Showing papers in "Metallurgical and Materials Transactions B-process Metallurgy and Materials Processing Science in 1977"

Crack formation in the continuous casting of steel

Abstract: This review examines the different types of internal and surface cracks that can form during the continuous casting of steel. For each crack type, the operating and metallurgical factors that are known to influence crack formation are assessed in the light of the high temperature mechanical properties of steel and a knowledge of the stresses generated in the solidifying shell. The importance of two zones of low ductility in steel is demonstrated by this approach. One zone exists above 1340°C and probably accounts for the formation of all internal cracks and surface longitudinal cracks. The other zone lies between 700 and 900°C and is related to the presence of soluble aluminum, niobium and vanadium. Transverse surface cracks in slabs can be related to the latter zone.

The rate of reduction of iron oxides by carbon

Abstract: The rate of reduction of Fe2O3 and FeO by coconut charcoal, coal char and coke, in an inert atmosphere within the temperature range 900 to 1200°C was investigated. The effects of pressure, particle size, and the amount of carbon were determined. The results indicate that the reaction takes place by means of the gaseous intermediates CO and CO2, and that the overall rate is controlled by the oxidation of the carbon by CO2. The rates of reduction of FeO and Fe2O3 by CO are relatively fast, and the CO2/CO ratio for the oxidation of carbon is determined by their equilibria. The reduction of Fe2O3 by carbon is accomplished in two stages, with FeO forming first. The reduction of Fe2O3 to FeO is faster than that of FeO to Fe because its CO2/CO equilibrium ratio is higher and hence the rate of oxidation of carbon is faster. A direct comparison was made between the rate constants for the reduction of FeO by carbon and those for the oxidation of carbon in the appropriate CO2-CO gas mixtures, and they are in good agreement. Apparently the iron formed by the reduction does not significantly catalyze the oxidation of carbon; whereas for the reduction of NiO by carbon, the Ni formed does catalyze the oxidation of carbon.

Studies on the reduction of hematite by carbon

Abstract: Single pellet experiments have been carried out in a nitrogen atmosphere to study the reduction of hematite by graphite in the temperature range 925 to 1060°C. The effect of variables such as c/Fe2O3 molar ratio, pellet size, and so forth, has been investigated. Gas analysis data show a continuous decrease in CO2/CO ratio during reduction, the values being far away from Fe/FeO equilibrium for wustite reduction by CO. The activation energies associated with different degrees of reduction appear to be widely different suggesting a possible changeover in reaction mechanism during the progress of reduction. X-ray diffraction studies confirm the stepwise nature of hematite reduction.

The structure of molten binary silicate systems CaO-SiO 2 and MgO-SiO 2

Abstract: The structure of the molten binary CaO-SiO2 and MgO-SiO2 was investigated by means of a high temperature X-ray diffraction technique. From an analysis of the X-ray scattering intensity, the radial distribution function was evaluated from which interionic distances and coordination numbers were calculated. The following features emerge from these results: Molten silicates consist mainly of SiO4 tetrahedral units, up to 57 mol pct alkaline earth Met.al oxides. The number of oxygens which surround Ca and Mg at the monosilicate composition is 7 and 5 respectively. The structure of molten silicates is insensitive to temperature within the present experimental temperature range. The ionic distribution and structural changes in the molten mixture of alkaline earth silicates are also discussed.

Oxidation of aluminum-magnesium melts in air, oxygen, flue gas, and carbon dioxide

Abstract: Oxidation rates of aluminum—1 to 14 pct magnesium melts in air, oxygen, flue gas, and carbon dioxide at temperatures from 600 to 1100°C were measured with an automatic recording balance. For most conditions, a protective amorphous film on the melt surface kept the oxidation rate low initially. After an interval that was shortened by increasing temperature or magnesium content, the film crystallized to magnesium oxide and magnesium aluminate, accompanied by a sudden increase in oxidation rate. Known as breakaway oxidation, this phenomenon could be produced by adding crystalline magnesium oxide or magnesium aluminate seed to melts protected with amorphous oxide. Flue gas or carbon dioxide in the atmosphere, sodium or beryllium in the alloy, or boron dusted on the surface delayed crystallization unless seed was added. Beryllium was the most effective. Flue gas from burning natural gas delayed breakaway oxidation of unseeded melts containing up to about 4 pct magnesium at normal melting furnace temperatures near 750°C. Slow melting of solid aluminum-magnesium alloy concentrated magnesium in the first liquid to melt, sharply decreasing the interval before breakaway oxidation after melting. Rapidly melting the solid increased the protective interval. Homogenizing the solid just below the melting range before rapid melting further extended the interval before the onset of breakaway oxidation.

