A mixture of molten aluminum alloy and aluminum crystals is present, for example, in the fractional crystallization of contaminated aluminum. Fractional crystallization is a known method for purifying a contaminated metal alloy and is described, inter alia, in US patent 3,840, 364. One of the forms of fractional crystallization is suspension crystallization. In suspension crystallization, the contaminated metal alloy is cooled slowly from the melt. As solidification begins, very pure metal crystallizes out, and as cooling continues metal which is progressively less pure crystallizes out. By separating the crystals from the uncrystallized liquid in the mixture, metal crystals consisting of an alloy with a significantly higher purity than the original metal alloy are obtained. In addition, suspension crystallization can also be used for a more highly contaminated metal, in which case the impure metal crystallizes out first.

One problem with carrying out this method is that the proportion of crystals in the mixture is difficult to control. However, this is necessary in order to allow industrial application of the method. In the case of suspension crystallization, measurement of the temperature of the mixture cannot be used or can scarcely be used to control the proportion of crystals. On the one hand, the temperature which has to be measured is high, approximately 700°C for aluminum, making the temperature difficult to measure with accuracy, and on the other hand pure aluminum will crystallize at one fixed temperature and therefore, at this temperature, may be either completely liquid or completely crystallized, as well as every possible state in between. Therefore, temperature measurement cannot be used to control the proportion of crystals in the mixture; in the event of excessive cooling, the mixture may solidify completely, without this being controllable with the aid of the temperature measurement. If the aluminum is slightly contaminated, there is a temperature difference between the completely liquid state and the completely crystallized state, but this temperature difference is very small and depends on the impurities present, which are not usually known with accuracy.

Therefore, measurement of the temperature cannot be used in the industrial suspension crystallization of contaminated aluminum to keep the proportion of crystals in the mixture constant within a reasonable margin.

It is an object of the invention to provide a method with which the proportion of aluminum crystals in a mixture of molten aluminum alloy and aluminum crystals can be controlled during, for example, fractional crystallization.

Another object of the invention is to provide a method of this type, which can be used accurately and on an industrial scale.

Yet another object of the invention is to provide a method of this type, which is simple and reliable.

In addition, it is an object of the invention to provide a device with which the method can be carried out.

One or more of these objects are achieved with a method for controlling the proportion of aluminum crystals in a mixture of molten aluminum alloy and aluminum crystals in which the electrical resistance of the mixture is determined with the aid of a four-point measurement.

The electrical resistance measurement makes it possible to accurately control the proportion of crystals in the melt, even if the percentage of crystals is not known, since the crystals have a significantly lower resistance than the melt (factor of 2.2).

With this method according to the invention, it is possible to keep the proportion of crystals in the melt constant with sufficient accuracy for industrial application without it being known precisely which impurities are present and without any knowledge of phase diagrams.

However, the resistance in molten aluminum is low and has to be measured very accurately. Therefore, the resistance is measured with the aid of a four-point measurement in which a current is passed through the mixture between two points and the voltage in the voltage field between the two current-carrying point is measured between two separate voltage-measuring points. The two voltage- measuring points generate a negligible current, so that measurement of contact resistance is avoided. The current intensity which has to be used between the current-carrying points is at least 5 amperes in order to allow a voltage of a few tenths of a millivolt to be measured over a path of 40 mm.

The four-point measurement can be carried out completely independently of any temperature measurement.

The mixture is preferably stirred continuously, in order to keep a constant ratio between crystals and melt throughout the mixture and in order to counteract as far as possible the growth of crystals on the walls of the mixing vessel in which the measurement is carried out.

The method is preferably carried out during the fractional crystallization of contaminated aluminum. Accurate control of the proportion of crystals in the mixture is very important in the fractional crystallization of contaminated aluminum.

The method according to the invention can be carried out in the batchwise fractional crystallization of aluminum in order to separate the mixture as soon as a defined percentage of crystals is obtained.

However, the method is preferably used if the crystallization is carried out as continuous crystallization. In the case of continuous fractional crystallization, molten contaminated aluminum is supplied continuously, and a mixture of crystals and molten aluminum alloy is discharged continuously. For this purpose, the mixture has to be cooled to a greater or lesser extent, inter alia as a function of the temperature of the contaminated aluminum supplied. The cooling of the mixture then has to be controlled by the four-point measurement.

The electrical resistance is preferably kept constant in continuous crystallization.

The proportion of crystals in the mixture discharged is then also constant and can be set to an optimum value. To keep the resistance constant, by way of example the cooling and/or supply and discharge of the aluminum can be regulated.

The proportion of aluminum crystals is preferably kept constant within a margin of plus or minus 4%, more preferably within a margin of plus or minus 2%. A margin of this nature means that in a subsequent process the mixture can easily be separated into crystals and molten aluminum.

The method is preferably used on aluminum in which the contamination consists at least in part of Fe. Particularly in the case of AlFe, the eutectic melting point is just below the melting point of pure aluminum, so that accurate control of the process and therefore accurate measurement are required.

The invention also relates to a device for controlling the proportion of aluminum crystals with the aid of the method as described above, comprising a vessel for

holding the mixture, a means for regulating the temperature of the contents of the vessel, and a four-point ohmmeter with two electrodes for current to pass through and having two electrodes for measuring voltage, which electrodes are preferably surrounded by a protective tube.

The method according to the invention is therefore carried out using a known vessel with temperature control, in which a four-point measurement is carried out using a four-point ohmmeter configuration which is known per se and in which, however, the current-carrying electrodes are preferably at least partially surrounded by a protective tube. Using the protective tube protects the metal electrode from being dissolved in the melt and also leads to the current being fed into and out of the melt at the correct locations.

