uno

Investigation of the broken beam pipe

Shoji Uno(KEK)

KEKB Review, Feb/10, 2003



• Structure of IP beam pipe

• Searching for leak points

• Element analysis for remaining white

powder

• Possible reasons for He leak

• Summary

Structure of the IP beam pipe in SVD1.4

Inner 10~30µm gold by 10µm gold by 200~230µm gold by To reduce

chemical plating vacuum sputtering chemical plating

Surface SR









To reduce particle background







Beryllium part is cooled

He by Helium gas.









Aluminum part is cooled by water.

Possible sources of He leak to vacuum







He



•crack in Be or Al









•brazing joint between Be-Al •welding joint between Al-Al



0.5mmt





0.5mmt

Search for leak points

• After dismounting the beam pipe from Belle, a leak

check was performed to identify the leak point.

– He leak was confirmed in a normal leak test.

– Bubbles could be seen on the inner gold sputtering surface

of Beryllium, when liquid was supplied.

Not joint parts.

Leak point

~2cm from edge of Beryllium





Cutting point









Cutting point

Picture of the inner cylinder

• A large amount of white powder was found on the surface

of the inner Beryllium cylinder.

• It looks like a pattern of water flow.

• There is white powder on the Al ring, also.

– Large amount at backward ring

– Small amount at forward ring



Direction of helium gas



B



F

Leak points

• Leak position was confirmed from outside

of inner beryllium beam pipe with various

methods.

– There are at least two

separated leak points

on Beryllium surface

( ~1cm distance).

- It matches to position

where was found from

inside.



Leak points

Close up view of a leak point

• Any cracks could not be found near leak points.

– Many tiny dips are seen.









Leak point

Element analysis

• We found two types of powder.

– Color of one powder is clearly white.

– Another one is slightly yellow.





• Element analysis is in progress, separately.

– Several methods are used for identify elements.

– Preliminary results are available, now.

Preliminary results of element analysis

• White

– Main components are Be and O.

– Probably, it is BeO.

• Yellow

– Main components are Al and O.

– Probably, it is Al2O3.

• Common

– C was also found.

– Small amounts of P, K, Ca, S, Cl, Si, Mn, Fe, and Cu were found.

• Si, Mn, Fe and Cu are components of Aluminum alloy.

– S and Cl are dangerous elements for corrosion of Beryllium.

– But, expert for element analysis says amounts of S and Cl are small

and are consistent with normal metal. No conclusion, now.

Check for inside of Al part

• We cut L angle connector in Helium line to

see inside of Aluminum part.

– One inlet connector was checked.

– Inside surface shows white color without a

metric luster.

• Nothing happens on the surface of SUS

connectors.

Measurement of thickness

• Thickness of the inner Beryllium cylinder

was measured.

– We hope thickness may tell us something.

– No significant difference was found.

• Center : tend to be thick.

• Both edges : tend to be thin.

• No clear phi dependence.

• After cutting of Beryllium cylinder, more

precise measurement will be done.

Photo before assembling

• Leaked Beryllium beam pipe is recycled from the

first beam pipe, which had been used in May-July

of 1999.

• Before assembling, a photo was taken.

– White powder can be seen clearly on the Beryllium

surface.

– Amount of powder is much

smaller.

– Pattern is similar.

• At that time, we did not

investigate the white

powder. After that,

we forgot this fact.

Corrosion

• Corrosion might occur on both of Al and Be.

– At present, we do not know why corrosion occur.

– Which components affect corrosion? Water, Cl or S? More careful

investigation is necessary.

• Analysis of circulation gas is in progress.

• Before the accident, we had not paid attention to corrosion.

– Due point had not been monitored in circulated Helium gas.

– We have never analyzed inpurity of the circulation gas.





• Present methods to avoid corrosion

– Due point is monitored(~-20•Ž ).

– Additional filter will be added.

– More frequently, fresh Helium gas is supplied to avoid

accumulation of unlike gas. Most effective.

Possible reasons of Helium leak



•Corrosion most suspicious reason

•Heat stress

•Two accidents

•Recycled Be pipe from BP#1

Large stress at machining process (?)

•Very high temperature (~300 °C)

by gold-spattering and welding with Aluminum section

•Defect of material (?)

