Title: MachineDetector Interface
1Machine-Detector Interface
- David Urner
- JAI / Oxford University
2Topics
- ILC energy spectrometer
- Force-neutral Anti-Solenoid
- QD0-Service Cryostat Prototype
- Luminosity Optimization
- Push-Pull
- Background-Driven SiD Design Decisions
- Extraction Line
- Radiation Shielding
3Proposed ILC energy spectrometer
Bino Maiheu
- Precision measurement dE/E 10-4
- Minimal impact on beam itself allowed emittance
growth from SR - Limited space budget in BDS 60 m
- Minimal impact on physics data taking for e.g.
calibration runs - Magnetic chicane with high resolution beam
position monitors cavity BPM - Max 5 mm dispersion at center chicane
determines resolution - Emittance growth determines chicane layout
- Diagnostics needed
- Gain drifts temperature
- Mechanical stability interferometer
- Magnetic fields (?B.dl ) NMR, Hall, fluxgate
magnetometers
4T474 test experiment at ESA, SLAC
- ESA comparable repetition rate, bunch charge,
energy spread as ILC - Possibility to vary bunch length, energy, charge
- Easy steering with feedback system
- Build an energy spectrometer prototype, using a 4
magnet chicane - Goal is to demonstrate the stability of this type
of energy measurement at 10-4 level, and
investigate how such a magnetic chicane can be
operated most efficiently at the ILC - Operate at 5 mm ?X at centre chicane as in
current ILC design - Need lt 1mm resolution on position measurement
(BPM) with position measurement stability over
multiple hours of 100 nm
5(No Transcript)
6(No Transcript)
7QD0-Service Cryostat Prototype
- BDS RD plan stresses importance of integrated
systems testing of an engineering prototype (QD0
with connection to service cryostat) - to establish the degree of coupling of external
vibration sources to the QD0 cold mass. - What noise is expected from technical systems?
- How well can the QD0 cold mass be isolated?
- What about bellows?
- Are there any internal modes to worry about?
- How well can QD0 be supported in the detector?
- How do we isolate the QD0 cryostat from external
vibration sources? - Particular attention is needed in this regard to
the design of the cryogenic transfer line between
the service and QD0 cryostats.
8Luminosity Optimization
- GuineaPig (beam-beam interactions)
- New features
- beam-beam effects on bhabhas
- C
- all keywords of guinea-pig are now available
- hadrons (do_hadrons)
- minijets (do_jets)
- pairs (do_pairs)
- abstract I/O interface
- separate algorithms and I/O
- plugging different format (ascii)
- plugging graphical interface
- FONT Simulation
- Fast feedback using bunch kick-angle to correct
for y and y. - Simulation of fast feed-back using PLACET (beam
transport), GuineaPig (beam-beam interactions)
and Matlab (fast feed-back) - Simulation based on Glen Whites code
incorporating - Newest Lattice
- Smaller wake fields
- Similar results as
- with old code
- Ready to incorporate
- additional effects and
- other feedback loops.
- Likely Result A much worse behaviour
9Luminosity Optimization
- MONALISA
- Relative position measurement of two QD0 magnets
with respect to each other using interferometers - Position changes few nm (requires continuous
uninterupted measurements) - Absolute distance below 1mm (can be done at any
time) - Use as input for FONT
- Test on FONT simulation how effective this is in
reducing convergence time particularly under bad
beam conditions (as one would expect for startup
or after a longer period with no beams). - Requires light path (half inch pipe) through SiD
yoke from QD0 to the ground.
