US3234609A - Method of making magnetically anisotropic permanent magnets - Google Patents

Description

Feb. 15, 1966 OSAMU MADONO 3,234,609

METHOD OF MAKING MAGNETIGALLY ANISOTROPIC PERMANENT MAGNETS Filed Sept. 26, 1963 Fig. 3

IN VEN TOR.

WE-N ERQT LIND m PoNAcK ATTORNEY5 United States Patent C) 3 234,609 METHOD OF MA KING MAGNETICALLY ANISOTROPIC PERMANENT MAGNETS Osamu Madono, Faitama, Japan, assignor to Riken Piston Ring Industrial Co., Ltd., Tokyo, Japan Filed Sept. 26, 1963, Ser. No. 313,407 Claims priority, application Japan, Jan. 28, 1960, 35/2,301 1 Claim. (Cl. 22211) The present application is a continuation-in-part of application Serial No. 82,515, filed January 13, 1961, now abandoned.

The present invention relates to a novel method of making permanent magnets of anisotropic Al-Ni-Co-Cu type which have a columnar crystalline structure in perfectly parallel alignment.

As is well known, the cast alloy of Alnico type is coarse grained, extremely hard, and brittle. It is, therefore, cast in the final shape to the closest tolerance. possible, and is finished by grinding and then subjected to a heat treatment in a magnetic field. It is also known that a considerable improvement in the magnetic properties of a permanent magnet is attained if the columnar crystals formed during solidification are aligned parallel to the direction of final magnetization. A number of methods for producing such anisotropic magnets have been proposed but none of them has been considered to he satisfactory in a commercial sense. In the conventional methods columnar crystalline structure has been produced by chill casting. Chills promote directional cooling to result in the growth of columnar crystals. It is, however, noted that, though a considerable improvement in magnetic properties has been obtained by conventional chilling methods, there is a limitation in the improvement so far as it depends upon chill effect alone. That is, energy products of magnets which have been thus obtained cannot be over 6.5 gauss-oersteds but less than 6x10 is more representative of the commercial products. As compared with these values, higher energy product values were sometimes reported by some researchers. However, none of these processes can be used commercially because of their low productivity and high cost. Therefore, a commercial method capable of producing permanent magnets with perfectly aligned columnar crystals is highly desirable because of the excellent magnetic properties.

I have found that a permanent magnet having a columnar crystalline structure in perfectly parallel alignment can be obtained by a novel method in which the casting is made in a characteristic multi-cavity mold wherein the chill efiect can be controlled and directional solidification promoted by making the length to diameter ratio of the casting at least 6: 1.

In order that the invention may be clearly understood and readily carried into effect, it will now be described more fully with reference to the accompanying drawing in which:

FIG. 1 is a sectional view of a solidified casting within a mold having a chill. The chill, per se, is not part of this invention, and can be any of a number of materials well known in the art for that purpose. For example, the chill plate may be of steel or copper of a thickness and volume such that the heat capacity thereof is sufficient to effect the chilling of the melt during solidification thereof, such as shown in Patent No. 2,380,616 to Snoek et al., and the chill plate may preferably be cooled by water as shown in the Japanese Patent Gazette of Sho 34/9572 issued on October 27, 1959. The solidified casting and mold consists of the mold wall 1, the casting 2, the riser 3, the chill 4 and the columnar 3,234,609 Patented Feb. 15, 1966 crystals perpendicular to the chill surface 5. When liquid metal is poured into this mold, heat is lost in three ways: By thermal conduction into the walls 1, by radiation and convection losses to the atmosphere from the top of the riser 3, and by thermal conduction into the chill 4 from the bottom. In order to enhance the direc tional solidification in a direction parallel to the axis of the cylinder heat loss by conduction into the mold walls must be prevented. When the mold cavity is filled with liquid metal, heat flows into the mold walls by thermal conduction, which rapidly increases the temperature of the wall surfaces. The surfaces are soon saturated with heat and their temperature becomes equal to that of the metal. Moreover, even after the surfaces are saturated with heat, heat will continue to flow from the surface into the inside if there is a temperature gradient. Therefore, in order to prevent this heat loss, the mold walls must be saturated with heat at both the surface and inward.

