THE
WORKING OF STEEL

ANNEALING, HEAT TREATING
AND
HARDENING OF CARBON AND ALLOY STEEL

BY

FRED H. COLVIN

Member American Society of Mechanical Engineers and Franklin Institute; Editor of the American Machinist, Author of "Machine Shop Arithmetic," "Machine Shop Calculations," "American Machinists' Hand Book."

AND

K. A. JUTHE, M.E.

Chief Engineer, American Metallurgical Corp. Member American Society Mechanical Engineers, American Society Testing Materials, Heat Treatment Association, Etc.

SECOND EDITION
THIRD IMPRESSION

McGRAW-HILL BOOK COMPANY, Inc.
NEW YORK: 370 SEVENTH AVENUE
LONDON: 6 & 8 BOUVERIE ST., E. C. 4

PREFACE TO SECOND EDITION

Advantage has been taken of a reprinting to revise, extensively, the portions of the book relating to the modern science of metallography. Considerable of the matter relating to the influence of chemical composition upon the properties of alloy steels has been rewritten. Furthermore, opportunity has been taken to include some brief notes on methods of physical testing—whereby the metallurgist judges of the excellence of his metal in advance of its actual performance in service.

NEW YORK, N. Y.,

August, 1922.

PREFACE TO FIRST EDITION

The ever increasing uses of steel in all industries and the necessity of securing the best results with the material used, make a knowledge of the proper working of steel more important than ever before. For it is not alone the quality of the steel itself or the alloys used in its composition, but the proper working or treatment of the steel which determines whether or not the best possible use has been made of it.

With this in mind, the authors have drawn, not only from their own experience but from the best sources available, information as to the most approved methods of working the various kinds of steel now in commercial use. These include low carbon, high carbon and alloy steels of various kinds, and from a variety of industries. The automotive field has done much to develop not only new alloys but efficient methods of working them and has been drawn on liberally so as to show the best practice. The practice in government arsenals on steels used in fire arms is also given.

While not intended as a treatise on steel making or metallurgy in any sense, it has seemed best to include a little information as to the making of different steels and to give considerable general information which it is believed will be helpful to those who desire to become familiar with the most modern methods of working steel.

It is with the hope that this volume, which has endeavored to give due credit to all sources of information, may prove of value to its readers and through them to the industry at large.

July, 1921.

THE AUTHORS.

CONTENTS

INTRODUCTION

THE ABC OF IRON AND STEEL

In spite of all that has been written about iron and steel there are many hazy notions in the minds of many mechanics regarding them. It is not always clear as to just what makes the difference between iron and steel. We know that high-carbon steel makes a better cutting tool than low-carbon steel. And yet carbon alone does not make all the difference because we know that cast iron has more carbon than tool steel and yet it does not make a good cutting tool.

Pig iron or cast iron has from 3 to 5 per cent carbon, while good tool steel rarely has more than 1¼ per cent of carbon, yet one is soft and has a coarse grain, while the other has a fine grain and can be hardened by heating and dipping in water. Most of the carbon in cast iron is in a form like graphite, which is almost pure carbon, and is therefore called graphitic carbon. The resemblance can be seen by noting how cast-iron borings blacken the hands just as does graphite, while steel turnings do not have the same effect. The difference is due to the fact that the carbon in steel is not in a graphitic form as well as because it is present in smaller quantities.

In making steel in the old way the cast iron was melted and the carbon and other impurities burned out of it, the melted iron being stirred or "puddled," meanwhile. The resulting puddled iron, also known as wrought iron, is very low in carbon; it is tough, and on being broken appears to be made up of a bundle of long fibers. Then the iron was heated to redness for several days in material containing carbon (charcoal) until it absorbed the desired amount, which made it steel, just as case-hardening iron or steel adds carbon to the outer surface of the metal. The carbon absorbed by the iron does not take on a graphitic form, however, as in the case of cast iron, but enters into a chemical compound with the iron, a hard brittle substance called "cementite" by metallurgists. In fact, the difference between the hard, brittle cementite and the soft, greasy graphite, accounts for many of the differences between steel and gray cast iron. Wrought iron, which has very little carbon of any sort in it, is fairly soft and tough. The properties of wrought iron are the properties of pure iron. As more and more carbon is introduced into the iron, it combines with the iron and distributes itself throughout the metal in extremely small crystals of cementite, and this brittle, hard substance lends more and more hardness and strength to the steel, at the expense of the original toughness of the iron. As more and more carbon is contained in the alloy—for steel is a true alloy—it begins to appear as graphite, and its properties counteract the remaining brittle cementite. Eventually, in gray cast iron, we have properties which would be expected of wrought iron, whose tough metallic texture was shot through with flakes of slippery, weak graphite.

