US2156019A - Permanent magnet steel alloy and method of making same - Google Patents

Description

G. B. JONAS 2,156,019

PERMANENT MAGNET STEEL ALLOY AND METHOD OF MAKING SAME Patented `Apr. A25, 1939 VUNITED STATES a l 2,156,019 PATENT 1 OFFICE PERMANENT MAGNET STEEL ALLOY METHOD OF MAKING SAME .Gottfried Bruno Jonas, Eindhoven, Netherlands,

assignor to N. V. Philips Gloeilampenfabrieken, Eindhoven', Netherlands Application February 10, 1936, Serial No. 63,236

In Germany March 5, 1935 2 Claims.

My'invention relates to alloy steelsfor permanentv magnets and more particularly to al'- loy steels containing nickel and aluminium.

It has been proposed, for instance inthe UL S. Patent #2,027,994 to Mishima,fto make permanent magnets from an Al-Ni-Fe alloy without qenching or annealing. Althoughv this elimina'- tion of special heat treatment simplified and re'- duced the cost of manufacture, I have found that for alloys containing to 25% Ni and 7% to 15% Al, the resulting magnets had'poor magnetic propertiesunless they were subjectedl to special heat treatments. f

The object of my invention is to improve the magnetic properties of such alloys withoutthe use of special heat treatment, and for this purpose I add to the alloy between about. 0.5 to 5% Ti, the actual quantity of Ti being dependent upon the dimensions of the casting.

In order that Vthe invention'can be clearly understood and readily carried into effect, I shall describe the vsame in more detail in comparison with the prior art and with reference to the accompanying drawing, in which Figures 1 to 3 are graphs giving the magnetic curves of permanent'm'agnets according to the invention. It is knownfto make permanent magnets from steel alloys containing nickel and aluminium, such steels being described for'instance, in the FrenchPatent No. 731,361, as weilas in an article byKotaro Honda in theperiodical Metallwirtschaft, Wissenschaft und Technik (Number 24 of June 15, 1934). The magnet steels described in the French patent are made from iron to which from about 5% Ito 40% of nickel and from 1 to 20% of aluminium are added, with the possible addition of cobmt, manganese, tungsten, molybdenum, vanadium or copper.

The French patent states that such nickelaluminium steels have very favourable magneti'cproperties without ybeing hardened, 'and thus have the advantage of being usable in their cast vstate without requiring any hardening by thermal treatment. Furthermore, from the figf ures given regarding the magnetic properties of these alloys it wouldappear that this patent givesthe remanence of all of these Ai-Ni steels as exceeding '9,000' gauss, whereas all of Athe maximum values of their coercive force ex-v ceeds 400 gauss. Y'

However, in practice it has been found that for the steels of such composition, such* high values could not be even approximated without hardening, land even with the aid of a heat treatment these values could not be obtained. Extensive tests have conclusively proven that the statement that such Ni-Al alloys have such excellent magnetic properties in their cast state, without subsequent heat treatment, is erroneous in all cases, or at least in the greater number of possible cases.

The patent neglects the fact that for a cast'- ing made of a steel of a definite composition, the size of the casting as well as the thermal properties of the mould are of great importance with respect to the resulting magnetic proper'- ties of the steel because the cooling speed hasa determinative influence on same.

I have found that when making magnets of the size used for example for magnets of dynamic speakers, from such Al-Ni alloy steels without hardening, the magnetic properties thereof are as a rule inferior to those made of the well-known cobalt steels and that only when subjected to a' definite thermal treatment, do Athe Ni-Fe-Al alloys show marked superiority over cobalt steels.

The above will be illustrated by a series of tests, made on test pieces of Ni-Fe-Al alloys having a length of 33 mm., a width of 27 mm. and a thickness of 10 mm.'; these pieces having been Acut from cast bars of a length .of 150 mm., a width of 27 mm. and a thickness of l0 The dimensions of the test pieces are even more favorable for reaching high magnetic values than are the dimensions of most magnets used in practice.

Test I Thisvtest was made on a piece of the above size having a composition .of 25% nickel, 11% aluminium, and the remainder iron. In its cast state and without thermal treatment, this piece had a remanence of 5200 gauss and a coercive force of 325 gauss. After being subjected to thermal treatment by hardening from a temperature 4 of 1200 C. in an oil bath at roomtemperature, and subsequent annealing at a temperature of from 650 C.' to '150 C., the test piece had a .remanence of 7200 gauss and' a coercive force of '400 gauss.

Test II A test piece of the above size and having a vcomposition of 27% nickel, 12% aluminium and A test piece of the above size and having a.4

composition of 30% nickel, 14% aluminium'. and

the remainder iron, in its cast state and without thermal treatment, had a remanence of 3600 gauss and a coercive force of 450 gauss. After being subjected to the same thermal treatment as given in Test I, this piece had a remanence of 5600- gauss and a coercive force of 620 gauss. Even if in certain special individual cases which seldom occur in practice, for instance when using very thin-walled castings, mould materials of a very high thermal conductivity, etc. higher values are obtained without thermal treatment, it appears that the magnetic values given in the French specification cannot be obtained without a special heat treatment. In any case, the French patent does not teach one skilled inthe art how to obtain am] appreciable advantage by using the Al-Ni-Fe steels.