Phase stability in ternary systems of transition elements with aluminum

Abstract: The results of the present study are consistent with the conclusion that /gb-Mn is an electron phase. The absence of the extended β-Mn and σ phases in the Mn-Ni-Al and Mn-V-Al systems, respectively, which can not be accounted for either by electron concentration or by atomic size, is, as suggested by the location and shape of the phase fields, due to the relatively strong stabilization of the bcc phase in these ternary systems. In all the four systems studied, pronounced stabilization of the bcc and the related CsCl type phase is observed. It appears from the above study that Al plays a significant role in the stabilization of the bcc phase in the transition element alloys.

Instabilities of the metal surface in electrolytic alumina reduction cells

Abstract: Transient waves on the metal surface in electrolytic cells are studied by simulation on a computer model. The model used is a semidynamic modification of a model for stationary flow and surface calculation. The simulation results clearly show that any wave once started by some disturbance, will turn into a transient wave rotating along the edge of the cell cavity, resulting in a tilting movement of the whole metal surface. The mechanism of the rotating wave is explained. It is further demonstrated that there exists a stability limit above which the rotating waves are amplified instead of being damped, leading to an unstable situation known as “shaky pot”. A simple empirical stability formula is presented, showing the interrelation of the decisive cell parameters. General estimation curves for typical cells are also given.

High voltage microscopy of the reduction of hematite to magnetite

Abstract: The microstructural changes associated with the formation of magnetite in hematite have been studied in specimens which have been partially reduced outside the microscope, thinned until electron transparent, and then examined in the normal way. Three types of structure have been observed in varying proportions which depend on the reduction temperature. At low temperatures, magnetite grows by the propagation of a cellular interface, the gas phase being transported to the cell boundary by a network of tunnels. At intermediate temperatures, magnetite plates are formed, whereas at high temperatures, both plate magnetite and blocky magnetite appear. It is proposed that the factor controlling the morphology which develops is the ratio of the cell boundary diffusivity to volume diffusivity of ferrous ions. It is noted that the decomposition morphologies of hematite and austenite have many similarities. The basic reason for this similarity is that both transformations involve substantial redistribution of elements in the solid state and the microstructures which develop are those that perform this redistribution the most efficiently at the temperatures involved.

A mathematical model of slag and metal flow in the ESR Process

Abstract: Through the statement of the turbulent Navier-Stokes equations and Maxwell’s equations a mathematical representation is developed for the electromagnetic force field and the velocity field in the slag phase and the metal pool of cylindrical ESR units. Computed results are presented for both industrial scale (0.5 m electrode diameter) and laboratory scale (0.05 m electrode diameter) units operating with direct currents. It was found that for industrial scale units, the computed slag velocities ranged from 5 to 10 cm/s, while the velocities in the metal pool were substantially lower, except at the slag-metal interface. At a given spatial position, the velocity was found to increase in an almost linear fashion with the current density. The flow was found to be predominately laminar in the laboratory scale units and for comparable current densities the melt velocities were very much smaller. Some 600 to 900 s were required on a CDC 6400 digital computer for the solution of each case involving turbulent flow.

Nitrogen solution and titanium nitride precipitation in liquid Fe-Cr-Ni alloys

Abstract: The solubility of nitrogen in liquid Fe-Cr, Fe-Ni, Ni-Cr, and Fe-Cr-Ni alloys up to 20 wt pct Ni and 40 wt pct Cr was measured by the Sieverts’ method The first and second order interactions in iron between nitrogen and chromium, and nitrogen and nickel were determined Chromium increases the nitrogen solubility at lower chromium concentrations but the second order interaction term which is of the opposite sign becomes significant at higher chromium levels and compensates partly for the effect of the first order interaction term Nickel decreases the nitrogen solubility in iron Titanium nitride formation in liquid Fe-Cr, Fe-Ni, and Fe-Cr-Ni alloys also was investigated The first and second order interactions between titanium and chromium or nickel were determined at 1600°C Chromium increases the solubility product of TiN, principally by decreasing the activity of nitrogen in the melt Nickel decreases the solubility product of TiN by increasing the activities of nitrogen and titanium