•Dark spots which were found on Gold-plating

Temperature distribution with maximum allowed heat load

(Red:inner Be Blue:outer Be Black:He)



100 W with He cooling







dTin=+23.0 Kº

dTout=+4.1 Kº

(He flow : 2g/s)









•If He flow is 1 g/s,

dTin=+25 Kº, dTout=+5 Kº

•Present situation: 1g/s,dTout=+3 Kº

⇒ 60W?

Maximum allowed temperature from stress analysis by IHI



•Sm ≡ Min (1/4×Tensile strength, 1/1.5×Proof stress)

Proof stress = stress at 0.2% yield point

•Stress should be smaller than 3×Sm for thermal case

•No worry(?) for heat cycle if stress 8733?j77kj777jj00jk08??ks?778ttu>310?jjj666>6333ii0it0iiii30iiiiip0>>s73700'3hh>hthh000?1>8>>>87770iii?311ii3i0#0i373t333j1t8700388ti0000u377u3>?f0fiffjfjfjf0011ji00j0j7thf0jjjj8iu7j3>tjj878jtj003i0ti>>7j3jj00ii0j0k0hjjijjjh7hh3jjjjjjh'j0000jij0hiiigikigggg004iiiigggggi8igggs8gj>g?uu0800ttggg730igff7817gsgt73>??8888????ue83ekje7jeji>>>>?1u733j0u73000011j33jj8jk88j88?>>>80iiii?ss?u7u8?8i7s7>s11337ue0ie0ie37e1ue13e1ge1j0;0000kjk10e?>>uu8e1311371330ukse1?iiiii8t888888i8s>>>uuuuuuu333030000j0j017>66668888>0673f??7t>>>i773?777ji07j117161jjj;jii11111k8tiiiiit113708?8jj377jj8st1i7tkkkkkggggg3kiiigggiiiiiiigjjjjjggijitggjgj8iiiii73;333u30>081jjjsh>8u888j>>1h37uuhhh1jh366666j8jjsjjjp0iiiiiu?t?7sgiiiiiit#;ggg877777001180?0t70>>7377jj3j20kj1uu7u0j>u08didi0idiihgdiikdi8d7k8>0i?888ij0jjkj>17j00ju7j1j1iiiiiij17jjgggggggggjkj8>jgt0jtgg8>gug3?3179300000008?>0>7>673333338088808188jk3j3>;jjk888iiiiiij7787??>>8>>fjfjfkfjfjfjfjfjfjfjfj07?uuu00800000000>j80uuuuiikii08033u0iiijj7??t???????j3iiiiuh;0iijk3j7jjjjjjjijjjiuij44jkj833iiiiiiiii6888iii>>>>iiiiiu?fiiiikffffgfgf7f1?777g71111gg0000?gggggggggjsggg>77711333gg4000iiiitii833373300888>8kjjjjj000078/8>>7777ii'11187880888808#;>kk800>u>>7777703t888333777jjji03313>iikkkijiiiii7ikfghjijiii1hhiihjkjjjjjjj0j>iiii0!07780jjj7jgt7ggtei>eieiei>>u7777.1'114t888811q>>t>>iii7777ii!111;0k0k0sii9i007j?ii78888888837>j01j3717t>>kkkk788jj7338u8888881j11!7tt>>h9j800tti60k7kkk7jjk3k7e7ke3he7he7tjjjj08jj837773jhhhhjikjoi10ggttggu8>>>00?>8?0hhhhhhhhhhhhhhh77633ti0i0#080000kkk>?1s0tu111t182177itt8jiiiijj6ikigggggggggggggggijjjj103j1j331388j8hhjjjgjgggttgug3333g3gggggiiijjiijiiiikikhhjhj66611d e

e

e

e

e

e

e

e

e

e

e

e

e

e

e

T0 system stops.

– 30 seconds after the

system.









e f e 7k e

ef e0 e

e d jj e

ef e e

ef e e

ef e e

eg8jh0000000000000000000000033333333333333333333333330gi0ji0i00000000000jjiiiiiiffe0iffe0fhffe3jffgiffgiffgijjgggggjjjjgggggggggggggijjggggggggggggihgiiiiifge0fg0ffhffhf3e0g3eig3hhhhfgfd6jhg0d4hgjd6hg6jd1hg6hdkhg1e0iekekeje3jhdjhihjdj0jhjdj0jhgejhie88806464443332jjjjjjjjjj00gikkkkjkgggggiiiiiiiiiiiiiiiiiihhhhhhhff6hhhhhhhhhhhd7kgjhf6dg6eg6jhgjdg6jhgjdg6jhe80jdg1jh8hdg0jhhdi183hd6ii2hdi78jhdk8jgggg2311311111113313jjjjjjjijjjjjjjjjiiiiiiiiiiiiihhhhhhhhhhhhhhh4gggggiggggggggggggg4i 8800000jhei1ekekeke6ke6i1e6je6jejhgei3ieii6ejhgejhgejhgeifhhhh8f0f6fsis?t11hhfe

