US3159480A - Austenitic chromium-nickel stainless steels resistant to stress-corrosion cracking - Google Patents

Description

United States Patent 3,159,480 AHETENETIC CHRQMTUM-NTCKEL STAINLESS STEELS RElSTANT T0 STRESS-CURROSiGN CRACKS Harry R. Copson, Cranford, and Frances S. Lang, White House Station, NJL, assignors to The International Nickel Company, lino, New York, N.Y., a corporation of Delawa e No Drawing. Filed Nov. 28, 1962, Ser. No. 240,737 45 Claims. (Cl. 75-128) The present invention relates to austenitic. stainless steels and, more particularly, to overcoming the vexatious problem of stress-corrosion cracking in hot concentrated chloride solution to which the austenitic chromium-nickel stainless steels have been so susceptible heretofore.

The austenitic chromium-nickel stainless steels have been adopted for a vast number of different industrial and commercial applications, a result undoubtedly attributable, inter alia, to their relatively hi h strength properties at both room and elevated temperatures combined with their inherent ability to be fabricated with relative ease and their resistance to various corrosive environments.

=With respect to corrosion resistance, there appears to be general agreement that the ability of such steels to resist corrosive media stems from their passive nature, i.e., films (generally chromium oxide films) form on the surface of the steels and these films tend to protect the surfaces from penetration by corrodents or to at least greatly inhibit corrosive attack. Unfortunately, these passive films are under certain environments too easily ruptured or removed.

Corrosive activity of austenitic nickel-chromium stainless steels has been classified into four major categories: (1) general corrosion (usually a complete breakdown of the passive or protective film), (2) pitting corrosion (generally the occurrence of localized attack coupled with depth of penetration), (3) stress-corrosion cracking and (4) intergranular corrosion. The present invention is particularly concerned with the prevention of stress-corrosion caching, a condition which generally arises without any significant occurrence of general corrosion.

As is indicated in the authoritative treatise, Metals Handbook, 8th ed. (1961), p. 566, and by practically all other authorities who have grappled with the problem, all commercially available austenitic stainless steels are subject to stress-corrosion cracking. This problem has existed for at least twenty-five years and has been thought to have existed for a longer period except that it went unrecognized as such by earlier investigators. Notwithstanding the numerous articles or treatises (there are at least 250 references to the problem) which have been written concerning the subject, there is no single generally accepted theory which explains the why of this phenomenon. To be sure, the theories advanced have been numero-us and the following are illustrative of but a few of them: (1) the quasi-martensite theory, i.e., plastic defor mation results in the formation of a quasi-martensite which is preferentially attacked so as to result in cracking, (2) the plastic deformation theory which involves the formation of crack-sensitive paths wherein precipitates (carbides and nitrides) are formed at stressed areas which are thought to act as cathodes and the contiguous area of the steel as anodes of electrochemical cells entering into progressive corrosion and crack propagation, (3) the restricted slip theory which essentially is a theory of predicting susceptibility to stress-corrosion cracking, i.e., alloys which manifest cross slipping rather easily are considered to possess immunity to cracking, whereas alloys which inherently exhibit a high degree of resistance to slip are considered to be susceptible to cracking. The last might be considered somewhat as a special case of one of the two most prominent broad theoretical explanations ice presently in vogue, to Wit, the periodic electrochemicalmechanical mechanism. There are other explanations of the problem such as those involving surface films or inliving fl ons, but no one mechanism has achieved general acceptance.

While there may be no single theory which conclusively explains the fundamental nature of stress-corrosion cracking, it can be stated with assurance that if an austenitic chromium-nickel stainless steel is under stress, Whether the stress .be externally applied and/or residually induced by some processing operation such as quenching or cold- Working, and the steel is exposed to certain corrosive media, the conditions are ready-made for failure by stresscorrosion cracking. Chloride solutions, as indicated in the Metals Handbook, are the Worst offenders and these include the chlorides and other halides of magnesium, calcium, lithium, sodium, etc. Some other solutions have also been known to promote stress-corrosion cracking.

