Window Design report v30 1 Michael MONTEIL 12 April 2010 Specifications v30 Interface between machine vacuum and Atmospheric pressure 10 8 mbar P atm Protective atmosphere ID: 540687
Download Presentation The PPT/PDF document "HiRadMat" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
HiRadMat WindowDesign report v3.0
1
Michael MONTEIL - 12 April 2010Slide2
Specifications v3.0Interface between machine vacuum and Atmospheric pressure
10
-8 mbar / Patm
Protective atmosphere !!!Aperture min 60 mmResist to a proton beam size on the window :
1s = 0.5 mm
“Beam Size at the TT66 Vacuum Window”,
C. Hessler, 26.02.2010
2
Michael MONTEIL - 12 April 2010Slide3
Solution #5 : Be + C-C
Same as solution #4 but the pressure load is supported by a C-C plate
Simple window assembly
Thin thickness (no differential pumping…)Be cannot pollute vacuum chamber unless C-C fail
Tight
Price of Be but no pumps
Michael MONTEIL - 12 April 2010
3Slide4
Solutions - Sum-up#1: C-C (Differential pumping)
Protective atm (Nitrogen ?)
Radiations?#2: C-C + Graphite foil (useless now)#3: Tight steel “ring” with a C-C plate
#4: BerylliumSafety problem
#5: C-C + Beryllium
Michael MONTEIL - 12 April 2010
4
TodaySlide5
Different grades of Be
5
Michael MONTEIL - 12 April 2010
Data: Brush WellmanSlide6
Different grades of BePF-60 ?Low rate of Beryllium oxide compare to PS-200
Good quality-price ratio (Next slides…)
1.5 to 2 time cheaper than IF-1Almost the same temperature distribution as pure Be and IF-1 (IF-1 a bit better…)Used in CNGS…
Michael MONTEIL - 12 April 2010
6
Collaboration: J. BlancoSlide7
DesignSpecificationBe & C-CAperture min. 60mmDN80 or DN60 conical flange connection
15 cm depth maximumRemark
Cannot machine Be at CERNMichael MONTEIL - 12 April 2010
7Slide8
DesignCommon design – ChoicesStandard flanges only (cheaper)Be window assembled in lab between 2 flanges (safety)
Conical flange (faster assembly once in experimental area)
DesignConical Flange (plug-in flange)
Tube (connection conical flange <--> conflat flange)
2 x Conflat (Window in-between)
Michael MONTEIL - 12 April 2010
8Slide9
“CNGS” likeCNGSHiRadMat
– Option 1
Michael MONTEIL - 12 April 2010
9
Nota: Those drawing are drafts. Above dimensions are not representative of the realitySlide10
“CNGS” likeMichael MONTEIL - 12 April 2010
10
Data: Brush WellmanSlide11
“CNGS” likeMichael MONTEIL - 12 April 2010
11
Data: Brush WellmanSlide12
“TED @ TI2, TT40” – Beryllium version“TED @ TI2, TT40”
HiRadMat – Option 2
Michael MONTEIL - 12 April 2010
12Slide13
“TED @ TI2, TT40” – Beryllium versionQuote from BW
Michael MONTEIL - 12 April 2010
13Slide14
2 design proposalsOption 1
Option 2
Not that much
Precautions for the assembly
Non Standard conflat
assembly (Tightness)
Might be careful to not cut (shear cut) the Be foil during assembly
Michael MONTEIL - 12 April 2010
14
Life warranty on Be + flange assembly
Easy to assembly
Standard
conflat
assembly
Tightness OK
Not that much
Nota: Those drawing are drafts. Above dimensions are not representative of the realitySlide15
2 design proposalsCost estimationBe Foil
Option 1
Option 2
Number of foil to order : 3Spare : 1Window installed : 1
“In case we break a foil while assembling” : 1
Michael MONTEIL - 12 April 2010
15
Number of flange to order : 2
Spare : 1
Window installed : 1
Nota: Those drawing are drafts. Above dimensions are not representative of the realitySlide16
2 design proposalsCost estimationBe Foil
Option 1
Flange
Option 2FoilMichael MONTEIL - 12 April 2010
16
Nota: Those drawing are drafts. Above dimensions are not representative of the realitySlide17
About thickness, how does BW design their own Be foils?With (Thickness 0.25mm, radius 35mm, pressure 1.01
kPa, E 303Gpa, Poisson 0.08
)Resultssedge=
305MPa > 275 Mpa !!s
center= 297Mpa > 275
Mpa
!!
Michael MONTEIL - 12 April 2010
17
Data: Brush WellmanSlide18
However…BW : “With confirm that your calculations with reference to the DB450277 assembly are correct and show over the recommended values, however, the assembly was designed using empirical data as well taking into consideration the calculated values.
