J-PARC Neutrino Primary Beamline Beam Induced Fluorescence Profile - - PowerPoint PPT Presentation

J-PARC Neutrino Primary Beamline Beam Induced Fluorescence Profile Monitor R&D

• M. Friend, S. Cao, K. Sakashita (KEK), M. Hartz (IPMU/TRIMUF),
• A. Nakamura, Y. Kosho (Okayama University)

September 19, 2018

Outline

• J-PARC and Neutrino Primary Beamline Overview
• BIF Monitor R&D
• Gas Injection System R&D
• Optical System R&D

2 / 29

J-PARC Accelerator

• J-PARC = Japan Proton Accelerator Research Complex
• Accelerates proton beam to 30 GeV by:
• 400 MeV Linac (linear accelerator) → 3 GeV RCS (Rapid Cycling

Synchrotron) → 30 GeV MR (Main Ring)

• Fast Extraction (FX) scheme into neutrino primary proton beamline

3 / 29

J-PARC FX/NU Beam Parameters

• Currently : 485 kW with 2.48 s repetition rate (2.4 × 1014 PPP)
• 500+ kW achieved during beam tests
• Plan to upgrade MR power supplies in 2020 to reach 1.3 s

repetition rate

• RF improvements can allow for further decrease to 1.16 s
• Further improve beam stability, reduce MR beam losses to increase

number of protons per pulse Bunch Structure: 8 bunches/pulse Bunch Timing: 80 ns/bunch (3σ) Bunch Spacing : 581 ns/bunch Pulse Timing : 4.2µs/pulse

4 / 29

Neutrino Primary Proton Beamline

5 / 29

Neutrino Primary Proton Beamline

c

• Arc section : superconducting

combined-function magnets

• Used to sharply bend the

beam towards the Super Kamiokande direction

• Preparation section : normal

conducting dipole and quadrupole magnets

• Used to bend and focus the

proton beam extracted from the MR accelerator

• Prepare the beam to be safely

transported through the superconducting Arc section

• Final Focusing section : normal

conducting dipole and quadrupole magnets

• Used to bend and focus the

proton beam correctly onto the neutrino production target

• Proton beam position, angle,

size at the target must be carefully controlled

6 / 29

Neutrino Primary Proton Beam Monitors

Beamline Final Focusing Section Beam Direction →

• Beam monitors are essential for protecting beamline equipment and

understanding proton beam parameters for neutrino flux MC

• 5 CTs (Current Transformers) – monitor beam intensity
• 50 BLMs (Beam Loss Monitors)
• 21 ESMs (Electrostatic Monitors) – monitor beam position
• 19 SSEMs (Segmented Secondary Emission Monitors) –

non-continuously monitor beam profile + Profile Monitor R&D

• 1 OTR (Optical Transition Radiation) Monitor – continuously

monitors beam at neutrino production target

• Working on non-destructive profile monitor R&D for continuous

profile monitoring + long-term robustness at high beam powers

7 / 29

Destructive Profile Monitoring

Segmented Secondary Emission Monitor (SSEM)

• Protons interact with Ti foils

(15µm/SSEM)

• Secondary electrons emitted from

segmented cathode plane

• Compensating charge in each

cathode strip is read out by ADC Optical Transition Radiation Monitor (OTR)

• Ti foil @target (50µm thick)
• Optical Transition Radiation

produced when charged particles travel between two materials with different dielectric constants

• OTR light proportional to

beam profile

• Light detected by rad-hard

camera in low-rad area 8 / 29

BIF Project Goals

Design and fabricate a non-destructive beam profile monitor to be used continuously in the neutrino primary extraction beamline

• Beam loss should be reduced < 5 × 10−8 for one monitor
• For <100µSv/hr on contact @1.3MW, 3months running
• c.f. 1 NU SSEM (15µm Ti foils), where beam loss is ∼ 5 × 10−5
• Loss of 1 × 10−9 corresponds to ∼1 m of N2 gas at 1 × 10−2 Pa
• Monitor should work well at > 1.3MW (3.2 × 1014 PPP)
• Because monitor is designed for high-power, continuous operation,

does not need to work well down to very low beam power

• Minimum PPP of > 1.3 × 1014 (250kW @2.48s repetition rate)
• Aim for bunch-by-bunch profile measurement, but may be difficult
• At least must make spill-by-spill profile measurement

