US3177074A - Cobalt base alloys - Google Patents

Description

United States Patent 3,177,074 COBALT BASE ALLOYS Walter A. Luce and Glenn W. Jackson, both of Dayton, Ohio, assignors to The Duriron Company, he, Dayton, Ohio, a corporation of New York No Drawing. Filed Apr. 23, 1962, Ser. No. 189,274 6 Claims. (Cl. 75171) This invention relates to cobalt base alloys and particularly to cobalt alloys which are unique in respect to corrosion resistance throughout a broad range of conditions, with particular emphasis on resistance to chemical attack in severely corrosive liquid media, while providing concomittantly therewith exceptional wear and abrasion resistance, high strength, adequate ductility and good thermal shock resistance.

A general object of the invention is the provision of alloys especially useful in bearing surfaces and similar applications involving mechanical wear, where the surfaces are unavoidably exposed to highly corrosive liquid media, for example strong mineral acid like nitric, sulfuric and hydrochloric. In order to satisfy the requirements of wear and abrasion resistance, it is essential that the alloys possess a high order of hardness, and for this invention a Brinell hardness number of approximately 300 in the alloy is considered to be a minimum which is adequate in actual service. As-cast alloys which exhibit slightly lower Brinells, say as low as about 270 but which can through appropriate treatment be hardened to the recommended level without adversely affecting other properties, are thus useful and are encompassed by the invention.

The versatility of the invention alloys in respect to corrosion resistance is outstanding. These alloys have broad corrosion resistance capabilities which are not only better than any other known cobalt alloy but are also better than most other known alloys for a similar broad spectrum of corrosive environments. Corrosion data given hereinafter for the noval alloys in some of the more common mineral acids under various conditions serves to verify their superiority. For test purposes, as presently explained, certain combinations of media, temperature and concentration have been deliberately selected to provide extremely severe conditions which produce fairly high rates of corrosion which would not generally be acceptable in commercial practice. By reducing either the temperature or concentration conditions slightly, however, acceptable resistance results can be obtained. And when the conditions are thus altered to the point where the invention alloy corrosion rates become commercially acceptable, the nearest equivalent alloys heretofore available are far inferior in one or more of these environments.

In respect to corrosion resistance, tests have established that acceptable alloys Within the purview of this invention should have mean corrosion rates which generally do not exceed the following limits: 15 mils per year in 65% boiling nitric acid; 20 mils per year in 10% boiling sulfuric acid; 150 mils per year in 30% "ice boiling sulfuric; 30 mils per year in sulfuric at 176 F.; 30 mils per year in 15% hydrochloric at 125 F.; and mils per year in 20% hydrochloric at F. The term mean corrosion rates as here used designates the mean rates of corrosion for five 48-hour periods, in accordance With A.S.T.M. procedure set out in Standard No. A279-44 of the Society, except that mechanical agitation of the specimen is used for more consistency. It is not to be understood that corrosion rates of as high as mils per year, as in 30% boiling sulfuric acid, would be tolerated in actual practice. As a rule a maximum acceptable over-all corrosion rate is around 15 to 25 mils per year. However, as mentioned above, during testing for evaluation purposes extreme corrosion conditions were deliberately picked, and the higher rate levels given for these conditions therefore serve simply as a reference against which to measure and compare the various compositions.

Alloys capable of meeting the foregoing corrosion and abrasion specifications are greatly needed in industry but have not been available heretofore. Specific examples of such need are to be found particularly in chemical processing equipment, for example, chemical pump components such as shaft sleeves and impellers, and valve seats for handling corrosive fluids at various temperatures. It has long been a problem in handling such fluids to design and construct an impeller shaft for chemical pumps which is capable of serving adequately and economically. The problem has become aggravated with the need for pumps of larger capacity and higher operating speed, wherein the strength requirements have largely outstripped the capability of most materials to provide practical periods of useful life in such service. Known alloys having suitable chemical corrosion resistance are unfortunately insufiicient in respect to tensile strength, hardness and ductility to withstand the type of mechanical and thermal shock to which such equipment is inevitably subjected. 1

