Hardfacing of Cutters with High-Speed Steel Using a Carbon Electrode

A. N. Dobrovidov
Professor, Doctor of Technical Sciences

N. A. Balakin
Associate Professor, Candidate of Technical Sciences

A. M. Rozenberg
Professor, Doctor of Technical Sciences

N. D. Tyuteva
Associate Professor, Candidate of Technical Sciences

Introduction

The shortage of high-speed steel, explained by the scarcity of its principal alloying element—tungsten—has led, on the one hand, to efforts to create such grades of steel as would contain a smaller amount of tungsten while preserving the same cutting qualities, and, on the other hand, to the search for such thermal and thermochemical methods of treating cutting tools which, by improving their cutting qualities, would thereby lead to a reduction in tool consumption. As a result of these investigations, low-alloy high-speed steels appeared, containing a significantly smaller amount of tungsten, as well as various methods of improved hardening and tempering of tools, cyaniding, nitrocementation, and chemical treatment.

The desire to reduce the amount of high-speed-steel waste in tool manufacture led to the creation of a wide variety of assembled tools (milling cutters, countersinks, reamers), in which only small parts directly performing the cutting operation are made of high-speed steel, while the body of the tool is made of inexpensive carbon steel. The same tendency led to numerous and long-term experiments on obtaining cast tools, whose basic form would be obtained by casting.

The solution of the problems of economizing high-speed steel would be incomplete if, alongside the tasks listed above, the question of utilizing high-speed-steel waste—available in large quantities at machine-building plants in the form of chips obtained in the manufacture of cutting tools, small scrap, and remnants of worn tools—were not solved. Worn large-size tools can be used to manufacture small-size tools, but this does not solve the question of utilizing small scrap and, still less, high-speed-steel chips. Utilization of these wastes for remelting into ingots of high-speed steel or for manufacturing cast tools is entirely possible, but is feasible only at a limited number of enterprises having special metallurgical equipment, and thus does not fully solve this question.

Shortly before the beginning of the war, proposals began to appear to utilize high-speed-steel waste by hardfacing the cutting elements of tools with this waste, using gas or electric welding for the hardfacing. A number of studies were also carried out in this direction,

for the most part schematic and incomplete, not making it possible to judge reliably the cutting qualities of the tool obtained. The common idea behind all these proposals and investigations was to create on the tool a deposited layer of high-speed steel by melting it with an oxyacetylene flame or with the flame of an electric arc struck between an electrode of high-speed steel and the body of the tool.

It may be assumed that both methods give fairly good results, but both have substantial shortcomings.

Hardfacing high-speed steel with an oxyacetylene flame makes it possible to utilize any kind of high-speed-steel waste, but it requires the use of scarce calcium carbide and oxygen and the presence of special gas-welding equipment, which is not widely available at machine-building plants.

The method of hardfacing high-speed steel with an electric arc has, up to the present time, been developed exclusively by using the Slavyanov method (welding with a metallic electrode). A rod-electrode is made from high-speed steel and is connected to one pole; the tool being built up is connected to the other; an electric arc is struck between them, in the flame of which the high-speed-steel rod melts and is deposited on the body of the tool.

The shortcoming of this method is, first of all, the need to manufacture an electrode from high-speed steel, which can be done by forging large pieces of high-speed steel. High-speed-steel turnings cannot be used for this purpose. The second shortcoming is the need to prepare a complex coating for the electrodes, without which a high-quality deposited layer cannot be obtained. Among other components, the coating includes scarce high-modulus water glass.

The method of hardfacing high-speed steel with an oxyacetylene flame was proposed, apparently, by engineer Lapidus.

The method of hardfacing with a high-speed-steel electrode was proposed by Vagin and Bessonov, as well as by Heller and Novikova.

Taking into account the shortcomings of these methods and considering that the solution of this question under the conditions of the Patriotic War was especially urgent, the authors of the present work (a group of research workers of the Tomsk Polytechnic Institute) in 1943 undertook the development and investigation of a method for hardfacing high-speed steel using a carbon electrode (welding by the Benardos method). The flame of the electric arc is struck between the carbon electrode and the body of the tool being hardfaced. Hardfacing can be performed with high-speed steel in any form (turnings, small pieces). This method was not supposed to require either scarce materials or special equipment; it was to be simple and accessible to production. When using this method, it was to be easy to alloy the deposited metal with desired elements and to mix various grades of high-speed steel in the deposited layer. Therefore, it was expected to justify itself not only in the hardfacing of high-speed steels of the RF-1 type, but also of their low-alloyed substitutes.

The first results of the authors’ investigations were reported in the TEKhSO card file (card No. 999/60 for 1943) and then in the journal Vestnik mashinostroeniya, No. 7 for 1944.

The present work gives the results of investigations for 1943 and 1944. The work consists of four chapters. The first chapter gives a theoretical analysis of certain questions connected with obtaining a deposited layer of high-speed steel.

The second chapter sets forth the technology, developed by the authors, for hardfacing high-speed steel with a carbon electrode.

The third chapter is devoted to the study of the microstructure of the deposited layer.

The fourth chapter presents the results of laboratory and production tests of hardfaced cutters.

All the work was carried out with cutters, as the most widespread, simplest tool and therefore the one most accessible for broad investigation. It may be asserted that the results obtained can also be extended to tools of other types.

It should be noted that until very recently a number of authors have believed that hardfacing of high-speed steel, regardless of the method used, cannot produce a high-quality tool. For example, Prof. Minkevich (Low-Alloy High-Speed Steels. Metallurgizdat, 1944) states that hardfacing “leads to the formation of a cast structure, more brittle and less satisfactory in cutting properties, and to a certain change in the chemical composition of the high-speed steel, difficult to take into account in each case, and consequently to nonuniformity of the cutting properties in the hardfaced tool.”

Until now this assertion could not be refuted, since there were no systematic investigations of the cutting properties of hardfaced tools. The authors of the present work have carried out sufficiently complete tests of hardfaced cutters, which showed that the use of the method developed by the authors for hardfacing high-speed steel with a carbon electrode produced cutters which, in their cutting properties, were not inferior to cutters made from bar high-speed steel.

The work was carried out in the laboratories of metallography and metal cutting of the Tomsk Polytechnic Institute named after S. M. Kirov.

CHAPTER I

HIGH-SPEED STEEL AND ITS USE FOR HARDFACING

1. General characteristics of high-speed steel

The prototype of modern high-speed steel is usually considered to be the alloy proposed in the seventies of the last century by Mushet. Owing to its considerable carbon content (about 2.2%) and alloying elements (approximately 6% tungsten and 2% manganese), this alloy should be called alloyed white cast iron or a hard alloy of tungsten, manganese, and iron carbides. It forged poorly and had the ability to harden in air.

The appearance of this steel passed unnoticed, since there were no prerequisites for its rational application; but in its time it became the point of departure from which began the development of the steel proposed by Taylor and White (1898), which they called high-speed steel. [1] The first alloys of these investigators were close to Mushet steel in their high carbon content (1.4%) and moderate tungsten content (8.4%).

The composition of almost modern high-speed steel was published by Taylor in 1903 (0.7% C, 14.7% W, and 2.9 Cr), and around 1906 he gave the formula of the famous steel, which has retained its importance to our time, “Link Belt,” with a content of C = 0.7%, W = 18%, Cr = 5.7%, and V = 0.3%.

Taylor approached all his investigations purely empirically and did not even attempt to give any theoretical substantiation, even approximately, either for the composition of his steel or for the method of its heat treatment. And indeed such work was hardly within reach at a time when metallogra-

...graphy as a science was only just coming into being. And it is not surprising that the first investigations of high-speed steel gave very little for the further development of this new material for cutters (for example, the work of Bela, 1904), and its subsequent improvement proceeded by trial and error (for example, the addition of cobalt proposed by Becker in 1912).

A typical high-speed steel with a high tungsten content (18 and even, in the steels of some foreign firms, 21% tungsten) is an expensive alloy and, precisely because of this element, a scarce one. Therefore, even before the Patriotic War, when the shortage of tungsten steel became especially acute, metallurgists began to seek the possibility of replacing this scarce alloying addition with a more readily available element.

The theoretical foundations for such replacement are fairly simple if high-speed steel in the annealed state is schematically regarded as consisting of a eutectoid, resembling ordinary pearlite in structure and properties, and of primary and secondary carbides.

These complex carbides include tungsten, chromium, vanadium, and sometimes molybdenum. An approximate formula of the complex carbide is \((W, Cr, V, Mo)_n nC_m\). Within the limits of this fundamental composition, changes are possible that depend both on the temperature of the steel at a given moment and on the rate of cooling up to the moment under consideration. The change in the composition of this complex carbide \((W, Cr, V, Mo)_n C_m\) in the temperature interval 1300–400° must obey the general formula:
\(\mathrm{Me}_n C_m + M'e_R \rightleftarrows M'e_R C_m + nMe\). The equilibrium of the system at a definite temperature is given by the value of the equilibrium constant \(K\)

\[ K=\frac{(\mathrm{Me}_n C_m)(M'e)^R}{(M'e_R C_m)(Me)^n}. \]

The composition of the primary carbide must be different from the composition of the secondary carbide, and will also differ from the composition of the eutectoid [2]. In addition, the phenomenon is somewhat complicated by the interaction of carbides with the elements entering into solid solution (for example, chromium and partly tungsten), and by the interaction of the carbon of the carbides and of the solid solution.

But the formula of the carbide of high-speed steel makes it possible to draw certain general conclusions about the mutual replaceability of the elements entering into its composition.

Tungsten, chromium, vanadium, and molybdenum have geometrically similar crystal lattices. In addition, tungsten and molybdenum exhibit a similar structure of electron shells, and the atoms (as well as the ions) are almost identical in size. Chromium has an analogous structure of the electron shell as well, but the size of its ion is smaller.

These elements give similar constitution diagrams in binary alloys with iron (a closed region of \(\gamma\)-solid solution), form simple carbides separately with carbon and, all together, the aforementioned complex carbide of high-speed steel. It is considered that in this carbide one tungsten atom may be replaced by an atom of any other of these elements, and then the replacement coefficient will obviously be equal to the ratio of the atomic weights of tungsten and of the substituting element [3]. In doing so, it is, of course, necessary to take into account that some elements are partly present in solid solution (in ferrite or austenite).

A cutting tool made of high-speed steel obtains the highest productivity only after correct heat treatment, consisting, as Taylor had already shown, in hardening from a temperature close to the melting temperature of the steel, and in subsequent tempering.

In high-speed steel, the austenitic grain grows slowly during heating; therefore the high heating necessary for transferring secondary carbides into solid solution does not spoil it as it would spoil an ordinary ha—

eutectoid steel. This is explained by the presence in the alloy structure of a large quantity of carbides, which purely mechanically restrain the growth of the austenite grain. But under prolonged heating this phenomenon has to be reckoned with, and in order to slow grain growth some authors recommend adding nickel to the steel; there are, however, contradictory opinions concerning its influence on the cutting properties of high-speed steel. At the same time, high heating, despite the ease with which it is carried out, is an undesirable circumstance, and sometimes quite a significant obstacle for many plants. This became especially evident during the war, when evacuated enterprises that had had a good heat-treatment base at their former location had, without the appropriate equipment, to build heat-treatment shops in new places.

A simple matter—the hardening of a cutter made of ordinary high-speed steel (RF-1 or a similar grade)—becomes a complicated operation when low-alloy high-speed steel is used (even EI-184), when an error in heating of ten degrees threatens inevitable hardening rejects.

For the best use of a high-speed tool, after proper hardening it must be tempered. Tempering is a less complicated operation in comparison with hardening. The tempering temperature is around 550°. At such a low temperature one does not encounter decarburization of the cutting edge, grain growth, or the development of excessive internal stresses during heating. The apparatus here is simpler. Nevertheless, to obtain good results, double, or even triple, tempering is necessary.

2. Hard and Superhard Alloys

People have become so accustomed to the necessity of hardening and tempering high-speed steel that the naturalness of such heat treatment even makes the question of its fundamental necessity in the manufacture of a cutting tool seem somehow strange. However, eliminating heat treatment would undoubtedly simplify and cheapen the manufacture of cutters.

One of the advantages of a tool with a cutting part made of hard alloys (of the “stellite” type) and superhard alloys (of the “pobedit” type) is precisely that they do not require annealing, hardening, and tempering.

If, in a cutter made of high-speed steel, the “carriers” of the ability to cut are the martensitic matrix and the carbides embedded in it, then in stellites and superhard alloys it is mainly the carbides, surrounded by a comparatively soft binding mass, that cut. Therefore, in stellites there are significantly more carbides than in high-speed steel. The exceptional role of carbides in the cutting process is manifested especially strongly in superhard alloys of the “pobedit” type, in which the metallic mass that binds the carbides cannot cut at all because of its softness.

Owing to the increased quantity of carbides in cutting hard alloys as compared with high-speed steel, their brittleness is considerably greater than that of high-speed steel; in return, however, they do not require hardening and tempering.

The cutting part of a tool made of hard alloy is produced by casting, and the final geometrical form is imparted to the blade exclusively by grinding on an emery wheel, since, owing to its brittleness, the alloy cannot be forged, and because of its considerable hardness it cannot be worked with an ordinary cutting tool.

The brittleness of cutters made of alloys of the “stellite” type is, of course, undesirable, but apparently it is not a major obstacle in their operation. Plants producing stellites nevertheless devote much attention in their brochures to the brittleness of these alloys. Before the war, great ra...

The German alloy “akrit” has found use. Its composition includes cobalt (about 38%), chromium (about 30%), tungsten (16%), molybdenum (4%), carbon (about 2%), then iron and unavoidable impurities [4]. In the brochure devoted to akrit [5] it is stated: “Akrit is a cast alloy and therefore possesses a certain brittleness, but to a considerably lesser degree than other such metals, which are brittle even after the most careful treatment.”

Such alloys are not free of blowholes, and sometimes every piece of good-grade stellite is, as a rule, examined by X-rays for inspection.

Stellites are much more expensive than high-speed steel.

Among the stellites should be included the alloys proposed by Kontorov [6]. In them the principal alloying element is chromium, the content of which varies from 16 to 24%; molybdenum and vanadium are present at 6 to 8% of each element.

The cutting speed of cutters made of stellites is only slightly higher in comparison with cutters made of good high-speed steel. The hardness of stellites differs little from the hardness of properly heat-treated high-speed steel.

3. Possible Use of High-Speed-Steel Waste

High-speed steel is delivered to the consumer—the tool shop of a machine-building plant—in the form of rolled stock (bars of various profiles) or in the form of forged blanks. The shape is imparted to the tool chiefly by chip removal. Less often the tool shop forges the rolled stock received, since forging high-speed steel without well-established thermal control gives a considerable percentage of rejects.

In tool manufacture, up to 50% of the steel goes into low-value waste—chips. It is true that high-speed-steel chips can be remelted, but this requires certain equipment, and doing this at a machine-building plant is somewhat difficult, especially since the ingots obtained after remelting the chips must be forged or rolled into a profile or blank. And this means that, within the system of the tool shop, steelmaking and rolling shops, a shop for heat treatment of rolled stock, etc., must be organized. It is possible, of course, to send the chips to special plants for quality steel, but this entails transportation of a comparatively low-value cargo.

It is considerably more rational to produce the tool by casting or stamping, reducing machining to a minimum. In this case the chips could be used on the spot, hardfacing them onto the cutting parts with an electric arc or an oxyacetylene flame.

It is usually considered that, in order to obtain high-speed steel possessing good cutting properties, it must necessarily be forged or rolled. Not long ago many metallurgists believed that forging imparts to tool steel the most favorable structure. But there are some grounds for believing that properly carried out rolling gives metal no worse than forged metal.

Forging or rolling of a high-speed-steel ingot is considered necessary for breaking up the crystals of primary carbides.

The role of structurally free carbides, not entering into the eutectoid, is quite definite and identical in ordinary carbon, low-alloy, and high-speed steels when we consider a cutter made from one or another steel. The carbides, distributed more or less uniformly in the form of the finest possible grains in the structure of the finished cutting tool, should impart additional hardness to the base metal—

…of the metallic mass structure (troostite or martensite). But at the same time they must not reduce the impact toughness of the cutting edge. This is possible only if the carbides have been transformed by forging (rolling) and subsequent spheroidizing annealing into small, rounded grains without sharp corners, uniformly dispersed in the basic structure of hardened and tempered steel.

In the case of ordinary carbon hypereutectoid steel, such a basic structure will be troostite. In carbon steel, the granular form can be imparted to structurally free (secondary) cementite without forging, purely by thermal means—by hardening and high tempering. Such an operation is troublesome, and it is used rarely, only in exceptional cases, it being preferable to break up the cementite mechanically, by rolling below the \(SE\) line on the iron–carbon diagram. After rolling (forging), annealing should follow, relieving the stresses from forging, if they have arisen, and imparting the final granular form to the fragments of cementite crystals obtained as a result of plastic deformation.

In steels with a carbon content only slightly exceeding the eutectoid composition (for example, in steel U.10), when the carbide network, owing to the small amount of cementite, is discontinuous, it is possible, even without forging, by annealing alone sometimes to obtain granular cementite. But since the manufacture of a profile of small section requires the indispensable rolling (or forging) of the ingot, plastic deformation of the steel will always be present and will necessarily enter into the technological process of producing bars of steel with granular hypereutectoid cementite.

In all high-speed steels there are always primary, ledeburitic carbides separated from the liquid phase. These carbides, as is known, cannot pass into solid solution and dissolve only in the liquid phase from which they separated during crystallization of the steel. Therefore there exists the assertion that heat treatment has no effect whatsoever on their structure, and that they retain their structure unchanged both after annealing and after hardening with subsequent tempering. Accordingly there is only a single way of transforming them into grains—plastic deformation of the steel (forging or rolling).

This is both true and not true. The dissolution of a carbide crystal in austenite is a surface phenomenon, and the rate of dissolution will depend on the development of the crystal surface in relation to its volume. A coarse-grained rounded crystallite of secondary carbide may not yet dissolve in austenite, whereas a branched, skeletal, fine-lace-like ledeburitic carbide, with a more favorable ratio of external surface to volume, has already passed into solid solution and thus outstripped the grain of secondary carbide. But such small ledeburitic carbide crystals are formed only during rapid cooling of the crystallizing melt and, understandably, are not formed in ingots of ordinary size (100–1000 kg). Coarse primary carbides must be broken up mechanically (by forging, rolling), followed by annealing—spheroidization.

It is generally accepted that the crystallization of a metallic melt obeys Tammann’s laws, i.e., the size of the precipitating crystals and their form depend on the number of crystallization centers per unit volume of liquid and on the linear rate of crystallization. The number of centers and the rate of crystal growth are, in turn, a function of the degree of supercooling. The degree of supercooling is significantly influenced by the cooling rate of the melt, its chemical nature, the degree of contamination by suspended particles (and their dispersity), and even the material and radius of curvature of the vessel walls, since at the first moment of crystalliza-

…centers of crystallization arise precisely at the cold wall in contact with the liquid metal.