But to return to the methods of making steel tools in use 100 years ago.

The iron bars, after heating in charcoal, were broken and the carbon content judged by the fracture. Those which had been in the hottest part of the furnace would have the deepest "case" and highest carbon. So when the steel was graded, and separated into different piles, a few bars of like kind were broken into short lengths, melted in fire-clay crucibles at an intense white heat, cast carefully into iron molds, and the resulting ingot forged into bars under a crude trip hammer. This melting practice is still in use for crucible steel, and will be described further on page 4.

THE WORKING OF STEEL

ANNEALING, HEAT TREATING AND HARDENING
OF
CARBON AND ALLOY STEEL

CHAPTER I

STEEL MAKING

There are four processes now used for the manufacture of steel. These are: The Bessemer, Open Hearth, Crucible and Electric Furnace Methods.

BESSEMER PROCESS

The bessemer process consists of charging molten pig iron into a huge, brick-lined pot called the bessemer converter, and then in blowing a current of air through holes in the bottom of the vessel into the liquid metal.

The air blast burns the white hot metal, and the temperature increases. The action is exactly similar to what happens in a fire box under forced draft. And in both cases some parts of the material burn easier and more quickly than others. Thus it is that some of the impurities in the pig iron—including the carbon—burn first, and if the blast is shut off when they are gone but little of the iron is destroyed. Unfortunately sulphur, one of the most dangerous impurities, is not expelled in the process.

A bessemer converter is shown in Fig. 1, while Fig. 2 shows the details of its construction. This shows how the air blast is forced in from one side, through the trunnion, and up through the metal. Where the steel is finished the converter is tilted, or swung on its trunnions, the blast turned off, and the steel poured out of the top.

OPEN HEARTH PROCESS

The open hearth furnace consists of a big brick room with a low arched roof. It is charged with pig iron and scrap through doors in the side walls.

Through openings at one end of the furnace come hot air and gas, which burn in the furnace, producing sufficient heat to melt the charge and refine it of its impurities. Lime and other nonmetallic substances are put in the furnace. These melt, forming a "slag" which floats on the metal and aids materially in the refining operations.

In the bessemer process air is forced through the metal. In the open-hearth furnace the metal is protected from the flaming gases by a slag covering. Therefore it is reasonable to suppose that the final product will not contain so much gas.

A diagram of a modern regenerative furnace is shown in Fig. 3. Air and gas enter the hearth through chambers loosely packed with hot fire brick, burn, and exit to the chimney through another pair of chambers, giving to them some of the heat which would otherwise waste. The direction is reversed about every twenty minutes by changing the position of the dampers.

CRUCIBLE STEEL

Crucible steel is still made by melting material in a clay or graphite crucible. Each crucible contains about 40 lb. of best puddled iron, 40 lb. of clean "mill scrap"—ends trimmed from tool steel bars—and sufficient rich alloys and charcoal to make the mixture conform to the desired chemical analysis. The crucible is covered, lowered into a melting hole (Fig. 4) and entirely surrounded by burning coke. In about four hours the metal is converted into a quiet white hot liquid. Several crucibles are then pulled out of the hole, and their contents carefully poured into a metal mold, forming an ingot.

If modern high-speed steel is being made, the ingots are taken out of the molds while still red hot and placed in a furnace which keeps them at this temperature for some hours, an operation known as annealing. After slow cooling any surface defects are ground out. Ingots are then reheated to forging temperature, hammered down into "billets" of about one-quarter size, and 10 to 20 per cent of the length cut from the top. After reheating the billets are hammered or rolled into bars of desired size. Finished bars are packed with a little charcoal into large pipes, the ends sealed, and annealed for two or three days. After careful inspection and testing the steel is ready for market.