Profile reconstruction precision should be comparable to SSEM precision

• SSEM position measurement precision is 0.07 mm
• SSEM position measurement stability is ∼±0.15mm
• SSEM width measurement precision is 0.2 mm
• SSEM width measurement stability is ∼±0.1mm

→ Plan is to develop Beam Induced Fluorescence (BIF) monitor

9 / 29

BIF Design Principle

Beam Induced Fluorescence (BIF) monitor

• Uses fluorescence induced by proton beam interactions with gas

injected into the beamline

• Continuously and non-destructively monitor proton beam profile

Basic details :

• Protons hit gas (i.e. N2) inside the

beam pipe

• Molecules are excited by interaction
• Gas may or may not be ionized
• Gas fluoresces when electrons fall to a

lower energy state

• Pattern of fluorescence light is

proportional to profile of proton beam that excited the gas

• Fluorescence light exits beampipe through viewport
• Observe 1D vertical profile from the side and 1D

horizontal profile from the bottom

10 / 29

BIF R&D Overview

Now doing R&D for various components :

• Gas injection :
• For ∼1000 photons/spill, need to **locally** degrade vacuum level

from ∼10−5 → ∼10−2 Pa while maintaining average vacuum level of ∼10−4 Pa at ion pumps and ∼10−6 Pa at SC section

• Use pulsed gas injection to reach 1 × 10−2 Pa during each beam spill
• Continue to use ion pumps throughout NU beamline for radiation

safety reasons

• Light transport and focusing :

Must be radiation hard (∼10 Gy/month)

• Light detection :
• Must work in/near radiation environment
• Must work down to very low light levels (maximize gain)
• Should be fast to compensate for drift of any ions in proton beam

space charge field

• 2 methods of photon detection under development :

1 Focus light onto optical fibers which direct the light to MPPCs in

sub-tunnel ← main option under development now

2 Focus light onto gatable image intensifier coupled to rad-hard CID

camera ← backup plan; could use for 1/2 readout planes

11 / 29

BIF Number of Detected Photons

• Number of detected photons (Ndet

γ

) should be 1000 to reconstruct Gaussian profile with reasonable stat. error (1/ √ N =< 3%)

• Note : Gaussian beam width is 4∼8mm at install position
• Number of detected photons is Ndet

γ

= NpNp

γ ǫ

• Np =∼ 2.5 × 1014 protons per spill
• Number of produced photons is Np

γ = dE/dx × αγ × ∆z × ρ

• Energy loss dE/dx = 0.2keVm2/g and number of photons per

deposited energy αγ = 0.278keV −1 for N2 (Nucl. Instrum. Meth., A492:7490, 2002)

• If profile is integrated over ∆z = 20mm along the beam direction,

ideal gas law (ρ = PM/RT = 0.011P for N2) gives number of produced photons Np

γ = 1.2 × 10−5P

• Reasonable light collection acceptance × efficiency

ǫ ≃ 4.1 ∼ 7.1 × 10−5 → Need pressure P = 5 ∼ 8 × 10−3 Pa to detect 1000 photons

→ Must safely degrade vacuum to ∼ 10−2 Pa at BIF interaction point → Should maximize optical collection acceptance × efficiency in order to reduce required beamline pressure

12 / 29

Gas System Design Goals

• Achieve 10−2 Pa during each beam spill while maintaining average

vacuum of 10−4 Pa at ion pumps and 10−6 Pa at SC section entrance

• Ion pump lifetime is proportional to average pressure and is ∼4

months @10−3 Pa or ∼3.5 years @10−4 Pa

• 10−6 Pa required to prevent quench at SC section
• Achieve gas non-uniformity of <5%
• Gas non-uniformity directly scales measured profile, so 5% fluctuation
• f gas density → 5% distortion of profile
• Could correct profile by expected gas uniformity, if necessary, but

prefer not to

• Achieve gas injection with 20% pulse-by-pulse stability
• Because BIF does not make a spill-by-spill absolute beam intensity

measurement (just position and width measurement), absolute pulse-by-pulse gas stability doesn’t affect measurement

• However, stability does need to be good enough to assure the

required pressure to ±20% during each beam spill

• In order to ensure enough photons are produced
• ie decrease of gas pressure of 20% would degrade statistical error