'I'he object of my invention is to provide a magnet steel alloy which is much less sensitive to the cooling speed and which, in its cast state, has much better magnetic properties than the castings cast under normal conditions from the alloys described in the French patent. Furthermore, castings of the alloy according to the in- Vventlon without being heat-treated have magnetic properties which closely approximate the magnetic properties of hardened castings made from the alloy of the French patent.

.According to thefinvention, I addrto an alloy containing iron as a base material, and in addition about 15% to 25% nickel and about 7% to 15% aluminium, a quantity of titanium-usually from .5% to %-which depends on the size Percent Nickel 22 Aluminium 11 Titanium 3 Iron. Remainder Magnets made from alloys of the above composition have a remanence of about 6400 gauss and a coercive force of about 500 gauss; for magnet sizes having a ratio between their surface area in sq. mms. and their volume in cu. mms. of about 1 to 3%. 'Ihe size of the magnet is specified because the remanence and coercive force are dependent thereon, and if the above ratio between the surface area and volume is changed, the values of remanence and coercive force will also change to some extent.

'I'he magnetic properties of magnets made according to the invention are indicated in the accompanying drawing, `in which Figures 1 to 3 are magnetization curves.

The results which can in practice be obtained by the present invention will be better understood by giving the following example from loudspeaker practice. Magnet rings for moving coil loudspeakers. for instance, were made which were given the following dimensions:

i Millimeters External diameter 72 Internal diameter 49 Height 22 netisation and gave the following magnetic values: v

Remanence Bi- 3400 Coercive force Hc 390 After a suitable heat treatment (quenching from a temperature of 12001300 C. inan oil bath to room temperature, and subsequently annealing for about half an hour at about 675 C.) and subsequent magnetisation the following values were obtained:

Br 6000 H.- '540 Thus it appears from the last-mentioned example of the invention that without using a special heat treatment, magnetic values at B1- 8150, Hc 540 can be obtained. These are of practical use in loudspeaker construction, and are furthermore'of the same order of magnitude as those values (Br:6000, Hc 540), which according to the previously stated example, are obtained only by heat treatment.

In the Figures 1 to 3 the abscissae represent coercive force in gauss and the ordinates repre'- sent remanence in gauss.

Referring to Figure 1, the magnetization curves i, 2, 3 and 4 represent the test results obtained on four pieces, all of the pieces being in their cast state without heat treatment. 'Ihe difference found between the four curves is principally due to the fact that the constituents of the alloy were added in different order and to some extent in different forms when making the castings for the test pieces.

For instance, titanium is added inits pure state, but preferably it forms a constituent of a commercial ferro-alloy which may have the following composition:

The use of titanium in this form is preferable because it is cheaper than using titanium in its pure state.`

It has been found that the other constituents of the mentioned ferro-titanium alloy are not of great importance for the magnetic propertiesl.

of the magnetic alloy. The improvedvresults obtained by using the ferro-titanium alloy instead of pure titanium are probably due to the fact that smaller consumption losses occur during the alloying and thus the desired composition of the magnetic alloy can be more accurately obtained. Furthermore, it has been found advantageous to add thevaluminium last, because somewhat improved magnetic properties may be obtained in this way.

Fromcurves I tolofFigure litwill benoted is, as a ru1e, also Vimpossim@ as it usuany results that the remanences of the test pieces range between about 5700 and 6400, whereas the coercive forces range between about 420 and 520.

'I'he curvesv I', 2', 3' and 4' of Fig. 2 wereobtained with the test pieces, which without-heattreatment gaveicurves I, 2, 3, and l respectively, after these test pieces had been annealed for one hour at a temperature of between 650 to 675 C. Comparison ofthe curves of Figs. 1

vand 2 shows practically no diierence in the magnetic properties oi' the test pieces before and after their annealing.

'I'hat annealing has practically no effect more clearly appears from Fig. 3 in which curve A represents the average of curves I, 2, 3, and 4 of Fig. l, whereas curve B represents the average of curves I', 2', 3', and 4' of Fig. 2. Comparison of curves A and B shows that the annealing of the test pieces has no appreciable effect upon either the coercive force or the magnitude BXHmax. Furthermore, annealing only slightly increases lthe 'remanence of the test pieces above that o! the cast pieces before heat treatment.

The Nl-Al-Ti-Fe alloys should not be subjected at all to the normal hardening operation, as this greatly deteriorates their magnetic properties. For instance, when the test piece corresponding to curves- I and I' was subjected to ,i a hardening and an annealing process, (quenching in oil or water, from 1150-1300 C. to room temperature and annealing to about 650-675 'C.) its remanence was reduced to about 5000 in fracture or cracking of same.

y Reduction of the average titanium content slightly increases the remanence but reduces the coercive force to such an extent asv to produce a steel of inferior magnetic properties. Also too small a titanium contentrenders the alloy more sensitive to thermal treatment. On the other hand, increasing the titanium content above 3% is objectionable, in that it renders the casting so brittlev'as to cause a certain percentage of waste by breakage.

Zirconium and antimonium have similar ef? 1. An unquenchedA cast permanent magnet consisting substantially-of 22% Ni, 11% A1, 3% Ti, and the remainder substantially Fe.

2. A method 4of making a permanent magnet casting comprisingforming an alloy consisting srmstantiany or 2,2% Ni. 11% A1, 3% 'n.and the an remainder substantially iron, casting the alloy inamold, andheat-treatingthesolely byslowly coolingthe samein themold.

GOITERIED BRUNO JONAS.

Images