RANGE a string in range format that specifies the name of the element to be moved.
REFFRAME defines the coordinate system with respect to which coordinates and angles are
specified.
Possible values are:
gcs global coordinate system (Default)
current current position
previouselement end face of the previous element
X, Y, Z coordinates of the front face of the magnet.
(Default: 0.0)
PHI, THETA polar (in xz plane, around z axis) and azimuthal (around x axis) rotation angles
respectively.
(Default: 0.0)
ONLYPOSITION a flag to perform only translation changes, and orientation of element is left
unchanged.
(Default: false)
ONLYORIENTATION a flag to perform only rotation changes and position of element is left
unchanged.
(Default: false)
AUTOPLACEDOWNSTREAM a logical flag: if true all elements downstream are placed at default
positions with respect to the moved element; if false the rest of the layout stays
untouched.
(Default: true)
SURVEYALL a logical flag used essentially for debugging. If true, a survey of the entire line
is performed after element placement at new position and orientation.
(Default: true)
Example
Dog leg chicane: position of quadrupoles is matched to obtain required R566 value.
![](asset-e5.png)
34.7. PTC PRINTFRAMES 227
34.7 PTC PRINTFRAMES
Print the PTC geometry of a layout to a specified file.
PTC PRINTFRAMES, FILE=filename, FORMAT=string;
Command parameters and switches:
FILE specifies the name of the file.
FORMAT specifies the format of geometry data. Currently two formats are accepted:
text prints a simple text file. (Default)
rootmacro creates ROOT macro that can be used to produce a 3D display
of the geometry.
Example
Dog leg chicane with some elements displaced with help of PTC EPLACEMENT.
34.8 PTC SELECT
Selects a map element to be stored in a user-defined table, or stored as a function (Taylor
series) of a defined knob. Both cases can be joined in a single PTC SELECT command.
PTC SELECT, TABLE=tabname, COLUMN=string,string,
POLYNOMIAL=integer, MONOMIAL=string, PARAMETRIC=logical,
QUANTITY=string;
Command parameters and switches:
TABLE the name of the table where values should be stored.
COLUMN the name of the column in the table where values should be stored.
POLYNOMIAL specifies the row of the map.
MONOMIAL a string composed of digits that defines the monomials of the polynomial in
PTC nomenclature. The length of the string should be equal to the number
of variables and each digit corresponds to the exponent of the corresponding
variable. Monomial ’ijklmn’ defines xipxjykpyl∆Tm(∆p/p)n.
For example, element=2 and monomial=1000000 defines coefficient of the sec-
ond polynomial (that defines px) close to x, in other words it is R21.
PARAMETRIC a logical switch. If true, and if any knobs is defined, the map element is stored
as the parametric result.
(Default: false)
QUANTITY ??? is that the element referred above ??
![](asset-e6.png)
228 CHAPTER 34. MAD-X-PTC AUXILIARIES
To store map elements in a user-defined table and column, the table with the named columns
should pre-exist the PTC SELECT command.
To store map elements as a function of a defined knob, the PARAMETRIC attribute must be set
to true.
Examples
dog leg chicane: strength of quads is matched to obtain required T112 value.
dog leg chicane: position of quads is matched to obtain required T566 value.
dog leg chicane: dipole and quadrupole strengths are matched with the help of knobs to
obtain required momentum compaction and Twiss functions.
34.9 PTC SELECT MOMENT
Selects a moment to be stored in a user-defined table, or stored as a function (Taylor series)
of a defined knob. Both cases can be joined in one command.
PTC SELECT MOMENT, TABLE=tabname, COLUMN=string,
MOMENT S=string,string, MOMENT=integer,
PARAMETRIC= logical;
Command parameters and switches:
MOMENT S a list of coma separated strings, each composed of up to 6 digits defining the
moment of a polynomial in PTC nomenclature:
the string ’ijklmn’, where i,j,k,l,m,n are digits from 0 to 9, defines the moment
< xipxjykpyl∆Tm(∆p/p)n>.
For example, MOMENT S=100000 defines < x1>
Note that for input we always use MAD-X notation where dp/p is always the
6th coordinate. Internally to PTC, dp/p is the 5th coordinate. We perform
automatic conversion that is transparent for the user. As the consequence
RMS in dp/p is always denoted as the string ’000002’, even in 5D case.