Ammonia, oxidation leaching of chalcopyrite —reaction kinetics

Abstract: The reaction for the ammonia, oxidation leaching of chalcopyrite, CuFeS2 + 4NH3 + 17/4 O2 + 2 OH- ⇌ Cu(NH3)+2 42 + l/2Fe2O3 + 2 SO4 + H2O was studied using monosize particles in an intensely stirred reactor under moderate pressures to determine the important chemical factors which govern the kinetic response of the system. The reaction kinetics were studied at dilute solid phase concentration so that oxygen transport at the gas/liquid interface would not limit the rate. A catalytic electrochemical surface reaction was shown to control the reaction kinetics with the reaction rate determined by the following equation derived from electrochemical considerations: dα/dl=127 f/do (OH- 1/2 (k1PO1/1+k2PO1 (k1+k2(Cu+2)o+k’2α)(1-α)2/3 (K P 1/2 Excellent agreement between theory and experiment was obtained both with regard to apparent reaction orders for oxygen, cupric, and hydroxyl, and with regard to geometric factors that influence the reaction rate. Further support for the reaction mechanism included an activation energy of approximately 10 kcal/mole obtained under a variety of experimental conditions and the fact that the initial reaction rate constant was several orders of magnitude less than predicted mass transfer coefficients.

Activities in the spinel solid solution, phase equilibria and thermodynamic properties of ternary phases in the system Cu-Fe-0

Abstract: A review of the structural and thermodynamic information and phase equilibria in the Cu-Fe-O system suggested that a consistent, quantitative description of the system is hampered by lack of data on activities in the spinel solid solution CuFe2O4-Fe3O4. Therefore the activity of Fe3O4 in this solid solution is derived from measurements of the oxygen potentials established at 1000°C by mixtures containing Fe2O3 and spinel solid solutions of known composition. The oxygen pressures were measured manometrically for solid solutions rich in CuFe2O4, while for Fe3O4-rich compositions the oxygen potentials were obtained by an emf technique. The activities show significant negative deviations from Raoult’s law. The compositions of the spinel solid solutions in equilibrium with CuO + CuFeO2 and Cu + CuFeO2 were obtained from chemical analysis of the solid solution after magnetic separation. The oxygen potential of the three-phase mixture Cu + CuFeO2 + Fe3O4(spinel s.s.) was determined by a solid oxide galvanic cell. From these measurements a complete phase diagram and consistent thermodynamic data on the ternary condensed phases, CuFeO2 and CuFeO2O4, were obtained. An analysis of the free energy of mixing of the spinel solid solution furnished information on the distribution of cations and their valencies between the tetrahedral and octahedral sites of the spinel lattice, which is consistent with X-ray diffraction, magnetic and Seebeck coefficient measurements.

Oxidation kinetics of iron alloy drops in oxidizing slags

Abstract: The kinetics of reactions between drops of Fe-C, Fe-C-P and Fe-C-S alloys and synthetic oxidizing slags at 1550°C have been studied. The reaction kinetics are obtained either from chemical analysis of quenched samples, or, for decarburization, * by continuous measurement of the flow rate of evolved gas. The marked effect, on the kinetics, of the metal emulsification during carbon removal is evidenced. A decrease in slag oxygen potential results in the slow-down of decarburization. Sulfur shows a tendency to limit interfacial reactions, but it enhances metal emulsification. Both mass transfer processes and interfacial reactions, as well as nucleation, are believed to play their part on the rate of decarburization.

The electrical conductivity of fluidized bed electrodes—its significance and some experimental measurements

Abstract: By applying a mathematical model (developed by Newman and Tobias for porous electrodes) to a fluidized bed electrode it can be concluded that the effective electrical conductivity of the bed is a key parameter affecting bed performance. Poor current and power efficiencies and operational difficulties are predicted for beds with low electrical conductivities and it is suggested that high conductivities may also cause operational difficulties. An apparatus for measuring bed conductivities is described and some results are presented for beds of copper particles in the size range 270 μ. to 385 μ. Bed conductivity was relatively insensitive to particle size and electrolyte conductivity but decreased sharply as the bed expanded. From the results it was possible to infer that the dominant mechanism of electrical conduction in the fluidized bed, under the conditions studied, was through “chains” of particles in momentary electrical contact.