As there have been a number of theories to explain stress-corrosion, so too have there been a number of proposals for overcoming the problem, none of which has been completely sattisfactory since the difiiculty is presently ever the more to the forefront of long-time unresolved problems. For example, it has been suggested to avoid using a material susceptible to stress-corrosion cracking under conditions of intended use. In this connection, the use of ferritic stainless steels has been given consideration since the ferritic types are ostensibly rather immune to stress-corrosion cracking. To do so, however, would be to lose all the advantages of the austenitic stainless steels. It has also been suggested to avoid contact between austenitic stainless steels and chlorides. This is hardly a panacea, although it might circumvent the problem. The use of inhibitors, e.g.,' chromates, phosphates, and oxygen scavengers, has been urged, but this approach has not proven successful and brings into play other undesirable problems. Changes in design, processing conditions and/or fabrication have been proposed. For example, the use of high temperature stress-relieving treatments has been advanced in an effort to minimize the effect of residual stresses, it being considered that residual stress are more causative of stress-corrosion cracking than stresses externally applied. This approach has not proven to be the answer, partly because it is diflicult to reduce and to maintain the stresses at a low level. Another proposal is the use of the well-known cathodic protection principle. This is based upon the consideration that stress-corrosion cracking is of an electrochemical nature and the use of protective currents should assist in preventing cracking. The difficulties arising from this approach, including the impracticability of carrying it out on a commercial scale, are well known and will not be dwelt upon herein. In addition, stress-corrosion cracking still occurs. Other proposals include change of alloy composition, e.g., maintaining the carbon and nitrogen contents at very low levels in 18-8 stainless steel and cold working. Actually, this results in a ferritic stainless steel or one very high in ferrite.

Of all proposals, considerable interest has been shown with regard to purity of composition with emphasis on keeping the nitrogen (a known subversive performer) content at a low level, e.g., 0.01% to 0.02% or lower. This involves the use of materials of very high purity plus the application of expensive processing techniques, e.g., vacuum treatment. This approach appeals to theory but involves substantial additional costs which render it un economical from a commercial viewpoint. The austeuitic stainless steels are relatively expensive. To increase the cost through the use of high purity materials and expensive vacuum treatments would be quite disadvanta- 3 geous. In addition, there is considerable conflict as to whether purity is the answer. In any event, a more practical solution is required and it is to this aspect that the instant invention is directed.

Although many attempts were made to overcome the problem of stress-corrosion cracking in austenitic chromium-nickel stainless steels, none, as far as We are aware, was entire.y successful when carried into practice commercially on an industrial scale. This includes those proposals involving the beneficial effects of nickel, carbon or silicon and low amounts of nitrogen.

It has now been discovered that the problem of stresscorrosion cracking of austenitic ainless steels can be overcome or greatly minimized while utilizing standard air-melting techniques (as opposed to expensive vacuum techniques) and commercially pure grades of materials (as opposed to high purity materials) provided the steels are of specially controlled composition as set forth hereinafter. It has been further found that certain elements, notably phosphorus, promote the susceptibility of austenitic stainless steels to stress-corrosion cracking. However, in accordance with the invention, the presence of controlled amounts of phosphorus is not harmful and can be readily tolerated, thus avoiding questionable recourse to the use of pure materials or to undesirable processing techniques to rid the steels of phosphorus. In addition, because of the specially balanced composition of the steels within the invention, larger amounts of nitrogen can be present than otherwise might be the case.

It is an object of the invention to provide austenitic chromium-nickel stainless steels which greatly resist stresscorrosion cracking notwithstanding that the steels are highly stressed and in contact with corrosive solutions, including chloride solutions, which are known to be causative of stress-corrosion cracking.

Generally speaking and in accordance with the present invention, it has been found that stress-corrosion cracking of austenitic chromium-nickel stainless steels can be eliminated or greatly minimized with steels of the following most advantageous composition: about to about nickel, about 16% to about 25% chromium, up to 0.03% carbon, about 1.9% to about 2.5% silicon, phosphorus in an amount up to not more than 0.015%, nitrogen in an amount up to 0.035%, the sum of the phosphorus plus nitrogen being not more than 0.045 and the balance essentially iron. Phosphorus, among other elements, has been found to exert a pronounced adverse influence in the resistance of austenitic chromium-nickel stainless steels to stress-corrosion cracking. However, provided that the phosphorus content is controlled in accordance with the invention and provided the composition of the steels is otherwise critically balanced, the detrimental effect of phosphorus is overcome. This avoids use of processing techniques to eliminate phosphorus and is thus beneficial in maintaining the cost of producing the steels at an economic level. In addition, it has been found that what might have heretofore been considered as a relatively high and detrimental amount of nitrogen, e.g., 0.035%, can be present in the austenitic steels of the instant invention provided that the composition is critically balanced. A particular feature and advantage of the invention is that air-melting practices and commercial grades of materials can be employed. These factors greatly contribute to maintaining the economic competitiveness of the steels.

