US2665412A - Analysis of nonferrous alloys - Google Patents

Description

Jan. 5, 1954 J, EDING ETAL 2,665,412

ANALYSIS OF NONFERROUS ALLOYS Filed March 31, 1952 5 Sheets-Sheet l f 3 l I f 3/ 9 1 Z y 30 l? 2/ Z r16 L CLARENCE GLASSBROOK ATTORNEY 1954 H. J. EDING ETAL 2,665,412

ANALYSIS'OF NONFERROUS ALLOYS Filed March 31, 1952' s Sheets-Sheet 2 A z 440's. Q 5 Q Q. 0 s k 450 '0. E E i MILL/VOL rs I N VEN TOR5 Jan. 5, 1954 H. J. EDING ET AL 2,665,412

ANALYSIS OF NONFERROUS ALLOYS Filed March 51, 1952 5 Sheets-Sheet 5 FIG. 3

I N VEN TORS HAROLD J. EDING CLARENCE I. GLASSBROOK ATTORNEY Patented Jan. 5, 1954 2,665,412 ANALYSIS OF NONFERROUS ALLOYS Harold J. Eding, Palo Alto, and Clarence I. Glassbrook, Hayward, Calif.,

poration, New York,

Delaware assignors to Ethyl Cor- Y., a corporation of Application March 31, 1952, Serial N 0. 279,662 8 Claims. (01. 324-71) ducing metallic alloy compounds for chemical purposes. Foremost among such instances are the alloys of lead, particularly with the alkali metals or the alkaline earth metals. Such alloys include,

for example, sodium-dead alloys, magnesium-lead alloys, calcium-lead alloys, ternary alloys containin lead, sodium and potassium, and potassium-lead alloys, and the like. Foremost among alloys of this nature is the monosodium alloy of lead, NaPb, containing 16 weight percent sodium and 90 percent lead. This monosodium alloy is widely used in the manufacture of the alkyllead compounds, particularly tetraethyllead.

The prior commercial methods employed for the preparation of the lead-sodium alloys may be considered typical of the time-consuming procedures necessary. Thus, in the manufacture of an alloy containing weight percent sodium, the ordinary procedure has been to mix molten lead and molten sodium to the approximate composition required, using the weights of materials as guides. Upon the completion of the mixing operation, it has been customary to take a sample and carry out the chemical analysis to determine what corrective action must be taken to arrive at the final precise composition necessary. The method of analysis heretofore used has been to contact the alloy sample with mercury and a standard amount of acid of known concentration. After complete reaction of the sodium content of the alloy, the residual or excess acid is back-titrated with a caustic solution to determine the amount of sodium metal present. This method, while accurate, involves the usual delay of chemical analytical procedures, and has caused the hold-up of production while such laboratory analysis was carried out. Obviously, the time required for such analysis to determine what corrective action was required is a bottle-neck in the manufacturing operations. A more rapid and precise analytical technique was therefore desired.

The object of the present invention is to provide a method and apparatus for the analysisof molten metal alloys without the necessity of chemical manipulation. A further object is to provide apparatus which is relatively simple and requires no special skill in operation thereof. A further object is to provide a method and apparatus therefor capable of a high degree of reproducibility and accuracy in a long series of repetitive determinations. Other objects will appear hereafter.

It has now been discovered that the composition of a molten alloy can be ascertained with a great degree of precision by determining the electrical potential difference or voltage difference between the alloy and a reference metal of known composition. The reference metal as well as the alloy should be in the molten state and preferably at the same temperature. A specimen of the referenc metal is maintained adjacent to the alloy being analyzed, and is contained within a capsule of non-metallic material, the capsule being P of thin glass or some other vitreous composition.

Broadly, therefore, the method of the invention comprises placing a specimen of the molten reference metal within a body of the alloy being analyzed, but separated from admixture therewith by a barrier or liquid tight capsul container which is permeable at operating temperatures to ions of the metal being determined. The reference metal specimens so placed is, however, electrically isolated from the surface of the alloy body. The voltage developed between the alloy proper and the reference metal specimen is measured and is a function of and an indication of the composition of the alloy.

The reference metal can be any of a number of different compositions. In most instances, a pure specimen of the metal being determined in the alloy will be used, but an alloy or mixture of such a metal with a different metal is equently used. In other instances, an equivalent metal will be used, that is, the reference metal will not include or consist of the metal being determined by the analysis.

