A. N. Dobrovidov and A. M. Rosenberg.

Detection of Plastic Deformations in Iron.

(The article contains 7 photographs and 2 drawings.)

I. Plastic deformations.

If a metallic specimen (for example, of soft iron) has undergone plastic deformations of any kind (tension, compression, torsion, etc.), then all its properties have changed to a greater or lesser degree¹—it is no longer the same specimen as before. The elastic limit and the ultimate strength of the specimen have increased, its plasticity has decreased, the specimen has become less dense, and its solubility in acids has increased.

But if we have at our disposal some iron specimen and we do not know its history exactly, then it is sometimes very difficult to determine why the metal is too strong and brittle: whether owing to the influence of previous mechanical treatment, or to the influence of some impurities. Of course, by applying special methods of investigation such a problem can be solved; but if there are nests in a component—that is, separate cold-worked regions—then obtaining an answer is extremely difficult, and for the practical investigation of iron that has been in service it is very important to determine, in good time and more or less simply, whether a given part of a structure has been subjected to stresses above the elastic limit and has acquired residual deformations as a result of its service. Ordinary macro- and microscopic investigation rarely provides an answer to such a question.

If one takes a well-polished piece of iron and subjects it to the action of a load exceeding the elastic limit and therefore causing residual deformations, the surface of the specimen becomes covered with ripples—Lüders or Hartmann lines. The nature of these lines may be explained as follows. A piece of metal, in particular the soft iron that interests us, is an aggregate, an assemblage of microscopic crystallites bound together by intercrystalline material. In deforming the specimen as a whole, we thereby change the arrangement of the individual crystallites relative to one another and the form of the individual crystallites of which the piece of metal is composed.²

Under the influence of an external force, the phenomenon, depending on the nature of the metal, may proceed in two ways: the crystallites first rotate and then are destroyed, or else changes first occur within the crystallites and then their rotation follows. An individual metallic crystal, under the influence of stresses exceeding the elastic limit, changes its form. In the crystal, slips occur along slip planes; depending on the crystal system, one or several directions of these planes may appear. In crystals of iron the slip planes coincide in direction with the faces of the rhombic dodecahedron,[^3] and in some cases, possibly, of the cube and the icositetrahedron.[^4] The rotation of the crystallites in a piece of metal and the slip within them—all this occurs in a limited volume and is accompanied by friction of the individual particles of the metal against one another. Surfaces rubbing under considerable pressure are polished, abraded, and at their boundaries

extremely small intercrystalline or, perhaps, amorphous fragments are formed.^4) If the deformations are large, a fairly considerable number of these fragments may be formed. In any case, in a piece of plastically deformed metal, grains of normal and very small size must be found side by side; furthermore, crystals, experiencing pressure from their neighbors, acquire so-called “internal stresses.” Their space lattice will be elastically deformed and, when the action of external forces ceases, will regain its former dimensions and shape. All this taken together explains why strain hardening and cold mechanical working (stamping, forging, drawing, etc.) have such a strong influence on the properties of the metal.

Pure iron is strain-hardened more strongly than steel. This must be explained by the fact that cementite is present in steel (in the free state if the carbon is more than 0.9%, and in pearlite at lower carbon content), which, owing to its brittleness, does not undergo plastic deformation. In addition, the iron crystals in steel are already strengthened, because the cementite plates penetrating them along the slip planes (such is the probable structure of pearlite) affect their strength, deforming by their influence the boundary parts of the space lattice of the iron crystallite. Some mutual influence of the lattices of iron and cementite undoubtedly exists, and this influence is most pronounced at the interface of the two crystallites. The strengthening of the metal under the influence of plastic deformation will be the greater, the greater the magnitude of the deformation \(\varepsilon\); it is quite natural that the strengthening, at least at the moment of transition through the elastic limit, is distributed through the specimen according to certain laws of the theory of elasticity and does not cover the whole specimen uniformly. Lüders lines indicate precisely those places where such a phenomenon (strengthening, strain hardening) has occurred.

II. Detection of Plastic Deformations.

Many observers have noted that iron rusts most readily along Lüders lines; oxidation, under suitable conditions, proceeds more rapidly at plastically deformed places, the greater the relative magnitude of the deformation. This phenomenon, somewhat modified, is used for etching the lines of deformation. It is, of course, necessary to bear in mind that the figures visible to the unaided eye represent a collection of microscopic lines on individual grains. Using a microscope, one can study the shear lines in individual crystallites, i.e., determine the deformations on a polished section, which is sometimes of great importance for practice.