Satisfactory results can also be achieved with austenitic stainless steels of the following composition: at least 19% and up to about nickel, about 15% to about 30% chromium, up to 0.1% carbon, from 1.7% to about 2.5% silicon, phosphorus in an amount up to not more than 0.018%, nitrogen in an amount up to 0.045% but preferably not greater than 0.04% the sum of the phosphorus plus nitrogen being controlled such that it does not exceed 0.055%, and the balance essentially iron. When the silicon content is on the low side, i.e., 1.7%, it is advantageous that the carbon content be at least 0.04%.

In carrying the invention into practice, it is most important that special attention be given to controlling the nickel, silicon and phosphorus contents. As will be demonstrated hereinafter, should the nickel content fall below 19%, e.g., 18%, stress-corrosion cracking can occur. Amounts of nickel higher than 35% are unnecessary and only increase cost. The minimum silicon content is most important. In the most preferred embodiment of the invention the silicon content should be at least 1.9% particularly when the carbon content is at a level of 0.03% or less. A particular advantage of this silicon-carbon relationship is that the occurrence of detrimental transgranular cracks is not only greatly minimized, but intergranular corrosion is also avoided. Thus, subsequent solution-annealing coupled with rapid cooling treatments to minimize intergranular cracking are unnecessary. The silicon content can be lowered to 1.7%, but in so doing it is advantageous that the carbon content should be at least 0.04%. Although carbon can be present in an amount up to 0.1%, carbon contents above 0.05% are unnecessary. The maximum silicon content should not exceed 2.5 because of difiiculties in forging the steels. Phosphorus, a subversive, should not exceed 0.018%. It has been found, for example, that even in high-purity, vacuum-treated austenitic stainless steels, 0.023% phosphorus is most harmful since it actively promotes stresscorrosion cracking. As those skilled in the art will appreciate, the A181 commercial grades of austenitic stainless steels contain up to 0.2% phosphorus. Phosphorus contents of the order of 0.2% and higher have been known to impart beneficial effects to austenitic stainless steels. For example, it has been indicated that even 0.12% phosphorus improves stress-rupture properties at high temperatures. However, in accordance with the invention, such amounts of phosphorus are detrimental. Nitrogen is present in austenitic chromium-nickel stainless steels and is known to act subversively. There are prior references indicating that nitrogen should be maintained at a very low level; however, it has now been found that in the steels provided in accordance With this invention, comparatively high amounts of nitrogen, i.e., up to 0.045% can be present. However, the total phosphorus and nitrogen contents should not exceed 0.055% and it is most advantageous that the sum not exceed 0.045%. Additionally, if the phosphorus content is on the high side, it is preferable to maintain the nitrogen content on the low side and vice versa.

For the purpose of giving those skilled in the art a better understanding and/ or appreciation of the invention, there is given hereinafter data which are illustrative of the advantages embodied by the strainless steels having compositions within the invention. However, it should be mentioned that the test employed in connection with the data, to wit, immersion of stressed U-bend specimens in boiling concentrated magnesium chloride, has been criticized as being too severe on the grounds that such severe conditions would not be encountered in using commercial applications. In addition, the tests conducted in accordance with the invention have been continued for longer time periods than are usually considered necessary. If the strainless steels will withstand what is considered to be a severe test, then it would seem that the results thereby obtained would be a more critical indicator as to reliability, particularly from the commercial viewpoint.

In addition to using air melting practice in combination with commercial grades of materials, e.g., Armco iron and ferrochromium, specimens for test were also prepared using materials of high purity (electrolytic iron, electrolytic chromium) and vacuum melting technique. Because of availability, carbonyl or electro-nickel was used in the tests. Initially, ingots were prepared from compositions as set forth in the appropriate tables hereinafter.

The ingots were machined and ground to remove all surface defects and contamination, were heated in air to a temperature of about 2150" F. and were then forged in air to /1 inch slabs; All surfaces were again machined to a depth of A; inch to insure removal of any contaminating effect that might have been introduced from the forging operations.

