Beam Window at TS entrance

1
Beam Window at TS entrance
T.Iwashita (KEK)

• Introduction
• Material
• Cooling
• Shape
• Future plan
• Summary

2
Introduction

• A beam window separates primary beam-line and
  target station (TS) helium-vacuum
• Must be thin (gt2mm Ti, beam loss lt 1)
• Must endure heat and heat stress from primary
  proton beam (3.3X1014ppp, 3.53sec rep. spill
  width 5msec)
• Must endure 1 atm. of pressure between beam-line
  vacuum and helium vessel
• Need remote maintenance system

Target Station
Helium vessel
Final focus
Beam-line
Decay volume filled with He
Beam window
3
Energy deposit on the window

• Al, SUS, Be, Ti etc.
• Assuming 10cm-f, 1mm-t window
• Energy deposit 34J/spill (Al 1mm-t) from GEANT
• 56W/g for thin material

4
Temperature rise of several metal

• 56.3W/g inside rlt0.5cm
• t1mm?temperature is 30oC at r10cm(assume
  cooling)
• Temperature is lower than melting point
• Must be cooled (by He) except for Al and Be

5
Thermal stress

• Ti alloy has large permissible stress
• (Permissible stress is the smaller one 1/3 of
  tensile stress)
• SUS cant be used because of large thermal stress
• Material except Ti alloy has smaller safety
  factor
• Ti alloy is our candidate

6
Cooling window

• Edge cooling by water etc. (30oC)
• Need piping around the window
• Surface cooling by He etc.
• Need double window system

Heat deposit by beam
He gas flow
Surface cooling
Edge cooling
7
Temperature risewith surface cooling
250oC
450oC
580oC
0W/m2/K
10W/m2/K
100W/m2/K

• Ti alloy window (t0.25mm)
• Assume surface cooling (0, 10, 100W/m2/K)
• 10W/m2/K is free convection by He
• 100W/m2/K is forced convection by He
• gt100W/m2/K is achieved in cooling test (by Belle
  group)
• 100W/m2/K is better (Edge cooling by water
  doesnt affect much when 100W/m2/K of surface
  cooling is assumed)

8
Window shape

• Stress from He pressure is very large for a flat
  window whose radius is 10cm, tlt1mm (gt100MPa for
  Ti)
• Shape like high-pressure vessel is essential

Beam
Beam
Beam
Flat window with taper
Hemispherical window
Half-cylindrical window
9
ANSYS results (thermal stress)
Half-spherical Ti-alloy window for t1mm, 0.5mm,
0.25mm
Permissible Stress (MPa)
390MPa
390MPa
390MPa

• Temperature reaches 300K in case of t1mm
• t0.25mm is better

10
Comparison of window shape
Stress (MPa) by heat and pressure
Displacement (mm) by heat and pressure
400
0.5
Permissible stress of Ti alloy (with safety
factor 3)
Flat r10cm
Hemisphere r10cm
Hemisphere r5cm
X0.7 (under high temperature)
0.25
200
X0.7 (fatigue)
Thin hemispherical Windows are safe
0 0.25 0.5 1.0 mm
0 0.25 0.5 1.0 mm
Thickness
Thickness
1 atm of He pressure and 100W/K/m2 surface
cooling are assumed
11
Which shape is the best?

• Taper
• Easy to be made
• Weakest for pressure (Considered to be a spare
  option)
• Spherical
• Considered to be strongest shape
• Cost much (difficult to make)
• Cylindrical
• Easy to be made by a thin Ti plate
• (Need to be established)
• No need to replace when OA changes
• Stronger than taper window for pressure
• Thermal stress is not estimated correctly

12
Dynamic analysis with ANSYS

• Ti-6Al-4V, t1mm
• No He pressure
• Atten. rate 0.005(default)
• Equiv. stress is consistent with analytic
  calculation

0mm
Displacement
?33mm
?60mm
0s
3.5s
100MPa
Equiv. stress
60MPa
Temp. rise
165ºC
0MPa
?60MPa
100MPa
30ºC
?100MPa
0s
0s
3.5s
3.5s
13
Double window (cylindrical)

• Window system needs helium path...double window

Beam
He out
He in
Now Printing
Cut view
3D view
14
Future plan
Stress test
Stress calculation
Prototype of window
Design confirmation
Thermal test
Research of manufacturing
Nov.04
Mar.05
Nov.04Mar.05
Aug.04
Calculation of Helium flow
Aug.04
Prototype of Double window
Design confirmation of double window
Double window helium flow Cooling test
Cooling test to check heat transfer coeff.
2005 Fall
Winter 2005
2005 Summer
Aug.Sep.04
15
Summary

• As a beam window at the entrance of target
  station entrance, Ti alloy (Ti-6Al-4V) is a
  promising candidate
• Several type of window shape are examined
• Hemispherical or half-cylindrical windows are
  safety
• A conceptual drawing of half-cylindrical window
  is ready

16
Supplement
17
Performance of Ti-alloy
Yield strength of Ti-6Al-4V at high temperature
MPa
Fatigue of Ti-6Al-4V
MPa
1000
1000
600
600
times
0 100 300 500 oC
104 105 106
3106 spill/year at J-PARCn target
70 at 300oC
70 at 106 times of hit
Yield strength become 0.70.70.49
18
ANSYS results (temperature)

• Initial temperature is 30oC (assume cooling at
  r10cm)
• Temperature rise in 1spill is 82K for Al, 108K
  for SUS
• Al is cooled by the next spill because of high
  heat cond.
• SUS continues to rise (reached 518K)

112
140
30
30
???
SUS
19
Window with remote flange

• Window with remote flange shield

Radial seal developed by Y.Yamanoi, et. al.
shield
Proton beam
Double wall
He gas
expanded by screws
Conceptual drawing
20
Frame for cylindrical window
Outer
Inner