This notation allows to define more then one moment with a single command.
In this case, the corresponding column names are built from the string argu-
ments to MOMENT S with a mu prefix. However, they are always extended to 6
digits, i.e. trailing 0’s are automatically added.
For example, with MOMENT S=2, defines < x2>and the corresponding column
name is mu200000.
This method does not allow to pass bigger numbers larger than 9. In order to
define such a moment, see the attribute MOMENT below.
![](asset-e7.png)
34.10. PTC MOMENTS 229
MOMENT a list of up to 6 coma separated integers that define the moment:
< xipxjykpyl∆Tm(∆p/p)n>being defined as MOMENT=i,j,k,l,m,n
(Default: 0)
For example: MOMENT=2 defines < x2>,MOMENT=0,0,2 defines < y2>,
MOMENT=0,14,0,2 defines < px14py2>, etc.
COLUMN defines the name of the column where values should be stored. If not specified
then it is automatically generated from the MOMENT definition:
< xipxjykpyl∆Tm(∆p/p)n>=>mu i j k l m n
(where numbers are separated with underscores).
This attribute is ignored if MOMENT S is specified.
(Default: none)
TABLE specifies the name of the table where the calculated moments are stored.
(Default: moments)
PARAMETRIC a logical flag to to store the element as a parametric result if a knob has been
defined.
(Default: false)
To store a moment in a user-defined table and column, the table with the named columns
should pre-exist the PTC SELECT MOMENT command.
To store a moment as a function of a defined knob, the PARAMETRIC attribute must be set to
true.
Examples
from ATF2:
!Here is sigmax**2
ptc select moment, table = momord2, moment s= 2;
!Below are example how to encode other moments
ptc select moment, table = momord2, moment s= 20,11,02,002,0011,0002,00002;
ptc select moment, table = momord2xy, moment s= 1010,0110,1001,0101,10001,
01001,00101,00011;
ptc select moment, table = momord4, moment s= 40,22,04, 004,0022,0004;
ptc select moment, table = momord6, moment s= 6;
34.10 PTC MOMENTS
The command PTC MOMENTS calculates the moments previously selected with the PTC SELECT MOMENT
command. It uses maps saved by the PTC TWISS command, hence, the SAVEMAPS switch of
PTC TWISS must be set to true (Default) to be able to calculate moments.
PTC MOMENTS, NO=integer,
XDISTR=string, YDISTR=string, ZDISTR=string;
The command parameters and switches are
![](asset-e8.png)
230 CHAPTER 34. MAD-X-PTC AUXILIARIES
NO order of the calculation, maximally twice the order of the last twiss.
(Default: 1)
XDISTR, YDISTR, ZDISTR define the distribution in x, y and z dimension respectively and
can take one of the following values:
gauss Gaussian distribution (Default)
flat5 flat distribution in the first of variables (dp over p) of a given
dimension and Delta Dirac in the second one (T)
flat56 flat rectangular distribution
Examples
ATF2
34.11 PTC DUMPMAPS
PTC DUMPMAPS dumps the linear part of the map for each element of the layout into the
specified file.
PTC DUMPMAPS, FILE=filename;
The only command parameter is:
FILE the filename of the file to which the matrices are dumped.
(Default: ptcmaps)
34.12 PTC SETCAVITIES
The PTC SETCAVITIES command adjusts cavities and sets appropriate reference momenta for
a layout containing travelling wave cavities.
PTC SETCAVITIES;
The main goal is to update the reference beam energy for the elements that follow a travelling
wave cavity. PTC traces the synchronous particle, that is the particle that has all its parameters
set to zero at the beginning of the layout under study.
When PTC reaches a cavity in the layout, the parameters of the cavity may be adjusted
according to the user-defined MAXACCEL switch previously set in PTC SETSWITCH.
If MAXACCEL=true the phase of the cavity is adjusted so it gives the maximum acceleration.
The phase lag is then added to this adjusted phase.
If MAXACCEL=false the cavity parameters are left unchanged.
The synchronous particle is then tracked through the travelling wave cavity and the energy
gain is calculated. This energy becomes the new reference energy for all elements downstream
of the cavity.