In a centrifugal pump, for example, some of the most troublesome operating or maintenance problems occur Where the shaft passes through the stuffing box. Attempt has been made to use a one-piece shaft of some corrosion resistant material such as a stainless steel. Not only does this become quite expensive but the austenitic stainless steels which are suitable from the standpoint of corrosion resistance cannot behardened beyond about Brinell which, as above indicated, is far below the desired level, with the result that wear resistance is poor. Alternatively, the use of a stainless or carbon steel shaft fitted with a separate sleeve of Duriron, which is a vary hard material (400 to 500 Brinell) having excellent corrosive resistance, runs into problems of low tensile strength and low tolerance to thermal shock in the sleeve material. Some of the known cobalt alloys, like the dental alloy Vitalium or Stellite 21, have excellent hardness and wear resistance but their corrosion rates in these severely corrosive media show them to be quite insutficient for the type of service here contemplated. The mean corrosion rate of Stellite 21 in boiling 65% nitric acid, for example, is around 164 mils per'year as compared to the desired limit of only mils per year in this media as specified above, and about 8 mils per year actually achieved by the invention alloys. As a further example, the invention provides' an alloy having a mean corrosion rate of 15 mils per year in a 10% nitric-3% hydrofluoric acid solution at 140 R, which is in contrast to around 75 mils per year from Durimet 20, heretofore considered to be the best commercial alloy available for this environment. Even such exotic materials as titanium and tantalum are completely unacceptable here, and the costly nickel-base alloys are considerably worse than Durimet i i We have discovered that the combined requirements of corrosion, Wear and abrasion resistance, ductility, adequate tensile strength and resistance to thermal shock;

needed for the severe type of service contemplated herein, are possessed uniquely by certain cobalt-base alloys having carefully combined and narrowly limited ranges num, mayalso be present for specific applications of the alloy, but the total of these must be held low and general 1y their inclusion in the alloy composition is undesirable if the alloy is to be used under'the more drastic operating conditions. Similar small percentages of other elements used principally as deoxidants, as for example calcium,

and as stabilizers, for example columbium, or hardeners,

as boron for example, may also be incorporated.

The unique results of the invention, however, are obtained only within compositional limits of the major components defined below, and then only by proper balance of the components within these ranges, as more fully explained hereinafter. Thelimiting ranges of the alloy compositions are as follows: i V

C From about 0.02 to,0.l0%. Ni Up to 10.0% max.

Fe Up to 6.0% max.

Mn Up to,2.0% max.

Si Up to 2.0% max.

Co Balance.

For shaft sleeve purposes, hardness is of paramount importance. Accordingly, a preferred alloy within the foregoingrange which is particularly adapted for such use has been found to have a nominal composition of about 30% chromium, 13% molybdenum, 3.25% copper,

Where an alloy of lower hardness can be tolerated, as for pump housings and impellers, or for valve bodies, gates, seats and similarcomponents, a somewhat more dutile composition within the foregoing over-all range is provided by the following compositional analysis: 28% chromium, 8.5% molybdenum, 3.25%. copper, 0.10% max. carbon, up to: 10% nickel, 2% iron, 0.5% each of manganese and silicon, the balance (approximately 48 to 5 8%) cobalt; Such an alloy has a hardness of 270-300 BHN, a tensile strength of approximately 75,000 p.s.i., and an elongation of 6%. 'Again this alloy falls within the corrosion rate test limits given; (See Alloy A in Table II.)

Such an-alloy aslast described, however, must in general be work hardened as by shotpeening to achieve the higher hardness level. For more intricate castings, therefore, an alloy having a molybdenum-chromium content intermediate the specified limits of these elementsprovides a hardness of around 340 BHN, and should be used. An alloy of such. intermediate hardness will have im-. proved mechanical properties relative to the .very hard (375-400 BHN) alloy, without serious reduction in Wear or abrasion resistance.

An alloy having as much as-6% elongation represents substantially the maximum in respect to dutility within the invention. Although significantly: less hard than others obtainable within the present teaching, it is still much superior tov stainless steel. But again. this maximum degree of dutility will not be realized unless all the alloying elements closely approximate the stated levels.

Theshaft sleeve (very hard) alloy provides mechanical properties which are considerably improved over that of high silicon iron alloys, e.g. .Duriron, heretofore found to be the best available for service under the conditions contemplated. ,Although it is necessary in these shaft sleeve alloys to approach a brittle composition in order to acquire a hardness of :375 t'o 400BHN, still the ductility obtained is, by comparison, a substantial improvement over that of the'high silicon irons.

