Effect of melt cleanliness on the properties

An experimental study was conducted to examine the effect of melt cleanliness with respect to the presence of inclusions on the properties of an Al-10 wt pct Si metal matrix composite (MMC) reinforced with 10 vol pct SiC particles. The occurrence of inclusions was controlled by filtra- tion, using ceramic foam filters of 10, 20, and 30 ppi sizes, under gravity and pressure. Test bars obtained from filtered and unfiltered melt castings, prepared from fresh (as-received) and recycled composite materials, were T6-tempered and tensile tested at room temperature. The casting quality was examined using X-ray radiography. The results indicate that various factors influence the casting quality and mechanical properties of the cast composite. The A1₂O₃ films and spinel MgAl₂O₄ — the main inclusions observed in the present composite — are chiefly responsible for the degradation in the mechanical properties. In addition, SiC sedimentation, Al₄C₃ formation, the hydrogen level of the melt, and the starting material used can also influence these properties. Fracture studies reveal that the inclusions and associated microvoids act as the crack initiation sites during composite fracture. Simple filtration using 10 ppi ceramic foam filters under gravity serves adequately in removing these inclusions and producing the desired mechanical properties.     

Deep filtration of molten aluminum using ceramic foam filters and ceramic particles with active coatings

The inclusions in molten aluminum were removed using the deep filtration of ceramic foam filters and ceramic particles with active coatings. The results of tensile tests showed that the elongation of the filtered tensile specimen S₆ increases by 17.93 pct, but the tensile strength does not. The scanning electron microscope (SEM) examination showed that the secondary cracks and dimples in the filtered tensile specimen S₆ were finer and more homogeneous than those in the unfiltered tensile specimen S₀. In addition, metallographic examination showed that there were only a few inclusions of approximately 6 μm in diameter in the filtered specimen S₆, but more inclusions of approximately 40 μm in diameter were found in the unfiltered specimen S₀. The filtration efficiency of the coated ceramic particles was higher than that of the coated ceramic foam filters. The active coating could effectively capture the inclusions and dissolve Al₂O₃ in them during filtration.     

Removal of Inclusions from Aluminum Through Filtration

Filtration experiments were carried out using both an AlF₃ slurry-coated and an uncoated Al₂O₃ ceramic foam filter to study the removal of nonmetallic inclusions and impurity elements. The results showed that the 30-ppi ceramic foam filter removed up to 85 pct inclusions from aluminum. Several pictures of two- and three-dimensional morphologies of both nonmetallic and intermetallics inclusions also have been presented. The following contributing mechanisms for the removal of nonmetallic inclusions in the deep-bed filtration mode are proposed: (1) collision with walls and interception effect and (2) the formation of both intermetallic and nonmetallic inclusion bridges during filtration. Fluid dynamics modeling of inclusion attachment to the filter walls showed that most inclusions, especially those with larger sizes, are entrapped at the upper part of the filter, whereas smaller inclusions are dispersed well throughout the filter. The calculated inclusions removal fractions for the 30-ppi filter showed that almost all inclusions >125 μm are removed, and inclusions ~5 μm in size are removed up to 85 pct. The interfacial energy between two collided same-size inclusions was calculated, indicating that a strong clustering of inclusions may result within the filter window. Magnesium impurities were removed up to 86 pct by the AlF₃ slurry-coated filter. The filter acted in active filtration mode in addition to the contribution of the air oxidation of dissolved [Mg], which was calculated to be 13 pct. The total mass transfer coefficient of dissolved [Mg] to the reaction interface was calculated to be 1.15 × 10⁻⁶ m/s.     

Solvent extraction of Ce(III) and Pr(III) with P507 using SiC foam as a static mixer

Extraction reactor is a major research area of interest within the field of rare earths extraction and separation. SiC foam offers excellent material characteristics as well as three-dimensional (3-D) reticulated structure; however, very little research has been carried out on its application in extraction reactor so far. In this work, a static mixer reactor based on SiC foam was designed and demonstrated to extract and separate Ce(III) and Pr(III) from nitric acid media by using 2-ethylhexyl phosphoric acid mono-2-ethylhexyl ester (P507) as extractant. The structure–performance relationship between SiC foam and extraction performance was studied by experiment combined with computational fluid dynamics (CFD) simulation. The experiment data are in good agreement with the simulation results. Contrast experiment by using a Kenics mixer was carried out, and SiC foam shows better extraction and mass transfer performance. Using the optimal structural SiC foam (pore size D = 2.3 mm, open porosity ε = 85%, foam length L = 80 mm), high extraction efficiency η (Pr(III): 94.6%, Ce(III): 88.5%) and separation factor β (2.27) between Ce(III) and Pr(III) is achieved at a high total throughput of 200 mL/min. Besides, overall volumetric mass transfer coefficient K_La of Pr(III) and Ce(III) are 0.519 and 0.378 s^?1 at the residence time τ of 3.6 s, respectively, which reach the high level of microchannel reactors and are better than conventional extractors and other static mixers. SiC foam is found to be applicable as a static mixer for efficient and high-throughput extraction and separation of rare earths.     

