METHOD FOR DETERMINING THE SPECIFIC GRAVITY AND RELATIVE VISCOSITY OF UNSTABLE SUSPENSIONS USED IN COAL BENEFICIATION

I. V. Gebler

As weighting agents for aqueous suspensions in coal beneficiation, magnetite, pyrite cinders, and barite are used. These substances, with a specific gravity of 4.5–5.0, even when finely ground—for example, to 0.06–0 mm—form unstable suspensions; the grains of the material settle rapidly when at rest, and the homogeneity of the suspension over its height, even over a short interval of time, is not the same.

Specific gravity and viscosity are the most important properties of suspensions in the beneficiation of coal in them. Specific gravity determines the floating and sinking of a given fraction of coal depending on the content of mineral substances and, in production, is set to a prescribed value in accordance with the character of the coal or middlings.

The viscosity of a suspension determines one or another rate of fall and rise of the particles of the material being beneficiated, i.e., the rate of stratification, on which in turn depend the time required for the material to remain in the suspension and, correspondingly, the dimensions of the apparatus. In this connection, the specific gravity and viscosity of a suspension are always determined in production control.

The specific gravity of suspensions is determined very simply by the usual method, namely by weighing a definite volume of suspension; the determination of the viscosity of a suspension, however, presents certain difficulties because of their instability and is possible only while the suspension is being stirred. Ordinary viscometers for determining the specific viscosity of liquids, based on the rate of outflow, are in this case entirely unsuitable, since in such viscometers it is difficult to carry out proper stirring, and compaction of the suspension occurs at the outlet opening at the bottom of the vessel; viscometers in which viscosity is determined on the basis of measuring the rate of fall of some body are also inconvenient; at low suspension viscosities, large dimensions of the apparatus would be required. The most suitable here should be viscometers based on measuring the force necessary to rotate, in the suspension at a definite speed, some body that can at the same time play the role of a stirrer (the torsion method); but they are comparatively complex and can become maladjusted owing to the entry of the finest particles of weighting agent into the parts of the apparatus.

In addition, suspensions of magnetite and pyrite, with the hardness of these minerals up to 6.5, have a strong abrasive action on metals, as a result of which the contours of the rotating member in instruments of the last of the indicated designs will change rapidly, and with this the correctness of the readings will be impaired.

Nevertheless, some authors still resort, in determining the viscosity of certain suspensions, to the outflow method as the simplest, using as the viscometer a glass tube, for example, 400 mm long,

with a diameter of 28 mm, ending in a capillary 2 mm in diameter and 20 mm long [1]. The suspension in such a tube, mounted on a rotating disk, is first shaken by rotating the disk, and then the time of its outflow from the tube is observed; it is assumed here that the settling velocity of the largest grains of weighting agents does not exceed 0.4 cm/sec for an amount of suspension of 200 ml.

Such a method for determining the relative viscosity of unstable suspensions with a dispersion of 100–120 microns must be considered incorrect and unsuitable for production control.

In fact, the rate of outflow of a magnetite suspension with a dispersion on the order of 100 microns and a specific gravity of 1.5 through an orifice 2 mm in diameter is, at constant head, as found experimentally, 1.0–1.1 ml per second; 200 ml of suspension will flow out in an average time of 190 sec. Suppose that the settling velocity of the largest magnetite particles is 0.3 cm/sec (less than 0.4 cm/sec); after 95 sec approximately half the suspension will have flowed out (neglecting the decrease in outflow velocity as the liquid level in the tube falls), i.e., 100 ml. The height of a column of 200 ml of liquid in a tube 2.8 cm in diameter is 32.5 cm; consequently, the liquid level will drop by 16.25 cm, and during this time the particle, settling, will travel a distance equal to \(95 \times 0.3 = 28.5\) cm, i.e., settling of the particle proceeds ahead of the outflow velocity, and, consequently, as outflow proceeds the suspension becomes compacted, which distorts the results of the determination.

The American Shelton–de Vany instrument is a capillary viscometer with a stirrer [2]; it consists of a vertical tube 38 mm in diameter and 165 mm high, in which a stirrer is placed, making 900 rpm; at the bottom the tube passes into a capillary 2.6 mm in diameter, from which the suspension flows out. The following shortcomings are noted for this instrument: 1) the suspension enters the capillary insufficiently smoothly from the space located directly beneath the stirrer; 2) the liquid in the instrument becomes strongly heated during the experiment; 3) the largest particles, at the high rotational speed of the stirrer, are thrown toward the periphery, as a result of which the homogeneity of the suspension is disturbed and the outflow process is distorted; 4) the flow in the instrument is turbulent, which in turn also impairs the regularity of the outflow.

Testing, with the aid of this instrument, the quality of various unstable suspensions with respect to their viscosity in practice gave readings completely inconsistent with the phenomena that were observed for the same suspensions in a separator during coal beneficiation.

More perfect is the Ginzetmet instrument for determining the dynamic resistance to shear and the viscosity of suspensions [3].

However, working with it, which requires the selection of capillaries of various diameters and lengths and the construction of calibration graphs, is rather troublesome, which makes the method not entirely convenient for use in the process of production control.

In addition, the attempt to give the value of the absolute viscosity of a suspension in poises from the value of the dynamic resistance to shear determined by the outflow method is fundamentally incorrect, because suspensions cannot be identified with true liquids, for which the absolute viscosity, determined in capillary viscometers and expressed by Poiseuille’s formula, is, in physical meaning, the coefficient of internal friction of the liquid.

