Identification of self-sealing processes in claystone-bentonite mixtures by digital image correlation

Processed claystone-bentonite mixtures are considered as potential backfill material for drifts and shafts in deep geological high-level radioactive waste repositories, foremost due to their strong self-sealing and radionuclide retardation properties. Self-sealing is mainly facilitated by the materials’ swelling capacity when getting in contact with water from the surrounding host rock. A timely sealing of fractures and voids is of high significance in the isolation of radioactive nuclides. However, a specific duration for a fracture to close upon wetting, the exact processes involved (next to interlayer water uptake in smectite) and the necessary width of a swelling zone are unclear. In order to investigate the self-sealing, past research focused almost exclusively on the (free) swelling capacity of bentonites or claystone and hence neglected other processes for permeability reduction upon saturation. More recently, µCT techniques have been used to overcome the lack of direct evidence of fracture closure in clay rocks and to provide a mechanistic understanding of self-sealing processes. To obtain a deeper understanding of the self-sealing behavior of crushed claystone-bentonite mixtures, we map the swelling induced strain in a simple 2D approach using digital image correlation (DIC). For this purpose, an artificial fracture in specimen with different claystone/bentonite ratios but defined initial bulk density and initial water content is saturated and a camera taking pictures of the sample surface every two minutes documents the wetting front as well as the developing strain. For DIC the free software GOM is used that allows tracking strains beneath the order of pixel resolution, which in our setup is about 10 µm. In previous experiments with Opalinus clay localized tiny fissures appeared when the wetting front spread into the specimen and contributed to the closure of an a-priori induced fracture. Further, a straining zone, advancing ahead of the wetting front, lead to a more evenly volume expansion that finally closed not only the artificial fracture but also the newly formed tiny fissures. In general, we assume a somehow similar self-sealing behavior in the claystone/bentonite mixtures, but we expect different responses depending on the claystone/bentonite ratio. Additional tests (e.g. wettability, BIB-SEM, zeta potential) could be used to understand the observed behavior.