1、Designation: F 140 98 (Reapproved 2003)Standard Practice forMaking Reference Glass-Metal Butt Seals and Testing forExpansion Characteristics by Polarimetric Methods1This standard is issued under the fixed designation F 140; the number immediately following the designation indicates the year oforigin
2、al adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon (e) indicates an editorial change since the last revision or reapproval.1. Scope1.1 This practice covers the preparation and testing ofreference gla
3、ss-metal butt seals of two general configurations:one applicable to determining stress in the glass and the otherto determining the degree of mismatch of thermal expansion(or contraction). Tests are in accordance with Test MethodF 218 (Section 1.1).1.2 This practice applies to all glass and metal (o
4、r alloy)combinations normally sealed together in the production ofelectronic components. It should not be attempted with glass-metal combinations having widely divergent thermal expan-sion (or contraction) properties.2. Referenced Documents2.1 ASTM Standards:F 47 Test Method for Crystallographic Per
5、fection of Siliconby Preferential Etch Techniques2F 79 Specification for Type 101 Sealing Glass3F 105 Specification for Type 58 Borosilicate Sealing Glass3F 218 Test Method for Analyzing Stress in Glass43. Summary of Practice3.1 Five seals of a standard configuration are prepared fromrepresentative
6、specimens of the glass and metal to be tested.The glass and metal are cleaned, treated, and sized to specifiedproportions. Plane-interfaced seals are formed, annealed, andmeasured for residual optical retardation. The stress parallel tothe interface in each seal is calculated from the opticalretarda
7、tion, and the average stress is computed for the sample.For disk-seals the thermal expansion mismatch is calculated.4. Significance and Use4.1 The term “reference” as employed in this practiceimplies that either the glass or the metal of the referenceglass-metal seal will be a “standard reference ma
8、terial” such asthose supplied for other physical tests by the National Institutefor Standards and Technology (NIST), or a secondary referencematerial whose sealing characteristics have been determined byseals to a standard reference material.5Until standard referencematerials for seals are establish
9、ed by the NIST, secondaryreference materials may be agreed upon between manufacturerand purchaser.5. Apparatus5.1 Polarimeter, as specified in Test Method F 218 formeasuring optical retardation and analyzing stress in glass.5.2 Cut-Off Saw, with diamond-impregnated wheel and No.180 grit abrasive bla
10、de under flowing coolant for cutting andfine-grinding glass rod.5.3 Glass Polisher, buffing wheel with cerium oxide polish-ing powder or laboratory-type equipment with fine-grindingand polishing laps.5.4 Heat-Treating and Oxidizing Furnaces, with suitablecontrols and with provisions for appropriate
11、atmospheres(Annex A1) for preconditioning metal, if required.5.5 Sealing Furnace, radiant tube, muffle or r-f inductionwith suitable controls and provision for use with inert atmo-sphere.5.6 Annealing Furnace, with capability of controlled cool-ing.5.7 Ultrasonic Cleaner, optional.5.8 Fixture for Fu
12、rnace Sealing, designed as suggested inAnnex A2.5.9 Micrometer Caliper, with index permitting direct read-ing accuracy of 0.02 cm.5.10 Immersion Mercury Thermometer.6. Materials6.1 MetalRepresentative specimen pairs of the metalfrom either rod or plate stock with dimensions satisfying therequirement
13、s of 7.2 or 7.3. The surfaces to be sealed should berelatively free of scratches, machine marks, pits, or inclusionsthat would induce localized stresses. The sealing surfacesshould terminate in sharp edges at the peripheral corners to actas a glass stop. Edges that are rounded, such as appear ontumb
14、led parts, will have the tendency to permit glass overflow.1This practice is under the jurisdiction of ASTM Committee C14 on Glass andGlass Products and is the direct responsibility of Subcommittee C14.04 on Physicaland Mechanical Properties.Current edition approved Oct. 10, 1998. Published January
15、1999. Originallyapproved in 1971. Last previous edition approved in 1995 as F 140 83 (1995)e1.2Annual Book of ASTM Standards, Vol 10.05.3Annual Book of ASTM Standards, Vol 15.02.4Annual Book of ASTM Standards, Vol 10.04.5See NIST SP 260.1Copyright ASTM International, 100 Barr Harbor Drive, PO Box C7
16、00, West Conshohocken, PA 19428-2959, United States.6.2 GlassRepresentative specimens of rod or plate glass,cut with either diamond-impregnated or other abrasive cuttingwheels under flowing water. Dimensions (volume) shall satisfythe requirements of 7.2 or 7.3.7. Test Specimen7.1 Two basic cylindric
17、al geometries are considered. Fordetermining only the stress in glass, a seal whose total lengthis at least twice its diameter must be used. For determiningexpansion mismatch (as well as stress) a seal whose totalthickness is equal to or less than one fifth of its diameter mustbe used.7.2 The design