It has been found essential that a truly representative voltage difference be ascertained between the alloy being analyzed and the reference metal. By this is meant that certain obscur lg Voltages, having no relation to the voltage difference between the alloy proper and the reference metal, are screened out or prevented from being impressed on the voltage between the alloy prop-er and the reference metal. It has been found that if this is not done, that not only is the representative voltage not ascertainable at the start,

but on continued usage of a particular apparatus,

the obscuring voltages increase with time so there is no opportunity for standardization.

The above mentioned obscuring voltages are apparently resultant from impurities or contaminants on the surface of the alloy being analyzed when in contact with the apparatus used. In order to negate the adverse effect of such obscuring voltages, it is essential to provide electrical isolation of the reference metal specimen from the surface of the alloy.

The method of the invention and the best manner of use will appear from the detailed description given hereafter and from the figures. Fige ure 1 illustrates typical requisite apparatus for carrying out the method, including a voltage measuring instrument, electrical conductors for transmitting the voltage of the alloy and of the reference metal, and a reference metal assembly. Figure 2 shows a typical relationship between the composition of an alloy, of sodium and lead, and a voltage developed between such an alloy and reference metal consisting of sodium. Figure 3 is a schematic illustration of auxiliary electrical apparatus receiving and indicating the voltage difference signal measured between the alloy and the reference metal and for providing a further refinement in accuracy.

Referring to Figure 1, the essential elements of apparatus for employing the method are shown, these including a reference metal assembly I I, an alloy potential wire l2 for ascertaining the electrical potential of the alloy being analyzed, and a voltage measuring instrument 13. The reference metal holder i I is partially immersed in the alloy 22, and the alloy wire [2 has its terminus, which may be a small coil 24, also immersed in the alloy sample. A voltage diiference between a reference metal specimen [8 and the alloy 22 is transmitted by wires i2, I? to the voltmeter 13 for measurement and indication,

The reference metal holder ll includes a cap sule Hi in the form of an elongated tube of circular cross section, a sleeve I5, a protective sheath [6 and an electrical lead wire [1. The lead Wire ll enters the capsule It through a gas tight seal 2? and is normally positioned concentrically in the capsule. The terminus of the wire H, usually a short coil 26, is immersed in the reference metal specimen l8. 7

The capsule It and the sleeve l are made of the same material, e. g., a heat resistant glass, the sleeve being permanently joined to the capsule wall in preparation. The annular space 28 between the capsule and sleeve is open to the atmosphere. In some instances the sleeve may be joined to the capsule at top and bottom. The protective sheath It is of a metal resistant to corrosive attack by the alloy being analyzed, iron or steel being suitable for most purposes. The protective sheath may be in a number of mechanical forms, a preferred shape being a shouldered cylindrical bushing as shown. The sheath l6 fits fairly snugly about the sleeve 15, and has a shoulder 3| at the topmost extremity which rests on the wall 36 of the alloy conduit or vessel. Usually a packing 32 is wedged in between the bottom of the sheath l6 and the sleeve I5 to provide a liquid and gas tight seal. The packing or seal material may be of asbestos fibers, or asbestos plus sauereisen cement or a similar sealing material which is inert to the alloy at the temperatures involved. In addition to providing a fluid seal, the packing material 32 provides for mechanical support of sleeve and capsule in operation.

In virtually all instances it is highly desirable that the free space 59 within the capsule It (not occupied by the reference metal specimen [8) should be filled with an inert gas. Suitable gases are argon, helium, nitrogen or an alkane hydrocarbon gas. The exclusion of any gaseous component which could react with the reference metal precludes the possibility of reaction products creating an obscuring voltage by contact with the lead wire H. In constructing and filling the capsule It of the reference metal holder, the lead wire is usually first inserted and a gas tight seal 2? is formed. A short branch or nipple is part of the unfilled capsule and serves as the filling port for the reference metal and for the inert gas protective blanket. After charging thecapsule this nipple is closed, preferably by fusion of the glass, leaving a slight projection 25 as shown.

In this form of reference metal assembly, it is essential that the seal or point of entry 21 of the lead wire I? into the capsule be gas tight. This may be accomplished in several ways. For example, a metal having an expansion coefficient near to that of glass, such as tungsten, may be used. In such cases, the portion of the wire i'i passing through the capsule wall is made of such a metal and is sealed in a conventional manner, by melting and pinching the glass nipple about the wire. An alternative and preferred mode of making this seal involves melting a supply of silver chloride about the junction of the lead wire and the capsule wall, which has been found to provide a highly efficient seal.