A number of test specimens were machined from the forgings and subjected to the U-bend, magnesium chloride test. For the most part, the steels tested in the forged condition were of high purity and vacuum treated (Tables I and LA). Other specimens of the forged material were cut into inch x inch slices and cold-rolled about 7 40% and subjected to test in the cold-rolled condition. In

addition, cold-rolled specimens were also annealed in an argon atmosphere for one hour at 1950 F. and water quenched. These too were subjected to the U-bend, magnesium chloride test. In some cases the forgings were hot-rolled prior to cold rolling. All specimens were finish ground to about 40 microinches or finer to avoid possible surface contamination. Still other specimens were sensitized, i.e., subjected to an intermediate temperature treatment, e.g., about 1250 F., and then subjected to test in the sensitized condition. This was conducted with the View that austenitic stainless steels, as a commercial proposition, are aften subjected to the so-termed sensitization treatment (as by an intermediate heat treatment or welding operation) and this could aflfect susceptibility to stress-corrosion cracking. Thus, specimens weretested in at least one or more of four different conditions so as to intensify and diversify test conditions,

- thereby avoiding reliance on the possibility that stresscorrosion cracking might not occur in one condition without knowing whether it would occur in one of the others.

The test specimens were formed in the configuration of a U-bend by taking specimen samples of about 6 inches long, /2 inch wide andabout Vs inch thick, and bending them over a mandrel using a tensile test machine. This induced considerable stress in the specimens and the specimens were then placed in a vise and the legs of the notwithstanding that highly pure materials and vacuum techniques are employed. In this connection, vacuum melting, electrolytic iron, electrolytric chromium and carbonyl nickel were used in the preparation of the austenitic stainless steels having compositions shown in Table I.

TABLE I Chemical Composition The results of the test are given in Table IA.- The period to crack is given as of the day cracking was observed, but since inspection was usually made every second or third day, cracking could have taken place before the day indicated. For example, if cracking time is given as 15 days, cracking could have occurred between the 13th Average Alloy Forged Cold Rolled Cold Rolled sensitized Percent Life-to- No. and Annealed Cracked Crack in Days K 15 15 OK OK 3 4 11 30 58 21. 5 K 21 21 21 28 10 20 30 30 83 20 K 24 OK OK OK 21 21 23 30 25.4

2 3 3 6 OK 6 6 0 -6 92 6 1 1 1 1 1 1 1 l 1 100 1 1 1 1 1 1 1 1 1 1 100 1 3 6 6 6 6 3 3 3 3 100 4 lJ-bends were drawn about parallel to each other by tightening the vise. Bolts were inserted through holes near the ends of the legs of the specimens. Boiling concentrated magnesium chloride was used as the corrosive agent because of its known severity in promoting stresscorrosion cracking. The solutions were of about 42% concentration and the boiling point was adjusted to 154 C. Five liter flasks equipped with reflux condensers held the magnesium chloride. Four test specimens of at least one of the groups (forged, cold-rolled, annealed or sensitized) under test were suspended in each flask with the apices of the bends and about half of the legs immersed in the solution. The time required for cracking was the criterion employed in determining susceptibility or resistance to stress-corrosion cracking. The specimens were usually removed from test every second or third day and examined under a magnifier for cracks. When cracking was evident, the specimens were removed from test. Whenno cracking occurred in a period of about two weeks, fresh magnesium chloride solutions were used to replace the old solution. Generally, if no stress-corrosion The data in Tables I and LA illustrate the detrimental eiiect of phosphorus under What might be considered as the most favorable conditions, to wit, vacuum melting and the use of extremely pure materials. When the phosphorus content was rather low as in Alloys 1, 2 and 3 the average life-to-crack was comparatively high (2023 days). However, as the phosphorus content was increased, cracks becarne evident almost immediately. For example, Alloy No. 4 containing 0.023% phosphorus cracked in less than 6 days and Alloy No. 5 containing 0.065 phosphorus cracked in less than a day. Regarding Alloys Nos. 1 and 2, it can be stated that these austenitic steels were of high purity; nevertheless, they cracked and this is indicative that the use of high purity materials alone in producing austenitic chromium-nickel steels will not provide resistance to stress-corrosion cracking as contemplated by the present invention.