18、 for measuring stress provides seals betweena cylindrical rod specimen of glass and metal of either rod orsheet (strip) form. The standard rod seal of Fig. 1(a) shall bemade from specimens so that the diameter of the metal, dm, is0.5 to 1.0 mm larger than the diameter of the glass, dg, beforethe sea
19、l is made; the lengths lgand lmshall each be at least dg.The standard sheet seal of Fig. 2(a) shall be made fromspecimens so that lgis at least 10 lmand a and b each exceeddgby at least 1.0 mm. In all cases dgshall be at least 5.0 mm;d is defined as the sighting line (or light path) through the glas
20、sat the interface after sealing.7.2.1 Record the dimensions of glass and metal.7.3 For determining the thermal expansion mismatch be-tween the metal and the glass, the standard disk seal shown inFig. 3(a) is made. Here dmmay exceed dgby 0.5 to 1.0 mm; dgshall be at least 10 mm. The metal to glass th
21、ickness ratio, tm/tg,may range from13 to 1; d is defined as the sighting line (orlight path) through the glass at the interface after sealing andmust be at least 5 (tm+ tg).7.3.1 Record the dimensions of glass and metal.8. Preparation of Specimens8.1 MetalChemically clean the specimens to removesurf
22、ace contaminants, especially lubricants and fingerprintsfrom fabrication and handling. Usually it is advisable topreoxidize parts as described in Annex A1. Preoxidationpromotes a better glass-to-metal bond and relieves cold-working stresses.NOTE 1The cleaned and heat-treated metal should be sealed w
23、ithin 24h and should be protected from surface contamination during this period.8.2 GlassUsing optical-glass techniques grind and polishthe sealing surface of the glass specimens with either wetabrasive wheels or water slurries of abrasive on a lap. Thepolished surface should be at 90 6 2 to the spe
24、cimen axis andwithout chips, nicks, or scratches. Remove any surface con-taminants which could produce bubbly seals. An ultrasonicwash may be used (Annex A1).8.3 Measure and record the dimensions (diameter, length,thickness) of each glass and each metal specimen.9. Procedure for Making the Butt-Seal
25、9.1 Record dimensions of metal plates and glass parts.9.2 Make the seal in a furnace, by flame, or by inductionheating of the metal, utilizing suitable specimen holders orsupports under controlled conditions of temperature and time(Annex A2).10. Annealing10.1 Once a symmetrical, bubble-free seal has
26、 been made,proper annealing of the seal becomes the most critical part ofthe procedure. It is by this operation that all stresses arerelieved except those due to the difference in thermal contrac-tion of the two materials from annealing temperature levels.FIG. 1 Rod SealsFIG. 2 Sheet SealsFIG. 3 Dis
27、k SealsF 140 98 (2003)2This process involves heating the seal to a temperaturesomewhat higher than the annealing point of the glass andmaintaining the temperature for a time sufficient to relieve theexisting strain. The test specimen is then cooled slowly at aconstant rate. As an alternative, anneal
28、ing can proceed directlyon cooling during the making of a seal.10.2 Seal stress and associated expansion mismatch can bevaried markedly by annealing schedule modification. For thisreason, when the test is used as an acceptance specification, itis strongly recommended that producer and user mutuallyd
29、efine the annealing schedule and establish rigid controls for itsmaintenance.11. Procedure for Measuring Optical Retardation11.1 For each specimen measure the retardation in theannealed seal due to the stress parallel to the interfaceaccording to Test Method F 218.11.1.1 Position the cylindrical axi
30、s of the glass (in animmersion liquid, if needed) in a direction 45 from thedirection of vibration of the polarizer and analyzer, so that theline of sight or light path lies in the plane of the interface andpasses through its center.11.1.2 Determine the retardation along the light path interms of de
31、grees of rotation of the analyzer. Rotate the analyzerin a direction that causes the curved black fringe seen withinthe glass to appear to move up to but not beyond theglass-metal interface (as though into the metal). Rotate theanalyzer so that any light or “gray” area which may existbetween the dar
32、kest part of the fringe (its center of width) andthe surface of the metal disappears; this condition is termed“extinction.” When extinction is achieved correctly, the widthof the black fringe should appear to be about half its initialvalue, the other half apparently being obscured by the metal.Recor
33、d the rotation of the analyzer required to produceextinction.NOTE 2Sealing combinations may exist in which the thermal expan-sion coefficients of glass and metal at room temperature may differsignificantly. In these cases it may be important to record the temperatureof the refraction liquid (or the
