US2564004A - Structural steel - Google Patents

Description

' LlTeS.

Patented Aug. 14, 1951 STRUCTURAL STEEL James W. Halley, Dune Acres, 1nd,, assignor to Inland Steel Company, Chicago, 111., a corporation of Delaware N Drawing. Application May 17, 1949, Serial N0. 93,827

8 Claims. 1

This invention relates to steel and more particularly to structural steel of the semi-killed type.

In the past infrequent but often disastrous failures of certain steel structures have occurred under normal service or operating conditions with little or no explanation being found for such fail- This puzzling problem has occupied the attention of the steel making and allied industries for some time. In more recent years with the advent of Widespread use of welding in the construction of tanks, bridges, ships, and the like, the problem has become particularly acute. For example, in World War II certain welded merchant vessels operating under what were considered to be normal conditions experienced unusual structural casualties as a result of extreme fractures of the steel in the vessels, the fractures occurring in many cases with explosive suddenness and the fractured surfaces'exhibiting a characteristic brittle condition not ordinarily associated with the behavior of normally ductile ship steel.

Asa result of extensive investigations of the circumstances surrounding such failures, it has been determined that brittle fractures of-this type are caused by an adverse combination of two fac tors: (1) a notched condition in the steel structure resulting either from design or workmanship,

and (2) an inherent notch sensitivity of the steel itself under operating conditions. Although by careful design notching may beminimized, it is virtually impossible to avoid sharp corners in all instances, and even when the best has been done from a design point of View notches may occur through inadvertent or careless workmanship, such as by a welder accidentally striking an are on the edge of a highly stressed plate. Thus, it becomes important to control the second of the above-named factors, namely, the inherent notch sensitivity of the steel itself.

It is now known that any structural steel, when sharply notched and placed under stress, will fracture in a brittle manner if cooled to a low enough temperature. This general characteristic of structural steels becomes of extreme importance in connection with large all-welded structures, as mentioned above, because the welding technique produces, in effect,- a one-piece structure. As a result, a brittle fracture once initiated continues from one steel plate to the next plate and so on until there is complete failure of the structure. 1 In the case of riveted ships and tanks, brittle fractures of this type while serious do not usually result in such extreme failures because a crack starting in one plate merely runs out the opposite edge of the plate.

Although the desirability of employing structural steels having a high resistance to brittle fracture has thus become apparent, yet designers, engineers, and various groups having the responsibility of establishing standards for structural steel have been reluctant to add this requirement to the specifications of steels for use in ships, pressure vessels, and the like because no satisfactory method of consistently making steel to meet such requirements has heretofore been available to the steel industry. It has been common experience that duplicate heats of steel, made apparently by identical practices to the same chemical specifications and which show identical properties in all other tests, will show marked differences in their resistance to brittle fracture.

Accordingly, a primary object of my invention is to provide a structural grade steel having high notched toughness, i. e. a high resistance to brittle fracture over a substantial range of temperatures such as are encountered under normal operating conditions.

A further object of the invention is to provide a structural steel having superior resistance to brittle fracture but which is at the same time simple and inexpensive to manufacture.

Another object of the invention is to provide a stee] characterized by a high resistance to brittle fracture and suitable for use in large welded structures and for other structural purposes.

Broadly speaking, I accomplish the foregoing objects by providing a semi-killed steel, i. e. an incompletely or partially deoxidized steel, containing zirconium in amounts insufficient to completely deoxidize the steel but sufficient to impart thereto greatly improved resistance to brittle fracture. As described hereinafter in greater detail, the zirconium is present in the steel as zirconium oxide.

In a more specific embodiment, my invention comprises a structural-steel containing'not more than about .35 weight percent carbon and having added thereto sufiicient zirconium to give a Zirconium content in the steel of from about .005 weight percent to about .075 weight percent, the amount of zirconium being such that the steel is merely semi-killed or incompletely de= oxidized. As hereinafter described in detail, the zirconium is preferably added to the steel in the form of an alloy with iron and silicon.

Structural grade steels are normally used in the as-rolled condition, i. e. without being sub jected to heat treatment such as quenching or drawing. Thus, it is necessary that a structural steel possess the desired ultimate tensile and mechanical properties in its as-rolled condition without benefit of heat treatment. For these reasons, it has been common practice in the steel making industry to manufacture semi-killed structural steel largely according to tensile requirements with only slight emphasis on the chemical composition of the steel. For example, AS'I'M specification A'7-46 for structural steel for bridges and buildings includes a tensile range of 60,000 to 72,000 pounds per square inch along with other physical requirements but the only significant requirement as to chemical composition is a maximum phosphorus content of .04.10%, dependent upon the steel making process used, and a maximum of .05% for sulfur. In addition, structural grade steels are ordinarily restricted to the lower carbon content steels, for example, not more than about carbon. However, I have found that by regulating the chemical composition of semi-killed structural steels by the addition of zirconium, as hereinafter described in detail, the resistance of the steel to brittle fracture is greatly improved.

