Microstructural Design Optimisation of Magnesia-Chromite Bricks for RH Degassers: Determining an Optimal Balance Between Thermal Shock and Corrosion Resistance

Magnesia-chromite refractories are widely used in various industries due to their exceptional thermal stability, minimal thermal expansion, and resistance to thermal shock. They are particularly favoured in applications such as RH degassers, rotary cement kilns, and nonferrous copper converters [1]. Compared to magnesia-based counterparts, they offer superior thermal shock resistance and corrosion resistance to acidic and basic slags, especially fayalitic slags [1–3]. To improve slag penetration resistance, it is crucial to densify the bricks, reduce apparent porosity, while maintaining thermal shock resistance. This can be achieved through microstructure design, focusing on fostering direct bonding between the refractory grains, including secondary spinel phases and fused magnesia-chromite (FMC), along with a carefully tailored grain size distribution (GSD).

The principal constituents of magnesia-chromite bricks, customised to specific applications, include sintered magnesia, fused magnesia, chrome ore, and FMC. These raw material compositions facilitate categorising magnesia chromite bricks into three discrete classes: Direct-bonded, semi-rebonded, and rebonded. Each classification presents distinct advantages and constraints, requiring detailed alignment with the specific requisites of the intended application. Notably, direct-bonded bricks exhibit superior thermal shock resistance, but comparatively inferior corrosion resistance compared to rebonded bricks, with the semi-rebonded class  filling an intermediate position within this range.

In direct-bonded bricks, the bonding mechanisms between sintered magnesia and chrome ore grains are determined by temperature-driven reactions during the firing process. At 1500 °C, a liquid phase forms around the chrome ore grains, subsequently infiltrating the adjacent magnesia matrix, enabling diffusion of Fe₂O₃, Cr₂O₃, and Al₂O₃ from the chrome ore into the magnesia at 1670 °C. Finally, at 1750 °C, complete dissolution of the chrome ore occurs, precipitating spinel-like exsolution phases within the MgO structure [4]. This behaviour is beneficial for thermal shock resistance owing to the microcracking induced by interdiffusion of chrome ore components into the magnesia matrix.

In the context of RH degassers, magnesia-chromite bricks primarily comprise FMC and are typically classified as rebonded bricks. The main bonding process occurs during the FMC electrofusion process, followed by firing the bricks at temperatures exceeding 1700 °C to form rebonded refractories. The FMC consists mainly of periclase and a solid solution spinel [(Fe,Mg)(Fe,Al,Cr)₂O₄], resulting from the fusion of various spinel structures (AB₂O₄-type phases), where A represents a divalent cation and B a trivalent one [5–7]. This complex structure, characterised by interstices and vacancies, enables the formation of solid solutions with broad solubility ranges and ionic radii ranging from 0.44– 1.00 Å [5,8]. Consequently, these bricks exhibit excellent resistance to corrosion and slag infiltration. Additionally, preferentially oriented microcracks within the FMC grains enhance the material’s flexibility and resistance to thermal shock damage. The FMC features a periclase matrix with embedded spinel precipitates, typically around 15 μm in size, exhibiting a consistent pattern with {110} planes as revealed by electron backscatter diffraction, aligning with the Zener-Stroh crack mechanism [9]. Figure 1 provides a visual representation of the FMC parameters described.

Laboratory magnesia-chromite bricks were formulated utilising Alfred’s model incorporating a customised particle size distribution coefficient tailored to the FMC content in the semi-rebonded formulation under investigation. Additionally, a novel GSD was developed and assessed to enhance the key properties of these materials during application. A spectrum of compositions ranging from low to high FMC content was systematically explored (Table I). To ensure comparison across all FMC compositions and mitigate any potential interference during firing, uniform binder systems and sintering additives were incorporated across all compositions.

To assess the impact of different levels of FMC and the two GSDs on bulk density and apparent porosity, 3 samples of each magnesia-chromite brick composition were tested. Based on the results obtained, it was apparent that increasing the FMC content led to an overall rise in bulk density, coupled with a decline in apparent porosity for both GSDs examined (Figure 2). Furthermore, alongside the benefits of increasing the high-grade raw material content on these properties, the new GSD emerged as a pivotal factor for increasing bulk density and decreasing porosity, apart from at the lowest level of FMC addition.

To visually examine the different brick compositions, optical microscopy was performed and micrographs of compositions featuring the standard GSD are provided in Figure 3. The images clearly demonstrate how each composition’s microstructure changes with the addition of higher FMC levels; most notably the porosity decreases, leading to improved densification.

In order to explore the influence of FMC content and GSD on cold mechanical properties, the cold crushing strength and cold modulus of rupture were measured for 3 samples of each magnesia-chromite brick composition. The results presented in Figure 4 show that both properties increased with FMC content for both GSDs. However, regarding the cold crushing strength, discernible differences in values between the two GSD only emerged from composition D onwards, with the magnitude of difference increasing with FMC content. In summary, it appeared that the novel GSD effectively optimised microstructural packing, resulting in highly favourable outcomes for practical applications when combined with higher amounts of FMC. In contrast, a similar trend was not observed for the cold modulus of rupture. Although at the lowest FMC content no distinction was observed between the GSDs, for all other compositions the standard distribution exhibited a higher cold modulus of rupture compared to the new GSD.