THE ELECTRIC PROCESS

The fourth method of manufacturing steel is by the electric furnace. These furnaces are of various sizes and designs; their size may be sufficient for only 100 lb. of metal—on the other hand electric furnaces for making armor-plate steel will hold 40 tons of steel. Designs vary widely according to the electrical principles used. A popular furnace is the 6-ton Heroult furnace illustrated in Fig. 5.

It is seen to be a squat kettle, made of heavy sheet steel, with a dished bottom and mounted so it can be tilted forward slightly and completely drained. This kettle is lined with special fire brick which will withstand most intense heat and resist the cutting action of hot metal and slag. For a roof, a low dome of fire brick is provided. The shell and lining is pierced in front for a pouring spout, and on either side by doors, through which the raw material is charged.

Two or three carbon "electrodes"—18-in. cylinders of specially prepared coke or graphite—extend through holes in the roof. Electrical connections are made to the upper ends, and a very high current sent through them. This causes tremendous arcs to form between the lower ends of the electrodes and the metal below, and these electric arcs are the only source of heat in this style of furnace.

Electric furnaces can be used to do the same work as is done in crucible furnaces—that is to say, merely melt a charge of carefully selected pure raw materials. On the other hand it can be used to produce very high-grade steel from cheap and impure metal, when it acts more like an open-hearth furnace. It can push the refining even further than the latter furnace does, for two reasons: first the bath is not swept continuously by a flaming mass of gases; second, the temperature can be run up higher, enabling the operator to make up slags which are difficult to melt but very useful to remove small traces of impurities from the metal.

Electric furnaces are widely used, not only in the iron industry, but in brass, copper and aluminum works. It is a useful melter of cold metal for making castings. It can be used to convert iron into steel or vice versa. Its most useful sphere, however, is as a refiner of metal, wherein it takes either cold steel or molten steel from open hearth or bessemer furnaces, and gives it the finishing touches.

As an illustration of the furnace reactions that take place the following schedule is given, showing the various stages in the making of a heat of electric steel. The steel to be made was a high-carbon chrome steel used for balls for ball bearings:

6-TON HEROULT FURNACE

The deoxidizing slag is now formed by additions of lime, coke and fluorspar (and for some analyses ferrosilicon). The slag changes from black to white as the metallic oxides are reduced by these deoxidizing additions and the reduced metals return to the bath. A good finishing slag is creamy white, porous and viscous. After the slag becomes white, some time is necessary for the absorption of the sulphur in the bath by the slag.

The white slag disintegrates to a powder when exposed to the atmosphere and has a pronounced odor of acetylene when wet.

Further additions of recarburizing material are added as needed to meet the analysis. The further reactions are shown by the following:

To form white slag there was added:

During the white-slag period the following alloying additions were made:

The furnace was rotated forward to an inclined position and the charge poured into the ladle, from which in turn it was poured into molds.

Electric steel, in fact, all fine steel, should be cast in big-end-up molds with refractory hot tops to prevent any possibility of pipage in the body of the ingot. In the further processing of the ingot, whether in the rolling mill or forge, special precautions should be taken in the heating, in the reduction of the metal and in the cooling.

No attempt is made to compare the relative merits of open hearth and electric steel; results in service, day in and day out, have, however, thoroughly established the desirability of electric steel. Ten years of experience indicate that electric steel is equal to crucible steel and superior to open hearth.

The rare purity of the heat derived from the electric are, combined with definite control of the slag in a neutral atmosphere, explains in part the superiority of electric steel. Commenting on this recently Dr. H. M. Howe stated that "in the open hearth process you have such atmosphere and slag conditions as you can get, and in the electric you have such atmosphere and slag conditions as you desire."

Another type of electric furnace is shown in Figs. 7 and 8. This is the Ludlum furnace, the illustrations showing a 10-ton size. Figure 7 shows it in normal, or melting position, while in Fig. 8 it is tilted for pouring. In melting, the electrodes first rest on the charge of material in the furnace. After the current is turned on they eat their way through, nearly to the bottom. By this time there is a pool of molten metal beneath the electrode and the charge is melted from the bottom up so that the roof is not exposed to the high temperature radiating from the open arc. The electrodes in this furnace are of graphite, 9 in. in diameter and the current consumed is about 500 kw.-hr. per ton.