Nucleation of solid aluminum on inclusion particles injected into Al-Si-Fe alloys

Systematic inoculation experiments were carried out to study the influence of various inclusions on the nucleation of the α-Al phase in Al-Si-Fe alloys at different cooling rates. The results showed that in dilute alloys, containing less than 1.5 pct Si+Fe, almost all the inclusion types have high percentages of occurrence within the α-Al phase, indicating that nucleation can be promoted on the surface of such inclusions. In a hypoeutectic Al-Si alloy containing 6.3 pct Si, the inclusion particles of MgO, TiB₂, TiC, α-Al₂O₃, and SiC become mostly inactive nucleants and are pushed to the interdendritic regions because of the dominating poisoning effect of Si. The current results were used successfully to explain the efficiency differences between the commercial grain refiners in the hypoeutectic Al-Si alloys. Silicon is observed to preferentially segregate to the liquid-Al/inclusion interfaces so as to lower the free energy of such interfaces. A theoretical analysis of the poisoning effect of Si showed that Si segregation to the liquid/nucleant interface alters the interfacial energy balance so that the catalytic efficiency of the nucleant particles is dramatically reduced. Careful analysis showed that the poisoning effect of Si in the hypoeutectic alloy is overcome when the nucleant particles have active surface characteristics, as represented by the high catalytic potencies of γ-Al₂O₃, CaO, and Al₄C₃ particles in nucleating the α-Al phase of the hypoeutectic Al-Si alloy. Although some inclusions have comparable or higher occurrence levels than TiB₂ in the α-Al phase, they cannot be used as efficient nucleants because of either their poor wettability with liquid aluminum or their chemical reactivity, which can change the alloy chemistry.     

Refining of Aluminum and Steel Melts by the Use of Multi-Cellular Extruded Ceramic Filters

Abstract

Laboratory prepared melts of steel containing Al₂O₃ inclusions and aluminum containing TiB₂ inclusions have been successfully filtered using multicellular extruded ceramic filters. Relatively high inclusion removal efficiencies have been achieved in both low temperature and high temperature melt systems —68 % inclusion removal efficiency in the Al-TiB₂ system (1020 K) using cordierite multicellular filters and 96 % inclusion removal efficiency in the steel–Al₂O₃ system (1873 K). The results have been analysed using Apelian and Mutharasan's kinetic model for filtration of the metallic melts [1]. The inclusion capture kinetics and filtration characteristics of the porous media used in this investigation are discussed.     

The interactive role of inclusions and SiC reinforcement on the high-cycle fatigue resistance of particle reinforced metal matrix composites

The effect of intermetallic inclusions on the fatigue crack initiation and growth in 2080 Al alloy and 2080/SiC _p composites was investigated. Using surface replication, it was determined that, in the high-cycle fatigue region, life is dominated by the initiation process. It was also determined that the majority of initiation sites were associated with intermetallic inclusions. While 2080/SiC/20 _p showed a definitive relationship between inclusion size and fatigue life, i.e., a higher inclusion size resulted in lower fatigue life, there was no correlation in 2080/SiC/30 _p . This was attributed to more of the load being shared by the higher volume fraction of SiC particles and smaller average inclusion sizes in the latter composite. A conceptual model is proposed that accounts for these observations and qualitatively shows the effect of reinforcement on stress enhancement in near-surface inclusions. N. CHAWLA, formerly Research Fellow, Department of Materials Science and Engineering, University of Michigan C. ANDES is former Research Fellow, Department of Materials Science and Engineering, University of Michigan. This article is based on a presentation made in the Symposium “Mechanisms and Mechanics of Composites Fracture” held October 11–15, 1998, at the TMS Fall Meeting in Rosemont, Illinois, under the auspices of the TMS-SMD/ASM-MSCTS Composite Materials Committee.     