In order to evaluate the resistance of various steels to brittle fracture, the so-called tear test or Kahn tear test has been found to be the most satisfactory test method since its results correlate well with large scale plate tests. This test method is described in detail by N. A. Kahn and E. A. Imbembo in the Journal of the Amer- 1 ican Welding Society, vol. 27, page 169-S (April 1948). Briefly, the test utilizes a 3" x 5 test specimen of fullplate thickness, the specimen being provided with a transverse keyhole notch 1" deep and having a 1 mm. bottom radius. By means of pins passing through holes adjacent the opposite ends of the test specimen, the specimen is asymmetrically loaded in static tension to complete failure under controlled temperature conditions. base of the notch and continues across the specimen, the resultant fracture being either of the ductile or shear type, or of the brittle or cleavage type, or a combination of both. When both types of fracture are present, the initial fracture at the base of the notch is always ductile and changes to brittle at some intermediate point during the test.

The load-extension diagram obtained for the specimen indicates the energy input required to start initial failure and the energy required to propagate the fracture to completion. It has been found that the energy input required to propagate the fracture to completion, once failure has been initiated, is roughly proportional to the percentage of the fracture which exhibits the shear or ductile type failure. Accordingly, either the absorbed energy to propagate tearing or the percentage of fracture in the ductile mode may be utilized as a basis of evaluating the change from ductile to brittle type failure. For purposes of describing my invention, I have reported tear test ratings in terms of inches of ductile fracture, these values being readily determinable by visual inspection and measurement of the fractured test specimen. In general, 1 of ductile fracture is considered to be an acceptable minimum requirement for structural steel purposes whereas 2" of ductile fracture represents a completely ductile type failure. It is considered that less than 0.3" of ductile fracture represents an essentially brittle type failure.

As hereinbefore suggested, a significant characteristic of brittle type fracture is that although any structural steel becomes brittle if sharply Tearing of the specimen starts at the notched and stressed, whether or not the notched condition will result in brittle fracture is determined largely by the operating or service temperature. In other words, for any structural steel having a notched and stressed condition, there will be some temperature at which the steel will fracture in a brittle fashion, and the tendency toward brittle fracture increases markedly with decreasing temperature. For any given steel, the tear test may be utilized for the purpose of defining or locating the critical transition temperature range within which the mode of failure of the steel changes sharply from ductile type failure to brittle type failure.

The nature of the transition temperature phenomenon may be illustrated by the following tear test results which I have obtained on typical semi=k illed structural steel plates:

Table I Sample l 2 3 4 Plate Thickness (in Chemical Analysis (weight per cent):

Carbon .26 .18 .21 24 The steels employed in the tests reported in Table I are typical prior art structural steels and it will be seen that each of them gave essentially brittle fractures at a temperature of 80 F. which includes all the average room temperature range and lower. Even at F. the steels were not all acceptable, and it was necessary to heat to F. or higher in order to obtain essentially ductile fractures with all four of the steels.

I have found that by the incorporation of zirconium in a semi-killed structural steel, greatly improved resistance to brittle fracture is imparted to the steel as evidenced by a marked lowering of the transition temperature range at which the change from ductile to brittle type fracture occurs. Steel made according to my invention generally exhibit high resistance to brittle fracture at temperatures well below room temperature.

My steel is obtained preferably by adding zirconium to the steel in amounts insufficient to completel deoxidize the steel but sufficient to accomplish the purpose of improving the resistance of the steel to brittle fracture. In general, the steel being intended primarily for structural purposes should contain not more than about 35% carbon, and I have found that zirconium should be present in the final steel in an amount of from about .005% to about .0'75% and preferably from about .01% to about 065%. To obtain these amounts of zirconium in the final steel I have found that zirconium should be added to the liquid steel in amounts ranging from a minimum of about .25 lb. per ton of steel to a maximum of about 1 lb. per ton of steel. Larger additions than 1 lb. per ton may cause excessive deoxidation of the steel. In most cases the amount of zirconium added will be slightly greater than the final zirconium content of the steel because of the usual process and handling losses. In certain cases the. zirconium content of the final steel as determined by analysis of a particular sample may actually be greater than can be accounted for by the zirconium added. Such discrepancies are the result of segregation of the zirconium in the mold which sometimes occurs.