As the nickel content is raised,.the alloy of course is softened, with consequent improvement in its ductility,

and for nickel contents up to 10% the corrosion resistance is generally maintained at adequate levels forv all but the very worst conditions. Here again. however the level within the allowable range. will be determined by the 0.03% max. carbon, 2% nickel, 2% iron, 0.5% each of silicon and manganese, the balance (approximately 48%) being cobalt. Such alloy has a Brinell hardness number of approximately 400, a tensile strength of 60,000 psi. 1 and an elongation of about 1%. As appears more fully hereinafter, the corrosion rates of such an alloy are all well within the maximum rates previously specified. (See Alloy M in Table VII which follows.)

hardness and corrosion resistance requirements to be met.

Although the prior art teaches that in general substitution of tungsten for some or allof the molybdenum may be made in other alloys, this teaching definitely does not apply to the alloys of this invention, and tungsten cannot be substituted for'molybdenum, nor simply added to the alloy, without seriously 2 reducing corrosion resistance and/or increasing crack susceptibility, even at theylower molybdenum levels.

A further critical point in respect of the novel'alloys is the carbon content, more particularly in the harder, i.e. shaft sleeve (375400 BHN), composition. In this com position, carbon must be maintained at or below 0.03% maximum in order to develop the optimum corrosion resistanceof which the alloy is capable. In. the case of alloys of lower hardness 1evels,'i.e. around 270-300 BHN, since these appear to respondsatisfactorily to quench'anhealing treatments-without adverse ,elfect upon corrosion resistance, carbon contents up to 0.10% can be used but a preferred level even here is.0.07% or less.

The criticality of ranges-of components of the novel alloys given above is shown by the following examples. Table I lists the compositional I analyses of all of the alloys whose characteristics .arelisted inthe subsequent tables. 1

Other Table I 5 O0U0O0000000 7500000000000000 llzalltlouswuaweaasasws 7 7 Alloy Nil 145 Nil I Periods 10 S, Boil.

Coluld not 0 otain samples* 11 10 Sf Boil.

Mean Corrosion Rate, Mils Per Year 548 Hour Test 7 65 N, Boil.

Table II Mean Corrosion Rate, Mils Per Year -48 Hour Test Periods 65 N,* Boil.

Table III tionally be considered a ductile material, nevertheless the mechanical properties are such that it is capable of Withstanding the shock and strain of mechanical stresses involved in shaft sleeve use.

Tungsten, as mentioned above, has commonly been considered in the prior art to be a potential substitute for molybdenum but not at the low carbon levels here concerned. In Table III below, an 8% tungsten addition as a complete substitute for molybdenum (Alloy V) produces severe cracking, and in fact this heat was so weak that it could not even be tested for corrosion rates. substitution of 4% tungsten for that same amount of molybdenum causes moderate cracking and lowers the corrosion resistance of the material considerably. Even the straight addition of tungsten to the optimum alloy composition, as in Alloy N, causes loss of corrosion resistance.

Hardness BEN Hardness BHN act 50 that the hardness changes from approximately 265 BHN Percent Mo and 20% hydrochloric acid. These abbreviations apply throughout the remaining tables.

Further increase in molybdenum up to 18% increases the hardness Percent Percent Alloy *65 N denotes 65% nitric acid. S, 30 S and 70 S denote, respectively, 10%, 30% and 70% sulfuric acid. H

Alloy and H denote 15 Although at the 13% molybdenum level the alloy is not one which would conven- *Samplcs could not be out without cracking severely.

Values considered nil since elements not intentionally added. I Balance cobalt in all cases except for incidental impurities which do not exceed 1% at most, generally less.

All heats quench annealed from 2250 F. (air or water quench, depending on the alloy as explained in the test).

Molybdenum is one of the most critical elements in the alloy and requires close control. In most media cor- See Table II for the effect on corrosion resistance atvarious molybdenum levels.

Quite as important as corrosion resistance is the f This increase in hardness is accompanied by gradual change from what, for present rosion resistance drops oif quickly below and above 8% and 15%, respectively. At the high end of the molybdenum range, satisfactory compositions are obtained by proper adjustment of the elements, especially the nickel content.

at 8% molybdenum to 390 BHN at 13%.

to approximately 500 BHN.

purposes, may be considered a ductile alloy at 8% to a very brittle material at 18%.