One must also not forget the possible presence in the melt of particles that play the role of centers of crystallization. They may be introduced artificially or may form as a result of chemical reactions that have taken place, for example, deoxidation and oxidation reactions. The crystals of aluminum oxide formed in steel when it is deoxidized with aluminum, if of suitable dispersion, are forced centers of crystallization for crystals of the solid solution of carbon in iron.

The questions touched upon have recently been studied quite successfully with respect to the crystallization of ordinary carbon steel. There is little doubt that the main conclusions can be transferred to alloys of the high-speed-steel type.

First of all, the influence of ingot size on the magnitude of the crystals of primary carbides is beyond doubt. The mass of the liquid alloy will determine one or another cooling rate. In a small ingot, ledeburitic carbides will separate out from a large number of crystallization centers.

Unfortunately, there are not enough practical data on the nature of bodies that may serve as artificial (forced) centers of crystallization for crystals of carbides and austenite in high-speed steels. Undoubtedly, such substances exist, and introducing them into the melt will replace spontaneous crystallization with forced crystallization.

The considerations set forth evidently lead to recognition of the fundamental possibility of manufacturing a cast cutting tool from tool steels.

As has already been mentioned, the weight of a cutting tool made by machining on machine tools, by removing chips, averages 50–60% of the weight of the forged blank or bar stock. It is not always possible to use hot stamping for tool manufacture, although where this can be done a considerable saving of metal is obtained, as, for example, in the production of twist drills by the stamping-twisting method. A much larger assortment of cutting tools can be obtained by casting.

4. Cast Tool

The question of the cast tool has its own history.

At the Siberian Institute of Metals [7], A. N. Dobrovidov and K. A. Bessonov carried out a study on manufacturing tools by casting from carbon steels with a carbon content of 1.0–1.5%. Casting was performed in cast-iron and dry clay molds. Centrifugal casting on a specially designed machine was also used. The steel was deoxidized with manganese and aluminum. The greatest casting weight reached 2.5 kg. The size of the austenitic grain of the casting could be varied in the process of deoxidation with aluminum by changing its amount, but cementite in the casting always separated out in mixed form, forming a network along the boundaries of austenite crystals and in the form of needles located inside the pearlitic grains that had arisen in place of austenite crystals. Such a structure was very brittle, and specimens made of the steel broke without noticeable plastic deformation.

The cast tool was heat-treated. First it was quenched from 1050° in oil, preliminarily tempered at 650°, and then annealed at 720°. After heat treatment the cementite had the form of fine grains uniformly scattered in granular pearlite.

The cutting properties of cutters made by casting did not differ in any way from those of cutters made from ordinary first-class bar steel.

Cutters made by centrifugal casting also showed good results.

Much more important is the manufacture of cast tools from high-speed steel.

Dietz [8] proposed casting cutting parts of alloy steel around a shank made of tool steel. The mold for casting was made of clay and, before pouring, was heated together with the shank core. The cast tool was annealed, machined, hardened, and brought to final dimensions by grinding. Dietz believes that his method is economical when casting tools of more than 40 mm diameter, and asserts that, in productivity, the tool is not inferior to the ordinary one.

Horwitz [9] gives data on a cast tool made of steel similar in composition to ordinary high-speed steel of the RF-1 type. According to his data, the productivity of the cast tool is almost the same as that of the forged one.

French [8], when cutting with a cast tool made of steel similar in composition to RF-1, obtained considerably worse results.

Extensive work with cast cutters made of high-speed steels of very varied composition was carried out by Ertel and Peltzguter [8]. They investigated fifteen grades of steel, attempting to determine the influence of cobalt and vanadium on molybdenum and tungsten steels. Eight molybdenum steels contained 5% molybdenum; cobalt from 3 to 9% was added to three of them. Four steels contained, in addition to 5% molybdenum, also 9% cobalt, and varying amounts of vanadium (from 0.8 to 2.6%) were introduced into them. The remaining seven steels were tungsten steels with a constant content of this element (11%); cobalt (3–8%) was added to them, and to two of them, in addition, vanadium (1.3 and 1.5%). The steel grades did not correspond to standard grades, and the authors’ conclusions, while of some theoretical interest, give little for practice.

Despite the large number (dozens) of works on cast tools, their manufacture on large factory scales was not established in the prewar period. Rapatz [10] writes that “experiments in obtaining cast tools from high-speed steel, carried out in recent years in the USA, remained unsuccessful.” He sees the explanation for this in the nonuniformity of the steel structure in the cast state, which causes the low quality of cast tools, chiefly their brittleness.

It seems to us that the matter is not only the poor structure. The production of shaped castings from ordinary carbon steel, as is known, is economically advantageous only in the case of castings of not very small weight. Castings of small weight (hundreds of grams) are the more expensive the smaller they are. Their high cost depends on the difficulty of casting (molding and the actual pouring). This explains why malleable cast iron—a material less strong than steel—is not replaced by steel in the manufacture of small castings. Of course, the difficulty of pouring in small portions does not decrease when switching to high-speed steel, but increases. This is the chief obstacle to the wide application of casting for the production of cutting tools.

Not long ago a work by Revis and Lebedev [11] appeared on cast tools made of high-speed steel. The authors proposed the following technological process. In order to obtain fine-grained primary carbides in the casting, one should avoid, when designing the tool and the casting mold for it, large masses of metal, since this will give slow cooling of the casting instead of the necessary rapid cooling. It is desirable to superheat the molten steel for the most complete possible liquidation of undissolved carbide crystals present in the melt. During solidification of the casting these crystals may give forced—

new crystallization centers and make the structure fine-grained. On the basis of these considerations, the recommended overheating temperature of the steel is about 1550°. Before casting, the alloy must be cooled to 1450°. The authors mention in passing the possibility of modifying the alloy.

After ordinary annealing, hardening, and double tempering, a cast tool made of RF-1 steel has a Rockwell hardness of 62–64 and in operation is not inferior in productivity to a forged one.

Experimental specimens of this tool were 40–50 mm in diameter. It seems to us that there is hardly any serious basis for counting on wide dissemination of cast tools produced by the methods described. In addition to what has been said above, it must be added that it is scarcely possible to manufacture a large tool because of unfavorable crystallization conditions. It is difficult to cast cutters, since the weight of the cutting part of this tool is small, not exceeding 20–30 g. The equipment required for melting the steel is expensive. When remelting chips and waste (broken tools), deoxidation is required, i.e., the work of an experienced steelmaker is necessary. In any case, the process can be carried out only at a well-equipped enterprise.

5. Hardfacing of Cutting Tools

An attempt to dispense with complex units for melting steel led to the invention of methods for producing cast tools by gas and electrode hardfacing of the cutting parts. In this case, the melting of high-speed steel takes place at the site where the cutting edge of the tool is formed; the bottom, and sometimes—depending on the shape of the tool—the walls of the crucible (and simultaneously of the mold) are served by the body (holder) of the tool, made of carbon steel. The volume of molten metal in this case corresponds to the volume of the cutting part and will be considerably smaller than in producing a cast tool by casting from steel melted in a furnace. Owing to its small dimensions, the cast blade will cool rapidly, which will have a favorable effect on the size and shape of the primary carbides. The cooling rate will also increase because of the rapid propagation of heat through the holder, which will be strongly heated only at the point of contact with the molten metal. The consumption of steel, as compared with the consumption for an ordinary cast tool, will be smaller.

In Soviet literature, a number of methods for hardfacing cutting tools have been proposed in recent years. All these methods may be divided into two groups. The first includes hardfacing with a gas flame [12]; the second, hardfacing with a voltaic arc using a metallic electrode.

Hardfacing with an oxyacetylene torch must be carried out with a slightly reducing flame, which does not change the chemical composition of the high-speed steel, usually deposited in the form of a rod. With careful work one can obtain good fusion and dense hardfacing.

But owing to the characteristics of gas welding, the holder is heated very strongly, and this has an unfavorable effect on the crystallization of carbides.

After hardfacing, the tool requires the hardening and tempering customary for high-speed steel. Sometimes slow cooling of the tool after hardfacing in an annealing furnace or in a heat-insulating material, and ordinary annealing, are recommended. As a result, a good tool is obtained.

One of the substantial disadvantages of this method of manufacturing a cast tool, in addition to those mentioned above, is the scarcity of oxygen and acetylene.

There are also methods of hardfacing a tool by atomic-hydrogen welding and by an electric arc according to the Slavianov method (the arc burns between an electrode made of high-speed steel and the body of the tool).

The first method is good for welding low-carbon steels. When welding high-carbon steels, decarburization occurs as a result of the interaction of hydrogen with the carbon of the steel. To compensate for the burning out of carbon, high-carbon alloys have to be introduced into the welding zone, which greatly complicates the whole process. At the present time this welding belongs to the category of methods not readily accessible to a large number of plants.

Hardfacing by the Slavianov method requires the manufacture of electrodes from high-speed steel, or the use of an alloying coating for electrodes made of ordinary welding wire, which greatly complicates the process and requires the introduction into the coating of scarce ferroalloys and materials.

The first to propose using electrodes made of high-speed steel for hardfacing cutters were K. A. Bessonov and V. A. Vagin [13] in 1937. An analogous proposal was then made by Geller and Novikova [3] (1941). The authors recommend forging electrodes from tool fragments (drills, taps, etc.) and using special coatings to protect the molten steel from the burning out of alloying elements.

Judging from the literature data, quite satisfactory cutters are obtained with a cutting edge hardfaced with such electrodes. The weak point of this method is the necessity of making, from pieces of high-speed steel, electrodes 4–6 mm in diameter. This is an expensive and troublesome operation.

None of the methods proposed up to now makes it possible to use such widespread and low-value waste as turning, milling, and similar chips, as well as small fragments of high-speed steel. Therefore the authors of the present work undertook the development of a method for manufacturing tools from such low-value wastes, which are available at all machine-building plants, by hardfacing the cutting edges by the Bernardos method.

Bernardos proposed using, for melting the filler metal forming the seam, the heat of a voltaic arc arising between a carbon electrode and the part being welded at the site of the future seam. Welding with a carbon electrode is the oldest variety of arc welding, but in ordinary cases it has been almost completely displaced by welding with a metallic electrode.

Almost always, in welding with a carbon electrode, direct current of normal polarity is used—minus on the electrode. An arc of straight polarity more than 4 mm long does not carburize the weld metal. With reverse polarity, intensive carburization occurs; reverse polarity is not used in ordinary welding work.

An arc of straight polarity is stable, does not go out even at a length of about 50 mm, and even an inexperienced welder can use it easily. The carbon arc is very stable also on alternating current.

As has already been mentioned, in making a cast tool one must take into account first of all the character of crystallization of a small-size casting. In hardfacing cutters by the authors’ method with a carbon electric arc, the casting weighs only a few tens of grams, which has a favorable effect on the structure of the hardfaced layer.

Crystallization, as shown by fractures and macrosections of the hardfaced metal, proceeds first of all at the surface of contact of the molten metal with the holder, and then at the place of contact with the side walls of the mold and at the top, where on the surface of the melt there is a layer of liquid slag.

Crystallization is columnar in character, but, owing to the rough surface of the mold and the surface of the seat selected in the cutter holder having become uneven as a result of melting, the parallelism of the axes of the columnar crystals is very weakly expressed, and the fine-grained nature of the fracture is so considerable that it differs little from the fracture of bar high-speed steel. The fine-grained structure can explain why transcrystallization is very often not detected at all either in the fracture or on an etched macrosection.

Some authors [12], [13] recommend forging the deposit immediately after hardfacing, while the steel is still sufficiently hot.

Forging is a superfluous operation in the manufacture of cutters, but it is recommended because there is an opinion that high-speed steel acquires good cutting properties only after rolling (forging), and that the greater the shape change of the ingot, the higher the quality of the steel obtained. In the process of plastic deformation of the metal, as we have indicated, the primary carbides are broken up and the ledeburitic crystallization is destroyed.

Thus, Khabakhpashev believes that when rolling an ingot of high-speed steel a 5–6-fold reduction is necessary. Sheinin recommends a reduction by a factor of \(3 \tfrac{1}{2}\); Krits speaks of the need for a tenfold reduction of the ingot section [14]. Hoxhage and Rollet came to the conclusion that, for complete destruction of the ledeburitic network, it is necessary to reduce the cross section by 12.5 times. Andrew eliminated the ledeburitic network by rolling an ingot of \(150 \times 150\) mm section down to a section of \(55 \times 55\) mm, which corresponds to a 7.5-fold reduction in section [15].

Chizhikov [15] investigated what reduction of the ingot should be considered minimal. An ingot of high-speed steel 210 mm in diameter was forged according to the scheme in Fig. 1. From various places of the forging, as shown in the drawing, specimens of square section (\(29 \times 29\) mm) and 250 mm long were cut out. From the specimens, turning tools were made, which were heat-treated in exactly the same way. The hardness of the tools after heat treatment ranged within 65–68 Rockwell units.

Fig. 1

The cutters were tested in rough turning of chromium-nickel steel with a temporary tensile strength of \(100\ \text{kg}/\text{mm}^2\) at a cutting speed of \(14\ \text{m}/\text{min}\). The feed was 4 mm, and the depth of cut was 5 mm.

The results of the experiments are presented in Table 1.

Considering Table 1, we come to the conclusion that forging with a 52-fold reduction increased the average durability of the cutters by 50%, while the durability of individual ...

Table 1

Cross-section size from which the cutters were made	Degree of forging	Cutting time to blunting of the cutter, in minutes and seconds: average for four cutters	Cutting time to blunting of the cutter, in minutes and seconds: maximum for one cutter
Initial cross-section	Not forged	8–05	10–05
Diameter 210 mm	Cast steel
Diameter 105 mm	4-fold	8–25	12–10
Square 59 × 59 mm	12-fold	8–45	14–05
Square 42 × 42 mm	25-fold	8–32	12–05
Square 29 × 29 mm	52-fold	12–30	18–15

of a single cutter (the best one) by 80%. Forging with a 25-fold reduction of the cross-section, according to the average data, did not improve the cutting properties of the steel at all. The best cutter (of the four) gave an increase in durability of only 20%. But this durability value (12 min., 5 sec.), apparently, is not typical for metal with this degree of forging, since the cutter from the place with 12-fold forging gave better results.

An increase in cutter durability (in time) by 50% corresponds to an increase in cutting speed of only approximately 5%.

Thus, Chizhikov’s experiments show quite clearly that forging even large ingots of high-speed steel with a considerable forging reduction has very little effect on increasing the cutting speed of lathe cutters. It must be noted that the ingot was taken from a rather small heat. Undoubtedly, its structure contained large ledeburitic carbides.

On the basis of the foregoing we came to the conclusion that there is no point in forging overlays that are ingots of very small size. The maximum possible reduction that can be given to the overlay will not exceed, in the best case, twenty percent, and therefore will not affect the structure or the cutting properties of the tool.

Forging may be useful in that, by changing the shape of the overlay, which always solidifies with rounded edges instead of sharp corners, it facilitates the subsequent mechanical machining of the faces, somewhat reducing the consumption of abrasives.

Proceeding from these considerations, we press the liquid overlay from above, before it has begun to crystallize, with a flat piece of metal (cast iron, copper, etc.). As a result of such pressing, the metal begins to cool very rapidly from above, which has a favorable effect on the size of the crystals growing in it, while the rounded corners of the overlay are transformed into sharp ones.

High-speed steel is less thermally conductive than ordinary carbon steel. Therefore, during cooling, internal stresses arise in it more readily. Their appearance is also promoted by the phase transformations occurring in the steel, accompanied by a change in volume. Especially significant stresses must appear at the place where the alloyed steel is fused with the backing steel of the holder. Therefore, after welding, welded tools are recommended to be cooled slowly and then annealed in order to facilitate subsequent mechanical machining by cutting and to remove internal stresses more completely.

But, on the other hand, the following cooling schemes exist for ingots of high-speed steel after casting [14]. After making sure that the ingot has solidified, it is stripped and cooled in air. According to another scheme, the ingot is stripped at 700° and also cooled in air. In this case there seems to be no fear of the formation of internal stresses.

Slow cooling of the hardfaced cutter will not improve the structure of the deposit, but will introduce an additional operation—it will make annealing necessary. Seeking to provide the simplest method for producing cutters with a hardfaced cutting edge of high-speed steel, we became convinced, after a number of experiments, that after hardfacing the cutters may be cooled in still air and even in an air jet.

Cooling in an air jet gives the same hardening as cooling in still air and offers no special advantages. In both cases, cracks appear neither in the deposit itself nor at the place where it is fused with the steel of the holder.

6. Some Considerations on Electric-Arc Hardfacing

Thus there was developed the process for hardfacing cutters which is described in greater detail below: hardfacing a layer of molten high-speed steel (chips, fragments of tools) onto the cutter holder, cooling in air, tempering, and sharpening.

Since the carbide-forming elements of high-speed steel are characterized, to a certain extent, by mutual substitutability, it has become possible in recent years to obtain a number of substitute steels. There is no doubt that such low-alloy high-speed steels will continue to appear in the future.

Some considerations may be expressed regarding their probable composition. Steel R contains 0.66–0.78% carbon, 17.0–18.5% tungsten, 0.5–0.8% vanadium, and 3.8–4.6% chromium. Steel EI-262 (RF 2) has the following composition: 0.71–0.77% carbon, 11.8–12.8% tungsten, 2.3–2.6% vanadium.

Factory experience shows that a tool made of steel R is approximately equal in cutting properties to a tool made of EI 262. A tool made of the latter steel in some cases even has greater durability.

It is possible to make the quite well-founded assumption that the 6–7% tungsten lacking in EI 262 steel, as compared with steel R, has been replaced by 1.8–2.0% vanadium, and that such a replacement has had absolutely no effect on the cutting properties of the steel.

It is evidently quite possible to obtain a series of intermediate steels between EI-262 and R. These intermediate steels should possess identical cutting properties. The transition from one of these steels to another can be carried out by reducing the vanadium content in the steel and simultaneously increasing the tungsten content. In such a replacement, one percent of vanadium corresponds to approximately three percent of tungsten.

This means that, by using mixtures of chips of EI 262 and R steels for hardfacing cutters, it is possible to obtain a full-grade high-speed cutter.

It is better to take RF-1 steel instead of R, since in electric-arc hardfacing vanadium is oxidized first of all, and the increased content of this element in RF-1 compensates to some extent for burn-off.

Among the low-alloy high-speed steels, grade EI-184 has proved itself well. Its composition is: tungsten 3.5–5.0%; chromium 7–9%; vanadium 1.1–1.3%; and carbon 0.8–1.0%. The cutting properties of this substitute are almost the same as those of high-speed steel R.

