This page was written by Helmut
Weick, 7thApril 2020, contact h.weick(at)gsi.de, Imprint (Impressum), Privacy
Policy (Datenschutzerklärung)
The FRS is a magnetic separator. The primary beam comes from the heavy ion synchrotron SIS, hits the target (TA) and the products are transported to the focal planes (F_i) and further to other experimental areas.
The FRS has large dipole magnets to deflect the beam, quadrupole doublets and triplets to focus the beam and hexapole magnets to correct aberrations and small steerers to deflect the beam in vertical direction.
Dipoles: nominal radius = 11.25 m, deflection angle = 30deg, inclined field boundary of angle 7.5deg on both sides. Bmax= 1.65T, BLmax= 9.8Tm with a current of 900A. The maximum field defines a maximum magnetic rigidity of the beam (Brho) of 18.5 Tm. The minimum Brho due to stability of the dipole power supplies is 1.97 Tm. The magnet half gap is 5.75cm. The aperture of the vacuum chamber is rectangular with half gap = 5.0 cm, half width = 12.7 cm.
Quadrupoles: path length = 1.0m, (central one in
triplet = 1.2m), B'Lmax= 10.8 T with a current of 600A.
This can add additional constraints on the maximum Brho. The
pole-tip radius is 8.5cm, while the vacuum chamber aperture is
star shaped with an inner radius (R) of 7.0cm and an outer
half-width of dX=12.40cm and dY=11.88cm.
Hexapoles: path length = 0.26m (eff. length=0.316 m), B''Lmax= 10.45 T/m with a current of 85 A. Pole-tip radius = 12.5cm, inner radius of round vacuum tube = 11.88cm, Note on polarity of hexapoles (pdf file).
Steerers: Steerers act only in vertical direction. Type 1 has BLmax= 0.044 Tm with a current of 100A and type 2 BLmax= 0.0187 Tm with a current of 21A. The aperture is like this of the surrounding quadrupoles.
Note, for ion optical calculations not the iron length of the elements is used but the effective length (Leff), i.e. the B-field integrated along the path divided by the field in the middle of the element (B0).
Usually the FRS is operated in an overall achromatic separator
mode (also called energy-loss spectrometer mode). This means the
dispersion due to the momentum spread coming from the target is
matched to zero in position and angle at the final focal plane F4,
F6 or F8. Different variants of this mode exist, characterized by
a different dispersion coefficient (x,δ) in the intermediate focal
plane F2.
The beam envelopes for a fragment beam starting at the target up
to F4 and a dispersion line (red) for δp=1% are
depicted below.
FRS standard mode (RUN81-TA1B, RUN81-TA2B), Dispersion
(TA-S2)= -6.5 m
There exist similar achromatic modes with lower dispersion down
to Dispersion (TA-S2)= -4.5 m. In addition the illumination of the
dipoles may be changed keeping the dispersion constant, but
increasing or decreasing the resolution.
Two fully equipped target ladders exist at different distance to
the first quadrupole of the FRS. With the target close to the
following quadrupole triplet (TA#2) the angular acceptance is
larger but the resolving power is reduced compared to the first
target (TA#1). Different settings of the beamline SIS to target
are required to focus the beam on TA#1 or TA#2.
Input files for the ion-optics program GICOSY
describing the modes can be found at the MOCADI download area.
Special modes are:
to become achromatic already at F3 to separate there
and to use the last dipole to create again dispersion for energy
bunching.
to add the dispersion of all dipoles to achieve
highest resolution without being achromatic at the end.
A mode with no dispersion at each focal plane for a
transport beamline of high momentum acceptance.
A mode to become achromatic from TA to F2 with the
help of a strongly wedge shaped degrader at F1.
Each experiment can ask for small modifications like the adjustment of the final focal position at F4. This is in principle easy to do and does not change the overall mode, the dispersion etc.
Here are good reference magnet settings with centered primary
beam for a defined magnetic rigidity:
These settings correspond to the usual operation mode with target
station #2.