For purposes of convenience and economy, I prefer to add the zirconium to the steel in the form of a ferro-silicon-zirconium alloy containing from about to about zirconium and from about 40% to about silicon, the remainder being essentially iron. The amount of zirconium which can be added either as the metal itself or in the form of an alloy with iron and silicon depends upon the oxygen content of the liquid steel and the presence of other deoxidizing elements since it is an important feature of my invention, as hereinafter shown, that the steel must not be completely deoxidized. The amount of oxygen in liquid steel increases as the carbon content decreases so that as a general guide it may be stated that larger amounts of zirconium can be added to lower carbon steels. For example, I have found that a steel containing about'.20% carbon and no other deoxidizer may have added thereto on the order of 2 lbs. per ton of a ferrosilicon-zirconium alloy of the above-described type without danger of completely deoxidizing or killing the steel. For lower carbon steels, more of the alloy can be added but in no case is the amount added sufficient to completely kill the steel. It will be apparent that since silicon is one of the most common and effective deoxidizers for steel, precautions must be taken when adding zirconium in the form of a ferro-silicon-zirconium alloy to insure that the combined deoxidizing effect of the silicon and zirconium does not completely deoxidize the steel. At the same time, however, the zirconium addition must be sufficient to obtain the desired degree of improvement in resistance to brittle fracture.

The aforementioned range of zirconium content represents an overall operable range, but it will be understood that a particular zirconium content selected from Within this range must be correlated with the oxygen content and the effect of other deoxidizing elements which may be present in order to avoid complete deoxidation of the steel. It will also be understood that although with any zirconium content within the abovenamed range and with the steel in semi-killed condition an improved result will be noted with respect to resistance to brittle fracture as compared with the same steel containing no zirconium, yet the degree of improvement in resistance to brittle fracture will obviously not be the same for all zirconium contents. In general, larger amounts of zirconium within the above disclosed range will depress the transition temperature to a greater degree than smaller amounts of. zirconium within the same range.

In addition to a maximum carbon content of about 35% and a zirconium content as above disclosed, the steels of my invention should generally have a silicon content of not more than about .10% by weight in order to insure the semikilled or incompletely deoxidized condition of the steels. In addition, it is known that phosphorus has a detrimental effect on the resistance of steels to brittle fracture and accordingly to obtain the maximum benefits from the addition of zirconium the phosphorus content of the steel should not exceed about 025% by weight. Sulfur may be present in the normal amounts but for most structural steel purposes should not exceed about .05%.

. In order to illustrate the remarkable efiect oi' zirconiumaddition on semi-killed structural steels, the following tear test data are presented in Table II for two steels prepared according to the principles of my invention and also. for a semi-killed steel containing no zirconium included for purposes of comparison. Typical tensile test data are also given to demonstrate the general suitability of the steels for structural purposes. It will be understood that the remainder of the steel in each case is essentially iron with incidental impurities in the normal amounts.

Table II Sample 5 6 7 Plate thickness (in) 1% 1 1 Chemical Analysis (weight per cent):

Carbon 20 .16 Manganese 97 83 Phosphorus. .011 .016 Sulfur 029 .032 silicon .06 .06 Zirconium 1 043 015 Inches of Ductile Fracture:

t40 1.25 At0 F 0. 25 2.00 0.25 At-40 F 0.25 2. 00 2. 00 At F. 0. 50 2.00 2. 00 Tensile Properties:

Yield point (p. i.) 34, 200 41, 000 31,150 Tensile strength p. s. i.) 62, 000 64, 300 57,850 Per cent Elongation in 2 35. 8 27.0 8. 0

1 Present as and determined as ZrOz.

It will be seen from Table II that sample 5 containing no zirconium was completely unsatisfactory even at an operating temperature of 90 Fl, whereas samples 6 and '7 containing zirconium in accordance with the teachings of my invention gave satisfactory fracture results at 40 F. Samplefi which contained the greater amount of zirconium gave a completely ductile fracture even at 0 F.