Reasoning as in the preceding case, one may suppose that here 13–14% tungsten has been replaced by 2–4% chromium and 0.5–0.8% vanadium. Therefore one may try to use mixtures of RF-1 and EI-184 steels for hardfacing cutters. Taking into account the inevitable loss of vanadium, one may think that the best deposits will be obtained with a significant content of RF-1 scrap.

If ferrovanadium is available, the simplest procedure is to add it to the charge and thereby compensate for burn-off.

A stronger deoxidizer than vanadium is aluminum. Experience has shown that it is very convenient to add aluminum in the form of filings, mixing them with the flux in electric-arc hardfacing of cutting edges. Aluminum must be added in such an amount that, while protecting the fused hardfacing layer from oxidation, it would not alloy the steel being remelted. One or two tenths of a percent of aluminum are harmless for high-speed steel. A more substantial content makes the tool brittle.

In making cutters by the method of hardfacing with a carbon electrode, it is very easy to alloy the remelted waste of high-speed steel. In wartime, the possibility of alloying is in most cases limited by the shortage of alloying additions. But it may be assumed prospectively that the costs of alloying will be repaid by an increase in tool life.

Among the additions that unquestionably improve the quality of the tool, cobalt should be included. High-speed steels contain from 5 to 15% cobalt. The greater the cobalt content within these limits, the higher the tool life.

The addition of cobalt in electric-arc hardfacing is advantageous in that cobalt oxidizes only slightly, and it can be introduced into the charge (the remelted chips) without adding an allowance for burn-off.

There are reports of successful experiments in alloying high-speed steel with nitrogen [21].

Nitrogen stabilizes austenite and, on the constitution diagram, expands the $\gamma$-region. In casting high-chromium steels, nitrogen is usually introduced in the form of ferrochrome containing about 2% nitrogen. Nitrogen can be added in the form of aluminum nitride (AlN). Nitrogen can also be added by introducing nitrided chips of high-speed steel into the charge. The chips are nitrided in ammonia at a temperature of 560°. Nitriding of chips can also be carried out in molten red prussiate of potash, or in any bath for low-temperature cyaniding of steel. Such a bath operates at a temperature of 530–560° and practically only nitrides the steel, but does not carburize it, as some investigators assert [16].

In high-speed steel, as in other steels, under the influence of nitrogen the austenitic grain grows more slowly.

There are very incomplete data on the effect of nickel on the cutting properties of high-speed steel. One of the works [17] indicates its positive effect.

Stellites of foreign origin contain significant amounts of cobalt, and only in Soviet stellites (“smena”) has cobalt been successfully replaced by nickel.

In superhard alloys the carbides are also bonded by cobalt (“pobedit,” “carboloy”). It may be expected that nickel, in certain combinations with the principal alloying elements of high-speed steel, can produce an improvement in the cutting properties of the steel. Its influence should be studied as applied to cast cutters.

It is known that part of the tungsten in high-speed steel can be replaced by molybdenum without lowering its cutting properties. Apparently, the replacement of tungsten by molybdenum takes place in a ratio of five to two; that is, five percent tungsten is equivalent to two percent molybdenum [18]. But even to ordinary high-speed steel (with 18% tungsten) molybdenum is sometimes added in amounts from 0.5 to 2.5%.

Molybdenum has a positive effect on the cutting properties of steel, but molybdenum steel is easily decarburized during heat treatment.

from the surface, and also loses part of its molybdenum in the form of a volatile oxide at comparatively low temperatures.

The latter circumstance, precisely in the case of cutters hardfaced with a carbon electrode, is not of special significance, since the tool may be not hardened at all.

Molybdenum does not burn out in liquid steel, because liquid iron, in the case of an alloy with a low molybdenum content, is oxidized more readily than this element.

It is necessary to mention one more alloying element—boron. Small additions of boron cause refinement of the steel grain because the borides that form are centers of crystallization [19].

In making cutters by our method, the possibility is not excluded that some amount of boron may pass from the flux containing borax.

Usually boron carbide (\(B_6C\)) is formed when boron oxide is heated with carbon in an electric arc furnace [20]. In melting cutters, just such a process may occur, and small amounts of boron carbide may pass into the molten steel.

It must be borne in mind that high-speed steel is a very complex alloy. The behavior of the individual alloying elements is not always entirely clear. It is possible to predict theoretically, only approximately, along what paths its further development will proceed. An in-depth study of the influence of various alloying additions will undoubtedly lead to the production of alloys with higher cutting properties.

It is precisely in arc hardfacing of cutters with high-speed steel that the possibilities for alloying are extremely great, and the process of adding various elements is especially simple when a carbon arc is used.

When hardfacing with an alternating-current carbon arc, some carburization of the remelted steel occurs, and the carbon content in the weld metal may increase.

Gulyaev [24] believes that in modern high-speed steel the carbon content may be increased somewhat. Some carburization of the hardfacing does not reduce the cutting properties of the tool.

An increased carbon content increases the stability of austenite; therefore, if cutters hardfaced by our method do not require hardening, tempering is necessary. The best results are given by double and triple tempering.

The described method of making cutters from high-speed-steel waste is very simple. Alloying and deoxidation are not very difficult: alloying additions may be introduced into the remelted charge (chips or fragments), and the deoxidizer (aluminum) into the flux. The hardfaced cutter does not require hardening; it requires only tempering.

As was stated, there is an opinion, and it is confirmed by experiment, that annealing does not affect the structure of the primary carbides of high-speed steel. But this is true only when the steel is cast into large ingots and the crystals of ledeburitic carbides are large.

The smaller the casting, the finer the carbides. In small crystals the highly developed surface is very large in comparison with their volume. During annealing, such carbides should assume a compact, granular form.

Even at the relatively low temperature of ordinary annealing (850–900°), spheroidization of the primary carbides occurs in arc hardfacings of cutters. This considerably increases the impact toughness of the steel.

If there is no need for spheroidizing annealing in making, by the authors’ method, turning and undercutting cutters, such annealing should significantly improve the quality of cut-off cutters, which operate under more severe conditions than turning cutters.

From all that has been set forth, several conclusions may be drawn concerning the particular case of manufacturing a cast cutting tool by various methods of electric-arc hardfacing:

a) owing to the small mass of molten high-speed steel crystallizing at the site where the cutting edges are formed, a structure arises that possesses very good cutting properties. The primary carbides in such a structure are precipitated in the form of fine-grained inclusions. An attempt may be made to refine the carbides still further by supercooling the liquid deposit, which is achieved by pressing upon it from above with a metal plate; by removing heat through direct contact, this plate accelerates cooling and causes additional centers of crystallization to appear in the solidifying deposit;

b) because of the small dimensions of the casting and the absence, in electric-arc hardfacing, of severe overheating of the molten steel, the deposited cutter is well hardened when it is cooled in air;

c) steel deposited by a voltaic arc has, in comparison with remelted high-speed steel, an increased carbon content. This circumstance, as is known, increases the thermal stability of austenite, which requires repeated tempering of the cutter;

d) double or triple tempering is the only heat-treatment operation for cutters deposited by our method;

e) to improve the quality of the tool, the deposits may be alloyed over a wide range;

f) ordinary annealing causes a certain spheroidization of the primary carbides, which improves the cutting properties of the cutters.

CHAPTER II

TECHNOLOGY OF HARDFACING CUTTERS WITH HIGH-SPEED STEEL

1. General Provisions

As was indicated above, the simplest method, one accessible for production and not requiring scarce materials, is the electric-arc method of depositing high-speed steel with a carbon electrode. This method requires neither the use of scarce calcium carbide and oxygen, as in the application of an oxyacetylene flame, nor the manufacture of electrodes from high-speed steel with a complex coating. Working with a carbon electrode is simpler than working with a metal one and requires a lower qualification on the part of the welder. Electric current is available at all enterprises; welding transformers of ordinary design or direct-current welding machines are widely used in industry and can be employed for the hardfacing of cutters. When a voltaic arc burns between the metal being deposited and the carbon electrode, a high temperature of about 3800° is obtained, which produces very powerful thermionic emission from the surface of the cathode spot on the carbon electrode, making it possible to obtain a current density reaching up to 1000 amperes per 1 cm² [23]. With the low thermal conductivity of the carbon electrode, it is easy to maintain a high cathode temperature even at low currents, on the order of one ampere, provided, of course, that the electrode diameter is appropriate. This property of the “carbon arc” is very valuable for hardfacing work and occurs not only when operating on direct current, but to a very considerable degree also on alternating current.

We set ourselves the task of using both direct and alternating current for hardfacing cutters, taking into account that, although when working on direct current the arc burns more steadily, and work on direct current

therefore simpler, but in production welding transformers are more widespread than DC welding machines, and carrying out hardfacing with alternating current is more accessible for production.

In developing the hardfacing technology it was necessary to resolve a number of questions connected with the choice of materials for the mold, preventing the liquid deposited metal from flowing off the body of the tool; with the choice of fluxes; with obtaining a dense, nonporous deposited layer; with the possibility of using for hardfacing not only high-speed steel but also its low-alloy substitutes; with the choice of the heat-treatment regime for the hardfaced cutters, etc.

All these questions were resolved by a large number of hardfacing experiments.

The experiments showed that hardfacing of cutters can be carried out equally well both with direct and with alternating current; moreover, alternating current can be used from any welding transformer with a reactor for current regulation.

When direct current is used, the carbon electrode should be connected to the negative pole and the tool being hardfaced to the positive pole, in order to avoid intensive carburization of the deposited layer.

It should be noted that when hardfacing in graphite molds, some carburization of the metal of the deposited layer is obtained, and when borax is used as the flux, boronizing of the deposit occurs. As a result, the hardness of the deposited steel after annealing is within the range of 35–40 units on the Rockwell “C” scale. A deposit of such hardness is difficult to machine with metal-cutting tools. In making cutters by hardfacing, this circumstance is not of essential importance, since the cutters can be worked with abrasives. But in hardfacing cutting tools which, in the course of manufacture, require machining with cutters, milling cutters, and other metal-cutting tools, this presents a certain obstacle. Hardfacing in a fireclay mold with a flux not containing borax is not accompanied by boronizing and accordingly gives a lower hardness after annealing.

Of especially great importance is obtaining a dense deposited layer without bubbles and pores. The first experiments that we carried out on hardfacing cutters gave, in this respect, completely unsatisfactory results. The deposited layer was porous, with a large number of bubbles, resembling a sponge, and in such a form could under no circumstances serve for the manufacture of a cutting tool. It was possible to eliminate the porosity after a series of experiments by introducing deoxidation of the deposited metal with the aid of silicon or aluminum. The best results were given by deoxidation with aluminum. The introduction of various amounts of aluminum was tested, and it proved that to obtain a deposited layer it is sufficient to add aluminum in an amount of 2–4% of the weight of the metal being deposited. A larger amount of aluminum leads to a reduction in the cutting qualities of the hardfaced tool.

It was found that a porous deposited layer can also be obtained when deoxidizing with aluminum if the hardfacing is carried out in insufficiently refractory fireclay.

Special experiments established that the deposited layer does not require quenching, receiving it automatically in the process of cooling from the molten state. To obtain high cutting qualities and red hardness, the tool must, after hardfacing, be subjected to repeated tempering. The tempering temperature was established for hard—

tempered high-speed steel of type RF-1 within the range 550–580°, and for hardfacing from low-alloy high-speed steels EI-184 and EI-262 within the range 540–560°. Thus, the tempering temperature for the hardfaced high-speed steel corresponds to the tempering temperature of the forged high-speed steel. Cutters subjected immediately after hardfacing to cooling in oil showed, during testing, no advantages in comparison with cutters cooled after hardfacing in still air.

The absence of any need for special hardening of the hardfaced cutters substantially simplifies their manufacture, freeing production from carrying out a complex heat-treatment operation requiring special thermal equipment.

In addition to hardfacing with steel of type RF-1, the possibility of hardfacing with low-alloy high-speed steels EI-184 and EI-262 was investigated. The experiments showed the full possibility of applying to the hardfacing of these steels the same technology as for RF-1 steel, and tests of the hardfaced cutters revealed their cutting properties and showed that a high-productivity tool can be obtained by using EI-262 steel in pure form for hardfacing and EI-184 steel in a mixture with RF-1 steel.

Below we give a description of the technology of hardfacing cutters.

2. Technological process for hardfacing cutters with high-speed steel using a carbon electrode

Hardfacing of the cutting elements of cutters is carried out with high-speed steel of types RF-1, EI-262, EI-184. Low-alloy high-speed steel EI-262 may be used for hardfacing in pure form and in this case gives high cutting qualities. Steel EI-184 in pure form is not suitable for obtaining a high-productivity tool by hardfacing. In order for cutters hardfaced with EI-184 steel to give high cutting qualities, from 25 to 50% (of the total weight of the deposited metal) of RF-1 steel must be added to it during hardfacing (see Chapter IV). The steel for hardfacing is used in the form of chips or in the form of pieces with a volume from 10 to 100 cubic mm (up to 5 mm in cross section).

The chips for hardfacing should be crushed in a cast-iron mortar so as to obtain a material that can be poured onto the shank in a more compact mass. This is necessary both to reduce oxidation and to reduce the number of deposited layers. The crushing of the chips should not be carried to a powdery state. The size of the individual pieces should be 1–3 mm.

The shank for hardfacing is made of carbon tool steel with a carbon content of 0.4–0.5%. The shape of the shank is the usual one, as for the manufacture of cutters with brazed or welded-on plates of high-speed steel. No special precision is required in making the seat, but smooth transitions are necessary, and sharp and right angles without rounding are impermissible (Fig. 2).

To retain the poured-on chips or pieces of high-speed steel and the liquid deposited metal at the end of the shank, it is necessary to surround the hardfacing area with a mold. This mold is best made in the form of plates of lump graphite; the plates may also be made from electric-machine brushes (graphite brushes of grades G1, G2, G3, G4, G5) or cut from old used graphite crucibles. With these plates of refractory materials the hardfacing area is surrounded on three sides, forming a closed bath for the metal being deposited. The plates are cut down in their upper part by 1.5–2 mm so that—

the deposited layer is wider than the holder, which ensures penetration of the edges and facilitates sharpening of the hardfaced cutters.

In Fig. 3 the mold is made of three pieces: а, б, в, but it may be made of two parts and even cut out as a single piece from one block of graphite. The latter provides faster and easier work for the welder, but requires more accurate manufacture of the holders. If the mold is made of two or three parts, then during work they are held together with a clamp of one design or another, or inserted into a fixture specially made for this purpose.

The mold can also be made of finely crushed fireclay, to which 15–20% refractory clay is added for bonding. From the mass mixed with water a mold is made, dried, and fired at a temperature of 700–800° to remove moisture. In this case a one-piece mold is obtained, which is simple to manufacture and, with careful work, can be used repeatedly. Fragments of molds are used for making new molding compound.

Fig. 2

Fig. 3

The mold may also be made from ordinary fireclay brick in the form of plates or as an integral mold.

It is most convenient to use fireclay molds for hardfacing cutters, made by the method of ordinary foundry molding. For this, the same molding compound is used as is employed for steel castings. Ground and screened fireclay (grain size 12 mesh) is mixed with clay diluted in water in an amount of 20% of the weight of the fireclay.

As experiments on hardfacing cutters in a fireclay mold have shown, the fireclay must be sufficiently refractory. Otherwise the hardfacing is porous. The cause of the porosity is gases released from the molten flux and penetrating the liquid deposited metal. If the fireclay is low-melting, then the flux, fusing with it, produces a viscous slag, which creates such great resistance to the escape of gases through its thickness that they prefer to escape through the metal. With sufficiently refractory fireclay, the flux almost does not fuse with the fireclay of the mold, the slag remains liquid, and the hardfacing is obtained completely dense.

We obtained good-quality hardfacing in a fireclay mold by using fireclay from the refractory-products plant of Glavogneupor at the Nizhne-Tagil station, grade B, melting temperature 1690°. The composition of the fireclay according to the plant analysis was: SiO₂—60%; Al₂O₃—36.54%; Fe₂O₃—2%; CaO—0.46%; MgO—0.56%. The moisture content of the molding compound is taken to be the same as for ordinary foundry molding. The prepared compound is poured into an iron box and lightly rammed. The cutter holder intended for hardfacing is joined to a wooden pattern made to the shape of the holder seat. Figure 4 shows a wooden pattern with pins а, preventing movement of the pattern along the holder.

In Fig. 5 the cutter holder is joined to the pattern, which is set into the socket of the holder. In this form the holder with the wooden pattern is pressed into the molding fireclay mass poured into a box, as shown in Fig. 6. After the holder with the pattern has been pressed in, the fireclay mass is rammed around the holder and the pattern; then the wooden pattern is removed from the fireclay mass, owing to which a space is formed above the socket of the holder intended to receive the high-speed steel deposited on the socket, the flow of which from the socket will be prevented by the surrounding mass. Figure 7 shows

Fig. 4

Fig. 5
holder

Fig. 6
holder; pattern

the holder placed in the fireclay mass with the wooden pattern removed. The holder must touch the wall of the box at point a in order to conduct the current. After the mold has been made, hardfacing can be begun at once. Drying of the mold is not required, since it takes place in the first moments of the operation under the action of the heat of the voltaic arc.

Excess moisture in the mold causes a slowing of the heating of the holder before hardfacing and increases the danger that the deposited metal will absorb hydrogen formed during the decomposition of water vapor emerging from the molding mass beneath the cutter.

Fig. 7
holder; space for hardfacing

In order to reduce the amount of water vapor released from the mold, the following molding method is recommended: the dry molding mass is poured in and rammed up to the plane AB (Fig. 7). On this plane the cutter with the wooden pattern on it is placed, as was indicated above in Fig. 5 and Fig. 6. Then the moist molding mass is poured around the pattern and rammed.

In this case no water vapor emerges from beneath the cutter, which gives an additional improvement in the quality of the hardfacing and reduces heat consumption; as a result, the productivity of the hardfacing work is increased.

Hardfacing may be carried out with alternating current from an ordinary welding transformer ST-2 with a reactor, or from any welding trans-

from a transformer capable of supplying a current of the required strength, and also with direct current from an ordinary DC welding set.

When using direct current, the carbon electrode should be connected to the negative pole, and the cutter being hardfaced to the positive pole, in order to prevent intensive carburization of the deposited layer.

A carbon electrode of small diameter is used so that it can work in a narrow space. An ordinary carbon electrode (used for arc lamps), when of small diameter, becomes strongly heated, burns away quickly, and its working end has to be dressed. Therefore it is better to use a graphitized electrode of the C-Kudinovo grade, which, owing to its higher conductivity, heats less. The electrode can also be made from lump graphite.

Heating of the electrode accelerates its burning and causes stoppages in the work because the electrode holder becomes heated. It may be recommended to cool the electrode and electrode holder by immersion in water. Unfortunately, not every carbon electrode withstands this operation. Some electrodes crack when soaked; in that case, of course, this method of cooling will be unsuitable.

The current for hardfacing is selected according to the diameter of the electrode and its material from the following tables.