10Luminosity Optimization
- Luminosity feedback
- Luminosity sensitive to bunch shape ? optimum at
y?0 - After IP position and angle feedbacks (based on
kick-angle monitor) have converged maximise
luminosity directly using luminosity monitor
based on BeamCal and GamCal - BeamCal Detector
- .003 lt ? lt .02 rad
- ?3.5m from IR
- Pairs curl in the magnetic field
- Measure the ?104 beam-strahlung ee- pairs/bunch
- GamCal Detector (B Morse)
- ?180m from IR integrated into polarimeter chicane
- ?10-4 X0 to convert beam-strahlung gammas into
ee- pairs - Converter could be gas jet or a thin solid
converter - Dipole magnet with PT kick 0.25 GeV/c separates
the pairs from beam electrons - Calorimeters outside vacuum after magnet measure
the 1-10 GeV positrons
11Luminosity OptimizationVertical Offset
M.Ohlerich
- complementary information from
- total photon energy vs offset_y
- BeamCal pair energy vs offset_y
ratio of E_pairs/E_gam vs offset_y is
proportional to the luminosity similar behaviour
for angle_y, waist_y
see also EUROTeV-Memo-2006-011
12Statistical Error for BX
- If at startup less than 10 nominal luminosity
- beamcal might not give much of a luminosity
measurement from a single bunch. - Conclusion BeamCal and GamCal provide robust
complimentary information.
13Push pull
- Assertions - Superconducting magnet (detector
final focus) warm-up/cool-down time scales are
long enough that these magnets have to be moved
while cold? - But they may be de-energized?
- If true, this drives a need for long umbilical
connections to each of the experiments that are
able to accommodate 20 m motion while cold. - In last few weeks lots of discussions about
possible layouts
14Push Pull
- Separate cryostats for QD and QF magnets
- QD0 carried by detector at optimized L
- Cantilevered support tube concept dead
- QF never moves and z-position of QF same for
all detectors - Distance between the 1st and 2nd SC quads after
IP is increased to provide sufficiently long warm
section for push-pull design. - For three options of L 3.51 m, 4.0 m, 4.5 m,
the SC extraction quad QDEX1 is placed at 5.5 m,
5.95 m and 6.3 m. The 2nd SC quad QFEX2A is at
fixed position, 9.6 m from IP. - A long drift after QFEX2A provides transverse
space for crab-cavity.
15Push-Pull Alignment after Move
- Assume repositioning accuracy of detector 1mm.
- Use MONALISA system to measure remaining
difference to O(mm) - Realign QD0 using MONALISA information
- Unclear yet how QD0 is supported
- Use y and y kickers to steer beam correctly into
QD0 based on MONALISA data - From here FONT should do the trick.
- Would expect that silicon Tracker and VXD have to
be realigned as well. - Is some method of position monitoring for the
tracking detectors foreseen?
16Background-Driven SiD Design Decisions
Takashi Maruyama
17Background-Driven SiD Design Decisions Current
Beam pipe is not compatible with the Low P or
High Lumi options.
R 1.2 cm ? 1.5 cm (Low P), and R 1.2 cm ? 1.8
cm (High Lumi).
18Background-Driven SiD Design Decisions New SiD
Geometry
- LumiCal
- Z156.75 168 cm
- Rinside6cm
- compatible with
- Nominal 5 Tesla
- Nominal 4 Tesla
- Low P 5 Tesla
- Beampipe
- Original 43 mrad cone cylinder
- BeamCal
- Study background as a function of BeamCal
z-position
LumiCal
43 mrad
Move BeamCal
Cylinder
19Tracker Hits vs. BeamCal DZ
Si Tracker photons
Barrel VXD e/e- hits
Barrel
- VXD hits
- No effects from LumiCal/BeamCal changes
- Si Tracker hits
- Less photons (20) due to smaller radius LumiCal
- photons increases by moving the BeamCal
forward. - But the rate is acceptable if ?Z lt 30 cm.
Endcap
20Extraction Line
- The quads focus the beam to the 2nd focus at
Compton IP with R22 -0.5 transformation. - The diagnostic contains energy and polarimeter
chicanes, two proposed GAMCAL bends and the fast
kickers for sweeping the beam on 3 cm circle at
the dump window. - Two collimators are included in the energy and
polarimeter chicanes, and three collimators are
in the final 100 m drift to protect the kickers
and limit the beamsize to within the R 15cm
dump window.
21Extraction Line
- The original 4-bend symmetric polarimeter chicane
is modified to increase the strengths of the 3rd
and 4th bends by 50 which improve acceptance of
Compton backscattered electrons in the Cherenkov
detector. - The 5th and 6th bends are added to the
polarimeter chicane which close the orbit bump
and can be used by the proposed Gamma Calorimeter
(B. Morse).