![](asset-e9.png)
34.12. PTC SETCAVITIES 231
This process is repeated at every cavity encountered further in the tracking trough the layout.
Parameters of the cavities are printed to a file named ”twcavsettings.txt”.
Attention:
in PTC the phase velocity of a cavity wave is always equal to the speed of light. Hence, if PTC
internal state TIME is true, which is the most correct setting, the voltage seen by a particle
is varying along the structure. If TIME=false, the tracked particle is assumed to propagate
at the speed of light (v=c) and the particle moves synchronously with the wave front.
Attention:
For programming reasons, any element that changes the reference momentum, i.e. travelling
wave cavities, must be followed by a marker. If a marker does not follow immediately each
of these elements, PTC detects an error and stops the program. Hence two cavities cannot be
placed one immediately after the other and a marker must be inserted in between.
Part VI
Trailing Material
232
Chapter 35. Known Differences to Other Programs
35.1 Definitions in MAD-8
MAD-8 uses full 6-by-6 matrices to allow coupling effects to be treated, and the canonical
variable set (x,px/p0), (y,py/p0), (−ct ,δE/p0c), as opposed to other programs most of
which use the set (x,x0), (y,y0), (−δs), δp/p0).
Like Dragt [38], MAD-8 uses the relative energy error p/p0, which is equal the relative momen-
tum error delta =δp/p0multiplied by β=v/c.
As from Version 8.13, MAD-8 used an additional constant momentum error deltasin all
optical calculations. The transfer maps contained the exact dependence upon this value;
therefore the tunes for large deviations could be computed with high accuracy as opposed to
previous versions.
The choice of canonical variables in MAD-X still leads to slightly different definitions of the
lattice functions. In MAD-X the Courant-Snyder invariants mentioned in [3] take the form
Wx=γxx2−2αxxpx+βxp2
x
Comparison to the original form
Wx=γxx2−2αxxx0+βxx02
shows that the orbit functions cannot be the same.
A more detailed analysis, using x0=px/(1+δ) shows that all formulas can be made consistent
by defining the MAD orbit functions (index M) with respect to the Courant and Snyder
functions (index C) as follows:
βxM =βxC ×(1 + δ),
αxM =αxC ,
γxM =γxC /(1 + δ)
For constant δsalong the beam line and δ= 0, the lattice functions are the same. In a
machine where δvaries along the circumference, e.g. in a linear accelerator or in an electron-
positron storage ring, the definition of the Courant-Snyder invariants must be generalised.
The MAD-8 invariants have the advantage that they remain invariants along the beam line
even for variable δ.
With the new method this problem occurs in Twiss module (see 19) only for non-constant δ.
35.2 Treatment of Energy Error in TWISS
It has been noted by Milutinovic and Ruggiero [39] that MAD-8 returned tunes which are
too low for non-zero δ. The difference was found to be quadratic in delta with a negative
coefficient. This problem has been eliminated thanks to the new treatment of momentum
errors from MAD-8 Version 8.13 onwards.
233
![](asset-ec.png)
Chapter 36. MAD-X pitfalls
Find a loose collection of pitfalls that may be difficult to avoid in particular for new users but
also experienced user might profit from this list.
Twiss calculation is 4D only! The Twiss command will calculate an approximate 6D
closed orbit when the accelerator structure includes an active cavity. However, the
calculation of the Twiss parameters are 4D only. This may result in apparently non-
closure of the beta values in the plane with non-zero dispersion. The full 6D Twiss
parameters can be calculated with the PTC TWISS command. The Thinlens Tracking
module presently suffers from this deficiency since it requires the true 6d closed orbit
and not the approximate one as calculated by Twiss. In this context one has to mention
that the coordinate system for the Twiss module is not x, px in the horizontal plane as
the advertised canonical coordinates instead x, x’ have been used (same for the vertical
plane).
Be careful that for TWISS with the CENTRE attribute activated, i.e. looking inside the
element, the closed orbit includes the misalignment of the element.