For suspensions, however, in addition to the friction of liquid (water) particles, there is mutual friction of solid particles, as well as friction between solid particles and liquid; therefore the recalculation of the relative viscosity, determined by the outflow method,

… readings, in centipoises, which is possible for true liquids by known formulas (Ubbelohde, etc.), cannot be done for suspensions.

Thus, the significance of the Ginzvetmet instrument should be assessed chiefly from the standpoint of determining the relative viscosity of suspensions, expressed, as usual, by the ratio of the times required for equal volumes of suspension and water to flow out through a tube with a channel of definite diameter (2.5–4.0 mm); for this purpose the instrument is quite suitable, but comparatively complex, since the installation requires continuous motion of the suspension, produced by means of a centrifugal pump.

The authors propose a method for the simultaneous determination of the specific gravity and relative viscosity of suspensions, which gave good results in testing.

The homogeneity and stability of the suspension are achieved in a trough-shaped bath 1 (Fig. 1), inside which a mesh drum 2, mounted on a shaft, rotates; the size of the mesh openings is 3 mm; the drum is open at the end faces. On the outer surface of the drum strips 3 are fastened, directed—

Fig. 1.

—along the generatrices of the cylinder; these six strips play the role of blades and serve to stir the suspension, which inside the drum, during its rotation, retains complete homogeneity. The circumferential speed required for this on the blades, as established experimentally, should be on the order of 0.1 m/sec.

For an instrument of the indicated dimensions the number of revolutions should be

\[ n=\frac{30\cdot 0.1}{3.14\cdot 0.028}=34 \text{ rpm}. \]

Into the bath are placed 200 ml of liquid. In one of the end walls of the bath there is a tube 4, opening into the interior of the drum; the tube is closed, as shown in Fig. 1, by a rubber stopper with a glass tube having an internal diameter of 2 mm.

This tube is connected by means of rubber tubing with a similar tube, which in turn connects the bath with a 100 ml measuring cylinder in the common setup for determining the viscosity and specific gravity of suspensions (Fig. 2).

Vessel 3 is filled with distilled water up to the mark from funnel 4, which serves as a pressure reservoir; then stopcock 7 is closed and stopcock 6 is opened, whereupon water from vessel 3 flows out into beaker 5; the pressure in the cylinder decreases, and the suspension is drawn into the cylinder. At the moment the jet begins to separate from the tube, the stopwatch is started and is stopped when the level of suspension in the cylinder reaches 50 ml; immediately thereafter stopcock 6 is closed; stopcock 8 serves for communication with the atmosphere.

Thus, the time of discharge of 50 ml of liquid is determined. Naturally, the experiment is first carried out with distilled water.

Fig. 2.

After determining the discharge time of the suspension, the cylinder, previously calibrated, is weighed to an accuracy of 0.1 g, and the volume of suspension is counted off; this volume is always somewhat greater than 50 ml, owing to the noncoincidence of the moments of stopping the stopwatch and the subsequent closing of valve 6.

Thus the specific gravity of the suspension is determined in the same experiment.

A series of determinations by this method gave the following figures.

Distilled water

\[ t = 20^\circ(\pm 2^\circ), \]

Time of discharge of 50 ml

\[ \begin{aligned} 1)&\ 30.5''\\ 2)&\ 30.2''\\ 3)&\ 30.0'' \end{aligned} \]

\[ \overline{\text{Mean}}.\ 30.43'' \]

\[ \Delta = \text{deviation from the mean value:} \]

\[ = \frac{(30.5 - 30.23)\,10}{30.23} = 0.89\% \]

Magnetite suspension.
Dispersion: (60–0)

\[ t = 20^\circ(\pm 2^\circ). \]

\[ \text{Sp. viscosity} = \frac{43.5}{30.23} = 1.44 \]

\[ \text{Sp. gravity} = \frac{82.90}{55.0} = 1.51 \]

\[ \text{Sp. gravity} = \frac{80.61}{54.0} = 1.49 \]

II

1) \(47.0''\)

2) \(46.5''\) \(\Delta = 0.53\%\)

Average \(46.75''\)

\[ \text{rel. viscosity}=\frac{46.75}{30.23}=1.55 \]

II.

1) \(55.0\ \text{ml}\) \(\qquad 96.26\ \text{g}\)

\[ \text{sp. gr.}=\frac{96.26}{55.0}=1.75 \]

2) \(56.0\ \text{ml}\) \(\qquad 97.04\ \text{g}\ \Delta=1.01\%\)

\[ \text{sp. gr.}=\frac{97.04}{56.0}=1.73. \]

Thus, the discrepancy in parallel experiments when determinations are made by the method described is, for the average figures, for viscosity: \(0.53\text{–}1.15\%\), and for specific gravity: \(0.67\text{–}1.01\%\), which may be considered sufficient for practical purposes.

The magnitude of the viscosity of suspensions, determined by the efflux method, as already noted, has no definite physical meaning and should be regarded as a conventional quantity, which nevertheless is of great importance as one of the parameters determining the character of suspensions in production control during the beneficiation of minerals.

REFERENCES

1. Zemlyanskii P. M. Coal Beneficiation in Heavy Media. Ugletekhizdat, 1953.
2. Handbook on the Beneficiation of Minerals. Vol. III, Metallurgizdat, 1952.
3. Mitrofanov S. I. Investigation of Ores for Beneficiability, Metallurgizdat, 1950.