Effect of Load and Sample Preparation Method on the Measured Hardness of LS-59-1 Brass

V. I. Karnov

(Presented by the scientific seminar of the Department of Machine Tools and Metal Cutting)

The Brinell hardness test was standardized in 1924. It was proposed that, in measuring hardness independently of the mechanical properties of the material being tested, the following ratio should be maintained:

\[ \frac{d}{D}=0.25\text{--}0.6, \]

where \(d\) is the diameter of the indentation and \(D\) is the diameter of the ball. At values of this ratio greater than 0.6, the measurement results are unsatisfactory. It was also proposed to make indentations with a 10 mm steel ball under a load of 500 kg for soft metals and 3000 kg for hard ones. But such a recommendation is entirely insufficient for an experimenter faced with the question of the degree of hardness to which a given metal should be assigned. According to the British technical specifications, the following table should be used [1]:

Initially we believed that the hardness of LS-59-1 brass, measured under a load of 3000 kg, was \(H_B = 110\text{--}115\ \text{kg/mm}^2\). The hardness obtained under a load of 500 kg proved to be \(H_B = 80\text{--}85\ \text{kg/mm}^2\). Despite the fact that, in determining hardness, we adhered to the recommended values

\[ \frac{d}{D} \leq 0.6, \]

the difference obtained was very large.

Does each load correspond to its own definite hardness, or can certain intervals be grouped with sufficient accuracy? For this purpose, experiments were carried out to measure the hardness of brass on a Brinell press with various loads for indenter balls of 5 and 10 mm diameter. The results for the dependence of hardness on load are given in Table 1 for a ball of 5 mm diameter and in Table 2 for a diameter of 10 mm. A visual representation of this dependence is given by the curves in Fig. 1.

For a ball of 5 mm diameter, when the load is increased to 500 kg, and for a ball of 10 mm diameter to 1000 kg, the hardness is practically the same; subsequently, as the load increases, it rises. In general, the curve characterizing the change in hardness as a function of load cannot—

will continue to increase without bound. There is a certain limit when the ball is pressed into the specimen up to its equator, i.e., when \(d = D\). The dependence of the numerical hardness values for a given specimen and a given ball on the load can be explained by the fact that the condition of geometrical similarity in Brinell testing is not fulfilled.

The principle of similarity was first formulated by Thomson and then finally formulated by Barba. “Similar bodies of a homogeneous

Table 1

Table 2

material, when loads equal in magnitude and direction are applied, deform to the same degree, both within and beyond the elastic

Fig. 1

limits, in proportion to the square of their linear dimensions.” From the principle of similarity the following consequences follow:

1. The results obtained cannot be comparable if geometrically similar impressions were made under different indentation conditions.

2. The loads for obtaining similar impressions must be proportional to the square of the linear dimensions of the impressions, i.e., the following condition must be fulfilled:

\[ \frac{P_1}{d_1^2} = \frac{P_2}{d_2^2} = \ldots \frac{P}{d^2} \quad \text{or} \quad \frac{P_1}{D_1^2} = \frac{P_2}{D_2^2} = \ldots = \frac{P}{D^2}. \]

The hardness of a given specimen with a given ball (Fig. 2, б) inevitably varies with the load, and at a given load it varies when balls of different sizes are used (Fig. 2, в). The impressions are consi-

are geometrically similar if the surface of each of them is bounded by identical so-called indentation angles \(\varphi\). Numerical expressions of hardness can be compared only when impressions with identical indentation angles are obtained (Fig. 2, \(a\)).

When determining the microhardness of the lateral surface of the cutting zone, it was necessary to decide whether to remove the work-hardened layer. It is known that, when hardness is measured with an indenter of conical shape or with a pyramid, owing to the similarity of the impressions the hardness should not depend on the applied load. The presence of a work-hardened layer leads to the fact that, as the load on the hardness indenter is decreased, the hardness increases. To remove the work-hardened layer, three methods were used. First, a method of careful grinding and polishing of the lateral surface of the specimen was tested. The final operation of the second method was electropolishing, and, finally, a method of chemical etching with a reagent of the following chemical composition was tested [2]: \(\mathrm{H_2SO_4}\)—50%, \(\mathrm{HNO_3}\)—36%, \(\mathrm{H_3PO_4}\)—14%.

The results of all three methods, when hardness was measured on a Vickers instrument, are given in the following table.

From the table presented it is seen that electropolishing almost does not remove the work-hardened layer, and only chemical etching leads to the desired result. This conclusion was also confirmed when measuring hardness on the PMT-3 instrument.

Fig. 2.

REFERENCES

1. O’Neill. The Hardness of Metals and Its Measurement. GNTI, 1940.
2. P. V. Shchigolev. Electrolytic and Chemical Polishing of Metals. Publishing House of the Academy of Sciences of the USSR, Moscow, 1955.