Numerical Modeling of Reacting Liquid Glass Layer in an Advanced Glass Melter

In the advanced glass melter small particles of batch materials are rapidly heated to glass formation temperatures while entrained in gas, and then deposited on a target surface. The resulting liquid slurry, consisting of sand, CaO, and MgO particles in the liquid formed by mixing of molten cullet and Na₂CO₃, flows down the center body. During this flow the solid particles dissolve in the liquid and diffuse to form glass. This paper describes a numerical model developed to describe the behavior of this liquid layer. The model is based on Navier-Stokes equations reduced for the case of a low Reynolds number liquid layer flowing under the action of gravity and shear from gases flowing on top. Experimental results agree with model predictions.     

Thermal Analysis of Reactions in Soda–Lime Silicate Glass Batches Containing Melting Accelerants: II, Multicomponent Systems

The glass melting reactions in a multicomponent system (sand–soda ash–calcite–dolomite–feldspar) were studied using data from DTA, TGA, and XRD interactively. The first-formed liquid phase occurred at 700°C from eutectic melting among CaCO₃, Na₂CO₃, and MgO. Further liquid phase formed at the CaCO₃–Na₂CO₃, eutectic at 785°C and a fusion reaction among SiO₂, CaO, and the molten phase at 812°C. Reactions between molten soda ash and silica grains to form a sodium disilicate coating also occurred in this temperature range. The effects of reaction accelerant additions (Na₂SO₄, NaNO₃, NaCI) on batch fusion were analyzed. Sodium chloride was found to be the most effective melting accelerant due to the formation of a NaCI–Na₂CO₃ eutectic liquid phase at ∼636°C, which effectively attacked the silica relic. CO₂ gas release terminated ∼80°C earlier with 1 wt% NaCI additions to the base glass.     

Influence of Batch Moisture and Atmosphere on Melting Behavior of Lead Oxide-Containing Glass

Melting behavior in the sand-potash-lead oxide system was found to depend strongly on the amount of H₂O added to the batch (between 0.1 and 9.0 wt%) and also on the melting atmosphere (air, N₂, and air loaded with 30 vol% H₂O). Pb₃O₄ and PbO (red), which were used as raw materials, behave similarly. The newly formed compounds are Pb₃O₄ (formed only when PbO (red) was used as a raw material), PbO (yellow), and K₂Pb₂Si₂O₇. The addition of 2 wt% As₂O₃ to the batch influences the processes of formation and dissolution of compounds only in some cases.     

Glass Formation and Polyamorphism in Rare-Earth Oxide–Aluminum Oxide Compositions

We report formation of single- and two-phase glasses from rare-earth oxide–alumina materials. Liquids with the Y₃Al₅O₁₂ and Er₃Al₅O₁₂ compositions underwent a liquid–liquid phase transition which resulted in glasses with a cloudy appearance due to spheroids of one glass in a matrix of a second glass. The two glasses were isocompositional within the limits of experimental error. Clear, brilliant, single-phase glasses were obtained from La₃Al₅O₁₂, ErLaYAl₅O₁₂, and compositions containing ≥5 mol% La₂O₃ substituted for the other rare-earth oxides. Formation of two glasses is attributed to nucleation and growth of the second liquid at a temperature below the equilibrium liquid–liquid transition temperature. Addition of lanthanum depresses the phase transition temperature below the glass transition temperature and the liquid–liquid phase transition is not observed. The results are discussed in the context of first-order liquid–liquid phase transitions (polyamorphism) and formation of single-phase glass from liquids that contain a high proportion of 4-coordinate aluminum ions.     

Controlled Phase Transformations by Glass Particle Size in Spodumene and Cordierite Glass-Ceramics

The effects of the particle size of initial glass powders on the phase transformations in Li₂O-Al₂O₃-SiO₂ and MgO-Al₂O₃-SiO₂ glass-ceramics were investigated. Both materials undergo crystallization of glass to a metastable phase and a subsequent metastable-stable transformation. As particle size is decreased, the crystallization of glass to the metastable phase is accelerated, because of the increase in the total number of nucleating sites on the particle surfaces. The metastable-stable transformation is restricted as the particle size is decreased. This phenomenon is explained by the decrease in the crystal size of the metastable phase as a result of decreasing the particle size. The metastable-stable transformation of the fine crystals is restricted by the constraint imposed by the surrounding matrix.     