We have performed tests on this design and found it to be reliable, with units sold to customers over the years performing well under real-life conditions
.”ExplanationBecause of plasticity effects, Be foil withstands 1
Atm (according to BW tests) even if Roark’s calculation says that it doesn’t withstand
Michael MONTEIL - 12 April 2010
18
Data: Brush WellmanSlide19
To knowBe have ultra high resistance to fatigue crackingHigh endurance strength level
Michael MONTEIL - 12 April 2010
19
Data: Brush WellmanSlide20
Solutions #5
stresses and deflection -
C-C+Be
under D
P = 1
atm
Linear circular fixed support
2 planes of symmetry
Geometry
Diameter f
80 mm
Thickness: 0.254 mm
Aperture:
f
60 mm
Pressure 1
atm
20
Michael MONTEIL - 12 April 2010Slide21
ANSYS Study - Solutions #5
stresses and deflection -
C-C+Be
under D
P = 1
atm
Beryllium foil study
Smooth and continuous temperature distribution
Through-thickness energy deposition
Coefficient of Thermal Expansion varying with temperature
Be (pure elasticity):
Poisson’s ratio = 0.08
High R
e
= 303
Mpa
21
Michael MONTEIL - 12 April 2010Slide22
Michael MONTEIL - 12 April 2010
22Slide23
Michael MONTEIL - 12 April 2010
23Slide24
Michael MONTEIL - 12 April 2010
24Slide25
Michael MONTEIL - 12 April 2010
25Slide26
Michael MONTEIL - 12 April 2010
26Slide27
Michael MONTEIL - 12 April 2010
27Slide28
Michael MONTEIL - 12 April 2010
28Slide29
Michael MONTEIL - 12 April 2010
29Slide30
Michael MONTEIL - 12 April 2010
30Slide31
Michael MONTEIL - 12 April 2010
31Slide32
Michael MONTEIL - 12 April 2010
32Slide33
Conclusion: influence of gap reducingSo if we flatter the foil on the C-C, we reduce the Max stress (as shows ANSYS calculation with non plasticity model), maybe also stay in elastic domain (Bellow 275Mpa at room Temp).
We will manage to reduce this gap (flattering the Be foil as much as possible on C-C plate)
Michael MONTEIL - 12 April 2010
33Slide34
Easiness to reduce Gap C-C / BeOption 1
Option 2
Michael MONTEIL - 12 April 2010
34
+
-Slide35
To do :Order Beryllium
Delivery: 4 Weeks ARO for flanges (Option 1)Delivery: 6 Weeks ARO fro foil (Option 2)
AssemblyTestMichael MONTEIL - 12 April 2010
35Slide36
Michael MONTEIL - 12 April 2010
36Slide37
V2.0 slidesMichael MONTEIL - 12 April 2010
37Slide38
Window geometry – C-C option
Carbon/Carbon composite: 1501 G
from SGLCylindrical window
Diameter f 80 mmAperture
f 60 mm
Thickness: 0.5 cm
Aperture (
flange internal diameter
): f 60 mm
38
Michael MONTEIL - 12 April 2010Slide39
Solutions #1 for C-C tightness problem:Differential vacuum
(V2.0)
1 Window C-C
Pumping speed needed: 2.3x108
l/s …2 Windows C-C with differential pumping
Pumping speed needed: 8.94
x
10
2 l/s OK !3 Windows C-C with differential pumpingPumping speed needed: 13 l/s Too low ?!
39
Michael MONTEIL - 12 April 2010Slide40
Solutions #1What about radiations in this area ?Possible maintenance needed on the roots pump…
Protective atmosphere
Decreasing pressure in Vacuumside with serial pumps
Michael MONTEIL - 12 April 201040Slide41
Michael MONTEIL - 12 April 201041
P2 : Roots pump
100 –> 1500 m
3
/h
10
-3
-> 10 Bar
P3 : Ion pump
400 l/s
ReferenceSlide42
Solutions #2 for C-C tightness problem: Add a Graphite foil (v1.0)
42
Michael MONTEIL - 12 April 2010
Solution #3 : Tight
steel“ring
” with a C-C plate (v1.0)
Solution #4 : BerylliumSlide43
Michael MONTEIL - 12 April 2010
43Slide44
ANSYS Study - Solutions #1stresses and deflection -
C-C
under
DP =
1.4
atm
Linear circular fixed support
2 planes of symmetry
Geometry
Diameter f 80 mm
Thickness: 5 mm
Aperture:
f
60 mm
Pressure 1.4 bar
44
Michael MONTEIL - 12 April 2010Slide45
ANSYS Study - Solutions #1
stresses and deflection -
C-C under
D
P = 1.4
atm
Orthotropic properties : 18 plies [0°/90°…]
Smooth and continuous temperature distribution
Through-thickness energy deposition
Coefficient of Thermal Expansion varying with temperature and directions
45
Michael MONTEIL - 12 April 2010Slide46
C-C - Pressure load - Deflection
46
Michael MONTEIL - 12 April 2010
