Which Bricks Are Best Suited For RH Furnaces: Magnesium-spinelle Bricks Or Magnesium-chrome Bricks?

RH furnaces play key roles in processes such as natural decarburization, composition adjustment, temperature elevation of molten steel, decarburization through oxygen injection, and desulfurization via powder injection. They have become essential equipment for refining high-quality steels such as electrical steel, pipeline steel, steel for shipbuilding plates, galvanized sheets, and cold-rolled sheets. During operation, the furnace maintains a vacuum state inside. Molten steel and argon flow at high speeds in a fountain-like pattern through a closed piping system that includes the upward insertion tube, lower trough, and downward insertion tube. At the same time, significant amounts of alloying elements or desulfurizing agents are added to adjust the composition of the molten steel. This imposes strict requirements on the stability under vacuum, high-temperature performance, erosion resistance, and slag resistance of refractory materials.

As intermittent metallurgical equipment, RH furnaces subject the insertion tubes and lower troughs to frequent thermal shocks caused by the alternating hot and cold conditions throughout their lifespan. Therefore, refractories must also exhibit excellent thermal shock resistance. Extensive operational experience has shown that acidic, neutral, and carbon-containing refractories are not suitable for RH furnaces. Only high-temperature sintered magnesium-chrome bricks are deemed appropriate as lining materials. However, the Cr₂O₃ present in magnesium-chrome bricks reacts with alkaline metal oxides to form hexavalent chromium compounds (R₂CrO₄). Under conditions where sulfur, chlorine, and alkalis are simultaneously present, solid solutions of R₂(Cr·S)O₄ may also form. These two compounds are toxic substances soluble in water. Whether released into the atmosphere or retained in demolished bricks after use, the Cr⁶⁺ infiltrating the environment via rainwater poses serious risks to humans, animals, and plants.

With increased environmental awareness and growing emphasis on pollution control, several major national steel mills have recently adopted chrome-free magnesium-spinelle bricks in their oxygen converters, achieving favorable operational results. However, many steel mills are still unfamiliar with magnesium-spinelle bricks. Therefore, it is necessary to compare the performance characteristics of magnesium-chrome and magnesium-spinelle bricks for oxygen furnaces in order to provide a reference for steel mills when choosing refractory materials for oxygen furnace applications.

01Test

Magnesium-chromium bricks (designated MC, fired bricks) from the lower trough of an RH furnace at a major domestic steelworks, prior to chromium-free modification, and magnesium spinel bricks (designated MS, unfired bricks) following chromium-free modification, were selected for testing. Spectrometer analysis determined the chemical composition of the MC and MS samples, while X-ray diffraction was used to analyze their phase composition. Testing was conducted according to the standards GB/T2997-2000, GB/T5072-2008, GB/T5988-2007, and GB/T3002-2004 to determine the apparent porosity, bulk density, compressive strength, thermal linear change rate, and high-temperature flexural strength (1450°C for 1 hour) of the MC and MS samples after heat treatments at 110°C for 24 hours, 1000°C for 3 hours, and 1600°C for 3 hours, respectively.

In accordance with standard GB/T7320-2008, the thermal expansion coefficient from room temperature to 1500°C was measured using an RPZ-04 high-temperature thermal expansion analyzer. Thermal shock resistance was evaluated according to standard YB/T 376.3-2004 (water-cooling method) at temperatures ranging from room temperature to 1100°C. Specimens were prepared to standard brick dimensions, and surface crack morphology was compared after thermal shock testing.

The slag resistance was evaluated according to standard GB/T8931-2007 (static crucible slag resistance method), with a comparison of the erosion morphology of residual bricks removed from the lower trough wall of the RH furnace. The chemical composition (wt%) of the test slag was as follows: SiO₂: 11.50%, Al₂O₃: 36.08%, Fe₂O₃: 10.15%, CaO: 22.50%, MgO: 7.71%, TiO₂: 0.76%, Cr₂O₃: 0.16%, MnO: 9.97%. Scanning electron microscopy (SEM) was employed to examine the microstructure at the reaction interface between the static crucible specimens and the slag. Energy dispersive spectroscopy (EDS) was used to determine the micro-area composition at the reaction interface between the slag and the brick, thereby analyzing the respective erosion mechanisms.

02 Results and Discussion

2.1 Chemical Composition and Phase Composition

The chemical composition and XRD patterns of the MC and MS specimens are presented in Table 1 and Figure 1, respectively. It can be observed that the main constituents of the magnesium-chromium brick (MC) are MgO, Cr₂O₃, Fe₂O₃, and Al₂O₃, with the primary phases being periclase and (Mg, Fe)O·(Al, Cr)₂O₃. The main constituents of the magnesium spinel brick (MS) are MgO and Al₂O₃, with the primary phases being periclase, spinel, and metallic aluminum.