In the present invention I have found that a mold having a multi-cavity construction is the most advantageous for heat saturation. FIG. 2 is a plane view of the multicavity mold which consists of the cavities 6, the mold walls 7, and the insulating mold frame 8. This mold has a great number of cavities closely formed side by side within the small body of the mold so that the volume ratio of the cavities to the mold can be made as large as possible. The larger the ratio of the cavities to the mold, the easier heat saturation occurs. Hence it is necessary that the mold wa-lls should be made as thin as possible. For example, when the cylindrical castings are 10 mm. in diameter, the cavity to mold ratio must be more than 50% and the minimum thickness of the walls between the cavities has to be about 2 mm.

However, if liquid metal is poured into a cold mold, growth of the columnar crystals is not favorable even though the mold has the said construction. The cold mold cools the liquid metal so rapidly that the temperature of mold cannot attain to the melting point of the metal after thermal equilibrium is established. Therefore, in order to increase the equilibrium temperature, preheating of mold at a high temperature is essential. I have found that, if the volume ratio is 50%, the mold must be preliminarily heated at least at 1000 C., and that the higher the temperature of pre-heating, the more favorably the columnar crystalline structure is developed. The preheating temperature may be lowered a little if the volume ratio of cavities is larger than 50%. For example, when the volume ratio is 70%, the pro-heating temperature may be lowered to 800 C.

There is also another factor which has an influence upon directional solidification. That is, chill effect is important for the growth of columnar crystals. Generally, when liquid metal is poured into a mold provided with a chill, it cools rapidly until nucleation of crystals begins to occur at the chill zone. These first-formed crystals do not have favorable orientations. That is, the chill effect is too strong to give suflicient time for the growth favorable oriented crystals to have enough time for growing. It is due to this fact that permanent magnets made by the conventional chilling methods, i.e. the length to diameter ratio is less than 5:1, contain a considerable amount of unfavorably oriented crystals, and that, in point of alignment of columnar crystals, some are parallel but some are deviated from the parallel alignment. In contrast thereto, if chill effects are controlled and cooling takes place slowly, growth of crystals occurs preferentially in the direction of of body-centered cubic system and the growth of unfavorably oriented crystals is suppressed. Deposition of new solid proceeds at an enhanced rate of the more favorably aligned crystals.

I have found that, for the purpose of controlling chill effects, it is necessary for the castings to have a length of diameter ratio of at least 6:1 and the greater the ratio the better. That is, the castings must be long enough for columnar crystals to completely grow no matter how long the finished magnets are. FIG. 3 is the section corresponding to the line AB of the multi-cavity mold shown in FIG. 2. FIG. 3 shows the cavities 6, the mold walls 7, the insulating mold frame 8, the chill 9, the riser 10, and the columnar crystals 11 grown in parallel alignment. In this mold the length of the castings is at least six times as long as their diameter. It should, further, be noted that, since the time needed for solidification to be completed depends upon the length of castings, it is desirable for castings to be as long as possible, so far as directional cooling takes place in a direction parallel to the axis of cylinder. For example, if castings are mm. in diameter, their length must be longer than 60 mm. Since the area of chill surface has less influence upon the length of columnar crystals, length of castings should always be as used for investment casting or ceramic molding is probably the best binder for this purpose.

The following example illustrates the presently preferred embodiment of the invention and the percentages are by weight.

An alloy containing 8%Al, 13.6% Ni, 24% Co, 3.2% Cu, 0.1% Si, with the balance being Fe was cast into a cylindrical form of constant diameter of 10 mm. with varying lengths. The molds were-made of a mixture of silimanite and silica sol solution and preheated at 1100 C. The sectional area of molds were 250x220 mm? and there were formed 390 cavities in each mold. The volume ratio of cavities to mold is 55%. The depth of risers was mm. and exothermic compounds such as a mixture of Al and Fe' O were usedto delay solidification. In order to prevent heat losses the outside of the mold was surrounded by an insulating frame. Liquid metal was poured into the molds at 1670 to 1700 C. The results are shown as follows:

Table 1 N0. of heat 1 2 3 4 5 6 long as possible no matter how large the diameter is. Practically, if the volume ratio of cavities is more than 65% and mold is pre-heated at a temperature higher than 1000 C., it is easy to obtain columnar crystals longer than 80 mm. notwithstanding change in diameter. Of course there is an upper limit to the length to diameter ratio and unfavorably oriented crystals are developed again, if the ratio is too great. Herein the upper limit of the length to diameter ratio depends upon the volume ratio, the temperature of preheating of mold, and the diameter. For example, the upper limit to 10 mm. diam- 'eter is around 20:1, while that to mm. diameter is about 10:1, so far as the temperature of pro-heating of mold is 1200 C.

