Representative results obtained using the slug calorimeter are presented in Figure 3 which provides a plot of the calculated effective thermal conductivity as a function of the mean FRM specimen temperature for one of the fiber/portland cement-based FRMs. Results for three consecutive heating/cooling cycles are presented. For the first two heating curves, the furnace temperature (setpoint) was 538 ºC after 45 min, 704 ºC after 70 min, 843 ºC after 90 min, 927 ºC after 105 min, and 1010 ºC after 2 h. For the third heating curve, the furnace temperature was raised linearly and much more gradually from room temperature to 600 ºC in 4 h, and then held constant for a period of time at 600 ºC, while the slug temperature gradually approached this value as well. All cooling curves were generated by simply turning off the furnace and monitoring the temperatures during its cooling back to room temperature.
Figure 3 Effective thermal conductivities (solid lines) for FRM B in comparison to previously measured data obtained using either a hot wire technique (2) or a transient plane source (TPS) method (3).
Several points are worth noting from Figure 3. First, in comparing the first and second heating curve results, some indication of the influence of reactions on the effective thermal conductivity can be deduced. For this particular FRM, both endothermic and exothermic events are indicated. The endothermic events indicated when the heating curve 1 values fall below those obtained for heating curve 2 likely correspond to dehydration reactions and phase changes from liquid or chemically bound water to gas. As opposed to corresponding to exothermic reactions (which are indeed present for some FRMs), for this FRM, the exothermic events indicated when the heating curve 1 values exceed those of heating curve 2 likely correspond to the mass transfer of hot gases through the highly porous three-dimensional FRM microstructure (for example, see the large connected grey pore that spans the middle of the microstructure shown on the right side of Figure 4). This would include the mass transfer of steam created both from the vaporization of free and loosely bound (gel) water at temperatures below 200 ºC and from the dehydration/vaporization of chemically bound water at a mean specimen temperature of about 350 ºC. In the latter case, the exterior faces of the specimens exhibit temperatures on the order of 600 ºC, well above the dehydration temperature of calcium hydroxide in portland cement-based materials (16). The mass transfer of these reaction gases along with their potential subsequent condensation may somewhat cancel out the beneficial effects of the endothermic reactions themselves. For example, for heating curves 1 and 2, times of 104 min and 102 min were required for the protected steel slug to reach a temperature of 538 ºC, respectively. Thus, in this case, the performance of FRM B was nominally identical whether (dehydration) reactions were present or absent during the high temperature heating exposure.
In Figure 3, the three cooling curves overlap one another over a large temperature range, suggesting that all of the reactions and mass transfer of vapors were completed during the course of the first heating curve. The cooling curves also correspond closely to the heating curve 3, obtained with the slowest heating rate. For the more rapid heating rates applied for heating curves 1 and 2, transient effects are more significant and the relatively constant temperature gradient across the specimen that is assumed in the analysis presented in reference (10) is only achieved when the mean specimen temperature exceeds about 400 ºC. Before this time, the effective thermal conductivities fall below their "true" values, as "steady state" conditions have yet to be established through the thickness of the FRMs. It can further be observed that by appropriately combining together the heating curves (for temperatures above 450 ºC) with the cooling curves (for temperatures below 450 ºC) an effective thermal conductivity curve that spans a larger temperature range than any individual heating or cooling curve can be obtained, providing values over a temperature range from 30 ºC to about 700 ºC. Finally, the slug calorimeter-measured values are observed to be significantly higher than the previously measured values for temperatures above 400 ºC, possibly due to heat transfer via radiation (17) between the retaining plates and the central slug, through this highly porous fibrous FRM.