A mathematical model of aluminum depth filtration with ceramic foam filters: Part I. Validation for short-term filtration

This work presents a mathematical model to compute the efficiency of depth filtration of molten aluminum using ceramic foam filters. In the model, the porous structure of foam filters was represented by a unit cell that takes into account the convergent-divergent nature of the flow field. The steady, two-dimensional, and fully developed flow field within the unit cell was obtained from the numerical solution of the continuity and Navier-Stokes equations. The assessment of the proper assumptions for the model was carried out by comparing the computed velocity field with that experimentally determined for a physical model of the unit cell with scale 10:1 and containing an aqueous solution of CaCl₂. The measurements were done using the particle image velocimetry (PIV) technique. The efficiency and the coefficient of initial filtration for foam filters were obtained from the determination of the particle limiting trajectory, resulting from a force balance on a spherical inclusion. This balance included the buoyancy and the viscous drag forces. The last force took into consideration the wall effect on the particle motion. The values of the computed initial filtration coefficient show an excellent agreement with the corresponding measured ones reported for laboratory and plant tests for short-term filtration. This comparison involves several combinations of particle sizes and downward fluid superficial velocities. This model is further extended to study long-term filtration in the second part of the article.     

<Emphasis Type="Italic">In-situ</Emphasis> formation of SiC-reinforced Al-Si alloy composites using methane gas mixtures

Theoretical and experimental studies on the in-situ formation of an Al-Si alloy composite using a methane gas mixture were carried out. An Al-Si alloy composite with in-situ formed SiC as a reinforced phase was produced by bubbling methane gas at temperatures from 1223 to 1423 K. An optical microscope, scanning electron microscope (SEM), and electron microprobe were used for the product characterization. Primary and eutectic silicon were observed in the samples taken from the top part of the crucible, and only eutectic silicon was observed in the samples taken from the bottom of crucible. The SiC formation rate increased with the decrease in the bubble size. A silicon concentration gradient existed at different vertical positions of the liquid alloy. The silicon concentration close to the top of the liquid alloy was higher than that at the bottom. The SiC concentration was very low in the bulk alloy. The bubbling of the gas mixture in the melt resulted in the formation of a layer of foam on top of the crucible. Formed SiC particles were enriched in the foam and carried out of the crucible by the overflow foam to a composite collector located under the crucible. The foam in the composite collector was broken, and composites in the foam contained up to 30 wt pct SiC. The particle size of the SiC is in the range of 1 to 10 μm. The bubbling process resulted in the unevenness of the silicon concentration and the different crystallizing processes. The SiC formation rate was found to be about 12.5 mg/(L·s). A kinetic model was developed. The model-predicted results are in very good agreement with the experimental results.     

Nucleation of Fe-intermetallic phases in the Al-Si-Fe alloys

Nucleation of Fe-intermetallic phases (i.e. binary Al-Fe, α-AlFeSi, β-AlFeSi, δ-AlFeSi, and q₁-AlFeSi phases) on the surface of different inclusions in six experimental Al-Si-Fe alloys was studied through a quantitative evaluation of the number of inclusion particles that have a direct physical contact with the nucleated phase as seen through the optical microscope. It was found that nucleation of each of the Fe-intermetallic phases was promoted on the surface of several inclusions under the same conditions of alloy composition and cooling rates. Some inclusions exhibited high potency for the nucleation of particular Fe-intermetallic phases under certain conditions and poor potency under other conditions. The potent nucleants for the primary α-Al phase such as γ-Al₂O₃ exhibited poor potency for the nucleation of the Fe-intermetallic particles that lie within the primary phase (intragranular particles). Reactive inclusions such as CaO and SiC are very potent nucleants for the intragranular Fe-intermetallic phase particles. The nucleation of the Fe-intermetallic phases in Al-Si-Fe alloys obeys the general features of nucleation, in particular, the effect of cooling rate and solute concentration on the potency of the nucleant particles: (1) it was observed that increasing the cooling rate enhances the heterogeneous nucleation of the Fe-intermetallic phases on the surface of different inclusions, and (2) the nucleation potency of inclusion particles in both α-Al and interdendritic regions improves with increasing solute concentration up to a certain level. Above this level, the solute concentration poisons the nucleation sites. Nucleation of the Fe-intermetallics in the alloys studied does not seem to be largely affected by the type of the nucleating surface.     

