US1826524A - Treatment of metallic filaments - Google Patents

Description

Oct. 6, 1931. J. H. RAMAGE 1,826,524

TREATMENT OF METALLIC FILAMEN'IS Filed Nov. 10, 1927 INVENTOR JbHN HE GE.

ATTOR EY Patented 0st. 6, 193i UNITED STATES PATENT orl-"lcr.

JOHN H. BAKAGE, OF BLOOHFIELD, NEW JERSEY, ASSIGNOR TO WESTINGHOUSE LAD COMPANY, A CORPORATION OF PENNSYLVANIA TREATMENT OF METALLIC FILAHENTB Application filed November 10, 1827. Serial No. 232,800.

This invention relates to the treatment of metallic filaments for incandescent lamps, electron discharge devices and similar devices and more particularly to the heat treatment of tungsten filaments of the non-sag character, such as produced by the process set forth -in Hall Patent No. 1,461,117 granted July 10, 1923, entitled'Filamentary body for electric lamps and method of making the same, although it may e applied to other types of refractory meta wires.

The life of a refractory metal filament is limited by alterations which take place in the filament during operation, such alteration comprising rearrangement and crystalline growth within the filament and eva ration of the particles of metal from the lament. Such internal changes render the filament brittle and cause it to become deformed, that is, to sag or offset. 1

One of the objects of this invention is to impart to refractory metal wires an internal structure which resists alteration during burning and which, therefore, materially adds to the operating lifethereof.

Another object is to provide a heat treatment fbr refractory metal filaments which will increase the crystal size many times 'and which will very materially increase the nonsag property of so-called non-sag wire.

A further object is to produce refractory metal filaments having an increased strength particularly when burned under vibratory conditions and which will have a long life.

A further object is to reduce the rate of evaporation of refractory metal filaments.

A still further object is to develop a more uniform crystal structure in a refractory metal wire.

Othe v objects and advantages will hereinafter appear.

If a definite internal crystalline structure is imparted to a refractory inetal wire before the filament is operated at or above the critical recrystallization temperature, further internal alteration is restrained and the strength and life of the material is increased.

In acordance with the present invention, 1 produce such'internal structure by subjecting the filament after it is placed in the envelope, to a form of heat treatment which Lhavediscovered has the desired beneficial effect upon the filament. This heat treatment may take the place of the usual flashing or seasoning 0 eration and consists in initially heating tfie wire in its strained drawn condition to temperatures which cause recrystallization. The heating is effected after mounting of the filament in the evacuated or gas filled device and is commenced at a temperature well below the recrystallization temperature. The temperature is then raised at a uniform slow rate to a point several hundred degrees above the maximum temperature at which the wire is to be heated in operation.

In metals, the grain structure of which has been deformed by mechanical working, such as drawn tungsten wire, there is a certain temperature, known as the critical or efquiaxing temperature, at which the metal assumes an unstressed or natural cell-like structure, known as the equiaxed condition. Immedi= ately above this equiaxing temperature there vis a temperature zone in which the enlarging or growth of the grains composing the metal structure is extremely rapid. This recrystallization temperature varies in different t as of wire depending upon the kind and purlty of the materials comprisin the same, the size thereof, etc., but genera ly speaking, recrystallization may start at about 1100 to 1400 C. for tungsten filaments of the non- Sag type" I have found that if the 1mt1al heating of the drawn wire is commenced at a tempera ture below this recrystallization temperature and slowly and uniformly raised therethrough that a crystal structure is developed in which the crystals are of very large size and which retain-their interlocking or non-sag character. The slow heat treatment gives the crystals sufiicient time to grow enormously in the zone of rapid grain growth and the uniform increase in temperature produces a temperature gradient between adjacent grains extending from the point or points of highest temperature to the point or points of lower temperature. Due to the loss of heat by conduction at the filament supports or leads,

there is always a point on the filament between such sulpports or leads where the filament attains t highest temperature. As the temperature of the filament is slowly raised from a temperature below the recrystallization temperature these points of highest temratureare the, first to reach the condition avorable for grain growth and due to the temperature gradient existing between the grams at these points and the adjacent grains,

there is an absorption of the cooler grains tact with another crystal of such size that it cannot be absorbed.

The rate of increase of temperature der upon the size of the wire being treated a: in general a shorter time than 3 minutes does not give the satisfactory results even with the smallest sizes of filaments when mounted in lam s as coiled filaments and for sizes above 3 mil diameter the time required to raise the temperature from below the recrystallization temperature to a point well above the normal operating temperature should be increased at the rate of about 1 minute per mil of increased diameter.