TA-S4 (TS-HFS)
SRMAG file: S323_03, for Brho (TA-S4) = 8.5601 Tm.
Was used for a 238U92+ beam centered up to
S4. SIS-Energy was 437.2 MeV/u for 238U73+.
Only Seetram + Cu-90 mg/cm^2 target in the beam, which means
E(TA-S4) = 429.14 MeV/u,
Target station #2, focus at S4 is 233 cm behind last quad in x +
y.
based on theory: RUN81_TA2B.DAT
with adjusted last triplet.
reference E084 logbook,
p.32, July 2010
-----------------------------------------
TA-ESR (TS-ESR)
SRMAG file: E082_01, for Brho (TA-ESR) = 7.7978 Tm.
Setting from oscillation experiment 132Xe54+
at E-SIS = 404.7 MeV/u, matter in FRS only Seetram.
Was derived from older setting for 7.900 Tm, may be adjusted a bit
to new TA2, matching to ESR in standard mode.
based on theory: frs-esr-matched-4.dat
reference in MMM logbook
2009, p.3, 23 March 2009
-----------------------------------------
TA-CaveC (TS-HTC)
SRMAG file: S389_03, for Brho (TA-CaveC) = 9.8241 Tm.
64Ni28+ at E-SIS = 660.0 MeV/u, matter in
FRS only Seetram + SIS-window (35 um Ti).
based on theory setting: frs2r3b_ta2_largerspot.dat
reference S389 logbook,
p.10, 24. Sept. 2010
sextupole TH4KS1 was offline (set by hand to B''L = +0.7088 T/m)
The code GICOSY is usually used to calculate values for the different magnets. As a result you obtain plots of the beam envelope and dispersion along the beam line, the matrix coefficients for the stages of the FRS and a list of magnet values. It can also provide matrix files of all single optical elements to be used in the Monte Carlo simulation program MOCADI. Some GICOSY input files can be found in the MOCADI download area web page others only in the personal directory of an FRS expert.
The magnet values are stored in th LSA data base. They can be
imported to this from a csv list of magnet set values. The values
are given as integrated gradients divided by Brho. For a dipole
simply BL/Brho, for a quadrupole B'L/Brho and for a hexapole
B''L/Brho. For a given B-field (B0) and a half-gap or
inner radius (G0) these can be calculated as simply BL
= B0 * Leff for a dipole, for a
quadrupole B'L = B0 * Leff / G0,
and a hexapole B''L = 2 B0 * Leff
/ (G02). In the header of these files I
put some commment lines, which may have to be removed before a
data base import. One important remarks ist the Brho at which the
setting was calculated because teh effective lengths vary slightly
with absolute B.
Example of a csv file with magnet values for the LSA control
system (sistshfs_run81_ta1b.csv).
The optical properties are characterized by transfer matrices. They are provided before every experiment by an FRS expert. Most of them can be found on the MOCADI download area, including matrices for each section of the FRS, plots of beam envelopes, trajectories and of the dispersion function.
The values derived for one magnetic rigidity can be scaled and used also at different magnetic rigidity, see section "adjust online". Even at higher field strength (close to the maximum Brho of 18 Tm) this approximation is still sufficient for FRS experiments. In this case the transfer matrices do not change.
A control system based on LSA (LHC software architecture) is
used at GSI. All settings on magnets are made by LSA. Different
apps are used to provide input or to modify values. The FRS is not
calculated directly in these apps, so we need different interface
programs running on different computer systems. The figure below
gives an overview.
Presentation on
LSA for GSI/FAIR by Jutta Fitzek.
Theory reference values
Each GSI beamline has at least one theory reference setting. This
basic mode is selected when an accelerator pattern is created.
In LSA the reference is stored in the LSA data base and only
privileged machine responsibles are allowed to upload data. For
upload two files are required in the format of a csv file with
data for each magnet named by the official nomenclature.
