US3401036A

US3401036A - Valve steel

Info

Publication number: US3401036A
Application number: US660007A
Authority: US
Inventors: Dulis Edward John; Kasak August
Original assignee: Crucible Steel Company of America
Current assignee: Crucible Materials Corp
Priority date: 1967-08-11
Filing date: 1967-08-11
Publication date: 1968-09-10
Anticipated expiration: 1985-09-10

Description

Sept. 10, 1968 E. J. DULIS ETAL VALVE STEEL Filed Aug. ll, 1967 7// //////A ,sa 7//A 7///// .3,. /V/A MV/// a 7/ mV E; m7/ e 7/ m7////////A a. 27/4 mV// %d a 3 .3

Ia ooo 5 ooo /N VEN rcms.

d n a mA p LS m UA 0 D M &W .A DM MU w D E United States Patent Office 3,40l,036 Patented Sept. 10, 1968 3,401,036 VALVE STEEL Edward John Dulis, Pittsburgh, and August Kasak,

Bridgeville, Pa., assignors to Crucible Steel Company of America, Pittsburgh, Pa., a Corporation of New Jersey Contimation-in-part of application Ser. No. 616,793, Feb. 17, 1967. This application Aug. 11, 1967, Ser. No. 660,()07

4 Claims. (Cl. 75-128) ABSTRACT OF THE DISCLOSURE This patent relates to valve steels having an excellent combination of properties with respect to (1) lead oxide corrosion resistance, (2) high-temperature creep strength, (3) machinability, (4) ingot soundness and (5) hot-workability. More specifically, this patent discloses that the above-mentioned desirable results are obtained in certain chrornium-containing austenitc stainless steels by using compositions that exhibit the following features: (1) the use of a carbon content within a restricted range of about 022% to O.5%, (2) the use of silicon contents of 0.05 to 0.25 (3) the use of a strong-carbide-forming strengthening element such as columbium, titanium, tantalum, molybdenum, or tungsten, singularly or in combination, columbium being preferred, within the range of 0.02 to 0.75%, (4) the optional use of boron up to 0.025% and preferably 0.00005 to 0.025%, (5) the use of nitrogen in amounts up to 025%, (6) the use of sulfur in substantial amounts up to O.25%, (7) the use of Copper in amounts up to about 3%, (8) chromium of 16 to 22%, (9) nickel 4 to (10) manganese of2 to 8%, (11) with the sum of manganese plus nickel within the range of 8 to and the sum of carbon plus nitrogen within the range of .025 to 06%.

This application is a continuation-in-part of our earlier filed copending application Ser. No. 616,793 filed Feb. 17, 1967, now abandoned, which is in turn a continuation-in-part of our application Ser. No. 417,240, filed Dec. 9, 1964, which is now abandoned.

This invention relates to high-temperature stainless steel especially suitable for the manufacture of valves, valve parts, and other internal-combustion-engine Components intended for use in hot, Corrosive atmospheres.

Among the objects of our invention is the provision of strong, tough, and durable' stainless steel, and various internal-combustion-engine valves and engine Components fashioned of the sarne for elevated-temperature use, which steel products function in a highly satisfactory manner in such fields as passenger-car, truck, aircraft, diesel and marine-Vessel engine use, and which offer both great hardness at the high temperatures encountered in such use and substantial resistance, in the heated condition, to hot Corrosive atmospheres such as those containing the combustion products of anti-knock gasolines, e.g., of the tetraethyl lead variety. In additon, the steel is characterized by good workability at the prevailing hardness levels.