Evaluation of organic additives for use in zinc electrowinning

Abstract: Organic additives are used extensively in zinc electrowinning to assist in controlling the process. A cyclic voltammetry technique has been developed to provide a rapid, quantitative evaluation of the effectiveness of selected organic additives in minimizing the deleterious effects that impurities, such as antimony, have on zinc deposition. Results have indicated that animal glues are more effective than the other organic additives tested, which included several gums, enzymes, and amino acids, in relation to the current efficiency of zinc production. Of the various animal glues compared in this research, the most effective appeared to have average molecular weights in the 25,000 to 30,000 range. The effects of certain process variables on the test results have also been evaluated; these included the acid concentration of the zinc sulfate solution and the cathode preparation.

Effect of fluorspar and other fluxes on slag-metal equilibria involving phosphorus and sulfur

Abstract: The effect of additions of fluorspar-and potential substitutes such as borate, manganese ore, and ilmenite—to basic steelmaking slags on the equilibrium phosphorus and sulfur distribution ratio has been studied at 1550δC. From laboratory slag-metal equilibrium studies it was found that the presence of CaF2 in concentrations between 1.5 and 5 wt pct in the slag (equivalent to 2 and 6 kg of fluorspar per ton of steel) lowered the equilibrium phosphorus content in the metal by about a factor of three. The use of B2O3, manganese ore, or ilmenite (the active ingredient in Sorel flux) did not affect the equilibrium phosphorus content of the metal. The effect of CaF2 is ascribed to a lowering of the activity of P2O5 in the slag. Good agreement was found between the metal phosphorus content obtained from open-hearth plant data (no fluorspar added) and the present laboratory data for slags not containing CaF2. None of the fluxes studied in this work, including CaF2, improved the equilibrium sulfur distribution between metal and slag. The present results indicate that the choice of a substitute for fluorspar in an actual steelmaking operation, except when phosphorus presents a problem, should be based primarily on the cost of using the fluorspar substitute, including any change in refractory costs resulting from using the substitute.

Distribution equilibria of Sn, Se and Te between FeO-Fe2O3-SiO2-AI2O3-CuO0.5 slag and metallic copper

Abstract: The distribution of tin, selenium and tellurium between alumina-containing fayalitic slags and metallic copper was measured at 1200 and 1300°C under controlled CO-CO2 atmosphere with oxygen partial pressure (pO2) in the rangePO2 = 10-6 to 10-11 atm (1 atm = 1.013 x 102 kPa). The solubility of Sn in slag was observed to increase linearly with increasing P1/2O2. It was deduced that Sn is present in the slag in the form of SnO or Sn2 and the activity coefficient of SnO in the slag was calculated to be 1.9 at 1200°C and 0.8 at 1300°C. The solubility of Se in the slag decreases with increasing oxygen partial pressure up topO2l = 4 x 10−8 atm, but above this oxygen partial pressure it becomes practically constant and the ratio (pet Se in slag/pet Se in copper) = 0.018 (at 1200°C) and 0.036 (at 1300°C). The solubility of Te shows a similar variation with oxygen partial pressure and the ratio (pet Te in slag/pet Te in copper) = 0.026 (at 1200°C) and 0.032 (at 1300°C) abovepol2 = 106 atm. A concept of molecular dissolution of chalcogen elements in slag was developed on the basis of thermodynamic properties of slag, and the observed solubility of Se and Te was explained in terms of the chemical stability of the molecular cluster FeSe and FeTe iη the slag.

Activities and free energies of mixing in binary silicate melts

Abstract: The thermodynamic properties of the Whiteway, Smith and Masson (WSM) models of polymerized silicate melts are calculated. The free energy is calculated as a function of composition and degree of polymerization of the silicate ions from heats of mixing and Guggenheim’s expression for the configurational entropy of a system of linear and branched chain polymer molecules in which random mixing of the mer units occurs. The variations, with composition, of the equilibrium degree of polymerization, the activities of solution components and the free energy of mixing are calculated as functions of the stability of the solid orthosilicate compound. Under the assumption of random mixing of the mer units, the thermodynamic properties are found to be independent of whether the SiO in4 su4 ion is considered to be a bifunctional or tetrafunctional polymerizing unit. It is found that the apparent differences between the dependencies, on composition, of the ion fraction of O2- in the WSM linear chain model and the WSM branching chain model are the result of the assumption, in these models, that random mixing of the anions occurs,i.e. that the systems exhibit Temkin behavior. It is suggested that the deviations of the theoretical free energies of mixing from experimental measurement, which begin when the mole fraction of silica is between 0.3 and 0.4, are due to the first significant appearance of ring ions in actual silicate melts at some composition within this range.

