US3527597A - Carbide suppressing silicon base inoculant for cast iron containing metallic strontium and method of using same - Google Patents

Description

United States Patent US. CI. 75-58 11 Claims ABSTRACT OF THE DISCLOSURE An inoculant for graphitizing cast iron, consisting basically of silicon and between 0.1% and strontium in metallic form, wherein any calcium present is kept below 0.35%; preferably calcium is kept substantially below that level and in the optimum case below 0.1%. The silicon may be present in the form of ferrosilicon or as an alloy of copper and silicon; also a straight strontiumsilicon alloy is disclosed.

This application is a cOntinuation-in-part of my copending application Ser. No. 303,582, filed Aug. 21, 1963, now abandoned.

This invention relates to the manufacture of cast irons and has for an object to provide means for improving certain of the mechanical or physical properties of the iron.

Cast iron generally contains, as a major alloying element, from 2 to 4 percent of carbon. This may be in the form of graphite, when the iron is referred to as grey iron or in the form of iron carbide, when the iron is referred to as a white iron. If the carbon is present partly as graphite and partly as carbide, the iron is termed mottled. A grey cast iron is relatively soft and machinable while a white iron is hard and brittle. The strength of grey iron may be controlled by changing the amount or form of the graphite. Usually the graphite occurs in the form of flakes, but processes have been de veloped whereby it assumes a nodular or spheroidal form which is associated with the highest strengths and with improved ductility.

The structure and properties of grey cast iron, especially when the graphite is present in nodular form, can be improved by employing the technique known as inocula tion. Inoculation may be described as an addition to the molten metal, just before casting, of a small quantity of a substance termed an inoculant, and the effect which it achieves is very much greater than would be anticipated from the change in composition resulting from its solution in the iron.

Inoculation is an essential feature of the production of high strength cast irons, particularly those having a low carbon content or containing nodular graphite. If inoculation is inadequate these irons will contain carbides, or chill, in thinner sections and may fail to achieve the required mechanical properties. One of the major effects of inoculation is the suppression of this tendency to form chill, the degree of suppression depending on the power of the inoculant and the amount added.

Almost all inoculants contain a large proportion of silicon, the one exception being graphite, which may be used very successfully for the inoculation of some flake graphite irons, but not of nodular irons.

Ferrosilicon alloy containing 75 to 80 percent silicon is probably the most widely used of the high-silicon inoculants, particularly in nodular irons, but pure ferrosilicon has very little inoculating effect when added to cast iron; commercial foundry grade ferrosilicon depends upon its content of small amounts of minor elements, notably aluminum and calcium, for stimulating the inoculating effect. The inoculation effect improves more or less proportionally to the aluminium and calcium contents when these contents are small, but any increase above these low content levels does not cause any useful further increase in inoculating effect. Furthermore, the use of such ferrosilicon containing a few percent of aluminum is a potential source of pinhole defects, particularly in nodular lIOl'lS.

The addition of an inoculant refines the grain of cast iron, which is recognised by those skilled in the art from the appearance of the surface at a fracture. Recent investigations have demonstrated that the grain of the cast iron can be revealed on a polished micro-section from the casting as a pattern known as the eutectic cell pattern, and it is now well-known that the magnitude of the inoculating efiect is related to the number of eutectic cells appearing on a unit area of such a polished surface. For nodular irons the number of nodules in a unit area of a micro-section represents the equivalent parameter.

In the production of nodular irons there is a greater tendency for white iron to appear in thin sections, and inoculation is normally carried out following the nodu larising addition as an important part of the process. For this purpose very effective inoculation is necessary.

For both flake graphite grey irons and for nodular graphite irons there is a great need for new inoculants which can have a more potent effect for a given quantity of addition than those at present in common use, or alternatively for additions which will have the same effect when added in smaller quantities. In recent years it has been shown that the presence of a small amount of aluminium or calcium is desirable in silicon-bearing inoculants in order to develop the full inoculating effect.

It has now been discovered that the addition of as little as 4% of strontium (about 2% by content), and even less, to a silicon-bearing inoculant greatly increases its efficiency when added to flake graphit grey iron or to nodular graphite iron. The strontium does not itself have an appreciable inoculating action, but has the effect of increasing the efficiency of such an inoculant. This effect is not directly proportional to the quantity of strontium added, and is operative for strontium contents in the inoculant from 0.1% to 10%. Strontium contents from 0.4 to 4% give particularly good results and strontium contents from 0.4 to 1% are highly effective while also being more economical as regards strontium. An inoculant having about a 1% strontium content has been found to be conveniently manufactured while providing good results.

Further, according to the present invention, the maximum permissible calcium content of the inoculant is critical at about 0.35% if the benefit of the strontium is to be obtained. It is preferred that the calcium content be kept below 0.15% in order to realise the full effect of strontium, and calcium contents below 0.1% have been found to be highly effective in a wide variety of tests.

