Thermophysical properties such as the thermal diffusivity, specific heat and thermal conductivity of coarse ceramics are crucial parameters for the optimization of production processes. For decades, such properties were determined by employing stationary methods (e.g., guarded hot-plate technique) or standardized transient techniques (e.g., the hot-wire method).

However, those methods are limited to large sample sizes and low thermal conductivities. Materials with thermal conductivity values above 25 W•m^-1•K^-1 can generally not be analyzed using these traditional methods. Furthermore, those methods are typically very time consuming, with tests between room temperature and 1000°C generally requiring measurement times of several days.

Flash methods, on the other hand, are non-contact measurement techniques and can handle high thermal conductivity materials without any difficulties.¹ Flash methods are absolute methods for the determination of thermal diffusivity. Modern instruments often allow additional measurement of the specific heat—and therefore the determination of the thermal conductivity—without further tests. However, the flash method has traditionally been somewhat limited to small sample sizes and homogeneous materials, so it has generally not been employed for inhomogeneous coarse ceramics.



Various improvements and modifications have been made to the flash method since its introduction by Parker et al.¹ New evaluation models were developed that account for heat loss² and finite pulse effects.³ Some analysis methods are already available that simultaneously take both of these into account in an optimum way.^4,5 New developments in the field of detectors and electronics carried out in recent years allow for using this technique for larger samples as well. Therefore, the flash technique can now be used for inhomogeneous materials such as coarse ceramics or refractory materials.

Concept

Enlarge this picture Figure 1. Measurement part of a laser flash system.

In a flash measurement, the front side of a small, usually disk-shaped plane-parallel sample is heated by a short light (laser) pulse. The temperature increases induced in a small surface layer of the sample by the laser diffuse through the sample and lead to a small temperature increase of the entire sample, including the back surface. This temperature rise on the rear surface is measured vs. time using an infrared detector.

By placing the sample into a tube furnace, temperature-dependent measurements can also easily be realized. Modern flash systems⁶ are fully automated top-loading devices equipped with a flash source (e.g., Nd:YAG lasers) on the bottom and an infrared detector on top (see Figure 1). However, other flash systems are equipped with an external laser connected to the measurement part by glass fiber optics.*

*Netzsch LFA 427

Testing and Results

Figure 2. Silica-magnesia-alumina sample block.

The measurements on the oxide ceramics and the magnesia-carbon refractory materials presented in this work were carried out using two laser flash systems.^† Several small samples of the same material were measured to check the homogeneity of the material and the reproducibility of the method.

The analyzed materials came from different sources. A sillimanite ceramic material containing approximately 75% alumina and 25% silica was measured. From a larger block, three samples were prepared and measured with a laser flash device up to 1200°C. A low-conductivity refractory brick (major contents were silica, magnesia and alumina) was also analyzed. Again, three samples were prepared from a larger brick and measured up to 800°C. The sample structure can be seen in Figure 2. Positions 1, 2 and 3 indicate the positions where the samples were prepared.

† NETZSCH models LFA 427 and LFA 457 MicroFlash

Enlarge this picture Table 1. Carbon content and density of different magnesia-carbon refractories.

Finally, carbon-containing magnesia refractories with different carbon content were tested. The carbon refractories had densities of approximately 2.9 g/cm³ and carbon contents between 8 and 20%. The samples contained 1 to 2% of organic binder that decomposed between 300°C and 500°C. The maximum grain size of the samples was in the range of 3 mm. The magnesia-carbon samples measured here were approximately 10 mm thick and had a diameter of approximately 24 mm. The carbon content and room temperature bulk densities of the different materials analyzed are summarized in Table 1.

All tests were conducted in a dynamic argon atmosphere. The tests on the sillimanite samples were carried out between room temperature and 1200°C, in steps of approximately 300 K. Three samples of 12.7 mm in diameter and approximately 3 mm high were prepared from the original brick and measured. The measurements on the low-conductivity silica-magnesia-alumina materials were done at room temperature, 300°C, 600°C and 800°C. Tests were carried out on three samples of 12.7 mm in diameter and approximately 3 mm high.

Enlarge this picture

For samples 1, 3 and 5 of the magnesia-carbon refractories, tests were conducted between room temperature in steps of approximately 100 K. The other two carbon refractories (samples 2 and 4) were tested at room temperature, 300°C and 1000°C. For the specific heat determination, the laser flash systems were calibrated using different alumina standards prior to the measurements on the ceramic/refractory materials. From the room temperature bulk density, the measured thermal diffusivity and the specific heat, the thermal conductivity was determined using the equation seen at the right.

The temperature dependence of the density was not considered because this influence is generally small (within a few percent) for such materials. Depicted in Figure 3 are the measured thermal diffusivity and specific heat of the first sillimanite sample. Also shown is the thermal conductivity calculated from the measured results and a room temperature bulk density of 2.542 g•cm^-3.

Enlarge this picture Figure 3. Thermophysical properties of the sillimanite material (first sample).