Complete deoxidation or killing of steel by the addition of silicon, aluminum, zirconium, or other suitable deoxidizers apparently also improves the resistance of the steel to brittle fracture. Evidence of this fact is seen in the following tear test results obtained on steels which were completely deoxidized with silicon and aluminum:

However, by comparing the tear test results for killed steels as shown in Table III with the results obtained on samples 6 and 7 as shown in Table II, it will be seen that the zirconium-containing semi-killed steels of my invention are superior to the usual killed steels. In addition, completely deoxidized or killed steels are considerably more expensive to manufacture than the semi-killed or incompletely deoxidized steels which are normally used for structural purposes. The killed steels do not give oif gas during solidification and consequently contain large shrinkage cavities. Although, the existence of shrinkage cavities in killed steels maybe minimized by the use of hot tops, this technique materially increases the cost of casting and in any event the killed steels have substantially lower ingot-toproduct yields than the semi-killed steels. With the semi-killed steels a controlled amount of gas evolution during solidification compensates for shrinkage thereby avoiding shrinkage cavities an resulting in high ingot-to-product yields.

With the zirconium-containing steels of my invention, there are numerous factors which demonstrate that the zirconium content is insufiicient to combine with all of the oxygen in the steel and that the steels are, therefore, semikilled. Perhaps the most obvious indication of the semi-killed status of such steels is the fact that the solidified ingots do not contain large shrinkage cavities and the further fact that considerable gas is evolved during solidification. Moreover, the ingot-to-product yield is invariably high and the finished product is sound and free from the defects normally attributable to shrinkage cavities.

Further proof of the semi-killed condition of the steels of my invention is found by chemical analysis. In steels such as samples 6 and 7 in Table II above, the zirconium is all contained in the steel in the form of zirconium oxide (ZlOz) with no metallic or free zirconium being present. Obviously, if an excessive amount of zirconium were added to the steel, the total oxygen in the steel would be taken up and there would normally be an excess of free or metallic zirconium as a clear indication of the completely killed nature of the steel. The fact that in the steels of my invention all of the zirconium present is in the form of zirconium oxide is indicative of the fact that the amount of zirconium used is insufficient to combine with all the oxygen in the steel.

A suitable chemical technique for differentiating zirconium oxide from zirconium metal consists in boiling a sample of the steel in 3-normal hydrochloric acid. This treatment dissolves all of the metallic components of the steel but leaves certain oxides as a residue. I have found that in all cases the total zirconium added to the steels of my invention is completely accounted for in the zirconium oxideresidues remaining in the hydrochloric acid solution test.

As described above, steels prepared according to the present invention are preferably deoxidized to the desired semi-killed state by the addition of I zirconium or a ferro-silicon alloy thereof whereby to accomplish simultaneous deoxidation and zirconium addition. However, it is also within the scope of my invention to achieve the same result by a series or sequence of steps. ample, partial deoxidation may be accomplished in a first step by means of silicon, aluminum, or other suitable deoxidizer, and following this step zirconium may be added, either as the metal per se or as an alloy thereof, in amounts suflicient to complete the desired degree of deoxidation while at the same time incorporating the desired quantity of zirconium in the steel. Similarly, zirconium may be added in a first step followed by continued deoxidation in a second step with some other deoxidizing agent. In other words, a portion of the partial deoxidation of the steel may be carried out effective'iy in an independent step without concomitant addition of zirconium. However, in any case, the total deoxidizing effect as a result of the addition of zirconium and the other deoxidizing agents must be such as to avoid complete deoxidation of the steel. Zirconium additions may be made to the steel either in the ladle or in the molds with equally good results.

For ex- 1 ill] Although all of the zirconium present in the steels of my invention is in the form of zirconium oxide as a result of chemical combination with the oxygen present in the semi-killed steel, I have not been able to obtain satisfactory results by the direct addition of zirconium oxide (ZIOz) to liquid steel. Apparently, the physical chemistry of liquid steel is such that zirconium oxide is not readily soluble therein and cannot be incorporated therein by simple addition of the oxide.

However, addition of metallic zirconium apparently results in the formation of zirconium oxide in situ and the oxide thus formed is held in solid solution or otherwise distributed throughout the steel.

In addition to the improvements in the resistance to brittle fracture of semi-killed steels by the addition of zirconium, it has also been found that small but significant improvements can be achieved by increasing the manganese content of a semi-killed steel above the amount ordinarily used in structural steels. The data presented in Table IV show tear test results as obtained on semi-killed steel samples of normal manganese Although it will be seen from Table IV that increasing the manganese content from .46% to 1.01% results in an improvement in the resistance to brittle fracture, the final results are still not satisfactory for general structural use. However, in order to realize the maximum benefits with respect to resistance to brittle fracture by controlling the chemical composition of semikilled steel, it is within the scope of my invention to employ substantial amounts of manganese in order to augment or supplement the effect of zirconium. In general, when manganese is so employed in conjunction with zirconium, the manganese content may range from about .25% to about 1.25% by weight and it is preferred to employ a manganese content in the upper portion of this range, for example, from about 175% to about 1.25%.

