Randy Ruchti
------------
work on advanced optical instrumentation: ultra-compact fast timing
RADiCAL = RADiation-hard innovative CALorimetry
radiation dose up to 1MGy and 30e15 1-MeV neutrons/cm^2
want to have excellent energy resolution, high efficiency, rapid response, shower position and fast timing, use for trigger
challenge: radiation, PU
resolution: 10%/sqrt(E) + 0.7% up to |eta|<4
technique: sampling calorimeter (interest is optical properties), dense (25X0 but <1lambda)
very compact: 12cm depth
crystal or ceramic scintillators and WLS
Geiger-mode pixelated photosensors (SiPM and large band-gap devices)
started with shashlik device: could do sampling in very compact size: 114mm
built prototype w/ LYSO:Ce
quartz capillaries filled with DSB1
scintillators under investigation
Cerium and Praseodymium, Calcium co-doping
note: LYSO+SiPM is part of CMS BTL
novel ceramics (LuAG:Ce) more radiation tolerant than LYSO
slide with fast/ultrafast inorganic scintillators from Ren Yuan
the best: LYSO:Ce and LuAG:Ce
they are bright; emission is different (LYSO is blue, 420nm; LuAG is yellow, 520nm)
revisited fiber optics profile
typical WLS fiber has thin cladding: not useful if radiation high, it is the core that gets damaged
looking at thick wall profile: quartz capillary with high-efficiency rad-hard WLS
the core is small fraction, so light transport (after shifting) happens in quartz, which is rad-hard
photosensors: Arjan Heering and Yuri Musienko experts working on them; in CMS: HCAL with Y11 fibers; BTL with LYSO
intention to exploit development of localized cooling to reduce noise and improve lifetime
more sensors: GaInP pixelated devices (large band-gap devices)
similar to SiPM, currently no commercial market, but interesting for high-radiation areas
4x4 prototype of tungsten/LYSO:Ce w/ DSB1 capillaries
used 100GeV electrons; resolution: stochastic term 15%, constant term <1%
since then, increased light output by factor of 3
energy resolution dominated by sampling, which is 18-19%: one could adjust by changing ratio of W and LYSO:Ce
light level helps in timing and low end of spectrum
added a quartz rod with a waveshifter sitting at shower max: that provides the timing measurement exactly where there are more particles
capillary thick: one can collect also Cherenkov light (with appropriate filter on photosensor)
one can also read downstream (Cherenkov + scintillation) or upstream (scintillation only)
shower max timing: simulated 50GeV electrons
goes down to 20-30ps with a couple hundred p.e.
waveshifter could also be LuAG:Ce: very rad-hard, and one could use small piece of LuAG to get the light from only sections of the shasklik; option for high-eta region,
where very high radiation doses
Franco: if you release request for very high radiation, can design be simplified or made cheaper; after all, the FCC-hh options happening late in time, while the e+e-
come first
RR: one can relax the idea of expensive rad-hard crystals, in favor of siloxanes or plastic
interest of ceramics: light efficiency (but cost)
slide 14: plots from CMS shashlik, 2014
propose to make new tests with new structures (slide 15) at FNAL; goal is to test idea of timing w/ capillaries, depth sampling and shower max measurement, rather than
energy resolution
Franco Bedeschi
---------------
IDEA: detector concept for e+e- colliders
initial idea was to use design of ILC detectors, but some notable differences emerged
plot of luminosities: in circular you have enormous statistics at Z and WW, while that is not the case in linear collider (which instead can go to higher energies)
implication on physics: stronger interest in EWK at Z pole and WW
heavy flavor physics at Z pole comparable to LHCb upgrade
circular vs linear: need detector for low energies (lower momentum: need higher transparency); high control of acceptances to match statistical uncertainty (silicon
wrapper/pre-shower); PID needed for heavy flavor; pi0 for tau and heavy flavors
constraints from physics
would like to do better on jets and photons; maybe tracker and HCAL too tight?
