Dynamic loading associated with earthquake shaking can lead to the liquefaction of loose, saturated sands, essentially transforming competent geological material into quicksand. Any structures - natural or man-made - that exist on these deposits will suffer catastrophic damage due to the loss of strength associated with liquefaction. Cemented soils, however, are less prone to liquefaction than loose granular materials. Cementation can occur naturally due to the precipitation of certain minerals or may be induced via chemical injections. More recently, novel biological techniques such as microbially induced calcite precipitation (MICP) have been used in the laboratory to mimic the natural cementation process. The biological cementation techniques have the potential to be more sustainable than traditional chemical injection methods. This work will explore the effects of bio-cementation using a tightly integrated numerical-experimental program. The expected outcome of the work is a modeling framework: if the level of cementation is known, then bio-cemented soil behavior can be predicted. Thus, the models dveloped as part of this work will help engineers use bio-cementation to prevent liquefaction during future earthquakes. The work will be broadly disseminated through course development and a collaboration with the Oregon Museum of Science and Industry.
Bio-cementation such as MICP has the potential to increase liquefaction resistance by increasing the cyclic strength of sand, reducing the generated excess pore pressures, and reducing settlements due to dynamic loading. The behavior of cemented soil is extremely dependent on the mineralogy of the cementing agent. Sands cemented artificially with chemical agents, such as lime, ordinary Portland cement, and gypsum behave differently than naturally cemented soils. Thus, existing models used to simulate the behavior of chemically-cemented sand are not appropriate for bio-cemented sand. Before implementing MICP for liquefaction mitigation, a better understanding of the underlying physics governing the constitutive behavior of bio-cemented sands is necessary. Measuring the underlying micromechanics (e.g., changes in particle roughness, calcite fines generation during shearing) is difficult with traditional experiments, so discrete element method simulations will be used to help study bio-cemented sand at the microscale. Element- and particle scale behavior of bio-cemented sands will be assessed through a combination of strength testing, particle-scale measurements, and X-ray computed tomography. Results from these experiments will be used to develop and calibrate numerical models to predict the bulk response of bio-cemented sands subjected to static and dynamic loading.
