Hoffman D.M., Singh B., Thomas J.H. (Eds). Handbook of Vacuum Science and Technology

Подождите немного. Документ загружается.

542 Chapter 4.8: Aluminum-Based

Vacuum

Systems

4.8.7.1 Avoiding Porosity and Particulate Defects

Hydrogen is the main cause of porosity in welds. Much more of it can dissolve in

molten metal than can stay dissolved when the metal cools and solidifies. As hy-

drogen comes out of solution, it forms voids that interconnect to form leak paths.

The molten/solid hydrogen solubility ratio for aluminum is 36 times higher than

for

iron.

This makes aluminum welds much more sensitive to this source of poros-

ity than those of stainless steel [93]. The principal sources of hydrogen are mois-

ture,

hydrocarbons, and hydroxides on the surface and in the surrounding atmo-

sphere. Fortunately, relatively simple and practical measures can reduce these

enough to produce good UHV welds.

Gloves should be worn at all times, because skin oils cause considerable con-

tamination. Oils and finger prints can be removed from the work piece and filler

rod with ethanol or acetone.

Oxides entrap moisture (a prime source of hydrogen) and have high melting

temperatures that can cause incomplete fusion. They grow rapidly on bare alumi-

num and must be removed within four hours of

welding.

The surfaces within

of a joint should be scraped with a hard-edged tool [94]. Wire brushes may drive

contamination into the soft surface and should be avoided. Since filler represents

most of the metal in a joint, it is a prime source of potential contamination and

worthy of special care in sourcing, packaging, storing, and handling—especially

in wire form. Filler rod can be cleaned off with several passes of Scotch-Brite®

immediately before welding.

Guard gas should be premium ultra-high-purity type (99.999%) and flushed

through the tip for 20 seconds before initiating an arc to assure a moisture content

below 10 ppm (preferably <3 ppm) [95]. Relative humidity in the welding area

should be less than 75% and preferably less than 50%, which may necessitate

working in an air-conditioned room. Fast weld speed has been shown to reduce

porosity [96]. Efforts should be made to assure that there are no leaks in the inert

gas lines, because these can be an inadvertent source of moisture build-up.

AC-TIG welding with argon gas should be used whenever possible, since

during its positive half-cycles, shielding gas is ionized and accelerated toward

the work piece, providing a sputter-etch cleaning action that removes residual

contamination.

4.8.7.2 Avoiding Crack Defects

Cracks usually occur when there is a wide temperature range over which an alloy

solidifies (freezing range). While the joint cools in this range, the bead is mushy

and easily fractured by small thermal contractions. This process is partially miti-

gated by the fact that alloys contain some eutectics that continue to flow at the

4.8.7 Welding Aluminum for

Vacuum

Applications

543

low-temperature end of the freezing range and tend to heal cracks. These then so-

lidify without cracks over a narrow temperature range [97].

Unfortunately, alloy compositions that produce good structural properties of-

ten also have wide freezing ranges. Weld crack sensitivity when joining these can

be greatly reduced by selecting filler alloys that are less susceptible to cracking

(i.e.,

narrower freezing range and/or high eutectic content), even when partially

diluted by the base metal alloy. Most cracks occur near the center of

bead, which

cools last. Fortunately, this area is least diluted by the base metal. Figure 13 shows

the effect of various alloy compositions on crack sensitivity. It should be noted

that for AlMg2Si (6000 series) alloys, crack resistance is increased by increasing

either Mg or Si content. Fillers that solidify at much higher temperatures than the

Fig.

13.

Relationship of aluminum craclc sensitivity to the presence

of various alloying elements.

Al-Si (4000 series)

AI-MggSi (6000 series)

» 2 3 4 5 6 7 8

Composition of weld, percent alloying element

[Courtesy of

H. Dudas and F. R. Collins, "Preventing Weld Cracks in High-Strength Alu-

minum Alloys,"

Welding Journal

(June 1966, p. 4).]

u-l

i-i

i§s

o p

< p <

O O Q

s 8

13^

s-.^

11 i

Is I

l!"

g-^ ;

ill

II i

s §

jlltlliiiilllilllt

«I s

I 1111 5

ill

lis

I I I I

o o o

111

S 5 S

III

o> r- o

p p

o p

4.8.7 Welding Aluminum for Vacuum Applications 545

base metal should be avoided, because thermal contraction forces will cause

cracking at the molten edge of the bead [98]. Recommended fillers for various

aluminum alloys are given in Table 7. Combinations, for which no filler is speci-

fied, should be avoided.

