Cherenkov detector for proton Flux Measurement Cp FM

Cherenkov detector for proton Flux Measurement (Cp. FM) for UA 9 experiment F. Iacoangeli, 1 D. Breton, 1 V. Chaumat, 3 G. Cavoto, 1 S. Conforti Di Lorenzo, 1 L. Burmistrov, 3 M. Garattini, 1 J. Jeglot, 1 J. Maalmi, 2 S. Montesano, 1 V. Puill, 2 R. Rossi, 2 W. Scandale, 1 A. Stocchi, 1 J-F Vagnucci 3 LAL, Univ Paris-Sud, CNRS/IN 2 P 3, Orsay, France CERN - European Organization for Nuclear Research, CH-1211 Geneva 23, Switzerland 3 INFN - Roma La Sapienza, Italy 1 2 1

Outline • UA 9 experiment at SPS • LUA 9 project • Cp. FM detection chain components • Optical simulations on the Cherenkov radiator • Beam tests at BTF of simplified prototypes (October 2013) • Beam test at BTF of the Cp. FM full chain (April 2014) • First preliminary results of the beam test • Conclusions 2

UA 9 experiments q The main purpose of UA 9 is to demonstrate the possibility of using a bent silica crystal as primary collimator for hadron colliders. q Bent crystal works as a “smart deflectors” on primary halo particles θch ≅ αbending amorphous <θ>MCS≅3. 6μrad @ 7 Te. V channeling θoptimal @7 Te. V≅ 40 μrad R. W. Assmann, S. Redaelli, W. Scandale, “Optics study for a possible crystal-based collimation system for the LHC”, EPAC 06 c UA 9 experiment runs in SPS since 2009 3

Crystal assisted collimation q If the crystalline planes are correctly oriented, the particles are subject to a coherent interaction with crystal structure (channeling). q This effect impart large deflection which allows to localize the losses on a single absorber and reduces the probability of diffractive events and ion fragmentation 4

LUA 9 project Use bent crystal at LHC as a primary collimator LHC beam pipe (primary vacuum) To monitor the secondary beam a Cherenkov detector, based on quartz radiator, can be used. Aim: count the number of protons with a precision of about 5% (in case of 100 incoming protons) in the LHC environment so as to monitor the secondary channelized beam. Main constrains for such device: - No degassing materials (inside the primary vacuum). - Radiation hardness of the detection chain (very hostile radioactive environment). - Compact radiator inside the beam pipe (small place available) - Readout electronics at 300 m Cherenkov detector for proton Flux Measurements (Cp. FM) 5

Cp. FM detection chain components Radiation hard quartz (Fused Silica) radiator Flange with custom viewport to realize UHV seal and optical air/vacuum interface Movable bellow Quartz/quartz (core/cladding) radiation hard fibers. üUSB-Wave. Catcher read electronics. For more details see : USING ULTRA FAST ANALOG MEMORIES FOR FAST PHOTO-DETECTOR READOUT (D. Breton et al. Photo. Det 2012, LAL Orsay) The Cherenkov light will propagate inside the radiator and will be transmitted to the PMT throughout the bundle of optical fibers. Radiator must work as waveguide. The first prototype of Cp. FM will be installed and tested in SPS 6

Geant 4 Optical Simulation - We performed many simulation of the optical behavior of detection chain to define the best configuration Different reflection coefficient At the end of detection chain Angle wrt the fiber axis At the radiator surface The number of p. e. is strongly dependent on reflection coefficient 7

BTF Test Setup (October 2013) LAL Cerenkov INFN Cerenkov e- Beam BTF Remote Control Table BTF Calorimeter 8

Radiator with fibers bundle Charge per Electron (normalized to electron path length into the radiator) 47° Optical grease at interfaces between fibers, PMT and radiator The width of the peak is compatible with the angular aperture of the fibers Radiator/beam angle We need the light arrive to the fibers with a angle compatible with angular acceptance The best geometry is with the radiator which cross the beam with 47° angle and the fibers coupled with the same angle so as to use all the angular acceptance of fibers 9

Best configuration for Cp. FM Beam § Radiator cross the beam perpendicularly, due to small place available § Fibers are coupled with 43° angle, so as the incoming light is at 47° angle Fibers Quartz 43˚ Ø Flange brazed configuration or commercial viewport configuration Ø “I” shape or “L” shape, so as to increase the quartz crossed by beam The “double bar” configuration will be useful to measure the diffusion of the beam and the background 10

New BTF Set-up of Cp. FM (April 2014) 4 Cherenkov bars (L and I shape) 47º end of the bars MCP-PMT Fibers bundle PMTs black boxes Fibers bundle 11

Cp. FM Test Setup @ BTF (April 2014) The setup was almost the same of last test beam , except for the long fibers bundle and for the readout provided by Wave. Catcher board. 12

Tested configurations • Configuration A : Trioptics Quartz bars (L curved bars) + bundle + PMT • Configuration B : Optico AG L + I bars + bundle + PMT • Configuration C: I bar + PMT (direct coupling) • Configuration D: I bar + glass plate (thickness=3, 85 mm) +PMT • Configuration E: I bar + glass plate (thickness=3, 85 mm) +bundle + PMT (simulation of a viewport) • Configuration F : quartz I bar with a black tape around it + bundle + PMT (simulation of an absorber around the bar) • Configuration G : bundle in the beam For all these configurations, The signal is recorded by the Wave. Catcher 8 channels module 13