e46jjkkiiiihhhhf111111jjjjjjiiiiiihhhhhhgggggge11111111hhhhhhhhfe

e8ue8;ees;te68t7776ggkkkggggggefke'et7e31ejueiuejeji8ejgf3ejh8eiiigggf e fe

eg"hje8>ett7e21her6ejte3jjjiiigggfhhh886k?6=7j0kggge61jkhjih71=tk>6hj7tk?jj7tiik?jj7thhhhh1ef?fs11jjjjiiiiii808888ss7777kk0jj01330000jjjjjjhhhhhhgggggge

e30feteiregtej0ehseg3ehse0se89etje6ue33e7hteijeis ikj8j0jgggf040jk8j37iii36iikjggg630>tigf01hfe

eg7eg3e0ke0je0je1htiijhhhf000111hhhhfe

e>ti00000000008044666631333133311310j00j0ji00j0jjijkkkjjijijjjiijjjijiiiijjjijjiiiiihihihiihihiihhhhhhh000/jstt9t0'.7ji747738r13f7ff&fi0j0f>f8jjfjjf0jkjkjfgff2ffu3>0jjfkfffhjji0ieet0hff8if3f'ffiiigfhfhffiii7if8f8if?iiifff1ffuffff7jhh>tf7f7f6f1h>htfjfk3itfjfgfkfgti1gggtg8ggggjgggggjtggh?i16teke8jfggeh>ei8euhe70fhetjekessejf>ei,e8tet7e78s81&>1e667jjkkjiiiii88>uu13i7jjjk00tht9331iii>#t000710000hhhhhfe

e f f fe

e jf fe

e f i fe

e f ki fe

e01ji6hf&0?11ji1jif&7>71kt8>!u633t0j>jjjstiifhis1hhhr86?>.9t8=>u77070kjjkhjihiihh9ffji6k46ji80jgff>f3jihg>t0hfe

eji8387jih?t1jt1ji18u03ji6igff>0jih#03j0t0h76jijigfff>1(13ji6kjige0?fshg00hfe

e860iii40t#t1jjjiiihhh00s'70>tt00000000000000300003000000000000000j00000jiiiiihhhhhhhhhhhhhhhii888808uuuufffffffjfjfjfjfjfjf3fhfh????7jj1110ks000008hhhhhggggggge

e ee

e ee

e ee

e 3!30jiiihhhf61iii0630jiiihhhet000000000hhhhhfe

ffe f e fe

ffe e fe

ffe f e fe

ffe e fe

ffe 3ihf6kj0jhge00hfe

ffe 3jhff0ijhge00hfe

ffe f e fe

ffffe 080s?utttj300iiijhhhhf1337hjjjhhhgggge?t0000000hhhhfe

ffe fe

ffe 6jhffijhge00hfe

ffe e fe

ffe jhff0fi6jhge00hfe

ffe e e

ffffe 6jkjhihffi66jjhhgge0000hhfe

ffffe 000sss000jjjhhhf0011jjjhhhggge000000hhhfe

fffe e fe

ffffe f e fe

ffffe ttt377ikkjjhhhff00600kjjjhhhggge000000hhhfe

ffffe f e fe

ffffe e fe

fffffe 8jhff8j6kjhge00hfe

fffffe 00jjhhff0jjiihhgge64kkhhfe

fffffe f e fe

fffffe 0000000jiiihhhhffj6iiiihhhhgggge4444kkkkhhhhfe

e e

e e

fffffe e fe

fffffe f e

fffe e fe

fffffe f e fe

e e

ffffe e

ffffe f e e

fffffe f e fe

ffffe e

e e

fffffe f e fe

fffffe 6j0033jijjjjjhhhhhhhf033j01113301iiiiihhhhhhhggggggge0tt4664477kkkkkkkhhhhhhhfe