One of the things which sometimes confuse regarding the contents of steel is the fact that the percentage of carbon and the other alloys are usually designated in different ways. Carbon is usually designated by "points" and the other alloys by percentages. The point is one ten-thousandth while 1 per cent is one one-hundredth of the whole. In other words, "one hundred point carbon" is steel containing 1 per cent carbon. Twenty point carbon, such as is used for carbonizing purposes is 0.20 per cent. Tool steel varies from one hundred to one hundred and fifty points carbon, or from 1.00 to 1.50 per cent.

Nickel, chromium, etc., are always given in per cent, as a 3.5 per cent nickel, which means exactly what it says—3½ parts in 100. Bearing this difference in mind all confusion will be avoided.

CLASSIFICATIONS OF STEEL

Among makers and sellers, carbon tool-steels are classed by "grade" and "temper." The word grade is qualified by many adjectives of more or less cryptic meaning, but in general they aim to denote the process and care with which the steel is made.

Temper of a steel refers to the carbon content. This should preferably be noted by "points," as just explained; but unfortunately, a 53-point steel (containing 0.53 per cent carbon) may locally be called something like "No. 3 temper."

A widely used method of classifying steels was originated by the Society of Automotive Engineers. Each specification is represented by a number of 4 digits, the first figure indicating the class, the second figure the approximate percentage of predominant alloying element, and the last two the average carbon content in points. Plain carbon steels are class 1, nickel steels are class 2, nickel-chromium steels are class 3, chromium steels are class 5, chromium-vanadium steels are class 6, and silico-manganese steels are class 9. Thus by this system, steel 2340 would be a 3 per cent nickel steel with 0.40 per cent carbon; or steel 1025 would be a 0.25 plain carbon steel.

Steel makers have no uniform classification for the various kinds of steel or steels used for different purposes. The following list shows the names used by some of the well-known makers:

Passing to the tonnage specifications, the following table from Tiemann's excellent pocket book on "Iron and Steel," will give an approximate idea of the ordinary designations now in use:

CHAPTER II

COMPOSITION AND PROPERTIES OF STEEL

It is a remarkable fact that one can look through a dozen text books on metallurgy and not find a definition of the word "steel." Some of them describe the properties of many other irons and then allow you to guess that everything else is steel. If it was difficult a hundred years ago to give a good definition of the term when the metal was made by only one or two processes, it is doubly difficult now, since the introduction of so many new operations and furnaces.

We are in better shape to know what steel is than our forefathers. They went through certain operations and they got a soft malleable, weldable metal which would not harden; this they called iron. Certain other operations gave them something which looked very much like iron, but which would harden after quenching from a red heat. This was steel. Not knowing the essential difference between the two, they must distinguish by the process of manufacture. To-day we can make either variety by several methods, and can convert either into the other at will, back and forth as often as we wish; so we are able to distinguish between the two more logically.

We know that iron is a chemical element—the chemists write it Fe for short, after the Latin word "ferrum," meaning iron—it is one of those substances which cannot be separated into anything else but itself. It can be made to join with other elements; for instance, it joins with the oxygen in the air and forms scale or rust, substances known to the chemist as iron oxide. But the same metal iron can be recovered from that rust by abstracting the oxygen; having recovered the iron nothing else can be extracted but iron; iron is elemental.

We can get relatively pure iron from various minerals and artificial substances, and when we get it we always have a magnetic metal, almost infusible, ductile, fairly strong, tough, something which can be hardened slightly by hammering but which cannot be hardened by quenching. It has certain chemical properties, which need not be described, which allow a skilled chemist to distinguish it without difficulty and unerringly from the other known elements—nearly 100 of them.

Carbon is another chemical element, written C for short, which is widely distributed through nature. Carbon also readily combines with oxygen and other chemical elements, so that it is rarely found pure; its most familiar form is soot, although the rarer graphite and most rare diamond are also forms of quite pure carbon. It can also be readily separated from its multitude of compounds (vegetation, coal, limestone, petroleum) by the chemist.

With the rise of knowledge of scientific chemistry, it was quickly found that the essential difference between iron and steel was that the latter was iron plus carbon. Consequently it is an alloy, and the definition which modern metallurgists accept is this:

"Steel is an iron-carbon alloy containing less than about 2 per cent carbon."