Table 2

Current for a graphitized electrode of the C-Kudinovo grade

Electrode diameter, mm	Current, A
5	0
10	180—700
12.5	210—250
15	250—280

Table 3

Current for a graphite electrode

Electrode diameter, mm	Current, A
5	100
10	190
12.5	250
16	325

Table 4

Current for a carbon electrode

Electrode diameter, mm	Current, A
5	25
10	70
12.5	90
16	120
22	170

The end of the electrode is filed to a conical shape over a length of 2.0 to 2.5 electrode diameters (Fig. 8). It is also possible, instead of filing the end of the electrode with a file, to grind it on a coarse-grained emery wheel. Grinding is easier to do and there is less danger of breaking the electrodes.

The electrode holder is most simply made from sheet red copper with a cross-section of \(2 \times 50\) mm (Fig. 9). The ends of the prepared holder can be inserted into an ordinary electrode holder for metal electrodes. As the carbon electrode burns away, it is advanced from the electrode holder. The holder in Fig. 10 is somewhat more difficult to manufacture, but it has greater elasticity and makes it possible to apply greater pressure to the carbon electrode when clamping it, which reduces the contact resistance.

carbon—the holder, and consequently will also reduce the heating of the electrode and the holder.

High-speed steel intended for hardfacing must be free of rust, oil, and contamination; the pieces of steel may have any shape, but their size must be convenient for pouring into the prepared holder; the volume of the pieces must be no more than 100 cubic mm, as indicated above. If the steel for hardfacing is used in the form of chips, then it must also be free of rust and oil; in addition, for a more compact charge during hardfacing, it is crushed in a cast-iron mortar.

Fig. 8

Fig. 9

The following are used as fluxes in hardfacing:

a) for hardfacing in graphite molds:
remelted and pulverized borax . . . . . . . . . . . . . . . . . . . 90% (by weight)
aluminum filings . . . . . . . . . . . . . . . . . . . . . . . . . . . 10%

The aluminum filings must be obtained from a rough file. Finer filings are unsuitable, since they may burn up uselessly.

b) For hardfacing in fireclay molds:
marble . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52% (by weight)
fluorspar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32%
aluminum filings . . . . . . . . . . . . . . . . . . . . . . . . . . 16%

The amount of flux used in hardfacing each cutter must be taken from the calculation that the aluminum contained in the flux constitutes, by weight, about 3% of the weight of the metal being deposited.

Aluminum may also be omitted from the flux and introduced directly into the metal being deposited. The introduction of aluminum with the flux or directly into the deposited metal is necessary in order to obtain a dense, pore-free layer. In this case aluminum acts as a deoxidizer of the liquid bath. Without aluminum, obtaining a dense layer is difficult. Increasing the amount of aluminum above 3–4% of the deposited metal reduces the cutting qualities of the cutters.

Fig. 10

When aluminum is introduced into a flux containing borax, the amount of free aluminum in the slag is greater than when aluminum is introduced into a flux without borax (marble and fluorspar). This is explained by the fact that, under the conditions of the welding focus, boron is reduced by carbon; the reduced metallic boron is a good deoxidizer, as a result of which the consumption of aluminum for deoxidation decreases. In the case of using a flux that does not contain borax, a more considerable amount of aluminum is used for deoxidizing the steel, and a smaller amount passes into the slag. However, in the metal the concentration of aluminum is approximately the same and, apparently, does not exceed hundredths of a percent in the solidified deposited layer, which was confirmed by spectral analysis.

The procedure for carrying out the process of hardfacing cutters is as follows. After placing the cutter shank in the mold, the surface of the shank, at the place where the cavity for hardfacing has been selected, is sprinkled with a layer of flux. Then the arc is struck at point $k$ (Fig. 3) and the cavity on the shank is heated. After heating the shank, which is determined by its relatively slow darkening after the arc is broken, a layer of chips 2–3 mm thick is poured onto the cavity of the shank, and a layer of flux is poured on top. The arc is again struck at point $k$, and as it is moved along the shank the chips melt. Special attention must be paid to good heating of the shank along the edges of the cavity. When the charge has melted and covers the entire surface of the cavity with an even layer, the arc is broken and a small amount of flux is poured onto the molten surface of the metal. Onto the layer of flux, chips are poured in a layer up to 10 mm thick. Onto this layer of chips a layer of flux is poured, lightly covering the entire surface of the chips. The arc is again struck at point $k$ and the charge is melted over the entire surface. In doing so, it is necessary to ensure that the surface of the metal is level and calm; for this purpose a certain interval of time should be allowed for gases to escape from the bath. If the gases do not escape completely from the molten metal, the hardfacing will be porous.

The third layer is deposited in exactly the same way. Usually three layers of hardfacing are sufficient to obtain a layer of high-speed steel 6–8 mm thick.

For greater homogeneity of the entire deposit, it is useful, before final cooling, to remelt it again to almost its full depth and to stir the entire molten part of the deposit.

This operation is useful both for the final removal of gases from the entire deposited mass and for reducing the quantity and size of carbide grains that may have separated out in the lower layers during deposition of the upper layers. This operation of stirring and remelting is carried out as follows. In Fig. 11, $a$ is the cutter, $b$ the molding mass, $c$ the edge of the fused molding fireclay mass, $d$ the slag located around the liquid molten bath $ж$. With electrode $э$, a voltaic arc is excited at the point of contact of slag $d$ with the surface of the liquid deposit, i.e., at point $и$, as shown in Fig. 11.

Then, after heating the bath, the arc is moved successively to points $к$, $л$, $м$ and back, doing this gradually. After the very first pass of the electrode along the indicated path, the bath is stirred very vigorously in the direction of the arrows, as is evident from Fig. 12, which is a section of the mold in Fig. 11 along plane $Q$.

In Fig. 12, $a$ is the shank; $d$ the slag bath; $ж$ the liquid bath of the deposit; $к$ the point of intersection of the surface of the deposited bath with the surface of the slag bath $d$.

Stirring is carried out until the bath of liquid metal is completely clean and smooth. After this the arc is interrupted, the metal is allowed to solidify, and the cutter is freed from the molding ...

material. The melting and mixing described above are, of course, also necessary when hardfacing in graphite molds.

When the deposited layer solidifies freely, its hardened surface takes the form of a convex meniscus, which makes it necessary, when sharpening the cutter, to remove a considerable amount of deposited metal and to use abrasives excessively.

Fig. 11

Fig. 12

To avoid this, it is recommended that immediately after hardfacing, while the metal is still completely liquid, a metal rod of steel or copper with a flat end be pressed against its surface, as shown in Fig. 13.

Fig. 13

Fig. 14-a

In Fig. 13, \(u\) is the cooling rod, \(ш\) is the fireclay molding compound, \(a\) is the holder, \(д\) is the slag displaced by the rod \(u\) to the sides of the rapidly cooling bath \(ж\). This operation, in addition to leveling the solidifying surface of the deposited metal, also causes accelerated cooling of the entire deposited layer and prevents the formation of coarse primary carbides.

By using a metal rod and giving it the desired shape, it is also possible to form shaped surfaces of the deposited metal.

After cooling to shop temperature, the cutters are tempered. For tempering it is necessary to use an electric resistance muffle furnace with temperature control, or salt baths with electric, ga—

with gas or other heating capable of providing the required temperature and maintaining it constant. The tempering of cutters hardfaced with high-speed

Fig. 14-b

steel of the RF-1 type is carried out at a temperature of 550–580°, with the best results obtained by triple tempering with a one-hour hold each time. Cool-

Fig. 14-c

ing after the third tempering should preferably be carried out slowly. The hardness of the cutters after hardfacing should ordinarily be no lower than 57–60 units

Fig. 14-d

Rockwell on the C scale. The hardness after triple tempering should be within the range of 63–66 Rockwell units on the C scale. The tempering temperature for cutters hardfaced with low-alloy high-speed steels must be changed accordingly. The tempering temperature for cutters hardfaced with a mixture of various high-speed steels must

approach the tempering temperature of the steel predominating in the deposited layer.

After tempering, the hardfaced cutters are sent for sharpening. It is recommended first to rough-sharpen the cutter on a wheel with a grit size of 24–36, and then to finish it on a wheel with a grit size of 60. Sharpening should be carried out with abundant cooling in order to prevent the possibility of overheating and burns.

Fig. 15

Fig. 16

Fig. 17

Fig. 18

The grinding angles for hardfaced cutters are taken to be the same as for normal cutters made of high-speed steel, and depend on the type of cutter and the grade of the material being machined.

If all the rules set forth are carefully followed, cutters are obtained with a dense hardfaced layer firmly bonded to the material of the holder, possessing high cutting qualities. No scarce materials are required to carry out the hardfacing.

Fig. 19

Fig. 20

To illustrate the hardfacing process, Fig. 14 shows a photograph of the mold before hardfacing, and Fig. 15 the mold after hardfacing. A comparison of them shows how much the molding compound is melted by the action of the heat of the electric arc. Fig. 16 gives a photograph of a cutter removed from the fireclay mold with slag adhering to it, and Fig. 17 shows the same cutter with the slag knocked off. Fig. 18 shows two cutters after hardfacing: “a” cooled with a copper rod, and “b” cooled freely, without a copper rod. On this cutter the rounded surface of the hardfacing is clearly visible.

Fig. 21

Fig. 22

Figures 19 and 20 show graphite molds. Figures 21 and 22 show a mold made of fireclay bricks.

3. Some Considerations on Further Improvement of the Hardfacing Process

We carried out the process of hardfacing cutters as described in Section 2 of the present chapter; moreover, the large number of cutters hardfaced by us in the course of the investigations (more than 500 cutters) enabled us to thoroughly verify the correctness of the technology developed and to become convinced of the reliable quality of the cutters obtained by hardfacing. In Chapter IV, exhaustive data will be presented on extensive tests of the hardfaced cutters, which showed their high cutting qualities. Most of the cutters were hardfaced by us using alternating current. The quality of hardfacing with direct current is all the more reliable because work with direct current, owing to the greater stability of the arc, is simpler.

Above we noted the valuable property of the “carbon arc,” which makes it possible to obtain a high temperature even at low current intensities. To this it should be added that the action of the arc can be further intensified if a longitudinal magnetic field is applied according to the diagram in Fig. 23.

Fig. 23                  Fig. 24

The solenoid a, by creating a longitudinal magnetic field, calms the arc; it ceases to wander from place to place over the welded part and contracts, reducing its transverse dimensions, i.e., instead of the dimensions indicated in the drawing by dotted lines, it becomes narrower, as indicated by the solid lines b, v, g, d. This stability of the arc is very important for hardfacing cutters, since it considerably facilitates manipulation of it, reduces the number of arc breaks and, consequently, thereby improves the quality of the deposited metal. For good stabilization, a sufficient number of turns of the solenoid will be from 8 to 12.

The structural design of the electrode holder with a solenoid is given in the book by Prof. Khrenov and Yarkho [22], from which the sketch of the holder in Fig. 24 has been borrowed.

In the holder shown, the solenoid is made of a copper tube, which at the same time serves both as a conductor and as a cooler, if a cooling liquid—water—is passed through it.

This improvement is very important under shop conditions, if hardfacing is performed while continuously maintaining a voltaic arc for an extended period of time. The conductor and the place of contact between the electrode and the holder become strongly heated by the Joule heat released during the passage of electric current, as well as by the action of radiant heat from the nearby burning arc. This heating of the holder forces work to be stopped and time to be lost in cooling it.

It should also be noted that the cooled tubular conductor can be made lighter and more compact, which is very important for convenience of work.

If, nevertheless, the considerable weight of the holder proves tiring for the welder, it may be recommended to suspend the holder on a cable with a counterweight. This is easy to do, since hardfacing of cutters is carried out on the welding table in one particular, quite definite place.

Serious attention must be paid to clamping the carbon electrode with sufficient pressure in order to reduce the contact resistance.

From contact theory it is known that contact resistance depends on pressure, decreasing as the latter is increased, and does not depend on the area of the contacting surfaces [23].

4. Possible Other Methods of Electric-Arc Hardfacing with High-Speed Steel

In addition to the simple method of hardfacing that we used and subjected to careful study, we also tested other methods, which we mention here, with the reservation that they were not subjected by us to careful development and investigation.

a) The Zerener method. This method consists in striking an arc between two carbon electrodes clamped in a special electrode holder. In this way an arc flame is obtained that burns independently of the metal being hardfaced, just as is the case in hardfacing with an oxyacetylene flame and with an atomic-hydrogen torch. The Zerener method gives somewhat less heat and requires more current than the method we used. The arc has a larger volume and gives a less sharp flame than the flame of an oxyacetylene torch.

The Zerener method is better when three-phase current is used, in which case the voltaic arc is struck between three carbon electrodes. Here the arc is more intense in its action, the flame is more pointed, and the electrodes are of smaller diameter. The apparatus required to carry out this method is more complicated, since no fewer than two welding transformers and three regulators are required.

It may be assumed that application of the Zerener method would be convenient for hardfacing surfaces of considerable area.

b) The method of hardfacing under a layer of flux. We noted that the disadvantages of the method of hardfacing with a high-speed-steel electrode are, in addition to the need to manufacture this electrode, also the need to prepare and apply a complex coating to the electrode. This can be avoided if hardfacing with a high-speed-steel electrode is carried out by the method of high-speed welding under a layer of flux. This method will have the advantage that here no carburization of the metal being hardfaced will occur at all.

Attempts to apply the two indicated methods gave positive results and are subject to detailed development.

CHAPTER III

METALLOGRAPHIC INVESTIGATION OF HIGH-SPEED STEEL HARDFACING

1. General Considerations

This section of the work is a study of the macro- and microstructures of cutters hardfaced with high-speed steel. First of all, we paid attention to the character of the fracture of the hardfaced layer.

As a result of the formation of a large number of crystallization centers, the primary ledeburitic carbides, which impart the basic structure to the frac—

…were precipitated in the form of fine crystals, and in grain size the fracture differed little from the fracture of forged steel. Only upon careful examination was it possible to detect individual zones of cast steel—the zone of transcrystallization and the inner zone, and even then not in all hardfacings. The inner part had no noticeable columnar crystallization. No pores, cavities, or other ingot defects were found in the fracture. The use of aluminum eliminated the possibility of boiling of the metal. In this case aluminum was not only a deoxidizer. The principal role of aluminum, in fact, was not only the deoxidation of the steel, but also the partial suppression of certain reactions accompanied by the evolution of gases. In any case, in the fine-crystalline character of the fracture, the hardfacings in no way resembled the fracture even of a small ingot and did not resemble the fracture of cast metal.

When the fracture of the hardfacing was compared with the fracture of stellite (a stellite plate of the same cross section and mass as the hardfacing was compared), the advantage in fine grain was on the side of the hardfacing. The fracture of the stellite was much coarser. The macrostructure of the hardfacings essentially repeats the fracture pattern.

To study the nature of the fracture, we broke the deposited layer away from the holder by blows of a hand sledgehammer. It should be noted that separation of the deposited layer always required a large number of heavy blows, and, as a rule, the deposited layer separated from the holder only after the holder had been severely deformed. This circumstance in itself indicates the high impact toughness of the hardfacing—considerably greater than that of cast steel.

When producing a hardfacing by our method, owing to the small dimensions of the ingot, a large number of crystallization centers is obtained. Therefore the ledeburitic carbides precipitate more uniformly and in a finer form than in cast steel; in ingots, even of small size—not more than 10 kg—carbide segregation was not observed, and there was almost no columnar transcrystallization. The uniformity of the carbide structure guarantees constancy of the cutting properties throughout the entire thickness of the deposited layer.

At the same time it should be noted that the carbides precipitate in the form of ledeburitic lace, but the individual areas occupied by this lace were small, and therefore the presence of ledeburitic carbides did not affect brittleness. The size of the islands of ledeburitic carbides varied within certain limits from tool to tool, but this, within certain limits, is not reflected in the durability of the tools. In any case, even slowly crystallizing hardfacing does not have such enormous carbide inclusions as any stellite alloy. Probably the modification of the alloy by a small amount of aluminum, and in some cases probably boron, has an effect here. Special experiments intended to check whether the cooling rate was sufficient for air hardening showed that there is no point in taking special measures to accelerate cooling—for example, by immersion in oil, blowing with compressed air, and so forth.

If the holder is heated normally, so that a good fusion zone is obtained and it is not excessively overheated, then quiet cooling in air gives a completely suitable normal structure. The fear of cracks appearing at the fusion site or in the hardfacing itself also proved unfounded. Special cooling conditions (cooling in a thermos, covering freshly hardfaced tools with a heat-insulating powder) are not required.

Cooling of tools in air, and even accelerated cooling of hardfacings with a metal rod, never caused, either in laboratory or in factory tests, the appearance of cracks, spalling of hardfacings, etc.

Therefore, the combination of air cooling after hardfacing with hardening proved very successful. Air cooling after hardfacing replaced hardening.

Special magnetometric experiments to determine the amount of austenite and martensite in the hardfaced layer cooled in air were not carried out. However, hardness measurements and cutting tests showed that one may speak of a normal ratio between the amounts of austenite and martensite in the hardfaced layer, i.e., of the ratio that we have in ordinary forged high-speed steel after hardening with cooling in an air blast or in oil. It must be noted, however, that owing to the increased carbon content in the hardfaced layers as compared with ordinary high-speed steel, the resistance of austenite to tempering in the hardfaced layers is usually higher than the resistance of austenite in ordinary high-speed steel. Boron, which passes in small quantities from the borax into the hardfaced layer, may probably also increase the stability of austenite, although there are no reliable data on this. Therefore, the only necessary heat treatment when hardfacing cutters with high-speed steel by means of a carbon electrode is triple tempering.

Our experiments showed that the best results are obtained with triple tempering at 550°–580° with a holding time of one hour each.

We sought the greatest possible simplification and reduction in the cost of cutter production, and obtained high-quality cutters using only a single heat-treatment operation—tempering. But during annealing the ledeburitic carbides must coagulate, and the structure must approach that of forged high-speed steel. Therefore, if it is necessary to obtain a cutting tool with increased impact toughness, an additional annealing operation may be introduced; in that case, however, special heating for hardening will of course be required.

In the process of hardfacing a cutter with high-speed steel by a carbon electrode, it is especially easy to alloy the hardfaced layer by introducing ferroalloys and pure metals into it. We carried out a number of experiments on alloying hardfaced layers with nickel, cobalt, molybdenum, and nitrogen, which gave very interesting results. Research in this field opens broad prospects for the future.

In all the normal hardfaced layers investigated, the structure is homogeneous. Segregation or liquation is practically not observed. This is quite natural, since with rapid cooling we have an accelerated passage through the liquidus and solidus lines, and during the hardfacing process there is vigorous mixing of the deposited metal. At the junction of the hardfaced layer with the base metal, when the hardfacing is properly conducted, gas bubbles and porosity are absent. The size of the transition zone at the point of fusion usually does not exceed 0.5 mm.