Not only must the phosphorus content be controlled, but, as mentioned above herein, the alloy composition as a whole must be critically balanced. This is particularly apropos regarding the nickel, silicon, nitrogen or nitrogen 7 plus phosphorus contents. As will be seen from data presented hereinbelow, small amounts of such elements can greatly alfect the results obtained. In this connection, several austenitic chromium-nickel stainless steels having compositions given in Table II were subjected to test and the results are given in Table II-A.

TABLE II Chemical Composition 1 3 any appreciable degree in (or at least are not intentionally added to) austenitic chromium-nickel stainless steels. However, to avoid stress-corrosion cracking, the amounts of bismuth, arsenic and antimony should not exceed 0.015%, 0.615%, and 0.015%, espectively. Molybdenum is quite often added to stainless steels for a number of reasons. In accordance with the invention, the amount of molybdenum should be controlled so as not to exceed 0.075% and preferably not more than 0.05%.

. 19 With respect to aluminum it has bee found that Alloy N Percent Percent Percent Percent Percent Percent amounts thereof in the range of FuJOllt 0.03% to about or P 0.66% adversely aiiect resistance to stress-corrosion 1 no cracking of austenitic chromium-nickel stainless steels. 3:3 815% 8:83 8:892 8 855 Amounts of aluminum below 0.03% or appreciably above 58-? 8%; 8-82; 8 8? 0.66%, e.g., 0.1% to 0.2%, can, in accordance with the 20.0 18.5 1121 0: 04 01007 0I022 lnvfintwn, be t a ed. @313 ii? g; 8-83: 8- 8 With regard to otherelements that can be present in 20.2 18.3 2. 50 0.02 0. 00s 0. 02 0 the stainless steels contemplated in accordance with 23 $1? 8:8; 8:38? 8:81? the invention, it is to be pointed out that manganese 16-10 119 0-016 (1007 0-010 should not be present in an amount greater than 0.7% 18.20 17.6 1.50 0.012 0. 008 0.010 I 18,29 65 M14 0,095 0,012 and advantageously sneuld not exceed 0.2% if stress-corg% 1;: 8-33 8-8}; 8- 8- rosion crac ing is to be avoided. in addition it is pre- 27. 41 17.4 as 0. 013 0. 013 0. 022 ferrcd that copper should not be present in an amount W03 Q0557 greater than 0.2% .id it is more referred to maintain co n-e '1 .ess ti i '0.

11111037 steels air melted using Armeo iron and ferrochromium except "3 l L n :i A v Alloy No. 20 in which electrolytic iron was used and Alloy No. 7 which IL 15 to 1153 (301135611 m L119 Prescnt mvemlon PTOVldes was vacuum melted using electrolytic iron and electrolytic chromium. austnific h i i k l 1 i 1 515C513 hi hl i TABLE lI-A Cracking Time Average Alloy No. Gold Rolled Cold Rolled and sensitized Percent Lite-to- .Annealc Crack Crack in Days 2 2 2 2 2 2 3 3 1 1 2 2 100 2 2 2 2 2 2 2 2 2 2 2 2 2 100 2 2 2 2 2 2 2 2 2 5 5 7 7 100 3.3 13 13 13 13 e 0 6 0 4 4 4 10 100 :13 OK OK OK or: 24 24 -4 OK OK OK OK OK 25 28.5 OK OK OK OK OK OK OK OK OK OK OK OK 0 OK OK OK OK OK OK OK OK or: OK OK 01g 0 30 OK OK OK or: OK OK OK OK OK OK OK on 0 30 01' OK OK OK OK OK OK OK OK OK OK OK 0 30 3 3 3 3 4 4 4 4 3 3 3 a 100 27 27 27 27 4 4 4 4 0 0 o s 100 12.3 21 30 OK OK 4 4 4 4 11 11 11 11 as 14.2 11 11 11 11 22 22 22 100 15.3 5 5 5 5 5 5 0 s 4 4 4 4 100 4.8 10 10 12 12 5 5 5 5 4 4 4 4 100 0.7 7 7 7 0 7 7 7 7 100 4.8 3 3 s 3 6 0 6 0 3 3 3 3 100 4.0

Tables II and Il-A illustrate the importance of main- 5() ant to stress-corrosion cracking when subjected to stress taining the required composition correlation among the elements which form the austen itic chromium-nickel stainless steels of the invention. Alloys Nos. 8 through 12 all manifested stress-corrosion cracking and in each instance the silicon content was at a level outside the present invention. When the silicon content was greater than 1.7% as in Alloys Nos. 13 through 16, stress-corrosion cracking did not occur. Alloys Nos. 17 through 20 reflect what happens when the nickel content is below the minimum of 19%. Cracking occurred in each of these steels. In this regard, it is to be observed that Alloy No. 20 which contained 18.20% nickel and 1.65% silicon, a steel quite close to those contemplated in accordance with the invention, cracked in each of the specimens tested. Increasing the nickel content above 20% as in Alloys Nos. 21, 22, 23 and 7 did not overcome the occurrence of stress-corrosion cracking.