As has been mentioned above, heat is not only lost to the chill 9 but also to the atmosphere from the top of the riser 10 by radiation and convection. Since the heat loss from the top of the riser takes place generally in a direction perpendicular to the surface of the riser 10, it will not disturb the directional solidification of the castings. However, if the riser is frozen before solidification of the castings is completed, it is not useful as a feeding head for compensating for shrinkage. Therefore, in order to obtain both longer columnar crystals and soundness of castings it is of fundamental importance to delay solidification in the riser thereby obtaining the necessary constant supply of liquid metal to the castings. The idea of maintaining the riser as a reservoir of completely liquid metal throughout the solidification of castings in effectively approached when a blind cover is used in place of open risers, or when exothermic compounds or other means of local heating are adapted to delay solidification in the riser.

Since the mold is subjected to extremely high temperatures, it should be made of refractory materials of excellent quality. For this purpose alumina, silimanite, zircon or magnesia is appropriate, while natural silica sand is inappropriate because of low heat-resistivity at temperatures above 1500 C. Further, attention should be paid to the binder materials used for molding refractories. In particular, difiiculties may arise in molding the thin walls of the cavities. Hence binder materials must carefully be selected in order that they be easy to mold and have high heat-resistivity. Silica solsolution which is commonly In this table the measurements of magnetic properties were made on the samples which were cut off from longer castings into the same length of 15 mm. Referring to this table it is interesting that the magnetic properties of permanent magnets were improved with increasing the length of castings if the volume ratio of cavities was constant. Though the length of the columnar crystals did not grow in proportion to the length of the castings, the energy products of the magnets always became larger as the length of castings was increased. That is, the maximum energy products of magnets was increased from the value of 6.5 10 gauss-oersteds to that of 7.5 X 10 gaussoersteds when the length to diameter ratio exceeded the critical value of 6:1. As was already mentioned, the value of 6.5 10 is the upper limit to the maximum energy product of anisotropic Alnico magnets made by the conventional chilling method. Accordingly, it is concluded that the 6:1 ratio of length to diameter shows a critical value needed for the magnets to be improved better than the conventional magnets. As is shown in Table 1, the maximum energy product of magnet was further improved till the length to diameter ratio was increased up to 10:1 but no longer changed even when the ratio was further-more increased.

Having thus disclosed the invention, what is claimed is:

A process for producing a magnetically anisotropic Al-Ni-Co-Cu type alloy magnet body which has a columnar crystalline structure with the crystals in parallel alignment which comprises heating a mold having a multicavity construction to a temperature greater than about 800 C., the depth of the cavities being greater than six times the diameter, transferring the heated mold to a chilling plate having a heat capacity sufiicient to maintain a temperature gradient lengthwise of a molten and solidifying body of said alloy, filling said mold cavity with said molten alloy so that said alloy has a length greater than six times its diameter, and cooling and solidifying said alloy under the influence of said temperature gradient while simultaneously retarding the cooling of the end of the alloy body remote from the chilling plate and substantially preventing the escape of heat laterally from the moldto produce a cast magnetbody'having columnar crystalline structure and a l gth greater than six times its FOREIGN IATENT S diameter. 652,022 4/1951 Great Britain. by 52:25: 31222 32:22:21: UNITED STATES PATENTS 5 743,636 1/ 1956 Great Britain.

2 3 0 1 7 1945 Snoek et 1 22 213 J. SPENCER OVERHOLSER, Primary Examiner. 2,578,407 12/ 1951 Ebling 222'13 XR MICHAEL V. BRINDISI, WILLIAM J. STEPHENSON,

Examiners.

Images