Specimens for metallographic examination were prepared perpendicular to the axis of the cutter; many cutters that had been in service were examined. The specimens were prepared by the usual method and etched with a 5% solution of nitric acid in ethyl alcohol. The structures were examined at various magnifications and photographed mainly at a magnification of 300.

2. Structure of the High-Speed-Steel Layer after Hardfacing

For hardfacing cutters, chips of high-speed steel grade RF-1 of the following chemical composition, in percent, were used (Table 5).

Technological conditions of hardfacing: the cutters were hardfaced by electric-arc welding with a carbon electrode, using alternating current. Hardfacing was carried out under various conditions of shaping of the part being hardfaced: in graphi-

Table 5

C%	Cr%	W%	V%
0.76	4.60	17.84	1.13

in powdered form, in the form of finely crushed chamotte with the addition of 20% clay. The cutting edges were hardfaced onto holders usually made of tool steel grade 45. In almost all cases, except where specially stated, aluminum was used as a deoxidizer, calculated as 3% aluminum filings by weight of the chips being hardfaced. The hardfacing of one cutter usually weighed 20–25 grams.

Figure 25 shows the microstructure of a cutter hardfaced with chips of high-speed steel grade RF-1, of the above-mentioned chemical composition. The microstructure was examined immediately after hardfacing; the cutter was not tempered. The structure is shown at a magnification of 300.

Fig. 25 ×300

Figure 26 shows the same structure at a magnification of 600. Hardfacing was carried out in a mold made of graphite plates, using a borax flux deoxidized with aluminum.

Fig. 26

The microstructure shown in Figures 25 and 26 consists of a fine network of ledeburitic carbides, chiefly iron carbides and carbides of the alloying elements—tungsten, chromium, and vanadium—against an austenitic-martensitic structure. This structure is typical of hardfaced cutters before tempering. It is related to forged high-speed steel by the fine-grained nature of the structure, and at the same time what it has in common with cast steel is the presence of a ledeburitic network. In the present case, the especially fine-grained character of the primary carbides is characteristic. Such a structure confirms the assumptions made earlier regarding the nature of crystallization in small ingots (see Chapter I). The structure is the same throughout the entire cross section of the hardfaced layer, with the exception of the transition zone, which usually has a thickness of approximately 0.5 mm. Particular emphasis should be placed on the absence of carbide segregation and of separate accumulations of carbide masses. This will undoubtedly ensure constancy of the cutting properties when the cutter is reground. The chemical composition of the hardfacing deposit, in percent, is given in Table 6.

Table 6

C	Cr	W	V
0.90	4.65	17.09	0.73

Figure 27 shows the structure of a cutter hardfaced with chips of the same chemical composition as in the first case. Hardfacing was carried out under the following conditions: a mold made of graphite plates; fluorspar and marble were used as the flux. Aluminum served as the deoxidizer.

Fig. 27

In the microstructure of the hardfaced cutter, several larger accumulations of ledeburitic carbides are noticeable, precipitated in the form of a network. In the transition zone there is a sharp boundary between the base metal and the hardfaced metal (see Fig. 28). The dark region in the microphotograph is the overetched material of the holder. Near the hardfacing we have a carburized layer with a carbon content of 0.9% and a pearlite structure. Because longer etching of the hardfacing is necessary, the material of the holder is, naturally, overetched.

In this case, with a sharp transition zone, we have a good weld, with complete absence of slags and gas bubbles.

Fig. 28

The chemical composition of the hardfacing (in percent) is given in Table 7.

Table 7

C	Cr	W	V
0.86	4.60	17.23	0.89

Figures 29 and 30 show microphotographs of a cutter hardfaced in a form made of crushed fireclay mixed with 20% clay. Fluorspar mixed with marble was used as the flux. Aluminum served as the deoxidizer. The cutting edge was hardfaced with chips of high-speed steel of grade RF-1, of the chemical composition already indicated. As a result of hardfacing carried out under these conditions, a series of transitional structures from the base metal to the deposited metal was obtained. The microstructure of the base metal is pearlite with a small amount of ferrite.

Fig. 29

Fig. 29 shows the transition zone from the material of the holder to the deposited layer, consisting successively of: a ferrite-pearlite structure of dendritic character, then a carburized layer with a carbon concentration of approximately 0.9%, and a pearlite structure. Further on, in some places the layer of deposited metal consists of a structure of clearly expressed acicular martensite against a background of polyhedral austenite.

Fig. 30

Subsequently, on passing to the deposit, inclusions of ledeburitic carbides appear in the structure, having the appearance of ledeburitic lace of rather fine structure.

Fig. 30 shows the microstructure of the deposited metal, on which ledeburitic carbides are clearly visible in the form of fine lace. The structure shown in Figs. 27, 28, and 30 is characterized by a somewhat coarser ledeburite morphology in comparison with the structure in Fig. 25.

Some coarsening of the network of primary carbides was reflected in an increase in the grain size in the fracture and, as special tests showed, did not affect the cutting qualities of the deposited tool.

The chemical composition of the deposited metal is given in %% in Table 8.

Table 8

C	Cr	W	V
1.00	4.50	17.28	0.90

Fig. 31 shows the microstructure of a cutter deposited in a fireclay mold using a flux of fluorspar and marble. In this case, deposition was carried out without deoxidation. The deposit consists of a martensitic-austenitic structure with a network of carbides in the form of a ledeburitic eutectic. In the structure there is a clear predominance of martensite over austenite. The structure is quite satisfactory. A good-quality cutter without the addition of aluminum can be obtained, but only provided that the work is carried out carefully by a very experienced welder. The introduction of aluminum, which protects the deposit from oxidation and degasses it, gives a good deposited layer practically in the hands of any welder. Po-

For this reason, we recommend in all cases of hardfacing cutters that aluminum be used in the quantities indicated earlier.

Fig. 31. ×300

The chemical composition of the hardfacing, in %, is given in Table 9.

Table 9

C	Cr	W	V
0.88	4.25	17.13	0.82

In Fig. 32, at higher magnification, martensite needles are visible. Structural inclusions of this kind are encountered in the hardfaced layer in those cases when the hardfacing has been carried out in several layers without substantial melting of the entire mass of deposited metal.

Fig. 32. ×300

In all the cases described, after the hardfacing process could be considered complete, a more sharply defined form of the cutting edges of the cutter was imparted to the solidifying metal bath by pressing with a metal rod whose cross section corresponded to the overall dimensions of the hardfacing. This pressing eliminated the convex meniscus that inevitably co-

accompanies every metallic pool formed in the process of hardfacing. The purpose of this operation was also to obtain a larger number of crystallization centers and a finer-grained structure through accelerated cooling associated with heat removal when the liquid hardfacing metal came into contact with the cold rod. In some cases the cutters were hardfaced without pressing with the rod, in order to investigate the effect of the process of additional cooling by the rod on the microstructure of the hardfacing obtained.

Fig. 33 ×300

Fig. 33 shows the microstructure of a cutter hardfaced without subsequent cooling of the liquid metal by the rod. In this case, RF-1 chips were hardfaced onto a shank made of steel 45.

The shank was molded in crushed fireclay. During hardfacing, borax mixed with aluminum, which is a deoxidizer, was used as the flux. The microstructure of the hardfacing obtained differed considerably from all subsequent structures by the rather coarse form of the separated ledeburitic carbides. In its structure it somewhat resembles the structure of cast steel of the carbide class (cast high-speed steel).

Fig. 34 ×300

Fig. 34 shows the microstructure of the metal under the same hardfacing conditions, but with cooling by the rod. When comparing the structures shown—

shown in Figs. 33 and 34, a finer austenite grain and a finer ledeburite network are clearly visible in the structure of the hardfacing with rod cooling. Thus, it may be considered proven that the use of rod cooling, in addition to leveling the surfaced layer, causes refinement of the structure as a result of an increase in the number of crystallization centers in the solidifying metal.

The chemical composition of the hardfacing is given in Table 10.

Table 10

C %	Cr %	W %	V %
0.86	4.20	17.02	0.90

3. Structure of the Surfaced High-Speed Steel after Annealing

To study the effect of annealing on the structure of the surfaced metal, a cutter hardfaced with high-speed steel of grade RF-1 was prepared. Surfacing was carried out in a fireclay mold with a flux of fluorspar and marble and with deoxidation by aluminum.

Figure 35 shows the microstructure of the cutter before annealing—some martensite, a large amount of austenite, and a coarse ledeburite network of primary carbides. A cutter with such a structure was annealed according to the following regime: heating together with the furnace to 870°, at a rate of 200° per hour, holding at this temperature for one hour, cooling together with the furnace to 720°, and holding at 720° for two hours.

Fig. 35 ×300

Figure 36 gives the microstructure of the specimen after annealing: grains of primary and secondary carbides uniformly distributed over the entire surface of the polished section. The structure of the main metallic mass is a sorbite-like eutectoid.

During annealing, owing to the small size and great branching of the primary carbides, the latter coagulate, forming grains, which does not occur during the annealing of large ingots, since the surface energy of coarse crystals is small in relation to the total energy of the lattice in comparison with small and especially very small branched crystals. Obviously, after quenching such annealed high-speed hardfacing, the primary car-

carbides retain their granular structure, and the impact toughness of the hardfacing should improve, as should the cutting properties. But annealing and subsequent hardening complicate the technological process of manufacturing cutters hardfaced with a carbon arc.

We did not use annealing followed by hardening, seeking the greatest possible simplification in the manufacture of cutters, and at the same time obtained a tool with extremely high cutting properties, as is evident from Chapter IV.

Fig. 36 × 300

In the case where increased impact toughness is required of the tool, it can be obtained by carrying out additional heat treatment—spheroidizing annealing followed by hardening and tempering.

Fig. 37 × 300

Incidentally, hardfacing with high-speed steel by an oxyacetylene flame, which is always recommended to be annealed and hardened, after annealing gives approximately the same structure (see Fig. 37).

4. Influence of Certain Conditions on the Structure of the Hardfaced Layer

During hardfacing, when the metallic bath is kept in the liquid state for an excessively long time, and when cooling is artificially slowed,

possibly yielding coarser ledeburite. This may also occur when the toolholder is excessively overheated.

A similar structure of a cutter subjected to threefold tempering at 550–560° C with a holding time of 1 hour is shown in Fig. 38. This structure

Fig. 38 ×300

is a coarser and more massive network of ledeburitic carbides and martensite.

Cutters with such a deposited-layer structure have somewhat reduced cutting properties. The microstructure of the cutter that showed the best

Fig. 39 ×300

cutting properties consists of ledeburitic carbides of finer structure and inclusions of low-carbon martensite. The structure differs significantly from the preceding one: the ledeburitic carbides are more fragmented, and the carbide network is less massive (see Fig. 39). The heat treatment of this cutter is similar to the preceding one.

5. Structure of the Layer Deposited with Low-Alloy High-Speed Steel

Since part of the present work was intended to determine the possibility of using low-alloy high-speed steels for hardfacing cutters, the structures of cutters hardfaced with low-alloy high-speed steels of grade EI-184 are presented below, as well as the structures of cutters,

hardfaced with a mixture of 50% EI-184 and 50% RF-1. After hardfacing, the cutters were tempered three times at a temperature of 560° with a holding time of 1 hour.

Fig. 40 shows the structure of a cutter hardfaced with low-tungsten high-speed steel grade EI-184. The structure is ledeburite and martensite. Characteristic here is coarse-needle tempered martensite, which is the predominant component of the structure against a background of a ledeburitic network. The ledeburite has a coarse carbide network.

Fig. 40 × 300

The chemical composition of the hardfacing is presented in Table 11.

Table 11

C %	Cr %	W %	V %
1.32	7.08	5.07	1.27

In cutting tests, cutters hardfaced with EI-184 showed reduced durability. It is difficult to find an exhaustive explanation of the reduced durability of the cutters in the nature of the microstructure. True, the form of ledeburite precipitation and the size of the austenite grain in the structure of this hardfacing cannot be considered entirely satisfactory; however, the sharp reduction in the cutting qualities of the EI-184 cutter can apparently be explained by the increased carbon content in the hardfacing compared with the usual composition of EI-184.

Fig. 41 gives the structure of a hardfacing made with a mixture of EI-184 and RF-1 steels in the quantitative ratio: 50% EI-184 and 50% RF-1.

The microstructure is characterized by a less massive network of ledeburitic carbides and medium-sized martensite needles.

The chemical composition of the hardfacing, in %, is given in Table 12.

Table 12

C %	Cr %	W %	V %
0.94	4.56	9.51	1.18

In cutting tests, the cutter showed cutting properties identical with a cutter hardfaced only with RF-1 steel. The microstructure is similar to the microstructure of a hardfacing deposit made with RF-1 steel in its pure form.

Fig. 41 ×300

6. Structure of Alloyed Hardfacing Deposits

In the present work, several experiments were carried out on introducing alloying elements into the deposited high-speed steel with the aim of improving the quality of the hardfaced tool. The hardfacing deposits were alloyed with nitrogen, nickel, molybdenum, and cobalt.

Alloying with nitrogen, nickel, and molybdenum did not give clear results. Alloying the deposits with cobalt in amounts of 5, 10, and 15% gave very interesting results.

On the basis of data in the literature, very interesting results may be expected from alloying high-speed steel with nitrogen. Under conditions of hardfacing with a carbon arc, nitrogen alloying is simplified by the possibility of low-temperature cyaniding of the chips. In low-temperature cyaniding, nitriding alone will take place.

Fig. 42 ×600

Cyaniding can be carried out by heating in any cyanide baths. Nitriding can be carried out in any installation for nitriding parts in an ammonia atmosphere.

Figure 42 shows the microstructure (at 600× magnification) of a cutter hardfaced with Rf-1 high-speed steel to which nitrided chips had been added. Isolated grains of primary carbides are visible on the polished section. This structure very closely approaches the structure of forged steel. Cutting tests with such cutters have not yet given a clear answer

Fig. 43 ×300

as to their advantages. But there is reason to continue these investigations, since the effect of nitrogen on the structure of ledeburitic carbides is favorable.

Figure 43 shows a micrograph of a cutter hardfaced with high-speed steel with the addition of 15% Co, based on the weight of the deposited metal. The cutter was tempered three times at 570° with a holding time of 1 hour each time.

Fig. 44 ×300

The structure of the deposit with 15% cobalt is very interesting. First of all, the relatively smaller amount of primary carbides is characteristic in comparison with any deposit not alloyed with cobalt. Owing to the smaller amount of primary carbides, the ledeburite network has the character of a light lacework of an exceptionally fine structure. The structure of the eutectic can be distinguished only at 1000× magnification. Here we have a more broken-up carbide network in comparison with all the structures presented earlier. The carbide network is interrupted, giving only slight outlines of the austenitic grain.

With the exception of small islands of a fine eutectic network, the principal metallic mass has the structure of fine-acicular martensite, which is the product of the decomposition of cobalt-alloyed austenite. From the standpoint of structure, the high cutting qualities of a cutter with 15% cobalt can be explained, on the one hand, by the fact that cobalt increases the solubility of carbides in the principal mass. Cobalt raises the carbon content in austenite, and the amount of primary carbides will be smaller.

On the other hand, martensite alloyed with cobalt is harder and more thermally stable.

In Fig. 44, for comparison with the structure of the hardfacings, the structure of a cast alloy of the stellite type (acrite) is shown. If in stellite the cutting part consists of carbides, then in hardfaced cutters the cutting properties are determined not only by the carbides, but also by the martensitic part of the structure. When comparing micrographs of stellite and of the hardfacings, the latter are characterized by a considerably smaller size of carbide grains and by their isolation from one another.

7. Change in the Chemical Composition of High-Speed Steel During Hardfacing

In addition to metallographic analysis, chemical analysis of the hardfacings was carried out. Without presenting all the analyses performed, apart from those included in this chapter, the following conclusions may be drawn. Under all conditions of hardfacing with a carbon electrode, with various kinds of molding and with the use of different fluxes, a somewhat increased carbon content in the hardfacing should be noted in comparison with the carbon content in the initial high-speed-steel shavings. This increase in carbon content does not exceed 0.2%. According to the literature, such an increase in carbon content is even desirable; the usual carbon content in modern high-speed steels may be considered too low. [24]

As laboratory and shop-floor experiments have shown, the increased carbon content is not reflected in the durability of cutters hardfaced with high-speed steel using a carbon electrode. If we turn to theoretical metallurgy and use data on the heats of formation of individual oxides, then, as a first approximation, it may be considered that, in terms of the rate of burning out, the individual elements, in accordance with the heats of formation of their oxides, are distributed according to the data given below [25].

It should be noted, however, that this assumption is made as a first approximation, since it does not take into account the concentration of the elements in the steel or the protective influence of aluminum.

Heat of Formation of Oxides

Element	Oxide	Heat of formation, cal/mol
Al	Al₂O₃	500
B	B₂O₃	352
Cr	Cr₂O₃	338
V	V₂O₂	263
Si	SiO₂	252
W	WO₂	158
Fe	FeO	79
Co	CoO	72
Ni	NiO	72

Proceeding from these approximate considerations, it may be assumed that during hardfacing Co and Ni will not burn out. W and V will burn out only slightly, and Cr most of all.

8. Conclusion

The question of the relationship between the cutting qualities of a cutter and its microstructure is very complex. If it can be asserted that coarse ledeburitic carbides cause brittleness in a cutter, the question of the influence of the degree of alloying of martensite on the cutting process is far more complicated. We did not set ourselves the task of relating the structure of the cutter to its cutting qualities. This would require an extensive and special investigation.

Our aim, by means of the microphotographs presented, was to show that our structures are finer than the structures of large ingots, and that in grain size they approach forged steel; however, elements of a cast structure are, of course, present. Testing hundreds of hardfaced cutters showed their excellent cutting properties. This compels us to pay especially close attention to the properties of small ingots and to carry out further work on studying the crystallization of small volumes of metal, which has its own specific and very important features.

CHAPTER IV

Testing of Hardfaced Cutters

The tests of hardfaced cutters were intended to reveal their cutting qualities. The tests were carried out both in laboratory and in production settings. Neither one nor the other, taken separately, could be considered sufficient for an exhaustive judgment of the quality of hardfaced cutters. Laboratory tests are conducted under ideal conditions: on a stable, well-adjusted machine tool, without vibrations, on a clean, previously rough-turned blank, with a constant depth of cut, with careful control of the cutting speed and with its minimum permissible fluctuations, with systematic control of the hardness of the blank both along its length and across its cross section, with careful sharpening and setting of the cutter, with precise determination of the moment of its final blunting, etc. Laboratory tests make it possible to determine accurately the cutting qualities of cutters and to compare them with any other cutters. Carrying out such tests in production, under shop-floor conditions, is impossible, if only because their performance requires a machine tool with stepless speed regulation and with constant control of those speeds; and also because, under production-work conditions—especially in roughing operations—one inevitably has to deal with scale, crust, and contaminated areas on the surface of the blank; with varying hardness of the material being machined in some regions; with runout of the blank in the centers and, as a result, with variable depth of cut; with shocks and vibrations; with varying quality of cutter sharpening, often performed without cooling and accompanied by burns; and with other circumstances that are difficult to take into account and that, to one degree or another, affect the service life of the cutting tool. Therefore, on the one hand, for an accurate judgment of the cutting qualities of cutters their laboratory testing is necessary; on the other hand, laboratory tests alone do not make it possible to judge the full suitability of cutters for work under production conditions, with all their attendant circumstances.

