On the Methodology for Testing Coals with Respect to Mechanical Strength and Thermal Stability

I. V. Gebler

I

The mechanical properties of fossil coals, i.e., their resistance to such effects as crushing, impact, and abrasion, are of great importance in coal technology. Brittle coals yield many fines (shtyb), which often creates inconvenience when coal is burned on an ordinary grate; non-brittle coals, meanwhile, provide lump fuel. In crushing coal, in certain cases the action of crushing machines must be proportioned to its brittleness so as, on the one hand, to ensure comminution to the required limit and, on the other hand, to avoid as far as possible the formation of excessively fine classes (flour), if for some reason this is undesirable. The mechanical properties of coals are also used in their dry beneficiation in friction separators, where the rounded shape of the lumps, characteristic of harder coals (anthracites), is significant, and in so-called selective crushing in special crushing apparatuses operating in such a way that brittle coal, when passing through them, is comminuted and screened out, whereas the harder and more difficultly comminuted rock passes through unchanged and is separated from the coal.

In this connection, the mechanical strength of coals, as well as their thermal stability, which partly depends on it—i.e., the ability to burn in a lump without its destruction during the combustion process—must be one of the classification criteria in the practical grading of coals.

The mechanical strength of coals cannot be placed in direct relation to their hardness, since coals sometimes possess considerable resistance to destructive effects by virtue of their toughness (sapropelites); in other cases the destructibility of coals is connected with their texture, the presence of cracks, etc. (crumpled seams). The hardness proper of fossil coals generally ranges from 1 to 3 on the mineralogical scale. The lowest hardness is possessed by brown coals—earthy brown coals; the highest by anthracites and some sapropelites. Hard coals proper have an intermediate hardness, and one that varies according to their petrographic character. Dull coals with a predominance of durite are harder than lustrous varieties, which, together with this, possess greater brittleness and often exhibit sharply pronounced cracking.

The following materials are available on the study of coals with respect to mechanical properties. At the Stalin Branch of the Coal Research Institute, under the direction of Academician A. N. Dinnik, determinations were made of the temporary compressive strength and the modulus of elasticity for certain coals of the Donbass (the Smolyaninovskii seam, the Budennovskoe deposit, and the Bugovskii—Stalinskoe deposit) [1]. These values—

values were determined in three directions, namely, with the compressive force directed: 1) perpendicular to the bedding, 2) along the bedding (edgewise), 3) perpendicular to the sides. The results of the determinations are given in Table 1.

Table 1

where

$K$ — temporary compressive strength, kg/cm²
$E$ — modulus of elasticity, kg/cm².

At the Prokopyevsk mine, experiments to determine the mechanical properties of coals were carried out as follows: pieces of coal $>100$ mm were dropped onto a cast-iron plate from heights of 2, 2.5, and 3 m. After each dropping, the coal was screened on a sieve with openings of $36 \times 36$ mm; pieces of the $>36$ mm class were dropped again, for a total of up to 10 times. The strength coefficient was taken to be the percentage of the $>36$ mm class relative to the sample taken. The results of these experiments are given in Table 2 (average figures).

Table 2

Table 3 also gives the following data concerning the strength of some Kuzbass coals in crushing tests and in a drum (of the “Mikum” type).

At the author’s request, Engineer I. K. Murzin, in the mechanical laboratory of the Tomsk Polytechnic Institute, determined the temporary compressive strength of coal from the Moshchnyi seam of the Prokopyevsk deposit (Kuzbass); this coal belongs to the especially strong and hard coals. The figures obtained are as follows (Table 4) [2].

According to data from Engineer Baranov, the resistance of coals to crushing in individual pieces ranges from 20 to 40 kg/cm², and for anthracite up to 90 kg/cm² [3].

Table 3

Table 4

As is evident from the figures given above in Tables 3 and 4, some Kuzbass coals deviate greatly from the indicated average values in crushing resistance, in this respect considerably surpassing even anthracites.

At the end of 1937, G. E. Bakanov and A. N. Postoevko carried out experiments at the Kaganovichugol trust to determine the relative strength of coal by the method of detonating standard charges of grisoutite in order to find the degree of resistance of coal to breaking [4].

