US1663564A - Refractory metal filament - Google Patents

Description

Patented Mar. 27, 1928.

v 1,663,564 PATENT OFFICE.

MALCOLM N. RICH, OF EAST ORANGE, NEW JERSEY, ASSIGNOR TO WESTINGHOUSE LAMP COMPANY, A CORPORATION OF PENNSYLVANIA.

I REFRACTORY METAL FILAMENT.

No Drawing.

This invention relates to the art of drawing refractory metal filaments and more particularly to the productionof drawn tungsten filaments for use in incandescent lamps, electron discharge devices, etc.

While the described invention is particularly applied to the art of drawn tungsten filaments, it is to be understood that itv can be e ually well applied in whole or in part toot er drawnrefractory metal bodies with similar metallurgical characteristics, such as molybdenum, chromium, tantalum, columbium and vanadium. Some ferrous and non-ferrous metals and alloys might similarly be treated in order to facilitate and in1- prove their drawing conditions, especially when subjected to hot working and drawmg processes. a This invention is in the nature of an i1ngrovement on the method of producing rawn tungsten filaments set forth in patent to Coolidge 1,082,933, granted Dec. 30, 1913. Briefly, the Coolidge process consists in pressing owdered tungsten metal into slugs, which a ter a preliminary sintering in hydrogen, are heat treated in vacuo or in hydrogen and then subjected to swaging operations at high temperatures in hydrogen or in gas fired furnaces until a sufficient reduction in diameter of the slugs is obtained, and the physical characteristics of the metal so developed as to permit drawing. 7

The temperatures employed during the heavy swaging operations of this process are relatlvely high, ranging from 800 C. to 1500 C., and as a carbonaceouslubricant is used to reduce ,the'wear on the dies and gas fired furnacesare used to heat the rods, considerable carbon absorption in the metal body takes place. While this carbon absorption is essentially a surface condition, the penetration is found to extend in wires of 36 mil diameter, reduced from the regular one-quarter inch slugs, approximately 1%, to- 2 mils in depth. Moreover, the surface of the wire is overlaid with a hard, coherent and non-plastic scale of carbon, carbides and other impurities, so compacted and driven into the surface of the metal body, as to become practically a art of it. -When this metal body is exposed as in the process described in my copending application Serial No. 4140, filed January 23, 1925, entitled Pliabletungsten and method of producing the same, and assigned to the Westinghouse Application filed October 30, 1925. Serial No. 65,881.

Lamp Company, the surface of the body of the metal is found to contain numerous dents and depressions caused by the uneven or unequal impacts of the swaging dies. These swaging die marks persist in wires even after considerable drawing, and it is apparent,

ingpperations, resists the leveling action of the dies during drawing.

' Heretofore, in the production of drawn tungsten filaments by the Coolidge method, it has been customary, to remove a portion of this surface carbon and carbide layer, by flashing the filament in hydrogen. This is effected usually at two stages. The first stage is after the wire has been drawn down to approximately seven mils. At this stage it has been found necessary to remove work hardening effects and carbon and carbide contaminations by a so-called' annealing flash which consists essentially of a preheating in air and subsequent high temperature flash in hydrogen.

One of the difficulties of this method is that often, due to the extreme variableness of carbon and carbide absorption in the different slugs as drawn down, a standard temperature and time interval of exposure to the hydrogen flash produces extreme variable effects upon the wire. 'For instance, if the carbon absorption be unusually heavy, in-

complete removal will be effected, and the wire will fall below the required physical standards. On the other hand, if the carbon absorption is lighter than usual, the extremely high temperature flash employed or the flashing for a longer interval of time than required, to complete the removal of the carbon carbide scale, will promote excessive absorption of hydrogen in the wire and cause brittleness to develop.

After this annealing treatment, the wires are drawn down to filamentary sizes and the 'residual carbon removed' therefrom before mounting, by flashing in forming gas or in the case of coil t pe lamp filaments, by baking the coils in ydrogen.

As will be noted, this process of producing tungsten filaments permits of variable amounts of carbon and carbides to "be removed, and consequently results in variation in the life, strength and maintenance of filament. Moreover, the full leveling effect of the drawing dies is not exerted during the initial and more important part of the wire drawing process. Consequently, the surface irregularities are carried down with the wire until the first hydrogen flash or annealing step is applied, and subsequent drawing is not sufficient to eliminate them entirely.

The effect of hydrogen upon tungsten and other metals, has not been entirely appreciated heretofore. Hydrogen, while used to a large extent in the metallurgy of tungsten, molybdenum and other metals, both as a reducing agent in the production of metal powder, and as a deoxidizer and decarbonizing agent, exerts on tungsten and other metals a distinct embrittling effect when absorbed in excess in the metal body. With metals like tungsten, molybdenum and chromium, this embrittling effect does not ap parently affect the ductility of the metal at elevated temperatures to any great degree, but does affect its strength and pliability' at room temperatures. With other metals such as tantalum, columbium and vanadium, hydrogen so materially affects the ductility, that these metals cannot be worked until this gas is completely removed.