Two methods for detecting plastic deformations are described in the literature: Oberhoffer’s^5) and Fry’s^6).

Oberhoffer’s etching method consists of the following. A part of iron deformed in the cold state is first heated to 200° for 30 minutes, polished on emery paper (the polishing is finished on 0000 emery paper), and immersed as the anode in an electrolyte of the following composition:

An iron or carbon plate, of the same or somewhat larger area than the specimen, is taken as the cathode. For one sq. cm. of anode area a current of 0.002 ampere is applied. The etching time varies around 30 minutes. The larger the bath and the greater the distance between the electrodes, the sharper the lines that appear on the specimen.

Since, despite its good results, this method is somewhat complicated for practice, lacking a laboratory and comparatively delicate instruments at hand, we settled on the aforementioned Fry etching method*), the essence of which we shall try to describe in greater detail.

A specimen suspected of having undergone plastic deformation in the cold state is held at \(200^\circ\) for \(1/2\) hour. This operation can be carried out in an electric furnace; for this purpose we made tubular resistance furnaces with an iron winding (of the Heræus type); small parts are most conveniently annealed in the vapors of boiling naphthalene, whose boiling point differs only slightly from \(200^\circ\)—(\(218^\circ\)). Naphthalene (ordinary, sold in apothecary shops) is poured into a flask of suitable size; the specimen is suspended on a wire or placed on the bottom. A vertical glass tube is used as a reflux condenser (diameter 12 mm, length 500 mm), fixed in the neck of the flask by means of an asbestos-cord stopper. The assembled apparatus is heated in a sand bath. After annealing, the specimen may be carefully machined on a planing machine or on some other machine, depending on the section in which we wish to observe the “figures of the action of force.” After removing the thinnest chip, the surface is ground, down to emery paper No. 0000. Such treatment after annealing is not revealed by etching. The ground surface is washed with water and alcohol, dried with filter paper or by blowing with bellows, and etched with a reagent of the following composition:

Fry reagent

In general this reagent does not have a strictly definite composition, but if there is too little hydrochloric acid in the solution, metallic copper may be deposited on the surface of the polished section. Copper is likewise precipitated, concealing the pattern, when, after etching, water gets onto the polished section covered with reagent. On the basis of our experience, we therefore recommend washing the specimen after etching not with alcohol, but with strong hydrochloric acid (technical acid may be used). Alcohol may also be used, but only if it contains absolutely no water. As soon as the reagent has been washed off with acid, the polished section may be rinsed with water and dried, as usual, with alcohol or filter paper. If, in spite of all precautions, a layer of copper is deposited on the surface, it must be dissolved in a saturated aqueous solution of potassium cyanide. One may also immerse the polished section in strong ammonia, so that a thin layer of liquid remains above the surface of the specimen. After several hours the copper will dissolve, but in the event of oversight the polished section becomes covered with rust, which is then washed off by the reagent.

The etching technique is as follows. The reagent is poured onto a horizontally lying polished section; after contact with iron it changes its green color to a dirty brown. The darkened liquid is poured off and the specimen is etched with a new portion. It is useful to wipe the polished section with cotton wool moistened with the solution. Since one usually has to work with large, heavy polished sections, it is very inconvenient to grip them with ordinary forceps. It is most convenient to take the specimen in the hands and wipe the surface of the polished section with a finger, but for this it is necessary to put on rubber gloves or finger cots; otherwise the skin of the hands is severely damaged by the hydrochloric acid.

*) Etching of “figures of the action of forces” during deformation in the cold state (Kaltwirkungsfiguren).

The duration of etching fluctuates around 20–30 minutes on average, although there are grades of iron that do not etch at all or require a considerable time for etching (80 hours or more). In such cases it is necessary to carry out the etching in automatically rocking baths. On a fully etched polished section the etched deformation lines are darker than the main mass of unchanged metal and lie below the plane of the polished section; they are, as it were, etched out. Therefore, if the difference in shades on the polished section after etching is not great, it can be intensified by a new, careful grinding; sometimes, in order to obtain a very distinct, contrasty pattern, for example for photography, it is necessary to etch and then lightly grind several times.