Judgment as to the degree of strength of the coal of one seam or another was obtained by comparing the sizes of the base areas of the blast craters formed after the explosion. In this way a “strength index” was established for coal,

$$ \gamma = \frac{1}{F_{\mathrm{av}}}, $$

where $F_{\mathrm{av}}$ is the mean area of the crater base in m², obtained from two experiments.

Table 5 gives the results of experiments conducted for various coal seams of the Kaganovichugol trust.

In accordance with the strength indices for coals, the authors give the following classification of them by strength:

a) very strong coal
when $F \leq 0.20$ m² and $\gamma = 5.0$;

b) strong coal
when $F > 0.20$ to $0.40$ m² and $\gamma < 5.0$ to $2.5$;

c) coal of medium strength
when $F > 0.50$ to $0.80$ m² and $\gamma < 2.5$ to $1.25$;

d) weak coal
when $F > 0.80$ to $1.33$ m² and $\gamma < 1.25$ to $0.75$;

e) very weak coal
when $F > 1.33$ m² and $\gamma < 0.75$.

Table 5

Thus, coals show very large differences with respect to mechanical strength, in connection with which there is a need to determine the mechanical properties of coals; for this purpose an appropriate laboratory method must be developed.

Recently K. I. Syskov [5] proposed a method for determining the strength of lump materials (coal, coke, semicoke), based on the work of destroying them, which can be calculated when a load is dropped onto the material being tested; the test is carried out in an apparatus consisting of a steel cup, on the bottom of which the material under test is placed in a definite granulometric composition, a tube entering the cup, and a load falling in the tube onto the material. In the test the latter is taken with a size of 25–5 mm in an amount of 25 g after preliminary crushing until it passes through a sieve with 25 mm openings, followed by screening off the fines through a sieve with 5 mm openings.

After the load has been dropped three times, the material is sieved through sieves with round openings of 1 and 3 mm. The resulting classes 25–3 mm, 3–1 mm, and 1–0 mm are weighed.

The surface area of the pieces of material is calculated by the formula:

\[ S=\sum \frac{6ad^2}{\gamma}\ \text{cm}^2, \]

where \(a\) is the yield of the given class,
\(d\) is the average diameter of the pieces, cm,
\(\gamma\) is the apparent specific gravity of the material, g/cm\(^3\).

Further, on the basis of Rittinger’s theory, the work of destruction is calculated in accordance with the surfaces of the charge taken and newly formed after impact on the material.

Such a method, developed for coke, gives rise to doubts as to the accuracy and comparability of the figures obtained when it is applied to coals.

As is evident from the formula, in calculating the surface area the grains of the material are taken as cubes, and if this is at all permissible for coke, then for coals, with their heterogeneity in petrographic composition and with the very different mechanical properties of the ingredients, such an assumption must be regarded as unfounded: here one may expect grains of elongated shape, which will lead to large errors.

Thus, for a cube with an edge equal to 10 cm, the surface is 600 cm²; for a parallelepiped with edges of 10 and 20 cm the surface will be 1000 cm², and with edges of 10 and 30 cm—1400 cm²; for a tetrahedron with dimensions corresponding to passage through the same sieve with openings of 10 × 10 cm, the surface will amount to only 200 cm². The obtaining of comparable figures for different coals thus becomes doubtful.

The micromethod developed by B. E. Koifman for testing the mechanical strength of coals [6] gives good agreement of results in parallel experiments.

However, the preliminary crushing of coals to separate the 5—4 mm fraction, which is taken for the determination, may cause weakening of the coal, so that the figures obtained may not fully correspond to the strength of coals in relatively large pieces, which is necessary for practical conclusions.

As regards thermal stability, in this respect coals have been studied still less, just as there is no established and tested method for such testing of coals.

In the literature on this question we were able to find only one reference to a procedure for determining the thermal stability of coals at the Kharkov Institute of Transport Engineers, which is carried out as follows [7]. Into an electric muffle furnace heated to 900°, a weighed piece of fuel weighing 50—70 g, of size 50—30 mm, is placed in a quartz cup. The furnace is closed with a mica shutter, through which the specimen is observed. The temperature of the furnace is measured by a thermocouple. The specimen is kept in the furnace at high temperature for 15 minutes, then removed from the muffle and, after cooling, weighed. In this way the percentage loss in weight is calculated. After screening through a sieve with four openings per square centimeter, the amount of fines formed is determined. The percentage loss in weight and fines is calculated in relation to the initial weight of the specimen. The degree of thermal stability is determined by the external appearance and strength of the piece removed from the muffle. To determine strength, the test specimen is placed on a horizontal plate, and above it is installed a hollow iron cylinder 30 cm high and 7.5 cm in diameter. A load of 250 g is dropped onto the test specimen, and after the impact the granulometric composition of the specimen characterizes its strength.