“L” and “I” bars + bundle + PMTs (R 7378 A) “ I ” bar configuration “L” yield less signal than “I” • Higher light signal, due to a better surface polishing • Number of detected p. e. quite linear “ L” bar configuration • Lower light signal, due to a worst surface polishing • Detected p. e. increase when beam came near to the fiber bundle • In principle more light produced in the 3 cm fused silica along the beam direction (“L” shorter arm) bu le d n The polishing, difficult because of the “L” shape, must be enhanced. Reflectivity is essential feature of radiator 14

Optical AG Fused Silica bars Well polished “I” bar: it is possible to distinguish the reflection points along the bar Worse polished “L” bar: the light appears more widespread and only few reflection points are visible 15

Mounting configurations Ø Another result was obtained by the study of mounting with viewport At the start we proposed 2 different mounting configuration Flange brazed configuration • Better light transport (no viewport interfaces) • Better mechanical strength • Loss of light in the brazed points • Technological problems to braze Fused Silica with iron flange Viewport configuration • Worst light transport (viewport interfaces) • More complex mechanical set-up • No technological problems The brazed flange is still under development 16

Effect of the viewport Configuration: “I” bar + quartz plate+ bundle + PMT - Cp. FM output with viewport= 1. 2 p. e/ incident electron - Signal per e- ( @ 800 k. V) : • without window: 10. 5 m. V • With window: 6. 0 m. V Few-particles regime (<4 e-) - Reduction of the signal is about 40 % The insertion of a quartz plate (thickness = 3. 85 mm) between the quartz output and the fibers bundle decreases the signal by a factor less than 2 The Cp. FM with viewport is a suitable solution 17

Cp. FM and BTF Calorimeter correlation Few-particles regime (<4 e-) The study of measured Cp. FM charge (normalized on the charge of single p. e. ) as a function of BTF calorimeter charge shows a rather linear dependence. 18

Conclusions • We have evidence that the full chain (radiator + quartz window + fiber bundle + PMTs) works well, also for low fluxes • We need more time to finish analysis of the data. All the measurements need to be compared with simulations as well • We chose the “I” shape bars and the mounting with viewport for the first Cp. FM • The obtained signal is ~1 p. e. per incident e-. It can be improved by a factor 5 or 6 by means of very well polished L bars (not yet available). • The brazing of the quartz bars with a flange needs of a long collaboration process with a specialized company. It’s not possible for the installation of the Cp. FM in SPS this winter but it has to be done for the future (LHC) • We have measured 250 ps of time spread for the best timing configuration of Cp. FM. This features can be improved using a faster readout electronics. 19

Thank you for attention 20

SPARE 21

Channeling effect of the charged particles in the bent crystal Mechanically bent crystal Using of a secondary curvature of the crystal to guide the particles 22

Multi stage collimation as in LHC q The halo particles are removed by a cascade of amorphous targets: 1. Primary and secondary collimators intercept the diffusive primary halo. 2. Particles are repeatedly deflected by Multiple Coulomb Scattering also producing hadronic showers that is the secondary halo 3. Particles are finally stopped in the absorber 4. Masks protect the sensitive devices from tertiary halo 0 beam core q 6 7 6. 2 10 tertiary halo & showers masks secondary collimator 1 m CFC secondary halo & showers tertiary collimator absorber 1 m W secondary halo & showers Sensitive devices (ARC, IR QUADS. . ) >10 Normalizes aperture [σ] primary collimator 0. 6 m CFC primary halo Collimation efficiency in LHC ≅ 99. 98% @ 3. 5 Te. V ü Probably not enough in view of a luminosity upgrade ü Basic limitation of the amorphous collimation system 23

Loss rate along the SPS ring Loss map measurement in 2011: intensity increased from 1 bunch (I = 1. 15 x 1011 p) to 48 bunches, clear reduction of the losses seen in Sextant 6. q Loss map measurement in 2012: maximum possible intensity: 3. 3 x 1013 protons (4 x 72 bunches with 25 ns spacing), average loss reduction in the entire ring ! 2011 data protons Reduction factor (Lam / Lch) q 2012 data protons (270 Ge. V) 24

25

The Wave Catcher board 26

Results of the simulations without fibers 27

Cp. FM detection chain components q Fused Silica HPFS 7980 (M. Hoek , “Radiation Hardness Study on Fused Silica”. RICH 2007) q Fibers with core and cladding made of fused silica (U. Akgun et al. , "Quartz Plate Calorimeter as SLHC Upgrade to CMS Hadronic Endcap Calorimeters", CALOR 2008) q Hamamatsu R 672 & R 7378 A (A. Sbrizzi LUCID in ATLAS ) 28

Time resolution - Time resolution measurements were performed with multi-particles events (Ne=236) so as to have at least 1 particles in the first microbunch. RMS=266 ps Ne=236 RMS=455 ps 10 ns We don’t know the cause of the 2 distinct distributions 20 ns -We measure the delay from trigger edge (LINAC NIM Timing signal) of the first particle of each event - The Cp. FM take out a distribution similar to the CALO’s one but by far less light 29

I bar with + bundle + PMT 1 last run 235 (low flux) Online analysis Preliminary We have a signal even in the single-particle regime 30