- Fast sweeping kickers
- Without beam collision, the undisrupted nominal
beam size is too small (2.42 x 0.29 mm2) at the
dump window. - A system of 5 horizontal and 5 vertical fast
kickers is included 90 m before the dump. The
rapidly oscillating x and y kicker field (1 kHz)
sweep the beam on 3 sm circle at the dump window.
This was shown to be sufficient to protect the
dump window and prevent water boiling in the dump
vessel (L. Keller).
22Neutrons from the Beam Dump
New FLUKA simulation (Darbha)
x (cm)
Quadrupole aperture
2.4 cm
3.7?1010 ns/cm2/y
2.4 cm
7 mrad
n
3.0 cm
VXD Layer 1 fluence (one beam) 1.8?1010
ns/cm2/y w/o BeamCal 4.3?108 ns/cm2/y w BeamCal
Be
Si VXD
W BeamCal
z (cm)
23Neutrons in VXD (FLUKA)
Neutrons from pairs
Neutron origins
BeamCal
Neutrons from radiative Bhabhas
Beampipe
M1
Z (cm)
- Neutrons that reach the vertex detector are
mostly generated in the BeamCal. - Anti-DID can reduce the neutron flux.
- Different L design should not affect the neutron
flux.
24Modelling of spent beam
(Nosochkov)
- 5T post-IP solenoid including anti-DID field with
orbit and focusing correction have been modelled. - Disrupted beam loss on magnets small in nominal
option - Manageable in a large energy spread option
- High loss at diagnostic collimators if at IP
- Large y-offset
- Non-zero y-angle
- Particularly in large energy spread option
- Driven by nonlinear dispersion from dipole
correctors - To reduce this effect
- a) the design IP y-angle is minimized,
- b) anti-DID field is increased,
- c) dipole corrector field is brought closer to
IP. - In machine operation, the beam running with large
y-offsets at IP should be efficiently detected
and prevented.
25Radiation Shielding
- Shaft and wall between detectors
- Shielding capability of a SID
- Effects of gaps for cables and tubes
- Pacman thickness
- Inner diameter
- Requirement of penetration
- Effects from upstream part (muon etc)
- 3D Monte-Carlo simulation ? MARS15-MCNP code is
used
26Other Topics
- Intensive discussions particularly in preparation
of the ILC Interaction Region Engineering Design
Workshop at SLAC 17-21 Sept. - 4 workgroups
- WG-A Overall detector design, assembly, detector
moving, shielding. - WG-B IR magnets design and cryogenics system
design. - WG-C Conventional construction of IR hall and
external systems. - WG-D Accelerator and particle physics
requirements
27The End
28Backup slides
29Preliminary spectrometer results
- Taking into account Taking into account ?B.dl and
deflection at center of chicane, can B.dl and
deflection at center of chicane, can - compute correct beam energy compute correct beam
energy - Have to Have to subtract incoming orbit subtract
incoming orbit in each event prove we measure
just energy ! in each event prove we measure
just energy ! - Further detailed analysis, Further detailed
analysis, spectrometer stability studies
spectrometer stability studies underway...
underway... - More and better data to come in July...
30Power loss for L 3.51 m with 5T solenoid
- Loss on magnets is still small or modest, but
loss on diagnostic collimators may become high
when both y-offset and y-angle at IP are large,
particularly in high disruption options c14
(c15). - Therefore, large IP orbit offsets need to be
efficiently controlled. Further improvement of
diagnostic collimation system should be studied. - The source of the higher loss is the non-linear
dispersion created by the corrector field. It
increases the orbits of low energy electrons
which may be consequently lost. - In order to reduce the non-linear dispersion, the
corrector fields need to be lowered and,
possibly, brought closer to IP for more local
correction. The following may be considered - The uncorrected IP y-angle is minimized.
- The anti-DID field is increased, so the
correcting fields on the SC quads are reduced.
31(No Transcript)
32Current SiD layout
33(No Transcript)