Numerous metallurgical compositions have hitherto been used, or proposed for use, in making valves for internal-combustion engines, and especially for internalcombustion-engine eXhaust valves, but the steels of our invention possess a combination of properties that, for certain purposes, makes them more suitable than any prior-art steel. Two important properties sought in a stainless steel for internal-combustion-engine exhaust valves are (1) resistance to corrosion by lead-containing fuelcombustion products, and (2) high-temperature stretch resistance (creep strength). Perhaps the foremost of the steels presently used for such purposes is a steel containing about 21% chromium, 4% nickel, 9% manganese, 050% carbon, 0.40% nitrogen, 0.25 maximum silicon, and the balance substantially iron. Of the iron-base materials hitherto known which exhibit sufficient creep strength to suit them for use in exhaust valves, this steel, called 21-4N, eXhibits the greatest resistance to corrosion by lead-containing fuel-combustion products. The lead oXide crucible test is presently the best laboratory method for preliminary evaluation of the corroson resistance of eXhaust-valve alloys. In the performance of this test, Specimens of 0.444 inch in diameter by 0.444 inch long are weighed, placed in a magnesia crucible with 40 grams of lead oxide, and heated to 1675 F. for one hour. Specimens are then cooled to room temperature, cleared, and weighed again. The loss in weight per unit surface area expresses the relative corrosion resistance of the alloy. In such a test, the above-mentioned 21-4N alloy exhibits a weight loss of 15 to 19 grams per square decimeter per hour. The preferred steels according to our invention exhibit weight-loss values considerably lower, on the order of 8 to 10 grams per square decimeter per hour, yet our steels are not only less expensive to produce than 'the "21-4N alloy but also exhibit a stretch resistance (creep strength) better than that of many other known alloys for use in exhaust valves and better than or not substantially inferior to the values for the "21-4N" alloy.

A further disadvantage of the "21-4N alloy now in common use is that it is quite difficult to machine. It is known that sulfur improves machinability of austenitic stainless steel and other steels, and that at the same time, it tends to make the steel diflicult to hot work; with the known "21-4N" steel, only relatvely small amounts, such as about 0.07%, of sulfur can be added to improve the machinability of the steel while still avoiding undue hot working difficulties. Machinability is an important property in a valve steel; the usual practice in the making of exhaust valves for automotive engines involves a twostep hot forming of a cylindrical slug of steel to give it the usual stern and poppet valve head, permitting the hot-formed piece to cool, and then finish machining the valve stem and face to a desired size. A final machining operation is important in order to obtan valves of uniform dimensions. The temperature to which the cylindrcal slugs are heated necessarily varies somewhat, so that the amount of contraction obtained upon cooling difl'ers from one valve to the next; moreover, scale forms on the valves as they are cooling and must be removed. The prior art has not furnished a valve steel with good creep strength and good lead oxide corroson resistance and, at the same time, anything more than marginal machinabiliity.

Yet another advantage of the steel of our invention over conventional valve steels, and particularly 21-4N, is that our steel exhibits good hot-hardness in combination with good hot-workability. In the manufacture of valves for this purpose, it is customary to form the same by a hot-forging operation, typically at a. temperature of, for example, 2100 F. Hence, it is desirable that the steel have good hardness at these high temperatures for good performance during service and yet permit forging to the required valve shape without undue resistance to material flow.

The steel of our invention overcomes the above-discussed disadvantages with respect to conventional valve steels and provides other advantages, which will be apparcnt from the following description and drawings, in which:

FIGURE 1 is a bar graph showing the desirable combination of hardness and hot-wor-kability achieved with our steel by observing our limitations with respect to nitrogen content as compared with conventional stainless steels, particularly 21-4N";

FIGURE 2 is a graph showing the effects of manganese and nickel With respect to providing the steel of our invention with a fully austenitic structure;

FIGURE 3 is a graph showing the relative austenite and martensite percentages at varying levels of manganese and nickel; and

FIGURE 4 is a bar graph showing the effects of various strengthening elements on the creep strength (stretch resistance) of steels of our invention that were austenitized at 2150 F. for one hour, water quenched and tempered at 1400 F. for eight hours.

