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

TL;DR

This study uses neutron tomography to directly demonstrate moisture accumulation ahead of the drying front in porous media, providing insights into explosive spalling phenomena in concrete and refractory materials.

Contribution

It introduces a novel experimental setup for truly one-dimensional drying, eliminating artifacts and enabling direct observation of moisture clog formation.

Findings

Moisture accumulates ahead of the drying front in porous media.

The experimental setup reduces artifacts like beam hardening.

Provides data for validating numerical models of drying processes.

Abstract

Structural failure of concrete buildings on fire and complete destruction of the monolithic refractory lining during their drying stage are dangerous examples of the effect of explosive spalling on partially saturated porous media. Several observations in both cases indicated the presence of moisture accumulation ahead of the drying front, which are in tune with the most common theories on the explosive spalling of concrete. Previous studies have shown evidence of the existence of this phenomenon, however, they were biased by artifacts and experimental limitations (such as the beam hardening effect and changes in the microstructure of the material due to the presence of pressure and temperature sensors). In the current work, rapid neutron tomography was used to investigate the in-operando drying behavior of a high-alumina refractory castable, proposing a novel experimental layout aimed at a truly one-dimensional drying front. This setup provided more realistic boundary conditions, such as the behavior of a larger wall heated from one of its sides, while also preventing some nonphysical artifacts (notably beam hardening). By eliminating these aspects, a direct proof that moisture accumulates ahead of the drying front was obtained. This work also lays the basis for further studies focusing on the response sensitivity analysis to boundary conditions and other parameters (e.g., heating rates and properties of the sample related to the moisture clog formation), as well as useful data for the validation and characterization stages of numerical models of partially saturated porous media.