Selection of the proper burner tile design and material requires knowing the operating temperature profile and thermal boundary conditions to ensure optimum performance.

Fig. 1. Flat-flame burner. Courtesy of John Zink Co.

A refractory burner tile is an essential component of a burner in a thermal processing system. Equipment downtime due to faulty operation of the burner tile can result in lost production with associated costs, which can far exceed the price of the burner equipment. Burner design has changed over the past 20 years due to increasingly stringent environmental regulations, leading to a compromise between efficient mixing of fuel and oxidizer and the need to reduce total fuel consumption and pollutant emissions, which requires moderately low flame temperatures to reduce NOx. Design strategies have led to the development of custom solutions for specific applications, including sophisticated burner designs using complex shaped tiles. For example, Fig. 1 shows a low-emission flat-flame burner consisting of a complex tile shape with sharp edges, nonuniform wall thicknesses and additional fuel jets to reduce NOx. The design is used to control a precise fuel oxidation process [1]. Burner-tile refractory material requirements and evaluation methods, as well as effects of different refractory binder systems are discussed in this article.

Burner tile design

Burner tile design affects burner performance by forming the desired flame pattern, preventing flame radiation toward the burner casing, forming the orifice that controls the flow of air for the combustion reaction and stabilizing the flame by means of thermal insulating properties of the refractory material. Factors that affect tile design include:

  • Thermal stresses in the tile due to thermal expansion restraint from the structural casing or anchoring
  • Stress risers at sharp inside edges
  • Thermal stresses in thinner tile wall sections
  • Refractory thermoplastic/elastic properties in the operating temperature range
  • Large thermal gradients (to 1000°C, or 1830°F) in the material
  • Thermal shock from burner cycling

Fig. 2. Axisymmetric burner tile used to analyze thermomechnical behavior during heat up; Fig. 3. Temperature contours (K) for cement-bonded castable HS burner tile after one hour heat up. Displacement scaling factor = 50.; Fig. 4. Circumferential stresses (Pa) for cement-bonded castable HS burner tile

Tile design considerations

Selection of the proper burner tile design and material requires knowing the operating temperature profile and thermal boundary conditions, which is influenced by burner type, fuel and air inlet composition and possible flue gas recirculation. Obtaining reliable data experimentally is time consuming and requires the availability of appropriate equipment. By comparison, burner modeling using computational fluid dynamics (CFD) can provide a good conception of the flame temperature and pattern.

Material testing must take into consideration the different dynamic temperature stages to which the tile is exposed including heat up, steady heat flow at operating temperature, thermal cycling and cool down. Also, the elevated temperature tensile properties of the refractory materials must be known.

The chemical composition of tile material for use in certain applications is specified in ISO Standard 13705. Normal burner tile must contain more than 40% alumina, and high intensity combustors, more than 85% alumina. Mineral composition also is important in thermochemical and mineralogical reactions. High mullite content imparts good thermal shock properties. Alkali and alkali earth metal oxide content influences modulus of elasticity, thermal shock properties and phase transitions that lead to stresses due to expansion or shrinkage.

Burner tile material must have good resistance to elevated temperature tensile and compressive strains resulting from severe thermal gradients during service. Cracking from these stresses can change the flame pattern or can result in the flame to be in contact with the tile causing problems like hot spots and total destruction of the tile and burner casing [2]. An ideal thermal shock resistant material should have low thermal expansion, high thermal conductivity and high strength under consideration of the stress/strain ratio. Design factors that can reduce susceptibility to thermal shock are uniform, thin wall, small overall dimensions of tile and tile components, and avoidance of sharp angular shapes.

Fig. 5. Calculated plastic strain in circumferential direction; Fig. 6. Temperature contours (K) for phosphate-bonded castable SB burner tile after one hour of heat up. Displacement scaling factor = 50.; Fig. 7. Circumferential stresses (Pa) for phosphate-bonded castable SB burner tile

Types of tile materials

Castable materials traditionally have been used for burner tiles, which minimizes the cost of manufacturing complex shapes in limited quantities. Castables also have better resistance to thermal shock than pressed and rammed products and are easier to install. However, castables have higher shrinkage than fired products, which affects dimensions at service temperatures.

Cement-bonded materials have evolved from high cement conventional formulations to low and ultralow cement materials over the past 30 years with an improvement in installation characteristics and high temperature stability. New additives and fillers have helped achieve water reduction to decrease porosity and increase strength [3].

High water content and low permeability of cement-bonded materials can result in explosive spalling or a steam explosion during a rapid increase in temperature. Curing conditions, heat up and dry-out methods and material properties can contribute to explosive spalling [4]. Therefore, precast burner tiles require accurate dry out or pre-firing before installation. Stiffness of castables increases with lower cement content and higher alumina content [5].

Phosphate-bonded burner tiles having complicated shapes can hold up better than conventional cement-bonded burner tiles in some critical process conditions. Phosphate-bonded castable materials have a range of properties and distinguishing characteristics. Binder systems for phosphate-bonded refractory castable products include heat setting, cold setting plus setting additive, one-component dry material with water addition and two-component mix with phosphoric acid solution or monoaluminum phosphate solution plus dry component.