Furthermore, the calcium content must not exceed the strontium content, and preferably the ratio of strontium to calcium is about 2:1 or more by weight.

The silicon-bearing, strontium-containing inoculant of this invention need not contain aluminium in order to be effective; however, aluminium can be present in amounts up to 5% or more.

Although the preferred form of the present invention is a ferrosilicon alloy (70%90% Si) containing stron- 3 tium and having low calcium content as previously mentioned, the silicon content can range from as low as 15% up to an alloy that is essentially free of all materials except silicon and strontium. Copper-silicon alloys containing strontium and having low calcium contents are also effective as inoculants as hereinafter described.

A series of practical experiments was conducted with various strontium-containing ferrosilicon alloys having different strontium additions. The alloys to which strontium was added were made in a vacuum furnace under an atmosphere of argon.

Special ferrosilicons were made from Armco iron and 99% pure silicon metal either in small, recrystallised alumina crucibles heated by a graphite inductor or directly in the acid-lined crucible fitted in the vacuum furnace. The vacuum chamber was always thoroughly evacuated during the relatively slow heating up period and then filled with argon to a pressure of one atmosphere just before melt-down. Minor additions of elements were made to the liquid bath while it was held under argon at a tem perature just above melting point and the alloys were then allowed to solidify in this atmosphere. These alloys contained less than 0.1% calcium and less than 0.5% aluminium unless otherwise indicated below.

Strontium was also added to commercial alloys by melting the alloys as described above and then adding strontium metal to the liquid alloy. These alloys contained about 0.8% calcium.

In addition, a strontium silicon alloy was made from a starting mixture of 60% Sr and 40% Si.

Finally, tests were made with very high silicon content alloys, and alloys containing approximately 70% silcon and 30% copper, some of these latter two groups The results were compared with the results using a normal foundry grade 75% ferrosilicon having the following analysis:

75% Silicon, 1.2% aluminium,

0.8% calcium, remainder iron.

The cast iron melts, to which the inoculants were added were prepared in acid-lined high frequency furnaces from pig iron charges or recarburised steel scrap. The castings were always made in greensand moulds, based upon natural red sand, 6% coal dust and 5% water, unless otherwise stated.

The following examples illustrate the improvement in inoculating effect which can be achieved by the use of strontium additions to ferrosilicon alloys having low calcium contents. Improvements of the order of 10% are regarded as significant.

Melt No. 1 was prepared consisting of flake graphite iron containing nominally 3% carbon, 1.6% silicon, 0.5 manganese, 0.1% sulphur and 0.1% phosphorus. Four taps were taken and the last three were inoculated at 1380 C. One pair of 1.2 in. diameter test bars and one natural chill test casting were poured from each tap 30 seconds after treatment. The chill test casting provided three small plates of A in., in., and in. thickness; these were fractured along their width and the amount of chill at the outer edge was measured. Tensile tests were carried out on the 1.2 in. diameter bars and eutectic cell counts were made on samples cut from the gauge-length of the broken test bars. Brinell hardness measurements were made on the shoulders of the tensile bars.

The treatments of the taps, the chemical analyses and the test results for melt No. 1 are given in Table 1.

TABLE 1.MELT NO. 1 (FLAKE) Chill depth in Karin. on plates Alloy No. Tensile Eutectic Si Si addn. or type Me Me strength, cells per Si increase percent of inoculant in. in. in. tons/in. l-ln. line percent percent The chill removal and eutectic cell count were marginally better in tap /2, inoculated with alloy No. 1 having a 3% strontium metal addition (about 1.5% by content) and less than 0.1% Ca, than with normal ferrosilicon having a calcium content of about 0.8%. The presence of 1% calcium together with the strontium had a detrimental effect.

Melt No. 2 was a nodular iron containing 3.7% carbon, 2.1% silicon, 0.5 manganese, 0.6% nickel and 0.06%

COMPOSITIONS OF INOCULANT ALLOYS Sr Ca Sr percent Al Ca percent Si percent content percent percent content percent Type of inoculant added approx. added added approx. approx. Alloy No.: I

1 Special FeSi 3 1. 5 0 0 0. 1 2 do 3 1. 5 0 1 0. 6 70 3 do 2 1 0 0 0. 1 4 do 1 0. 5 0 0 0. 1 75 5 dn 4 2 0 0 0. 1 75 6 do 2 1 1 0 0. 1 75 7 do 2 1 2 0 0. 1 76 R dn 2 1 0 1 0. 6 75 Remelted commercial FeSi 2 1 0 0 0. 8 70-80 10 Special SrSi 60 O 0 40 magnesium. Three taps were taken. The details of taps are shown in Table 2.