Of course there are other elements contained in commercial steel, and these elements are especially important in modern "alloy steels," but carbon is the element which changes a soft metal into one which may be hardened, and strengthened by quenching. In fact, carbon, of itself, without heat treatment, strengthens iron at the expense of ductility (as noted by the percentage elongation an 8-in. bar will stretch before breaking). This is shown by the following table:

Just why a soft material like carbon (graphite), when added to another soft material like iron, should make the iron harder, has been quite a mystery, and one which has caused a tremendous amount of study. The mutual interactions of these two elements in various proportions and at various temperatures will be discussed at greater length later, especially in Chap. VIII, p. 105. But we may anticipate by saying that some of the iron unites with all the carbon to form a new substance, very hard, a carbide which has been called "cementite." The compound always contains iron and carbon in the proportions of three atoms of iron to one atom of carbon; chemists note this fact in shorthand by the symbol Fe₃C (a definite chemical compound of three atoms of iron to one of carbon). Many of the properties of steel, as they vary with carbon content, can be linked up with the increasing amount of this hard carbide cementite, distributed in very fine particles through the softer iron.

Sulphur is another element (symbol S) which is always found in steel in small quantities. Some sulphur is contained in the ore from which the iron is smelted; more sulphur is introduced by the coke and fuel used. Sulphur is very difficult to get rid of in steel making; in fact the resulting metal usually contains a little more than the raw materials used. Only the electric furnace is able to produce the necessary heat and slags required to eliminate sulphur, and as a matter of fact the sulphur does not go until several other impurities have been eliminated. Consequently, an electric steel with extremely low sulphur (0.02 per cent) is by that same token a well-made metal.

Sulphur is of most trouble to rolling and forging operations when conducted at a red heat. It makes steel tender and brittle at that temperature—a condition known to the workmen as "red-short." It seems to have little or no effect upon the physical properties of cold steel—at least as revealed by the ordinary testing machines—consequently many specifications do not set any limit on sulphur, resting on the idea that if sulphur is low enough not to cause trouble to the manufacturer during rolling, it will not cause the user any trouble.

Tool steel and other fine steels should be very low in sulphur, preferably not higher than 0.03 per cent. Higher sulphur steels (0.06 per cent, and even up to 0.10 per cent) have given very good service for machine parts, but in general a high sulphur steel is a suspicious steel. Screw stock is purposely made with up to 0.12 per cent sulphur and a like amount of phosphorus so it will cut freely.

Manganese counteracts the detrimental effect of sulphur when present in the steel to an amount at least five times the sulphur content.

Phosphorus is an element (symbol P) which enters the metal from the ore. It remains in the steel when made by the so-called acid process, but it can be easily eliminated down to 0.06 per cent in the basic process. In fact the discovery of the basic process was necessary before the huge iron deposits of Belgium and the Franco-German border could be used. These ores contain several per cent phosphorus, and made a very brittle steel ("cold short") until basic furnaces were used. Basic furnaces allow the formation of a slag high in lime, which takes practically all the phosphorus out of the metal. Not only is the resulting metal usable, but the slag makes a very excellent fertilizer, and is in good demand.

Silicon is a very widespread element (symbol Si), being an essential constituent of nearly all the rocks of the earth. It is similar to carbon in many of its chemical properties; for instance it burns very readily in oxygen, and consequently native silicon is unknown—it is always found in combination with one or more other elements. When it bums, each atom of silicon unites with two atoms of oxygen to form a compound known to chemists as silica (SiO₂), and to the small boy as "sand" and "agate."

Iron ore (an oxide of iron) contains more or less sand and dirt mixed in it when it is mined, and not only the iron oxide but also some of the silicon oxide is robbed of its oxygen by the smelting process. Pig iron—the product of the blast furnace—therefore contains from 1 to 3 per cent of silicon, and some silicon remains in the metal after it has been purified and converted into steel.

However, silicon, as noted above, burns very readily in oxygen, and this property is of good use in steel making. At the end of the steel-making process the metal contains more or less oxygen, which must be removed. This is sometimes done (especially in the so-called acid process) by adding a small amount of silicon to the hot metal just before it leaves the furnace, and stirring it in. It thereupon abstracts oxygen from the metal wherever it finds it, changing to silica (SiO₂) which rises and floats on the surface of the cleaned metal. Most of the silicon remaining in the metal is an excess over that which is required to remove the dangerous oxygen, and the final analysis of many steels show enough silicon (from 0.20 to 0.40) to make sure that this step in the manufacture has been properly done.

