T2K Target and Beam Window Upgrades for 1.3 MW Operation Chris - - PowerPoint PPT Presentation

T2K Target and Beam Window Upgrades for 1.3 MW Operation

Chris Densham, Mike Fitton

(STFC Rutherford Appleton Laboratory)

• T. Nakadaira, T. Ishida, M. Tada, T. Sekiguchi

(KEK Beam Group)

A.Wilkinson, J. Gong

(Oxford University Materials Science)

P.Hurh (Fermilab, RaDIATE collaboration)

1

T2K Target & horn

• Helium cooled graphite rod
• Design beam power: 750 kW
• Beam power so far: 435 kW
• 3% beam power deposited in

target as heat

• 1st target & horn replaced after

4 years, 6.5e20 p.o.t.

• 2nd target OK after 2.2 e21 p.o.t.

Next target under construction

Beam and Window Parameters

3

750 kW 1.3MW

Design T2KII path Beam Energy [Gev] 30 30 Protons per spill [-] 3.30E+14 3.20E+14 Energy deposited per kg per proton [J/kg/proto n] 2.52E-10 2.52E-10 Energy deposited per kg per pulse [J/kg/pulse] 83300 80640 Cycle time [s] 2.1 1.16 Spill length [s] 4.13E-06 4.11E-06 Number of bunches [-] 8 8 Bunch length [ns] 58 40 Gap length [ns] 523 541 Peak Heat Generation [J/m^3/s] 8.15E+14 1.14E+15 Beam sigma [mm] 4.24 4.24 Heat load per spill [J/cc/pulse] 378.18 366.11 Heat load per sec [W/cc] 180.09 315.61 Peak Temp per bunch [C] 19.78 19.15 Thermal stress per bunch [MPa] 61.27 59.32 Peak Temp per pulse [C] 158.27 153.22

1.3 MW beam window and target studies

• Q: Can we push existing beam window and

target design from 0.75 -> 1.3 MW?

• Q: If so, what changes are needed for

target and beam window materials, design, manufacture, and helium plant (pressure/flow rate)?

• What is expected lifetime for upgraded

target and, more crucially, beam window?

4

Target upgrade status

• The increase in beam power from 0.751.3MW will

increase the integrated heat load on the target (approx. 23kW  41kW).

• Just increasing the flowrate will lead to very high

velocities and pressure drops (current velocities already exceed 460m/s).

• It is proposed that the helium cooling systems operating

pressure is increased to reduce velocities and pressure drops.

• Increasing the pressure will increase the helium density and therefore

heat transfer coefficient.

• Higher operating pressure has the advantage of reduced pressure drop.
• Running at higher pressure reduces the pressure drop and max velocity
• Compressor will be cheaper to purchase and operate (less power

consumed).

CFX model heat inputs (MARS)

Potential solution for 1.3 MW operation

→ 5 bar helium

7

0.75 MW 1.3 MW

Heat load 23.5 kW 40.8 kW Helium pressure 1.6 bar 5 bar Helium mass flow 32 g/s 60 g/s Pressure drop 0.83 bar 0.88 bar

• Velocity contours

‐> max 400 m/s (OK for helium)

Conjugate Heat Transfer analysis

Increased beam power, flow rate & pressure

• 60g/s helium, 5barG outlet pressure – 1.3MW beam power

Helium velocity flow lines

Thermal analysis for 1.3 MW operation

9

0.75 MW 1.3 MW

Helium pressure 1.6 bar 5 bar US window temp 105 °C 157 °C DS window temp 120°C 130°C Max graphite temp. (for 1/4 conductivity) 736°C 900°C

400°C 600°C

Reduction in thermal conductivity(from fast neutrons)

900°C

Thermal analysis assuming x4 reduction in thermal conductivity

It’s better hot

Some iterating still to do

Pressure on upstream window (0.5 mm)

• Pressure stress and deformation in current beam window (Ansys)
• 5 bar pressure inside of target (proposed operating pressure)

Stress in thin dome section agree with hand calculations (10.13MPa @ 1.6bar, 0.5mm) Max stress  75MPa

Reducing pressure stresses in upstream window at 5 bar

• A parameterised model of the window has been optimized using a genetic

algorithm.

• Pressure stress can be halved from 75 -> 34 MPa by increasing outer plate

thickness from 7 -> 10 mm

• Window thickness remains unchanged at 0.5mm thick.

Parameter ranges (limits)

Optimum found (34MPa) Plate thickness = 10mm External radius = 7mm Internal radius = 25mm Current design (75MPa)

Effect of pulsed beam on T2K target

Inertial ‘violin modes’ Stress distribution after

• ff-centre beam

spill Radial stress waves – on centre beam spill

8 MPa 0.5 µs beam spill

The ultimate destiny for all graphite targets?

(T2K: c.2 x 1021 p/cm2 so far)

LAMPF fluence 10^22 p/cm2 PSI: fluence 10^22 p/cm2 NuMI target

Remote Target Exchange System

14

Target exchanger and manipulator

system

Helium cooling pipe ceramic development

• New bolted design for ceramic isolator, metal seals

– To make more resilient to thermal cycling

Helium temperature cycle for 0→380 kW Ti brazed to ceramic

T2K beam window upgrade plans for - 1.3 MW

Pressure drop 0.06bar @ 1.1g/s Max velocity 230 m/s Strong recirculation zones driven by high speed jet

16

• Helium flow lines
• 1.3 MW beam

(steady state/CW simulation)

Through-thickness Stress Waves in Beam Window

• Constructive interference of bunch structure (8

bunches) possible in existing 0.3 mm window.