2.2 Physical Properties

The physical properties of magnesium-chromium bricks (MC) and magnesium spinel bricks (MS) after different temperature treatments are shown in Table 2. As indicated in Table 2:

The apparent porosity of the MC specimens (both as-made and after treatment at 1000°C for 3 hours and 1600°C for 3 hours) is significantly greater than that of the MS specimens, while their strength is considerably lower. Due to differences in composition, the bulk density of the MC specimens exceeds that of the MS specimens. Since the MC specimens are fired bricks and the MS specimens are unfired, the linear expansion rate of the MC specimens after treatment at 1000°C for 3 hours and 1600°C for 3 hours is lower than that of the MS specimens. The flexural strength of the MC specimens at 1450°C is considerably lower than that of the MS specimens.

The strength of Specimen MC is lower than that of Specimen MS, primarily due to the less compact microstructure of Specimen MC compared to Specimen MS. Furthermore, the matrix of Specimen MS contains a certain amount of metallic aluminum powder, which forms a metal-ceramic bond during heating. This significantly enhances the high-temperature strength of Specimen MS.

2.3 Thermal Expansion Coefficient

The thermal expansion rate curves of magnesium-chromium brick (MC) and magnesium spinel brick (MS) from room temperature to 1500°C are shown in Figure 2. It can be observed that the linear expansion rate of Sample MC steadily increases with rising temperature, reaching a maximum of 1.78% at 1500°C. This is primarily due to the inherent thermal expansion and contraction properties of the material. For Sample MS, in addition to the inherent thermal expansion and contraction of the material, sintering reactions within its matrix also occur. Prior to 600°C, its linear expansion rate remains comparable to that of Sample MC. However, beyond 600°C, the linear expansion rate of Sample MS begins to exceed that of Sample MC due to the sequential occurrence of aluminum oxidation reactions and spinel formation reactions between Al₂O₃ and MgO within the matrix. This causes the linear expansion rate of Sample MS to surpass that of Sample MC. The increase culminates at 1500°C, where the maximum value of 1.96% is reached.

The residual linear expansion rate and high-temperature linear expansion rate of magnesia spinel bricks after high-temperature treatment are both greater than those of magnesia-chromium bricks. Therefore, larger expansion joints must be provided during installation compared to magnesia-chromium bricks to prevent spalling due to excessive expansion during service.

2.4 Thermal Shock Resistance

The morphology of magnesium chromite bricks (MC) and magnesium spinel bricks (MS) after undergoing 1, 4, and 7 cycles of 1100°C water-to-hot thermal shock testing is shown in Figure 3. As illustrated in Figure 3: After one thermal shock cycle, neither the magnesium-chromium brick nor the magnesium spinel brick exhibited surface cracks. After four thermal shock cycles, the surfaces of both bricks developed reticular and dendritic cracks, respectively, with magnesium-chromium bricks exhibiting a higher prevalence of cracks. After seven thermal shock cycles, the number of cracks on both magnesia-chrome bricks and magnesia-spinell bricks stopped increasing. However, the width of the cracks expanded to varying degrees, with magnesia-chrome bricks showing the most pronounced increase.

Magnesium spinel bricks exhibit superior thermal shock resistance, due to the plastic bonding of metallic aluminum powder within their matrix and the ceramic bonding formed by in-situ spinel reactions.

2.5 Resistance to Slag Attack

Cross-sectional photographs of magnesium-chromium bricks (MC) and magnesium spinel bricks (MS) after 3-hour static crucible resistance testing at 1600°C are shown in Figure 4. The intact crucible walls indicate that both MC and MS samples exhibited good resistance to slag erosion. However, based on the residual slag volume within the crucible, the MC sample demonstrated poorer resistance to slag penetration compared to the MS sample.

Samples of residual magnesia-chromite bricks (MC) and residual magnesia-spinel bricks (MS) were collected from the lining removed during the replacement of the immersion tube and excavation/repair of the lower channel beneath the RH furnace at a certain steelworks. As illustrated in Figure 5 (original thickness: 300 mm, with surface layers removed to reveal a clear slag-brick reaction interface), both MC and MS specimens were examined.

After 100 furnace cycles, the average wear rates, calculated from the residual thickness, were 0.9 mm per furnace for MC and 1.0 mm per furnace for MS, respectively. It was further observed that sample MC exhibited a through-crack parallel to the working surface, approximately 40 mm from the working face, whereas sample MS showed no such destructive cracking. It is inferred that sample MC was prone to slag penetration, forming a metamorphic layer. The significant difference in thermal expansion coefficients between this layer and the original brick layer led to markedly disparate stresses during the RH furnace's frequent thermal cycling operations. Consequently, the through-crack manifested at a critical point. The emergence of this through-crack in sample MC ultimately resulted in structural spalling during subsequent use, accelerating its rate of deterioration.

Microstructural photographs of specimens MC and MS following static crucible slag resistance testing are shown in Figures 6 and 7, respectively. The figures reveal that:

(1)Specimen MC exhibits no discernible reaction interface (Figure 6(a)), with the molten slag having largely infiltrated the interstitial spaces between its aggregate and matrix to form a relatively dense alteration zone. Conversely, the aggregate and matrix within the original brick layer appear loose and porous (Figure 6(b)).