On average, 5 cu. cm of reagent are required for one sq. cm of the etched surface. If the iron belongs to a grade that gives good results when etched by this method, then it can also be etched with an old reagent that has already once been used. The most suitable temperature for etching is approximately 15°, although it must be thought that the lower the temperature of the reagent, the sharper the image obtained.

Unfortunately, if the iron is not homogeneous, as is almost always observed, then the etching will emphasize this heterogeneity. Interlayers rich in phosphorus, sulfur, etc., and segregation phenomena become very clearly visible; this can sometimes give occasion for erroneous interpretations.

Meyer and Eichholz ⁷) proposed a reagent of somewhat different composition. These authors propose adding ferric chloride to Fry’s reagent, and they proposed the following recipe:

But it seems that this solution is capable, to a greater degree, of revealing segregation bands than the original Fry reagent. Sometimes it is recommended to apply preliminary etching of the specimen in hydrochloric acid.

At the beginning of the article we mentioned the probable dependence of deformation lines on changes at the surface of crystals cut by the plane of the polished section. If one takes a sufficiently coarse-grained specimen of soft iron, anneals it as described earlier, and carefully etches it, then on the surface of individual grains, under the microscope, one can notice certain parallel lines, apparently slip lines. For this purpose it is best to use the following reagent:

The reagent acts slowly, but there is nevertheless a danger of over-etching. Therefore, during the etching process it is necessary to observe the polished section under the microscope. More careful grinding and polishing of the specimens is required; one cannot be limited to emery paper. Wet polishing with crocus or alumina on cloth disks is necessary until all scratches have completely disappeared.

III. The Authors’ Experiments.

We tested all the methods described for revealing plastic deformations. We think that the most suitable method is Fry’s method, since it requires almost no apparatus, current sources, measuring instruments, etc.

We were very interested in the question of how small a plastic deformation can still be detected by etching. For this purpose, [[unclear: continuation follows on next page]]

specimens of variable cross-section were made from soft boiler iron. They were subjected beforehand to heat treatment consisting of annealing up to 950° followed by slow cooling (about 50° per minute). On the surface of the specimen, scratches were made 10 mm apart by means of a dividing machine. Before and after stretching (on a 50-ton Amsler machine), the distances between the marks were carefully measured with a dividing microscope. Then the Shore number was determined for various places on the specimen after etching. A sketch of the specimen and the data of one series of measurements are presented in drawings 1 and 2 and in Table I.

TABLE I.

As was found, a plastic deformation of 0.2–0.3% (calculated over a length of 10 mm) already gives a clear picture upon etching. On closer examination of the specimen (photograph 1), two systems of bands are visible on it, inclined at equal angles to the axis of stretching; first one system of bands appears, and then, as the plastic deformation increases, the other system appears. At deformations above 5%, both systems of bands merge into a continuous dark field (photographs 2 and 3). The last photograph represents two longitudinal sections, perpendicular to one another and parallel to the axis of the specimen. It is clearly seen that on the planes of the sections there are lines equally inclined to the axis of the specimen. The outermost lines bound the volume that underwent deformation. Photograph 3 also gives two perpendicular sections of one and the same specimen, but here one section is parallel to the axis of the specimen, and the other is perpendicular. The polished section is the head of a ruptured specimen. Incidentally, on the head of the specimen (photograph 1) zigzag lines are noticeable, formed as a result of the crushing of the material by the notching of the grips of the testing machine. This compression extends through almost the entire depth of the specimen and is especially clearly visible in photograph 2. It should be noted that the perspective parts of the specimens in photographs 2 and 3 were drawn schematically by hand, and not photographed, as were sections a and b. In photographs 3 and 4 it is visible how the deformation lines, like tentacles, catch onto

for the undeformed metal in the head. The straightness of the lines was disrupted owing to the presence of roundings (fillets) at the transition from the body of the specimen to the head.

From the specimen in photograph 4, before annealing (200°), a chip 0.5 mm deep was removed on a lathe. Although such treatment could hardly have produced strong deformations, nevertheless traces of cutting, after grinding the machined surface to a depth of 0.2 mm,*) appeared in the form of clearly visible concentric circles. This once more confirms that, even with the most careful machining by chip removal, the metal still receives surface residual deformations, hardened to a certain depth. With a cutting depth of 4 mm (as our experiment showed), plastic deformations spread to a depth of 1–2 mm, and one must suppose that part of the energy during cutting goes into imparting plastic surface deformations to the part being machined. In this specimen, photograph 4, segregation lines in the form of vertical dark bands are especially noticeable.