A closed and sealed reference metal capsule as shown is the preferred mode of protecting the purity of the reference metal. However, the same degree of protection can be realized even if the capsule container is open at the top end. For such embodiments, a continuous small stream of a pure, dry inert gas is passed into the capsule. This method of operation of course requires more frequent attention than necessary when a sealed capsule is employed, but nevertheless provides reproducible, accurate results.

In operation, the reference metal holder is partially immersed in a sample or stream of molten alloy 22. The protective sheath it projects through the surface of the alloy. It has been found that alloy melts are apparently covered by a thin, almost imperceptible film 21 of products of corrosion of the alloy, or slag impurities, which tend to attack the material of the sleeve l5. The sheath thus prevents such contact and corrosion.

It will be noted that the surface 23. of the speciment N3 of the reference metal occupies an elevation, when the assembly is in operation, which is somewhat below the seal of-the sleeve 55 with the capsule l4. This disposition of the reference metal specimen is not particularly significant with respect to the operability of the method, but-is desirable in minimizing thermal strains at the seal 28.

In contrast to the reference metal elevation, the disposition of the seal point 23 with respect to the sheath 5% is of great importance in the apparatus. It has been found that adequate dis" tanoe between these two points is essential for providing the necessary electrical isolation. This separation distance is not precisely ascertainable in advance for all installations, as the magnitude of obscuring voltages, apparently associated with th surface of an alloy sample, are variable according to the concentration and identity of the components of the alloy. In general, however, it is found that providing a separation distancethat is, the linear distance between the sheath is lower extremity and the seal 28of at least ten times the wall thickness of the capsule, will provide a high degree of isolation. As the capsule will ordinarily be made of material having a wall thickness of less than 1 millimeter, a separation distance of the order of about one-half to one inch provides an ample margin of safety. In those instances wherein a metal sheath is not used, the separation distance requirement refers to the distance from the alloy surface to the joint of the sleeve with the capsule.

Surprisingly, the desired electrical isolation cannot be achieved by merely extending the length or the capsule M or by other more or less obvious measures. For example, it would be theorized that if capsule M were sufiiciently long in dimension, and the wire I7 did not touch the interior walls other than at the point of seal 21, that merely by having the reference metal specimen sufficiently far below the level 2| of the alloy sample, that electrical isolation would be achieved. However, numerous tests have shown that this not the case, and that an inaccurate and variable voltage is measured. It is believed that at the high temperatures of operation, a thin film of reference metal is vaporized and deposited on the interior walls of the capsule. This thin film, even though of extremely thin proportions, is suflicient to transmit obscuring voltages to the reference metal specimen proper or directly to the wire I1. A sleeve I i, such as is illustrated, joined to the capsule at a point 28 removed from the above-mentioned separation distance is thus essential to the method.

As the above described combination provides electrical isolation of the entire capsule, it is immaterial whether the wire I! touches the interior walls of the capsule. This makes possible particularly compact assemblies when desired. Thus, the capsule may have a restricted diameter portion, say about one-fourth inch, above a bulb portion holding the reference metal specimen. As heretofore mentioned, the wire ll may be in contact with the walls so that concentricity is not essential. In addition the reference metal specimen may be of such volume that it occupies a portion of the restricted diameter section of the capsule. No particular benefit is realized from such an increase in volume of the reference metal specimen, however, and in all instances wherein a sealed capsule is used a certain amount of free space is necessarily provided. Such free space is needed to prevent stresses and cracking of the capsule owing to expansion of the metal specimen with temperature changes.

The two lead wires 12, I 1 are made of the same metal, pure soft iron thermocouple wire being a preferred material. The terminus 24 of the wire 12 to the alloy sample is located quite near to the reference metal holder, and at about the same immersion, so that the temperatures at these two points will be the same.

An alternative but generally less satisfactory mode of obtaining the voltage of the alloy proper is to connect the conductor I2 to the metal container or conduit for the alloy at a point adjacent to the reference metal specimen holder. This procedure is less satisfactory because of the likelihood of errors being introduced because ofvoltages being generated through differences in composition of the metal container for the alloy and sheath I 6 by the' of the conductor H for ascertaining the potential of the reference metal specimen.

