Based on years of research on products before and after operation, as well as systematic monitoring of refractory materials under specific conditions, the magnesium spinel refractory materials produced by the Magnezit Group have been successfully applied in rotary kilns for cement clinker production. In recent years, there has been a significant increase in enterprises using more economical alternative fuels, which reduces the main costs of target products and the accumulation of regional waste. However, when alternative fuels are used, the load on the refractory lining increases. Alkali metal sulfates, carbonates, and chlorides act as aggressive components, whose physical effects include the release of salts from the gas phase. These salts then condense and penetrate into the structure of the refractory materials with volume expansion, generating internal stress. Under chemical action, refractory components react with alkali metal salts. The length of the rotary kiln overhaul period depends on the intensity of interaction between the sintering zone lining and the fired materials.
The processes in the sintering zone can be divided into three groups:
(1) interactions where the dissolution or corrosion of refractory materials generates new minerals or liquid phases;
(2) penetration of clinker liquid phases and interaction products into refractory materials through pores;
(3) mechanical wear of refractory materials by erosive gas flow or melt.
Experiment and Results
The characteristics of the RMAG series refractory materials with the highest parameters (Group T) developed at the Satka production site are listed in Table 1. This material is used under extremely harsh conditions, mainly in dry furnaces.
| Parameters | Grades | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| RMAG T1 | RMAG T11 | RMAG T1Z | RMAG T12 | RMAG T1C | RMAG T2 | RMAG T21 | RMAG T2Z | RMAG T3 | RMAG T31 | |
| Ultimate Compressive Strength/(N·mm⁻²) | 60 | 60 | 60 | 75 | 55 | 60 | 70 | 60 | 55 | 70 |
| Open Porosity/% | 15.0 | 15.0 | 15.0 | 14.4 | 14.6 | 15.0 | 15.0 | 15.0 | 15.0 | 15.9 |
| Apparent Density/(g·cm⁻³) | 3.00 | 3.00 | 3.05 | 3.03 | 3.04 | 3.00 | 3.00 | 3.00 | 2.99 | 2.96 |
| Refractoriness Under Load/℃ | ||||||||||
| GOST 4070-2014 | ＞1700 | ＞1700 | ＞1700 | 1700 | 1690 | 1680 | ＞1700 | 1700 | 1670 | 1680 |
| ISO 1893-2005 | ＞1700 | ＞1700 | ＞1700 | ＞1700 | ＞1700 | ＞1700 | ＞1700 | ＞1700 | ＞1700 | 1700 |
| Thermal Resistance, Thermal Cycle | ||||||||||
| 1300℃ Water | 25 | 15 | 25 | 10 | 27 | 25 | 15 | 25 | 25 | 15 |
| 950℃ Air | ＞100 | ＞80 | ＞100 | 80 | ＞100 | ＞100 | ＞80 | ＞100 | ＞100 | ＞80 |
| Thermal Conductivity at 1000℃/(W·m⁻¹·K⁻¹) | 3.1 | 3.0 | 3.0 | 3.7 | 3.5 | 3.1 | 3.0 | 3.0 | 3.1 | 3.1 |
| Relative Elongation at 20～1000℃/% | 1.3 | 1.3 | 1.6 | 1.3 | 1.3 | 1.3 | 1.3 | 1.6 | 1.3 | 1.3 |
| Mass Fraction/% | ||||||||||
| MgO | 86.0 | 88.0 | 85.5 | 94.5 | 88.0 | 87.5 | 90.0 | 85.5 | 87.1 | 89.0 |
| Al₂O₃ | 12.0 | 10.0 | 11.0 | 3.6 | 6.3 | 10.2 | 7.5 | 11.0 | 10.5 | 7.7 |
| SiO₂ | 0.3 | 0.3 | 0.3 | 0.3 | 0.3 | 0.32 | 0.5 | 0.4 | 0.6 | 0.5 |
| CaO | 0.8 | 0.8 | 0.8 | 0.8 | 0.8 | 0.73 | 1.0 | 0.7 | 1.0 | 1.0 |
| Fe₂O₃ | 0.6 | 0.6 | 0.6 | 0.8 | 2.2 | 0.8 | 0.8 | 0.8 | 0.8 | 0.8 |
| ZrO₂ | - | - | 1.8 | - | - | - | - | 1.0 | - | - |
| Cr₂O₃ | - | - | - | - | 2.1 | - | - | - | - | - |
Table 1 Characteristics of High - Purity Magnesium Spinel Refractory Materials of RMAG Series in Group T
To improve the resistance to corrosive clinker components, the structure of the refractory material should not be penetrated by alkalis, and there should be no significant changes in its mineral composition. As a possible protection method, the pores of the product can be additionally filled with a suspension or a chemical solution, with which the alkali metal salts should react. In this study, an aluminum - containing agent (A) or a carbon - containing agent (C) was used as the impregnation solution to fill the pores. Al₂(OH)ₙA → H₂O↑ + A↑ + Al₂O₃ (where A is acid slag), and it decomposes after heating and then reacts with the clinker components. According to the phase diagram of the binary system CaO - Al₂O₃ (Figure 1), at the temperature for clinker firing in the converter, high - calcium aluminates of the 3CaO·Al₂O₃ and 12CaO·7Al₂O₃ types are first formed (with decomposition temperatures of 1535 °C and 1415 °C respectively).