TABLE 2.-MELT NO. 2 (NODULAR) Alloy N0. and Si percent Type of Total Si addition inoculant percent percent 0. No. 1 3. 72 2. 11 0.5 N onnal FeSi 2. 14 0.5 No. 2 3. 55 2. 08

Short bars of 0.6 in., 0.875 in. and 1.2 in. diameter were poured from each tap and their micro-structures were examined. In the 0.6 in. bars no carbide was found in tap 2/1 inoculated with alloy No. 1 containing strontium as the only additive, a considerable amount of carbide was found in tap 2/2 inoculated with normal ferro-silicon and tap -2/ 3 inoculated with alloy No. 2, con- Melt No. 4 was prepared from a charge similar to Melt No. 3. Five taps were taken and additions of alloys Nos. 3, 7, 8, 9 and one of normal ferrosilicon (0.8% Ca) were made; but the silicon addition was higher than in Melt No. 3, thus encouraging a higher cell count (other conditions remaining similar) and hence higher tensile strength. The same types of test were made as before.

TABLE 4.MELT NO. 4 (FLAKE) [Si addition, all taps, 0.35%]

Chill depth in Sr Ca 34 2 in. on plates Percent percent Tensile Eutectic Si in- Alloy No. or content content %t1 A6 strength cell per Si crease,

type of inoculant approx. approx. in. in. in. tons/in. 1 in. line percent percent Tap No.:

4/1 3(2% S1) 1 01 4 0 0 17. 9 58 1. 63 0.20 4/2 7(2% Sr plus 2% Al) 1 0. 1 5 0 0 20. 3 56 1. 79 0. 36 4/3 8(2% Sr plus 1% Ca) 1 0. 6 3 0 l8. 1 46 1. 80 0. 37 4/4 Normal FeSi 0.8 1 0 17.6 52 1. 77 0.34 4/5 9(nor. FeSi plus 2% Sr). 1 0. 8 1 0 18. 9 53 1. 75 O. 32

taining strontium and calcium was almost white. A similar variation occurred in the larger diameter bars except that the amount of mottle became lessas the diameter increased. As in the flake graphite iron, alloy No. 1 containing strontium alone had an inoculating effect stronger than the normal ferrosilicon while alloy N0. 2 containing strontium and calcium had a weaker effect.

THE EFFECT OF STRONTIUM DISSOLVED IN FERROSILICON (i) Flake graphite irons:

These results again illustrate the advantages which arose from the use of the low calcium alloys to which 2% strontium had been added (1% by content) (see taps 4/ 1 and 4/2). The addition of 2% aluminum with the 2% strontium (1% by content) to the alloy had little further effect on the eutectic cell count or chill removal but it did give the expected improvement in tensile strength (see tap 4/2). The presence of calcium in the alloy together with the strontium was a definite disadvantage (tap 4/3) and this likewise was a definite disadvantage provement which occurred when strontium was added to normal ferrosilicon which contained 0.8% calcium (tap 4/5).

(ii) Nodular graphite irons Two experiments were carried out using the special alloys Nos. 3 to 9 to inoculate nodular iron. The nickelmagnesium alloy was added to each tap at 1400 C. and

TABLE 3.-MELT NO. 3 (FLAKE) [Si addition, all taps, 0.25%]

Sr Ca Chill depth in percent percent lz-in on plates Tensile Eutectic S1 Alloy No. or type of content content strength, cells per S1 lncrease, inoculant approx. approx. M6 in. m. 346 in. tons/1n. l in. line percent percent 1 White.

No'rE.-Brinell Hardness (H13 10/3000) was measured in each case and was found to be lowest (196) for tap 3/3 and highest (204) for tap 3/6. Tap 3/4 showed a value of 198, taps 3/1 and 3/2 showed 199, and tap 8/5 showed 200.

The notable feature of the results was the remarkable effectiveness of alloys Nos. 3 and 6 (taps 3/2 and 3/5) both of which had received 2% additions of strontium (about 1% by content), and contained less than 0.1% calcium. The eutectic cell count was higher and the chill removing effect much greater than for the equivalent additions of normal ferrosilicon which contained about 0.8% calcium. The alloys to which 1% and 4% strontium had the inoculants were added as soon as the magnesium flare subsided.

Melt No. 5 (similar to Melt No. 2) was divided into six taps and alloys Nos. 3 to 6 were compared with additions of normal ferrosilicon. Keel blocks having 1 /2 in. square keels and short bars of 0.6 in., 0.875 in. and 1.2 in. diameter were poured from each tap. Mechanical properties were determined on bars cut from the keels of been added (taps 3/3 and 3/4) were not so effective as 7 the keel blocks and the various section bars were examined as before. The numbers of nodules per square inch were counted on samples cut from the tensile bars and from the 0.6 in. diameter bars. Results are shown in Table 5.

by content). A trace of mottle occurred at the edge of the A in. plate from tap 6/ 1 but it was completely free from mottle elsewhere. A considerable amount of mottle occurred in all the other in. thick plate and it was TABLE 5.MELT NO. 5 (NODULAR) [Si addition, all taps, 0.5%]