New austenitic stainless valve steels having improved hot-hardness, hot-workability and machinability as compared with the 21-4N steel and other conventional stainless steels, together with lead oxide corrosion resistance and creep strength at least about as good as 2l-4N steel, and in some instances, substantally better results as respects those properties, are obtained by making a steel with about 16% to 22% chromium, suflicient nickel and manganese to make the steel completely austenitic, 0.05 to 025% silicon, about 0.22% to 05% carbon, and a strengthening addition of a carbide-forming element selected from the group consisting of columbium, tantalum, molybdenum, tungsten, and titanium, with Columbium being preferred, and balance of the steel being iron, impurities in small amounts not substantally detrimentally affecting the properties, and in some instances, small amounts of other elements such as boron, copper, and sulfur, which may be added in particular cases to improve the properties of the steel further in certain respects. Use of the carbon-content range indicated above is important for obtaining lead oxide corrosion resistance and strength.

The optional addition of boron to the steel of the invention serves to improve the creep strength. The addition of nitrogen in combination with a controlled carbon content provides an optimum combination of hot-workability and hardness. Sulfur may be added to improve machinability. Also, if desired, copper may be added to further improve the creep strength.

The following Table I presents the composition limits, in percent, of steels according to our invention:

TABLE I Carbon 0.22-0.5 Chromium 16-22 Silicon 0.05-0.25 Nickel 4-15 Manganese 2-8 Manganese--nickel 8-20 Boron Up to 0.025 Sulfur Up to 0.25 Nitrogen Up to 0.25 Carbon--nitrogen 0.25 to 0.6 Copper Up to 3 At least one carbide-forming element selected from the group consisting of columbium,

molybdenum, tungsten, titanium and tantalum 0.02-0.75 Iron Balance In addition, boron may be present in an amount within the range of 0.00005 to 0.025%. Also, the preferred carbide-forming element is Columbium, and such may preferably be present in an amount within the range of 0.05 to 025%. Columbium may be present within this range with tantalum within the range of 0.005 to 0.25 To insure good hot-Workability, it is preferred that the nitrogen limit be 02%. The steel in accordance with the invention is fully austenitic and substantally free of ferrite.

With carbon contents too low, the strength of the steel is inadequate, and with carbon contents too high no benefit is observed and the steel tends to become diflicult to hot-work. The optimum carbon content for this purpose is about 0.357. The carbon content also influences the lead oxide corrosion resistance, optimal values being observed at aicarbon content of about 036%. It must also be borne in mind that carbon is an austenite-promoting element. With carbon contents too low, the steel Will become partially ferritic unless compensating increases are made in the contents of other austenite-promoting elements.

As respects silicon, the resistance of the steel to corrosion in the presence of the products of combuston of leadcontaining fuels is substantally lowered if too much silicon s used. Hence, for use in valves of engines burning leaded fuels, steels containing not over 025% silicon, and still more desirably not more than 012%, are preferred.

As respects chromium, the steel must contain sufficient of this element to insure the stainless character of the steel, but the use of an amount too great must be avoided in order to prevent the formation of substantial amounts of ferrite and/or sigma phase, which would detrimentally affect the creep resistance of the steel.

As respects nickel, the steel must contain an amount sufiicient to insure that the steel will possess an essentially austenitic microstructure, which is required for good creepstrength properties. The use of amounts of nickel greater than those indicated above is without substantial benefit justifying the addition and is contraindicated on account of cost.

As respects Columbium, a small addition of this element, or of another appropriate carbide-forming element is required in order to achieve adequate creep strength. The use of amounts of Columbium greater than that indicated above is considered undesirable because there is no observed improvement in properties that would justify the cost of using more Columbium.

When they are used in place of, or in addition to, Columbium, the elements titanum, tantalum, molybdenum, and tungsten are generally to be employed in substantially equivalent amounts on the basis of their atomic weights. That is, 0.1% columbium' is equivalent -to 0.05 of titanium, or 0.2% of tantalum, or 0.1% of molybdenum, or 02% of tungsten. As will be explained below, however, our data with respect to titanium show an anomaly; when titanium is used, it is preferred to use a relatively large amount, such as about 036%.