Figure 1 Phase diagram of the CaO - Al₂O₃ system
The pores of the refractory material are filled with calcium aluminate, which crystallizes in the direction of decreasing temperature gradient, preventing the further penetration of aggressive gases and clinker components into the structure. The reason for choosing the carbon - containing agent type is that, in addition to reducing the wettability of the furnace lining surface, after being oxidized in the furnace atmosphere, it generates coke residues, which partially fill the large pores or insulate the small pores.
Before impregnation, regardless of the reagent used, the product is heated to 50 - 60 °C. Impregnation with agent C is carried out in a special vacuum chamber: under vacuum (about 0.08 N/mm²), and the impregnated product is kept for 2 hours. Impregnation with agent A is to soak the product in a container with a solution of a certain density and keep it until the release of bubbles stops. All impregnated products are heat - treated at 220 °C, and the gas permeability coefficient of the refractory material after impregnation is significantly reduced. The performance indicators of the samples before and after impregnation are listed in Table 2.
| Sample | Ultimate Compressive Strength/(N·mm⁻²) | Open Porosity/% | Apparent Density/(g·cm⁻³) | Gas Permeability Coefficient/μm² |
|---|---|---|---|---|
| Non - impregnated | 79.8 65.6 Average: 72.7 |
15.2 15.8 Average: 15.5 |
3.00 2.97 Average: 2.99 |
0.294 0.293 Average: 0.294 |
| Impregnated with A | 73.2 69.3 Average: 71.3 |
10.7 10.5 Average: 10.6 |
3.05 3.06 Average: 3.06 |
0.035 0.035 Average: 0.035 |
| Impregnated with C | 173.6 195.8 Average: 184.7 |
8.1 7.2 Average: 7.7 |
3.07 3.09 Average: 3.08 |
0.013 0.011 Average: 0.012 |
Table 2 Performance Indicators of Samples Before and After Impregnation
The crucible (static) method is one of the operating methods for determining the clinker resistance of refractory materials and is used to evaluate the degree of integrity of their chemical interaction with reagents. This method was used to study the corrosion resistance of the structure of high-purity magnesium spinel refractory materials impregnated with different types of impregnating agents. For the test, the product is made into a rectangular crucible with a concave hole drilled in the center, as shown in Figure 2. The sample crucibles are filled with corrosive substances, and the impregnated samples are compared with the non-impregnated samples.
Figure 2 Schematic diagrams of the crucible of the test sample before and after sectioning
A mixture of raw meal from a cement plant with K₂SO₄ and NaCl salts (mass ratio 80:10:10) was used as the corrosive agent for the test. The chemical composition of the raw meal: MgO 2.30%, Al₂O₃ 7.16%, SiO₂ 18.60%, CaO 66.9%, Fe₂O₃ 3.90%, Na₂O 0.43%, K₂O 0.99%; △ma is 35.1%.
Under the same conditions, crucibles with corrosive agents were fired simultaneously at a maximum temperature of 1,680 °C. The furnace was covered with a protective cover made of crucible material (10 mm thick). All crucibles after firing showed intact appearances (see Figure 3). In the working holes of the samples, the reagents were preserved in the form of cylindrical sintered bodies. The maximum amount of residual reagents was observed in the crucibles impregnated with carbon-containing impregnants. In both cases, the reagents only contacted the refractories at the lower part of the working holes (see Figure 4), where the color of the samples turned dark brown. The reagents fell off during the sawing process (see Figure 5).