• For 0.3 mm window, the stress wave travels from
• ne surface to the other and back in ~100 ns  10

MHz stress cycle

• 0.32 mm  Peak stress in Z = ~20 MPa
• 0.30 mm  Peak stress in Z = ~ 270 MPa

17

Comparison of stress in Z (through window stress) as function of time at window centre (max stress point) for 0.3, 0.5 and 0.7 mm thick windows @ 1.3 MW

‐300 ‐200 ‐100 100 200 300 1000 2000 3000 4000 5000 6000 SZ [MPa] Time [ns] 0.7 mm 0.5 mm 0.3 mm

Beam Window Stress Wave Analysis

18

NB. SZ = through-thickness stress SX = radial stress

Current window thickness tolerance

• too tight
• c. 0.01 mm

Proposed upgrade window thickness = 0.4 ± 0.05 mm

• much better

tolerance for manufacture 0.5 mm also good

• But more

activation, higher heat load

• Good for target

window upgrade

• 100
• 50

50 100 150 200 250 300 350 400 450 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 Stress [MPa] Window thickness [mm] Equivalent Stress Z Stress X Stress

1.3 MW

Next beam window (update)

750kW, 0.5mm thick Ti-6Al-4V window, long duration transient analysis

10MHz through thickness oscillation,

• approx. 2.7MPa

Reflected radial wave from free surface (simplified geometry) Stress increasing from thermal expansion (conduction) Stress at end of 8th bunch Bunch-to-bunch stress amplitude ~20MPa

100 MPa

Transient Analysis – 750kW

20

• New proposed beam

parameters:

• 2.0 x 1014 ppp @ 1.28 s rep

rate (750 kW)

• HTC = 886 W/m2.K@ 300 K
• Results summary:

Thickness [mm] Peak temp [K] Peak stress [MPa] Quasi-static average stress [MPa] ± stress limits [MPa]

0.3 418.7 56.1 32.2 23.9 0.4 436.8 71.2 44.8 26.4 0.5 442.2 79.5 50.9 28.6 0.7 463.5 95.2 65.5 29.7

Effects of elevated temperature, fatigue and radiation damage on beam window

• N. Simos

(BNL) Current window 22.4x1020 pot  4.5 dpa (c/o T.Davenne) 0.24 dpa

Significant loss of ductility at 0.24 DPA Existing window entirely brittle? Does it matter? Low stress at moment. Time to ask the materials scientists… RaDIATE

1.3 MW, 0.5 mm 0.75 MW, 0.3 mm

21

Next (or not) beam windows

• Domes made from Ti-6Al-4V ELI (Grade 23)
• Plate used instead of bar – maybe better properties at

centre (or not?)

• Spare material used for material characterisation.

Electron Back‐Scatter Diffraction

Characterises microstructure of crystalline materials

NB Following figures show grain orientations in through-thickness direction (of beam window)

Department of Materials University of Oxford

AJW 2016

EBSD: 2” thick grade 23 plate

• Larger grains but less texture,

macrozones less evident in the bar

• Current window (from bar) has

performed well so far

• Recommend staying with bar
• Irradiation samples taken from 200

mm (8”) diameter bar

EBSD: Centre of 8” dia. bar

• Plate: fine grains but large

macrozones (= regions with similar crystal orientations inherited from large prior beta grains).

• Microstructure not as refined as it

first appears.

• The effective structural unit size may

be much larger than it initially appears

• Could impact badly on fatigue

properties.

100 µm 100 µm

• 6 x 0.3 mm thick discs machined from same 8”

Ti-6Al-4V bar purchased for next T2K window domes

• 2 x sample foils polished to 0.25 mm and laser

cut using very fine scanning laser

• Foils installed at BNL/BLIP for 10 weeks

irradiation at c.180 MeV, c.1 DPA

Centre ‘pip’ taken for SEM, EBSD

Irradiation starting tomorrow

Department of Materials University of Oxford

AJW 2016

Measurement of deflection during resonant vibration (Oxford Univ. materials)

Input laser beam

160 mm ~0.65°

Specular Reflection Movement of beam due to deflection Need to operate near but not on resonant frequency (c.20kHz) to generate stress range ‐> need to measure amplitude

Tresca stress

Chris Densham

Plan to reproduce test equipment at Culham lab. for active samples testing

3 mm

Department of Materials University of Oxford

AJW 2016

Ti-6Al-4V Fatigue Life Data

from Lutjering & Gysler Titanium Science and Technology

Summary

• Plan to upgrade beam window thickness from 0.3 mm -> 0.4

mm to increase tolerance of beam window to thickness/bunch structure (and increase helium pressure)

• Operation at 1.3 MW appears feasible for both target and

beam window with incremental design changes

• Much work to do – particularly for high pressure helium

cooling for target:

– windows, welds, helium plant etc – thermal & pressure stresses

• Radiation damage in window material is main question mark
• Ti6Al4V meso-fatigue samples installed in BLIP facility
• Mesoscale fatigue tests planned at Culham for irradiated

material 28