Percent Percent Brinell Nodules Sr Ca TGl'lSllB Elongahardness per in. Si Alloy No. or type content content strength, tion, 1113 in 0.6 S1 mcreese, of inoculant approx. approx. tons/in.'- percent /3,000 bars percent percent The micro examination of the bars showed that a trace of mottle occurred in the 0.6 in. diameter bars from taps 5/1 and 5/6, inoculated with normal ferrosilicon, but not in any other bars. The nodule counts on the small bars show that all the taps treated with the special low calcium alloys containing strontium (taps 5/2-5/5) were more strongly inoculated than those treated with normal ferrosilicon (0.8% Ca) (taps 5/1 and 5/ 6). This is supported by the results of the microexamination.

Melt No. 6 was prepared in a similar manner to melt No. 5 except that only five taps were taken. Additions of alloys Nos. 3, 7, 8 and 9 were compared with an addition of normal ferrosilicon. Similar test castings were 4 made and one additional casting device giving small plates 2 in. x 1 /2 in. with thicknesses ranging from in. up to in. was poured from each tap. This was done to enable the chilling tendency of the taps to be examined in sections thinner than the 0.6 in. diameter bars.

The chemical analyses and mechanical test results are shown in Table 6, together with the nodule counts on the 0.6 in. diameter bars all of which were free from mottle. In this melt the silicon content of the furnace metal was higher than in melt No. 5, thus encouraging the castings to be free from mottle and the nodule count to be higher.

TABLE 6.-1\/IELT NO. 6 (NODULAR) [Si addition, all taps, 0.5%]

very severe in plate 6/3 inoculated with alloy No. 8 containing strontium and calcium (0.6% Ca by content), and in plate 6/4 inoculated with normal ferrosilicon (0.8% Ca by content). The powerful inoculating effect of strontium containing alloy No. 3 (less than 0.1% Ca) was again confirmed by the nodule counts on the 0.6 in. diameter bars. Tap 6/1 treated with this inoculant had a far higher nodule count than any other tap. Alloy No. 7 to which 2 percent strontium and 2 percent aluminium had been added was the next most powerful inoculant although it was significantly weaker than alloy No. 3 with no aluminium (tap 6/ 2). The 1 percent calcium addition (0.6% by content) with the 2 percent strontium added to alloy No. 8 reduced its inoculating power to a level slightly below that of normal ferrosilicon (compare tap 6/ 3 with 6/4). The addition of strontium to normal ferrosilicon (alloy No. 9) had a slight improving eflect on its inoculating power but the results were not as good as those from alloys No. 3 or 7. This is accountable to the presence of calcium (0.8%) in the commercial ferrosilicon.

THE EFFECT OF ADDING STRONTIUM METAL ALONE A flake graphite iron charge in which the silicon content had been raised to compensate for the absence of a ladle addition of silicon was used for melt No. 7. This was Sr Ca Nodules percent percent Tensile Elonper in. Si Ta Alloy No. or type content content strength, gation, in 0.6 Si, increase, No. of inoculant approx. approx. tons/in." percent bars percent percent 6/1 1 0. 1 28. 9 26 246, 000 2. 0. 47 6/2 1 0. 1 28. 3 25 204, 000 2. 46 0. 40 6/3 1 0. 6 28. 4 23 129, 000 2. 42 0. 36 6/4 Si 0. 8 29. 0 23 139, 000 2. 48 0. 42 6/5 9(nor. FeSi plus 2% Sr)- 1 0.8 28. 7 23 179, 000 2. 53 0. 47

divided into five taps and additions of strontium metal ranging from 0.01 percent to 0.25 percent addition were made. Chill test plates and 1.2 in. diameter bars were cast and tested as before except that only eutectic cell counts were carried out on the 1.2 in. bars. From the results in Table 7 it is clear that only very slight inoculation occurred with even the heaviest addition of strontium.

TABLE 7.MELT N0. 7 (FLAKE) [No Si addition] Chill depth in 52 in. on Plates Eutectic Type of inoculant Mu m. 3 in. Me in. 1 in. line percent .9 THE EFFECT OF STRONTIUM METAL ADDED WITH FERROSILICON The addition of calcium and aluminium together with ferrosilicon is known to increase the inoculating power of the ferrosilicon in flake graphite irons by an amount similar to that resulting from the solution of the same amounts of these elements in the ferrosilicon.

The possibility of strontium metal having the same eifect was tested in Melt No. 8 which was a flake graphite iron melt divided into five taps. An addition of special ferrosilicon having a low calcium content (less than 0.1% Ca) was made to each tap and increasing additions of strontium metal were made mixed with the ferrosilicon. Chill test castings and 1.2 in. diameter bars were cast and tested as before except that only eutectic cell counts were carried out on samples from the 1.2 in. diameter bars. The tap details are given in Table 8.