As respects boron, the use of this element is optional, but to provide better creep strength, it is desrable to include an effective addition of boron. The use of quantities greater than those indicated above may cause incipient gram-boundary melting at temperatures of use. Amounts of boron up to 0.1% have a favorable effect upon creep strength, but boron additions of more than about 0.04% tend to worsen resistance of the steel to corrosion by lead oxide, so that it is preferable that the boron not exceed 0.025

The prim'ary role of manganese in the steel, and particularly the combination of manganese and nickel, is to control the phase relationsh'ps so that an essentially fully austenitic structure is maintained during heat treatment and service. The presence of delta ferrte and/or martensite in the structure has an adverse effect on the stretch resistance of the steel. For this purpose, at least 4% nickel is necessary to prevent the formation of delta ferrite at solution annealing temperatures, e.g., 2150 F., (FIGURE 2). Manganese alone is not effective for this purpose; however, manganese is effective for the purpose of stablizing the austenitic structure against transformation to martensite when used in combination with nickel (FIGURE 3). Without manganese, at least about nickel is necessary for a stable structure. With 4% nickel, at least 5% manganese is needed for a stable structure.

The nitrogen content, in combination with the carbon content, must be controlled to achieve the desired combination of hardness and hot-workability (FIGURE l). Although the hardness of the steel increases With increasing nitrogen at a given carbon level, e.g., about 030%, when the nitrogen content is increased substantally above 0.25 percent the hot-workability of the steel is substantially decreased. As may be seen from FIGURE 1, at

about 021% nitrogen, in combination with a carbon content of about 0.31%, the hot-workability of the steel is far superior to that of the conventional "21-4N. When nitrogen is increased to 031%, with carbon remaining substantially the same, the hardness is only slightly im- The efi'ect of using various strengthening elements (co lumbum, vanadium, titanium, zirconium, tantalum, molybdenum, tungsten, and aluminum) upon the lead oxide corrosion -resistance of steels of this kind was studied,

5 comparison being made with the known 21-4N steel proved, eg., increased from 28 R to 30 R however, and with a steel consisting of 21% chromium, 6% mangathe hot-workability is decreased to that typically exhibited nese, 6% nickel, 035% carbon, balance: iron, i.e., a steel by 21-4N." At even higher nitrogen levels, eg., 0.53 that would be in accordance with the instant invention, and 063%, the hot-workability is substantially decreased save for the fact that it contains no strengthening-element over that of 21 4N." addition. This steel, designated 1733, is not suitable as a These data presented in FIGURE 1 were obtained by valve steel, exhibiting a plastic strain of 1.62% after 100 investigating the following steels: hours at 1350 F. under a Stress of 16,000 p.s.i. as comc Mn s Si Cr Ni Cb :B

0.34 5. 95 0. 061 0. 07 20.00 5. 87 11 0. 001 0.31 5. 97 0. 055 0. 20. 24 5. 80 .08 0.001 0. 28 5. 85 0. 059 0. 17 20. 10 5. 87 10 0.001 0.31 5. 88 0. 058 0. 14 20.34 5. 80 11 0.001 0. 31 6. 04 0. 057 0. 14 20. 14 5. 93 11 0. 001 0.37 5. 74 0. 050 o. 14 20.29 5. 95 11 0.001 0. 30 6. os 0. 055 0. 16 20. 29 5. 87 09 0. 001 o. 04 1. 32 0. 020 o. 51 18.53 8. 87 0. 05 1. 07 0. 020 0. 28 16. 44 13.24 o. 57 9. 0. 040 0. 15 20. 87 4. 23