To assess the influence of FMC addition and GSD on thermomechanical properties, the hot modulus of rupture test was conducted on 3 samples of each brick composition at test temperatures of 1250 °C, 1400 °C, and 1485 °C (Figure 5). At 1250 °C, the hot modulus of rupture showed an upward trend with the increasing addition of FMC, irrespective of the GSD, indicating incorporation of FMC had a beneficial effect on this thermomechanical property, while the GSD did not significantly influence the obtained values. With an elevation of the test temperature to 1400 °C, the hot modulus of rupture was lower than at 1250 °C, but also rose with increasing amounts of FMC, in a similar manner for both GSDs, until a certain content was reached (i.e., composition D), whereupon this upward trend stabilised. It is established that at 1400 °C, liquid phase formation originating from in situ generated silicate minerals (e.g., monticellite and merwinite) is more prevalent in the material’s microstructure, which would account for the observations. At a testing temperature of 1485 °C all the samples showed the lowest hot modulus of rupture, with similar values irrespective of FMC content and GSD.

The microcracking effect induced by direct-bonding reactions within the magnesia-chromite brick matrix is an approach to enhance thermal shock resistance, when appropriately balanced against corrosion by slag and molten metal. To examine the influence of FMC content and GSD on thermal shock behaviour, 3 samples of each magnesia chromite brick composition were heated to 725 °C, held for 30 minutes, and rapidly cooled to room temperature. After each thermal cycle, the elastic modulus was determined and after 5 cycles the residual elastic modulus was calculated. In addition, the cold modulus of rupture was measured at the end of the 5 cycles and the residual cold modulus of rupture determined. The residual elastic modulus results (Figure 6) indicate that increasing the FMC content was not detrimental to the thermal shock behaviour and that the new GSD positively impacted this property. However, the residual cold modulus of rupture showed a decreasing trend with increased FMC addition, which was more pronounced with the standard GSD.

Having determined that the thermal shock resistance was not detrimentally affected in materials featuring high FMC levels corresponding to direct-bonded compositions, it was essential to evaluate the impact of FMC addition on other pertinent properties, notably corrosion resistance. However, a critical aspect to enhancing this essential property entails understanding the material’s permeability and abrasion characteristics, which in conjunction with chemical corrosion are key factors contributing to the compromised performance of magnesia-chromite bricks. Figure 7 shows the permeability and abrasion values obtained for 3 samples of each magnesia-chromite brick composition, indicating the permeability slightly decreased with FMC addition, irrespective of the GSD, and that abrasion was notably lower for both GSDs with increasing FMC content.

The final analyses performed on all the different magnesia chromite brick compositions were corrosion tests conducted at 1700 °C for 5 hours with steel and fayalitic slag. It is important to mention that it was not possible to evaluate all 12 brick compositions simultaneously due to the experimental setup; however, multiple tests were conducted and wear in both the steel and slag zones was assessed. The results presented in Figure 8 show that in both the slag and steel zones, typically an increased FMC content positively influenced the corrosion resistance, with a significant benefit observed for the new GSD in the slag zone.

Following the culmination of research and development activities aimed at creating a new magnesia-chromite grade endowed with optimal properties and an engineered microstructure, composition D was selected for plant trials. One-tonne mixes of both the standard and new GSD were processed in an industrial mixer, pressed into RH degasser throat shapes, and fired according to standard plant procedures. Subsequently, the properties of five bricks for each GSD were determined and the results compared to those of the laboratory-produced bricks (Table II). In the majority of cases, properties of the plant trial bricks were either equivalent or better than the laboratory-produced bricks, with the only notable exceptions related to the hot modulus of rupture. Furthermore, properties of the plant trial bricks with the new GSD were typically better than the standard GSD. Microscopy was also performed and showed that bricks produced in the plant with the new GSD had a dense microstructure (Figure 9). Consequently, composition D featuring the new GSD (i.e., RADEX TB016) was selected for field trials in the throat area of RH degassers at six Brazilian customers.

Currently, only one of the six Brazilian customers has completed field trials with the developed material. In this case, the initial test was carried out in the outlet throat (down leg) of the RH degasser. Throughout this campaign, which ran for 78 heats and ended based on the customer’s program, the field trial was documented in detail, as shown in Figures 10–12. Given that the outlet throat typically experiences lower stresses than the inlet throat (up leg), conducting a field trial in the less demanding region was an appropriate first step to evaluate the new grade. During the trial, the RADEX TB016 exhibited minimal wear with residual thicknesses ranging from 225 mm (initial brick dimension) to 190 mm. Furthermore, the RADEX TB016 bricks showed significantly less wear compared to the standard magnesia chromite bricks from another supplier used to line this area. Figure 11 shows the visual appearance of the outlet throat lined with RADEX TB016 during the campaign. The calculated wear rate of RADEX TB016 was 0.385 mm/hour, indicating a potential throat lifetime of up to 415 heats with this new grade, while maintaining a minimum safety thickness of 60 mm. Following the initial field trial, RADEX TB016 was successfully implemented in the inlet throat for a campaign of 104 heats, which ended according to the customer-defined process.