TABLE 8.MELT NO. 8 (FLAKE) [Si addition, all taps, 0.5%]

Chill depth in :;2

low calcium (less than 0.1% Ca) ferrosilicon alloys to which 2 percent strontium had been added, (1% Sr by content). The following two melts were designed to give a comparison between normal ferrosilicon and alloy No. 3 (2% Sr added) at three levels of silicon addition.

Melt No. 10 was of typical nodular iron composition except that the silicon content was lower than normal. Three taps were taken, and nickel-magnesium alloy was added to each tap. They were then inoculated with l percent, 0.5 percent and 0.25 percent silicon respectively, added as alloy No. 3 (2% Sr added). Suitable additions of silicon metal were made to the furnace before taking taps 2 and 3 so that the final silicon contents of each tap were similar after inoculation. Keel blocks, various section bars, and small plates of various thicknesses were cast from each tap and tested as previously described. Nodule counts were made on samples cut from the 0.6 in. diameter bars all of which were free from mottle.

in. on plates V 3/ Eutectic S 1 Si us 18 cc s er 1, ncrease, Alloy No. or type of inoculant in. in. in 1 in. line percent percent Tap No:

8/1 Low-Ca FeSl White-.- 8 4 29 1.93 0.49 8/2.-.. Low-Ca FeS1 plus 2% Sr metal -do 5 1 33 1. 98 0. 54 8/3--.. Low-Ca FeSi plus 5% Sr metal d0--. 7 2 32 2.03 0. 59 8/4.... L0w Ca FeSl plus 10% Sr metal 13.-. 0 0 55 1. 93 0. 49 8/5 Low-Ca FeSi plus Sr metal White... 1 0 52 1, 93 0, 49

Note-The Sr metal content was measured as a percentage of the FeSi addition.

The addition of the appropriate amount of strontium metal with the low calcium ferrosilicon can produce a very powerful inoculating effect. In the best case the amount of strontium was equivalent to 10 percent of the ferrosilicon addition (tap 8/4). This is very much more than the earlier tests showed to be best when the strontium was dissolved in the ferrosilicon. However, strontium metal is a very volatile and inflammable substance at liquid iron temperatures and very considerable losses must occur when it is added directly to cast iron.

Melt No. 9 was similar to Melt No. 8 except that the normal foundry grade ferrosilicon (0.8% Ca) was used instead of the grade having low calcium contents.

The tap details and results given in Table 9 show that no improvement resulted from the addition of the strontium metal with this grade of ferrosilicon.

TABLE 9.MELT NO. 9 (FLAKE) [Si addition, all taps, 0.5%]

The details and results are given in Table 10. It should be noted that the nodule number is almost directly proportional to the silicon addition.

plate from tap 10/1 inoculated with l percent silicon as alloy No. 3, was very slightly ductile but the other in. plates were brittle and fully white. The in. plate from Chill depth in $62 in. on Plates Eutectic Si V M6 cells per Si, increase, Alloy N 0. or type of inoculant in. in. 1 in. line percent percent Tap No: 9 to 1.- Normal FeSi 2 0 1. 90 0. 47 9 to 2.-- Normal FeSi plus 2% Sr metal.... 2 0 49 1. 93 0.50 9 to 3. Normal FeSi plus 5% Sr metal. 1 0 51 1. 93 0. 50 9 to 4. Normal FeSi plus 10% Sr metal... 2 0 50 1. 95 0. 52 9 to 5. Normal FeSi plus 40% Sr meta-L. 1 0 48 1. 89 0. 46

No'rE.The Sr metal content was measured as a percentage of the FeSi addition.

A pair of experiments similar to the previous two were carried out in which strontium metal was added with both grades of ferrosilicon when used to inoculate nodular graphite irons. A small improvement in the inoculating power of the low aluminium and calcium ferrosilicon occurred as the strontium addition increased but at no time was the eifect comparable with that which is generally achieved with normal foundry grade ferrosilicon. No further improvement in inoculating power occurred when strontium metal was added with the normal ferrosilicon.

THE EFFECT OF VARYING ADDITIONS OF FERROSILICON CONTAINING STRONTIUM The most striking and probably the most valuable of the preceding results is the improvement in the inoculatap 10/1 was very ductile and that from tap 10/2 fairly ductile. The equivalent plate from tap lO/ 3 was brittle. The in. plate from the tap having the 1 percent silicon addition was free from mottle, whereas the plates treated with 0.5 percent silicon and 0.25 percent silicon contained progressively increasing amounts of mottle. The in.,plates from taps lO/l and 10/2 showed con siderable bending before the first crack appeared, whereas the plate from tap 10/3 was completely brittle.