The hydrometallurgical treatment of zinc silicate ores

Abstract: A novel hydrometallurgical procedure is described for the recovery of zinc from zinc silicate ores which have high acid-soluble zinc and silica contents. The process is conducted in a continuous concurrent manner at atmospheric pressure and temperatures of 50 to 95°C. The ore is leached with spent electrolyte from an electrolytic zinc plant to a final pH of approximately 2 to dissolve zinc and the soluble silica. The pH of the leach solution is then raised to 4 to 5.5, using a neutralizing agent, to precipitate and coagulate the colloidal silica. Finally the coagulated silica is filtered from the solution and washed. The resulting filtrate is treated conventionally for electrolytic zinc production. This process solves in a simple way the difficult problem of precipitating large concentrations of colloidal silica(e.g. 25 g SiO2 per liter) in a readily filterable and easily washed form. The process has been tested at up to the 5 tons per day scale on ores containing willemite and on a laboratory scale with ore containing hemimorphite and smith-sonite.

Fluid velocities in induction melting furnaces: Part II. large scale measurements and predictions

Abstract: Measurements of melt surface velocities on a twelve ton Ajax induction furnace were compared with the predictions of a mathematical model for melt flow in induction furnaces. Agreement was found to be fairly good. The results of calculations of the effect of changing power supply frequency, coil and melt geometry, current and phasing are presented.

Sulfuric acid pressure leaching of nickeliferous limonites

Abstract: This laboratory study examines the sulfuric acid pressure leaching characteristics of nickeliferous limonite samples from South New Caledonia. The effect of temperature (225°C to 300°C), acid-to-ore ratio (0.21 to 0.30), pct solids (10 to 45), particle size (−20 to −270 mesh), and agitation (50 to 600 rpm) on nickel extraction and selectivity are discussed. For overall improvement of the leach response, leaching is best conducted at temperatures higher than 250°C and at solids less than 33 pct. Variation of goethite, hematite, and basic aluminum sulfate content of the leach residue was determined semiquantitatively as a function of leaching time. Sulfate content of the residue rises sharply during the first several minutes of leaching, then falls, and finally equilibrates at five to six percent. This change in sulfate analysis can be explained by assuming a series of consecutive iron and aluminum hydrolysis reactions.

Mathematical model of the SL/RN direct reduction process

Abstract: A mathematical model has been developed to predict the operating behavior of an SL/RN direct reduction kiln from a knowledge of the main process variables. The model is based on steady state principles and is capable of quantitatively describing the different chemical reactions in the kiln such as reduction, Boudouard reaction, coal devolatilization, combustion in the freeboard together with the mass and heat flows. Results of the model, which are in the form of axial temperature profiles in the freeboard gas, solids bed and wall, and axial concentration profiles in the gas and solids, are in good agreement with measurements made on a 200 ton per day pilot plant at the Steel Company of Canada. The model has been used both to investigate the influence of process variables on the kiln performance and to elucidate operating features of the process. The model has shown that highly reactive coals such as subbituminous and lignite, and highly reducible pellets are most suitable for the SL/RN process. It has also demonstrated that the most important process variable from the standpoint of control is the air profile in the kiln; and consequently that heat transfer is the rate controlling step. The model has also been employed to examine scale-up of the SL/RN process by calculating the operating behavior of a large commercial kiln. It has shown that the large kiln should run cooler than the pilot-size unit so that accretion problems may be significantly reduced.

Solubility of nitrogen and carbon in CaO-Al2O3 melts in the presence of graphite

Abstract: CaO-Al2O3 slags were melted in graphite crucibles under N2–CO–Ar gas mixtures at 1600°C. The contents of total nitrogen, cyanide and total carbon of the slags were determined by chemical analyses of quenched samples taken by suction from the melt. The nitrogen is present in the melt as nitride N−3 ion and cyanide CN−1 ion, and carbon as cyanide and carbide C 2 2− ion. The equilibrium constants for the respective reactions were evaluated. It is found that the nitride capacity of the melt decreases whereas the cyanide and carbide capacities increase with increasing CaO/Al2O3 ratio.