Several other techniques help minimize cracks. Fillet and Vee-groove joints are

preferred, because they maximize crack-insensitive filler in the joint and mini-

mize its dilution by crack-sensitive base metal. The weld bead should be slightly

convex to promote flow of molten metal in healing incipient cracks. If a bead is

allowed to become flat or concave, surface tension can inhibit flow. This can be

accomplished by tilting the electrode 10 to 20 degrees and using the force of the

arc to "dam up" the molten bead [99]. A crater must be avoided when terminating

a seam using any of a number of established techniques. In especially problem-

atic seams, trenches can be cut on either side of the joint (Figure 14). These trap

Fig.

14.

Weld joint nomenclature.

Square Butt Joint

Trench to trap heat

and reduce stiffness

^H K Root Gap

Vee-6roove Joint

Fillet Weld

546

Chapter

4.8:

Aluminum-Based Vacuum Systems

heat around the bead and minimize thermal expansion of the rest of the structure.

They also relieve stress by reducing stiffness and permitting some mechanical

deflection.

4.8.7.3 Joint Design

Except for a few special considerations, joint design for aluminum is basically the

same as for stainless steel. See Figure 14 for nomenclature. Simple, square butt

joints (touching) are suitable for thicknesses up to about 6 mm (0.25 in) and are

especially desirable because they are easy to scrape clean. Thicker joints should

employ Vee-grooves (60-90 degrees) or fillets to assure arc penetration and to

minimize dilution of filler by the base metal. A 3 mm

{-^

in) root opening, 1.5 to

2 nmi from the bottom, helps ensure full penetration. It is desirable for both sides

of a joint to be machined to the same thickness for balanced heating that mini-

mizes stress and distortion. Examples of joints between a thin flange neck and a

thick chamber wall are shown in Figure

15.

Tack welds can be made after fit-up to

hold a joint in place. These are essential to prevent distortions that build up dur-

ing welding from closing root gaps (when used) before a seam is completed. Tack

size may be according to operator preference, since during the final weld they must

be fully remelted and incorporated into the seam

[100].

Seams may be crossed

since this does not affect metallurgical soundness in aluminum as it does in other

metals

[101].

There is a "heat-affected zone" of about 25 mm on either side of

joint where

the temper is affected and strength permanently reduced. This is important when

welding highly tempered flanges but can be mitigated by the use of a heat sink on

Rg.l5.

Chamber

Recommended joint design for aluminum flanges.

Trench Chamber

-TIG Weld

uihM

Flange

(Best)

(Good)

4.8.7 Welding Aluminum for

Vacuum

Applications 547

the flange face (such as a mating flange) to keep the temperature down. If a weld

seam is required on both sides of

joint for structural reasons, one side should be

only skip-welded to prevent trapped gas from causing virtual leaks. After weld-

ing, shrinkage will be about twice that of stainless steel (about 6% of the width of

a seam). This can cause distortion and should be accounted for by welding pieces

together in a symmetrical sequence and machining after welding if strict dimen-

sional tolerances are required.

4.8.7.4 Welding Techniques

Excellent vacuum-tight seams can be achieved with manual TIG welding. An ex-

perienced stainless steel welder can acquire the necessary skills in a matter of

hours through demonstration and practice. The most important difference is that

temperature cannot be gauged by color. Aluminum simply becomes shiny when it

melts,

and only this quality can be used to determine the extent and temperature

of the bead. Welding style is also affected by the fact that molten aluminum tends

to hang together like jelly instead of flowing like a liquid, as does stainless steel

[102].