This embrittling effect of hydrogen on tungsten filaments and wires, seems to be due to absorption and can be removed by exposing the filament to oxidizing conditions. Any process that does not enable exact control over this hydrogen content of the finished filamentto be obtained, introduces a variable that afieots the strength and maintenance of its product- One of the objects of this invention 1s 'to provide a method of obtaining uniform filaments of the refractory metals.

Another object is to provide a refractory metal filament substantially free from the variable effects of carbon, metallic carbides and other volatile constituents in the filament.

Another object is to eliminate the variable embrittling and excessive grain growth effects of hydrogen, in refractory metal filaments. 7

Another object is to improve the working gualities of refractory metal bodies and to acilitate the drawing thereof.

Another object is to provide a strong, rugged and uniform filament free from hot spots and impurities, and which will have a long life, high operating efliciency and a low rate of evaporation.

Another object is to provide a method of rendering the refractory metals pliable and clean preparatory to drawin thereof and to reduce the resistance to the fibw of the metal through the drawing dies.

Otner objects and advantages will appear as the description proceeds.

In accordance with this invention the refractory metal rod at the conclusion of the swaging operation and prior to the drawing thereof is cleaned to remove the adherent scale of carbon, carbides, and oxides, etc., therefrom and the hydrogen is substantially removed therefrom to render the rod soft and pliable. The treating of the rod and removal of the hydrogen may be accomplished by the method set forth in my aforesaid application. Briefly, such method consists'in making the rod the anode of an electrolytic bath which will liberate oxygen at the anode. Such bath may consist of a solution of caustic soda through which the rod is passed in acontinuous manner, care being exercised to prevent the hydrogen from the cathode from dispersing to the anode and thereby nullifying the beneficial effects of the nascent oxygen. The current density on the anode is so maintained that the closely adherent scale and surface impurities are completely removed without a substantial amount of the tungsten going into solution. I have found that by suitably regulating the currentdensity, the wire may become polarized upon exposure of the metal by the removal of the surface scale, and in this manner the metal may be effectively prevented from going into solution, while the hydrogen. is being eliminated. The'proper current density varies with different types of wires, such as thoriated tungsten or pure tungsten wire and is also dependent on the size of the wire. The proper current density can readily be determined by those skilled in the art. The time of treatment also varies according to the size of the wire, its nature, extent of surface carbomzatmn and to some extent, on its grain structure.

It is understood, of course, that the carbon and other surface impurities may be removed by other methods such as scrubbing with hot alkali solutions, etc., and the hydrogen elimination may be achieved in other Ways as by heating in air or in rarefied oxygen or by subjecting the rod to the action of ionized oxygen but I prefer to use the anodic method due to its simplicity and ease of operation.

After the wire has been cleaned and rendered soft and pliable by elimination of the embrittling agent, it may be coated with the usual graphite lubricants and drawn down in the ordinary manner.- Due to the fact that the depressions and'irregularities are now open or uncovered and the metal is more plastic and less brittle, it flows more readily under the dies and the surface dentsand depressions appear to be completely filled up and eliminated and an extremely uniform wire is produced early in the wire drawing process. Wires so treated at 36 mils and drawn to 14 -mils show no surface depressions.

The regular wire drawing process is practiced on these filaments with the exception drogen should be done as quickly as possible to avoid undue hydrogen absorption and consequent embrittling. The standard physical tests heretofore employed can be used as a guide. The wire is then drawn down to the required filament size according to the usual process,

The great difficulty heretofore experienced in producing lamps of uniform quality, life and maintenance has been thought to be due to the extensive detail work involved in the fabrication of the lamps. However, I have found that many of these variable factors can be attributed to non-uniformity in wire product itself. This non-uniformity in wire product is the result of unequal or non-uniformity in cross-sectional area, incomplete and variable carbon or carbide removal, excess absorption of hydrogen, etc. These variables are due to irregular and uncontrolled :onditions during the wire manufacture. By

eliminating these variables, as by this process, and maintaining a uniform wire product, greater consistency and uniformity in lamp manufacture is attained;

The removal of the surface scale of carbon and tungsten carbides, formed during the bination of. the factors of non-uniformity in heavy s wagin'g operations, prior to drawing and the treating of the'wire anodically to, improve the physical characteristics of .the metal body, eliminates a large proportion of the variables from the wire.

As the temperatures during the drawing operations are relatively low and the time of exposure" is relatively short, carbon and car bide formation in the tungsten metal body is confined largely to the surface and can be readily removedvby flashing in hydrogen or otherwise. Furthermore, by increasing the number of available dies to perform the surface leveling and by softenin the metal body through the elimination 0 hydrogen and surface scale, uniformity in cross-sectional area is promoted. Moreover, by removing the variable factor of carbide and carbon elimination from the function of the annealing step, which is usually applied at seven mil size, and confining the annealing ste to its true function of removing work-bar e ing effects, alone or with the ossible elimination of mere surface carbi es, a uniform and constant time and temperature of the hydrogen flash at this stagecan be applied with v the result that the ph sical characteristics-of e wire are more un1 or'm.

