US1862365A - Magnetic alloys and method of making the same - Google Patents

Description

it the same.

Patented June 7, 1932 PATENT OFFICE WILLIAM E. RUDER, OF SCHENECTADY, NEW YORK, ASSIGNOR TO GENERAL ELECTRIC COMPANY, A CORPORATION OF NEW YORK MAGNETIC .ALLOYS AND METHOD OF MAKING- THE SAME No Drawing. Application filed July 6,

The present invention relates to magnetic material and more particularly to iron-nickel alloys which are characterized by a very low total watt loss and to a method for making Iron-nickel alloys have been known for some time and have been used as transformer core material and as loading coils in telephone transmission circuits.

These alloys when employed under such C11' e1 cumstances are known to have a relatively low total watt loss. So far as I am aware, however, it has been impossible prior to the present invention to obtain magnetic material of any composition and of normal thick- 515 ness, such as 14 mils, which, when used as the core of a transformer, has a total watt loss as low as .4 watts per pound when tested at B=10,000 gausses and at 60 cycles.

It is one of the objects of the present invention to provide an improved iron-nickel tion furnace and in a reducing or non-oxidizing atmos here, such as hydrogen, or in a vacuum. he resulting ingot is rolled into sheets of-a desired thickness and annealed for about two. hours in a reducing or non-oxidizing atmosphere such as hydrogen, or in a vacuum and at a temperature of about 1300 C. The alloy produced by this process is characterized by an unusually low total watt loss. For example, in sheets of normal thickness, as 14 mils, the material has atotal watt loss substantially less than .4 watts per pound and a hysteresis loss substantially lower than .23 watts per pound at 10,000 B. and 60 cycles. This may be compared with the best silicon steel which, under the same conditions, has a total watt loss varying from .45 to .55 watts per pound and with the best nickel-iron alloys heretofore made, the latter. having a total watt loss slightly over .4

1928. Serial No. 290,889.

watts per pound and a hysteresis loss greater than .23 watts per pound. Heretofore it has been usual to melt nickellron alloys in an arc furnace. I have found however that if such alloys are melted in a furnace of this type they usually contain a small amount of nitrogen and as the amount of nitrogen present in the alloy increases there is a corresponding increase in the total Watt loss obtained. It will therefore be found, generally, that the most satisfactory results are obtained if the melting operation is carried out in an induction furnace and 1n a reducing or non-oxidizing atmosphere, such as hydrogen, or in a vacuum. Under these conditions the presence of nitrogen, and more especially the nascent nitrogen which occurs in the operation of an arc furnace, is avoided and the magnetic characteristics of the alloy are substantially improved as compared with the alloy melted in the arc furnace.

The ingot obtained from the melting furnace may be rolled hot or cold into sheets of the desired thickness. Good results have been obtained with both rolling processes. While the best silicon steel sheets cannot be rolled to a thickness of less than approximately 14 mils without materially increasing the total watt loss per, pound due to the rapid increase in the hysteresis loss in such thin sheets, I have found that, with the present material, sheets having a thickness of about 6 mils may be easily rolled and that as the thickness of the sheets decreases the total watt loss likewise decreases. For example, with a sheet 14 mils thick the total watt loss was about .36 watts per pound for 10,000 B. at 60 cycles. With sheets 13 mils, 9 mils and 7 mils in thickness the total watt losses under the same conditions were respectively .335, .285 and .272 watts per pound. While it is my opinion that sheets as thin as 1 mil may be rolled and that such sheets for certain frequencies, as for example frequencies substantially above 60 cycles and radio frequencies, may be used advantageously, such a thickness could not be produced without considerably increasing the cost of the alloy. Sheets about 10 to 12 mils in thickness give a very satisfactory low watt loss and this thickness may be rolled and handled economically.

After rolling the alloy to the thickness desired, it is annealed in a reducing or nonoxygenous atmosphere, such as hydrogen, or in a vacuum, and at an elevated temperature, the annealing period varying from about two to about ten hours. The anneallng temperature employed aflects the subsequent electrical characteristics of the alloy and for the best results the temperature employed should not be materially lower than 1200 C. A temperature of about 1300 C. is very satisfactory. The temperature should not vary greatly from this latter figure. If the annealing temperature is materially below 1200 (3.; it will be found that the total watt loss per pound is much higher than the total watt loss obtained when the alloy is annealed for the same period of time at 1300 C.