Because of its high thermal conductivity, aluminum requires about three times

more heat to weld than does stainless steel. AC-TIG (alternating current tungsten

inert gas) welding with argon gas is recommended for thin joints because it pro-

vides a cleaning action and is relatively insensitive to variations in arc length. A

ball tip should be used, since the high arc power tends to melt pointed tips, caus-

ing contaminating particles of tungsten to be thrown off. However, for thicknesses

of 1 cm and more, the high power required makes AC-TIG welding difficult to

control. Instead, DC-TIG with helium should be considered, since it produces

much more concentrated heating for better control and smaller beads for less dis-

tortion. The presence of helium increases the heat concentration and appears to

have a cleaning action that partially makes up for the absence of the plasma

cleaning provided by AC-TIG. Preheating is not recommended, because it can de-

stroy the tempered properties of the metal. Trial welds on test pieces are highly

recommended.

A dip and transfer motion should be used with AC-TIG welding. After a spot is

melted, filler rod is added which, in turn, cools the spot. The tip is held in place

while the rod is withdrawn to allow the spot to thoroughly reheat. The tip is then

moved forward enough to get good penetration but not so much that the arc burns

through the work piece. In DC-TIG welding, filler is supplied continuously, be-

cause the intense, concentrated heat is sufficient to maintain bead temperature.

In both cases, the tip should be normal to the seam or slightly leading the bead.

Filler rod should be 25 to 50% of the width of the bead, depending on the opera-

tor's personal preference. Thicker rod provides a little extra stiffness for pushing

548 Chapter 4.8: Aluminum-Based

Vacuum

Systems

filler into square joints. Thin rod requires a high feed rate that can be difficult to

control. In general, the amount of heat used in welding should be minimized since

excessive heat causes loss of material strength, weld defects, and porosity.

4.8.7.5 Quality Assurance

Compliance with MIL-STD-2219 Class A weld requirements has been found to

produce welds that are suitable for UHV applications

[103].

These are similar to

several other commercial specifications [104,105,106]. The main requirement is

for subsurface pores and inclusions (as seen in X-ray photographs) relative to

weld thickness (T) to be as follows

[107]:

Individual size must be less than

0.33T

0.060

in, whichever is less

Spacing must be at least four times the size of the larger adjacent imperfection

Accumulated length in any 3 in of weld must be less than 1.33T or 0.24 in,

whichever is less.

4.8.7.6 Repairing Defects

Defects that cause leaks can nearly always be repaired by removal with a point

grinder, such as a Dremel® tool, and rewelding. Material must be removed 25 nun

(1 in) on either side of the defect but need not fully penetrate the seam. The tool

used should have a metal cutting tip rather than an abrasive tip, because the latter

tends to force contamination into the surface. The cleaning action of AC-TIG

welding is particularly useful for removing entrenched contamination. It can also

be used to "boil out" hydrogen from porous areas. Care should be taken not to

overheat the repair, because this can cause cracking further along the seam that

sometimes necessitates a series of repairs that "chase" a leak down the seam

[108].

REFERENCES

H. F. Dylla, D. M. Manos, and P. H. La Marche, Correlation of outgassing of stainless steel and

aluminum with various surface treatments, J. Vac. Sci.

TechnoL,

All(5) (September-October

1993) 2629.

John F. O'Hanlon, A User's Guide to

Vacuum

Technology,

2nd ed. (Wiley, New York, 1989), Ap-

pendices C.2 and C.3, pp. 445-446.

M. Suemitsu, H. Shimoyamada, N. Miyamoto, T. Tokai, Y. Moriya, H. Ikeda, and H. Yokoyama,

Ultrahigh-vacuum compatible mirror-polished aluminum-alloy surface: Observation of surface-

roughness-correlated outgassing rates,

Vac.

Sci.

TechnoL,

A10(3) (May-June 1992) 570.

References 549

M. Mohri, S. Maeda, H. Odagiri, M. Hashiba, T. Yamashina, and H. Ishimaru, Surface study of

Type 6063 aluminum alloys for vacuum chamber materials, Vacuum, V34(6) (1984) 643.

F. J. Norton, Gas permiation through the vacuum envelope. Trans. 8th AVS Vac. Symposium

(Pergamon Press, New York, 1962), p. 14.