The chart of Figure 2 illustrates a typical voltage-composition relationship such as is utilized by the present method. Referring to Figure 2, these relationships are given for an alloy comprising sodium and lead, curve A being for a uniform temperature of 440 C. and curve B being for a temperature of 450 C. It will be seen that ascertaining the voltage difference between a sodium-lead alloy and pure sodium metal affords a rapid and precise method for determining its composition. Thus, at 440 C., for example, a millivoltage of 197 shows that the alloy contains 10 weight percent sodium, while an increase in voltage to 200 mv. indicates that the alloy contains 9.88 percent sodium. The curves of Figure 2 were prepared by determining the voltage difference exhibited between a series of alloys of known composition and pure sodium. For other systems, similar operational data are readily determined.

As previously indicated, the usual shape of the reference metal capsule is an elongated tube having a hemispherical bottom. The shape and size of the reference metal specimen has been found to be of little importance with respect to accuracy of the method. An examination of performance of dence of a potential of asymmetry. In all cases, for practical reasons of assembly, it will be desirable to provide a capsule of about one-half inch internal diameter at the locus of the reference metal specimen.

- With reference to the vitreous material utilized for the capsule and the sleeve of the reference metal holder, aheat and shock resistant glass is preferred. The precise mode of operation of the vitreous material with respect to establishing the measurable voltage difference is not completely understood. It has been found that such materials are dielectrics, that is, electron flow is factory glass of this type is as follows, compositions in weight percent: 80.5 SiOz, 2.2 A1203, 3.8 NazO, 0.4 K20 and 12.9 B203. Other glasses have been used with very satisfactory results. How ever, not all glasses have adequate resistance to thermal shock. For example, although a capsule made of a glass containing about 19 percent sodium oxide has proven ver acceptable with respect to analysis of a sodium-lead alloy, it is fairly rapidly attacked owing apparently to its low physical strength at the elevated temperatures of operation. Consequently, its effective service life is limited. As a general rule, the capsule material should include as a component the oxide of the metal being determined in the alloy, although such oxide may be present in low concentrations, As an alternative to the construction here illustrated, the capsule and the sleeve walls may be strengthened by circumferential wire bands within the wall. Ordinarily, however, such elaboration is not necessary.

Generally, the wall thickness of the capsule does not affect the potential accuracy of determination of the voltage difference to be ascertained. The wall thickness does, however, affect the ease with which the voltage difference is measured. This arises because an increase in the capsule wall thickness. increases the elec trical resistance of the system and requiresfurther refinements in th external electrical circuit. For these reasons, a capsule wall thickness of about one-half to one millimeter is preferred.

Each of the several components of the reference metal holder are important for obtaining thefull benefits of the invention. The protective metal sheath it may be dispensed with when desired; its function being to provide chemical and mechanical protection of the capsule and sleeve. In addition, the protective sheath provides a mounting or support whereby the assembly is positioned within a conduit, pipe or vessel holding the alloy.

Although the protective metal sheath maybe eliminated, the integral combination of the capsule and the sleeve is highly desirable in all practical forms of the apparatus to provide the isolation of the reference metal Specimen. The importance of this combinationis illustrated by the following examples.

Example I A reference metal holder resembling the holder two weeks of stead operation. It was found that the voltage difference measurement failed to remain constant, but steadily increased with time. Thus, the initial voltage was 212 mv., and after 5, l0, and 12 days operation the voltage measured was 4.30, '70, and 11,68 mv., respectively. It

is apparent that the reference metal holder of this test would be entirely unsatisfactory for analyzing an alloy sample.

Example II A reference metal holder such as is described heretofore, and as illustrated by Figure l, was made of bore-silicate glass and charged with pure sodium metal. As in the previous example, the reference voltage wire made occasional contact with the capsule walls above the level of the alloy specimen. The instrument was used for measuring the voltage difference between the sodium specimen and sodium-lead alloy having 9.6.3 weight percent sodium and maintained at about 440 C. Measurements were taken continuously over a period approaching four weeks. The voltage readings varied less than one-half percent during this period.

The foregoing examples illustrate the-importance of providing electrical isolation of the reference metal specimen from the alloy sample. surface and the efficacy of our apparatusin this re-. gard. It has been found that the method of analyzing an alloy is extremely versatile. For example, it is not essential that the reference metal be identical with the component being determined in the alloy. As an example of such usage, sodium can be. employed as the reference metal in determining the amount of potassium in a ternary alloy of sodium, potassium and lead. In such instances, it will be necessary to ascertain the amount of sodium present, by application, of the method to the binary alloy of sodium and lead, before addition of the potassium component.

The addition of potassium to the alloy produces the same effect on the voltage difference as the addition of a stoichiometric equivalent of sodium metal. It will therefore be apparent that the procedure is applicable to the analysis of multicomponent alloy systems.