25 pared with a plastic strain of 0.56% in the "21-4N" steel For the purposes of this evaluation, 0.65 in. square by under the same conditions; but its performance in lead 6 in. long specimens of the above steels were hot-rolled oxide corrosion resistance tests gives a valuable indcation in six passes to nominally 0.11 in. long by 1 in. wide of the effect of the addition of the various strengthening flats by a reduction about 25% per rolling pass by the elements mentioned above upon the property of lead use of 13.5 in. diameter work-rolls. The roll-separating oXide corrosion resistance. Our tests show that all the force, entry thickness and exit width of the specimens above-mentioned strengthening elements were detrimental, were measured. These data were used to obtain the flowmore or less, to the lead oxide corrosion resistance. Tistress values presented in FIGURE 1 of the drawings by tanium and zirconium impaired the lead oxide corrosion means of a standard computer technique utilizing the wellresistance of the 1733 steel to the point Where there was known Hill-Longman roll-separating force equation, no improvement in comparison with the 21-4N steel. which is a measure of hot-workability. The result with an addition of aluminum in the amount of As respects sulfur, a noticeable improvement in maabout 0.23% was nearly as unfavorable. Tantalum addichinability is obtained with amounts of sulfur as small as tions of 0.12% and 1.57% gave the same result, as did 005%. Although amounts as large as 025% can be used, the additions of vanadium. Additions of molybdenum, the improvement in machinability obtained by adding tungsten and columbium were less detrimental and left ulfur levels ff at li h above hi amount H the steel with lead oxide corrosion resistance substantially amounts in substantial excess of said amount should not superior to that Of the S b d As explained above, only with small additions of co- Copper improves the creep strength of the steels. lumbum was there a substantial improvement over It has been discovered, moreover, that additions of 21-4N steel in creep strength, and the same small addivanadium, aluminum and zirconium worsen the creep tion of columbium at the same time, yielded a substantial strength of the steel, in comparison to th same steel withimprovement in lead oxide corrosion resistance, for which out any addition of a strengthening element as may be reason columbium is a preferred strengthening element. seen from the data presented in FIGURE 4. These steels, It is important, however, not to lose sight of the fact that i i composition only with respect to the strengththe improved lead oxde corrosion resstance of 1733 and ening element, Were creep tested for relative Stretch resstance at a temperature of 1350 F. under 16,000 p.s.i. for 100 hours. In these tests the resulting plastic strain (permanent set) is the evaluating criterion. Low plasticstrain values (in percent) are indicative of good creep strength (Stretch resistance).

the same steel with an addition of 011% columbium is also in part dependent upon the low nitrogen content of those steels. i

Data tending to substantiate the statements made above are presented in the following tables, from which the value of the invention will be further apparent.

TABLE II.-COMPOSITIONS AND RESULTS OF LEAD-OXIDE CORROSION TEST C Mr S Si Cr Ni Cb B N 0. 20 l. 33 1 NA 0. 08 20. 15 15. 85 0. 13 0. 001. 0. 004 0. 29 1. 41 1 NA 0. 10 20. 15 15. 95 0. 12 0. 001 0. 007 0. 36 1. 35 1 NA 0. 10 20. 32 15. 80 0. 09 0. 001. 0. 005 0. 45 1. 29 1 NA 0. 07 19. 92 15. 58 0. 10 0. 002 0. 005 0. 71 0. 86 1 NA 0. 06 19. 97 15. 52 0. 11 0. 001 0. 003 1. 00 1. 33 1 NA 0. 05 20. 02 15. 94 0.12 0. 001 0. 005 0. 36 1. 27 1 NA 0. 10 20. 28 15. 95 0. 13 1 NA 0. 005 0. 36 1. 30 1 NA 0. 08 20. 10 16. 10 1 NA 1 NA 0. 004 0. 35 1. 42 0. 013 0. 09 20. 52 3. 08 0. 12 0. 003 0. 01 0. 34 1. 38 0. 011 0. 10 20. 46 5. 06 0. 15 0. 003 0. 01 0. 33 1. 40 0. 014 0. 09 20. 38 6. 98 0. 07 0. 002 0. 01 0. 32 1. 34 0. 011 0.10 20. 40 8. 92 O. 08 0. 003 0. 01 0. 33 1. 37 0. 011 0. 07 20. 34 11.10 0. 10 0. 002 0. 02 0. 33 1. 40 0. 012 0. 08 20. 52 12. 94 0. 10 0. 002 0. 02 0. 33 1. 30 0. 012 0. 08 20. 32 15. 0. 11 0. 005 0. 02 O. 34 14. 13 0. 008 0. 08 20. 12 3. 03 0. 06 O. 001 0. 01 0. 34 10. 00 0. 014 0. 10 19. 96 G. 88 0. 08 0. 001 0. 02 0. 33 6. 20 0. 016 0. 08 20. 06 6. 30 0. 08 0. 002 0. 02 0. 33 6. 11 0. 055 0. 09 19. 98 5. 71 0. 02 0. 001 0. 03 0. 34 5. 95 0. 061 0. 07 20. 00 5. 87 0. 11 0. 001 0. 03 0. 35 6. 03 0. 052 0. 12 19. 5. 0. 62 0. 001 O. 04 0. 35 6. 16 0. 049 0. 08 19. 80 5. 83 I 0. 01 0. 001 0. 02