Corrosion rates of alloys 1 As already pointedout, the carbon level is critical. When increased quantities of chromium and molybdenum are employed in the presence of carbon much above 0.10%, hardness is increased but corrosion resistance is seriously lowered, ,especially in oxidizing media. Addi tions of up to 0.10% carbon, or slightly more, have no appreciable eifect on corrosion resistance of the'relatively ductile alloys (e.g. Alloy A), provided they are properly heat treated; however, an addition of 0.20% carbon reduces the corrosion resistance of all invention alloys in- Table IV Hardness Mean Corrosion Rate, Mils Per Year -48 Hour Test Periods Percent Alloy r BHN Re 65 N, 10 S, 30 S, 70 S. H. H,

Boil. Boil. Boil. 176 F. 125. F. 100 F.

23. 0 255 14 17 136 28 72 50, 27. O 270 28 8 11 114 14 Nil '35 30. 0 276 29 7 9 159 29 30 80 35.0 382 41 20 30 359 30 476 182 The amount of nickel in the alloy composition up to and including 10% does not significantly change the corrosion resistance of the alloys to various mineral acids such as nitric, sulfuric and hydrochloric. Best over-all results occur at about 2% nickel, and this is accordingly a preferred level. However, the hardness of the alloys drops cit sharply with increase-in nickel content, and at 10%, nickel this becomes approximately 200 BHN for the low molybdenum level alloys. By increasing the molybdenum content it is possible however to meet the desired BHN level even With 10% nickel, especially with work hardening, as by shot peening.

Corrosion rates in various media and hardness values for different levels of nickel content are shown inv Table V.

almost all media. Since the harder compositions wthin the invention cannot be adequately heat treated to elim inate corrosion-Producing carbide formation, at least not the molybdenum content-itself increases, and since such Table V Hardness Mean Corrosion Rate. Mils Per Year 5-48 Hour Test Periods Percent Alloy Ni BHN Rc 65 N, 10 S, 30 S, 70 S, 15 H. 20 H.

Boil. Boil. Boll. 176 F. 125 F. 100-F.

Nil 290 31 14 6 200 38 60 66 2.0 I 270 28 8 11 114 14 Nil 35 4. 0 260 24 11 6 152 3G Nil 72 5 5 l 0 1 5 16 25 23 1 8 140 18 6 6 *This value is recorded as Rockwell B.

Copper additions upto 4.5% do not noticeably affect the corrosion resistance except inmedium and concentrated sulfuric acid and in this corrosive a level of 1.75 to. 4.5% appears to providegood results, as seen in Table VI. Acceptable results for many uses are still obtainable at alow of 1.5% copper. Corrosion resista-nceto medium concentrations of sulfuric acid sufiers considerably as the copper content falls substantially below 1.5% or goes above 4.5%, as shown by Alloys R and T. Hardness of the alloys is slightly increased by removing all of the copper, but the slight benefit to hardness is more than carbide formation adversely affects corrosion resistance, proper. control of carbon'is essential. It may be noted that the use ofstrong carbide formers such as columbium to preferentially tie up excess carbon is a definitepossi bility. This mustbe balanced in turn against deteriora offset by the decreased corrosion resistance. Table VII compares the corrosion resistance of what Table VI Hardness Mean Corrosion Rate, Mils Per Year 5-48 Hour Test Periods Percent I Alloy Cu BHN Re N, 10 s, 30 s, s, 15 H, 20 H,

Boil. Boil. Boil. 176-F'.' 125 F. 100 F.

Nil 305 32 s 11 238' I 123 Nil 44 1. 285 30 s 12 103 25 N11 49 3. 25 210. 28 a 11 114 14 Nil 35 4. 5 285 24 6 s 83. 22 Nil 61 5. 0 27s 29 9 12 171 53 V 143 e9 '9 may be termed the normal carbon (0.06%) and low carbon (0.025%) alloys, as well as the effect at higher carbon contents.

used only for the milder applications. The efiect of the addition of elements like aluminum and titanium on the hardness of typical alloys within the invention range is Table VII Hardness Mean Corrosion Rate, Mils Per Year -48 Hour Test Periods All Perent BHN Re 65 N s, 30 s, 70 s, H, H,

Boil. Boil. Boil. 00 0. 125 F. 100" F.

