Nathaniel Bryant University of Northern Iowa Let us know how access to this document benefits you Copyright ©2019 Nathaniel Bryant Follow this and additional works at: https://scholar
INTRODUCTION
Casting involves pouring molten material into a mold cavity to create a desired shape After cooling and solidification, the cast part is removed and undergoes post-processing for its intended use This study focuses on steel casting, a practice dating back to 1200 B.C.E (McClellan III & Dorn, 2006) Advancements in metallurgical knowledge and technology have significantly improved steel quality over time.
Bessemer and open-hearth processes, the electric-arc furnace, and the basic oxygen process
Despite significant technological and scientific advancements improving steelmaking, certain steel properties still pose challenges in casting today Modern casting facilities often use simulation software to diagnose and address casting defects before prototype development; however, these tools are not infallible While current simulation packages can effectively model common issues such as porosity, they lack models for some complex defects, forcing engineers to rely on experience and literature In this particular case, a commercial steel foundry encountered a persistent, unexplained quality defect at the mold-metal interface where liquid steel interacts with bonded chromite sand This defect appeared as a large metal mass consuming part of the mold's porous structure, requiring extensive post-processing (as shown in Figure 1) The foundry engineers sought assistance from the University of Northern Iowa to identify and resolve the problem.
Figure 1: Presentation of the sporadic defect experienced by the commercial foundry
The metal has intruded into the marked section between the two teeth
The defect was initially identified as a metal penetration defect, which occurs when liquid metal intrudes into the mold's interstices Metal penetration can result from mechanical forces overcoming resistance or chemical reactions producing low-melting-point compounds like fayalite, leading to capillary-driven infiltration In heavy-section steel casting, the double-skin defect involves both mechanical and chemical mechanisms, creating a complex, hybrid penetration phenomenon.
This study aims to gain a fundamental understanding of the double-skin defect and the critical conditions leading to its formation A mathematical model will be developed to predict the occurrence and likelihood of the double-skin defect, allowing customization based on specific process variables in any production foundry This predictive tool enables foundry professionals to make informed decisions regarding materials and processes, ensuring high-quality large steel castings without flaws.
The primary cause of the double-skin defect is the specific penetration mechanism during the manufacturing process, which leads to structural inconsistencies Developing an accurate predictive model is essential to identify the conditions that foster this defect, enabling preventative measures By understanding the key factors influencing the defect formation, engineers can implement strategies to mitigate its occurrence, ultimately improving product quality and manufacturing efficiency.
This investigation aims to understand the formation conditions of double-skin penetration defects in materials By analyzing and quantifying these conditions, we can develop a predictive mathematical model to identify and prevent such defects The resulting model will serve as a valuable tool for industrial facilities, enabling engineers to effectively mitigate double-skin penetration issues and improve manufacturing quality.
Current process simulation software does not fully address all casting defects encountered by foundries This research aims to develop a mathematical model that enhances understanding of the causes of double-skin penetration, enabling foundries to implement effective prevention strategies By improving defect prediction, this model can reduce costly mistakes and improve casting quality in the industry.
The recent trend shows that the average age of foundrymen is decreasing, leading to a knowledge gap as experienced workers retire This loss of technical expertise poses a challenge for the industry, emphasizing the need to develop innovative technologies that can substitute traditional experience-based problem-solving Investing in technological advancements is essential to preserve institutional knowledge and ensure continued industry growth This investigation represents a crucial step toward leveraging technology to bridge the experience gap in the foundry industry.
Hypothesis and Research Questions What are the favorable conditions for the formation of double-skin penetration?
1 What is the primary penetration mechanism for double-skin formation?
2 Can double-skin penetration be predicted using a mathematical model?
3 What are some possible solutions to prevent double-skin penetration?
502 Bad GatewayUnable to reach the origin service The service may be down or it may not be responding to traffic from cloudflared
502 Bad GatewayUnable to reach the origin service The service may be down or it may not be responding to traffic from cloudflared
REVIEW OF LITERATURE
502 Bad GatewayUnable to reach the origin service The service may be down or it may not be responding to traffic from cloudflared
502 Bad GatewayUnable to reach the origin service The service may be down or it may not be responding to traffic from cloudflared
These phenomena have been thoroughly investigated for both iron and steel casting, and research conducted by the University of Alabama (Stefanescu et al., 1996;
Research by Giese, Stefanescu, Barlow, and Piwonka (1997), along with Pattabhi, Lane, and Piwonka (1996), indicates that cast iron metal penetration defects are primarily linked to mechanical penetration mechanisms In contrast, steel casting defects can result from both mechanical and chemical penetration processes, influenced by several critical factors outlined in Table 1 These findings emphasize the importance of understanding the distinct mechanisms behind metal penetration to improve casting quality (Hayes, Barlow, Stefanescu, & Piwonka, 1998; Barlow, Owens, Stefanescu, Lane & Piwonka, 1997).