The interactive role of inclusions and SiC reinforcement on the high-cycle fatigue resistance of particle reinforced metal matrix composites

The effect of intermetallic inclusions on the fatigue crack initiation and growth in 2080 Al alloy and 2080/SiC _p composites was investigated. Using surface replication, it was determined that, in the highcycle fatigue region, life is dominated by the initiation process. It was also determined that the majority of initiation sites were associated with intermetallic inclusions. While 2080/SiC/20 _p showed a definitive relationship between inclusion size and fatigue life, i.e., a higher inclusion size resulted in lower fatigue life, there was no correlation in 2080/SiC/30 _p . This was attributed to more of the load being shared by the higher volume fraction of SiC particles and smaller average inclusion sizes in the latter composite. A conceptual model is proposed that accounts for these observations and qualitatively shows the effect of reinforcement on stress enhancement in near-surface inclusions. This article is based on a presentation made in the Symposium “Mechanisms and Mechanics of Composites Fracture” held October 11–15, 1998, at the TMS Fall Meeting in Rosemont, Illinois, under the auspices of the TMS-SMD/ASM-MSCTS Composite Materials Committee.     

Direct simulation of initial filtration phenomena within highly porous media

Initial filtration phenomena within highly porous filter media—ceramic foam filters (CFF) —were simulated by numerically solving the Navier-Stokes equations and the general transport equation for suspended particles. In this approach, a “piece” of the highly porous filter was modeled directly by constructing a calculation domain such that the main average geometrical properties of the real filters were embodied therein. The governing equations were then solved by imposing appropriate boundary conditions on the solid surfaces of the filter webs. The influences of Reynolds, Peclet, and Gravitational numbers, as well as filter porosity, on initial filtration efficiencies were investigated in this simulation. Predicted results were spotchecked against data from industrial filtration trials conducted by the authors for aluminum melts and found to be in good agreement.     

Processing and Stability of LM25/SiCP Composite Foams

A simple method of metal foam production is to introduce a blowing agent (e.g. TiH₂) into an aluminium melt containing foam stabilisers such as oxides (usually Ca-based) and/or particles (e.g. SiC, Al₂O₃). In this work, Al/SiC composites (in-house and commercial Duralcan) [both of them with LM25 matrix (Al?C7Si?C0.3Mg)] containing particles of various sizes and contents were foamed at different temperatures using TiH₂. Foamability is characterised through their expansion and collapse. It is observed that high expansions and good quality foams could be obtained upon manipulating SiC particle size and content. However, irrespective of particle size/vol.% combination, significant effect of foaming temperature is noticed on the fundamental stability of the liquid foam until solidification. Both cell size and foam density varied along the ingot height. The distribution of SiC_P within the cell wall is random with no preferential segregation to gas/metal interface. The evolution of foam, and the role of SiC on foam stability are discussed based on macro and cell wall microstructural results.     

Foamability of MgAl2O4 (Spinel)-Reinforced Aluminum Alloy Composites

A novel foamable aluminum alloy has been developed. It contains sub-micron-sized MgAl₂O₄ (spinel) particles that are generated in situ by a reaction of SiO₂ with a molten Al-Mg alloy. The study involves an optimization of parameters such as Mg concentration, SiO₂ particles size, and reaction time and shows that a composite containing MgAl₂O₄ particles as chief reinforcement in the matrix leads to effective foaming. Composites containing large sized transition phases and particle agglomerates in the matrix yield poor foam structure. The best foamable composite obtained contained 3.4 vol. pct of ultrafine (80 nm to 1 μm) MgAl₂O₄ particles uniformly distributed in an Al-Si alloy matrix. The corresponding metal foam contained 75 pct porosity and exhibited a uniform distribution of cells.     

Finely dispersed Si3N4−SiC powders and materials based on them

The objectives of the present research were to investigate the preparation of homogeneous ultrafine composite Si₃N₄−SiC powders by a plasmochemical process and the properties of ceramics produced from them. The chemical and phase compositions of the powders depended on the particle size of the initial powder, silicon input rate, and ratio of ammonium and hydrocarbon flow rates. The particle size and specific surface area of the compounds depended on the concentration of particles in the gas jet, and the cooling rate of the products. Composite powders containing from a few up to 90 mass % SiC, with specific surface areas of 24–80 m²/g and free silicon and carbon content less than 0.5 mass % were obtained. The main phases present were α-Si₃N₄, β-Si₃N₄, β-SiC, and X-ray amorphous Si₃N₄. Dense materials were prepared both by hot pressing at 1800°C under a load of 30 MPa and gas-pressure sintering at 1600–1900°C under a pressure of 0.5 MPa nitrogen. The plasmochemical composites had smaller pore sizes, were finer grained, and densified more rapidly than materials sintered from commercial powders. Institute of Inorganic Chemistry, Latvian Academy of Sciences, Salaspils. Translated from Poroshkovaya Metallurgiya, Nos. 1–2(405), pp. 7–12, January–February, 1999.     