ffffe 6jjhf01ihge4khfe

ffffe 6jjhf00ihgekjjiijjjjjjjjjjj008888888726!!,#=-11!?6j22232j;ttgk%%kk%%t-jj01j11-10-10j!j7175ki6iiii1101kiikiiiiiiiiiiiiikhjjjjjhjhhhhhhhhhhhhhhhhhhhhhhhhhhhhh8t;t+u9sis888s800>>t?ttttttt00000000000000000800000000070033333333333333636777777777667676676666666676666666666666766666666666667666666664366766666666446jjjjjjjh000000hjj01jjjjjjjjjjjjjjj001gggjjjjjjjjjjjjjjjjhhhjjjjhhhhkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkjijkkkkkkkkkkkjjkjjhhhihjiiihhhhhhhiiijihhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhjiiiiihhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhiiiijiiiihiiiiiiiiiiihiiiiihiiiiiiiiii00088018?8>?>?8hh8ff=f0fjfjf7f7f7fkf4fhfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfjfiffiffhffgjf1f0f1f3f1f3f277?>?jf3307773177777j30k1jjjjjjj310000000>?80kk7>kkkkk88888888888888tt311jkkkhhjjhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhigggggggggggggggggggggggggggggggggggggggggggg80608300018kt8iiiigkhigiiiiisssjhhhiiiii>ttiggiiiiiigghhuiiiiiiiiutiiiiighhitgig3igghhhhhgggg07hg71ggggggggggg0hhhhhggggghhhhh031hgggggghhhhh6k6hhjjjjfi73f8fkf0f80jkf0fjf7jff6iiiiiiiii86iii080i7jjk610kffififififif6f86310f0fjfjfjfjf0ff7f1fkfhf0fkfh063kffhfhff18f0f00f8jjjjjfhfhfhiiiiii0731hhhhhhhhhhh0888gggggigjjigggggg7gitggggggggggggggggghkiiggggghsjggggh0hhht1sjgig>78s>0ijjjj88i8004373g730007662100jk0j0ki878886jhj633k464464g0jii000e

1e

se

je

ue

1e

3e

0e

te

7e

ie

e

ie

je

ie

0e

je

te

je

e

se

0e

7tji3jj>3jjrf0tf0f#31>880007777h7100011jijikik8jjikk$6jjjjjjjijp3;8j1j0tj888j8jjjj03ihhhhhhhhhhhhhhf3ffufkftftf;fjf3f7f#fsfqjgsgg9hii34irkhg$6=hk>$?t4?5jt8?tt0tj71tiu!>0fj0870heh?iiiiiijjjiii?>et8eshs6188h73,ggdpd

•dTout = 0 → +6 Kº with





fhj006606661i71iiikjjjjjjjjkjjjjkf0f8hhhhhhhhhhhhhhhhh88100jj31100011jjjjiiiiiiiiiiiiiiiiiggggggggggggggggg00000000000077771111111111111iiiikkjiiijjjiiijjjjji086ggjggj0jjjgggugghk3?0h?10jjjjj'88
270fif6300000ig00jj0jk6iii700011jtiikkh6611111jgggkk8687hh6j0kukktjjtj08jj70030010j80g6337686001kjjjktkjg7g08gg7gg13hfi11101kjjjijjjjjijiiiiiii861hhhj0377868jjj1tjjj0ju830073676660jejkj8kkkk887311133j>>e01e10ej0e6je0eje3e600je66e1jkjjjjjje17e1ue1e1fheh1ejj07j36663j310jk0680jjjjj0j0t0000063000ij08jfhf0fhfkfhfhfhfsf8fif7f0f6fjf3f1f7fhfkff6ff0ff8f3f1fkfjfjf1fjf0f0f0fjff6ff6fjffjfjftf?fk1kt0k07j8j3jj1h1j7jj k117j8k30jhkhed