For ceramic materials, thermal diffusivity typically decreases with temperature up to 800°C.^7,8 At temperatures above 800°C, the values increase. This increase is most likely due to the contribution of radiation heat transfer. The absolute values for the thermal diffusivity are between 1.4 and 0.6 mm²•s^-1. The specific heat increases vs. temperature over the entire temperature range, as can be expected from the Debye theory.⁹

Thermal conductivity decreases vs. temperature up to 600°C, which is what can be expected for a pure phonon-conducting material.⁸ At higher temperatures, the values increase again. This is mainly due to the influence of the radiation heat transfer. Compared to other ceramic materials,⁸ the results show only slight temperature dependence. All thermal conductivity results are between 2 and 3 W•m^-1•K^-1. The weak temperature dependence, as well as the low thermal conductivity results, are typical for such alumino-silicate ceramics.

The low thermal conductivity, as well as the temperature dependence, measured on the first sample was found for all three samples. All results were within 5%, which is within the general uncertainty range of the instrument for direct thermal conductivity determinations.

Enlarge this picture Figure 4. Thermal conductivity results of the silica-magnesia-alumina ceramic (first sample).

Figure 4 shows the thermal diffusivity, specific heat and thermal conductivity of the first sample of the silica-magnesia- alumina refractory. Thermal conductivity was determined using a room temperature bulk density of 0.727 g•cm^-3. The low density of the material is caused by a significant porosity of more than 50%. It should be pointed out that the main pore size was less than 0.1 mm. Therefore, the porosity had no negative impact on the uncertainty of the flash tests.

The measured specific heat increased vs. temperature. As can be expected from the composition, the results for the specific heat were close to the results measured for the sillimanite material. The thermal diffusivity decreased slightly with temperature up to 600°C. Above this temperature, the results increased again. The thermal diffusivity results were low, which is attributable to the composition of the material and the high porosity. The results at room temperature were around 0.3 mm²•s^-1. The calculated thermal conductivity increased vs. temperature over the entire temperature range. The values were between 0.15 and 0.25 W•m^-1•K^-1. These values are typical for porous silica ceramics.⁸

The thermal conductivity values of the three different samples of the porous silica-magnesia-alumina refractory ceramic showed a nearly linear increase vs. temperature between 300 and 800°C, while the room temperature results were slightly below the linear trend. For the three individual samples, slight deviations were detected (within ± 4%), which can be explained by the high degree of inhomogeneity of the material.

Enlarge this picture Figure 5. Thermophysical properties of refractory material 1 (20 wt% carbon content).

Presented in Figure 5 are the measurement results for sample 1 of the magnesia carbon refractory material (20% carbon content) between room temperature and 1000°C. The specific heat increased vs. temperature, as expected from the Debye theory.⁹ The thermal diffusivity decreased vs. temperature over the entire temperature range. Between 300°C and 500°C, a step was detected in the thermal diffusivity that can be explained by the decomposition of the organic binder.

The decomposition of the binder yielded a slightly higher porosity and level of carbon black between the grains of the refractory. This resulted in an increased thermal resistance between the grains, causing a decrease in the effective thermal diffusivity of the entire material. Above 500°C, the thermal diffusivity was nearly constant.

Enlarge this picture Figure 6. Thermal conductivity of all five analyzed magnesia-carbon refractories vs. temperature.

The thermal conductivity showed a slight increase up to 100°C, which can be explained by the strong increase in specific heat in this temperature range. This increase compensates for the decrease in thermal diffusivity. At higher temperatures, the changes in thermal conductivity were mainly caused by changes in thermal diffusivity. The slight drop in the thermal conductivity above 800°C was probably due to contact/surface reactions between the carbon and the oxide material inside the refractory. The values changed from ~25 W•m^-1•K^-1 at room temperature to ~16 W•m^-1•K^-1 at 1000°C. The high thermal conductivities were typical for magnesia refractories with high carbon content.

A comparison of the thermal conductivity of all five of the measured magnesia-carbon refractories clearly showed that thermal conductivity decreased with decreasing carbon content (see Figure 6). Materials 2 and 4 were only measured at room temperature, 300°C and 1000°C. However, the results for these samples fit quite well with the results for the other materials tested.

Enlarge this picture Figure 7. Thermal conductivity vs. carbon content of the different magnesia-carbon refractories.

Fast and Reliable Measurements

Laser flash systems are fast and reliable instruments for the determination of thermophysical properties such as specific heat, thermal diffusivity and thermal conductivity of homogeneous ceramic materials. By using larger sample sizes or analyzing different samples of the same brick, the method can easily be extended to inhomogeneous materials such as coarse ceramics or refractories. The results achieved and discussed here for a sillimanite material, a silica-magnesia-alumina brick and magnesia-carbon refractories have proved that the non-contact flash method can be used for such materials.

For additional information regarding the laser flash technique, contact NETZSCH-Gerätebau GmbH, Wittelsbacherstrasse 42, 95100 Selb, Germany; e-mail andre.lindemann@netzsch.com; or visit www.netzsch.com.