In general, it is possible to prepare or manufacture completed reference metal assemblies which are interchangeable. As a practical matter, it is highly desirable to make at least one calibration determinati n before using a new assembly. The necessity of such a check run arises because of the possibility of inclusion of small amounts of impuritie in the reference metal specimen which would have the eifect of altering the voltage readings slightly.

Theinstrument for measuring the voltage difference may be of several different types. Thus, a manually operated potentiometer may be used; A conventional voltmeter is, however, not suitable, inasmuch as the high internal resistance of the capsule wall will introduce an error in the voltmeter reading. In order to obtain the maximum accuracy, it will frequently be desirable to employ a bucking voltage, so that a voltage measuring instrument of greater sensitivity can be used. Figure 3 illustrates schematically an electrical circuit for such cases.

Referring to Figure 3, elements are shown of an embodiment for attaining a record of composition with a high degree of precision. Such elements include of course the reference metal assembly inserted in place. in a body of the alloy and appropriate electrical leads for transmitting the voltage developed. In addition electrical apparatus is supplied for-applying a suppressing or bucking voltage, opposite in direction to the voltage developed, and appropriate. indicating recording apparatus.

The suppression voltage has been found to contribute to the accuracy of the record of analysis obtained in the following manner. In a particular operation, ordinarily, the alloy componentscan be easily blended in approximately the desired proportions. Therefore, the analyses'will usually be of alloys deviating only minor amounts from the desired final precise composition. Accordingly, it will then not be necessary to measure voltages over the entire scope of voltages developed by the reference metal-alloy combination. Therefore, by applying a constant bucking voltage corresponding to the voltage developed-v when the alloy is at or near the desired, composition, the residual voltageidifference between the actual developed and the bucking voltages). can be measured'most accuratelyby employing an instrument providing the highest sensitivity for the small variation in residual voltage.

As an illustrative example, considering themanufacture of monosodium-lead alloy, a Voltage of the order of. millivolts is approximately thevoltage developed by this alloy at its desiredcomposition. By applying a suppressin voltage of slightly less than 195. millivolts and measuring the residual voltage with an instrument capable of. measuring over a span of only five millivolts with an accuracy of about onev percent, a net accuracy in voltage measurement of 0.005 percent, with corresponding precision in determining alloy composition, is readily attained.

It will be readily apparent that, in order to obtain the benefitsof a suppression voltage, it must be applied with a. high degree of precision and stability. The embodiment of Figure 3 illustrates circuiting providing such precision.

Referring to Figure 3, the electrical lead 43 from a reference metal specimen assembly 42 is connected to a resistor 45. Also joined at this point is a lead wire 46 from a conventional dry cell 4'! (or other semi-permanent direct current source) in opposite polarity to the voltage developed. A standard cell 48 is connected across a fixed resistor 69 in opposite polarity to the dry cell :17. The standard cell 48 is connected to the resistor 49 through an amplifier 50.

A third resistor of variable capacity is adjusted by a balancing motor 52 in response to impulses from the amplifier 50. The resistor 49 is so chosen that when sufficient current flows in the battery circuit to create a voltage drop in resistor 45 exactly equal to the desired suppression voltage, then the concurrent voltage drop in resistor 49 is equal to the voltage or the standard cell 43. Any unbalance with respect to cell 48 and resistor 49 is detected by amplifier 50 Which transmits a power impulse to balancing motor 52 which adjusts variable resistor 5| to provide the desired voltages across resistors 49 and 45.

It has been found that the above described circuiting is capable of extremely high precision in providing a suppression voltage. A stability of better than one part in 10,000 has been obtained in providing a suppression voltage of 192.5 millivolts.

The discharge wire 53 from the voltage suppression section and the wire 44 from the alloy sample therefore provide a residual voltage to a recorder 54, the said residual voltage being practically instantaneously responsive to the composition of the alloy. The recorder 54 may be any of a number of readily available instruments and is fitted with proper adjustments to provide maximum sensitivity over the voltage range to be applied thereto.

Having described our process and apparatus and the preferred embodiments thereof, What we desire to claim is:

l. The process for analyzing a molten nonferrous alloy comprising immersing a liquid reference metal specimen in a body of the alloy but separated therefrom by a barrier of a vitreous material permeable to ions of the metal being determined, electrically isolating the reference metal specimen from the surface of the alloy, establishing an electrical connection between the alloy and the said reference metal specimen and measuring the voltage developed between the alloy and the reference metal.

