INFLUENCE OF THE SHAPE AND DIMENSIONS OF PARTS ON THE DEPTH OF THE DIFFUSION LAYER DURING THERMOCHEMICAL TREATMENT OF STEEL

I. T. Tikhonov and I. V. Slosman

(Presented by Professor, Doctor A. N. Dobrovidov)

Introduction

The great industrial importance of thermochemical treatment processes for steel has caused, and continues to cause, numerous investigations devoted to the study of the basic regularities that occur in these processes. It may be considered that, at present, the influence on the results of thermochemical treatment of the process temperature, its duration, and the composition of the steel being treated has been sufficiently studied.

There is, however, one question that has been little studied. We have in mind the influence, on the depth of the diffusion layer, of the curvature of the parts being treated. Usually, when specifying the thermochemical-treatment regime for obtaining a diffusion layer of a definite depth, the shape and dimensions of the parts subjected to treatment are not taken into account.

M. M. Zamyatnin [1] showed that this factor cannot be neglected, and that the depth of the diffusion layer depends on the shape and dimensions of the parts. Since his work is the only one in this respect, we set ourselves the aim of experimentally verifying the character and degree of the influence of the shape and dimensions of specimens on the depth of the diffusion layer in certain types of thermochemical treatment of steel.

Experimental Part

Carburizing was carried out on specimens of Armco iron in the form of cylinders 10 mm high and 12, 6, 3, and 1.5 mm in diameter, and also in the form of hollow cylinders with an outside diameter of 12 mm and inside diameters of 6, 3, and 1.5 mm. Repeated experiments were performed on the same kind of specimens, but cut in half along the axis. In addition to cylindrical specimens, in each experiment a specimen in the form of a cube with a side of 10 mm was carburized. The specimens were carburized with carbon in a solid carburizer consisting of 90% charcoal and 10% barium carbonate.

The specimens and carburizer were packed in an iron cup 60 mm in diameter and 150 mm high. Carburizing was carried out in a muffle—

in an electric furnace with a nichrome heater. The carburizing temperature in the first experiment was 900°, and the holding time at this temperature was 6 hours. After carburizing, the specimens were embedded, for preparation of microsections, in a low-melting alloy and were ground down from the end to a depth known to exceed the depth of the carburized layer. The depth of the layer was determined from the microstructure of the sections after etching them with a 4-percent solution of nitric acid.

Data on the depth of the diffusion layer are presented in Fig. 1.

Fig. 1. Influence of surface curvature on the depth of the diffusion layer. Carburizing temperature—900°, duration—6 hours.

By the depth of the layer here is meant the sum of the eutectoid and hypereutectoid layers.

The graph illustrates quite clearly the dependence of the depth of the diffusion layer on the shape and dimensions of the specimen.

In the carburizing of hollow cylinders, the inner and outer surfaces are not under quite identical conditions: on the inner surfaces, contact with the carburizer and with carbon monoxide, formed according to the reaction,

\[ \mathrm{BaCO_3 + C = BaO + 2CO}. \]

is hindered.

To eliminate the influence of this factor, a repeated experiment was carried out in which hollow cylinders cut into two equal parts along their axes were subjected to carburizing.

The results of this experiment are shown in Fig. 2.

The experimental data obtained show quite definitely that the depth of the diffusion layer on cylindrical specimens is greater than on a plane; moreover, the depth of the layer on cylindrical specimens is the greater, the smaller the cylinder diameter. The depth of the layer in holes is less than on a plane. As the diameter of the hole decreases, the depth of the layer decreases.

In the already mentioned work of M. M. Zamyatnin [1], this dependence is explained both by means of a rigorous mathematical solution and by a simplified solution, under the assumption that the amount of element entering the surface and its average concentration in the layer do not depend on the shape of the surface.