For example, in the case of a 50 watt, 115 volt tungsten filament, which normally burns at a current of .435 amperes, the initial lighting current was about .087 amperes and by means of an adjustable resistance the current was raised at a uniform rate over a period of 3 minutes to about .582 amperes. Lamps having their filament treated according to this schedule had a life under vibratory operation conditions of twice that of lamps flashed according to the schedule heretofore employed'. The schedule heretofore employed consisted of a short flash at 100 and 130% of normal operating voltage.

As an example of the treatin schedule employed on large wire, a goo interlocking crystal of large size was obtained in a 25 mil wire operating at a normal current of 24 and peres, y raising the current from 2 amperes to 26 amperes in about 2% minutes, and in the case of a 40 mil wire, normally operating at 52 amperes, by raising the current from am'peres to 58 amperes in 30 minutes, although good results were obtained in this last instance with a minute schedule.

The efi'ects of this initial slow uniform increase in temperature appear to be to develop a. larger grain structure and a more uniform od crystal structure having interlocking grain boundaries which are more resistant to offsetting and furthermore, to the development of this structure in the arts of the fila mentwhich are ordinarily chllled, i. e., at the.

lca'dconnections and across the filament supports. The high finishing temperature promotes the last feature and is not necessary when no cooled parts are present in the span to be treated. However, to prevent further crystal alteration during the operation of the filament, it shouldbe heated a temperature above the maximum temperature to which it will be subsequently subjeizted during operation.

In order to insure that the heat treatment will commence well below the critical recrystallization temperature and continue well above the. maximum temperature which the filament will attain during use, i prefer to start the treament at a temperature of about 800 C. and to continue the treatment up to a temperature several hundred'degrees above the normal operating temperature of the filament. The maximum treating temperature may vary between 2200 C. sin "00 C. depending upon the use to'which the filament is to be subjected.

- The treating current must be applied continuously and without interrupt on until a. temperature is obtained sufllcientl; high to carry the crystallization around the cold portions of the filament supports. An intermittent schedule does not produce the same results in crystal formation, since when the filament is heated by intermittent steps, the crystallization takes place at the step at which the temperature is within the critical grain growth range and immediatel ceases again when the current is cut ofi in switching to the next step and the filament momentarily cools. v The subsequent steps have little effect as the grain growth and temperature gradients are interrupted at each step change.

v In order that the effect of the heat treatment herein described can be more fully understood, reference will be had to the accompanying drawings in which:

Fig. 1 is a longitudinal sectional view of a recrystallized non-sag filament flashed by the method heretofore employed and showing the usual crystal structure the-rot; and,

Fig. 2 is a longitudinal sectional view of a recrystallized non-sag wire treetri in accordance with my invention and shrwing the exaggerated crystal structure obtained thereby.

ferring to Fig. 1, it will be noted that the crystal structure comprises a lar e number of very fine grains 10. The s" of the grains can be better appreciated mean it is understood that the figures represent photomicrographs of a mil wire taker. eta magnification of about 75 diametem. ibs crystals shown at 11 in Figure 2 il iterate the growth that has taken place during the slowcontinuous increase in temperature which is described above. It should be noted that many of the crystals 11 have an extent of substantially-half the diameter of the wire or about 10 mil in diameter and many times the diameter. In many instances, I have obtained single 0 stals .extending over the entire diameter 0 the wire, and in larger size wires of as high as 40 mil or greater.

Referring to Figure 1', the dark portions 12 shown extending alon each side of the wire are not a representation of single crys tals but illustrate the reced' g curved por-' tion or rounded edge ofthe wire, as it ap pears in the original photo-micrograph.

It is vunderstood that many changes may be made in the treatment and that it may be applied to various t pes of wire and I do not desire to be .limite to the exact details described but contemplate such modifications as come within the scope of the invention as defined in the appended claims.

What is claimed is:

1. The method of treating tungsten wire to promote the development of large crystals comprising initially subjecting the drawn wire to a temperature about 800 C) and increasing the temperature at' a slow'rate up to a point above the normaloperating temperature.

2. The method of treating non-sag tungsten wire to increase the crystal size and augmentthe non-sag property of the wire comprising elevating the temperature of the wire frombelow 1100 C, to above 2200? C. at a substantiallyvunifonn rate,over a period of approximately one minute per mil diameter of the wire.

I In testimony whereof, I have hereunto subscribed my name this 9th day of November, 1927. a JOHNHrRAMAGEr.

a length of

Images