In this method the weak point appears to be that for the thermal action one piece of coal is taken, which after heating may change its shape and dimensions; this must affect the accuracy of the results under the subsequent impact, since pieces generally of unequal shape and size will be subjected to this impact for different coals. Therefore, one should expect difficulty in obtaining by this method any sufficiently accurate figures, and moreover figures comparable for different coals. In connection with this, we did not find it necessary to reproduce the corresponding experiments.

II

For testing coals with respect to mechanical strength, it would be desirable to be able to carry out the corresponding determinations in the laboratory with small quantities of coal and with simple equipment, obtaining indices that would characterize the relative mechanical strength of different coals. Determination of compressive strength appears insufficiently convenient and practical, since in this case it is necessary to prepare from a piece of coal a specimen of regular geometric shape (a cylinder or cube) with precise cross-sectional dimensions, which is rather laborious for practical work; all the more so because, as is evident from the data presented above, such determinations must be made each time in three directions relative to the bedding planes, which increases the number of determinations and complicates preparation of the specimen.

It would be more expedient to apply mechanical action, under specified conditions, to several pieces of coal simultaneously, with subsequent accounting of the degree of destruction of the pieces by sieve analysis. Such a procedure is simpler, and at the same time, when several pieces of coal are tested, random factors that may occur when testing a single piece are eliminated, and average values should be obtained already in one experiment.

Fig. 1

In accordance with this, for mechanical testing of coals a crusher recommended by Broche and Nedelmann [8] was used. It consists of a drum 200 mm in diameter and 75 mm wide. On the inner surface of the drum there are, symmetrically arranged in each quarter, 4 iron strips bent in the form of a trough for gripping pieces of coke. The rotor of the crusher consists of three hammers, hinged to the shaft (Figs. 1 and 2).

The rotational speed of the crusher and the gear ratios were somewhat changed.

As is known, the rotational speed of the drum must not exceed a certain critical value at which the pieces of material, owing to the development of centrifugal force, will be pressed against the surface of the drum and rotate together with it, in which case the crushing action will be reduced to zero.

Fig. 2

The possibility of crushing action is determined by the following expression:

\[ n < \frac{30}{\pi R}\sqrt{Rg}, \]

where \(n\) is the angular speed of rotation of the drum (number of revolutions per minute), \(R\) is the radius of the drum, m, and \(g\) is the acceleration of gravity, equal to \(9.81 \, m/sec\).

For our case we have:

\[ n < \frac{30}{3.14 \cdot 0.1}\sqrt{0.1 \times 9.81} < 91.4, \]

i.e., the number of revolutions of the drum per minute must be less than 91.

The rotational speed of the drum was set at 60 revolutions per minute; the corresponding angular speed of rotation of the rotor was 810 rpm; the peripheral speed of rotation of the hammers themselves, \(n'\),

\[ (R' = 0.081 \, m) = \frac{\pi n' R'}{30} = 6.8 \, m/sec. \]

After a series of experiments, the following operating procedure was established. Pieces of the coal to be tested are split with the aid of a chisel and hammer into such pieces as would pass through a sieve with round holes 25 mm in diameter and would be retained on a sieve with holes 15 mm in diameter.

Thus, for the mechanical test the 25–15 mm size fraction is always taken; 100 g of such pieces, amounting to 20–23 pieces, are placed in the crusher, where they are subjected to crushing for 10 seconds (timed with a stopwatch).

After this the crusher is emptied, and the coal is subjected to sieve analysis on sieves with round holes of diameters 15, 10, 7, 5, 3, and 1 mm. The yield of the fractions in percent characterizes the resistance of the coal to destruction under the given conditions, or its relative mechanical strength. For a clear representation of the sieve composition of the crushed

a screening curve of the product is plotted. Since the amount of the class larger than 15 mm is very characteristic, the strength of the coal may be expressed briefly by the amount of this class in percent, which is taken as the coefficient of mechanical strength of the coal \(K_m\) (in round numbers¹).