Dispersion for machines with small relativistic beta MAD-X uses the PT coordinate as
the canonical momentum in the longitudinal plane. The derivative of e.g. dispersion
is therefore not taken wrt delta-p over p but PT. Therefore one unfortunately finds
the dispersion being divided by the relativistic beta which is annoying for low energy
machines. PTC allows to change the coordinate system to delta-p over p with the
”time=false” option of the PTC CREATE LAYOUT command which delivers the proper
dispersion with the
Non-standard definition of DDX, DDPX, DDY, DDPY The MAD-X proper defintion
of DDX, DDPX, DDY, DDPY is not the second order derivative with respect to deltap/p
but multiplied by a factor of 2. The corresponding values from PTC NORMAL and in
PTC TWISS are the proper derivaties to all orders.
Chromaticity calculation in presence of coupling Chromaticity calculations are typi-
cally in order and agree with PTC and other codes. However, it was recently discovered
that in presences of coupling MAD-X simply seems to ignore coupling when the chromatic-
ity is calculated. This is surprising since the eigentunes Q1, Q2 are properly calculated
for a given (small!) dp/p. The issue is under investigation.
Field errors in thick elements Only a very limited number of field error components are
considered in TWISS calculations for some thick elements. Find below a complete list
of all those field error components that are taking into account for a particular thick
element. It should be mentioned that BENDs also allow a skew quadrupole component
K1s but NOT in the body of the magnet. It is only active in the edge effect for radiation
(expert use only).
234
![](asset-ed.png)
235
Magnet Type Normal Field Components Skew Field Components
Dipole —
Bend Quadrupole —
Sextupole —
HKicker Dipole —
VKicker — Dipole
Quadrupole Quadrupole Quadrupole
Sextupole Sextupole Sextupole
Octupole Octupole Octupole
MAD-X versus PTC The user has to understand that PTC exists inside of MAD-X as a library.
MAD-X offers the interface to PTC, i.e. the MAD-X input file is used as input for PTC.
Internally, both PTC and MAD-X have their own independent databases which are linked
via the interface. With the PTC CREATE LAYOUT command, only numerical numbers
are transferred from the MAD-X database to the PTC database. Any modification to the
MAD-X database is ignored in PTC until the next call to PTC CREATE LAYOUT For example,
a deferred expression of MAD-X after a PTC CREATE LAYOUT command is ignored within
PTC.
When introducing a cavity with the HARMON attribute instead of the FREQ attribute
(highly discouraged!) a problem arises for PTC TWISS due to the fact that internally
HARMON is transferred to FREQ too late. A simple TWISS command executed before PTC
start-up will help. However, avoiding HARMON is advantageous.
SLOW attribute in matching The SLOW attribute enforces the old matching procedure
and is considerably slower. Therefore we did not make it the default option. Recently a
number of parameters, like RE56, have been added to the list of matchable parameters
in the default and fast version. Nevertheless, some parameters are only available when
using the SLOW attribute. Therefore it is advisable to check with the SLOW attribute if
there are doubts about the matching procedure.
Validity of Twiss parameters The standard Teng-Edwards Twiss parameters suffer from
a deficiency near full coupling: i.e. the ”donuts” of linear motion in x-x’ and y-y’ phase
space have no hole anymore. This means that all energy is transfered from one plane
to the other. In this case the Twiss parameters and the coupling matrix (R11, R12,
R21, R22) become large or even infinite or the beta functions might become negative.
The Ripken-Mais Twiss parameters are always well defined (they are the ”average”
amplitude functions of their proper phase space region), i.e. at full coupling we have:
beta11 ∼beta12 and beta21 ∼beta22. Using the RIPKEN flag TWISS calculates the
![](asset-ef.png)
Chapter 37. Contributors to MAD-X
Lists are provided in alphabetical order.
Disclaimer: any omissions in these lists are accidental.
Feel free to contact mad support (mad@cern.ch) if you have been left out or some of your
contributions are not listed.
37.1 MAD team
The list includes all the persons officially working on the MAD-X project.
Laurent Deniau: project manager since june 2011.
Andrea Latina: since october 2011.
Ghislain Roy: since march 2013.
Piotr Skowronski: since january 2012.
37.2 Module keepers and contributors
This list includes all the volunteers who accepted to maintain and develop some of the MAD-X
modules.
Fanouria Antoniou: ibs.