Manganese is a metal much like iron. Its chemical symbol is Mn. It is somewhat more active than iron in many chemical changes—notably it has what is apparently a stronger attraction for oxygen and sulphur than has iron. Therefore the metal is used (especially in the so-called basic process) to free the molten steel of oxygen, acting in a manner similar to silicon, as explained above. The compound of manganese and oxygen is readily eliminated from the metal. Sufficient excess of elemental manganese should remain so that the purchaser may be sure that the iron has been properly "deoxidized," and to render harmless the traces of sulphur present. No damage is done by the presence of a little manganese in steel, quite the reverse. Consequently it is common to find steels containing from 0.3 to 1.5 per cent.

Alloying Elements.—Commercial steels of even the simplest types are therefore primarily alloys of iron and carbon. Impurities and their "remedies" are always present: sulphur, phosphorus, silicon and manganese—to say nothing of oxygen, nitrogen and carbon oxide gases, about which we know very little. It has been found that other metals, if added to well-made steel, produce definite improvements in certain directions, and these "alloy steels" have found much use in the last ten years. Alloy steels, in addition to the above-mentioned elements, may commonly contain one or more of the following, in varying amounts: Nickel (Ni), Chromium (Cr), Vanadium (Va), Tungsten (W), Molybdenum (Mo). These steels will be discussed at more length in Chapters III and IV.

PROPERTIES OF STEEL

Steels are known by certain tests. Early tests were more or less crude, and depended upon the ability of the workman to judge the "grain" exhibited by a freshly broken piece of steel. The cold-bend test was also very useful—a small bar was bent flat upon itself, and the stretched fibers examined for any sign of break. Harder stiff steels were supported at the ends and the amount of central load they would support before fracture, or the amount of permanent set they would acquire at a given load noted. Files were also used to test the hardness of very hard steel.

These tests are still used to a considerable extent, especially in works where the progress of an operation can be kept under close watch in this way, the product being periodically examined by more precise methods. The chief furnace-man, or "melter," in a steel plant, judges the course of the refining process by casting small test ingots from time to time, breaking them and examining the fracture. Cutlery manufacturers use the bend test to judge the temper of blades. File testing of case-hardened parts is very common.

However there is need of standardized methods which depend less upon the individual skill of the operator, and which will yield results comparable to others made by different men at different places and on different steels. Hence has grown up the art of testing materials.

TENSILE PROPERTIES

Strength of a metal is usually expressed in the number of pounds a 1-in. bar will support just before breaking, a term called the "ultimate strength." It has been found that the shape of the test bar and its method of loading has some effect upon the results, so it is now usual to turn a rod 5½ in. long down to 0.505 in. in diameter for a central length of 2-3/8 in., ending the turn with 1/2-in. fillets. The area of the bar equals 0.2 sq. in., so the load it bears at rupture multiplied by 5 will represent the "ultimate strength" in pounds per square inch.

Such a test bar is stretched apart in a machine like that shown in Fig. 9. The upper end of the bar is held in wedged jaws by the top cross-head, and the lower end grasped by the movable head. The latter is moved up and down by three long screws, driven at the same speed, which pass through threads cut in the corners of the cross-head. When the test piece is fixed in position the motor which drives the machine is given a few turns, which by proper gearing pulls the cross-head down with a certain pull. This pull is transmitted to the upper cross-head by the test bar, and can be weighed on the scale arm, acting through a system of links and levers.

Thus the load may be increased as rapidly as desirable, always kept balanced by the weighing mechanism, and the load at fracture may be read directly from the scale beam.

This same test piece may give other information. If light punch marks are made, 2 in. apart, before the test is begun, the broken ends may be clamped together, and the distance between punch marks measured. If it now measures 3 in. the stretch has been 1 in. in 2, or 50 per cent. This figure is known as the elongation at fracture, or briefly, the "elongation," and is generally taken to be a measure of ductility.

When steel shows any elongation, it also contracts in area at the same time. Often this contraction is sharply localized at the fracture; the piece is said to "neck." A figure for contraction in area is also of much interest as an indication of toughness; the diameter at fracture is measured, a corresponding area taken out from a table of circles, subtracted from the original area (0.200 sq. in.) and the difference divided by 0.2 to get the percentage contraction.