34、seal) at the time the retardation is measured.NOTE 3In certain glasses, especially those compositions containingmore than one alkali oxide, part of the retardation observed may not beassociated with the mismatch stress of interest. In these cases somestructural birefringence is caused by temporary s
35、tresses at elevatedtemperatures. The exact analysis of mismatch stress should be evaluatedby completely removing the metal member by acid immersion. Theretardation should again be read at the same glass surface. Any residualretardation should then be algebraically subtracted from that previouslyobse
36、rved.NOTE 4If it is desired to minimize any uncertainties about measuringthrough the curved surfaces, these may be ground after annealing toconform to the alternate shapes of Fig. 1(b), 2(b), or 3(b). Opposing facesshould be ground so as to be parallel to each other and normal to the planeof the sea
37、l interface each within12 . For rod seals or sheet seals, grindingshould be such that in Fig. 1(b) and 2(b) the dimension d is not less than0.8 dg. In the case of the alternative disk seal of Fig. 3(b), d must still beat least 5(tm+ tg). Grinding should be followed by reannealing beforemeasuring ret
38、ardation. It should be borne in mind that grinding mayproduce micro or macro cracks at the interface with the uncertaintiesassociated with these conditions.11.1.3 If an immersion liquid is used record the nominalindex of refraction, nD, of the liquid, and measure and recordto the nearest 0.1C the te
39、mperature of the liquid using animmersion mercury thermometer.11.1.4 Record the type of light source and the effectivewavelength, L, in nanometres of the light for which theretardation has been measured. Record the interface extinctionangle and sense (tension or compression) as defined in TestMethod
40、 F 218.11.1.5 Measure the length d along the light path (Fig. 1, 2,and 3) using a micrometer caliper with an index permittingdirect reading of 0.002 mm.12. Calculations12.1 Calculate the retardation per unit length of each speci-men as follows:R5LA/180d (1)where:R = retardation per unit length, nm/n
41、m,L = wavelength of light source, nm,A = rotation of analyzer, deg, andd = length of the light path through the interface, nm.12.2 Calculate the average, R, of the values of R for thespecimens in a test lot.12.3 For each test lot, calculate the average seal stressparallel to the interface using the
42、relationship:S5R/K (2)where:S = stress parallel to interface, Pa,R = average retardation per unit length of the test speci-mens, nm/nm, andK = stress-optical coefficient of the glass, Pa1.NOTE 5The stress-optical coefficient K of any reference glass shall besupplied by the manufacturer. Values for t
43、ypical sealing glasses are foundin Table A1 of Specifications F 79 and F 105.12.4 Calculate the thermal expansion mismatch (or differ-ential thermal contraction of glass and metal between tempera-tures in the annealing range of the glass and room temperature)for the disk seals using the equation:6DL
44、/L!T5S12!/EgF Note 6! (3)where:(DL/L)T= total expansion mismatch between setting point of glass and roomtemperature, m/m, = Poissons ratio for glass,F = shape-modulus factor (kr4+3r2+4r)/(kr4+4r3+6r2+4r +1/k),k = Em/Eg,Em= Youngs modulus for metal, Pa,Eg= Youngs modulus for glass, Pa,r = tm/tg,tm= t
45、hickness of metal, mm, andtg= thickness of glass after sealing, mm.NOTE 6Use of this equation is valid only if d is a minimum of5(tm+ tg), the measurement is made at the glass-metal interface, and the6Ondracek, M., “Magnitude and Distribution of Stresses in Test Seals Used inthe Photoelastic Study o
46、f Joints Between Two Materials and in the Padmos Test”,Silikaty, SITKA, Vol 7, 1963, pp. 118. (In Czechoslovakian; English translationavailable from SLA Translation Center, 35 W. 33rd St., Chicago, IL 60616.)F 140 98 (2003)3unsealed faces of the glass and metal are parallel to the interface within 1
47、.12.4.1 The shape-modulus factor, F, may be estimated fromFig. 4.13. Report13.1 Report the following information:13.1.1 Type of metal and identification,13.1.2 Type of glass and identification,13.1.3 Diameter and length of glass for each specimen,13.1.4 Diameter and length (or length, breadth, and t
48、hick-ness) of metal rod (or sheet) for each specimen,13.1.5 Average oxide thickness for specimens in a test lot interms of gain in weight per unit surface area after oxidation,13.1.6 Number of specimens tested,13.1.7 Annealing schedule,13.1.8 Length of the light path through glass at interface forea
49、ch specimen,13.1.9 Average and range of calculated retardation per unitlength,13.1.10 Stress-optical coefficient of glass,13.1.11 Type of light source and effective wavelength,13.1.12 Nominal index of refraction of immersion liquidand its temperature at the time of retardation measurements or,if no immersion liquid is used, the temperature of the seal,13.1.13 Average value and sense (tension or compression)of the stress in the glass,13.1.14 Average thermal expansion mismatch (or differen-tial thermal contraction) between metal and glass in the case ofdisk-