A similar series of tests on melt No. 10 was carried out in exactly the same manner except that normal ferrosilicon (0.8% Ca) was used instead of alloy No. 3. The test results showed that this time an increase in the silicon addition from 0.5 percent to 1 percent gave only about 6 percent increase in nodule number. All the thin plates tion of nodular irons resulting from the use of the special up to in. thick had the brittle fracture of white iron 12 except that the A in. plate treated with 1 percent silicon The strontium silicide alone (tap 11/2) had very little showed very slight ductility. inoculating effect, but when it was added together with The results of these two series of tests again illustrate the low aluminium and calcium ferrosilicon (tap 11/ 4) the superior inoculating properties of the special strontia very powerful inoculating efiect was produced.

um-containing, low calcium ferrosilicon alloy No. 3 when Melt No. 12 had a lower silicon content in order to used to inoculate nodular iron. In particular, an increase 5 make complete removal of chill more diiiicult. Additions in the addition of the normal ferrosilicon above 0.5 perof low aluminium and calcium ferrosilicon were made to cent silicon gave only a small increase in inoculation as each tap, together with increasing amounts of strontium measured by the nodule number, whereas an almost linear silicide (alloy No. 12). increase occurred with increasing additions of alloy No. 3 The same tests as before were carried out on similar containing strontium and having a low calcium content. castings and the results are given in Table 12.

TABLE 12.MELT No. 12 (FLAKE) [Si addition, all taps, 0.25%]

Chill depth in lz in. on plates Eutectic Si i V3 iii cells per Si, increase, 7 No. Alloy No. or type of inoculant in. in. 1 in. line percent percent Low-Ca FeSi White-.- 14 31 1.61 0.20 12/2. Low-Ca FeSi plus 4% Sr/Si do...- 12 34 1. 55 0. 14 12/3 Low-Ca FeSi plus 10% Sr/Si .do 11 39 1. 63 0. 22 12/4. Low-Ca FeSi plus Sr/Si .do 8 44 1. 59 0.18 12/5 Low-Ca FeSi plus 40% Sr/Si -do 4 48 1. 65 0. 24

Norm-The Sr/Si content was measured in a percentage of the FeSi addition. The two series of tests are illustrated graphically in the 25 A progressive increase in inoculating eifect resulted accompanying drawing, where curve A relates to the series from increasing additions of strontium silicide made with Using alloy 3 and Curve B to the sefles 115mg normal the ferrosilicon having low aluminium and calcium conferrosilicon. tents.

THE EFFECT OF STRONTIUM SILICIDE ADDED Melt No. 13 was identical to Melt No. 12 except that SEPARATELY TO THE MELT SIMULTANEOUSLY the normal grade ferrosilicon was used in place of the WITH FERROSILICON low aluminium and calcium grade. The same tests were Alloy inoculant No. 12 was prepared by making a small m 3 the reults g1ven m Table Agam a crucible melt of pure silicon and adding 60 percent of nificant increase n the inoculating effect occurred as the strontium metal to the melt. The reaction from the stronti- 35 e h of strontlum 111c1de mcreased: This is Surprising um addition was very violent, but a strontium-rich silicon VIGYV of the lastpf P Y Whlch Occurred when alloy containing approximately 65% strontium was b. strontium was dissolved 1n the normal ferrosilicon tained and its effectiveness as a means of adding strontium (Table 9 with ferrosilicon was tested.

TABLE 13.MELT No. 13 (FLAKE) [Si addition, all taps, 0.25%]

Chill depth in in. on plates V Eutectic Si Alloy No. or type tie cells per Si, increase, oiinoculant in. in. 1 in. line percent percent Tap No 13/1 Normal FeSi 2 44 1.63 0.20 13/2... FeSiplus 4% Sr/Si 4 1 48 1.61 0.18 13/3..-

FeSi plus 10% Sr/Si 2 0 52 1.67 0.24 13/4-.-

FeSi plus 20% sum... 1 0 56 1.68 0.25 13/5 FeSi plus Sr/Si.-- 1 0 58 1.65 0.22

Nor1-:.-The Sr/Si content was measured as a percentage of the FeSi addition.

Four taps were taken from Melt No. 11, which was Additions of strontium silicide alone and together with of flake graphite iron, and additions of alloy No. 12 both the grades of ferrosilicon were used to inoculate were made both alone and in conjunction with low alumin nodular graphite irons. No inoculation resulted from the ium (less than 0.5%) and low calcium (less than 0.1%) addition of the strontium silicide alone and no improveferrosilicon. Chill test castings and 1.2 in. diameter bars ment occurred when it was added with either grade of were cast and tested as before, the results being shown in ferrosilicon. Similarly, barium and other silicides were Table 11. In this table, the amount of alloy No. 12 added tested in the same way as strontium silicide and found to tap No. 11/ 2 was equal to the Weight of FeSi inoculant to h e no i ro ed i o ulatin ffe t, added to p 11/1- The lnoculaht 111 p 11/ 4 was the Same The following additional examples will further illusas in tap 11/ 3 with the addition of a quantity of SrSi equal tram the present invention in weight to two thirds of the weight of FeSi.