1ed

1ed

?ed

3ed

7ed

8ed

8ed

7ed

ked

jed

ed

ted

8ed

8ed

red

8ed

ed

jed

0s087331008k6jjjjjjjjjjjfrued

0ejj8ekie71ge1tgekfhhd

e0j1'7?#7'kjeh1s0?ed

jt28hhh7t>7jjjed fd

fd

fd

fd

fd

fd

fd

fd

fd

fd

fd

fd

fd

fd

t00067iiiiiiikkjjhhhhhhhhh000000jjjjjjjjjiiiiiiiiiggggggggg0>7377=s3711ig17880888jj1iikrh!ik830i00ihjt08877777733j7;3hiiiiihiiikkiiihii800000088733311000hhhh0iji10877h8qjjjjjjjjjjiiiiiiijt6k1h80st0s111000jjjjiiiijjjjji#00886708810jki8>3677f8i67h777k1tjjj3iijiiiiiihhhhhhh>ggggifj3333tg80!93ggggihh0773ghkkggf09h133310gggggghhkk!ggghujjjjjjjjjji3qggg60iiiiiiiiii78iiiiiiiiiiiiihhhhhhhhh6hh08733110h008khhhhht17jjjjjjjjjjjfg087310008k177ej>e
fd

0010kiiiiiiiiie0d

fd 87d

f1id

f0id

g0d

gd

7d

5d

j6d

ftid

8d

ud

0kd

73d

kgd

gjhd

h0d

jhd

ihd

j3d

f3id

i0d

4d

gjid

htd

0td

87d

j0d

17d

ihd

j3d

40d

fid e

e

e

e

e

e

e

•dTout = +2 Kº → +5 Kº

•At 6:10 On Sep. 30, 2002 •At 8:45 On Nov. 3, 2000

Two temperature accidents

Dark spots on the Au-plating surface of forward Aluminum

section were found after the Helium leak accident.

Large dynamic pressure?

Summary

• Leak points are Beryllium itself, not joint parts.

• Corrosion is most suspicious reason of He leak.

– In order to avoid corrosion, several actions were

done.

• Further investigation is in progress.

– We will cut the Beryllium cylinder, soon.

– Then, another element analysis will be performed

and we can see detail of surface around leak points

with microscopes.

• New beam pipe with SVD 2.0 will be installed

in this summer. T.Abe’s talk.

History of the beryllium beam pipe at KEKB

We constructed three Beam Pipes: BP#1, BP#2 and BP#3.

Basic mechanical structure is same in three pipes.



•BP#1 had been used with SVD1.0 from May 1999 to July, 1999.

30 µm gold plating both for fwd/bwd Al sections.



•BP#2 had been used with SVD1.2 from Oct. 1999 to July, 2000.

30 µm gold plating both for fwd/bwd Al sections.

20 µm gold foil outside the Outer Be cylinder



•BP#3 had been used with SVD1.4 from Oct. 2000 to Nov. 2002.

200/30 µm gold plating for fwd/bwd Al sections.

10 µm gold sputtered inside the Inner Be cylinder

Hydrogen pressure in vacuum pipe increased on Oct. 31. 2002

Many black spots were found in fwd Al section on Nov. 3. 2002

He leak to vacuum happened on Nov. 8. 2002

Broken beam pipe (BP#3) in SVD1.4 was replaced

• with old one (BP#2) inside SVD1.2 in B4 clean room

• with new one (BP#4) for SVD2

BP#3 BP#2 BP#4

cooling for Be He He Paraffin

Au on Be 10µm inside 20µm outside 10µm inside

fwd/bwd Aluminum Aluminum Tantrum

Au in fwd 200µm inside 20µm inside no

Au in bwd 20µm inside 20µm inside no

Res. HOM 5 buckets 5 buckets No?

Saw tooth bwd no bwd

Material (IP) 0.6% X0 0.9% X0 0.7% X0

Particle mask standard tolerable? better

Installation of BP#2 & SVD1.2 Resonant HOM heating

Installation of BP#3 & SVD 1.4 Trouble of He cooling system



dTout=+3K





dTout=0K

Temperature rises at 1.5A

• Temperatures are plotted at LER+HER=1500mA

20.0 5.0 0.7 2.5 2.1 3.0 2.5 3.5 3.3 3.0 1.5 2.5 11.0









2.0 11.0 9.0 2.5 3.3 4.0 7.5 9.0









9.0 0.5 2.7 0.5 2.4 2.2 0.9 1.1 0.9 0.9 0.8

Direct comparison of the temperature rises (∆T) of

the Be chamber against the beam currant

• If we conservatively 5

allow the maximum

∆T experienced in

SVD1.4, the

SVD1.6

maximum current in 3

SVD1.6 will be 2.2 A. SVD1.4

• If we allow ∆T=5, 2 (29 Oct. 2002)

current limit will be

2.7 A. (need 1

confirmation by

further study) 0

0 1 2 3

LER+HER current (A)