Table 1: Factors that influence metal penetration defects (Stefanescu et al., 1996)
Increased penetration Decreased penetration Metallostatic pressure Surface energy Dynamic pressure Metal/mold contact angle Sand grain size Sand coating Pouring temperature Mold density
Metallostatic pressure is generated by the sheer weight of the molten metal, which is exerted directly onto the mold interface during pouring and solidification, defined by the equation
Increasing the height of the metal section above the mold interface (h) raises the metallostatic pressure, which enhances the likelihood of metal penetration during casting Since the only variable we can control in this equation is the height (h), creating tall casting geometries and risers significantly increases the risk of defects due to higher metallostatic pressure Therefore, managing the metal section height is crucial to minimize penetration issues and ensure casting quality.
This condition occurs exclusively during pouring when molten metal enters the casting cavity turbulently It is important to monitor this phenomenon, as turbulence can impact the quality of the casting Dynamic pressure, a key factor during this process, is calculated using the specified equation to assess its effects Understanding and controlling this dynamic pressure helps optimize pouring quality and minimize casting defects.
The instantaneous velocity of molten metal upon impact with the mold interface, denoted as "V," plays a crucial role in the casting process An increase in dynamic pressure elevates the likelihood of metal penetration, making velocity a key factor in mold filling quality Factors such as sprue height, pouring height, and the cross-sectional area of the gating system directly influence this velocity, as explained by Bernoulli’s principles (Stefanescu et al., 1996) Optimizing these parameters is essential to control metal flow and prevent defects in casting.
Sand Grain Size and Mold Density
The average grain diameter of sand used to create molds is typically determined using the American Foundry Society’s AFS-1105-00-S method, which helps calculate the surface area at the sand interface following AFS-1109-12-S standards Increasing the sand particle size generally enhances the permeability of the sand, indicating larger open spaces on the mold surface A study by Bryant and Thiel (2017) demonstrated a direct relationship between grain diameter and permeability, concluding that larger grain sizes allow greater metal penetration due to increased open spaces, as illustrated in their research findings.
Figure 2: The relationship between AFS-GFN and sand permeability (Bryant & Thiel,
Surface Energy and Contact Angle
Stefanescu et al (1996) explained that atoms in the same physical state are attracted to each other due to molecular forces, resulting in balanced forces within a homogenized mixture However, when materials of different physical states come into contact, their atoms experience varying forces, leading to differences in interfacial energy This variation in energy at the interface is known as surface energy or surface tension, which characterizes the unique properties of atoms at the boundary between different phases.
In surface chemistry, these variables are measured using the sessile drop methodology, which involves placing a liquid droplet on a solid substrate in open atmosphere, as seen in Figure 3
Figure 3: The surface energy balance for a theoretical vapor-liquid-solid system (Liu &
The shape that the droplet forms is the result of the equilibrium balance of surface energies, which is described by Young’s equation
The equation γLV cos(θ) = γSV − γLS describes the relationship between surface energies at different interfaces, where γLV represents the liquid-vapor surface energy, γSV the solid-vapor surface energy, γLS the liquid-solid surface energy, and θ the contact angle Since both θ and γLV can be measured easily, this formula allows for the calculation of capillary pressure at the mold-metal interface Understanding these surface energy interactions is essential for optimizing processes involving wetting and adhesion in manufacturing and material science applications.
The average pore size between sand particles, denoted as de, plays a crucial role in capillary behavior Capillary pressure is highly dependent on the material's wettability, with negative Pγ indicating that capillary forces promote penetration when the contact angle (θ) is less than 90°, signifying a wetting condition Conversely, if the contact angle exceeds 90°, indicating a non-wetting scenario, Pγ becomes positive, meaning capillary pressure opposes metal penetration into the mold (Stefanescu et al., 1996).