Fig. 2. Influence of surface curvature on the depth of the diffusion layer. Carburizing temperature—950°, holding time 4 hours. Hollow specimens, sectioned along the axis (a—hypereutectoid layer; b—sum of the eutectoid and hypoeutectoid layers).

Then the volumes of the layers per unit surface area must be equal for products of different shapes. If in this case the layer depth for a plane is denoted by \(\delta\), then the layer depths for a cylinder and for a cylindrical hole will be, respectively,

\[ \delta_{\text{cyl}} = R - \sqrt{R^{2} - 2R\delta}, \]

\[ \delta_{\text{hole}} = \sqrt{R^{2} + 2R\delta} - R. \]

Investigation of the microstructure showed that the curvature of the surface affects not only the depth of the carburized layer, but also the concentration of carbon in it: on the outer surface of a hollow cylindrical specimen the carbon concentration is greater than on the inner surface.

Figure 3 gives a photograph of the microstructure of two specimens, one of which had an internal hole 6 mm in diameter, while the other was a solid cylinder also 6 mm in diameter. The specimens were carburized simultaneously. Whereas in the first specimen the carbon concentration does not exceed the eutectoid concentration, in the second a small hypereutectoid layer is clearly visible. This is explained by the fact that the conditions for removal of carbon into the interior of the specimen from the inner surface of the cylinder are more favorable than from the outer surface.

In order to determine whether, in other types of thermochemical treatment, the regularities described...

above for the case of carburizing with carbon, experiments were carried out on aluminizing (cementation with aluminum).

In this case, the cementing medium used was a mixture consisting of 70% ferroaluminum (50 percent), 29% aluminum oxide, and 1% ammonium chloride. Samples of Armco iron and grade 15 steel were subjected to aluminizing. The samples were heated to a temperature of 950° and held at this temperature for 10 hours.

Measurement of the depth of the aluminized layer showed that the dependence of the depth of the diffusion layer on the shape and dimensions of the samples that occurs during carburizing with carbon is not observed in the case of aluminizing. The layer depth in flat samples and in hollow cylindrical samples on the outer and inner surfaces was practically the same (Fig. 4).

Fig. 3. Different carbon concentration on the outer and inner surfaces.

This suggests that the principal role in the process of aluminizing in powder mixtures is played by the exchange reaction occurring on the surface of the sample and yielding atomic aluminum, which diffuses into the steel [2]:

\[ \mathrm{Fe} + \mathrm{AlCl}_3 = \mathrm{FeCl}_3 + \mathrm{Al} \]

It is evident that, for this reaction to proceed, diffusion of iron atoms toward the surface of the sample is necessary. In this case, with an increase in the curvature of the sample surface, the concentration of iron atoms in the surface layer should decrease and, consequently, the aluminizing process should be slowed.

Thus, an increase in the curvature of the sample surface, if the process of formation of the diffusion layer is considered from the standpoint of diffusion of only aluminum atoms, should increase the layer depth. If, however, the process is considered from the standpoint of counter-diffusion of iron atoms, which is necessary for the successful course of the exchange

reaction with the formation of atomic aluminum, an increase in the curvature of the specimen surface should reduce the layer depth. The opposite effect of the curvature of the specimen surface on these two processes led to the result that, in our experiments, the depth of the diffusion layer remained constant, within the limits of measurement error, for specimens of different shape and size.

Fig. 4. Effect of surface curvature on the depth of the aluminized layer of Armco iron. Aluminizing temperature—950°; duration—10 hours.

Conclusions

1. The influence of surface curvature on the depth of the diffusion layer during carburizing with carbon has been confirmed. The greater the curvature of the specimen surface, the greater the depth of the diffusion layer.

2. During cementation with aluminum, the curvature of the surface of the specimens, within the investigated limits, does not affect the depth of the diffusion layer.

Literature

1. Zamyatnin M. M. DAN SSSR, No. 4, vol. 68, 1949.
2. Minkevich A. N., Chemical-Thermal Treatment of Steel, Mashgiz, 1950.