1/ √ N ∼ 3% → 3.5%

13 / 29

Test Vacuum Chamber Studies

• Performed studies using 2 different small test vacuum chambers
• 150mm diameter, 840mm and 1020mm length
• Compared results of 2 chambers to understand gas flow dependence
• n chamber shape and benchmark gas injection simulations

14 / 29

Gas System Simulation

• Ran simulation of steady-state gas injection by COMSOL (FEM),

Molflow (MC) with same test vacuum chamber configuration

• Difference in expected pressure for 2 chambers is small, but may be

possible to resolve

15 / 29

Gas System Test vs Simulation

• Confirmed that steady-state simulation of test chamber matches

measured pressure in test chamber

• Measurement and simulation are consistent for 2 different test

chamber configurations at 2 different vacuum gauge positions for both simulation methods

• Unfortunately, systematic error on vacuum gauge measurement is

high

Near Inlet Near Pump

16 / 29

NU Full Beamline Simulation

• Simulation of steady-state gas flow in full NU FF beamline by

COMSOL

• Model present full NU FF

beamline configuration + Gas injection inlet + 2 optical windows at BIF interaction region for x and y readout + 2 new ion pumps upstream of gas injection point

• SC section treated as 1000

L/s pump

• Downstream pump @beam

window treated as 100 L/s pump

• Inject 2 × 10−7 kg/s N2 gas to reach steady-state 1 × 10−2 Pa at

BIF interaction region

17 / 29

Gas Injection Calculation

• Based on COMSOL steady-state simulation of full FF
• If 2 × 10−7 kg/s N2 continuous gas injection

→ Reach 1e-2 Pa at BIF interaction region (but pressure at ion pumps, SC section is unacceptable) → Can have acceptable average vacuum level at pumps and SC if gas is injected at 1Hz with 0.5% duty factor (5ms pulse)

• Probably actually want ∼0.05% duty factor (500µs pulse)
• Now working on pulsed gas injection tests + Molflow time-dependent

simulation to understand if this simple scaling is reasonable

2 × 10−7 kg/s Gas Injection at Steady State (During 4µs Beam Spill) Pressure (Pa) Molar flux (mol/m2/s) BIF Interaction Region 9.7e-3 9.4e-4 2 × 10−7 Gas Injection Pulsed w/ 0.5% Duty Factor Upstream (New) Pump 6.5e-5 6.3e-6 Downstream (Orig.) Pump 1.6e-5 1.6e-6 SC Sec. 4.6e-6 4.5e-7

18 / 29

Pulsed Gas Injection Tests

• Tests show we can inject ∼0.001kg/s N2 using pulse valve
• Testing pulse valve for stability, etc
• Trying to understand gas pulse shape

compared to hand calculation and Molflow time-dependent MC simulation

• Found that changing shape of connection between valve and

beampipe can change gas pulse peak by factor of 4 !

• Tested using different orifice plates – finalizing valve shape now

19 / 29

Vacuum Chamber Installation in Beamline

• Plan to install BIF in 3.9m straight, empty section of NU FF

beamline

• Will install gas injection valve, viewports in beamline during current

shutdown

• Use modular design for easy re-configuration if necessary

20 / 29

Space Charge Effects

• J-PARC beam has ∼ 3 × 1013 protons/bunch × 8 bunches (largest

protons-per-bunch in the world ?)

• Major challenge if ionized gas drifts in the space-charge field

induced by the passing beam

• Any drift of the ionized gas before the gas fluoresces would cause

distortion of the measured profile → Lifetime of the excited state + percentage of excited gas ionized (vs non-ionized) is very important !

• Transverse field induced by

proton beam space-charge in neutrino primary line (for ∼1/2

• f the current 1-bunch beam

intensity) :

• More than 4 kV/mm !
• Possible to mitigate by fast
• ptical readout – gate readout

timing to only use light from ions which fluoresce before drifting in the space-charge field

21 / 29

Optical Fiber + MPPC Optical System

• Focus light from viewport on

beampipe onto array of optical fibers

• Transport light away from high

radiation near beampipe to

• ptical sensors (MPPCs) in

lower-radiation subtunnel (behind shielding)