(a) Crucible without impregnation; (b) Crucible impregnated with Agent A; (c) Crucible impregnated with Agent C
Figure 3 Photo of a fired crucible sample
(a, b) Without impregnation; (c, d) Impregnated with Agent C; (e, f) Impregnated with Agent A
Figure 4 Pictures of samples after the test
(1, 2) From crucibles impregnated with Agent A; (3, 4) From crucibles impregnated with Agent C; (5, 6) From non - impregnated crucibles
Figure 5 Reactants (after cutting) of the crucible composed of magnesium spinel after the test
For petrographic studies, the fired crucible was sawed vertically from the center into two equal parts along the vertical section, and polished sections were prepared. An optical microscope was used to observe the working areas (corrosion and impregnation areas) generated in the samples after the clinker resistance performance test, and manually transferred to the images of the sawn surfaces. At the bottom and on the sides of the inner hole, the depth of the identified areas was measured every 3 mm using Image - Pro software. The evaluation of the impregnated area and the study of the microstructure of the changed areas were carried out using a Tescan electron microscope (see Figure 6).
(a, b) Non - impregnated crucibles; (c, d) Impregnated with Agent C; (e, f) Impregnated with Agent A
Figure 6 Cross - section of the crucible after the test: (■) Corrosion area and (■) Impregnation area
When conducting microscopic examination of the crucibles after the test, the following regions were identified: (1) Working region. Divided by mineralogical composition and porosity: the corrosion region, which is the area where the refractory material is in direct contact with the slag; the impregnation area formed by the secondary minerals impregnating the refractory material. (2) The region with the least change, that is, the area farthest from the working contact surface. The depth of the working region of the pre - impregnated samples is 10 - 11 mm, which is less than the 15 mm of the non - impregnated samples (see Table 3).
| Index | Samples | ||
|---|---|---|---|
| Non - impregnated | Impregnated with Agent A | Impregnated with Agent C | |
| Depth of working area /mm | 14.6 | 10.8 | 10.6 |
| Corrosion area | 3.9 | 3.1 | 2.8 |
| Impregnation area | 10.7 | 7.7 | 7.8 |
| Area of working area /% | 34.2 | 27.4 | 26.5 |
| Corrosion area | 4.9 | 3.8 | 3.5 |
| Impregnation area | 29.3 | 23.6 | 23.0 |
Table 3 Area and Depth of the Working Zone of Samples After the Test
The interaction mechanism between the corrosiveness and the structural components of the sample crucible is the same, including the penetration of high - calcium silicate through the open pore space. This leads to the local corrosion of aluminum - magnesium spinel (AMS) grains and the subsequent formation of new impregnated phases (calcium aluminate, limonite, sodium - potassium aluminosilicate).
The regions formed in the samples after the clinker penetration test have characteristics of the mineral phase composition (see Table 4 and Figure 7) and the pore structure (see Table 5 and Figure 8).
| Index | Samples | ||
|---|---|---|---|
| Non - impregnated | Impregnated with Agent A | Impregnated with Agent C | |
| AMS | |||
| Region with the least change | 15 - 16 | 15 - 16 | 15 - 16 |
| Working region | ~ 12 | ~ 14 | ~ 14 |
| Calcium aluminate | |||
| Region with the least change | ~ 3 | ~ 3 | ~ 3 |
| Impregnation region | ~ 7 | ~ 5.5 | ~ 4.5 |
| Corrosion region | ~ 6 | ~ 4 | ~ 3 |
| High - calcium silicate | |||
| Region with the least change | 2 - 3 | 1 - 2 | 1 - 2 |
| Impregnation region | 2 - 3 | 1 - 2 | 1 - 2 |
| Corrosion region | ~ 1 | 2 - 4 | 2 - 3 |
| Limonite | |||
| Region with the least change | - | - | - |
| Impregnation region | < 2 | < 1 | < 1 |
| Corrosion region | < 10 | 7 - 8 | 6 - 8 |
| Na - K aluminosilicate | |||
| Region with the least change | < 1 | < 1 | < 1 |
| Impregnation region | < 3 | < 2 | < 2 |
| Corrosion region | < 2 | < 1 | < 1 |
| Sum of silicates and aluminates | |||
| Region with the least change | ~ 7 | ~ 5.5 | ~ 5 |
| Impregnation region | ~ 13 | ~ 9.5 | ~ 9 |
| Corrosion region | ~ 17 | ~ 15 | ~ 13.5 |
Table 4 Mineral Phase Composition of Samples After the Test (in %)
| Pore Type | Pores in Samples | |||||
|---|---|---|---|---|---|---|
| Non - impregnated | Impregnated with Agent A | Impregnated with Agent C | ||||
| μm | % | μm | % | μm | % | |
| Impregnation Region | ||||||
| Closed, Isolated Pores | 4 - 5 | < 100 | 2 - 3 | < 70 | 2 - 3 | < 40 |
| Open Pores | ||||||
| Interconnected, Single - sided Open | 5.5 | < 70 | 6 | < 70 | 6 | < 50 |
| Annular | 3.3 | < 100 | 3.5 | < 40 | 3 | < 50 |