The protective tubes preferably consist of ceramic material, more preferably A1203. Ceramic material is able to withstand molten aluminum, and A1203 is a readily available and relatively inexpensive material.

According to an advantageous embodiment, the current-carrying and voltage- measuring electrodes in the protective tubes contain molten aluminum during use.

The use of molten electrodes means that there will be no oxide layer between the electrodes and the melt, so that the contact resistance is low.

According to an advantageous embodiment, the current-carrying electrodes are positioned at a distance from the walls of the vessel, which is at least equal to half the distance between the current-carrying electrodes. As a result, in the event of any growth of crystals on the inner wall of the vessel, the current will nevertheless pass almost completely through the mixture between the current-carrying electrodes, and will not pass or will scarcely pass through the crystals on the vessel wall.

It is preferable for each of the voltage-measuring electrodes to be positioned at a distance of at least 5 mm from the associated current-carrying electrode. Any growth of crystals on the current-carrying electrodes then has little or no effect on the voltage measurement via the voltage-measuring electrodes.

The invention will be explained on the basis of an exemplary embodiment and with reference to the appended drawing, in which: Figure 1 provides a highly diagrammatic illustration of an exemplary embodiment

of the device according to the invention.

Figure 2 provides a highly diagrammatic illustration of the current circuit of the four-point measurement according to the invention.

Figure 1 provides a highly diagrammatic illustration of an embodiment of a device 1 for controlling the percentage of crystals in a mixture of a molten aluminum alloy and aluminum crystals during the fractional crystallization of contaminated aluminum. The device 1 comprises a vessel 2 for holding the mixture 3, protective tubes 4,5, 6 and 7 for electrodes for carrying out a four-point measurement, and equipment for regulating the temperature of the mixture (not shown) and stirring equipment (not shown). Obviously, an inlet and outlet for the mixture, insulation materials and the like may be fitted to the vessel, as is known to a person skilled in this field.

The protective tubes 4,5, 6 and 7 consist of ceramic material and each have a current-carrying electrode, Il and I2, or a voltage-measuring electrode, Ul and U2.

The current-carrying and voltage-measuring electrodes are illustrated in Figure 1 as solid electrodes made from aluminum wire with a diameter of 2 mm, which project beyond the ceramic material.

During use, the current will flow from current-carrying electrode Il to current carrying electrode I2. The voltage-measuring electrodes Ul and U2 are positioned between the current-carrying electrodes Il and I2, since that is where the voltage difference is greatest. To ensure that the current will not or will scarcely flow via (crystals on) the wall of the vessel, the electrodes must be positioned at a distance from the walls of the vessel, which is at least half the distance between the current- carrying electrodes. The distance a between the voltage-measuring electrodes is, for example, approximately 50 mm. The distance b between the current-carrying electrode and the associated voltage-measuring electrode is, for example, 5 mm.

The possibility of the electrodes being partially melted is not shown. In that case, the electrode which in the solid state projects out of the protective tube will be at least partially melted at its end and will form a hollow meniscus in the corresponding protective tube. Since the electrode is (partially) melted at its end, the contact resistance between the mixture 3 and the electrodes will be very low, and consequently the four-point measurement will have a greater accuracy.

Figure 2 provides a highly diagrammatic illustration of how the four-point

measurement is carried out. A current I from a current source is successively passed through a reference resistance R1 and through the melt. The melt is obviously also a resistance, denoted by R2. The voltage is measured across the two resistances R1 and R2. The current I has to be measured constantly with a very high level of accuracy. The resistance R2 can then be determined on the basis of the measured values for the voltage across Rl and R2 and the known value for the reference resistance R1.

Obviously, calculation and control equipment will have to be present in order to measure and process the current and voltages and, on the basis of this information, to control the temperature-regulating equipment.

The method according to the invention will now be described with reference to Figure 1.

A slightly contaminated aluminum is introduced into the vessel 2 and heated until it has completely melted. The protective tubes 4,5, 6 and 7 with the current- carrying electrodes Il and I2 and the voltage-measuring electrodes Ul and U2 are placed into the molten aluminum. It is preferable to wait until the ends of the electrodes have (partly) melted. Then, a current with an intensity of, for example, 10A is passed through the melt, in accordance with the arrangement shown in Figure 2. Then the melt is slowly cooled with the aid of the temperature-regulating equipment. At a given moment, crystals will precipitate in the melt, which crystals will be distributed as uniformly as possible through the melt by the stirring equipment. The crystals have a composition, which is purer than the melt. Since the crystals have a lower resistance than the melt, the resistance of the mixture will drop. In theory, the percentage of crystals present in the melt can be determined on the basis of the voltage between the voltage-measuring electrodes Ul and U2, which is measured by means of the four-point measurement, and a known composition of the contaminated aluminum. In practice, it will be necessary to determine, on the basis of experiments for a defined arrangement of specific electrodes, what percentage of crystals is present at what voltage.

The above text provides a description of a batch process. In practice, the method will preferably be carried out as a continuous process, in which completely melted contaminated aluminum is supplied continuously and a mixture of (more intensively contaminated) molten aluminum and crystals is discharged continuously. In this case, it is important, above all, to keep the resistance in the mixture constant by means of cooling or heating, so that the percentage of crystals which is discharged is constant. The percentage of crystals which is obtained can be adjusted slowly in a continuous process.