The contact angle is significantly affected by temperature, with higher temperatures leading to increased wetting For instance, Stefanescu's study reports that the contact angle of liquid iron on silica sand decreases as temperature rises, indicating enhanced wetting behavior This temperature-dependent change in contact angle plays a crucial role in capillary pressure, as higher pouring temperatures heighten the likelihood of metal penetration due to amplified capillary forces.
Figure 4: The relationship between contact angle of liquid iron against silica and temperature (Stefanescu et al., 1993)
As liquid metal enters the mold, it encounters frictional resistance that hinders its penetration into the interstices This frictional pressure, which impacts the flow behavior of the molten metal, can be accurately calculated using a specific formula Understanding this resistance is crucial for optimizing casting processes and ensuring high-quality metal molds.
The liquid metal viscosity, denoted as "à," plays a crucial role in the metal penetration process Molding sand permeability, represented by "K," is directly related to the average sand particle diameter, influencing how easily the metal can infiltrate the mold The depth of penetration, "Lp," and the speed of the penetrating metal, "Vp," are key parameters determining the quality of casting Additionally, frictional pressure varies according to changes in sand permeability, as described by the relationship shown in Figure 2, affecting how metal flows into the mold cavity.
Pressure calculations based on diverse process and material variables are essential for ensuring proper mold-metal interface balance These pressures are incorporated into governing equations that accurately depict the pressure equilibrium during molding Optimizing these parameters helps improve process reliability and product quality, making the understanding of mold-metal pressure dynamics vital for successful manufacturing.
In casting processes, the pressure balance is described by the equation Pst + Pdyn + Pexp ≥ Pγ + Pf + Pgas, where "Pexp" represents the pressure from graphite expansion during solidification—though this is irrelevant in steel castings—and "Pgas" accounts for pressure from decomposing resin binders, which is typically negligible When the combined metallostatic, dynamic, and expansion pressures exceed the opposing forces, penetration is likely to occur, affecting the quality and integrity of the cast product.
EXPERIMENTAL METHODOLOGY
A specialized test casting was developed to accurately measure the temperature-dependent CO/CO2 ratio within a mold, providing valuable insights into mold atmosphere chemistry This test design is based on the proven test mold used in Barlow’s investigation (Barlow et al., 1997), enhancing reliability and comparability of results The detailed schematic, shown in Figure 10, highlights the core geometry dimensions, with hidden or dashed lines denoting the specific measurements critical for understanding the mold environment This innovative approach facilitates precise monitoring of gas composition changes during casting, optimizing quality control and process efficiency in metallurgical applications.
Figure 10: Schematic illustration of the test casting design The cavity dimensions are representative of the core All dimensions provided are in inches (Barlow et al., 1997)
In this study, 50 lb (22.68 kg) of commercially available three-screen, 57 GFN silica sand was prepared by mixing in a vibratory mixer The molds were fabricated using a commercial ester-cured phenolic resin system, with three different resin contents tested at 1%, 1.25%, and additional varying levels This research aims to evaluate the impact of resin content on mold properties and performance, providing valuable insights for optimizing silica sand mold manufacturing in foundry applications.
The mixture, consisting of 1.5% additive based on sand weight, was carefully placed into the test mold pattern After stripping the molds from the pattern, they were left idle for 24 hours to ensure proper curing before further testing.
For preparing chromite core samples, 3” x 5” cylindrical test cores were created using an ester-cured phenolic resin system similar to the mold material High-purity chromite sand, with less than 0.5% quartz content, replaced silica sand to ensure minimal impurities The sand was split with a 16-way splitter to obtain a representative sample, then mixed with the resin co-reactant at 30% based on resin weight in a desktop mixer for 60 seconds, followed by flipping for thorough homogenization Three resin contents (1%, 1.25%, and 1.5%) were used to match those in mold preparation, with the resin added accordingly and mixed again The mixture was poured into core patterns and cured until strip time was reached before being stored in a desiccator for 24 hours, ensuring optimal preparation for subsequent testing.