Helmut Burkhardt: makethin.
Riccardo De Maria: match2.
Laurent Deniau: survey.
Valery Kapin (FNAL, Chicago): ptc track.
Emanuele Laface (ESS, Sweden): match.
Andrea Latina: trac, twiss.
Yannis Papaphilippou: dynap.
Ghislain Roy: aperture, error, cororbit, touschek, c6t, plot.
Piotr Skowronski: (ptc trac), ptc twiss, ptc normal, ptc module.
Rogelio Tomas: emit.
We are looking for volunteers, any help is welcome!
37.3 Special contributors
This list includes all the persons who have contributed exceptionally to the project in the
past.
237
![](asset-f0.png)
238 CHAPTER 37. CONTRIBUTORS TO MAD-X
Frank Schmidt: project custodian from 2002 to 2011.
Etienne Forest: author of PTC & FPP (written in F90).
Hans Grote: author of the core (written in C).
37.4 Other contributors
The list below includes the persons who have contributed significantly to the MAD-X project
in the past. Unless noted they were affiliated with CERN at the time of their contributions.
Ralph Assmann (emit)
Oliver Bruning (match)
Hans Grote (Core in C, plot)
Werner Herr (error, cororbit)
Bernard Jeanneret (aperture)
Alex Koschik (thintrack)
Nikolay Malitsky - BNL, New-York (sxf)
Eric McIntosh (memory leaks)
Jean Luc Nougaret (twiss, ptc twiss, C/Fortran wrappers, bug tracker)
Thys Risselada (threader, closed orbit)
Frank Schmidt (c6t, twiss, sodd, ptc twiss, ptc module, ptc normal)
Yipeng Sun (thintrack)
Frank Tecker (survey)
Andr´e Verdier (survey)
Lingyun Yang (tpsa)
Frank Zimmermann (dynap, touschek, ibs)
MAD-9 contributors
Christoph Iselin is the author of MAD-9 and a major contributor to the Classic library.
MAD-8 contributors
Hans Grote and Christoph Iselin are the authors of MAD-8.
![](asset-f1.png)
Change Log
since version 5.02.00
The following changes have been made to the code and documentation since August 15th,
2014 in version 5.02.02
The changes are indexed by date (most recent first) and provide the MAD-X version number
where the change applies as well as the SVN release number for the change.
2016-Feb-14 version 5.02.08, r5673
fixed a bug in EMIT whereby the coordinates of the orbit were mixed up
(through Fortran equivalence statements) between the vertical and longitu-
dinal plane, leading to wrong results in damping partition numbers and emit-
tances. In all tests in our test-suite the difference turned out to be very small
but other users might experience otherwise.
2015-Nov-03 version 5.02.07, r5484
the keyword VERSION has been introduced.
2015-Sep-15 version 5.02.07, r5407
Removed the precedence information in the BEAM command between the longi-
tudinal emittance (ET) and beams sizes (SIGT and SIGE) since this is actually
not implemented in the code.
2015-Sep-15 version 5.02.07, r5399
A new option NOEXPR has been added to the SAVE command to allow saving
sequences with only values and no expressions in variables and commands.
2015-Sep-04 version 5.02.07, r5380
Change of the trailing message printed on output that no longer mentions the
version number and version architecture. The same information can already
be found in the header message.
2015-Aug-31 version 5.02.07, r5367
TEAPOT is now the default style for MAKETHIN. The previous default style that
was used when the STYLE attribute was not specified in the MAKETHIN com-
mand has been given the name HYBRID and can still be used with the explicit
STYLE=hybrid attribute.
All MAKETHIN commands where STYLE is not specified now use the
TEAPOT style instead of the previously unnamed HYBRID style
2015-Jul-15 version 5.02.06, r5336
corrected a typo in equation 1.10 reported by Michael Severance (Stony Brook).
The B2factor of the expansion of the Bxfield was reading (xy −h3
6y3+···)
and has been corrected to (xy −h
6y3+···).
2015-Jun-09 version 5.02.06, r5250
added a guard against negative sequence length and negative element lengths
at the time of sequence expansion triggered by a USE command, with MAD-X
239
240 CHAPTER 37. CONTRIBUTORS TO MAD-X
then finishing with fatal error. No checks were performed so far on these
attributes, assuming that all length were positive.
