Effect of the pore volume fraction on the thermal conductivity and mechanical properties of kaolin-based foams

Abstract

Candidate materials for thermal insulation combine a solid phase of low thermal conductivity with porosity. Very porous ceramics based on a commercial clay containing 79% kaolinite were made by a direct foaming method using methylcellulose, drying at 70 °C before a consolidation at 1100 °C. The pore volume fraction ($v_p$) was varied in the kaolin-based foam by modifying the amount of clay incorporated in the starting mixture. Measurements of the thermal conductivity with the hot disk method revealed a decrease from 0.23 W m⁻¹ K⁻¹ at $v_p$ = 0.57 to 0.054 W m⁻¹ K⁻¹ at $v_p=0.95$ in close agreement to predictions by the Hashin–Shtrikman upper bound, a cubic pore model and numerical simulation of an artificial microstructure representing the polyhedral pore shape with Voronoï mosaics. The effective Young's modulus, obtained from mechanical compression tests, also decreases with pore volume fraction as described by Ashby's relation. However the mechanical strength was sufficient for handling even the more porous kaolin-based foams.

Introduction

Thermal properties of ceramic materials are important for applications involving heat transfer such as housing, substrates for electronics and high temperature processing. One of the key parameters which determines performance is thermal conductivity. In the case of polycrystalline oxides, room temperature values typically vary from 35 to 40 W m⁻¹ K⁻¹ for large grain ceramics of alumina or magnesia down to 1 W m⁻¹ K⁻¹ for silica glasses. Clay based materials also exhibit very low values of thermal conductivity <1 W m⁻¹ K⁻¹ suitable for thermal insulation. Further decrease in conductivity can be achieved by the incorporation of a high volume of porosity into the material. This paper examines the influence of pore volume fraction on thermal conductivity and mechanical properties of very porous kaolin-based ceramics.

In this context, the effective thermal conductivity of a polycrystalline ceramic material depends notably on three main factors: (i) the intrinsic thermal conductivity, (ii) the thermal resistance due to interfaces called grain boundaries and (iii) the presence of pores. In the temperature range of interest to the present work, heat is carried within a grain (or crystallite) by lattice vibrations and limited by mutual interference in the form of phonon–phonon interactions. Related to the symmetry of the crystalline phase, the grain exhibits isotropic or anisotropic thermal conductivity. In the latter case, texturing can result in significant effects at the macroscopic scale. Additional scattering sites in the propagation of lattice vibrations are found at grain boundaries where crystallites of different orientation meet. Work on alumina, tin oxide and zirconia has yielded values for the grain boundary thermal resistance in the range 0.5–1.2 × 10⁻⁸ m² K W⁻¹.1, 2 Smaller grain size, which increases the number of grain boundaries per unit length of heat path, decreases the thermal conductivity of the solid phase. But this effect is less significant for very insulating oxides like zirconia³ or clay.⁴ Lastly the presence of pores in the microstructure leads to a strong drop in effective thermal conductivity due to the low thermal conductivity of the gas confined in pores compared to the solid phases.

To fabricate porous materials starting from kaolin clay, different processing routes can be investigated such as the replica technique, use of pore formers (templating) or direct foaming methods. The present work is devoted to the preparation and characterization of foams due to the low cost and short processing times. Furthermore, this choice promotes mechanical strength of the porous structure due to the continuity of the solid phase. Foams based on a commercial clay (BIO Kaolin) containing 79% of kaolinite have thus been prepared using a surfactant to yield pore volume fractions from 57% to 95%. Following heat treatment at 1100 °C, the final thermal and mechanical properties of these samples are examined. The experimental values for thermal conductivity are compared with predictions made with a simple parallel/series resistor approach, the Hashin–Shtrikman upper bound and also numerical simulations taking into account the effect of pore volume fraction. A similar approach, based on two phase analytical expressions for the mechanical properties, is adopted in order to emphasize the compromise between low thermal conductivity and sufficient mechanical strength.