TABLE 11.MELT NO. 11 (FLAKE) 1 About 0.8% Ca.

13 EXAMPLE I Flake graphite iron of nominal composition 3.0% total carbon, 1.5% silicon, 0.5% manganese, 0.08% sulfur, 0.1% phosphorus, was inoculated with 0.45% silicon using normal ferrosilicon (0.8% Ca) in tap 1 and low calcium (less than 0.1% Ca) ferrosilicon containing 0.4% strontium in tap 2. The results observed on A; in. plates are as follows:

CHILL DEPTH IN & IN.

Tap 1 2 Tap 2 0.

EXAMPLE II Base iron of nominally 3.4% total carbon, 1.8% silicon, 0.2% manganese, minimal phosphorus and sulfur was divided into two taps and each tap treated with sufficient cerium-free nickel-magnesium alloy to produce a nodular graphite structure. Tap 1 was post inoculated with normal ferrosilicon (0.8% Ca), tap 2 was post inoc ulated with low calcium (less than 0.1% Ca) ferrosilicon containing 0.4% strontium. in. thick plates and 0.6 in. diameter bars were cast from each tap. These were fractu-red and the fracture examined. The A in. plate from tap 1 (normal ferrosilicon) was completely white and brittle whereas the plate from tap 2 was strong and ductile. Metallographic examination of the 0.6 in. diameter bars showed that the bar from tap 1 (normal ferrosilicon) was mottled containing considerable quantities of carbide whereas the bar from tap 2 was free from mottle.

EXAMPLE III The following alloys were prepared for use as inoculants:

Alloy l1.6% Al, less than 0.1% Ca, balance Si Alloy 20.4% Sr. less than 0.1% Ca, balance Si Alloy 3-1.4% A1, 0.9% Ca, balance Si Alloy lWhite Alloy 2-Grey and ductile Alloy 3-White Pure siliconWhite EXAMPLE IV The following alloys were prepared for use as inoculants:

Alloy 1-70% Si, 30% Cu, less than 0.1% Ca Alloy 2Allo-y 1+2.2% Al Alloy 3--Alloy 1+0.55% Sr Alloy 4Alloy 1+ 1.13 Ca Alloy 5--Alloy 1+2.5% Ca and 3 Al Chill tests performed on flake graphite iron using the foregoing alloys as inoculants gave the following results:

Chill depth in %2 in.

plate plate Alloy 1 16 8 Alloy 2 16 6 Alloy 3 1 Alloy 4 4 1 Alloy 8 1 Similar tests performed with nodular iron using the same alloys gave the following results for A in. plates:

Alloy lWhite Alloy 2White Alloy 3-Ductile Alloy 4White Alloy 5White As can be seen from the foregoing examples, low calcium silicon containing small amounts of strontium is highly effective as an inoculant as is low calcium silicon containing small amounts of strontium and up to 30% copper.

The results of all the foregoing tests show that it is possible to use a low calcium silicon-containing inocu lant with both flake and nodular graphite cast irons which, in the presence of strontium, has a significantly more powerful inoculating effect than normal inoculating grade ferrosilicon. Furthermore, this inoculant can be virtually free from aluminium, thus avoiding the potential hazards which arise from the small additions of this element which are automatically made with normal ferrosilicon. However, as previously noted, aluminium can be present in amounts up to 5% and more without affecting inoculation.

The preferred inoculant of this invention can be produced by dissolving a small amount of strontium metal in pure ferrosilicon (less than 0.1% Ca, less than 0.5 Al) or in some cases just by adding strontium metal or a strontium-rich silicon alloy together with fer-rosilicon. In any event, the calcium content of the final alloy should be less than 0.35%. The strontium content should be at least 0.1% and not more than 10% and preferably from 0.4 to 4%.

The effect of the strontium in the present invention is analogous to, but stronger than, the well-known effect of aluminium in ferrosilicon. However, strontium differs from aluminium in that the simultaneous presence of calcium with the strontium nullifies its effect, whereas calcium enhances further the effect of aluminium. Aluminium has not been found to significantly affect the action of the strontium addition.

The amount of strontium required to be dissolved in the ferrosilicon alloy is quite small. As shown by the foregoing data, strontium additions of between 1 percent and 4 percent have been found to be effective with a maximum inoculating effect occurring with a 2 percent addition (about 1% by content) when added under the conditions described herein. However, strontium is a very volatile and reactive element when added to liquid ferrosilicon, therefore the best amount to add is likely to vary somewhat with the circumstances of the addition. With the alloying process described earlier in this specification, about 50% of the strontium added is retained in the alloy. The strontium content of the ferrosilicon alloy should, in any event, be in the ranges previously stated e.g. 0.1 to 10% and the calcium content should be less than 0.35% as also previously mentioned.