. 025 376 40 12 12 130 20 Nil 35 06 375 40 35 10 130 27 Nil 40 07 270 28 s 11 114 14 N 1 35 10 255 27 10 5 s1 N11 35 20 270 28 35 7 153 25 20 20 The addition of iron to the novel alloys produces pro- 20 shown in Table X. gressive loss in corrosion resistance, although the change Table X is primarily significant only in the more severe media. Hardness BHN For best resistance a maximum of 6% iron has been Treatment found to be a desirable upper level to be observed and Alloy A AHOYH AHOY I again this may be resorted to for increased hardness 1n the 25 alloy if the accompanying loss in corrosion resfistafice can i l g fi f t I? nnea 2 rs.a 250 *.,wa er queue PE tolerated or can be compensited for by urt Annealfihrs. at 1200F.,air cool 265 265 265 ustment of other elements, e.g. nickel. Table VIII gives finneal 61 1 5. at 1400" 1 air 000%.-

275 227 290 5 nnealfi rs. at 1600 air coo 265 3 2 210 examples of relative corrosion resrstance at various 1mm Annea16hrs at 5 air c001 260 370 350 levels.

Table VIII Hardness Mean Corrosion Rate, Mils Per Year 5-48 Hour Test Periods Percent Alloy Fe BHN Re 65 N, 10 S, 30 s, 70 s, 15 H, 20 H,

Boil. 13511. Boil. 80 0. 125 F. 100 F.

2 270 2s 8 11 114 14 Nil 5 27s 29 s s 151 22 Nil 55 10 290 30 8 19 223 22 Nil 81 Aluminum and titanium do not influence the alloy microstructure in the as-cast condition or when the alloys are quench annealed from approximately 2250 F. But these additions do have a definite effect when the resulting alloys are annealed at temperatures within the range of 1400 F. to 1800 'F. Hardening is influenced by the formation of sigma phase in the standard or normal (0.06%) carbon alloy. Diffusion rates of chromium and other elements are slow in the presence of cobalt, and even annealing the base composition (Alloy A) at 1800 F. for 6 hours will cause very little sigma phase to form.

Table IX Hardness Mean Corrosion Rate, Mils Per Year 5-48 Hour Test Periods Percent Alloy Si BHN Re N, 10 S, 30 S, S, 15 H, 20 H,

Boil. Boil. Boil. 125 F. F.

0. 5 270 28 8 11 114 14 N11 35 1. 5 322 34 9 13 148 22 Nil 66 3. 0 362 39 43 30 388 33 93 182 Manganese in small percentages up to 2% may be used for deoxidizing and hardening. Other deoxidants such as calcium may be employed. Also, small additions of other hardening agents such as aluminum and titanium can be employed. In general such additions are kept low, not in excess of about 1% each, and preferably below this. Boron is also a good hardening agent but since it has a deleterious effect on corrosion resistance, it may be In fact it requires about 16 hours to cause a significant 0 change in hardness in such alloy. The addition of small amounts of aluminum and titanium, however, results in large quantities of Sigma phase being formed (see Alloys H and I of Table X) in relatively short times.

The shaft sleeve alloy (e.g. Alloy M) shows a tendency to increase in hardness in the 1400 to 1800 F. range, probably because of the observed greater amount of 1 1- secondary phasewhich accompanies this treatment. It isthis secondary. phase in the midrostructure which accounts for varied hardness.

The effects of heat treatment and cold working of the alloys of the invention are important to their corrosion resistance. Quenching from temperatures as high as 225 F. does not affect their onginalharduees; hOWfiYfiL alloys of the harder composition (e.g. shaft sleeve Alloy M) may tend to crack because of lower thermal shock resistance if a severe quench, such as water, is used. Therefore, it is generally recommended that these harder compositions normally be air cooled from the elevated temperature. ciently rapid to maintain the same corrosion resistance as obtained from water quenching.

Cold working the harder alloys rapidly lowers their corrosion resistance, even when a relatively mild amount of working, such as shot peening, is used. The more ductile compositions, however, are not appreciably affect: ed in corrosion resistance by. shot peening although more severe deformationyeg. a 5% reduction ofarea, produces very serious changes, particularly in respect to lowered resistance in hydrochloric acid. As mentioned, work hardening of the softer alloys (e.g. 270 BHN) provides an important method of increasingtheir usefulness.