Within the altered layer, two primary phases were identified. Energy dispersive spectroscopy (EDS) analysis of the white material at point "3" revealed the following main components (wt%): MgO 1.51%-3.68%, Al₂O₃ 15.75%-26.22%, SiO₂ 2.42%-4.53%, CaO 41.80%-48.65%, TiO₂ 2.17%-4.31%, Cr₂O₃ 0.91%-1.57%, MnO 3.27%-6.66%, FeO+Fe₂O₃ 18.37%-19.88%. It is inferred that the phases primarily consist of low-melting-point phases, such as calcium aluminate and calcium ferrite.

Energy dispersive spectroscopy indicates that the main composition (wt%) of the grey material at point "4" is: MgO 27.30%-27.90%, Al₂O₃ 61.96%-68.37%, Cr₂O₃ 0.61%-3.26%, MnO 1.07%-1.69%, FeO+Fe₂O₃ 1.71%-5.55%. This confirms that the phase is primarily magnesium-aluminum spinel, with a composite spinel phase doped with manganese, iron, and chromium elements within its lattice.

(2) Sample MS exhibits a distinct reaction interface, with a dense layer approximately 0.5 mm thick forming at this interface. This layer consists of magnesia aggregate and numerous magnesia-alumina spinel crystals of varying sizes, as shown in Figure 7(a). The slag layer primarily consists of two phases, "5" and "6". Energy dispersive spectroscopy revealed the main constituents (wt%) of point "5" as follows: MgO 25.26%–27.14%, Al₂O₃ 64.69%–68.75%, MnO 1.45%–3.68%, FeO + Fe₂O₃ 2.28%–5.20%, CaO 0–0.60%, confirming that its phase is primarily magnesium-aluminium spinel, with manganese and iron-doped composite spinel phases in its lattice. Energy dispersive spectroscopy also revealed the main components (wt%) at point "6" as follows: MgO 2.04%–2.84%, Al₂O₃ 40.07%–47.96%, SiO₂ 5.88%–8.72%, CaO 37.11%–42.02%, TiO₂ 0–1.44%, MnO 0.68%–1.81%, FeO + Fe₂O₃ 2.46%–6.79%. This suggests that the phase is predominantly composed of low-melting-point phases, such as calcium aluminate and calcium ferrite. Observation of the original brick layer after the MS slag resistance test, as shown in Figure 7(b), reveals that it still exhibits a dense structure, where the spinel matrix tightly encapsulates the magnesia aggregate.

Comparing the physical phases and chemical compositions at the reaction interface between samples MC and MS reveals that, apart from the absence of Cr₂O₃ in sample MS, the two are fundamentally similar. However, the microstructural morphology at their reaction interfaces differs markedly. This discrepancy stems from the significantly higher apparent porosity of the original MC brick, which facilitates slag penetration into its matrix. Although a high-melting-point composite spinel phase formed during the reaction, rendering the slag viscous, it failed to consolidate into a cohesive structure. Consequently, it proved ineffective in halting further slag infiltration. Conversely, the original MS brick exhibited very low apparent porosity, making it inherently difficult for steel slag to infiltrate its matrix. Moreover, the composite spinel formed at the reaction interface readily aggregated and fused, creating a dense layer with a high melting point. This layer effectively prevented further slag penetration.

03 Conclusions

(1)Magnesium-chromium bricks are fired bricks with relatively high porosity in the finished product. Their room-temperature strength after treatment at 1000°C and 1600°C, as well as their high-temperature flexural strength at 1450°C, are both low. The residual linear expansion rate after high-temperature treatment and the hot linear expansion rate at elevated temperatures are also low.

(2)Magnesium spinel bricks are unfired bricks with very low finished porosity. Their room-temperature strength after treatment at 1000°C and 1600°C, and their high-temperature flexural strength at 1450°C, are both high. Due to the presence of in-situ spinel reactions, their residual linear expansion rate after high-temperature treatment and their linear expansion rate at elevated temperatures are relatively high. Consequently, appropriate expansion joints must be provided during masonry construction.

(3)Magnesium spinel bricks exhibit superior thermal shock resistance compared to magnesium-chromium bricks due to the composite metallic plastic bonding and ceramic bonding within their matrix.

(4)Both magnesia-chromium bricks and magnesia-spinel bricks exhibit good resistance to slag erosion. However, magnesia-chromium bricks possess relatively high apparent porosity, resulting in poorer resistance to slag penetration. Magnesia-spinel bricks, with their lower porosity, can form a high-melting-point dense layer composed of magnesia aggregate and spinel grains at the slag-brick reaction interface, thereby offering superior resistance to slag penetration.

(5)Magnesia spinel bricks demonstrate potential to replace magnesia-chromium bricks as the next-generation environmentally friendly refractory lining material for RH furnaces.