TAB LE II- Continued Heat No.

A mmumummnn 65 42 0 & 22

91 &mn u 11 3221 0000 03&

11111 00000 00000 0 nmn 0 0 29771 IUOOIM Plastic strain after Rupture Heat No. lfrhr. life, hours exposure, percent n% a a a a aga ag NNNNNNNNNNNNNNNNNNNNNNNNN aaLauaqmzLoaaLLuuLNaNNua TABLE III.-RESULTS OF CREEP AND CREEP-RUPTURE TESTS AT 1,350 F. UNDER A STRESS OF 16,000 P.S.I.

Plastc strain after Rupture 100-hr. life, hours exposure, pel-cent Heat No.

Weight Ioss after immersion in Heat No. molten PbO for 1 hr. at 1675 F. (gJ

All spociens were solution-created at 2150 F. for 1 hour, water 8 hours, except for the "21-4N" steel,

N. D.-Not determed.

065 90789080896667987mwm9 1 None added. 2 +Contaned also 192% copper.

TABLE lL continucd Weight loss after immersion in Heat No. molten PbO or 1 hr. at 1675 F. (g./m.

quenched, and aged at 1400 F which was aged for 16 hogrs.

l Not determined. 2 Estmated.

TABLE IV.-RESULTS OF TIARDNESS TESTS Hardness (Rockwell C") Heat No. Compositional Solution Solution treated variable(s) treated 1 plus aged at 1,400

F. 8 hr.

3230 020% C l 8 3229. 029% C i 9 12 036% C. 8 20 045% C 14 18 031% C.. 20 22 100% 16 21 No B 17 3261 No Gb, No 12 14 2150 F. 1 hour, water-quench.

From the foregoing results, it is apparent that the preferred steels of our invention exhibit a weight loss, after immersion in molten lead oxide for 1 hour at 1675 F., of less than two-thirds that exhibited by the comparison steel 21-4N." Table II shows the eifect of carbon on the lead oxide corrosion resistance, a maximum in such resistance being had at a carbon content of about 0.38% or 040%. Further inspection of data on Heat Nos. l7l8A, 1716, 1719A, and 1720 in Table II shows the effect of the boron content on the lead oXide corrosion resistance. Table II further shows the eifect on the lead oXide cor rosion resistance that is obtained when columbium is used.

Table III shows that, unexpectedly, the creep strength reaches a maximum at about 038% or 0.40% carbon, which is the carbon range providing maximum resistance to lead-oxide corrosion. With such a carbon content it is possible to obtain a steel that closely approximates the known 2l-4N steel in creep strength, yet is consider&- bly superior in its rcsistance to corrosion by lead oxide. This is especially remarkable because, as aforesaid, the 21-4N steel has never, among the iron-base materials that have suificient all-around oxidation resistance and high temperature creep strength to be useful in automotive and other internal-combustion-engine valves, been hitherto surpassed in lead oXide corrosion resistance. Moreover, comparison of Heat No. l714A with Heat Nos. 1716 and 1718A reveals that creep strength can be obtained without any boron additions, but Heat Nos. 1719A and 1720A show that a still further improvement in strength is obtained when, as is preferred, boron is added.