In addition to the foregoing and in accordance with the invention, elements other than phosphorus have to be controlled in order to avoid the subversive influence which they exert with respect to the resistance of austenitic chromium-nickel stainless steels to stress-corrosion cracking. These elements include aluminum, molybdenum, bismuth, arsenic and antimony. As a general rule, the latter three elements are not normally found present to and in contact with halide solutions. Thus, the steels, in addition to being highly useful for high temperature applications, are especially suitable in the chemical processing industry Where stress-corrosion cracking has heretofore been so prominent. It is to be further pointed out that no impairment in mechanical properties of the austenitic steels within the invention is encountered. Actually, the mechanical properties show an improvement over the properties of the typical AISI grades of austenitic chromium-nickel stainless steels.

It is to be noted that the present invention is not to be confused with the published compositions of the welllrnown AISI grades of austenitie chromium-nickel stainless steels. These AISI grades are usually reported as nominal compositions. Metals Handbook lists the A181 grades in terms of nominal composition, but, as referred to before herein, indicates that they suffer from stress-corrosion crackin".

Although the present invention has been described in conjunction with preferred embodiments, it is to be understood that modifications and variations may be resorted to without departing from the spirit and scope of the invention, as those skilled in the art will readily understand. Such modifications and variations are considered For example, the aforementioned to be Within the purview and scope of the invention and appended claims.

We claim:

1. An austenitic chromium-nickel stainless steel highly resistant to stress-corrosion cracking when subjected to stress and in contact with halide solutions, said steel consisting essentially of about 20% to about 30% nickel, about 16% to about 25% chromium, about 1.9% to about 2.5% silicon, carbon in an amount up to 0.03%, phosphorus in an amount up to not more than 0.015%, nitrogen in an amount up to 0.035%, the sum of the phosphorus plus nitrogen not exceeding 0.045%, up to 0.7% manganese, and the balance essentially iron.

2. The austen-itic stainless steel set forth in claim 1 wherein the manganese content is not greater than 0.2%.

3. An austenitic chromium-nickel stainless steel highly resistant to stress-corrosion cracking when subjected to stress and in contact with halide solutions, said steel consistin g essentially of from 19% to about 35% nickel, about silicon, carbon in an amount up to 0.1%, phosphorus in an amount up to not more than 0.018%, nitrogen in an amount up to not more than 0.045%, the sum of the phosphorus plus nitrogen not exceeding 0.055%, up to 0.7% manganese, and the balance essentially iron.

4. The austenitic stainless steel set forth in claim 3 wherein the manganese content is not greater than 0.2%. 5. The austenitic stainless steel set forth in claim 3 wherein when the silicon content is at about 1.7%, the carbon content is at least about 0.04%.

6. The austenitic stainless steel set forth in claim 4 wherein when the silicon content is at about 1.7%, the carbon content is at least about 0.04%.

No references cited.

Claims (1)

1. AN AUSTENITIC CHROMIUM-NICKEL STAINLESS STEEL HIGHLY RESISTANT TO STRESS-CORROSION CRACKING WHEN SUBJECTED TO STRESS AND IN CONTACT WITH HALIDE SOLUTIONS, SAID STEEL CONSISTING ESSENTIALLY OF ABOUT 20% TO ABOUT 30% NICKEL, ABOUT 16% TO ABOUT 25% CHROMIUM, ABOUT 1.9% TO ABOUT 2.5% SILICON, CARBON IN AN AMOUNT UP TO 0.03%, PHOSPHORUS IN AN AMOUNT UP TO NOT MORE THAN 0.015%, NITROGEN IN AN AMOUNT UP TO 0.035%, THE SUM OF THE PHOSPHORUS PLUS NITROGEN NOT EXCEEDING 0.045%, UP TO 0.7% MANGANESE, AND THE BALANCE ESSENTIALLY IRON.