Heat-setting phosphate-bonded castables are based on the Al2O3-SiO2-P2O5 system. Cold-setting phosphate-bonded castables are based on the Al2O3-SiO2-P2O5-MxO system (MxO = alkali or alkali earth oxide). The most used phosphate binder in the refractory industry is monoaluminumphosphate (MAP) in liquid or spray-dried form. An additional setting agent (MxO like sodium, magnesia, calcium or compounds thereof) is required to obtain setting properties at ambient temperature. A disadvantage of a spray-dried MAP additive in a single-component material is poor shelf life due to its strong hygroscopic behavior.

Alternatively, phosphoric acid can be used as a liquid in a two-component product. Depending on the reactivity, both MAP and phosphoric acid need to be diluted to slow down the exothermic reactions. Phosphoric acid is a tri-basic acid and reacts with different metal oxides to form salts, some of which function as a refractory binder.

Phosphate-bonded products consisting of a two-component mono- and single-component diphosphate binder system are commercially available. The advantage of a two-component monophosphate bond is the faster setting and faster heat up time. Monophosphates are less susceptible to explosive spalling than diphosphate-bonded materials.

Thermomechanical properties of phosphate-bonded materials are different than those of a cement bond. Additional curing time is not needed after setting. Poly and metaphosphate compounds formed during condensation have a ring and chain structure, which is stable to a temperature of 800°C (1470°F) [6]. The flexible structure has a low modulus of elasticity, which contributes to thermal-shock and abrasion resistance.

MxO-phosphate compounds in the binder system start to melt at a temperature greater than 900°C (1650°F). Heat-setting phosphate-bonded materials behave similarly due to a reaction of phosphates with MxO impurities. Thus, phosphate-bonded castables with alkali additions can have lower hot modulus of rupture and refractoriness-under-load values compared with cement formulations. A certain amount of melt phase at service temperature reduces material stiffness and improves thermal shock properties, which is desired in many burner tile applications. At temperatures between 900 and 1600°C (1650 and 2910°F), amorphous phosphates and phosphate melts form, which can attack raw-material aggregates at process temperatures above 1350°C (2460°F). Resulting volume changes are due to the formation of additional melt (shrinkage) or formation of new compounds such as Mg-spinel and mullite (expansion). Aluminum phosphate starts to decompose to aluminum oxide and P2O5 above 1600°C (2910°F), which is volatile at this stage.

Testing of phosphate-bonded materials

Phosphate melts are temporary in nature and change to different compounds at higher temperatures. Critical tests to indicate this material characteristic are:

  • Thermal expansion under load at temperatures higher than 1100°C (2010°F)
  • Hot modulus of rupture at temperatures higher than 1100°C (2010°F)
  • Refractoriness under load

Test results can be significantly influenced by the occurring melt phase at test temperature. For example, samples prefired at 1100°C with a holding time of 24 hours show a 10% increase of HotMOR, and a 65% increase with a holding time of 168 hours. Published data typically is for only a five hour holding time. In the thermal expansion under load test, plastic flow and coefficient of thermal expansion are mixed due to the load, and the resulting high permanent linear change is overstated.

Calcium phosphate also can covert to Anorthite. Calcium-monophosphate (part of the binder component) is formed at ambient temperatures by the reaction of phosphoric acid and calcium oxide. During heat up, the monophosphate converts into calcium-metaphosphate, which has a melting point of 975°C (1790°F). After firing at 1600°C with 10 hours holding time, all phosphate is converted to AlPO4. On the other hand, calcium oxide reacts with mullite and converts into Anorthite, which has a melting point of 1550°C (2820°F).

Test procedures of phosphate bonded materials should report this characteristic, because burner tiles are exposed to elevated temperatures with holding times up to many years. A material assessment based on standard test procedures would lead to the wrong conclusion.

Burner tile numerical analysis

FEM analyses of axisymmetric burner tiles used in an annealing furnace (Fig. 2) were conducted to determine thermomechanical behavior during heat up. The modeling is based on tiles made of commercially available cement-bonded and phosphate-bonded castables. A heat up time of 48 hours was considered. The cement-bonded material is an andalusite-based LCC in a dried, ready-to-install condition. The phosphate bonded material is a mullite-based two-component monophosphate castable containing 61% alumina and 4.5% P2O5. Physical properties are listed in Table 1, while mechanical properties under tensile loading conditions are listed in Table 2. The yield stress under compressive loading is assumed to be a factor of ten higher in this simulation.

Most of the damage in the LCC Andalusite HS occurs within the first hour of heat up, as shown by the calculated temperature contours (Fig. 3) and calculated stresses (Fig. 4). The highest stresses occur in circumferential direction. Calculated stresses are close to the yield stress limit for much of the outer burner block region. Calculated plastic deformation (Fig. 5) indicates where cracks are most likely to initiate. Crack deformation in the actual tile could be more severe due to the brittle nature of the ceramic materials.

No damage occurs to the phosphate-bonded SB in the first hour of heat up (Fig. 6). Calculated stresses in the circumferential direction (Fig. 7) are lower than for the LCC Andalusite HS tile.

FEM analysis shows that the phosphate-bonded material allows a higher temperature increase without exceeding the failure stress compared with cement-bonded material. Although this FEM analysis focuses only on the heat up process of the tile, it clearly shows material behavior at a very critical state of the temperature cycle. Thus, FEM analysis is a practicable tool for basic materials ranking. IH