Table III and FIGURE 4 also show that favorable creep strength results can be obtained by the addition of titanium, tantalum, molybdenum, or tungsten in place of columbium, but not with vanadium, zirconium, or aluminum. Table III also shows that the creep strength may be improved by additions of copper (Heat No. 1814) or nitrogen (Heat No. 1X35).

As respects machining, the machinability of a steel in accordance with the invention was evaluated in direct comparison with the 21-4N" valve steel in a lathe cut-oil? operation. The steel thus tested constitutes a preferred composition: 036% carbon, 603% manganese, 005% sulfur, 0.007% phosphorus, 007% silicon, 19.5l% chromium, 584% nickel, 012% columbium, 0.002% boron, 007% nitrogen, balance iron. A standard AISI Type Ml high speed tool was used to cut off one-inch diameter bar stock of each steel in the solution-treatedand-aged condition. The toolng and machining conditions were as follows:

MACHINING CONDITIONS Lathe American P a c e m a k e r 14 X 30 (equipped with a variable-speed drive for spindle-speed control).

skilled in the art.

10 TOOLING Tool material AISI type Ml Tool blank size in 6 x 1 X %s Tool geometry:

Rake angle degrees- 0 Front relief angle do 11.5

Side relief angle do- 3 Nose angle do 11.5

The tool used to cut 21-4N steel failed before even one cut was completed. In contrast, the tool used to cut the steels of this invention Was still cutting satisfactorily after 32 cuts, when the test was discontinued.

Additional data demonstrating that the preferred steel of our invention is superior to 21-4N steel in hot-workability is presented in the following Tables V and VI. Said preferred steel had the composition of Heat No. 1604 in Table II and the 2l-4N" steel had the composition given in that table. One-inch-diameter bars of each steel were heated to 2150F. and then hot-rolled to fiats /s-inch thick, using a substantially identical pass schedule. The roll-separating force and the exit thickness after each pass were determined, and the thickness reduction achieved in each pass was calculated. Table V presents the data in the order in which they were obtained, and Table VI shows the relationship between roll-separating force and thickness reduction achieved for each of the steels.

TABLE V.-HOT-WORKABILITY TESTS Exit thiekness, Thickness reduc- Roll-separating nch tion, percent force, 1,000 lb. Pass No.

Our steel "21-4N" Our steel "2l-4N Our steel "2l-4N" steel steel steel TABLE VI.-HOT-WO RKABILITY-COMPARISON OF THICK- NESS REDUCTION AND ROLL-SEPARATING FORCE Our steel "21-4N" steel Thickness reduo- Roll-separating Thiekness'edue- Roll-separating From table VI it will be seen that in each case with the "21-4N" steel it required a greater roll-separating force to achieve a given amount of thckness reduction.

Further data dem-onstrating the improved workability of preferred steels in accordance with *the instant invention is presented below.

The hot ductility of a preferred steel in accordance with our invention, in comparison with that of the known 21-4N steel, was further evaluated in an elevated-temperature tension-impact test. In this test, cylindrcal button-head specimens (0.375-nch-diameter by 0.8125-inchlong gage section) were heated to a test temperature, rapidly transferred to an impact machine with appropriate special fixtures, and then broken in tension under impact loading. The striking hammer weighed 60` pounds and was dropped from a height of 48 inches. The initial strain rate achieved in this test is about 256 in./in../sec., which is comparable -to that which occurs during hammer-forging Operations. The primary criterion for evaluating results in this test is the fracture ductility, which is expressed as the percent reduction of area of the test specmen, higher percentages being more favorable. The force in foot-pounds required to break the specimen is also measured; in this test, lower numbers indicate superior ductility and workability.