Determination of Core Atmosphere CO/CO2 Ratio during Pouring and Solidification
An 8mm (0.315in) stainless steel sample extension tube was placed within the test core of the assembled mold, 0.5in (12.7mm) from the mold-metal interface (see Figure
11) The molds were cast with WCB steel at 2900℉ (1593℃), and gases produced during casting were drawn into the sampling line at a rate of 1 liter per minute and analyzed for
Using the Testo 350 portable flue gas analyzer, CO, CO2, and O2 levels are measured to assess combustion efficiency A polyethylene filter in the sampling line ensures accurate readings by eliminating moisture and particulate matter from the system An extension tube equipped with a Type-K thermocouple positioned 0.5 inches (12.7 mm) from the mold-metal interface captures precise temperature data This thermocouple is connected to a data acquisition unit, enabling temperature-dependent analysis of gas concentrations.
Figure 11: Testing arrangement for collection of CO/CO2 emissions during casting
The oxygen fugacity was then calculated using the CO2/CO ratio results according to the following formula (Fegley, 2012):
● 𝑋 𝐶𝑂 2 is the carbon dioxide content
● 𝑋 𝐶𝑂 is the carbon monoxide content
The equilibrium constant, 𝐾𝑝, is expressed as 𝐾𝑝 = e^(-ΔG°/RT), where ΔG° represents the Gibbs free energy change associated with the fayalite reaction This free energy change is defined by the equation ΔG° = -39,140 + 15.59T J/mol, as reported by Jacob, Kale, & Iyengar (1989) The gas constant R is 8.31441 J/K·mol, and T denotes the temperature in Kelvin Understanding these thermodynamic parameters is essential for accurately determining the equilibrium state of mineral reactions Additionally, the study involves the determination of the specific heat capacity for high-purity chromite sand, a key factor in assessing its thermal properties.
Using a differential scanning calorimeter with the IsoStep DSC temperature program, the specific heat capacity of bonded chromite samples containing 0.17% silica was measured The IsoStep DSC method is preferred for assessing specific heat during chemical reactions and physical transitions, as it dynamically increases temperature at 15℃ per minute with isothermal holds every 200°C to ensure thermal equilibrium This approach requires three tests per sample: a blank with an alumina crucible, a calibration sample with a lid to account for mass differences, and the actual bonded chromite sample sealed in the same crucible Fourier analysis is used to evaluate the amplitude of each sample, allowing precise determination of specific heat capacity as a function of temperature through established formulas.
● CpAl is the specific heat capacity of the alumina
● As is the amplitude associated with the sample
● AAl is the amplitude associated with the alumina
● Ab is the amplitude associated with the blank
● mAl is the mass of the alumina
● m is the mass of the sample
Determination of Sinter Temperature of Chromite Sand with Different Concentrations of
A methodology proposed by the University of Northern Iowa (Thiel & Ravi,
Research by Tardos, Mazzone, and Pfeffer (1984), built upon in 2014, focused on measuring surface softening of granular materials at high temperatures Thiel explains that surface viscosity is determined by analyzing sand particle compaction under a compressive load during controlled heating, highlighting the porosity of sand which initially expands with heat before contracting due to sintering at contact points This sintering results from surface softening and deformation at the confinement points of pressure Additionally, these soft particles are modeled as Newtonian fluids, enabling Thiel and Ravi (2014) to define surface viscosity based on this behavior, providing a crucial understanding of high-temperature granular material dynamics.
Using a methodology similar to Ravi et al (2018), the high-temperature characteristics of silica sand-doped chromite samples were analyzed with a high-temperature aggregate dilatometer The cylindrical samples, approximately 1.6 inches (4.06 cm) in height and 1.1 inches (2.8 cm) in diameter, were prepared with a commercial ester-cured phenolic resin system at 1.5% addition by weight Sample deformation was recorded during heating at a rate of 2℃ (3.6℉) per minute, providing data to calculate surface viscosity and determine the sinter temperature.
Calculation of Surface Viscosity and the Associated Sintering Temperature
The deformation recorded from the dilatometer was used to calculate the viscosity of the bonded samples The following formula was used to calculate the viscosity (Thiel
The coefficient ‘n’ represents the slope of Δl versus time on a log-log plot, consistently determined to be between 0.47 and 0.5 across experiments Assuming ‘n’ equals 0.5 simplifies the governing formula to (Thiel & Ravi, 2014): ηs = KFpDp -2 / [∂ (fΔl/2lo) / ∂t], providing a reliable basis for analyzing the material's behavior.
