Interfacial contact topology governing molten iron–slag dripping in coke-packed beds under hydrogen-enriched blast furnace conditions

Technical Summary: Interfacial Contact Topology Governing Molten Iron–Slag Dripping in Coke-Packed Beds under Hydrogen-Enriched Blast Furnace Conditions

Problem Statement
The iron and steel industry is transitioning toward hydrogen-enriched blast furnace operations to reduce carbon emissions. While the effects of hydrogen on reduction kinetics and phase transformation are well-documented, the impact of these changing conditions on the coupled flow of molten iron and slag through the lower coke-packed bed remains insufficiently quantified. In hydrogen-enriched environments, the degree of reduction of iron-bearing materials varies, altering the FeO content in the slag, the iron-to-slag volume ratio, and the physical properties of the melts (viscosity, density, surface tension, and interfacial tension). Previous studies often assumed fixed slag compositions or liquid supply conditions, failing to account for how reduction-dependent property changes might shift the governing mechanism of slag drainage from a viscosity-controlled regime to one dominated by interfacial contact morphology and liquid hold-up.

Methodology
The study employed three-dimensional multiphase Smoothed Particle Hydrodynamics (SPH) simulations to model the dripping behavior of molten iron and FeO–CaO–SiO₂–Al₂O₃–MgO slag within a packed bed of irregular coke particles.

• Physical Property Estimation: Thermodynamic calculations (using FactSage) were used to determine slag compositions and liquid fractions based on three degrees of reduction ($R = 0.70, 0.90, 1.00$) and temperatures (1300–1500 °C). Key properties—density, viscosity, surface tension, and iron–slag interfacial tension—were estimated as functions of these variables. Notably, molten iron properties were held constant to isolate the effects of slag property variations.
• Simulation Setup: The coke bed was modeled with an average particle diameter of 0.0237 m. Molten phases were supplied intermittently from the top inlet. The gas phase was treated as an empty void to isolate hydrodynamic and interfacial effects.
• Analysis Metrics: The study quantified phase holdup, center-of-mass dripping velocities, apparent contact areas (between iron, slag, coke, and void), and an iron–slag mixing index. A regression analysis was performed to identify the descriptors governing the relative dripping velocities of the two phases.

Key Results

1. Dual Regimes of Dripping Behavior: The simulations revealed two distinct regimes governing slag drainage:
   • Viscosity-Controlled Regime ($R = 1.00$): Under fully reduced conditions, slag viscosity was high (750–5500 mPa·s). In this regime, increasing temperature reduced viscosity, directly increasing the slag dripping velocity. The iron and slag phases remained spatially separated due to high interfacial tension.
   • Morphology-Controlled Regime ($R = 0.70, 0.90$): Under lower reduction degrees, slag viscosity was low ($\le 110$ mPa·s), and the slag volume fraction was high. Counter-intuitively, lower viscosity did not always lead to faster dripping. At $R=0.70$ and 1500 °C, the slag exhibited the lowest viscosity but the slowest apparent dripping velocity. This was attributed to extensive slag–coke contact, where low-viscosity slag spread over the coke surface or was retained within the pore network rather than draining vertically.
2. Role of Interfacial Topology: A regression model using only slag viscosity, interfacial tension, and volume ratio failed to predict the slag-to-iron velocity ratio ($R^2 = 0.148$). However, incorporating the ratio of slag–coke to iron–coke contact areas significantly improved the model's accuracy ($R^2 = 0.943$). This demonstrates that the geometric contact state between the melt and the coke is a primary determinant of drainage behavior in low-viscosity, slag-rich conditions.
3. Iron–Slag Mixing: The specific iron–slag contact area and mixing index decreased as the degree of reduction increased. Lower reduction degrees ($R=0.70, 0.90$) promoted higher iron–slag contact due to larger slag fractions, while fully reduced conditions ($R=1.00$) resulted in distinct, phase-separated flow structures. Interestingly, at $R=0.70$ and 1500 °C, despite high fluidity, the mixing index dropped significantly, suggesting that enhanced fluidity promoted spreading on coke surfaces rather than maintaining contact with molten iron.

Significance and Claims
The paper claims that interfacial contact topology significantly influences melt drainage and reaction area in low-carbon blast furnace operations. The primary contribution is the demonstration that slag dripping behavior cannot be explained solely by viscosity, particularly under hydrogen-enriched conditions where reduction degrees vary.

• The study establishes that in low-viscosity, slag-rich regimes, the apparent dripping velocity is governed by the morphological interaction between the slag and the coke bed (specifically the slag–coke contact area) rather than viscous resistance alone.
• It provides a framework connecting the degree of reduction, slag properties, liquid-phase fraction, and interfacial contact morphology.
• The authors conclude that evaluating dripping-zone permeability and liquid hold-up in hydrogen-enriched blast furnaces requires combined descriptors (such as slag–coke contact area and iron–slag mixing) rather than relying on slag viscosity alone.

Limitations
The authors note several limitations: the dripping velocity is an apparent migration velocity based on center-of-mass motion rather than a true cross-sectional flux; the contact area is approximated using cubic cells rather than true curved interfaces; coke–liquid wetting was simplified to a non-wetting condition (180° contact angle); and molten iron properties were assumed constant. Future work is suggested to examine the effects of molten iron surface tension, wetting variations, and solid fractions.