Nucleation and Crystallization of a Lithium Aluminosilicate Glass

An aluminosilicate glass of composition 61SiO₂6Al₂O₃10MgO6ZnO·12Li₂O·5TiO₂ (mol%) has been prepared by a melting process and investigated as far as crystallization is concerned. Glass-ceramic is easily obtained because glass shows a high tendency to crystallize starting from 700°C. The crystalline phases evolve with temperature, showing the aluminosilicates to be the main phase up to 1050°C, followed by metasilicates and silicates, some of which have lower melting points. The titanates of Mg and Zn develop from the phase-separated glass, soon after T _g, and grow to form nucleation centers for the other crystalline phases. The evolution from phase-separated glass to glass-ceramic has been followed by many thermal, diffractometric, spectroscopic, and microscopic techniques.     

Sulfate Decomposition and Sodium Oxide Activity in Soda–Lime–Silica Glass Melts

Reaction equilibrium constants for the sulfate decomposition process, which releases oxygen and sulfur oxide gas in soda–lime–silica glass melts, have been determined. The chemical solubility of SO₂, probably in the form of sulfite ions in soda–lime–silica melts, has also been determined. The chemical solubility value of SO₂, dissolving as sulfite, ranges between 0.02 and 0.06 wt% SO ₃ ²⁻ at 1 bar SO₂ pressure in the temperature range of 1600–1800 K. Results of square-wave-voltammetry studies and measurements of the temperature-dependent sulfur retention after the fining process of commercial float glass melts and a model soda–lime–silica melt, with 74 wt% SiO₂, 16 wt% Na₂O, and 10 wt% CaO, are presented. The measured sulfur retention data and the results of the square-wave-voltammetry studies are used to determine the equilibrium constant of the sulfate decomposition reaction in the temperature range of 1600–1800 K. The thermodynamic relations and properties found for sulfate decomposition are used to derive activities of sodium oxide in soda–lime–silica melts. Literature values for sodium oxide activities in these glass melts are rare. In this study, these activities have been determined by a method, based on the measurement of sulfate decomposition equilibrium constants and the residual sulfate concentrations in glass melts, equilibrated with almost pure sodium sulfate galls.     

Molten Glass Corrosion Resistance of Immersed Combustion-Heating Tube Materials in Soda-Lime-Silicate Glass

The corrosion resistance of molybdenum, molybdenum disilicide, and a SiC_(p)/Al₂O₃ composite to molten soda-lime-silicate glass was studied. The ASTM-C621–84 corrosion test method was modified because of inherent inaccuracies in the method and Si attack of platinum crucibles. Specimen-glass interfacial regions were characterized using XRD, SEM, and EDS. After 48 h of exposure at 1565°C, the half-down corrosion recessions of Mo, MoSi₂, and SiC_(P)/Al₂O₃ were 0.11, 0.316, and 0.26 mm, respectively. Mo oxidized to form a MoO₂ surface scale which cracked, allowing glass seepage and further oxidation. Silicon was leached out of MoSi₂ into the glass, leaving a Mo₅Si₃ interface and particles of Mo near the interface. For the SiC_(P)/Al₂O₃ composite, bubbles observed at the interfacial regions formed from oxidation of SiC to form CO. Thermodynamic modeling corroborated these experimental observations.     

Influence of Batch Moisture on Melting Behavior of Glass

Melting behavior in the system sand-potash-dolomite was found to depend strongly on the amount of H₂O added to the batch, which was varied from 0.5 to 9.0 wt%. This dependence holds primarily for the formation and dissolution of K₂Ca(CO₃)₂, but other compounds are also influenced, e.g., α-K₂MgSi₃O₈, CaMgSiO₄, Ca₂MgSi₂O₇, andCaSiO₃.     

Damage Behavior in Ceramic Plasma-Coated and Uncoated Glass with Steel-Ball Impact

A ceramic plasma-coating method that reduces damage from particle impact in a ceramic material is suggested. A steel-ball impact test was performed to investigate and compare the damage behavior of uncoated and ceramic-coated glass. Cone and lateral crack lengths were compared and damage volume analyzed and compared for different steel-ball diameters and types of coating materials, Al₂O₃-TiO₂, Al₂O₃, and Cr₂O_3. Damage resistance of the coating and Al₂O₃-TiO₂ was found to be best because of its mechanical properties and microstructure.     

