Direct observation of the moisture distribution in calcium aluminate cement and hydratable alumina‐bonded castables during first‐drying: An NMR study

Abstract

The drying behavior for various calcium aluminate cement and hydratable alumina‐bonded refractory castables was investigated in the first‐drying temperature range (100°C‐300°C). Using a specialized high‐temperature Nuclear Magnetic Resonance setup, we were able to directly and nondestructively measure the spatially and temporally resolved moisture distribution, while simultaneously measuring the temperature distribution as well. These measurements show that the drying front position is a linear function of time, which can be explained on the basis of a simplified model where only vapor transport is considered. Based on the measurements and the model, one can directly determine the permeability at high temperatures. Moreover, the results demonstrate that the drying front speed and temperature strongly correlates with the control of key material parameters (eg, water demand, binder content, etc). In particular, microsilica fume‐containing low‐cement castables displayed the highest vapor pressures, while regular castables generated the lowest vapor pressures reflecting the permeability of these materials.

Towards a single-phase mixed formulation of refractory castables and structural concrete at high temperatures

“…One of the critical points towards a robust and accurate implementation of a dryout solver is the evaluation of the sorption isotherm and its derivatives, which is pointed out in Equation (20). In the present work, the material reported by Bazant et al [2], in which the sorption is defined by a piecewise function depending on whether the concrete is saturated or not (see Equation 8and Section 2.2), was used as a reference.…”

“…To gain further insight on the model behavior, a sensitivity analysis that tracks the evolution with time of the maximum pressure, the temperature on the cold side and the evaporable water content was also carried out. Figure 5 shows all these quantities as a function of time for two situations: in the left column, K0 was fixed to a value of 10 -12 m/s and the thermal conductivity ⋋c was varied the interval [1,20] W/(m K), whereas, in the right column K0 was considered in the range [10 -14 , 10 -10 ] m/s and the thermal conductivity had a fixed value of 4 W/(m K).…”

“…For instance, the placement of thermocouples and pressure sensors within the prepared samples may completely change the local properties and, thus, their overall behavior [17]. Indirect measurements may overcome such limitations [18,19,20], however, they are limited to small samples (i.e., cylinders of 60 mm in height and 30mm of diameter for neutron tomography analyses [18]).…”

Experimental proof of moisture clog through neutron tomography in a porous medium under truly one‐directional drying

“…Specifically, neutron imaging has the advantage of providing an ideal contrast between wet and dry regions, as neutrons are highly sensitive to hydrogen atoms (much more abundant in the wetter portions of the sample, making them more neutron opaque) [4]. It also enables the 3D reconstruction of the water content distribution, a feature that is not available for regular NMR based techniques proposed in the literature [18,20].…”

“…4 It also enables the 3D reconstruction of the water content distribution, a feature that is not available for regular NMR-based techniques proposed in the literature. 18,20 Using such a method, Dauti et al was able to observe the effect of the aggregate size on the moisture mass transfer inside the samples, as well as to validate its mesoscopic thermohygromechanical model based on the collected data. 4,13 In their work, an increase of neutron attenuation ahead of the drying front was also detected, which suggested the occurrence of the moisture clog phenomenon, a key aspect to the current understanding of explosive spalling.…”