Pressure dependent effective thermal conductivity of pure and SiC-opacified expanded perlite between 293 K and 1073 K

The effective thermal conductivity of pure and opacified expanded perlite is investigated at gas pressures between 10-4 hPa and 103 hPa and temperatures between 293 K and 1073 K by guarded hot plate and transient hot wire measurements. The examined materials are a low-density (p = 56 kg m-3) and a high-density (p = 180 kg m-3) commercial expanded perlite powder, denoted as P1.5 and P0.14, respectively, as well as a mixture of P0.14 with 40 weight-% SiC as opacifier, denoted as P0.14_O (p = 263 kg m-3). To obtain the gas pressure dependent contribution of the measured total effective thermal conductivity, the thermal conductivity in the fully evacuated state is determined experimentally and subtracted. A model for the gas pressure dependent contribution is developed, which takes into account all temperature dependencies of the occurring gas-kinetic parameters and successfully reproduces all measurement data within the experimental uncertainty (ca. 5%). The model includes four structural parameters determined by fitting, which are the effective pore diameters both inside and between the grains as well as the respective porosities. The obtained values agree well with the results of structural analysis by SEM and optical microscopy. Furthermore, the model contains an improved description of the coupling effect between solid-body and gas heat conduction by means of a single novel coupling parameter, which expresses the exceedance of the gas pressure dependent contribution of the effective thermal conductivity relative to the thermal conductivity of the gas inside the pores. At ambient pressure and 673 K mean temperature, the effective thermal conductivity of the opacified powder mixture amounts to 0.087 W m(-1 )Ksks, thus being competitive with commercial high temperature insulation materials like calcium silicate, but at a lower price (ca. 2-3 euro kgsks). Experiments with the transient hot wire method at ambient pressure yield a temperature dependent overestimation of the effective thermal conductivity by 4-22% compared to the guarded hot plate method, which is traced back to the pressure dependent contribution and discussed in appropriate terms of the difference between the coupling parameters obtained from both methods.