7.4
μmSlide47
C-C - Pressure load – Von-Mises
47
Michael MONTEIL - 12 April 2010
5.9
MpaSlide48
C-C - Pressure load – Tsaï-Wu
48
Michael MONTEIL - 12 April 2010Slide49
C-C - Thermal load ANSYS input =
FLUKA output
Radial
C-C | 1
s
=
0.5
mm
| 1.7e11 p+ | 288 bunches
Axisymmetrical
radial temperature field
Depth
49
Michael MONTEIL - 12 April 2010
T (°C)
R (cm)
T (°C)
Z (cm)Slide50
C-C - Pressure + Thermal load – Deflection
50
Michael MONTEIL - 12 April 2010
10.6
μmSlide51
C-C - Pressure + Thermal load – Von-Mises
51
Michael MONTEIL - 12 April 2010
31
MpaSlide52
C-C - Pressure + Thermal load – Tsaï-Wu
52
Michael MONTEIL - 12 April 2010Slide53
Michael MONTEIL - 12 April 2010
53Slide54
ANSYS Study - Solutions #4
stresses and deflection -
Be under
D
P = 1.4
atm
Linear circular fixed support
2 planes of symmetry
Geometry
Diameter
f
80 mm
Thickness: 0.254 mm
Aperture:
f
60 mm
Pressure 1.4 bar
54
Michael MONTEIL - 12 April 2010Slide55
ANSYS Study - Solutions #4
stresses and deflection -
Be under
D
P = 1.4
atm
Smooth and continuous temperature distribution
Through-thickness energy deposition
Coefficient of Thermal Expansion varying with temperature
Be:
Poisson’s ratio = 0.1
High R
e
= 275
Mpa
High
R
m
= 551
MPa
55
Michael MONTEIL - 12 April 2010Slide56
Be - Pressure load - Deflection
56
Michael MONTEIL - 12 April 2010
0.81 mmSlide57
Be - Pressure load – Von-Mises
57
Michael MONTEIL - 12 April 2010
319
MpaSlide58
Be - Pressure load – Safety factor Ult. Strength
Michael MONTEIL - 12 April 2010
58
1.7Slide59
Be - Thermal load ANSYS input =
FLUKA output
Be | 1
s
=
0
. 5
mm
| 1.7e11 p+ | 288 bunches
Axisymmetrical
radial temperature field
Z (cm)
T (°C)
59
Michael MONTEIL - 12 April 2010
Z (cm)
Radial Be
T (°C)Slide60
Be - Pressure + Thermal load – Deflection
60
Michael MONTEIL - 12 April 2010
0.8
mmSlide61
Be - Pressure + Thermal load – Von-Mises
61
Michael MONTEIL - 12 April 2010
315
MpaSlide62
Be - Pressure + Thermal load – Safety factor Ult. Strength
Michael MONTEIL - 12 April 2010
62
1.7Slide63
ANSYS Study - Solutions #5stresses and deflection -
C-
C+Be under
DP
=
1.4
atm
2 Studies
C-C
(See Solution #4)
Pressure load
Pressure + Temperature loads
Be
(Following slides)
Flattered on a C-C plate (Fixed support)
and apply pressure load on the other side
Flattered on a C-C plate (Fixed support)
and apply pressure load on the other side
+ Temperature load
2 planes of symmetry
Geometry
Diameter
f
80
mm
Thickness
C-C: 5 mm
Be: 0.254 mm
Aperture:
f
60
mm
Pressure 1.4 bar
63
Michael MONTEIL - 12 April 2010Slide64
ANSYS Study - Solutions #5
stresses and deflection -
C-C+Be
under
DP =
1.4
atm
Smooth and continuous temperature distribution
Through-thickness energy deposition
Coefficient of Thermal Expansion varying with temperature
64
Michael MONTEIL - 12 April 2010Slide65
Michael MONTEIL - 12 April 201065
Be (flatter on C-C) - Pressure load – DeformationSlide66
Be (flatter on C-C) - Pressure load – Von-Mises
Michael MONTEIL - 12 April 2010
66Slide67
Thermal load
ANSYS input =
FLUKA output
Radial C-C
C-C + Be | 1
s
=
0.5
mm
| 1.7e11 p+ | 288 bunches
Axisymmetrical
radial temperature field
T (°C)
Z (cm)
67
Michael MONTEIL - 12 April 2010
Z (cm)
Radial Be
Depth C-C
Z (cm)
T (°C)Slide68
Be (flatter on C-C) - Pressure + Thermal load – DeflectionMichael MONTEIL - 12 April 2010
68
x 2.6e+002
5.4 umSlide69
Be (flatter on C-C) - Pressure + Thermal load – Von-Mises
Michael MONTEIL - 12 April 2010
69Slide70
Be (flatter on C-C) - Pressure + Thermal load – Safety factor Ult. Strength
Michael MONTEIL - 12 April 2010
70
x 2.6e+002Slide71
To do :Rough mechanical design
Solution #1 C-C with differential pumpingMaybe coating
15 cm length between upstream and downstream sidesSolution #5 C-C + BeOrder quotes of Be
Same design that window in TI8, TI2, TT41 (Design by Kurt Weiss, Luca Bruno and Jose Miguel Jimenez) but replacing the Ti foil by a Be foilNickel-coating to prevent oxidation on Be ?15 cm length between upstream and downstream sides
71
Michael MONTEIL - 12 April 2010