● The denominator is the slope of the deformation versus time
● Dp is the particle diameter,
The inter-particle compression force, denoted as Fp, is determined using the formula Fp = (4ε Dp² L) / (π (1 - ε) Ds²), where ε represents the porosity of the sample, L is the applied load, and Ds is the diameter of the sample holder This equation highlights the relationship between porosity, load, and sample dimensions in calculating the inter-particle forces within the material Understanding Fp is essential for analyzing the compression behavior and mechanical properties of porous samples in various engineering and material science applications.
● Δl/lo is the linear expansion/contraction measured from the dilatometer
● The coefficient f is given by 3β/2 o β is the layer spacing and is given by β= √ (2/3) * {π/[3√2 (1-ε)]} 1/3
Quartz content in the samples was accurately measured using XRF spectrometry, which involved pulverizing the specimens into a fine powder with a mortar and pestle The prepared powder was then placed into the spectrometer’s sample holder, and quartz concentration was determined through analysis with three-point calibration curves, ensuring precise and reliable results.
RESULTS AND DISCUSSION
The data from the initial sample containing 1% ester-cured phenolic resin showed a sharp increase in both carbon monoxide (CO) and carbon dioxide (CO₂) levels immediately after pouring, with peaks occurring around 200 seconds The CO concentration was notably higher, reaching a maximum of approximately 140 ppm, whereas CO₂ levels barely exceeded 40 ppm Both gas concentrations stabilized after approximately 1600 seconds, indicating the solidification process's impact on gas emissions during resin curing.
Figure 12 CO and CO2 concentrations measured from the 1% ester-cured phenolic chromite sample
Using the concentration data from the 1% ester-cured phenolic sample, the
The CO2/CO ratio was monitored over time, as shown in Figure 13, revealing a significant increase immediately after pouring The ratio peaked at approximately 250 seconds, indicating the maximum amplitude, and then gradually stabilized thereafter.
The median stabilization time observed in the concentration data between carbon monoxide and carbon dioxide is approximately 1300 seconds This result closely aligns with findings reported by Barlow et al (1997), particularly the resin-bonded sample depicted in Figure 6 (D) Notably, the ratio exceeding 0.2 suggests that iron oxidation may have occurred at the metal-interface during this trial, highlighting key insights into the oxidation process and material interactions.
Figure 13 CO2/CO ratio calculated from the concentration data measured from the 1% ester-cured phenolic sample
The 1.25% ester-cured phenolic sample exhibits a similar trend, as shown in Figure 14, with slight increases in the amplitude of carbon monoxide and carbon dioxide curves, reaching approximately 180 ppm and 60 ppm respectively The peak concentrations were maintained for about 200 seconds, indicating a prolonged emission period After 300 seconds, both gases gradually decreased in concentration and stabilized around 1500 seconds, demonstrating the different emission dynamics of this cured phenolic sample.
Figure 14: CO and CO2 concentrations measured from the 1.25% ester-cured phenolic chromite sample
The CO2/CO ratio results were similar across samples, with a slightly higher peak value of 0.32 in this trial, likely caused by minor variations in pouring temperature The sand sample exceeded the maximum temperature achieved in the 1% trial, leading to a stabilized ratio of 0.07 compared to 0.03 in the previous sample Despite these differences, both samples stabilized at similar times after pouring commenced, indicating consistent reaction dynamics.
Figure 15: CO2/CO ratio calculated from the concentration data measured from the
The gas concentration data for the final trial with 1.5% ester-cured phenolic resin is illustrated in Figure 16, revealing slight variations in peak amplitudes compared to other samples Carbon monoxide levels peaked at approximately 160 ppm around 180 seconds after pouring, while carbon dioxide showed a similar trend with a peak of 45 ppm Interestingly, the 1.25% sample displayed higher amplitudes than the 1.5% sample, an unexpected finding that suggests further investigation into curing effects and gas emissions.
Figure 16: CO and CO2 concentrations measured from the 1.5% ester-cured phenolic chromite sample
The CO2/CO ratio for the final 1.5% ester-cured phenolic sample was analyzed, with results shown in Figure 17 The peak ratio reached 0.28, indicating a response similar to the concentration data for this sample After reaching stabilization at 1200 seconds, the ratio decreased to 0.07, displaying a trend comparable to the 1.25% sample.