| Large - sized | 1 - 2 | < 250 | 1 - 2 | < 250 | 1 - 2 | < 200 |
| Total Open Porosity /% | 15 | 13.5 | 13 | |||
| Corrosion Region | ||||||
| Closed Pores | ||||||
| Irregular Shape | 6 - 8 | < 800 | 4 - 5 | < 1,000 | 6 - 9 | < 800 |
| Elongated and Tortuous | 4 - 5 | < 100 | 4 - 6 | < 100 | 2 - 4 | < 100 |
| Total Open Porosity /% | 12 | 10 | 10 |
Table 5 Pore Structure of the Sample Matrix After the Test in the Working Zone
All refractory materials have the same initial AMS content. In the working area of the impregnated samples, its content is slightly higher than that of the non - impregnated samples, that is, impregnation slows down the corrosion of spinel particles. In the impregnated refractory materials, the high - calcium silicate phase is more concentrated in the corrosion area; in the non - impregnated samples, it migrates deeper into the impregnation area. According to the test results, the refractory materials impregnated with the impregnating agent have the best corrosion resistance, which is characterized by a low content of newly formed compounds (calcium silicate and aluminate, limonite, sodium - potassium aluminosilicate) in all areas (corrosion). In the non - impregnated samples, the content of silicates and aluminates in these areas is the highest.
(a) Impregnation zone
(b) Corrosion zone
1-Periclase; 2-Calcium aluminate; 3-Limonite; 4-Na-K aluminosilicate; 5-AMS; 6-High - calcium silicate
Figure 7 Typical Structures of the Impregnation Zone and Corrosion Zone in the Working Area of the Specimen After the Clinker Penetration Test (×1000)
(a) Non - impregnated
(b) Impregnated with Agent A
(c) Impregnated with Agent C
1-Periclase; 2-AMS; 3-Limonite + Secondary Impregnation Component; 4-Pores Corroded by AMS; 5-Open Connected Pores
Figure 8 General Microstructures of Crucibles Without Impregnation, Impregnated with C, and Impregnated with A (×20)
As shown in Figure 7, the working area is divided into the following two sub - regions:
(1) The corrosion region (dark brown) is characterized by the complete dissolution of AMS films and particles in the impregnated high - calcium silicate material, resulting in large pores of 400 - 800 μm. The secondary products of AMS dissolution are brownmillerite Ca₂(Al, Fe³+)₂O₅ and calcium aluminate mCaO·nAl₂O₃. Generally, the specified region is dense, and it is difficult to distinguish the boundaries between the charge components. In the spinel periclase particles and the matrix, impregnated minerals in the form of films with a width of 20 μm can be observed everywhere.
(2) The impregnation region is formed by newly generated minerals (calcium aluminate, and less commonly Na - K aluminosilicate) filling the open pore space. It is characterized by the retention of AMS grains and part of the AMS films. Its structural features include the generation of terminations in the open interconnected pores. The region with the least change is characterized by the partial replacement of silicate components with impregnation components. In terms of microstructure, it is practically equivalent to a typical refractory material.
The reagent is mainly represented by high - calcium silicate crystals of 20 - 50 μm, and there are few deformed crystals of calcium aluminate, brown silicate, and Na - K aluminosilicate films of 20 - 70 μm.
The results of petrographic studies show that: the proportions of different types of open pores in all crucibles are quite similar, the sizes of the impregnated samples are smaller, and the total open porosity of the impregnated samples is slightly lower than that of the non - impregnated samples.
Conclusion
Crucibles made of high-purity magnesium spinel refractory materials were pre-impregnated with different types of impregnating agents. The samples after the corrosion test were compared with the non-impregnated samples. The results show that pre-impregnation with an aluminum-containing crucible (A) or a carbon-containing (C) agent prevented and slowed down the penetration of aggressive components into the product structure, reducing the depth and area of impregnation.
Although the structural protection mechanisms of using impregnating agents A and C are different (interaction of active Al₂O₃ to form high-calcium aluminate at one time or filling pores with coke residue), the corrosion depths of the working areas of the magnesium spinel sample crucibles after impregnation are comparable, both being approximately 11 mm compared to the non-impregnated samples (~15 mm). This indicates the effectiveness of the pre-impregnation method, which is a promising approach to reduce the erosion of the structural strength of refractory materials by aggressive components.