There appears, on first glance, to be a general similarity in chemical composition between alloys of this invention and others in the prior art. However'thespecific (life ferences and refinemen s intr duced y he n esent ashi ing result in very decided differences in corrosion resistanceand hardness, which are unique and are not merely a matter of degree. It may be significant to point out in this connection that the teaching of the prior art in respect to obtaining increased hardness in the prior alloys is generally by way of promoting the formation of various carbides through addition of strong carbide formers in the presence of adequate carbon to permit this. This however adversely affects corrosion resistance. The discovery of the particular combination of alloying elements herein defined, on the other hand, provides a way of increasing hardness by the formation of other hard, inter-metallic compounds without sacrificing corrosionresistance. This is accomplished by carefully controlling the composition of. the alloys as described above.

What is claimed is:

1. Acast cobalt base alloyihaving substantiallythefollowing. composition:

cidental impurities.

Such air cooling has proved to be sufii-;

12 2. A cast cobalt base alloy as defined in claim 1, wherein'said alloy has aBrinell of about-375 3. A cast cobalt base alloy having substantially the following composition: 7

v 7 Percent by weight Cr 30. Mo 13: C ....V--.. Y 0.03 max. Cu V 3.25. Mn 0.5. Si 1 0.5. Fe [2.0. Ni 2.0. Co H Balance except for incidental impurities.

, 4. A corrosion. and abrasion resistant alloyas defined in claim 3, wherein the alloy has a Brinell'of at least 270. i V

5. A corrosion and abrasion resistant alloy. as defined in claim 3, wherein the alloy has a Brinell of at least 300.

6. A corrosion and abrasion resistant cobalt base alloy containing by Weight from approximately 26% to 32% chromium, from 8% to 15% molybdenum, not exceeding about-0.10%carbomfrom 1.5 to. 4.5%. copper; up to 10% nickel, .up to,6% ironand-conventional hard:

ening, deoxidizing and carbon stabilizing modifiers which aggregate. not more than about 3%, the balanceheing substantially all cobalt 'exceptxfor incidental impurities, said alloy. hayingit-s carbon content adjusted within the statedrange in accordance with the chromium-molybdenum content to maintain an as-cast hardness of at least about 270' Brinell, thecarbon beingv below about 0.03% at the higher levels of molybdenum and chromium to compensate for the increased tendency toward carbide formation and accompanying decreased corrosion resistance.

References Cite l y he Ex mi er UNITED STATES PATENTS DAVID, L. RECK, Primary Examiner.

UNITED STATESPATENT OFFICE CERTIFICATE OF CORRECTION Patent No. 3,177,074 April 6, 1965 Walter A. Luce et al.

It is hereb; certified that error appears in the above numbered patnt requiring correction and that the said Letters Patent should read as correctedbelow.

1, line 22, for "acid" read acids line 42, read novel column 4, lines 27 and 31, for

each occurrence, read ductility column 5,

Table I, in the footnote, line 5 thereof, for "test" read text column 8, line 26, for "wthin" read within column 10, Table X, fourth column, line 5 thereof, for

"210" read 310 Column for "noval" "dutility",

Signed and sealed this 21st day of September 1965.

(SEAL) Attest:

EDWARD J. BRENNER Commissioner of Patents ERNEST W. SWIDER AI testing Officer

Claims (1)

1. 6. A CORROSION AND ABRASION RESISTANT COBALT BASE ALLOY CONTAINING BY WEIGHT FROM APPROXIMATELY 26% TO 32% CHROMIUM, FROM 8% TO 15% MOLYBDENUM, NOT EXCEEDING ABOUT 0.10% CARBON, FROM 1.5% TO 4.5% COPPER, UP TO 10% NICKEL, UP TO 6% IRON AND CONVENTIONAL HARDENING, DEOXIDIZING AND CARBON STABILIZING MODIFIERS WHICH AGGREGATE NOT MORE THAN ABOUT 3%, THE BALANCE BEING SUBSTANTIALLY ALL COBALT EXCEPT FOR INCIDENTAL IMPURITIES, SAID ALLOY HAVING ITS CARBON CONTENT ADJUSTED WITHIN THE STATED RANGE IN ACCORDANCE WITH THE CHROMIUM-MOLYBDENUM CONTENT TO MAINTAIN AN AS-CAST HARDNESS OF AT LEAST ABOUT 270 BRINELL, THE CARBON BEING BELOW ABOUT 0.03% AT THE HIGHER LEVELS OF MOLYBDENUM AND CHROMIUM TO COMPENSATE FOR THE INCREASED TENDENCY TOWARD CARBIDE FORMATION AND ACCOMPANYING DECREASED CORROSION RESISTANCE.