6. H. Ishimaru, Developments and Applications for All-Aluminum Alloys Vacuum Systems, MRS

Bulletin, 15(7) (July 1990) 24.

H. J. Halama and J. C. Herrera, Thermal desorption of gases from aluminum alloy Al 6061, their

rates and activation energies, J.

Vac.

Sci. Technol, 13(1) (January-February 1976) 466.

8. June-Rong Chen, Chih-Hao Lee, Jyh-Cheng Chen, Huan-Lurn Hsieh, and Yuen-Chung Liu,

Outgassing behavior of A6063-EX aluminum alloy and SUS 304 stainless steel, J. Vac. Sci.

Technol., A5(6) (November-December 1987) 3422.

9. June-Rong Chen and Yuen-Chung Liu, A comparison of outgassing rate of 304 stainless steel

and A6063-EX aluminum alloy vacuum chamber after filling with water, J. Vac. Sci. Technol.,

A5(2) (March-April 1987) 263.

10.

G. Moraw, The influence of ionization gauges on gas flow measurements. Vacuum, 24 (1974)

128.

11.

J. R. Young, Outgassing characteristics of stainless steel and aluminum with different surface

treatments, J.

Vac.

Sci. and Technol., 6(3) (1969) 400.

12.

H. Iguchi, T. Momose, and H. Ishimaru, C12 gas corrosion and outgassing of various surface

treated aluminum alloys, J.

Vac.

Sci. and Technol., All(4) (July-August 1993) 1708.

13.

Chen et al. 1987, Iguchi 1993.

14.

Yuko Hirohata, Satoshi Fujimoto, Tomoaki Hino, and Toshiro Yamashina, Hydrogen desorption

behavior of aluminum materials used for extremely high vacuum chamber, /

Vac.

Sci. Technol.,

All(5) (September-October 1993) 2637.

15.

Ishimaru 1990.

16.

Francis J. Norton, Gas permiation through the vacuum envelope. Trans. 8th AVS Vac. Sympo-

sium (Pergamon Press, New York, 1962) 13.

17.

Ishimaru 1990, p. 27.

18.

R. N. Peacock, Practical selection of elastomer materials for vacuum seals, J.

Vac.

Sci. Technol.,

17(1) (January-February 1980) 330.

19.

A. Roth, Vacuum Technology, 2nd ed. (Elsevier Science, New York, 1982) 398, Equation 7.12.

20.

Aluminum Association, Aluminum Standards and Data, 1990, Aluminum Association, 900 19th

St., N.W., Washington, DC 20006 (Tel: 301/645-0756), p. 30, Table 2.1.

21.

J. Holden, L. Holland, and L. Laurenson, Bakeable vacuum seals using aluminum wire gaskets.

Journal

Scientific

Instruments,

36 (June 1959) 282.

22.

J. D. Redman, et al.

Vacuum

systems techniques and material studies. Oak Ridge National Lab-

oratory Report 3472. As referenced in Alex Roth,

Vacuum

Sealing Techniques (1966), Reprinted

by American Institute of Physics, 1993, Ref. 1036a, p. 309.

23.

Cefilac Etancheite, Seals for

Ultra-Vacuum

and Cryogenics, catalog p. 5.

24.

I. Sakai, H. Ishimaru, and G. Horikoshi, Sealing concept of elastic metal gasket "Helicoflex,"

Vacuum,

32{1){19S2) 34.

25.

Aluminum

Standards and Data, p. 30, Table 2.1.

26.

J. D. Gonczy, R. C. Ferry, R. C. Niemann, and B. Roop, UHV seal studies for the advanced pho-

ton source storage ring vacuum system. Conference Record of the 1991 IEEE Particle Accelera-

tor Conference, May 6-9, 1991, San Francisco, California, p. 2304.

27.

W. Unterlerchner, Non-standard vacuum hardware for an accelerator vacuum system, CERN/

LEP-VA/89-50.

28.

H. Ishimaru, Bakeable aluminum vacuum chamber and bellows with an aluminum flange and

metal seal for ultrahigh vacuum, J

Vac.

Sci. Technol., 15(6) (November-December 1978).

29.