In any case, one may think that the distribution of plastic deformations at the moment of transition beyond the elastic limit is similar to the distribution of elastic deformations previously seen, and therefore this method of detecting residual deformations can be applied to the solution of certain problems in the theory of elasticity. In photograph 5 a specimen of iron wire is given (diametral section), having the special feature that it has a fine-crystalline core (30 grains per 1 linear mm) and a coarse-grained crust (on average 1 grain per 1 mm). As can be seen, the deformation lines run only in the fine-crystalline part. The large crystallites, although they showed (photographs 6 and 7) internal slips, did not give total deformation lines visible to the naked eye. Some crystallites showed clear traces of destruction; as a result of intracrystalline slips and owing to the rotation of the shifted parts, new, finer grains are formed and the number of interlayers increases. In photograph 6, sliding lines are noticeable in the form of parallel curved bands; in photograph 7 the middle crystallite has, as it were, split into a number of elongated grains, which, with an increase in the degree of deformation, having rotated, will turn into independent grains.

During etching, the boundaries between grains and the slip lines appear; that is, the part of the specimen that is most fine-grained or even amorphous is etched. And it is precisely these places, most susceptible to the action of the reagent, that apparently give the macro-pattern. With stronger deformation, the general darkening of the surface during etching depends on the fact that the number of interlayers increases through mutual friction of the crystallites, as does the number of slip lines on individual grains.

The least clarified circumstance is the role of half-hour annealing up to 200°. It must be assumed that here phenomena of recrystallization and regrouping of atoms at elevated temperature take place. But in order to speak more definitely on this question, a series of additional experiments is necessary.

IV. Conclusions.

1. In the authors’ opinion, the best method of etching with the reagent is as follows. The specimen is heated in vapors of boiling naphthalene for 30 minutes, ground to emery No. 0000, and etched. After etching, the reagent is washed off with strong hydrochloric acid. The best results are obtained with soft iron.

2. Deformation lines appear in places with a deformation of 0.2–0.3%. The deformation at the place where the lines appear is obviously more significant. The entire deformed field darkens at deformations above 5%.

*) Of course the specimen was ground after 30 minutes’ heating to 200°.

1. Apparently, the lines appearing under small deformations are formed through the etching out of the intercrystalline interlayer separating crystallites that have begun to rotate, and which is newly formed as a result of this rotation.

The experiments were carried out in the Metallographic Laboratory of the STI; the treatment of the literature belongs to the first of the authors.

List of literary sources cited in the work

1. A. N. Dobrovidov, Vestnik Inzhenerov, 1926, Nos. 7–8.

2. A. N. Dobrovidov and N. P. Lyamzin, Journal of Applied Physics, 1927, vol. 4.

3. H. O’Neill, The Journal of Iron and Steel Inst., 1926, V. C XIII.

4. Z. Jeffries and S. Archer, The Science of Metals; Russian translation by N. F. Bolkhovitinov, Metallovedenie, Moscow, 1927, p. 23 ff.

5. F. Oberhoffer und M. Toussaint, Stahl und Eisen, 1924, No. 43.

6. Ad. Fry, Stahl und Eisen, 1921, No. 32.
   J. D. Jevons, The Journal of Iron and Steel Inst., 1925, V. C XI.

7. Meyer und Eichholz, Verein deutsch. Eisenhüttenleute, “Werkstoffs Ausschuss.” Ber. 20, 1922.

Photo 1. ×3/4.

Photo 2. ×3/4.

Photo 5. ×2 1/2.

Photo 3. ×3/4.

Photo 4. ×3/4.

Photo 6. ×100.

Photo 7. ×100.

Sketch of the specimen

①

Graph labels: vertical axis \(H_c?\); horizontal axis \(t\).
Scale marks: \(15.5,\ 17.5,\ 19.5,\ 21.5\) on the vertical axis; \(0.5,\ 1,\ 2,\ 3,\ 4,\ 5,\ 6,\ 7\) on the horizontal axis.

②

2. In contrast to crystals with natural faceting, the grains in a piece of metal, possessing all crystalline properties with the exception of regular faces, are called crystallites. ↩