While the annealing period will usually var from 2 to 10 hours it may be longer if esired since the increased heating period does not appear to have. any harmful effect on the total watt loss, in fact, lower losses may be obtained by increasing the time, particularly at the lower temperature range, due to the continued purifying action of the hydrogen. For example, if the sheets are annealed in hydrogen at 1200 C. for 24 hours the total watt loss obtained approaches close to but does not equal that obtained when the sheets are annealed in hydrogen for two hours at 1300 C.

I have found that an alloy having a nickel content of about 47% and an iron content of about 53% when annealed at about 1300 C. for two hours gives a very low total watt loss per pound. Although an alloy having a nickel content of about 51% when annealed in the same manner will also give a very satisfactory total low watt loss, it will be found that the alloy containing 47% nickel will give a substantially lower watt loss per pound than the 51% alloy when tested at 10,000 B. and cycles.

It is my opinion at the present time that the best and most economical range of the nickel content which may be employed in the alloy lies between about 40% and about 50% for I have observed that the lowest hysteresis losses occur with a nickel content varying from about 35% to about while the maximum resistance (80 microns per Cm) is obtained with a nickel content of 35% so that the best combination for low hysteresis and low eddy current losses will lie somewherein the range between 35 to nickel. The resistance of the alloy decreases very rapidly as the nickel content is increased above 35%, and as the minimum hysteresis loss occurs with a nickel content varying from about 40% to about 60% the composition having the lowest total Watt loss in any given thickness of sheet will be obtained by properly annealing an alloy having a nickel content varying from about 40% to about 50% nickel. Between these points the lowest hysteresis losses are obtained and at the same time a relatively high resistance, and, there- .fore, low eddy current loss and low total watt loss. A lower nickel content in the alloy will provide a higher resistance and therefore a lower eddy current loss but at the same time the hysteresis loss will increase so that the total watt loss will be higher than that obtained in the range from 40 to 50% nickel. Likewise, for a nickel content higher than 50% the hysteresis loss will be low but the resistance will likewise be so low that the total watt loss will be higher than that of the alloy with 40 to 50% nickel.

I have found that most of the nickel-iron alloys containin from about 40 to about 50% nickel when annealed in hydrogen at a temperature above 1100 C. have a comparative- 'ly large grain size and the structure of the alloy shows a large number of rectilinear twins that resemble plagioclase. While this structure itself is not indicative that, in sheets 14 mils thick, the total watt loss of the alloy will be applicable lower than .4 watts per pound, it is true that for the best magnetic conditions the nickel-iron alloy will exhibit this structure. In order therefore to obtain the best magnetic results, i. e., lowest total watt loss and highest permeability, the alloy should be annealed so as to give a homogeneous structure and a comparatively large grain size (about 10 to 40 grains per mm") with a strong tendency to twinning.

Since nickel is the expensive element in the alloy, it is very desirable to keep the nickel content as low as possible consistent with a minimum h steresis loss. At the same time it is desira le to employ a nickel content which will improve the total watt loss by reducing' the eddy current loss without the necessity of incurring additioual expense by rolling the material into very thin sheets. An alloy sheet of normal thickness (14.- mils) and having a nickel content varying from 40 to 50% not only provides a satisfactory low total watt loss and a low hysteresis loss but at the same time provides a maximum saturation value and so allows high working densities which 'is an important consideration in any power apparatus.

Alloy sheets prepared in the above manner and having a thickness of about 6 to 7 mils and containing about 47% nickel give a total watt loss of about .272 per pound for 10,000 B. at 60 cycles. I prefer, however, to employ sheets having a thickness of about 10 to 12 mils, since sheets of this size may be rolled and handled more economically than the thinner sheets and give a total watt loss which is Well under .4 watts per pound and a hysteresis loss less than .23 watts per pound.

For example, a sheet having a thickness of about 10 mils has a total watt loss of less than .290 watts.

While the initial permeability of the present alloy is not as high as certain other nickel alloys, it is relatively high compared with the initial permeability of the best silicon steel. For example, the above alloy containing 47% nickel and annealed at 1300 C. gave an initial permeability of about 6000 which is about three to four times as high as the initial permeabilit of silicon steel. The term initial permea ility is one generally employed in connection with magnetic material and may be defined as that permeability which would be shown for zero magnetizing forces and is obtained by plotting a series of values of permeability obtained by magnetic measurement of the material at exceedingly low magnetiz'n forces ofthe order of H=.0l to .02 GGr units and drawing through these plotted points a curve which will intersect the H=0 axis of the diagram.