The following Kuzbass coals were subjected to testing according to the above-described procedure (Table 6).

Table 6

In what follows, for brevity, the coals are designated by numbers in accordance with Table 6.

Coals Nos. 3, 4, and 5 are completely non-caking; the others possess negligible caking capacity.

The results of processing the coals in the crusher are given in Table 7 and in Figs. 3–9.

Table 7

| Coal No. | Experiment | \multicolumn{7}{c}{Yield of classes in %} |
|---:|:---:|---:|---:|---:|---:|---:|---:|---:|
| Coal No. | Experiment | mm > 15 | 15—10 | 10—7 | 7—5 | 5—3 | 3—1 | 1—0 |
| 1 | 1 | 38.5 | 17.2 | 14.1 | 7.1 | 5.8 | 9.1 | 5.9 |
| 1 | 2 | 34.8 | 23.9 | 11.2 | 6.6 | 6.5 | 9.6 | 5.9 |
| 1 | average | 36.6 | 20.5 | 12.6 | 6.8 | 6.6 | 9.3 | 5.9 |
| 2 | 1 | 47.1 | 21.8 | 7.9 | 4.0 | 5.0 | 7.8 | 5.1 |
| 2 | 2 | 48.6 | 23.6 | 6.7 | 3.3 | 4.4 | 7.4 | 4.9 |
| 2 | average | 47.8 | 22.7 | 7.3 | 3.6 | 4.7 | 7.6 | 5.0 |
| 3 | 1 | 22.4 | 24.7 | 13.4 | 7.9 | 9.9 | 13.5 | 7.0 |
| 3 | 2 | 26.4 | 13.8 | 11.7 | 7.4 | 13.3 | 17.5 | 7.0 |
| 3 | average | 24.6 | 19.2 | 12.5 | 7.6 | 11.6 | 15.5 | 7.0 |
| 4 | 1 | 22.2 | 22.1 | 15.6 | 10.5 | 9.8 | 12.9 | 6.2 |
| 4 | 2 | 21.0 | 26.0 | 15.5 | 8.9 | 9.5 | 12.2 | 5.6 |
| 4 | average | 21.6 | 24.0 | 15.5 | 9.7 | 9.6 | 12.6 | 5.9 |
| 5 | 1 | 21.0 | 23.2 | 10.9 | 10.4 | 11.8 | 15.1 | 6.4 |
| 5 | 2 | 17.6 | 24.7 | 13.3 | 13.9 | 11.5 | 13.1 | 5.9 |
| 5 | average | 19.3 | 23.9 | 12.1 | 11.8 | 11.6 | 14.1 | 6.0 |
| 6 | 1 | 55.5 | 16.2 | 4.9 | 5.7 | 4.7 | 6.4 | 4.2 |
| 6 | 2 | 58.4 | 13.8 | 5.0 | 4.8 | 3.5 | 5.2 | 3.5 |
| 6 | average | 56.9 | 15.0 | 4.9 | 5.2 | 4.1 | 5.8 | 3.8 |
| 7 | 1 | 40.4 | 21.2 | 8.9 | 5.5 | 7.3 | 9.0 | 6.0 |
| 7 | 2 | 37.1 | 19.8 | 7.3 | 6.6 | 5.9 | 9.9 | 4.8 |
| 7 | average | 38.7 | 20.5 | 8.1 | 6.1 | 6.6 | 8.9 | 5.4 |

Fig. 3

Fig. 4

Axis labels visible on the graph: \(-1\ \mathrm{mm}\), \(+1\), \(+3\), \(+5\), \(+7\), \(+10\), \(+15\); \(0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100\%\).

Fig. 5

Axis labels visible on the graph: \(-1\ \mathrm{mm}\), \(+1\), \(+3\), \(+5\), \(+7\), \(+10\), \(+15\); \(0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100\%\).

Fig. 6

Axis labels visible on the graph: \(-1\ \mathrm{mm}\), \(+1\), \(+3\), \(+5\), \(+7\), \(+10\), \(+15\); \(0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100\%\).