Therefore laboratory and production tests do not always give identical quantitative results.

Taking this into account, we carried out three types of tests of hardfaced cutters:

1) laboratory tests to determine the cutting properties of the hardfaced cutters and to compare them with the cutting properties of cutters with tips made of high-speed steel RF-1, subjected to optimal heat treatment;

2) laboratory tests to compare the hardfaced cutters with the cutters ordinarily used in production, having tips of high-speed steel RF-1 with factory heat treatment;

3) production tests of the hardfaced cutters under shop conditions on normal production operations.

Such a testing procedure provided exhaustive information both on the cutting properties and on the production suitability of the hardfaced cutters.

I. Laboratory tests of hardfaced cutters

1. CONDITIONS OF THE LABORATORY TESTS

The tests were carried out in the metal-cutting laboratory of the Tomsk Polytechnic Institute on a DIP-30 lathe. The machine is equipped with a direct-current motor with stepless speed regulation within the range 1.3:1, which covers the denominator of the progression of the series of spindle speeds \((G = 1.26)\) and makes it possible to obtain any spindle speed within the known limits and to maintain its required constancy. The spindle of the machine is equipped with an accurate stationary tachometer with a large scale, which makes it possible to set and maintain the constancy of the speed with deviations not exceeding 1%. The motor power is 20 kw; the motor is connected to the machine by a V-belt drive; the friction clutch of the machine is strengthened in comparison with the usual one in the DIP-30 model. All this gives complete stability of speeds. We dwell on these circumstances because they are decisive for the success of tool-life tests.

The geometry of all cutters that underwent testing was as follows:

Front angle	\(\gamma = 18^\circ\)
Main clearance angle	\(\alpha = 10^\circ\)
Auxiliary clearance angle	\(\alpha_1 = 10^\circ\)
Main plan angle	\(G = 75^\circ\)
Auxiliary plan angle	\(G_1 = 15^\circ\)
Nose rounding radius	\(r = 1.0\) mm
Cutter body section	\(16 \times 24\) mm

The main plan angle was taken to be of considerable magnitude in order to reduce metal consumption during testing and to eliminate the possibility of vibration of the workpiece.

In almost all tests, the tool life of the hardfaced cutters was compared with the tool life of cutters with brazed tips of high-speed steel RF-1. For this purpose a batch of cutters was made with RF-1 tips of the same dimensions and with the same geometry. These cutters were subjected to the following heat treatment:

Hardening temperature \(1280^\circ\), oil quenching.

Tempering temperature \(550^\circ\). Tempering three times for 1 hour.

Hardness after tempering \(R_c = 64\text{—}66\).

Cutters with hardness below \(R_c = 64\) were not admitted to the tests. The tips for this batch of cutters were made from a single bar of RF-1 steel, which ensured high uniformity of their cutting properties.

All cutters were sharpened and resharpened on a “GISHOLT”-type machine with abundant cooling. No honing after sharpening was performed. No traces of burning after sharpening were permitted. During resharpening after dulling, the entire affected layer was removed.

Laboratory tests of the hardfaced cutters were carried out on the following materials: steel 45, with hardness \(H_b = 180\text{–}200\); carbon steel with hardness \(H_b = 240\text{–}250\); chromium steel 40X with hardness \(H_b = 200\). It should be noted that the various tests were carried out on different blanks, in some cases even at the same hardness but with different machinability. Therefore, any comparisons and conclusions could be made only within the limits of tests carried out on one blank, which was subsequently taken into account throughout. Constancy of hardness was checked both along the length of the blank and across its cross section; the permissible deviation in hardness within a single blank was within \(\pm 5\) Brinell units.

The cutting conditions in all tests without exception were as follows: depth of cut \(t = 2\) mm, feed \(S = 0.2\) mm/rev.

The tests were carried out without cooling.

Work was performed on a previously turned surface. Thus, a semifinishing regime was adopted. This was done in order to economize on the steel consumed in the tests.

It must be borne in mind that the ratio of the cutting qualities of different high-speed steels (especially low-alloy steels) under different test conditions may be different. Thus, for example, steel EI-262, when machining steel 45 with hardness \(H_b = 183\text{–}192\), shows higher cutting qualities on finishing chips than steel RF-1, while on roughing chips it is the same as steel RF-1. When machining OKhM steel with hardness \(H_b = 217\text{–}228\), steel EI-262 is inferior to steel RF-1 both on finishing and on roughing chips, and on roughing chips more significantly than on finishing chips [26].

Taking this into account, we carried out most of the tests on steels of high hardness, thereby placing the hardfaced cutters under perhaps less favorable conditions in comparison with RF-1 cutters. We may be reproached for the fact that all the tests were carried out on semifinishing chips and that cast iron, as the material being machined, was entirely absent from the tests. These gaps were filled by us in production tests, which, as will be seen below, were carried out both on cast iron and under roughing conditions on steels.

The criterion of dulling in the laboratory tests was taken to be the complete failure of the cutting edge of the cutter.

In all tests we did not content ourselves with comparing tool life at identical cutting speeds, but carried out tests at a series of speeds and constructed \(VT\) curves (speed—tool life), which made it possible to judge the cutting qualities of the cutters being compared most fully. We became convinced that, with homogeneous cutters, a constant blank with only slight fluctuations in hardness, and provided that the constancy of speed was carefully maintained throughout each test, all tests with tool lives beginning at 2–3 minutes fit very well into a general law. Therefore, we did not follow the instruction that, in carrying out tests, the tool life of a cutter should be no less than 10 minutes, as stated in the “general methods for investigating cutting conditions” [27], and consider it not always correct.

We now proceed to the direct presentation of the results of laboratory tests of the hardfaced cutters.

2. Testing of Cutters Hardfaced with RF-1 Steel in a Graphite Mold

Figure 45 presents the results of tests of cutters hardfaced with alternating current in a graphite mold, with borax used as the flux. The cutters included in these tests were the very first to be made by hardfacing. In their manufacture aluminum had not yet been used as a deoxidizer, and the cutters usually had some porosity.

The heat treatment of these cutters consisted of triple tempering at a temperature of 550° with a holding time of 1 hour each. It should be noted that the 10 experimental points on the graph were obtained by testing 8 cutters. The graph shows a good arrangement of the experimental points with respect to the drawn straight line, which indicates the quite stable quality of the cutters obtained already in the first hardfacing experiments.

In testing these cutters we did not compare them with cutters with RF-1 inserts. But if one takes the latest standard data on cutting conditions for cutters made of RF-1 steel, one can see that the standard straight line exactly coincides with the line drawn on the graph from the experimental points. This indicates that the cutting properties of the hardfaced cutters correspond to the cutting properties of cutters with RF-1 inserts. All experiments with these cutters were carried out on billets of steel 45, taken from one bar and possessing the same hardness.

Fig. 45

In view of the porosity obtained, attempts were made to find measures for combating it. As indicated earlier, the introduction of aluminum as a deoxidizer made it possible to obtain a layer of hardfaced metal that was dense and free of pores. A test was conducted of the first cutters obtained with the introduction of aluminum. The results of these tests are presented in Fig. 46. Here the results of tests of four cutters are denoted by different symbols (circle, square, triangle, cross). The graph shows that here as well, just as in Fig. 45, sufficiently stable results were obtained for all cutters participating in the tests. These cutters were hardfaced in a graphite mold with borax as the flux. The heat treatment of the cutters was the same as indicated above. The amount of aluminum introduced was not precisely controlled in this case.

Subsequently, experiments were undertaken to determine the influence of the amount of aluminum introduced during hardfacing on the cutting properties of the cutters. This was necessary because there are considerations indicating the harmful effect of aluminum on the cutting properties of high-speed steel. Gulyaev indicates that “aluminum has a very harmful effect on the cutting properties of steel, since its main influence is manifested in reducing the solubility of carbides in austenite at high temperature: 0.3% aluminum is already sufficient to noticeably reduce the durability of steel.”

Good results in terms of obtaining pore-free, dense hardfaced layers were reliably achieved when 2–3% aluminum, by weight of the hardfaced metal, was introduced.

deposited metal. Therefore cutters were made with aluminum introduced into the hardfacing in amounts of 0%; 0.2%; 0.7%; 1.0%; 2.5%; 3.5%; 4.0%. Hardfacing of these cutters was carried out with alternating current, in a graphite mold, with borax as the flux. After hardfacing, the cutters underwent the normal tempering schedule—3 times for 1 hour at 550°. The test results for these cutters are presented in Fig. 47. The figure shows an exceptionally good arrangement of the experimental points and makes it possible to draw a quite

Fig. 46

Fig. 47

confident conclusion that introducing aluminum from 0% to 4% of the weight of the deposited metal has absolutely no effect on the cutting qualities of the tool. All the aluminum, apparently, burns out, and only traces of it were detected by spectral analysis.

In addition to aluminum, we tested the introduction of silicon in order to obtain a dense hardfaced layer. The introduction of silicon likewise gave a deposited layer without pores, but with less reliability. In addition, it turned out that cutters hardfaced with silicon as a deoxidizer gave lower cutting qualities. This is seen from Table 14, which gives the lives of cutters hardfaced with one or the other deoxidizer, obtained as a result of tests on steel 40Kh at the same cutting speed.

On this basis, in all subsequent experiments on hardfacing cutters in a graphite mold, aluminum was introduced in an amount of about 3% of the weight of the deposited metal, which gave a dense layer and had no effect on the cutting qualities of the cutters. We also recommend this amount of deoxidizer for production. We drew the conclusion that aluminum has no harmful effect on the cutting qualities of hardfaced cutters on the basis of experiments carried out with only one metal. But subsequently all hardfaced cutters were deoxidized with aluminum and

Table 14

Cutting speed, m/min	Life in minutes: Cutters hardfaced with aluminum	Life in minutes: Cutters hardfaced with silicon
48.5	11.0	1.75
48.5	10.3	3.2
48.5	10.9	2.5
average life	10.7	2.5

as will be seen, in tests on other steels as well, and in production tests on cast iron and brass, they gave equally good results.

In all the experiments cited above, the hardfaced cutters did not undergo special hardening; after the hardfacing operation, having been removed from the mold, they cooled in still air. We carried out an experiment on hardening hardfaced cutters in oil. For this purpose a cutter hardfaced in a graphite mold was quickly released from the mold and, immediately after the molten metal had solidified, was immersed in oil. This was followed by the usual triple tempering at 550°. The results of comparative tests of such cutters on 40 Kh steel are given in Table 15.

Table 15

Cutting speed, m/min	Life: Cutters cooled in air	Life: Cutters hardened in oil
48.5	10.0	10.9
48.5	10.8	12.0
48.5	10.9	10.0
Average life	10.7	10.9

Table 15 shows that hardening in oil did not produce any appreciable increase in life. Therefore all subsequent experiments were carried out by us without special hardening of the hardfaced cutters.

Next, in Figure 48, we present the results of testing hardfaced cutters on hard steel with \(H_b = 240\). The cutters were hardfaced in a graphite mold with borax as a flux and with deoxidation by aluminum. Before tempering these cutters, they were tested for hardness with a Rockwell durometer (diamond cone under a load of 150 kg); then the untempered cutters were subjected to a life test, after which the same cutters were subjected to the usual triple tempering at 550° and again, after tempering, subjected to a life test.

Table 16 gives the results of testing these cutters immediately after hardfacing (before tempering) and after triple tempering at 550°.

Table 16 shows that cutters obtained by hardfacing in a graphite mold, immediately after hardfacing (before tempering), in most cases acquire very high hardness, which apparently may be explained by the increased carbon content entering the metal from

Table 16

Cutter No.	Hardness Rc: before tempering	Hardness Rc: after tempering
11	63–66	68–70
12	65–67	65–67
13	65–66	68–69
14	63–64	67–68
16	62–63	67–68
17	58–59	64–67
111	64–65
112	64–65

graphite molding, as well as by alloying the surfacing with boron, which can be recovered from borax flux under welding-focus conditions, as has already been mentioned. Tempering further increases the hardness, and the more significantly the lower the hardness was before tempering, i.e., the more austenite was retained after cooling (hardening) in the structure of the high-speed steel.

Fig. 48

Fig. 49

Figure 48 gives very interesting results of durability tests of these cutters. Despite the fact that the hardness before tempering is very high, for most of the surfaced cutters it lies within the range 63–65, i.e., it is not inferior to the hardness of finally heat-treated cutters with inserts of RF-1 steel (hardened and tempered), the cutting qualities of surfaced cutters that have not undergone tempering are considerably lower than those of cu—

the best qualities after tempering. Figure 47 also shows the results of testing cutters with RF-1 inserts from a batch specially made for these comparative tests. Figure 48 shows that hardfaced cutters subjected to tempering are in no way inferior to normal high-speed cutters. In the same case, when tempering is absent, durability falls by approximately 100 times, and the speed must be reduced by 25% in order to maintain constant durability. Earlier we showed that hardfaced cutters do not require special hardening and, even if it is carried out, this changes nothing in their cutting qualities. Figure 48 indicates the extraordinary importance of tempering for hardfaced cutters, as also for ordinary high-speed steels.

Besides the normal threefold tempering at a temperature of 550°, we tried on these same cutters fourfold and fivefold tempering according to the following schemes: fourfold—three temperings for 1 hour at 550° and one hour at 575°; fivefold—three temperings for one hour at 550° and two for one hour each at 575°. The results of testing these cutters are given in Fig. 49, from which it follows that these tempering schemes give an additional appreciable increase in durability, especially fourfold tempering. We have, of course, by no means exhausted with this test all possible tempering schemes, but this experiment indicates that raising the tempering temperature to 575° does not impair the cutting qualities of the hardfaced tool. Thereby broader limits are established for the tempering temperature, which is convenient for production.

Fig. 50

All the test results presented above related to cutters hardfaced in a graphite mold with alternating current. It was of undoubted interest to compare these cutters with cutters hardfaced with direct current. This comparison was made in Fig. 50. The hardfacing and tempering conditions were identical for both sets of cutters; the test was carried out on steel 40Kh. It should only be noted that, whereas the cutters hardfaced with alternating current were hardfaced in the metallurgy laboratory of the Tomsk Polytechnic Institute, the cutters hardfaced with direct current were taken for testing from one of the plants where the method of hardfacing cutters with a carbon electrode had been introduced and mastered. There these cutters were hardfaced by the usual process developed by us in a graphite mold, with borax as the flux, with deoxidation by aluminum, and underwent normal threefold tempering at 550°. The cutters submitted for testing were not specially prepared for this purpose, but were taken from a large batch made for ordinary production work. Figure 50 shows that cutters hardfaced with direct current exhibited high cutting qualities, even somewhat exceeding the cutting qualities of cutters hardfaced with alternating current. In practice it may be considered that the type of current has no effect on the cutting qualities of the cutters.

3. Testing cutters hardfaced with RF-1 steel in a fireclay mold

Hardfacing in finely crushed fireclay with the addition of refractory clay may be carried out both with borax as a flux and with a mixture of marble and fluorspar. The second type of flux gives a more stable

arc, and in the case of alternating current the hardfacing process is carried out more easily. Aluminum is also added as a deoxidizer in order to obtain a dense deposited layer.

Of primary interest is a comparison of the cutting qualities of cutters hardfaced in a graphite mold and in a fireclay mold. Cutters hardfaced in a fireclay mold were not subjected to quenching; they underwent the normal threefold tempering at \(550^\circ\) and had the hardness indicated in Table 17.

Table 17

Cutter No.	Hardness \(R_c\), before tempering	Hardness \(R_c\), after tempering
21	60—61	64
22	61—62	65
31	61	64—65
32	58—59	63—65
33	62—64	65—66
44	60—62	66—67

Table 17 shows that the hardness obtained when hardfacing in fireclay, both before tempering and after tempering, is lower than when hardfacing in a graphite mold.

Fig. 51

Testing of cutters hardfaced in fireclay was carried out on steel with hardness \(H_B = 240\) and is presented in Fig. 51. Here circles indicate the test results for cutters hardfaced with marble and fluorspar as flux; triangles indicate those hardfaced with borax. The type of flux does not alter the cutting qualities. The same figure also gives the test results for cutters with RF-1 inserts. The figure shows that there is no difference between the hardfaced cutters and cutters with RF-1 inserts.

The experiments presented in Fig. 51 were carried out on the very same blank on which the tests of cutters hardfaced in graphite molds had been conducted, the results of which were shown in Fig. 48. A comparison of these two figures shows that there is absolutely no difference between one set of cutters and the other with respect to their cutting qualities, despite the observed difference in hardness.

It was next necessary to investigate the influence, on cutting qualities, of the amount of aluminum introduced during hardfacing in fireclay. For this purpose cutters were made with aluminum introduced during hardfacing in amounts of 1%, 3%, 6%, and 10% of the weight of the deposited metal. The hardness of the cutters thus obtained is illustrated by Table 18.

Table 18 shows that the hardness after tempering decreases noticeably when a large amount of aluminum is added. These cutters were subjected to the normal threefold tempering at \(550^\circ\). The results of testing them for durability are presented in Fig. 52, from which it follows that

cutters with the addition of aluminum during hardfacing up to 3% coincide exactly in their cutting qualities with cutters with RF-1 inserts. Increasing the aluminum addition to 6% already very sharply reduces the cutting qualities, bringing them close to those of carbon steel; a further increase in the aluminum addition worsens them still more. Therefore, also when hardfacing in chamotte, the optimum amount of aluminum added should be considered

Table 18

Cutter No.	Aluminum addition, %	Hardness Rc: Before tempering	Hardness Rc: After tempering
42	1.0	56–58	66–67
43	3.0	60–62	66–67
45	6.0	60–62	63–65
47	10.0	60	60–62

Fig. 52

Fig. 53

about 3%. This amount of aluminum, on the one hand, gives a dense deposit without pores, and on the other hand does not reduce the cutting qualities of the cutters.

Above we saw (Fig. 49) that, when hardfacing in a graphite mold, additional tempering at 575° somewhat increases the durability of the cutters. Investigation of this question as applied to cutters hardfaced in chamotte gave somewhat different results. Besides the basic one, there was also tested

two tempering schedules: after triple tempering at 550°, one additional tempering at 575° and double tempering at 600° plus one tempering at 550°. The results of testing such cutters are given in Fig. 53, which shows that practically all three schedules give cutters with identical cutting qualities. The second schedule gives an insignificant reduction in the cutting qualities of the cutters.