2015-Jun-05 version 5.02.06, r5247
added the cardinal sine SINC(x) to the list of available operators in arithmetic
expressions.
2015-Mar-31 version 5.02.05, r5181
added a SHRINK command to remove rows at the end of an existing table.
2015-Mar-11 version 5.02.05, r5162
Major change to the definition of emittances in the BEAM command. For
historical reasons, there was a factor 4 in the relation between normalised
emittance nand geometric emittance :n= 4βγ , where βand γare the
usual relativistic factors.
The common definition n=βγ is now used across all MAD-X modules.
The APERTURE command now gets the geometric emittances from values in-
put or calculated in the BEAM command; the attributes EXN and EYN of the
APERTURE command have been removed together with their default value of
EXN = 2.75E-6 and EYN = 2.75E-6 corresponding to the standard normalized
emittances for LHC beams in collisions.
2015-Mar-10 version 5.02.05, r5161
Major change to definition of RACETRACK aperture type. The RACETRACK
aperture now refers to a generalized shape with rounding of corner with el-
lipse instead of circle. The APERTURE array now takes four arguments for the
RACETRACK shape: maximum horizontal extent, maximum vertical extent, hor-
izontal semi-axis and vertical semi-axis of ellipse for rounding the corner.
Note also that the definition of the first two arguments has changed
from horizontal and vertical offsets to horizontal and vertical maxi-
mum extensions.
Removed also all references in the code and the manual to the MARGUERITE
aperture type (two RECTCIRCLEs crossing at right angle) that has been depre-
cated for some time already.
2015-Feb-19 version 5.02.05, r5143
added the OCTAGON in the list of predefined APERTURE types.
2015-Feb-11 version 5.02.05, r5128
clarified that the NMASS constant is the unified atomic mass unit and not the
neutron mass. None of the constants have changed in 2014 PDG publication
with respect to the 2012 version [6]. Updated the reference to PDG publica-
tions to include 2014 version [5].
2015-Jan-28 version 5.02.05, r5118
clarified in the definition of magnetic elements that the effect of defined mag-
netic strengths is always the same, irrespective of the CHARGE of the particles
declared in the BEAM command. It is agreed in the literature that a positive
quadrupole (positive K1) focuses positive particles in the horizontal plane and
37.4. OTHER CONTRIBUTORS 241
defocuses negative particles in the same horizontal plane, for the same direc-
tion of propagation.
Currently MAD-X ignores the CHARGE attribute and focuses both positive and
negative particles in the horizontal plane when going through a quadrupole
with positive K1.
THIS MAY CHANGE IN THE FUTURE TO CONFORM TO EX-
ISTING CONVENTIONS
Electrostatic elements (ELSEPARATOR,RFCAVITY,CRABCAVITY, and the RF part
of the RFMULTIPOLE) handle the CHARGE attribute appropriately and provide
opposite effects for opposite charges travelling in the same direction.
2014-Dec-19 version 5.02.05, r5111
added the Gauss error function ERF and the complementary error function
ERFC to the list of available operators in arithmetic expressions. Added docu-
mentation in the same section for the FLOOR,CEIL and ROUND functions that
were already implemented.
2014-Dec-10 version 5.02.04, r5093 and r5101
clarified the global coordinate system figure 1.2 with colors and representa-
tions of projections of planes onto the horizontal Cartesian plane as well as
intersections of local coordinate planes with horizontal Cartesian plane.
2014-Nov-25 version 5.02.04, r5092
removed the GLOBAL matching constraints DDQ1, DDQ2 from the documentation
since they are not implemented in the code.
2014-Nov-14 version 5.02.04, r5081
added a COPYFILE command. Changed the attribute name for the destination
for the RENAMEFILE command from NAME to TO.
2014-Nov-13 version 5.02.04, r5078
fixed figure 24.3 where the x-axis was pointing in the wrong direction and the
orientation of the element for positive DPHI was also not conforming to the
text for the EALIGN command.
2014-Nov-13 version 5.02.04, r5080
documented a bug occurring when LINE or MACRO constructs appear within a
IF ... ELSEIF ... ELSE or a WHILE construct. This bug will not be fixed
now.