The strength file (*.csv) contains the integrated magnet setting
normalised to Brho for each magnet (dipole KL = BL/Brho,
quadrupole K'L=B'L/Brho, sextupole K''L).
The Twiss files also contains the strengths but in addition beam
parameters of a typical beam, phase advance, dispersion and
nominal beam position off center. However, so far this additional
information is only used to create plots for information not for
setting (as of May 2018).
Preliminary document on
optics definition by David Ondreka.
Save / Load setting
The app ParamModi can be used to change individual magnet
values but also to load and save many values and even for setting
a section of a beamline completely.
The "File -> Export" or "Import" function in ParamModi is the
main tool for saving and loading data sets.
The save file contains values normalized in Brho. After loading
these values will be scaled with the calculated or defined Brho
values for each zone. You can enter Brho values in the folder
"Beamlinematter".
from old saved setting
Old data sets still saved in the IBHS format can be converted to
the new ParamModi format. However, in old times the data sets
lacked the information on the actual Brho setting. Here the old
logbook with the printout of keyword, assumed Brho values together
with the handwritten real Brho values are the only help. IBHS save
files are on asl740.acc.gsi.de, to find files use "grep
from GICOSY reference files
The theory data sets described on the MOCADI download page are
also stored in parammodi format in
/home/rifr/scheidenb/gicosy/*.txt
More files can be converted with a script:
/home/rifr/scheidenb/lnx/gicosy/lsa2parammodi.pl .
Status report
The magnet status including Hall probe readout can be written to
file and printed with the program FMGSTAT. This program
also saves magnet status files periodically.
Usually saves go to the local disk of the console, but when
FMGSTAT is started directly on "scheidenb@asl740" files can be
written to the user disk and forwarded to normal GSI Linux. The
Linux/Windows account "frs-phys@lx-pool" will be used for this.
After the reference setting or loading a complete ParamModi setting you will often need adjustments of the beam.
Trims with ParamModi
Individual magnets can be adjusted by a so-called trim in
ParamModi. In ParamModi choose the pattern and the list of magnet
values in the folder "extraction line". You can overwrite the
value of each magnet, for example to shift the beam to adjust the
focusing.
The values are written as KL (or K'L, K''L), normalized in Brho.
This trimmed value then is the new reference and will be scaled
with Brho to obtain the wanted BL, B'L, B''L. Any new setting
derived from this trimmed setting will contain the same trim.
Brho setting
Shifting the beam can be done by setting the beamline to a new
Brho value. Then all magnets in this section will be scaled in BL
(B'L or B''L) and the beam will move according to the dispersion
of this section.
The main tool for Brho definition in each section is ParamModi.
After selecting the pattern (here SIS18_Ring_HFS...), in the
folder "beamline matter" the Brho values can be entered directly
in the overall Brho values on top. Alternatively they could also
be calculated based on all the selected matter defining the energy
loss and the type of nuclide after each possible target.
Scaling (FMGSKAL)
Due to the hysteresis of the magnet iron a simple change in
current will not lead excatly to the wanted scaling factor. To
always reach a BL value in the same way as during the initial
calibration of the magnets the final currebt values are always
approached in a certain ramping procedure called "pre-cycling".
The corresponding app to do this is called FMGSKAL.
By clicking you can select the section you want to pre cycle and
whether you only want to precycle the dipole magnets are also the
quadrupoles (usually not needed). First the ramp always goes to
maximum current, then to zero, back to maximum, back to zero and
then waits for the new set value. The whole procedure takes ~2
minutes.
In this way the new set value is always approached from below.
This also means that actually for increasing fields a precycle is
not neccessary. Also for very small changes (like 0.1%) it is not
needed as there practically is no difference.
-> Note on scaling and ramping of FRS magnets (pdf file).