All the hardfacings described above in fireclay were carried out with alternating current, since it had previously been established that the type of current has no effect on the quality of the cutters.

4. Testing cutters alloyed during hardfacing with nickel and cobalt

The process of alloying the deposited layer during hardfacing with a carbon electrode is extremely simple. If, when hardfacing with a metallic electrode, a special coating of the electrodes must be prepared for alloying, then, when hardfacing with a carbon electrode, the alloying element is simply introduced in comminuted form into the metal being deposited.

In view of this simplicity, the introduction of alloying elements can be used in hardfacing with a carbon electrode not only for the purpose of directly obtaining alloyed deposits on a cutting tool, but also for the purpose of selecting and investigating special compositions of cutting alloys, accompanying the hardfacing in this case by mandatory chemical and metallographic analyses, hardness testing, determination of cutting qualities, etc.

We carried out several experiments on alloying steel during hardfacing of cutters with nickel, molybdenum, and cobalt; of course, these experiments do not yet claim any exhaustive completeness or systematic character. This question should be the subject of special work. We tested cutters with additions of 1% and 2% nickel and 5%, 10%, and 15% cobalt. The base metal for hardfacing was RF-1 steel. The hardfacing was carried out in a fireclay mold with a flux made from a mixture of marble and fluorspar; deoxidation was performed by adding aluminum in an amount of about 3% of the weight of the metal being deposited. After hardfacing, no special quenching was carried out; the hardfaced cutters cooled in still air. After hardfacing, triple tempering at 570° for one hour each followed. The hardfacing was carried out with alternating current.

The results of hardness measurement are given in Table 19.

Table 19

No.	Composition of alloying element	Amount of alloying element, in %% of the weight of the metal being deposited	Hardness Rc before tempering	Hardness Rc after tempering
1	Nickel	1.0	59–60	65–66
2	Nickel	2.0	60	63–65
3	Cobalt	5.0	61–63	63–67
4	Cobalt	10.0	64–66	65–66.5
5	Cobalt	15.0	62–64	64–66

The results of testing the cutting qualities of cutters with alloyed deposits are illustrated by Figs. 54 and 55.

It follows from Fig. 54 that the addition of nickel in amounts of 1 and 2% does not change the cutting qualities of the cutters at all, and their endurance characteristic coincides exactly with cutters with RF-1 inserts, and therefore also with cutters hardfaced with RF-1 steel.

Fig. 55 shows the effect of alloying with cobalt on the cutting properties of hardfaced cutters. It follows from the figure that an increased cobalt content considerably increases the durability of hardfaced cutters and substantially raises the permissible cutting speeds.

Fig. 54

In the figure: steel \(H_B = 250\); insert RF-1; hardfacing: \(1\%\ N\), \(2\%\ N\). Axes: \(T\), min; \(V\), m/min.

Fig. 55

In the figure: steel \(H_B = 250\); insert RF-1; hardfacing \(5\%\ Co\), hardfacing \(10\%\ Co\), hardfacing \(15\%\ Co\). Axes: \(T\), min; \(V\), m/min.

Fig. 56

The difference in the effect of additions of \(5\%\) and \(10\%\) cobalt is hardly noticeable, whereas an addition of \(15\%\) cobalt at once gives a noticeable improvement in cutting qualities in comparison with \(5\%\) and \(10\%\).

In Figure 56, on the basis of Figure 55, a graph is given illustrating the increase in tool life and cutting speed for hardfaced cutters alloyed with cobalt, in comparison with cutters with RF-1 inserts.

It is evident from the figure that the addition of cobalt in amounts of 5%, 10%, and 15% increases tool life, in comparison with RF-1 inserts, by 2.5 times, 3.5 times, and 14 times respectively, and the cutting speed by 5%, 9%, and 15.5%. This approximately corresponds to the data of Gudremont and Schrader concerning the influence of cobalt on the productivity of high-speed steel. [28] It may be expected that, with heavy roughing chips, the influence of cobalt would be still more substantial.

5. Testing of Cutters Hardfaced with Low-Alloy High-Speed Steels

It was of considerable interest to test cutters hardfaced with low-alloy substitutes for high-speed steel. We tested the most widely used low-alloy steels EI-184 and EI-262 for hardfacing cutters.

The hardfacing of cutters with these steels was carried out in a fireclay mold with marble and fluorspar as flux and with the addition of aluminum for deoxidation. The hardfacing was performed with alternating current. After hardfacing, triple tempering followed at a temperature of 560°.

For hardfacing, steels EI-184 and EI-262 were used both in pure form and in the form of mixtures with steel RF-1. The hardfacing compositions and the results of hardness measurements of the hardfaced cutters are given in Table 20.

Table 20

No.	Hardfacing composition	Hardness Rc before tempering	Hardness Rc after tempering
1	EI-184	52–54	62–63
2	75% EI-184 + 25% RF-1	57–59	63.5–65
3	50% EI-184 + 50% RF-1	61–62	65–66
4	EI-262	56–57	62–64
5	75% EI-262 + 25% RF-1	61–62	64–65

Cutters hardfaced both with pure low-alloy steels and with mixtures with RF-1 were tested on blanks of steel with HB 240. Tests of these cutters showed that cutters hardfaced with steel EI-184 in pure form have cutting properties reduced in comparison with RF-1.

A different effect is obtained if steel RF-1 is added to EI-184 during hardfacing. Figure 57 gives the results of testing cutters hardfaced with a mixture of 50% EI-184 and 50% RF-1, and Figure 58 gives the results of testing cutters hardfaced with a mixture of 75% EI-184 and 25% RF-1. It follows from the figures that even the addition of only 25% RF-1 gives cutters approaching cutters made of RF-1, while the addition of 50% RF-1 gives hardfaced cutters in no way inferior to cutters made of RF-1.

This circumstance is of very great importance, since mixing steels during hardfacing with a carbon electrode presents no difficulties and makes it possible to use the substitute EI-184 to obtain high-quality cutters. If it is taken into account that cutters with inserts of EI-184, with their best heat treatment, give 85–90% productivity in comparison with cutters with RF-1 inserts, it may be concluded that hardfaced cutters made from a mixture of 75% EI-184 and 25% RF-1 completely replace cutters with EI-184 inserts, and with addi—

with the addition of 50% RF-1, hardfaced cutters are obtained with cutting qualities superior to them.

Figure 58 also gives the test results for cutters hardfaced with EI-262 steel and with a mixture of 75% EI-262 and 25% RF-1. The figure shows that there is no difference in the cutting qualities of these cutters. In comparison with cutters having RF-1 inserts, however, these hardfaced cutters show a certain slight reduction. This reduction in cutting speed may be estimated at approximately 5–7%, i.e., it will be approximately the same as for cutters with EI-262 steel inserts when machining hard steels, which is what we had in the present case.

Fig. 57

Fig. 58

Proceeding from this, it may be considered that hardfacing with EI-262 steel gives quite high-quality cutters, fully replacing cutters with inserts made of this steel.

It is important to note here that, in order to obtain high cutting qualities, low-alloy high-speed steels require especially careful heat treatment. The limits of the hardening temperatures for them are extremely narrow (usually ±10°), and this creates considerable difficulties in their heat treatment in production; deviation from the optimum hardening temperature has a significant effect on the deterioration of the cutting qualities of a tool made of low-alloy high-speed steels. Our experiments on hardfacing these steels show that cutters hardfaced with these steels possess high cutting qualities, without requiring the special hardening operation, which substantially simplifies the process of their manufacture in production. This circumstance shows that, in cases where production does not have heat-treatment equipment by means of which the narrow limits of hardening could be reliably maintained, using

high temperatures in the heat treatment of low-alloy high-speed steels, it is expedient to make hardfaced cutters from bar steel EI-184 and EI-262 (and not only from their scrap).

6. Nature of the dulling of hardfaced cutters and preservation of cutting properties during regrinding

The nature of the gradual wear and dulling of hardfaced cutters in the cutting process essentially does not differ at all from the dulling of cutters with RF-1 inserts. A wear crater is formed on the rake face; it gradually deepens and widens and gradually reaches the cutting edge. From approximately this moment, noticeable wear begins to appear on the flank face. With the chip dimensions that occurred in our tests, wear on the flank face is localized at the nose of the cutter and is observed both on the principal and on the auxiliary flank faces. Wear on the flank faces increases rapidly from the moment the crater reaches the cutting edge, and by the moment of final destruction of the cutting edge it usually reaches a value approximately equal to 1 mm (the width of the land on the flank face). In our tests, destruction of the cutting edge usually occurred at the nose of the cutter.

Of extraordinary importance is the constancy of the cutting properties of hardfaced cutters as the hardfaced layer is ground away. During regrinding we removed substantial layers in order to remove completely the metal damaged when the cutter became dull; moreover, we observed that, up to the moment of almost complete removal of all the hardfaced metal, the cutting properties of the cutter do not change at all. This is apparently explained, on the one hand, by the complete homogeneity of the hardfaced layer and, on the other hand, by the circumstance that, with good bonding between the hardfaced metal and the cutter shank, no appreciable mixing of the metals of the hardfacing and the shank is observed; this is expressed by the sharp boundary between the one and the other metal on the polished section (see Fig. 28).

Fig. 59

We measured the hardness on the rake face of a hardfaced cutter during continuous grinding away of the hardfaced layer. For this purpose a cutter with a considerable thickness of hardfaced layer was specially selected, since in this case any nonuniformity of the layer through its thickness should have appeared most distinctly. The results of the hardness measurement are presented in Fig. 59. The figure shows that, within a layer thickness of 7–8 mm, hardness fluctuations are observed within the limits of 1–1½ Rockwell units, i.e., within the accuracy and sensitivity of the instrument itself, while in the last half-millimeter the hardness decreases abruptly from 63–64 to 42 units. Here there is already a transition layer of metal, whose thickness is thus determined as only fractions of a millimeter. If one takes into account that, for the first hardness measurement, a layer at least 0.5 mm thick was removed from the hardfaced cutter, then the total thickness of the hardfaced layer on this cutter was 9–9.5 mm. Such a thickness was obtained by hardfacing several portions of metal. Nevertheless, the hardness measurement shows exceptional sta-

... its stability over the thickness. This is in complete agreement also with observations on the preservation of cutting properties when cutters are reground.

7. General results of laboratory tests of hardfaced cutters

Briefly summarizing the results of the laboratory tests of hardfaced cutters, it may be considered established that:

a) the method of hardfacing with a carbon electrode gives cutters of high quality;

b) the kind of current used in hardfacing (direct current or alternating current) has no effect on the quality of the hardfaced cutters;

c) the material of the mold has no effect on the quality of the hardfaced cutters;

d) the composition of the flux used in hardfacing (borax or marble with fluorspar) has no effect on the quality of the hardfaced cutters;

e) the optimum amount of aluminum introduced during hardfacing, which reliably eliminates porosity and cavities in the hardfaced layer and does not reduce the cutting properties of the hardfaced cutters, is about 3% of the weight of the metal being deposited. The introduction of silicon as a deoxidizer gives cutters of lower quality, both from the standpoint of the density of the hardfaced layer and from the standpoint of their cutting properties;

f) special hardening after hardfacing does not increase durability and from this standpoint is not necessary;

g) tempering of cutters, especially repeated tempering, after hardfacing very greatly increases their cutting properties and must be regarded as absolutely necessary;

h) the cutting properties of cutters hardfaced with RF-1 steel and of cutters with RF-1 steel plates are exactly the same;

i) alloying hardfaced cutters with cobalt considerably increases their cutting properties;

j) hardfacing cutters with the low-alloy steel EI-184 with an addition of from 25 to 50% RF-1 steel gives cutters with high cutting properties;

k) hardfacing cutters with the low-alloy steel EI-262 gives cutters of high quality, not inferior to cutters made of EI-262 steel with optimum heat treatment;

l) the hardfaced layer in cutters possesses high homogeneity and, when ground down, does not cause a reduction in hardness or cutting properties.

In summing up the results of the laboratory tests of hardfaced cutters, it is appropriate here to make the following additional remarks.

Prof. Minkevich, in his book Low-Alloy High-Speed Steels (Metallurgizdat, 1944), is extremely skeptical about the possibility of obtaining a high-quality cutting tool by hardfacing. In assessing various methods of reinforcing cutters (casting plates with liquid cast iron, welding-on plates, mechanical fastening, etc.), as compared with hardfacing with high-speed steel, he writes:

“From a different point of view, the question of the use of high-speed-steel hardfacing should be considered. This method leads to the creation of a cast structure, more brittle and less satisfactory in cutting properties, and to a certain change—difficult to take into account in each case—in the chemical composition of the high-speed steel, and consequently also to nonuniformity of the cutting properties in the hardfaced tool, which is the principal defect of hardfacing. Therefore, the hardfacing method cannot be regarded as a more advanced method of reinforcing a cutting tool, intended to replace other methods within sufficiently broad limits.”

This categorical conclusion is entirely correct with respect to hardfacing with Kontorov’s alloys [29], which, as shown by ...

special investigations [30], give a nonuniform hardness of the hardfaced layer, varying at individual points within the range from 60 to 64 Rockwell units on the C scale, a heterogeneous structure, and, despite their higher alloying in comparison with high-speed steel RF-1, are inferior to it in cutting qualities.

The validity of this conclusion with respect to a layer of high-speed steel hardfaced by an acetylene–oxygen flame or by an electrode made of high-speed steel cannot be verified, since, unfortunately, there are no systematic studies of tools hardfaced by these methods. As for the method of hardfacing with a carbon electrode, which we have studied quite fully, Prof. Minkevich’s conclusion is clearly not justified here, since our investigations, carried out with several hundred cutters, showed their complete uniformity both with respect to cutting qualities and with respect to the hardness distribution through the thickness of the hardfaced layer. In addition, it was shown quite clearly that their cutting qualities are in no way inferior, and in the case of additional alloying with cobalt are considerably superior, to the cutting qualities of steel RF-1.

II. LABORATORY TESTS OF PRODUCTION CUTTERS

Laboratory tests of production cutters were carried out in order to compare the cutting qualities of ordinary production cutters with RF-1 steel tips with the cutting qualities of hardfaced cutters. These comparative tests were of interest for the following reasons.

The hardfaced cutters did not require the special quenching operation, receiving it automatically as the hardfaced layer cooled from the molten state. Receiving only tempering, the hardfaced cutters, as all our previously described investigations showed, were in no way inferior in cutting properties to cutters with RF-1 high-speed-steel tips that had undergone good heat treatment.

In production, the quenching operation for high-speed steel encounters considerable difficulties, requiring, for proper execution, the mandatory availability of thermal equipment with precise regulation and temperature control.

In the absence of this equipment, quenching of high-speed steel in production exists, at best, as an operation of cooling the cutter after the operation of brazing the tip to the holder. This is usually associated with a lowering of the quenching temperature relative to the optimum, since the melting temperature of the brazing powder is always below the normal quenching temperature. Such quenching entails a substantial reduction in cutting qualities. In the case where, after the tip is brazed to the holder, the cutter is not quenched but undergoes an operation of slow cooling and annealing, and only then is specially heated for quenching, the absence of temperature control can also lead to overheating, since the brazing powder, upon solidifying after the tip has been brazed, forms a solid solution with the material of the holder and the tip and, during subsequent heating, already has a considerably higher melting temperature. Overheating during quenching has an even more harmful effect on the cutting qualities of the tool. In many cases in production, tempering of the high-speed tool is not carried out, which also substantially worsens its cutting qualities.

Hardfacing of high-speed steel eliminates the need to carry out quenching and thereby eliminates (in the case where the tool—

does not require annealing for carrying out mechanical machining) one of the operations whose improper execution has the most harmful effect on the quality of a cutting tool. Laboratory tests were to show to what extent hardfaced cutters, which do not require a special hardening operation, surpass ordinary factory-made cutters in durability, since the quality of the latter inevitably reflects all the irregularities of heat treatment. In addition, these tests were to show how greatly cutters from different plants, having different standards of heat treatment of tools, differ among themselves.

For the tests, from the tool-distribution storerooms of four large plants, through-cutting cutters with welded-on plates of high-speed steel RF-1 were taken. The cutters were taken without any selection. In the laboratory, before the tests, they underwent sharpening on a Gisholt machine, by which they were given the same geometry and quality of sharpening as the hardfaced cutters had. After this they were tested on steel 45 with hardness \(H_b = 200\), on the same blanks on which the cutters hardfaced with RF-1 steel in a graphite mold had been tested, the test results for which were presented by us in Fig. 45.

The results of testing factory-made cutters obtained from four plants are illustrated in Fig. 60. In this figure, the straight line for the hardfaced cutters (hardfacing with RF-1 is taken from Fig. 45) and three more straight lines through the experimental points of the tests of factory-made cutters are drawn. First of all, the extremely different quality of the cutters obtained from different plants attracts attention. Still more significant is the circumstance that the cutters of all plants are very markedly inferior, in their cutting qualities, to the hardfaced cutters. For the hardfaced cutters the cutting speed for a 60-minute life is 36.5 meters per minute; for the factory cutters it is 31, 26.5, and 25 meters per minute, i.e., the factory cutters give productivity respectively 15%, 27.5%, and 31.5% lower than the hardfaced cutters. This is undoubtedly the result of their unsatisfactory heat treatment. If the lives of factory cutters are compared with hardfaced cutters at a constant cutting speed, the figures obtained are still more expressive. The life of the hardfaced cutters relates to the life of the factory cutters as

\[ \frac{8}{1},\quad \frac{30}{1},\quad \text{and}\quad \frac{240}{1}. \]

If one takes into account that the ratio of lives is proportional to the ratio of expenditures on the tool, then these figures acquire substantial economic significance.

Fig. 60

The introduction at plants of hardfaced cutters made from high-speed-steel waste, with their simpler manufacturing technology and higher cutting qualities, should yield a considerable economic effect.

III. Production Tests of Hardfaced Cutters

The production tests of hardfaced cutters were intended to test them in normal factory operations, under all the conditions that usually accompany the work of a cutting tool in production, such as: the presence of scale on the workpiece, the not always uniform hardness of the material being machined, eccentric rotation of the workpiece during roughing operations and the associated variation in depth of cut, possible vibrations, operation with shocks and impacts, etc. In addition, our aim was to compare hardfaced cutters with factory-made cutters under production conditions, in order to demonstrate, not by laboratory experiments but by work on normal factory operations, the advantages of hardfaced cutters over factory-made ones.

The production tests of hardfaced cutters were carried out at four different plants. For these tests the cutters were hardfaced with RF-1 steel by means of a carbon electrode in a graphite mold, with borax (flux) and aluminum (deoxidizer). After hardfacing, three temperings at 550° for 1 hour each followed.