Fabrication of Al-3.7?Pct Si-0.18?Pct Mg Foam Strengthened by AlN Particle Dispersion and its Compressive Properties

Al-3.7 pct Si-0.18 pct Mg foams strengthened by AlN particle dispersion were prepared by a melt foaming method, and the effect of foaming temperature on the foaming behavior was investigated. Al-3.7 pct Si-0.18 pct Mg alloy containing AlN particles was prepared by noncompressive infiltration of Al powder compacts with molten Al alloy in nitrogen atmosphere, and it was foamed at different foaming temperatures ranging from 1023 to 1173 K. The porosity of prepared foam decreases and the pore structure becomes homogeneous with increasing foaming temperature. When the foaming temperature is higher than 1123 K, homogeneous pores are formed in the prepared ingot without using oxide particles and metallic calcium granules, which are usually used for stabilizing a foaming process. This stabilization of the foaming at high temperatures is possibly caused by Al₃Ti intermetallic compounds formed at high temperature and AlN particles. Compression tests for the prepared foams revealed that the absorbed energy per unit mass of prepared Al-3.7 pct Si-0.18 pct Mg foam is higher than those of aluminum foams strengthened by alloying or dispersion of reinforcements. It is remarkable that the oscillation in stress, which usually appears in strengthened aluminum foams, does not appear in the plateau stress region of the present Al-3.7 pct Si-0.18 pct Mg foam. The homogeneity in cell walls and pore morphology due to the stabilization of pore formation and growth by AlN and Al₃Ti particles is a possible cause of this smooth plateau stress region.     

Carbothermic Reduction of Amorphous Silica Refined from Diatomaceous Earth

Aimed at developing solar-grade Si (SOG-Si) resources, amorphous silica (AS) refined from diatomaceous earth was reduced carbothermically. The reactivity of quartz—typically crystalline silica—also was investigated for comparison. Preliminary experiments confirmed an intermediate phase of SiC during the carbothermic reaction. SiC was produced more easily by heating AS mixed with graphite within 2 hours at 1773 K in a resistance furnace, whereas quartz remained unreacted under the same condition. The AS mixed with SiC then was heated in an electrode impulse furnace. An Si peak was identified in the X-ray diffraction (XRD) pattern of the sample reacted within 30 seconds at 2273 K. Chemical analysis indicated that the mole ratio of reduced Si to initial SiO₂ increased with a heating time of 15–30 seconds. It almost reached a constant depending on the heating temperature. The initial stage may correspond to a significant reduction from SiO₂ to Si in the solid–solid or solid–gas reaction systems. The next stage probably is a slow vaporization of SiO(g). Once the reduced Si melts with SiO₂ at the high temperature, the melt partially covers the surface of SiO₂ to prevent contact with SiC. A better reactivity for refined AS is observed than for quartz.     

The Morphology and Chemistry Evolution of Inclusions in Fe-Si-Al-O Melts

This study aims to elucidate the process of inclusion precipitation in Fe-Si and Fe-Si-Al melts. Deoxidation experiments were carried out in a vacuum induction furnace (VIF) at 1873 K (1600 °C). In the Si-deoxidation experiments, spherical SiO₂ of 1~2 μm diameter was dominant. When 3 wt pct Si and 300 ppm Al were added, such that Al₂O₃ and mullite were thermodynamically stable, the resulting inclusions depended on the addition sequence. When aluminum was added before silicon, spherical aluminum oxides were dominant after the Al addition, but after the Si addition, the number and size of alumina decreased and Al-Si oxides and mullite appeared with increasing time. When silicon was added before aluminum, spherical SiO₂ was dominant after the Si addition, but after the Al addition, spherical and polygonal alumina inclusions were dominant. When Al/Si was added simultaneously, polygonal alumina inclusions were dominant initially, but with time, Al-Si oxide and mullite inclusions increased in numbers. If the Al amount in the Al/Si addition was increased to 600 ppm, only alumina was found. This study shows how, under similar thermodynamic conditions, the transient evolution of inclusions in iron melts in the Si-Al-O system differ depending on the alloy addition sequence.