While the cost of the present alloy is in excess of the cost of the best silicon steel,

. which has heretofore been used almost exclusively in the construction of the ordinary distribution transformers, it will be found advantageous to employ the present alloy in such transformers since the latter carry an all-day load and the increase in eficiency in the operation of such transformers obtained by the use of my improved alloy more thaln ofisets the increased cost of the materia What I claim as new and desire to secure by Letters Patent of the United States is:

1. The method of maki a nickel-iron alloy which comprises melting the materials comprising, the alloy, rolling the alloy into sheets and annealing the sheets at a temperature not materially less than 1200 C.

2. The method of making a nickel-iron al loy which comprises melting the materials comprising the alloy, rolling the alloy into sheets and annealing at a temperature oi about 1300 C.

3. The method of making an alloy which comprises melting about to 50% nickel with about to 50%. iron, rolling the resulting alloy into sheets and annealingthe sheets at a temperature not materially less than 1200 C.

4. The method of making an alloy which comprises melting about 40% to about 50% nickel with'about 60% to about 50% iron, rolling the resulting alloy into sheets and annealing them in a hydrogen atmosphere at a temperature not materially less than 1200 C.

5. The method of making a nickel-iron alloy which comprises melting the materials comprisin the alloy' in a reducing atmosphere, rol ingthe alloy into sheets and annealing the sheets under non-oxygenous conditions and at a temperature about 1300 C.

6. The method of making a nickel-iron alloy which comprises melting the materials comprising the alloy, rolling the alloy into sheets and annealing the sheets under nonoxygenousconditions at a temperature in the neighborhood of 1300 C.

' 7. The method of making a nickel-iron alloy which comprises melting the materials comprising the alloy, rolling the alloy into sheets and annealing the sheets in a hydrogen atmosphere and at a temperature in the neighborhool of 1300 C.

8. The method of making a nickel-iron alloy which comprises melting the materials comprising the alloy in an atmosphere substantially free from nitrogen, rolling the alloy into sheets and annealing the sheets at a temperature not materially less than 1200 C.

9. The method of making anickel-iron alloy which comprises melting the material comprising the alloy in an atmosphere substantially free from nascent nitrogen, rolling the alloy into sheets and annealin the sheets at a temperature of about 1300 10. A nickel-iron alloy in sheet form having a nickel content varying from about 40% to about 50%, the remainder of said alloy consisting substantially of iron, said alloy having, in sheets about 14 mils in thickness, a total watt loss less than .4 watts per pound when tested at 10,000 gausses and 60 cycles.

11. A nickel-iron alloy in sheet form annealed at a temperature not materially lower than l200 G, said alloy when in sheets about 14 mils thick having a total watt loss less than .4 watts per pound when tested at 10,000 guasses and 60 cycles. I

12. An alloy in sheet form annealed at a temperature not materially less than 1200 (3., said alloy containing from about 40 to about 50% nickel, the remainder of the alloy being iron except for incidental impurities, said alloy when in sheets about 14 mils or less in thickness having a total watt loss less than .4 watts per pound when tested at 10,000 gausses and 60 cycles.

13. A nickel-iron alloy containing from about 40 to 50% nickel and from about 60 to 50% iron,\said alloy being in sheet form having magnetic properties substantially identical with the magnetic properties of a nickeliron alloy of the same composition in a sheet of substantially the same thickness which has been annealed at about 1300 C. for at least two hours in a hydrogen atmosphere.

14. A nickel-iron alloy containing from about 40 to 50% nickel and from about 60 to 50% iron, said alloy being in sheet form having magnetic properties substantially identical with the magnetic properties of a nickel-iron alloy of the same composition which has been melted in 'an atmosphere substantially free from nitrogen, rolled into a sheet of substantially the same thickness and annealed at about 1300 C. for at least two hours in a hydrogen atmosphere.

15. A heat-treated nickel-iron alloy containing from about to nickel and from about to 50% iron, said alloy having a homogeneous structure and comparatively large grain size.

16. A heat-treated nickel-iron alloy in sheet-form containing from about 40 to 50% nickel and from about 60 to 50% iron, sa1d alloy having a homogeneous structure and comparatively lar e grain size, and in sheets about 14 mils thic having a total watt loss less than .4 watts per pound when tested at 10,000 gausses and 60 cycles.

In witness whereof, I have hereunto set my hand this 5th day of July, 1928.

WILLIAM E. RUDER.