Glass Formation Range, Acid Resistivity, and Surface Charge Density of ZnO-B2O3-SiO2 Passivation Glass Containing Al2O3

The glass formation range in the system ZnO-B₂O₃-SiO₂ increases when 5% Al₂O₃ is added and then decreases with further Al₂O₃ additions. The acid resistivity of the glass also increases when Al₂O₃ is added. An observed increase in negative charge with Al content until the system contains equal amounts of Al and Si (in forms of mole %) is explained by the formation of AlO₄⁻ tetrahedra which substitute in the SiO₄ network. Alkaline-earth oxides cause a positive charge which compensates for the negative charge formed by Al₂O₃. Antimony oxide and lanthanum oxide result in a negative charge in the glass. The formation of a negative or positive charge in the glass is thought to reflect the acidity or basicity of the glass, respectively.     

The Kinetics of Gamma-Ray Induced Coloring of Glass

When glasses are colored by ionizing radiation the induced optical absorption increases as the radiation progresses and appears to be due to the superposition of a number of individual absorption bands. A detailed study of this process has been made using a specimen of Corning boro-silicate glass colored by exposure to ⁶⁰Co gamma rays. This particular sample was chosen because only four bands are formed. If it is assumed that each band is Gaussian shaped the spectrum may be separated into four absorption bands. The peak energy E₀ and full width U of these bands are, in electron volts: E₀= 4.85, U = 1.19; E ₀= 3.95, U = 1.30; E _o= 2.58, U = 0.58; E₀= 2.02, U = 0.52. For each band growth curves may be constructed showing how the density of absorption centers increases as a function of dose. These growth curves have been fitted with theoretical curves based on the following considerations: the radiation field creates ionization electrons in the glass; for each ionization electron one electron deficient region or hole is formed; absorption bands observed are due to centers formed by electron trapping although the possibility that some of the bands are due to hole trapping is not ruled out; and competition for ionization electrons exists between holes and the various kinds of electron traps. Satisfactory agreement between the observed and calculated curves is obtained. The theory indicates that the "radiation protection" imparted to glass by materials such as CeO₂ may arise in several different ways and that it would be possible to decide between them from rather simple experiments.     

Solubility of Refractory Oxides in Soda-Lime Glass

A method for measuring the solubility limit of refractory oxides in glass is presented, and the solubilities of Al₂O₃, ZrO₂, SnO₂, and Cr₂O₃ in soda-lime glass are given as a function of temperature. The order of solubility in soda-lime glass is Al₂O₃ZrO₂SnO₂Cr₂O₃ at 1300° to 16Q0°C. The solubility of ZrO₂ in Al₂O₃-saturated soda-lime glass is much lower than that in pure soda-lime glass.     

Effect of Alcohols on Crack Propagation in Glass

The effect of straight-chain alcohols having chain lengths of 6 to 12 carbon atoms on crack propagation in glasses was studied. The test was a modification of the double-cantilever-beam technique. Plots of crack velocity vs stress intensity factor, K , were trimodal, similar to those for glass tested in N₂ gas of varying relative humidity. Crack velocities in the two regions of lowest K could be explained using a model derived by Wiederhorn for the effect of water on crack propagation and were independent of alcohol chain length. The chain length of the alcohol affected the results only in the water-independent high- K region, where crack velocity increased monotonically with decreasing chain length at a given K. There was no systematic effect of glass composition within the Na₂O-CaO-SiO₂ system, but crack velocities at a given K in a 3BaO-5SiO₂ glass were higher under all conditions than those in Na₂O-CaO-SiO₂ glasses.     

Stress in Leached Phase-Separated Glass

When a phase-separated glass is leached, stresses develop because of release of thermal stresses, creation of surface area, ion exchange, and hydration. Analyses are presented for the thermal stresses, including the portion that develops on cooling from the heat-treatment temperature to the setting temperature of the less viscous phase. During leaching, the interfacial energy of the residual phase increases, so that phase tends to contract. A more important effect is the contraction caused by removal of alkali and B₂O₃ from the residual phase during leaching. The extent of removal of B₂O₂ decreases with heat-treatment time, t_H , because the scale of the microstructure increases as t^1/3_H. The change in residual B₂O₂ content with t_H is shown to be consistent with diffusion-controlled ion exchange. The dependence of stress on t_H in partially leached glasses, measured by Drexhage and Gupta, results principally from the change in extent of ion exchange; the reduction in surface area with increasing t_H also has a significant effect on the stresses.     