ll The two steels tested in ithis way had the following Chemical compositions:

HEAT NO. 70413 C 0.36 Mn 6.03 Si 0.07 S 0.050 P 0.007 Cr 19.51 Ni 5.84 Cb 0.12 B 0.002 N 0.07

21-4N" STEEL C 0.53 Mn 9.28 Si 0.20 S 0.070 'P 0.020 Cr 20.55 Ni 3.67 Cb N.A. B N.A. N 0.39

1 None added.

TABLE vlL-RESULTS OF TENSION-IMPACT TESTS Reduction of area, Ruptufre 1sltrengtl,

Heating tem- Test tempe'apcrcent peratu'e, F. ture, F.

The above results indicate that the preferred steel of our invention is superior in hot Workability, as respects both -mill processing and valve manufacture. A hot-working temperature of about 2150 F. to 2250 F. appears most favorable.

There are other properties that may prove important in a valve steel. Among these are (1) tensile strength at room temperature and elevated temperature, (2) Charpy V-notch impact strength, (3) amenabili ty to hardfacing, and (4) flash-weldability. Preferred steels of our invention exhibit at least satisfactory properties in the above 'espects. In room-temperature Charpy V-notch impact tests, the steel of Heat No. 7 0413 had impact values of 75 and 62 ft.-lb. in the sol'ution-treated-and-aged condition, as Compared with values of 4 and 5 ft.-1b. for !the "21- 4N" steel with which it was compared.

While we have shown and described certain embodiments of our invention, we intend to cover as well any change or modification therein which may be made without departing from the spirit and scope of the invention.

What is claimed is:

1. An austenitic stainless steel characterized by -good high-temperature creep strength along with good hardness and formability consisting essentially of, in percent, about .22 to .5 carbon, 16 to 22 chromium, .05 to .25 silicon, 4 to 15 nickel, 2 to 8 manganese, with the combination of manganese plus nickel ranging from 8 to 20, up to .025 boron, up to .25 sulfur, up to .25 nitrogen, with !the combination of carbon plus nitrogen ranging from .25 to .6, up to 3 copper, .02 to .75 of at least 'one element selected from lthe group consisting of columbium, molybdenum, tungsten, titanium and tantalum, and the balance iron, said stainless steel being fully austenitic and substantially free of ferrite.

2. The stainless steel of claim 1 further characterized by boron of .00005 to .025.

3. An austenitic stainless steel characterized by good high-temperature creep strength along with good hardness and formability consisting essentially of, in percent, about .22 to .5 carbon, 16 to 22 chromium, .05 to .25 silicon, 4 to 15 nickel, 2 to 8 manganese, with the combination of manganese plus nickel of 8 to 20, up to .025 boron, up to .25 sulfur, up` to .25 nitrogen, with the combination of carbon plus nitrogen of .25 to .6, up to 3 Copper, .05 to .25 columbium, 'and the balance iron, said stainless steel being fully austenitic and substantially free of ferrite.

4. The stainless steel of claim 3 further chartacterized by tantalum of .005 to .25.

References Cited UNITED STATES PATENTS 2,429,800 10/1947 Briggs -128 2,495,731 1/ 1950 Iennings 75-128 2,657,130 10/1953 Jennings 75-128 2,671,726 3/1954 Jennings 75-128 2,745,777 5/1956 Clarke 75-128 XR 2,799,577 7/ 1957 Schempp 75--128 2,839,391 6/1958 Loveless 75-128 XR 2,892,703 6/1959 Furman 7S-128 3,1S2,934 10/1964 Lula et al. 75-128 XR 3,310,396 3/1967 Hochmann 75-128 FOREIGN PATENTS 831,728 3/ 1960 Great Britain.

12,814 7/1963 Japan. 12,502 8/1962 Japan.

HYLAND BIZOT, Pr'mary Examier.

P. WEINSTEIN, Assistant Exam iner.