In the tests of the hardfaced cutters nothing was changed in the production operation on which the test was carried out. The depth of cut and the feed remained unchanged; no measures were taken to eliminate or reduce the eccentricity of rotation of the part, to clamp it more rigidly, to prevent possible vibrations, etc.; that is, all the usual working conditions of the tool remained unchanged. The only thing that we usually had to change was to increase the cutting speed. This was done in order not to be tied to an excessively long period of operation of the cutters before dulling. The factory cutters that were compared with the hardfaced ones were necessarily given the same geometry of the cutting elements as the hardfaced cutters had, in order to create completely identical test conditions.

The testing of hardfaced cutters in production was carried out mainly on roughing operations, since it is precisely in these operations that the difference between laboratory and production test conditions is revealed most sharply. The tests were carried out on steels, cast iron, and cast brass, with and without cooling. In all cases without exception, the hardfaced cutters showed considerably higher cutting performance than factory-made cutters. In some cases, the hardfaced cutters successfully operated in those operations in which, in production, cutters tipped with pobedit were used (which, of course, indicates an irrational use of pobedit). It is important that in this case the hardfaced cutters could operate at the same cutting conditions at which pobedit operated, whereas factory-made cutters with high-speed-steel tips could not do this.

Below we give excerpts from a number of reports on production tests of hardfaced cutters.

1. Cutters with a cross-section size of 10 × 12 mm, hardfaced with high-speed-steel chips, right-hand turning cutters, were tested in turning part 2480 made of steel 45.

Factory cutters with a welded-on high-speed-steel tip operated at a spindle speed of 410. For the test, the speed was increased to 660.

The factory cutter worked for 1 hour before dulling.

The hardfaced cutters, operating for three shifts, required resharpening once per shift.

The endurance of the hardfaced cutters is 10 times higher than that of factory cutters with high-speed-steel plates.

1. Hardfaced through and boring cutters of round cross section, 15 mm in diameter (factory standards I-15, I-23, I-90), were tested on “Pittler” turret lathes in the following operations:

a) Turning spindles made of steel 3.
b) Turning brass castings on the crust.

In the indicated operations the plant uses cutters with “Pobedit” plates. The tests were conducted without changing the factory cutting conditions. The tests gave the following results:

In turning spindles, cutters hardfaced with high-speed-steel waste have an endurance higher than that of cutters with “Pobedit” plates.

In machining brass on the crust, the hardfaced cutters showed normal endurance and can replace cutters with “Pobedit” plates.

1. Hardfaced cutters, with a cross-sectional size of \(16 \times 24\) mm, and factory cutters of the same size with high-speed-steel plates were tested in roughing the “AD” shaft made of steel with hardness \(H_B = 180\). The work was performed without cooling. In the shop, this operation is performed with cutters with “Pobedit” plates at 100 rpm. For the test, the number of revolutions was increased to 200. In this case the cutting speed was 34.6 m/min. Depth of cut 4–8 mm. Feed 0.3 mm/rev.

The factory cutter with a high-speed-steel plate showed an endurance of 30 minutes in the test. The hardfaced cutter—67 minutes. The endurance ratio was

\[ \frac{2.23}{1}. \]

1. Cutters hardfaced with high-speed steel RF-1, and factory cutters with RF-1 high-speed-steel plates, were tested in a roughing pass on the part “jackhammer barrel” (steel 50), on the crust, under the following cutting conditions:

\[ \begin{aligned} \text{cutting speed} \ . . . . . . . . &\ 44.5\ \text{m/min};\\ \text{depth of cut} \ . . . . . . . . &\ 1\text{–}2\ \text{mm};\\ \text{feed} \ . . . . . . . . . . . . &\ 0.65\ \text{mm/rev.} \end{aligned} \]

Work without cooling.

In the shop, this operation is usually performed with cutters with “Pobedit” plates.

In the test the cutters showed the following endurance:

\[ \begin{aligned} \text{Factory cutter} \ . . . . . . . &\ 4.3\ \text{minutes};\\ \text{Hardfaced cutters} \ . . . . . &\ 19.5\text{–}43\ \text{minutes}; \end{aligned} \]

Endurance ratio . . .

\[ \frac{4.55}{1} — \frac{10}{1}. \]

1. Cutters hardfaced with RF-1 high-speed steel, with a cross-sectional size of \(12 \times 18\) mm, and factory cutters with RF-1 steel plates, size \(16 \times 24\) mm, were tested on a turret lathe, type 136 of the Ordzhonikidze plant, in the operation of roughing on the crust the part “intermediate link,” made of steel 45.

Test conditions:

\[ \begin{aligned} \text{cutting speed} \ . . . . . . . &\ 56.5\ \text{m/min.}\\ \text{depth of cut} \ . . . . . . . &\ 2.0\ \text{mm.}\\ \text{feed} \ . . . . . . . . . . . &\ 0.22\ \text{mm/rev.} \end{aligned} \]

Work with cooling.

Endurance of factory cutters from 13.3 to 23 minutes.
Endurance of hardfaced cutters from 69 to 82.5 minutes.

If one takes into account the difference in the dimensions of the cutters and the influence of cutter size on its durability (Chart 71 of the standards for cutting conditions in turning, issued by the People’s Commissariat of Heavy Machine Building of the USSR, 1941), then the ratio of durabilities should be increased by a factor of 1.5. In this case the ratio of the durabilities of the hardfaced cutters and the factory-made cutters will be within the limits from

\[ \frac{4.35}{1} \text{ to } \frac{9.3}{1}. \]

1. Cutters hardfaced with high-speed steel and factory-made cutters with high-speed-steel plates of identical dimensions (\(16 \times 24\) mm) were subjected to testing in the operation of rough turning a seamless steel pipe over the scale.

Test conditions:

cutting speed . . . . . . . . . . \(41.5\) m/min.
depth of cut . . . . . . . . . . \(2.5\) mm.
feed . . . . . . . . . . . . . . \(0.3\) mm/rev.
Work with cooling.

The work in the shop on this operation is usually performed with cutters with “Pobedit” plates at the indicated cutting condition.

Test results:
Durability of the factory-made cutter—0.5 min. (almost instantaneous blunting).
Durability of the hardfaced cutters . . . . . . . . . . 36–45 min.

Ratio of durabilities . . . . .

\[ \frac{72}{1} - \frac{90}{1}. \]

1. Cutters hardfaced with high-speed steel and factory-made cutters with high-speed-steel plates of identical dimensions (\(16 \times 24\) mm) were subjected to testing in face turning of a cast-iron “cover” part.

Operation—rough turning over the scale without cooling.

Test conditions:

number of revolutions . . . . . . . . . . . . \(120\);
depth of cut . . . . . . . . . . . . . . . \(3\) mm;
feed . . . . . . . . . . . . . . . . . . \(0.15\) mm/rev.

The work in the shop on this operation with the indicated cutting condition is usually performed with cutters with “Pobedit” plates.

Test results:
Durability of the factory-made cutter—0.5 min. (almost instantaneous blunting).
Durability of the hardfaced cutter 43 min.

Ratio of durabilities

\[ \frac{86}{1}. \]

In addition to these tests, documented by reports, a considerable number of factory tests were also carried out, showing essentially the same positive results for the hardfaced cutters.

Comparison of the ratios of the durabilities of the hardfaced and factory-made cutters obtained in production tests and in laboratory tests shows that these figures are of the same order.

The low durabilities of the factory-made cutters are the result of their poor heat treatment. The following circumstance convinces us of this. One of the comparative tests was carried out at the Tomsk Instrument Plant. This test was performed in the operation of roughing blanks for drills made of high-speed steel. In this case the factory-made cutters had high cutting qualities, having been heat-treated under conditions—

conditions of a well-organized heat-treatment shop of a tool plant. The test showed that in this case the lives of the hardfaced and factory-made cutters were the same. This is in full agreement with the results of laboratory tests, which showed that the hardfaced cutters have the same cutting properties as cutters with R18 plates that have been properly heat-treated.

IV. GENERAL RESULTS OF TESTS OF HARDFACED CUTTERS AND SOME CONSIDERATIONS ON THE EFFECTIVENESS OF INTRODUCING THEM INTO PRODUCTION

As the material set forth in the present chapter shows, the tests of hardfaced cutters were carried out with great completeness and thoroughness. Hundreds of hardfaced cutters passed through laboratory and factory tests, and in all cases without exception they showed high cutting qualities, not inferior to cutters made with high-speed-steel plates that had undergone careful heat treatment. In those cases where cutters with high-speed-steel plates did not receive normal heat treatment, the hardfaced cutters showed many times greater durability. The hardfaced cutters were tested in machining various steels, cast iron, cast brass, and in roughing and finishing operations.

The extensive tests carried out on hardfaced cutters confirmed the correctness of the technology developed by the authors for hardfacing high-speed steels with a carbon electrode. It may be considered proven that highly productive cutters can be reliably obtained by hardfacing both high-speed steel of the R18 type and its low-alloy substitutes EI-184 and EI-262.

The introduction of cutter hardfacing into production may be justified by the following considerations.

Performing hardfacing with a carbon electrode is a simple operation that does not require scarce materials, any special equipment, or complicated preparatory work. It requires neither gas-welding equipment nor scarce calcium carbide and oxygen, as in hardfacing with an oxyacetylene flame, nor the special manufacture of high-speed electrodes with a complex special coating, as in hardfacing by the Slavyanov method with a metallic electrode made of high-speed steel.

The method of hardfacing with a carbon electrode makes it possible to use, for hardfacing, waste high-speed steel and its substitutes in the form of chips, which are always available in tool shops, and scrap remaining from used tools. In cases where production does not possess reliable heat-treatment equipment for carefully hardening cutters with high-speed-steel plates, it appears advisable to make only hardfaced cutters and to use for this purpose not only waste, but also new high-speed steel and its substitutes. The advisability of this is dictated by the fact that hardfaced cutters, while requiring no hardening, possess cutting qualities considerably superior to the cutting qualities of not very carefully hardened cutters with high-speed-steel plates. This is especially important as applied to low-alloy high-speed steels, which have narrow ranges of hardening temperatures and therefore require especially careful heat treatment. Our laboratory and production tests showed that hardfaced cutters, requiring no special hardening, in re-

…in terms of cutting qualities they fully correspond to cutters with plates subjected to careful heat treatment, and in many cases they surpass in durability factory-made cutters that have undergone unsatisfactory heat treatment. This applies to an even greater extent to low-alloy high-speed steels.

The technology for manufacturing cutters by hardfacing is simpler than the technology for manufacturing cutters with welded-on plates.

The schemes of the normal manufacturing process for a cutter with a welded-on plate and for a hardfaced cutter are presented in Table 21.

Table 21

Cutters with a welded-on plate	Cutters with a welded-on plate	Hardfaced cutters	Hardfaced cutters
No. of operation	Name	No. of operation	Name
1	Forging the holder	1	Forging the holder
2	Milling the seat in the holder	2	Preparation of steel for hardfacing
3	Preparation of the plate	3	Preparation of molds and flux for hardfacing
4	Fitting the seat and plate	4	Hardfacing
5	Fastening the plate to the seat with wire	5	Tempering
6	Preparation of welding powder and flux	6	Grinding
7	Welding the plate to the holder	7	Sharpening
8	Annealing
9	Hardening
10	Tempering
11	Grinding
12	Sharpening

As the table shows, a cutter with a plate goes through 12 operations, while a hardfaced cutter goes through 7 operations. For the hardfaced cutter the following operations are eliminated: milling the seat in the holder, fitting the seat and plate, fastening the plate to the seat, annealing, hardening. Milling the seat is not required, since it can be produced for hardfacing when forging the holder and will go to hardfacing in the black condition, which is impermissible in the manufacture of a cutter with a plate, where not only a clean surface of the seat is required, but also one accurately fitted to the surface of the plate, since otherwise a strong joint between the plate and the holder will not be obtained. Milling the seat in the holder, preparing the plate, and fitting the seat and plate are operations that occupy machine tools, and their elimination when switching to hardfacing of cutters frees machines and the workers who service them. Still more significant is the elimination of the hardening operation, insufficiently careful execution of which, especially in the manufacture of tools from low-alloy high-speed steels, leads to a substantial loss of cutting qualities. The absence of this operation in the manufacturing technology of hardfaced cutters, on the one hand, considerably simplifies their manufacture, and on the other hand eliminates the possibility of producing rejects because of its unreliable execution.

Important issues are productivity in the manufacture of hardfaced cutters and the throughput capacity of the welding workplace engaged in hardfacing cutters. As experience at one of the plants that mastered the manufacture of hardfaced cutters shows, one welder with an assistant, with normal organization of the workplace and with…

with normal provision of all the necessary materials prepared in advance (molding materials, flux, steel for hardfacing, etc.), produces 15–20 hardfaced cutters per hour. If one assumes 20 hours of work per day, then one welding station will thus produce 300–400 hardfaced cutters per day.

The economic effectiveness of introducing hardfaced cutters into production is determined not only by the simpler and cheaper manufacture of them and by the possibility of using, in their manufacture, waste from deficient high-speed steel, but also by their generally higher productivity, since they have higher cutting qualities that do not depend on the thoroughness with which the complex hardening operation is carried out. The director of one of the plants that have mastered the manufacture of hardfaced cutters writes on this subject: “It is important to note that cutters began to be made from waste high-speed steel by the method developed by a group of research workers of the Polytechnic Institute. The great durability of these cutters has considerably increased labor productivity in all machining operations” [31].

GENERAL CONCLUSIONS AND SUMMARY

1. In this work, for the first time, the full possibility has been demonstrated of hardfacing high-speed steel of various grades by means of a carbon voltaic arc. It is possible to use both the classical Benardos method—the direct-current carbon arc—and alternating current from any welding transformer. In both cases, hardfacings of equally high quality are obtained, and the simplicity of the procedures is ensured, allowing even a welder with little experience to work.

2. The proposed method makes it possible to utilize as filler material both rods of high-speed steel and all kinds of its waste: machine turnings, remnants of worn-out tools, and various scrap.

3. By the authors’ method it is possible to hardface high-quality cutters from steels R, RF-1, EI-262, and mixtures of steels EI-184 and RF-1. A tool hardfaced with a mixture of steel EI-184 (50% by weight) and RF-1 (50% by weight) is not inferior in durability to a solid tool made of steel RF-1.

4. For hardfacing, holders made of ordinary tool steel are used. The recess for hardfacing is made by hot stamping, which considerably reduces the cost of preparing the holders. The technological process for manufacturing hardfaced cutters is simpler and cheaper than the process of manufacturing cutters with plates of high-speed steels.

5. A method has been developed for obtaining bubble-free, calmly solidifying hardfacings by deoxidizing them with aluminum. Simple methods are given for introducing deoxidizing aluminum into the bath, both through metal shavings and through the flux.

6. Various molds for making hardfacings have been investigated. Very simple and inexpensive molds have been developed in graphite plates and in a mixture of chamotte powder with clay.

7. In durability and cutting speed, hardfaced cutters are equivalent to cutters made of high-speed steel that have been properly heat-treated, and in most cases they considerably surpass ordinary factory cutters with welded-on plates of the corresponding high-speed steels.

8. The hardfacings are properly hardened during cooling of the manufactured cutters in air. For hardening, accelerated cooling in an oil bath or in an air jet is not required, and when cooled in still air without any protection, dangerous internal [[unclear: continuation cut off on this page]]

...of internal stresses and any cracks in the hardfacings. This considerably simplifies the manufacture of cutters by the proposed method.

The only thermal operation required for hardfaced cutters is triple tempering, with a hold of one hour at each tempering.

1. Under laboratory conditions, hundreds of cutters were tested on a special experimental lathe. A large number of speed–tool-life curves were obtained, confirming the high cutting qualities of the hardfaced cutters.

2. At plants, tens of thousands of hardfaced cutters were tested in production operations, and their manufacture under factory conditions was mastered. The simplicity of the technology for making them and their higher cutting qualities produced a significant economic effect in production.

3. Metallographic study of the hardfacings showed that, as a consequence of the special features of their crystallization, their structure, although it has some features of a cast structure, is distinguished by the absence of segregation, by fine grain, and by uniformity.

4. There are very broad possibilities for alloying the hardfacings by adding ferroalloys or pure metals to the charge. In the postwar period, alloying with cobalt, which makes it possible to increase the tool life of cutters by tens of times, is acquiring special significance. Alloying with molybdenum, nitrogen, and boron is also of importance.

5. In the postwar period, further expansion of the research promises new achievements of a scientific and economic order: the development of new grades of steel for hardfacing, broad mechanization of hardfacing processes and similar work, and the introduction into production of new methods of hardfacing cutters (under a flux layer, hardfacing by the Zerener method, etc.).

6. The work was carried out in the laboratories of metal science and metal cutting of the Tomsk Polytechnic Institute.

Cited Literature

1. Taylor F.—The Art of Cutting Metals, Berlin, 1922.
2. News of Foreign Metallurgy, 1937, No. 6, p. 71.
3. Minkevich N. A., Ivanov A. G., Ivanov O. S.—New Low-Alloy High-Speed Steels, Metallurgizdat, 1940; second edition, 1944.
4. Brokhin I. S. Alloy “Smena”, Metallurgizdat, 1933.
5. “Akrit.” Das Schneidmetall für die moderne Bearbeitungswerkstatt, prospectus, 1931.
6. Bulletin of the Metal Industry, 1940, No. 4–5, p. 38.
7. Proceedings of the Siberian Institute of Metals, vol. 1, issue I, 1931.
8. W. Oertel u. Grützner. Die Schnelldrehstähle. Stahleisen, 1931.
9. Iron Age, 1919, p. 1711.
10. Rapatz F. High-Quality Steel, GONTI, 1938.
11. Bulletin of Mechanical Engineering, 1943, No. 11–12.
12. I. Autogenous Welding, 1937, No. 2, Lapidus V. A.
    II. Autogenous Welding, 1937, No. 8, Moroz L. S.
    III. Autogenous Welding, 1939, No. 4, Shorin P. I.
    IV. Autogenous Welding, 1940, No. 7, Smirnov I. S.
13. I. Bulletin of the Metal Industry, 1937, No. 18, Smichkov M.;
    II. Autogenous Welding, 1938, No. 1, Lapidus V. A.;
    III. Autogenous Welding, 1939, No. 5, Ragazina M. F.;
    IV. Machines and Tools, 1937, No. 15, p. 4, Kuks;
    V. Autogenous Welding, 1941, No. 2, Chistoserdov;
    VI. Vagin V. A., High-Quality Tool from Chips and Scrap, ITEIN, 1942;
    VII. Stefanovskii, Use of High-Speed-Steel Waste and Pobedit for Hardfacing Tools, ITEIN, 1943.
14. Ginzburg Ya. S. Thermomechanical Treatment of High-Alloy Steel, ONTI, 1937.

Errata Noted

Page	Line	Printed	Should read
23	14 from top	at point a	at point 8
24	Table 2, column 2, first row	0	50
32	18 from top	elektroderzhatel	elektrododerzhatel
33	16 from bottom	modiritsirovanie	modifitsirovanie

Izvestiya TPI, vol. 61, issue 4