Fig. 7

Fig. 8

Fig. 9

As can be seen, the coals are distinctly characterized by granulometric composition after treatment in the tumbler. The coefficients in the above expression have the following values for them (in round numbers):

The strongest of these coals proves to be the coal of the upper bench of the Volkov seam; next comes the coal of the Moshchnyi seam; the commercial coal of grade “SS” from the Chernaya Gora mine and the coal of the Bezymyannyi seam from the same mine are close to one another in strength; the coals of the Kuibyshevugol trust possess the least strength.

The character of the figures obtained agrees with the notions of coal strength that exist on the basis of separate tests of them by other methods.

III

As already indicated above, thermal stability is understood as the resistance that coal offers under combustion conditions, without at the same time breaking down into smaller pieces. This property is of special importance for non-caking or very slightly caking coals under various conditions of their combustion. When burned on a grate, the fines formed during combustion contribute to an increase in fuel losses in the form of fall-through and carry-over. In shaft furnaces, as well as in gas generators, the accumulation of fines impedes the movement of gas and disrupts the proper operation of the furnaces; accordingly, thermally stable coals are the most suitable for the latter purpose.

The disintegration of a lump of coal under the influence of high temperature under combustion conditions must be caused by internal stresses arising in the lump of coal under thermal action, which entails the appearance of cracking and, as a consequence, the disintegration of the lump and the formation of fines in the very process of combustion. Low elasticity and, consequently, high brittleness of coal promote this phenomenon, in view of which it should be assumed that coals with greater mechanical strength will also possess greater thermal stability.

If the combustion of coal is interrupted after a certain time of exposure of the coal lumps to high temperature, the latter, obviously, should show lower strength in comparison with the initial strength, and in this way the thermal stability of coals can be determined.

After a number of preliminary experiments in this direction, the following working procedure was established.

Coal in lumps—the same coal from which samples were taken for the mechanical test (size class 25–15 mm)—in an amount of 200 g is placed in a flat porcelain dish and then put into a muffle furnace, in which a temperature equal to 800°C is set beforehand. Suitable dimensions of the muffle itself, for example, are as follows: width—20 mm, length—290–300 mm, height—120 mm. The coal is kept in the muffle for exactly 10 minutes, after which the dish is removed, and the burning coal is immediately covered with another porcelain (inverted) dish of corresponding (somewhat

(smaller) size. After cooling, the pieces of coal are screened through a sieve with holes 25 mm in diameter¹, and the class larger than 15 mm is selected; 100 g of pieces of this class are subjected to treatment in the crusher and subsequent screening under exactly the same conditions as in determining the mechanical strength of coals (time of treatment in the crusher: 10 seconds; screening on sieves of 15, 10, 7, 5, 3, and 1 mm).

The granulometric composition of the crushed product gives an overall picture of the behavior of the thermally treated coal under mechanical action; for calculation, however, the percentage yield of the class larger than 15 mm is taken, which characterizes the mechanical strength of the coal and gives the value of the coefficient of mechanical strength \(K'_m\) for the thermally treated coal. The thermal stability of coals will be judged from the strength of the coals after heating, expressing it as a percentage of the original strength; this gives the coefficient of thermal stability of coals \(K_T\).

Thus we have:

\[ K_T=\frac{K'_m}{K_m}\times 100. \]

For an overall characterization of the mechanical properties of coals, all three coefficients are therefore taken into account, i.e., \(K_m\), \(K'_m\), and \(K_T\).

For the coals investigated, when tested under the specified conditions, the figures given in Table 8 and in Figs. 10–16 were obtained.