Thermal Analysis of Reactions in Soda–Lime Silicate Glass Batches Containing Melting Accelerants: I, One- and Two-Component Systems

To identify each glass melting reaction in a multicomponent system, one-component and two-component reaction processes were studied using DTA, TGA, and XRD. Two-component mixtures were prepared by choosing pairs in the same ratio as in a commercial container glass batch composition (sand-soda ash-calcite-dolomite-feldspar). The presence of silica in the silica-calcite system decreased the termination temperature of the decomposition of calcite, but did not alter the onset of decomposition. Similar behavior was found in the dolomite-silica system. A double carbonate (Na₂Ca(CO₃)₂) formed via solid-state reaction in the calcite-soda ash system, and metasilicate/disilicate phases were detected during the fusion process in the soda ash-silica system. The effects of reaction accelerant additions in the soda ash-silica system were investigated using 1 wt% additions of sodium sulfate, sodium nitrate, and sodium chloride. Sodium chloride was the most effective melting accelerant, lowering the termination temperature of CO₂ release by ∼80°C compared with the soda ash-silica system with no additives. NaCl additions caused complete reaction and/or fusion of Na₂CO₃ prior to its melting temperature. Sodium sulfate additions acted to suppress metasilicate/ disilicate formation by coating quartz grains and shifted consequent CO₂ release to higher temperature.     

Crystallization of Copper Metaphosphate Glass

The effect of the valence state of copper in copper metaphosphate glass on the crystallization behavior and glass transition temperature has been investigated. The crystallization of copper metaphosphate is initiated from the surface and its main crystalline phase is copper metaphosphate (Cu(PO)₃), independent of the [Cu²⁺]/[Cu_total] ratio in the glass. However, the crystal morphology, the relative crystallization rates, and their temperature dependences are affected by the [Cu²⁺]/[Cu_total] ratio in the glass. The crystal formed in the reduced glass is oriented and needleshaped and less aligned at higher crystallization temperature. On the other hand, the totally oxidized glass crystallizes from all over the surface. The relative crystallization rate of the reduced glass to the totally oxidized glass is large at low temperature, out small at high temperature. The glass transition temperature of the glass increases as the [Cu²⁺]/[Cu_total] ratio is raised. This dependence may be used to explain the relative crystallization rates. It is also found that the atmosphere used during heat treatment does not influence the crystallization of the reduced glass, except for the formation of a very thin CuO surface layer when heated in air.     

Decomposition of Ruthenium Oxides in Lead Borosilicate Glass

Behaviors of apparent phase changes of Ru-containing oxides in lead borosilicate glass at high temperature have been investigated using X-ray diffraction and transmission electron microscopy in application to thick-film resistors. During firing of thick films containing Ru oxide powder and lead borosilicate glass frits, apparent phase changes of Ru oxides have been found to occur both ways between ruthenium dioxide and lead ruthenate pyrochlore via decomposition of one phase in glass and subsequent formation of the other. The formation of pyrochlore occurs in a lead-rich form, Pb₂(Ru_{2− x} Pb _x )O_6.5, whereas the formation of RuO₂ is characterized by a platelike morphology instead of initial globular morphology. A general tendency is observed that RuO₂ is stable in low-PbO glass compositions and at high temperatures, while Pb₂(Ru_{2− x} Pb _x )O_6.5 is stable in high-PbO glass compositions and at low temperatures, with the implication that the stability of these phases is dictated by the chemical activity of PbO in the glass melt.     

Faceting Behavior of Alumina in the Presence of a Glass

The faceting of alumina interfaces in the presence of a glass affects both grain growth and grain-boundary mobility during liquid-phase sintering. The geometry and movement of facets that form during this sintering process are expected to play an essential role in the development of the final microstructure, in particular, by their influence on the topology of the grain boundaries which ultimately control the properties of Al₂O₃ compacts. A new method for studying the interaction between Al₂O₃ and a glass has been developed. A thin sample of Al₂O₃ suitable for examination in a transmission electron microscope is prepared and examined and then reacted with SiO₂ and CaO via the vapor phase. This experimental approach allows the faceting behavior of glass/Al₂O₃ interfaces to be studied systematically without introducing unnecessary complications during subsequent sample preparation. Faceting occurs almost exclusively on the (0001) and {1 1 02} planes. The interaction between glass and certain structured grain boundaries in alumina has been studied using polycrystalline thin films.     

Glass/Substrate Interaction in Model RuO5-Glass Double-Layer Thick Films

Chemical interaction between RuO₂-glass thick-film resistors and substrates was studied by observing microstructural changes in model double-layer thick films, which were composed of a top glass layer and an RuO₂ layer. The concentration profiles of Al, Pb, and Ru in the film were analyzed. A new glass layer formed on the alumina substrate during firing, the formation of which resulted from dissolution of A1₂O₃ into glass. A mechanism of microstructural changes has been proposed based on the results of EPMA.