• MPPCs are cheap and fast
• But not radiation hard
• Simulations + tests show sufficient light collection
• However, should optimize transmission and collection efficiency as

much as possible

22 / 29

Optical Fiber + MPPC Simulation

• Light transport through optical components simulated by GEANT4
• BIF light focused by 2x plano-convex lenses + 45◦ mirror on array of
• ptical fibers → readout by MPPCs
• Readout in y to get horizontal profile
• 12 channels w/ 1.4mm pitch in this example
• Integrate over z (beam direction) to increase number of detected

photons

23 / 29

Optical System Simulation Result

24 / 29

Optical Fiber + MPPC Optical System Installation

• Installed prototype lens/mirror system in the beam line
• 800µm silica core optical fiber placed at focal point of installed
• ptical system (single channel measurement of light in beam pipe)

25 / 29

Beam-Induced Noise On Optical Fibers

• Installed optical fibers coupled to 3x3mm MPPC in planned BIF

location

• + New test FADC for readout with beamline DAQ
• Found very large beam-induced background on the optical fibers :
• ∼150 p.e.’s/spill in-bunch timing + ∼150 p.e.’s out-of-bunch timing

(c.f. expected 1000 detected BIF photons per spill / 30 fibers)

• Suspect due to Cherenkov light (on-timing) and neutrons (off-timing)
• Correlated w/ BLM44 – seems to be due to scattering from FQ2
• Must mitigate by shielding or moving fibers or optical filtering or...

26 / 29

Optical Background Simulation

• Optical simulation of BIF profile measurement including

beam-induced background on fibers

• Measured width decreases as background increases
• True width = 4mm 1σ, measured width = p2
• Aim to keep background light level <20 photons per fiber to have

accurate profile measurement w/ unknown background level

• If the background light level is known and fixed, the width

measurement becomes more stable

• Larger background may be OK if we can measure it accurately

→ Need to reduce background light + understand background stability

27 / 29

Image Intensifier + CID Camera Option

• Other photon detection option :

gated image intensifier (w/ MCP) coupled to radiation-hard (CID) camera by fiber taper

• Image intensifier+CID

resolution is much higher than

• ptical fiber array
• Radiation hardness should

also be better than fibers

• MCP being tested has

minimum gate time of 30ns – can’t cut slow background components

• Testing radiation hardness, response of single stage MCP + CID
• Single stage MCP gain is too low (∼5000) – would need to use dual

stage MCP in final monitor

• So far, radiation hardness looks good, but need further tests of

beam-induced noise near beamline

28 / 29

Conclusion

• R&D for BIF for the J-PARC neutrino extraction beamline is

underway

• Use pulsed gas injection to achieve 10−5 → 10−2 Pa pressure bump

during beam timing

• Fast optical system using optical fibers coupled to MPPCs under

development

• Methods to mitigate beam-induced noise on optical fibers are under

study

29 / 29

Backup Slides

30 / 29

The T2K Experiment

• Primarily νµ or ¯

νµ, 2.5◦ off axis neutrino beam produced at J-PARC

• ND280 Near Detector

– 280 m from ν source

• Constrains systematic errors
• Measures ν cross sections and beam backgrounds
• Neutrino interactions detected at the Super-Kamiokande (SK) far

detector – 295 km from ν source

• 22.5 kT fiducial volume water Cherenkov detector
• Good performance of νe/νµ particle ID for sub-GeV energy ν’s
• νe appearance and νµ disappearance ν oscillation information

31 / 29

How to Make a Neutrino Beam

• Slam high-energy high-intensity proton beam into long carbon target
• Focus outgoing hadrons (pions, kaons, etc.) in electro-magnetic

focusing horns

• Switch between ν- or ¯

ν-mode by changing the horn polarity

• Pions decay to muons and νµ’s in long decay volume
• Stop interacting particles in beam dump; neutrinos continue on to

near and far detectors

• Number of neutrinos is proportional to number of protons

→ Increase beam power to increase neutrino flux

32 / 29

Why Is Proton Beam Monitoring Important?

• Required for beam diagnostics and tuning
• Required to correctly steer the proton beam/protect beamline

equipment

• Even one shot of mis-steered high-intensity beam can seriously

damage equipment – need continuous monitoring

• Information from proton beam monitors is used as input into the

T2K neutrino flux prediction simulation

• Need well-understood and well-controlled proton beam for world-class

neutrino physics results

33 / 29

Why Is Non-Destructive (+ Minimally-Destructive) Proton Beam Monitoring Important?