low-field detector to maximize luminosity: 2T
vertical emittance (pm) blows up as the B field goes from 2 to 3T
low-field -> need large tracking volume
calorimeter outside coil -> need very thin coil
beam time structure: short bunch spacing, no large time gap: issues of cooling for PF calorimeters
low field, only 2m tracking volume: return yoke can be small (<1m), it is non-negligible part of other detectors
momentum resolution: CLIC-like detector, fully silicon vs IDEA
smaller slope (asymptotic resolution is better), but in region below 80GeV IDEA is better: that is where we will have most of tracks
vertex detector
struggle to get to low power (so that one can use air cooling) <20mW/cm^2
resolution: 5um (ALICE ITS)
asymptotic value of resolution: 2um
beam pipe is only 1.5cm radius, very small: helps here
PID: maximum drift time is 400ns, cluster counting in DCH good for PID
good pi/K separation, except for 0.75 to 1.05GeV momentum
MIP region (where separation is not good) is narrow
plot of K/pi separation vs resolution of a 2m TOF
calorimeter: want 10%/sqrt(E) for ECAL, and 30-40%/sqrt(E) for jets
produced GEANT study with test beam from RD52 (they ran dual-readout calorimeter), encouraging results: 11%/sqrt(E)+0.8% ECAL, and seems to work well with jets
jet invariant mass for W, Z, H: can get some separation, before any attempt to add PF algorithm to separate c, b, semi-leptonic decays
studies ongoing about crystal option for ECAL
PbWO4, 20cm, expect could have 3%/sqrt(E) resolution
proposal for LYSO timing layer, 20-30ps
R&D
silicon: vertex: low-power, high-speed CMOS MAPS; outer Si: CMOS passive strips, long pixels
ARCADIA collaboration working on this
drift chamber: studies on mechanics, carbon fiber wires coated with metal, can take large stress
cluster counting electronics on board, online
calorimeter: scalable mechanical options under study
building prototypes on capillary brass tubes; seem to work well
readout: SiPM and dedicated chips
CAEN working on readout-chip system with large number of channels (O(1e3))
micro-RWELL for muons
good for industrialization (better than micro-mega and GEMs)
based on achievable technologies, but still need R&D and software simulation
ample space for additional contributions to R&D effort
Nicola de Filippis
------------------
ultra-light drift chamber
summary slide of technologies in detectors; plenty of experience with drift chambers
pT resolution improvement when using He instead of Ar in past
comparison of dEdx of DC vs TPC
cluster counting could provide better separation between particles in several regions of momentum
expect dEdx resolution of 4.5%
KLOE: DC was made in carbon fiber, X0<5% (was >=10% before)
tracker requirements: large coverage, high granularity, high angular and momentum resolution
could reach dp/p^2 ~ few 1e^-5 (small w.r.t. 0.136% beam spread)
useful for reconstruction of Higgs recoil mass from higgstrahlung
lepton flavor violation processes: momentum resolution helps here, expect limits (1e-6) to improve by five orders of magnitude (SM BR 1e-54, 1e-60)
PID for flavor physics
main issues of DCH
multiple scattering: dominates resolution for full range
redundancy: would be better if more layers...
main issues on TPC
genesis of IDEA DCH
KLOE at Frascati, CluCou chamber for ILC, I-tracker for Mu2E
list of innovations from KLOE to IDEA
mechanical structure with separated gas containment and wire support structures; wire tension compensation
larger number of thinner wires
Carbon-monofilament wires
cluster counting for PID, cluster timing for spatial resolution
cluster counting: amplitude of signal vs time
goal is to identify time of arrival of each cluster along the track of the particle
dEdx usually is truncated mean; here each cluster read independently, their distribution helps PID
He-isobutane mixture
mu/pion separation with cluster counting: expect 5sigma, measure 3.2sigma separation (with usual method, expect 2 and measure 1.2)
cluster timing: use drift time of first cluster (or all?) to determine the most probably impact parameter
need electronics that can detect all the (small) peaks in the signal distribution
mechanical structure: 112 layers
full GEANT simulation
single-track efficiency w/ >60% hits: 99.5%
momentum resolution vs angle
d0, z0 resolutions: few um, for large momenta (above 100GeV)
PID: comparison of truncated-mean method and cluster counting
in some ranges of momentum, one can gain factor 2 (in sigmas) in separation with cluster counting w.r.t. dEdx
note: TOF can recover the interval in which DCH PID does not work well
There are minutes attached to this event.
Show them.