# Experimental feasibility of dual-energy X-ray tomography for two-phase density analysis in bentonite during water infiltration

## TL;DR

This study shows that dual-energy X-ray tomography can accurately track water infiltration in bentonite over time when corrected for imaging errors.

## Contribution

The study demonstrates the experimental feasibility of DECT with correction methods for 4D water content analysis in bentonite.

## Key findings

- DECT with corrections can quantitatively estimate 4D water content in bentonite with reasonable accuracy.
- Both scattering and beam hardening corrections are essential for reliable results in material decomposition.
- Gravimetric and deformation measurement methods validated the DECT-derived water content distributions.

## Abstract

The feasibility of dual-energy X-ray computed tomography (DECT) with experimentally implemented scatter and beam hardening corrections to quantitatively determine the 4D (3D spatial with time) evolution of partial densities of bentonite and water was examined. Compacted bentonite samples were imaged using an X-ray microtomography system with various X-ray spectra before and after two days of water infiltration. The effects of scattering and beam hardening in the projection images were corrected using beam-stop array measurements and signal-to-thickness calibration. A post-reconstruction material decomposition (MD) technique was applied to obtain the partial density distributions of bentonite and water, from which the water content distributions were subsequently derived. The results were validated by physically slicing partially saturated bentonite samples and measuring the water content of the slices gravimetrically. Additionally, a previously developed deformation measurement-based X-ray tomography method was used to derive reference results. Comparing the results from MD with those from the other techniques demonstrated that DECT can yield quantitative estimates of 4D water content distribution in bentonite with reasonable accuracy in this experimental set-up. When testing the impact of correction methods on the results, it was found that both scattering and beam hardening must be corrected, as the post-reconstruction MD is sensitive to errors in spectral measurements.