The measurement and prediction of the melt velocities in aturbulent,electromagnetically driven recirculating low melting alloy system

Abstract: Experimental measurements are reported describing the velocity field in an inductively stirred low melting alloy. The molten metal was held in a cylindrical container, 500 mm high and having an inside diameter of 250 mm; induction stirring was supplied by a three phase coil, which provided a maximum field strength of 350 Gauss (0.035 Wb/m2). The velocities in the melt were measured by a mechanical force reaction probe and were found to range up to about 0.5 m/s. Theoretical prediction of the melt velocities was made by solving Maxwell’s equations, together with the turbulent Navier-Stokes equations, using a digital computer. The experimentally measured and theoretically predicted velocities were found to agree within about 30 pct, thus providing direct experimental proof for the validity of modelling electromagnetically driven flows using this technique.

Die thermal behavior in machine casting of partially solid high temperature alloys

Abstract: Experimental measurements and computer simulations of die thermal behavior during machine (die) casting of fully liquid and partially solid bronze alloy 905 were carried out. Ingots of the alloy were heated to temperatures above the liquidus and in the liquidsolid range in a reheat furnace. The partially solidified charge was previously made in a continuous slurry producer. Castings were made in both a low pressure laboratory machine and a high pressure commercial die casting machine. In both casting machines used, the maximum die temperature and the initial rate of temperature change in the die, at ∼0.014 in. from the casting-die interface, were significantly lower when a partially solid charge material was used. For example, measured die temperatures in the high pressure commercial machine were 475°C and 165°C above the initial die temperature when the bronze alloy was cast in the 100°C superheated state and the partially solid (volume fraction solid ∼0.57) state, respectively. Correlation of computer predictions and measured die temperatures have been used to calculate values of the heat transfer coefficient at the casting-die interface. Using these values, the maximum die surface temperature and die surface temperature gradients are calculated. The values of maximum surface temperature obtained are 800°C and 315°C for a superheated (100°C) liquid charge and a partially solid (volume fraction solid ∼0.57) charge, respectively. The corresponding surface temperature gradients are 5640°C/cm and 718°C/cm, respectively. The reduced thermal shock experienced by the die when a partially solid metal alloy is cast should improve die life considerably over conventional practice.

Solubility of nitrogen in liquid Fe-Cr-Ni alloys containing manganese and molybdenum

Abstract: The nitrogen solubility in liquid Fe-Cr-Ni alloys containing Mo or Mn was determined by the Sieverts’ method. The first and second order mutual interactions among nitrogen, chromium, nickel, molybdenum, and manganese in iron were determined as a function of temperature. The heat and entropy of solution in these alloys were correlated as functions of the logarithm of the activity coefficient of nitrogen at {dy1873} K independent of the composition of the alloys. An equation was derived to predict the nitrogen solubility in liquid multicomponent iron alloys for the range from logfN, 1873K = 0 to -1.4 as, log (wt pct N)T = (-247/T - 1.22) - (4780JT - 1.51) (logfn, 1873K)-(1760/T -0.91) (logfN,{dy1873}K )2.

Ammonia, oxidation leaching of chalcopyrite —surface deposit effects

Abstract: The reaction for the ammonia, oxidation leaching of chalcopyrite, CuFeS2 + 4NH3 + 17/4 O2 + 2 OH- = Cu(NH3)4 +2 + l/2Fe2O3 + 2 SO4 + H2O was studied using monosize particles in an intensely stirred reactor at moderate oxygen pressures. For dilute solids concentration, the rate is controlled by an electrochemical surface reaction. Under conditions of low stirring speeds and low oxygen pressure, the hematite reaction product passivates the surface and the reaction virtually stops. Even though stirring speed influences the rate of the electrochemical reaction, this effect is due to changes in the morphology of the hematite deposit which alters the surface reaction kinetics, rather than being indicative of mass transfer limitations.

Determination of sodium in molten aluminum and aluminum alloys using a beta alumina probe

Abstract: The development of a probe using beta alumina as an electrolyte and a reference material consisting of a mixture of α-alumina and β-alumina is presented together with a description of its application to the control of the sodium modification of aluminum silicon alloys.