Clarified also that IF ... ELSEIF ... ELSE and WHILE constructs can be
nested to at least six levels deep.
2014-Oct-14 version 5.02.03, r5013
fixed a documented feature of SURVEY where the first KSL component of thin
MULTIPOLE elements, representing a vertical angle for a thin dipole, was not
taken into account. Both KNL and KSL are now properly taken into account.
Another change was to make SURVEY take into account the RFMULTIPOLE ele-
ments in the same way that it treats MULTIPOLE elements.
2014-Aug-27 version 5.02.03, r4947
changed the behaviour of FILL to accept as parameter a row number equal to
![](asset-f4.png)
242 CHAPTER 37. CONTRIBUTORS TO MAD-X
the current number of rows in the table plus one, with the effect of creating a
new row and filling it.
2014-Aug-25 version 5.02.03, r4942 and r4943
harmonized the behaviour of FILL,SETVARS and SETVARS LIN with respect to
negative row numbers, and updated the default values. Added documentation
sections for SETVARS and SETVARS LIN that were hitherto undocumented.
2014-Aug-18 version 5.02.03, r4932
a single element can now be repeated in a beamline expansion: 2*S and -2*S
are of course identical (single elements are not reversed head to tail), and also
equivalent to 2*(S) and -2*(S) if Sis a single element.
Documentation updated; see 12.3
Bibliography
[1] Karl L. Brown. A First-and Second-Order Matrix Theory for the Design of Beam Trans-
port Systems and Charged Particle Spectrometers. Technical Report SLAC-75, SLAC,
1972. Revision 3.
[2] Karl L. Brown, David C. Carey, Christopher Iselin, and Franck Rothacker. TRANS-
PORT - A Computer Program for Designing Charged Particle Beam Transport Systems.
Technical Report CERN 73-16, revised as CERN 80-04, CERN, 1980.
[3] E. D. Courant and H. S. Snyder. Theory of the alternating gradient synchrotron. Annals
of Physics, 3:1–48, 1958.
[4] Brian W. Montague. Linear Optics for Improved Chromaticity Correction. Technical
Report LEP Note 165, CERN, 1979.
[5] K.A. Olive et al. (Particle Data Group). The Review of Particle Physics. Chin. Phys.
C, 38:090001, 2014.
[6] J. Beringer et al. (Particle Data Group). The Review of Particle Physics. Physical
Review, D86:010001, 2012.
[7] V. Danilov and S. Nagaitsev. ??? Phys. Rev. ST Accel. Beams, 13:084002, 2010.
[8] Philippe Defert, Ph. Hofmann, and R. Keyser. The Table File System, the C Interfaces.
Technical Report LAW Note 9, CERN, 1989.
[9] Frank Schmidt. SixTrack, User’s Reference Manual. Technical Report SL/95-56 (AP),
CERN, 1994.
[10] SixTrack website. http://cern.ch/sixtrack.
[11] Frank Schmidt. SODD: A computer code to calculate detuning and distortion function
terms in first and second order. Technical Report CERN SL/Note 99-009 (AP), CERN,
1999.
[12] Mark Hayes and Frank Schmidt. Run Environment for SixTrack. Technical Report LHC
Project Note 300, CERN, July 2002.
[13] Hans Grote, J. Holt, N. Malitsky, Fulvia Pilat, Richard Talmann, and C. G. Trah-
ern. SXF (Standard eXchange Format): definition, syntax, examples. Technical Report
RHIC/AP/155, BNL, August 1998.
[14] D. A. Edwards and L. C. Teng. Parametrisation of linear coupled motion in periodic
systems. IEEE Trans. on Nucl. Sc., 20:885, 1973.
[15] F. James. MINUIT, A package of programs to minimise a function of n variables, compute
the covariance matrix, and find the true errors. Technical report, CERN, 1978. Program
Library Code D507.
[16] A. Chao. Evaluation of beam distribution parameters in an electron storage ring. Journal
of Applied Physics, 50:595–598, 1979.