When the wire has been drawn down to final filament size, and the forming. of segments and'winding of coils is accomplish excessive temperatures and time intervals in flashing in hydrogen .to clean these parts of carbon and accumulated surface impurities Will t be .mqwed, and sensequenfly, loss from brittle filaments, over baked coils, nonuniformity in cross-sectional area, variable life and maintenance due to variable filaments, willbe substantially eliminated.

\Vith the elimination or reduction of the carbon and carbide content of the incandescent lamp filament and the increase in uniformity of cross-sectional area, a lowering of the rate of evaporation of the filament will be obtained.

ration is hastened by cyclic reactions within the evacuated envelope, taking place between the residual gases of the envelope, the hydrogen liberated fromthe tungsten filament, and residual carbon and carbides in the filament-s. I have found that a reduction of 10% to 20% in the rate of evaporation is obtained with tungsten filament treated in accordance with this invention. Not only is the rate of evaporation of the wire lowered over that of the usual product but a certain temperature This is believed to be ex-' plained on the theory that the rate of evaporise usually observed in filaments as prepared by the prior method is missingand the anodically treated filaments display a constant temperature curve throughout life. The rise in temperature observed in the filaments prepared by the usual wire drawing process is attributed to residual carbon and carbides, and the higher rate of evaporation to a comcross-sectional area, and the residual carbon and carbides reacting with residual gases of the evacuated envelope.

.The improvement instrength greater refinement in grain structureas a result of better drawing conditions, greater uniformity in cross-sectional area, freedom from excess hydrogen, which has a tendency to cause the formation of large grain structure in these metals, and absence of carbides.

While'thisinv'ention specifies a particular method of removing tungsten carbides, and hydrogen, from the metal body, prior to drawing, for the purpose of improving the quality and uniformity of the fi ament, it is to be understood that this invention is not limited to the exact details described but that "both the cleaning operation and the rendering of the metal soft and more ductile and pliable may be carried outiin different ways, and various other changes and modifications may ,be made without departing fromv the invention.

For instance, the anodic ttr'e atment of these filaments may be applied at any stage 'in the wire drawing1 process-and as many times as desired, or t e removin of the carbon and carbide scale me be e ected in any other desired manner. hydro en. ma also be accomplished by other meth s,bi1t haveifound that the eombina- Y tion of the two effects produced 'by" the treatment aecomphshes the and maintenance is attributed to a development of e removal offthe results in a more pronounced manner than where the operations are performed independently.

What is claimed is:

1. The method of treating refractory metal filaments which consists in freeing the surface thereof from impurities and eliminating hydrogen therefrom during the early part of the drawing operation.

2. The method. of treating refractory metal filaments which consists in removing the outer surface thereof and rendering the filament pliable prior to the drawing of the filament to final size and subsequently c0mpleting the drawing operation.

3. The method of producing refractory metal filaments of uniform cross section which consists in removim impurities from the surface thereof and eliminating hydrogen therefrom and subsequently drawing the filement to a smaller diameter.

4. The method of treating refractory metal wire during manufacture which consists in removing the surface impurities therefrom, subjecting the wire to the action of nascent oxygen to render it pliable and thereafter drawing the wire to a smaller size.

5. The method of producing refractory metal wire of uniform cross section which consists in removing the surface impurities therefrom and eliminating from the filament the agents which tend to render the same brittle and thereafter passing the wire through a die to effect a smoothing out of the irre ularities caused by the removal of the surface impurities.

6. The method of treating refractory metal filaments which consists in cleansing the surface thereof and eliminating hydrogen therefrom subsequent to the swaging operation, reducing the wire to filamentary size and prior to reduction of the filament to its final size, retreating the same to eliminate impurities which have been introduced during the reducing process.

7. The method of producing tungsten wire of uniform cross section which consists in removing the hard surface from the wire, treating the wire to render it pliable and thereafter working the wire to distribute the metal-evenly over the surface.

8. The method of producing tungsten wire of uniform cross section which consists in removing from the surface thereof, carbon, arbides, oxides and other impurities accumulated during the swaging'operation. eliminating hydrogen therefrom and passing the wire through a die to effect a redistribution of the metal therein.

9. The method of lowering the rate of evaporation of a tungsten filament which consists in removing carbon and metallic carbide from the surface thereof, during the early part of the drawing operation and again removing from the finished filament, the carbon and carbides accumulated during the subsequent drawing operation.

In testimony whereof, I have hereunto subscribed my name this 29th day of October 1925.

MALCOLM N. RICH.