T. Ishigaki, K. Shibuya, H. Sakaue, and S. H. Be, Sealing technique for aluminum to stainless

steel CF flange pairs, J.

Vac.

Sci. and Technol, A12(4) (July-August 1994) 1705.

550

Chapter

4.8:

Aluminum-Based Vacuum Systems

30.

Greg Andronaco, Lawrence Berkeley Laboratory, private communication, 1994.

3L Ishimaru 1990, p. 23.

32.

M. Suemitsu, H. Shimoyamada, N. Matsuzaki, and N. Miyamoto, Photoemission studies and

outgassing-rate measurements on aluminum alloy surfaces lathed with various alcohols,

Vac.

ScL

TechnoL,

AlO(l) (January-February 1992) 188.

33.

A. G. Mathewson, J. P. Bacher, K. Booth, R. S. Calder, G. Dominichini, A. Grillot, N. Hilleret,

Latorre, F. Le Normand, and W. Untcrlerchner, Comparison of chemical cleaning methods

of aluminum alloy vacuum chambers for electron storage rings, J. Vac. Sci.

TechnoL,

A7(l)

(January-February 1989) 77.

34.

N. Kaufherr, A. Krauss, D. M. Gruen, and R. Nielsen, Chemical cleaning of aluminum alloy sur-

faces for use as vacuum material in synchrotron light sources, /

Vac.

Sci. and

TechnoL,

A8(3)

(May-June 1990) 2849.

35.

R. A. Rosenberg, M. W. McDowell, and

R. Noonan, X-ray photoelectron spectroscopy analy-

sis of aluminum and copper cleaning procedures for the Advanced Photon Source, J.

Vac.

Sci.

and

TechnoL,

A12(4) (July-August 1994) 1755.

36.

John Philip Durrant and Beryl Durran, Introduction to Advanced Inorganic Chemistry (Long-

mans,

Green, London, 1962).

37.

Dylla 1993, p. 2636 (Appendix).

38.

Y. Tito Sasaki, A survey of vacuum material cleaning procedures: A subcommittee report of

the American Vacuum Society Recommended Practices Committee,

J. Vac.

Sci

TechnoL,

A9(3)

(May-June 1991) 2031, describing SLAC cleaning procedure.

39.

Dylla 1993, p. 2636 (Appendix).

40.

Tito 1991, p. 2031.

41.

Mathewson 1989; Kaufherr 1990; and Rosenberg 1994.

42.

Tito 1991, p. 2031.

43.

Mathewson 1989; Kaufherr 1989; Rosenberg 1994.

44.

Tito 1991, p. 2031.

45.

AUiminum

Standards and Data.

46.

Metals Handbook, 8th ed., Vol. 3: Machining, American Society of Metals, Metals Park, OH,

439.

47.

George A. Goeppner, APS storage ring vacuum chamber fabrication, AIP Conference Proceed-

ings 236 of the American Vacuum Society Series 12:

Vacuum

Design of

Synchrotron

Light

Sources, p. 127 supported by Reference 3, which was unpublished.

48.

E. A. Avallone and T. Baumeister, Mark's Standard Handbook for Mechanical Engineers,

9th ed. (McGraw-Hill, New York, 1978) pp. 5-52.

49.

Lloyd E. Brownell and Edwin H. Young, Process Equipment Design: Vessel Design (Wiley,

New York, 1959), p. 143.

50.

Ibid., Figure 8.3, p. 144.

51.

Mark's Standard

Handbook,

pp. 5-52.

52.

Joseph E. Shigley and Charles R. Mischke, Mechanical

Engineering

Design, 5th ed. (McGraw-

Hill, New York, 1989), p. 59.

53.

Mark's

Standard

Handbook,

pp. 6-11 and 6-44.

54.

Aluminum

Standards and Data, p. 40.

55.

A. Roth,

Vacuum

Sealing

Techniques

(AIP Press, Amsterdam, 1966), p. 752.

56.

Donald R. Pitts and Leighton E. Sissom, Heat

Transfer

(McGraw-Hill, New York, 1977), p. 17.

57.

Mark's

Standard

Handbook,

pp. 4-68.

58.

