|
In order to improve the safety of the planned MEGAPIE experiment at SINQ, an additional method
has been proposed to avoid proton beam bypassing target E and perforating the MEGAPIE target with too
much non-scattered beam [1]. The functioning of this method is based on the energy
dispersion of the 90° bending system consisting of the magnets AHM and AHO and the field-lens
duplet QHI29 / QHJ30. A model has been developed to compute the halo at the location of the
maximum dispersion (inside QHJ30) as accurately as possible. The required
Turtle Input File
shows the necessary parameterization of the beam line between Target M and the SINQ target. Results
from these computations are shown in Fig. 1 and Fig. 2 below.
Fig. 1: The histogram depicts the computed proton beam distribution in the middle of QHJ30 (the
location of the slit) for a well centered beam at target E. The total halo is only of the order of
10 nA for a 1 mA proton beam intensity at the SINQ target. The low-energy halo tail (to the
right = upper jaw) is longer than the high-energy halo tail (to the left = lower jaw). The
prognosis is that a few nA of beam should already hit the upper jaw even if it is in the
out-position. This could also be seen with the first tests with beam. If the lower jaw is put
into the working position (-25 mm) only about 5 nA of the proton beam would hit it and produce
some permanent but moderate losses around the slit. Fortunately, this region is well shielded by
the quadrupole iron surrounding the slit.
Fig. 2: In this histogram the resulting beam at the slit position is shown, if in the model computations
the beam on Target E is shifted by 1.5 mm from the center, so that 0.14 % of the protons are
bypassing target E. Most of these protons would be hitting the lower jaw, if it is put into the
working-position. The charge of the protons hitting the electrically insulated jaw is collected
and measured with a MESON unit (see control diagram of Fig. 6). But it should be mentioned here, that the
currents measured with the MESON units are at least twice the charge of the stopped protons per time
unit, because a lot of electrons are knocked out of the copper jaws and flowing to the surrounding vacuum chamber
[2]. The produced beam spill is also ionizing the air inside the closely down streams
placed ionization chamber (MHI37) and producing a current signal registered by the
LOGCAM2 unit (see control diagram in Fig. 6). Both signals are strong enough to produce already an
interlock signal with only 0.1 % of the beam bypassing target E.
Fig. 3: This picture shows a simplified drawing of the cross section through the vacuum chamber of the
quadrupole lens QHJ30 with the slit plug-in unit. Fortunately the vacuum chamber of the
QHI20/QHJ30 field-lens was build as a so-called 'cross-slit' type chamber as commonly preferred
for the pion/muon beam lines. The initial reason for this was the better transmission of the
beam halo because of the presence of dispersion inside this lens. This space could now be exploited
for the outer parking position of the slit jaws in case they are not used. First experience with
the slit jaws in the out-positions showed, that the halo hitting the copper jaws is negligible.
Fig. 4: This 3-dim design drawing shows all the important components of the slit collimator
unit. The support tube containing the jaws with the rod linkage and the cables (see Fig. 5)
is connected to a vacuum flange with the 2 stepping motors and the vacuum feedthroughs attached
to it. The whole unit is plugged into the quadrupole's 'cross-slit' vacuum chamber from the
right side. In order to do this, the AHO bending magnet (weight = 60 t) and the adjacent vacuum
valve to the right side had to be moved away temporarily.
Fig. 5: This photo shows the inside of the slit plug-in unit in opposite direction of the proton beam.
Well visible are the 2 closed copper jaws with the gear needed to open and close them. Also
visible are the coaxial cables of the 2 thermo-couples (wound to spirals because of the longitudinal
movement of the jaws) and the ceramic-pearls-insulated leads for the current measurement at the
jaws.
Fig. 6: This block-diagram shows the principle of the control-electronics for the slit. All units
are connected with the control computer via the CAMAC bus. Currents, temperatures, voltages,
interlock-limits etc. may be read or set via this bus. Not shown in this diagram are the cables
needed for transmitting the interlock signals.
Fig. 7: This screen shot shows on the left side the signals from the ionization chambers along
the proton beam line from target E to the SINQ target for a well centered beam. On the right
side the same display for a proton beam shifted near the edge of target E so that about 0.1 %
of the protons are missing the graphite. The lower KHNY30 jaw is put in working position. The
amount of beam bypassing the target E was estimated with the profile monitor MHP56 (see [1] Fig. 3)
The signal of MHI37 is very close to the interlock level (relative maximum).
Fig. 8: This screen shot is a control-room display of all relevant control parameters of the KHNY30
slit collimator. Both signals (red columns) MHI37 (=97.7 nA) and KHNY30u.ILOG (=80.4 nA) are
higher than the usual interlock levels (they have been temporarily over-bridged for this screen
shot).
[1] U. Rohrer, A Novel Method to improve the Safety of the planned
MEGAPIE Target at SINQ. PSI Large Research Facilities Scientific and Technical Report 2001, Volume VI, p.34-35
[2] L. Rezzonico, private communication.
Back to: Monitoring the beam transmission
Last updated by
Urs Rohrer on 25-Jan-2006