2. The process for analyzing a molten alloy of lead with at. least one alloying metal selected from the group consisting of alkali and alkaline earth metals, comprising immersing a molten ref erence metal specimen in a body of the alloy but separated therefrom by a barrier of a vitreous material permeable to ions of the metal being determined, electrically isolating the reference metal from the surface of the alloy, establishing an electrical connection between the alloy and the said reference metal specimen and measuring the voltage developed between the alloy and the reference metal.

3. The process of claim 2 further defined in that the alloy consists essentially of lead and sodium and the reference metal consists essentially of sodium.

4. The process of claim 2 further defined in that the alloy consists essentially of lead, sodium, and potassium, and the reference metal consists essentially of sodium.

5. The process for analyzing a molten alloy of lead and sodium comprising immersing a molten specimen of sodium in a body of the alloy but separated therefrom by a barrier of borosilicate glass, electrically isolating the specimen. of sodium from the surface of the alloy, establishing an electrical connection between the alloy and the sodium specimen to transmit the voltage developed between the alloy and the sodium, applying a constant suppression voltage to said developed voltage, and measuring the difference in voltage between the developed and suppression voltages.

6. A reference metal assembly for placing a specimen of reference metal Within a body of molten alloy but electrically isolated from the surface of the alloy comprising a capsule adapted for partial immersion in the alloy; a surrounding sleeve defining an annular space therewith and circumferentially sealed to the capsule to prevent entry of the alloy to the annular space; a molten reference metal specimen within said capsule; and an electrical conductor into the capsule and in electrical contact with the reference metal specimen; said conductor being electrically isolated from the molten alloy, and the capsule and the sleeve being of a dielectric vitreous material permeable to ions of a metal component of the alloy.

7. A reference metal assembly for placing a specimen of reference metal within a body of molten alloy but electrically isolated from the surf-ace of the alloy comprising an elongated closed capsule adapted for partial immersion in the alloy; a surrounding sleeve defining an annular space therewith and circumferentially sealed to the capsule to prevent entry of the alloy to the annular space; a, molten reference metal specimen within said capsule; and an electrical conductor extending through a Wall portion of the capsule and in electrical contact with the reference metal specimen, said wall portion being remote from the reference metal specimen and electrically isolated from the molten alloy, and the capsule and sleeve being of a dielectric vitreous material permeable to ions of a metal component of the alloy.

8. A reference metal assembly for placing a specimen of molten sodium metal within a body of molten sodium-lead alloy but in electrical isolation from the surface of the alloy, comprising an elongated, closed capsule adapted for partial immersion in the alloy; a surrounding sleeve d fining an annular space therewith and circumferentially sealed to the capsule to prevent entry of the alloy to the annular space; a molten speci men of sodium within said capsule; and an electrical conductor extending through a wall portion of the capsule and in electrical contact with the molten sodium specimen; said wall portion being remote from the molten sodium specimen and electrically isolated from the molten alloy, and the capsule and sleeve being of borosilicate glass.

HAROLD J. EDING. CLARENCE I. GLASSBROOK.

References Cited in the file of this patent UNIED STATES PATENTS Number Name Date 2,342,029 Zubko Feb. 15, 1944 2,366,844 Doschek Jan. 9, 1945

Claims (2)

1. 5. THE PROCESS FOR ANALYZING A MOLTEN ALLOY OF LEAD AND SODIUM COMPRISING IMMERSING A MOLTEN SPECIMEN OF SODIUM IN A BODY OF THE ALLOY BUT SEPARATED THEREFROM BY A BARRIER OF BOROSILICATE GLASS, ELECTRICALLY ISOLATING THE SPECIMEN OF SODIUM FROM THE SURFACE OF THE ALLOY, ESTABLISHING AN ELECTRICAL CONNECTION BETWEEN THE ALLOY AND THE SODIUM SPECIMEN TO TRANSMIT THE VOLTAGE DEVELOPED BETWEEN THE ALLOY AND THE SODIUM, APPLYING A CONSTANT SUPPRESSION VOLTAGE TO SAID DEVELOPED VOLTAGE, AND MEASURING THE DIFFERENCE IN VOLTAGE BETWEEN THE DEVELOPED AND SUPPRESSION VOLTAGES.
2. 6. A REFERENCE METAL ASSEMBLY FOR PLACING A SPECIMEN OF REFERENCE METAL WITHIN A BODY OF MOLTEN ALLOY BUT ELECTRICALLY ISOLATED FROM THE

Images