243
244 BIBLIOGRAPHY
[17] H. Burkhardt, R. De Maria, M. Giovannozzi, and T. Risselada. Improved TEAPOT
method and tracking with thick quadrupoles for the LHC and its upgrade. In Pro-
ceedings of the 2013 IPAC Conference, number MOPWO027 in International Particle
Accelerator Conference, 2013. http://accelconf.web.cern.ch/AccelConf/IPAC2013/pa-
pers/mopwo027.pdf.
[18] D. E. Knuth. The Art of Computer Programming, volume 2. Addison-Wesley, second
edition edition, 1981. Semi-numerical Algorithms.
[19] Johann Bengtsson and John Irwin. Analytical Calculation of Smear and Tune Shift.
Technical Report SSC-232, SSC, February 1990.
[20] Frank Schmidt. SODD: A physics Guide. Technical Report Beam Physics Note 60,
CERN, 1999.
[21] E. T. d’Amico. SODD: another MAD-X module. Technical Report AB Note-2004-069
(ABP), CERN, 2004.
[22] A. Piwinski. The Touschek Effect in Strong Focusing Storage Rings. Technical Report
DESY-98-179, DESY, 1998.
[23] Alex Chao and Maury Tigner, editors. Handbook of Accelerator Physics and Engineering.
World Scientific, 1999.
[24] Matthew Sands. The Physics of Electron Storage Rings. Technical Report SLAC-121,
UC-28, SLAC, University of California, 1970.
[25] J. D. Bjorken and S. K. Mtingwa. Intrabeam Scattering. Technical Report FERMILAB-
Pub-82/47-THY, FNAL, July 1982.
[26] M. Conte and Michel Martini. ??? Particle Accelerators, 17:1, 1985.
[27] Fanouria Antoniou and Frank Zimmermann. To be published. Technical report, CERN,
2012.
[28] R. Bartolini, A. Bazzani, M. Giovannozzi, W. Scandale, and E. Todesco. Tune evaluation
in simulations and experiments. Technical Report CERN SL/95-84 (AP), CERN, 1995.
[29] Valery Kapin and Frank Schmidt. MADX-SC Flag Description. Technical Report CERN-
ACC-NOTE-2013-0036, Nov 2013.
[30] Etienne Forest, Frank Schmidt, and Eric McIntosh. Introduction to the Poly-
morphic Tracking Code. Technical Report KEK Report 2002-3, CERN-SL-2002-
044-AP, KEK, CERN, July 2002. http://ccdb4fs.kek.jp/cgi-bin/img/allpdf?200302020,
http://cern.ch/madx/doc/sl-2002-044.pdf.
[31] Frank Schmidt. MAD-X PTC Integration. Number MPPE012 in Particle Accelerator
Conference, page 1772, Knoxville, USA, 2005. http://cern.ch/madx/doc/MPPE012.pdf.
[32] V. Kapin and F. Schmidt. PTC modules for MAD-X code. Technical report, CERN,
2006. To be published.
[33] G. J. Roy. A new method for the simulation of synchrotron radiation in particle tracking
codes. Nuclear Instruments & Methods in Phys. Res., A298:128–133, 1990.
BIBLIOGRAPHY 245
[34] Gerhard Ripken. Untersuchungen zur Strahlf¨uhrung und Stabilit¨at der Teilchenbewegung
in Beschleunigern und Storage-Ringen unter strenger Ber¨ucksichtigung einer Kopplung
der Betatronschwingungen. Technical Report DESY Internal Report R1-70/4, DESY,
1970.
[35] Gerhard Ripken and Ferdinand Willeke. Methods of Beam Optics. Technical report,
DESY, 1988.
[36] V. A. Lebedev and S. A. Bogacsz. Betatron motion with coupling of horizontal and
vertical degrees of motion. Technical report, Thomas Jefferson National Accelerator
Facility, 2010. PDF.
[37] E. T.d’Amico. Nonlinear parameters from PTC. Technical Report MAD-X Meeting 7,
CERN, November 2004.
[38] A. Dragt. Lectures on Nonlinear Orbit Dynamics. In 1981 Summer School on High
Energy Particle Accelerators. American Institute of Physics, July 1982.
[39] J. Milutinovic and A. G. Ruggiero. Comparison of Accelerator Codes for a RHIC Lattice.
Technical Report AD/AP/TN-9, BNL, 1988.