Table 8

| No. of coals | Tests | \multicolumn{7}{c}{Yield of classes, in %} |
|---:|:---:|---:|---:|---:|---:|---:|---:|---:|
| | | mm > 15 | 15–10 | 10–7 | 7–5 | 5–3 | 3–1 | 1–0 |
| 1 | 1 | 18,3 | 55,8 | 12,2 | 3,4 | 2,4 | 3,2 | 2,4 |
| 1 | 2 | 27,3 | 41,2 | 13,7 | 6,7 | 3,3 | 4,1 | 3,0 |
| 1 | average | 22,8 | 48,5 | 12,9 | 5,0 | 2,8 | 3,7 | 2,7 |
| 2 | 1 | 25,8 | 43,1 | 16,0 | 5,0 | 3,7 | 4,9 | 3,0 |
| 2 | 2 | 24,0 | 43,4 | 14,6 | 4,6 | 3,4 | 4,5 | 3,15 |
| 2 | average | 24,9 | 43,2 | 15,3 | 4,8 | 3,5 | 4,7 | 3,7 |
| 3 | 1 | 0,0 | 18,5 | 28,4 | 20,1 | 13,7 | 12,9 | 4,1 |
| 3 | 2 | 1,6 | 23,2 | 25,8 | 19,8 | 13,8 | 14,1 | 4,6 |
| 3 | average | 0,8 | 20,8 | 27,1 | 19,9 | 13,7 | 13,5 | 4,3 |
| 4 | 1 | 1,9 | 20,3 | 31,2 | 16,3 | 12,2 | 13,0 | 4,6 |
| 4 | 2 | 1,4 | 19,7 | 27,8 | 18,3 | 13,4 | 13,2 | 4,1 |
| 4 | average | 1,6 | 20,0 | 29,5 | 17,3 | 12,8 | 13,1 | 4,3 |
| 5 | 1 | 3,5 | 16,4 | 29,3 | 18,2 | 14,1 | 12,7 | 4,1 |
| 5 | 2 | 0,0 | 18,7 | 26,2 | 22,3 | 14,4 | 12,7 | 3,9 |
| 5 | average | 1,7 | 17,6 | 27,8 | 20,3 | 14,3 | 12,7 | 4,0 |
| 6 | 1 | 24,4 | 48,3 | 8,7 | 3,8 | 3,2 | 5,6 | 4,3 |
| 6 | 2 | 23,9 | 45,0 | 8,2 | 3,5 | 3,5 | 5,1 | 4,3 |
| 6 | average | 24,1 | 46,7 | 8,5 | 3,6 | 3,4 | 5,3 | 4,3 |
| 7 | 1 | 32,2 | 31,1 | 9,6 | 6,7 | 5,1 | 7,8 | 5,8 |
| 7 | 2 | 33,2 | 31,6 | 9,9 | 5,7 | 4,3 | 5,5 | 4,8 |
| 7 | average | 32,7 | 31,4 | 9,8 | 6,2 | 4,7 | 6,7 | 5,4 |

Fig. 10

Fig. 11

Fig. 12

Fig. 13

Fig. 14

Fig. 15

The values of the coefficient of thermal stability for the coals will be as follows:

Thus, the lean coals of the Kuibyshevugol trust have the lowest thermal stability; at the same time they are also the weakest with respect to mechanical strength.

Fig. 16

The stronger coals have much greater thermal stability. Coal No. 7 shows a certain deviation: with average mechanical strength it has the greatest thermal stability. It is possible that this circumstance is connected with its somewhat greater caking capacity; after heating, greater fusion of the pieces and a greater change in the appearance of their surface are observed. The coked pieces have greater mechanical strength and, consequently, a greater value of $K_T$.

Conclusion

A rapid and simple method has been developed for testing, under laboratory conditions, non-caking or very weakly caking coals with respect to mechanical strength and thermal stability.

Indices have been established for expressing the indicated properties of coals.

Descriptions are given of installations for the mechanical testing of coals.

LITERATURE

1. Peskovatskii A. A. Coal (Coal and Iron), No. 99, p. 144, 1933.
2. Murzin I. K. Coal (Coal and Iron), No. 106—107, pp. 123, 1934.
3. Levenson L. V. Machines for the Beneficiation of Minerals, p. 387, 1933.
4. Bakanov G. E. and Postoenko A. N. Proceedings of the Tomsk Industrial Institute, vol. 60, issue 2, 1940.
5. Syskov K. I. Factory Laboratory, p. 1202, 1947.
6. Koifman B. E. Factory Laboratory, vol. XIII, p. 711, 1947.
7. Heat Engineering, edited by Prof. V. P. Naumov, vol. I, p. 26, 1938.
8. Broche G. T. and Nedelmann G. Glückauf, No. 35, p. 769, 1932.

1. If the pieces of coal have become somewhat fused together, which may occur even with very weakly caking coals, the fused pieces are carefully separated by hand. ↩↩