In flake graphite irons the effect of the strontium can be achieved by making an addition of strontium metal or strontium silicide simultaneously with the ferrosilicon addition. In these cases the amount of strontium required to give the best effect is much greater than when it is dissolved in the inoculant. This is doubtless due to the volatility of the strontium addition. When strontium is added with the inoculant in this manner, the suppressing effect of calcium again occurs but even so a definite improvement was observed when the larger additions of strontium silicied were made with the normal ferrosilicon which contained a small amount of calcium.

The potent inoculating power of the ferrosilicon containing strontium and having a low calcium content is best illustrated by its use in nodular graphite irons. The aim in making nodular iron is to produce castings which are free from mottle in the as-cast condition and which therefore do not require to be annealed. When thin section castings are being made by the conventional techniques, a very high degree of inoculation is required to achieve this, and even so many cast-ings of /2 in. section or less need to be annealed. Attempts are usually made to improve the inoculation by adding very large amounts of ferrosilicon, but the advantage gained is only slight. This is demonstrated in the tests of melt No. 10, where increasing the addition of normal ferrosilicon above 0.5 percent had only a slight effect on the nodule number and the removal of mottle. Furthermore, very large additions of normal ferrosilicon greatly increase the risk of pinholes occurring due to the addition of the alumi- Ilium in the ferrosilicon.

Strontium-containing low calcium ferrosilicon additions of 0.5 percent silicon gave higher degrees of inoculation of nodular graphite iron than with normal ferrosilicon and furthermore, increasing the addition up to at least 1 percent gave a directly proportional increase in inoculating effect. In practice this means that it should be possible to produce much thinner nodular iron castings free from mottle by using additions of this type.

Similarly, in flake graphite irons it should be possible to produce very thin castings with less risk of chilled edges by using a low calcium strontium-containing inoculant and thus assist in satisfying the ever increasing demand for thinner and lighter castings.

In similar tests to those reported above (Tables 11 to 13), additions of silicides of tungsten, cerium, titanium, zirconium, molydbenum or barium, either alone or in conjunction with ferrosilicon, were not observed to have any significant inoculating elfect in flake graphite irons.

I claim:

1. A carbide-suppressing silicon-bearing inoculant for cast iron consisting essentially of from about 15% to 90% silicon and from about 0.1% to 10% metallic strontium, up to 30% copper, up to aluminum, balance iron, with residual impurities in ordinary amounts wherein any calcium present is not in excess of about 0.35% or the amount of strontium present, whichever is lower.

2. A carbide-suppressing silicon-bearing inoculant for cast iron as defined in claim 1, wherein the strontium content is from about 0.4% to 4%.

3. A carbide-suppressing silicon-bearing inoculant for cast iron as defined in claim 2, wherein any calcium present is not in excess of about 0.15%.

4. A carbide-suppressing silicon-bearing inoculant for cast iron as defined in claim 1, wherein the ratio of stron- 16 tium to calcium content within said maximum limit is a minimum of 2:1 by weight.

5. A carbide-suppressing silicon-bearing inoculant for cast iron as defined in claim 1, wherein the strontium content is approximately 1% and any calcium present is not in excess of about 0.1%.

6. A carbide-suppressing silicon-bearing inoculant for cast iron as defined in claim 1, consisting essentially of ferrosilicon containing from about 0.1% to 10% strontium.

7. A carbide-suppressing silicon bearing inoculant for cast iron as defined in claim 6, wherein the strontium content is from about 0.4% to 4% and any calcium present is not in excess of about 0.15%.

8. A carbide-suppressing silicon-bearing inoculant for cast iron as defined in claim 6, wherein the strontium content is about 1% and any calcium present is not in excess of about 0.1%.

9. A carbide-suppressing silicon-bearing inoculant for cast iron as defined in claim 6, wherein the silicon content is from to 90%.

10. A carbide-suppressing silicon-bearing inoculant for cast iron as defined in claim 6, wherein the silicon content is about strontium about 1%, and any calcium present is not in excess of about 0.1%.

11. The method of making cast iron which includes the step of adding to the melt, before pouring to suppress carbide formation in castings produced therefrom, a silicon-bearing inoculant consisting essentially of from about 15% to silicon and from about 0.1% to 10% metallic strontium, up to 30% copper, up to 5% aluminum, remainder iron except for incidental impurities wherein any calcium present is not in excess of about 0.35%.

References Cited UNITED STATES PATENTS 2,221,783 11/1940 Critchett et a1. 75-58 2,280,283 4/ 1942 Crafts 75-58 X 2,444,354- 6/ 1948 Kinnear 75-130 2,610,911 9/1952 Udy 75-130 X 2,676,097 4/ 1954 Strauss 75-130 X 2,932,567 4/ 1960 Evans 75-130 X 3,374,086 3/1968 Goehring 75129 X HENRY W. TARRING H, Primary Examiner U.S. Cl. X.R. 75-123, 129,