Example: Centering the beam
The beam is centered at S2 but at S3 the mean value of the beam
position is +5mm shown by the analysis program. In ion optical
coordinates this means -5mm (the x axis is reversed). The
dispersion coefficient for S2 to S3 is +2.2m. If you think the
reason is a not well known energy loss at S2, then the Brho afer
S2 should be redefined in ParamModi. Old Brho = 10.0 Tm, new Brho
= 10.0Tm * (1 -5/2200) = 9.9773 Tm. After the scaling of the
magnets the beam should be centered.
"Hall probe calibration (HP Cal)" and selection of "Beamline
Destination" are independent sequences which can be started from
FMGSKAL.
Monitor optics setting (GICOSYBACK)
The program gicosyback.pl can be used to read the values from a ParamModi export file and to run a GICOSY calculation of the FRS (TA-S4) with these values. It consists of a Perl script running on accelerator linux (asl74x.acc.gsi.de).
1.) work on scheidenb account (or weick).
2.) Export the status from ParamModi to directory ~/parammodi, wait max. 1 min
3.) run GICOSYBACK
>gicosyback
or versions gicosyback-esr, gicosyback-htc (for beam to ESR or Cave C)
This is only an alias to a script hidden in the respective subdirectories of ~/gicosy.
You can also start it directly with more parameters
> ./gicosyback.pl <export filename> <opt1> <opt2>
The parameter opt1=1 tells gicosyback also to consider the steerers,
and opt2 can be =1 or =2 for target stations #1 or #2.
Only "gicosyback" uses the last saved parammodi export file.
4.) The output will be graphic files gicosyplot-*.pdf, and
gicosyplot-*.png. An okular viewer for the pictures is
opened automatically.
The output is for two independent calculations: 1.) SIS - TA,
2.) TA - HFS (ESR / HTC).
MIRKO
MIRKO is an ion-optical program written by Bernhard Franczak at
GSI. It can calculate and display the beam in FRS and also modify
the settings.
ParamModi and FMGSTAT files are copied on demand to the frs-phys
Linux account on lxi from were also the frs-phys windows user can
read them via SAMBA.
MIRKO export files can also go in the other direction. The default
input is "last_param_in.txt" and default output is
"last_param_out.txt", both in Paramodi export format.
Start MIRKO with the Mirko-LSA icon on the desktop of
KSPC014. 
In the menue "Beamlines" -> choose "HEST". In the new menue
"HEST" choose "SIS18->FRS".
In the "Graphics" menue you can choose beam envelope plots, phase
space plots, or a view on the beamline from top.
In the dialog window you can enter more commands.
Data exported from param_modi can be read into MIRKO and
exported again. In the menue "device-access" mark the input source
[online] and then click on [read all] or [set all]. The
input/output runs via two files with names "last_param_in.txt" and
"last_param_out.txt".
Due to a bug in file access so far only
mode internal values is possible (from file
D:\mirko\last_param_in.txt).
Copy file from Samba to local dir by hand (copy /Y
H:\FRSsettings\mirko\last_param_in.txt
D:\mirko).
Similar with last_param_out.txt in opposite direction.
A special feature is the "Beam alignment". Usually it is used at
the FRS targets looking at the profile grids GTS1DG5 and GTS2DG2.
1. First read in the old steerer and dipole values
[Steerer lesen]
2. Then enter the current beam position. Most often the
wanted position (Sollposition) is 0 in x and y on both detectors.
3. Then click on [Korrekturen berechnen + setzen], to
calculate the correction.
In the envelope plot you should see a shift of the envelopes.
After tuning the new steerer values are exported directly to a
version of "last_param_out.txt" containing only the modified
steerer values. It can be read into param_modi
Simulation with LISE++ and MOCADI
For a detailed simulation of beams insiode the FRS with realistic
distributions, nuclear reactions and energy loss in matter the two
main tools are LISE++ and MOCADI.
For LISE++ in the simplified block mode and fast calculation by
convolution of shapes of distributions we provide standard
configuration files:
For more detailed simulations you should use MOCADI.