It will be seen that the zirconium-containing semi-killed steels of my invention are far superior to the usual semi-killed structural steels and are substantially better than the killed steels which re also more expensive and more difficult to manufacture. The addition of zirconium in amounts insufiicient to completely deoxidize the steel, with the result that all of the zirconium is present in the steel in the form of zirconium oxide, results in an important and marked decrease in the transition temperature range within which the change from ductile to brittle type fracture occurs. Thus, it will be apparent that the effective operating temperature of the steel with respect to its resistance to brittle fracture is greatly increased by employing zirconium in accordance with the principles hereinbefore described.

The present invention provides a simple and inexpensive method of manufacturing steel having superior resistance to brittle fracture in the tremendous quantities needed for structural purposes. Relatively small amounts of zirconium are required and since zirconium addition may be accomplished satisfactorily by the use of ferrosilicon-zirconium alloys, the net increase in cost per ton of steel is slight. Even when the manganese content of the steel is increased over the normal manganese level for structural steels in order to obtain the combined benefits of manganese and zirconium, the further increase in cost is not prohibitive, and the total increase in cost of the finished product as compared with the cost of the conventional structural steels is generally less than $1.00 per ton. Furthermore, the steels of my invention require no special handling in the rolling mills nor are special steel making tech niques needed. In all cases, it has been found that in open hearth practice the yields obtained with the steels of my invention are entirely commensurate with the yields obtained on the regular grades of structural steel.

I claim:

1. A semi-killed steel characterized by its high resistance to brittle fracture and containing not more than about 35% carbon and from about 005% to about .075% zirconium as zirconium oxide.

2. A semi-killed steel containing not more than about 35% carbon and having zirconium added thereto in an amount sufiicient to partially deoxidize the steel but insufiicient to completely kill the same, said zirconium being present in the steel as zirconium oxide and in an amount of from about 005% to about 075% whereby to impart to the steel a high resistance to brittle fracture.

3. A semi-killed structural steel characterized by high resistance to brittle fracture as evidenced by a Kahn tear test value of at least 1.5 inches of ductile fracture, said steel containing not more than about .35% carbon and from about 005% to about 075% zirconium as zirconium oxide.

4. A semi-killed structural steel characterized by high resistance to brittle fracture at temperatures below room temperature and containing not more than about .35% carbon and from about 01% to about 065% zirconium, the latter being present in the steel as zirconium oxide.

5. A semi-killed steel characterized by high 10 resistance to brittle fracture and containing not more than about 35% carbon, not more than about .10% silicon, and from about 005% to about 075% zirconium, the latter being present in the steel as zirconium oxide.

6. A semi-killed steel characterized by high resistance to brittle fracture and containing not more than about 35% carbon, from about 25% to about 1.25% manganese, and from about 005% to about .075 zirconium, the latter being present as zirconium oxide.

7. A semi-killed steel characterized by high resistance to brittle fracture and containing not more than about .35% carbon, from about .75% to about 1.25 manganese, and from about 005% to about .075 zirconium, the latter being present as zirconium oxide.

8. A semi-killed steel characterized by high resistance to brittle fracture and containing carbon in an amount not more than about 35%, silicon in an amount not more than about .10%, not more than about 025% phosphorus, not more than about 05% sulfur, from about 25% to about 1.25% manganese, and from about 005% to about 075% zirconium present in the steel as zirconium oxide, the remainder being iron and incidental impurities in the normal amounts.

JAMES W. HALLEY.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 1,546,831 Becket July 21, 1925 1,550,489 Becket et a1 Aug. 18, 1925 1,550,490 Becket et al Aug. 18, 1925 2,046,168 Kinzel et al June 30, 1936 2,069,758 Hayes et al Feb. 9, 1937 FOREIGN PATENTS Number Country Date 477,982 Great Britain Jan. 3, 1938 574,427 Germany Apr. 13, 1933 OTHER REFERENCES Metals Handbook, 1939 edition, pages 786 and 787. Published by the American Society for Metals, Cleveland, Ohio.

Claims (1)

1. A SEMI-KILLED STEEL CHARACTERIZED BY ITS HIGH RESISTANCE TO BRITTLE FRACTURE AND CONTAINING NOT MORE THAN ABOUT .35% CARBON AND FROM ABOUT .005% TO ABOUT .75% ZIRCONIUM AS ZIRCONIUM OXIDE.