Bruce Craig,

Handbook

of Corrosion Data (ASM International, Metals Park, OH, 1989), p. 146.

59.

Ibid., pp. 162-163.

60.

Ibid., p. 147.

61.

Ibid., p. 227.

62.

DECHEMA-Werkstoff-Tabelle/Chemische

Bestdndigkeit (DECHEMA Deutsche Gesellschaft

References

5S1

fur Chemisches Apparatewesen, Frankfurt

Main, Germany, December 1982),

pp.

Chlor B1.4.

63.

Ibid.

64.

Handbook

of Corrosion Data, p. 227.

65.

Ibid., p. 229.

66.

Philip A. Schweitzer, Corrosion Resistance Tables, 2nd ed. (Marcel Dekker, New York, 1986),

466.

67.

D. J. De Renzo, Corrosion Resistant Materials Handbook, 4th ed. (Noyes Data Corporation,

Park Ridge, New Jersey, 1985), p. 618.

68.

Handbook

of Corrosion Data, p. 273.

69.

DECHEMA, p. Fluor B1.3.

70.

DECHEMA, p. Fluor B1.3.

71.

Corrosion Resistance

Tables,

p. 466.

72.

Handbook

of Corrosion Data, p. 273.

73.

Corrosion Resistance

Tables,

p. 466.

74.

Corrosion

Resistant

Materials

Handbook,

p. 620.

75.

Corrosion Resistance

Tables,

p. 568.

76.

Ibid., p. 572.

77.

Handbook

of Corrosion Data, p. 326.

78.

Corrosion Resistance

Tables,

p. 572.

79.

Handbook

of Corrosion Data, p. 326.

80.

Stanslaw Mrowec and Teodor Werber, Gas Corrosion of

Metals,

Translated from Polish and

Published for the National Bureau of Standards and the National Science Foundation, Washing-

ton,

D.C., by the Foreign Scientific Publications Department of the National Center for Sci-

entific Technical and Economic Information, Warsaw, Poland, 1978, Chapter 4.

81.

Metals

Handbook,

p. 583.

82.

S. M. Fine, Design and operation of UHP low vapor pressure and reactive gas delivery systems.

Semiconductor

International,

October 1995, p. 138.

83.

Metals Handbook, 9th ed.. Vol. 13: Corrosion (ASM International, Metals Park, OH, 1987),

65.

84.

Robert H. Perry and Cecil H. Chilton, Chemical Engineers'

Handbook,

5th ed. (McGraw-Hill,

New York, 1973), Section 3.

85.

Ibid.

86.

Monika Kuhn and P. Bachmann, Demands for turbo-molecular pumps in the aluminum etching

process,

Vacmmi,

41(7-9) (1990) 2028.

87.

DECHEMA, Fluor

1.2.

88.

Kuhn, p. 2031.

89.

Rosemarie Becker and Monika Kuhn, Heated turbomolecular pumps in metal etch applications,

Solid

State

Technology,

November 1994, p. SI

90.

Aluminum Association,

Welding Aluminum

Theory and Practice, 2nd ed. Aluminum Associa-

tion,

900 19th St., N.W., Washington, DC 20006 (1991) (Tel: 301/645-0756).

91.

Aluminum Association,

Welding

Aluminum,

Aluminum Association, 900 19th St., N.W., Wash-

ington, DC 20006 (1984) (Tel: 301/645-0756).

92.

Let's

Talk

About

Aluminum Welding

(videotape). Published by ALCOA, available through dis-

tributors of AlcoTec, 2750 Aero Park Drive, Traverse City, MI 49684 (Tel: 616/941-4111).

93.

Weiping Liu and L. Dorn, Improved filler wires for aluminum alloy welding —

review. Part I,

Aluminum,

70(9-10) (1994), 583.

94.

Joseph Gagliano, Argonne National Laboratory, private communication regarding advanced

photon source welding procedures.

95.

Procedure used for advanced photon source storage ring welding at Argonne National Lab-

oratory.

96.

M. Kutsuna,

Suzuki, S. Kimura, S. Sugiyama, M. Yuhki, and H. Yamaoka, C02 laser welding