Figure 17: CO2/CO ratio calculated from the concentration data measured from the 1.5% ester-cured phenolic sample
The CO2/CO ratio was used to calculate the temperature-dependent oxygen fugacity based on the experimental methodology These fugacity values were plotted logarithmically against temperature, revealing their relationship as shown in Figure 18 The data was compared with the fayalite stability region, following the approach of Fisler and Mackwell (1993), with dashed lines indicating extrapolated boundaries of the QFI and QFM phases.
Despite increases in total binder content, oxygen fugacity remained constant across all samples Using logarithmic regression curves on three datasets, the intersection point between fugacity series and the extrapolated QFM phase boundary was determined mathematically The average intersection temperature for the three samples was 1434℃, indicating a consistent phase transition point.
Figure 18: The relationship between the logarithmic oxygen fugacity and sand temperature with relation to the fayalite stability field defined by O’Neill (1987)
Differential Scanning Calorimetry Specific Heat Capacity Results
The specific heat capacity of the high-purity chromite sample was approximately 650 J/kg∙℃ (157.9 Btu/lb∙℉) at room temperature, as shown in Figure 19 It steadily increased with temperature up to 925℃ (1697℉), reaching a peak value of 1526 J/kg∙℃ (364.47 Btu/lb∙℉) After reaching this maximum, the specific heat capacity decreased until the end of the trial at 1600℃.
Figure 19: Specific heat capacity results presented as a function of temperature for the high-purity chromite sample
The dilatometry analysis revealed how quartz content influences linear expansion, surface viscosity, and sinter temperature in ceramic samples XRF spectroscopy determined the quartz percentages, with the baseline sample containing only 0.17% quartz, similar to the material used in heat capacity testing Subsequently, samples were intentionally contaminated with higher quartz levels—0.84%, 1.78%, and 2.56%—covering most of the industry’s quartz content range (up to 3%) These findings highlight the correlation between increased quartz content and changes in key properties relevant to ceramic processing and performance.
Table 4: Quartz concentration results of linear expansion samples obtained through XRF
Quartz Concentration Results from XRF
The linear expansion behavior of bonded chromite sand samples as a function of temperature closely aligns with previous findings reported by Ravi et al (2018) All samples exhibited similar expansion patterns up to 400℉ (204℃), after which their behaviors diverged The high-purity sample with 0.17% contamination demonstrated superior thermal stability, reaching peak expansion around 2200℉ (1204℃) before rapidly contracting at 2550℉ (1339℃) In contrast, the 0.84% and 1.78% contaminated samples behaved similarly up to 1000℉ (538℃), but beyond this point, the 0.84% sample continued to expand at a higher rate, peaking at around 2100℉ (1149℃) with double the amplitude of the high-purity sample Increased quartz contamination in the samples led to a decrease in contraction temperatures, with rapid contraction occurring at approximately 2200℉ (1204℃), 2000℉ (1093℃), respectively, indicating that contamination adversely affects the thermal stability and expansion characteristics of chromite sand samples.
Samples containing 0.84%, 1.78%, and 2.56% quartz exhibited peak expansion at 1800℉ (982℃) Notably, the 2.56% quartz sample demonstrated a more pronounced expansion profile at lower temperatures compared to the other trials Additionally, this sample maintained its maximum expansion over a 400℉ (204℃) temperature increase, indicating higher thermal stability and expansion resilience.
Figure 20: Dilatometry results for chromite samples with different concentrations of quartz
Surface Viscosity and Associated Sinter Temperature Results
Using dilatometry data, surface viscosity was calculated for each sample, revealing peaks that indicate surface softening of the aggregate at contact points The data shows that increasing quartz content causes surface softening to occur at progressively lower temperatures For example, the high-purity control sample with 0.17% silica sintered at 2660℉ (1460℃), while the sample with 2.56% quartz sintered at 1969℉ (1076℃).
Figure 21: Surface viscosity and associated sinter temperature results for chromite samples
The sinter temperature increased linearly with quartz concentration, as shown in Figure 22, with equations provided for both imperial and metric units A strong correlation was observed, indicated by a high coefficient of determination (R² = 0.9646), demonstrating that quartz concentration significantly influences sintering temperature This dependency highlights the variability in performance characteristics among different chromite sand distributors, emphasizing the importance of controlling quartz levels for consistent sintering outcomes.
Figure 22: The relationship between sintering temperature and quartz content within chromite sand