• Standard monitors measure the beam profile by intercepting the

beam – they are destructive and cause beam loss

• Absolute amount of beam loss is proportional to beam power and

volume of material in the beam

• Beam loss can cause :
• Irradiation of and damage to beamline equipment
• Increased residual radiation levels in the beamline tunnel
• Foils in the beam may degrade
• Rate of degradation may increase as the beam power increases
• The beam profile must be monitored continuously
• So, R&D for J-PARC proton beam profile monitors that work well at

high beam power is ongoing

34 / 29

Measured Beam Loss Due to SSEMs

• Beam loss when SSEMs are IN is quite high
• ∼0.005% beam loss at each SSEM
• Can cause radiation damage, activation of beamline equipment
• SSEMs upstream of the neutrino target station cannot be used

continuously

35 / 29

Measured OTR Degradation

• J-PARC neutrino beamline OTR generally working well, but ...
• Gradual decrease in OTR signal size with integrated incident POT

has been observed

• Foil darkening where the beam hits also observed
• Materials properties study of previously used OTR foils ongoing

→ poster by T. Ishida, et al.

36 / 29

General BIF Design

• Plan to install BIF monitor in 3.9m straight, empty section of NU

FF beamline, 12.5m downstream of SC section entrance

• Basic design fixed :
• Use pulsed gas injection to reach 1 × 10−2 Pa during each beam spill
• Continue to use ion pumps throughout NU beamline for radiation

safety reasons

• Optical transport of photons away from beamline used to reduce

radiation on photon detection equipment

• Radiation hardness of optical components is essential
• 2 methods of photon detection under development

1 Focus light onto optical fibers which direct the light to MPPCs in

sub-tunnel ← main option under development now

2 Focus light onto MCP (gatable image intensifier) coupled to rad-hard

CID camera ← backup plan; could use for 1/2 readout planes

37 / 29

BIF Parameter Summary

• Number of generated photons (@2.5 × 1014 PPP) : 3 × 107
• Gas pressure to achieve that : 1 × 10−2 Pa
• Number of photons detected : 1000
• Optical system acceptance × efficiency to achieve that : 4.1 × 10−5
• Position measurement precision : 0.2mm
• Width measurement precision : 0.3mm
• Optical sensor resolution to achieve that : 1mm

38 / 29

Pulsed Gas Injection

• Using Parker Hannefin Series 9

pulse valve

• Normally closed valve
• Valve opened by electronic pulse

from associated controller (IOTA ONE)

• Requires homemade 30m long

cable to control from subtunnel (tested)

• Requires homemade 100m

long cable to control from NU2 (will test soon)

• 160µs minimum pulse time
• 2 pulse valves in series as safety

mechanism in case one fails

• Connected by small, 0.03L

chamber

• PEEK (rad-hard) poppet seems

to be commercially available for this valve

Pulse Valve Cutaway Pulse Valve Safety Chamber

39 / 29

Gas Injection Stability and Uniformity

• Gas pulse-by-pulse injection stability :
• Under testing, confirmation now
• However, limit on pulse-by-pulse fluctuation of absolute amount of

injected gas isn’t so strict (±20%) as long as uniformity is <5%

Gas uniformity :

• Based on simulations,

uniformity of gas is OK

• Uniform to < 4%
• ver full chamber

cross-section

• Non-uniformity < 2%

at interaction region (4∼8mm beam σ)

• But anyway, plan to

continue checks of gas uniformity with simulation + vacuum injection tests

40 / 29

Black Paint for Beampipe Walls

Black paint is needed on some parts of beampipe walls

• To eliminate distorted reflection of BIF light
• Now considering Diamond-Like Carbon (DLC) coating on small

portions of the beampipe (not full coating)

• Or black anodized beampipe walls or baffles
• Can use optical simulation to determine points on beampipe walls

where reflections contribute to background

• Vacuum safe
• Outgassing of 1 × 10−8 Pa m3/s/m2
• Used in space applications and for radiation protection of silicon solar

cells

• May be suitably radiation-hard
• Aim to get samples and start testing soon

41 / 29

Optical System Requirements

42 / 29

MPPC Specifications

43 / 29

MPPC Amplifier Specifications

44 / 29