HUO Yu-long, HU Ming-jian, TANG Jian-jian. Experimental and predictive modeling of coral sand infiltration in different pH environments considering time series[J]. 2025, 46(12).
HUO Yu-long, HU Ming-jian, TANG Jian-jian. Experimental and predictive modeling of coral sand infiltration in different pH environments considering time series[J]. 2025, 46(12).DOI: 10.16285/j.rsm.2025.0022.
The permeability of coral sand is related to the formation of freshwater lenses in blown-in islands and reefs
and is an important parameter describing the permeability characteristics of reef sand strata
which is of great significance to the sustainable development of islands and reefs. The permeability of coral sand is affected by the percolating solution pH and permeation duration in addition to the particle size
gradation
and pore characteristics. To investigate the effects of permeate duration
particle size and gradation on the permeability of coral sand in different pH environments
we carried out conventional permeability tests
and analyzed the changing rules of mineral content and particle fraction of the sand columns before and after the tests by using D8Advance X-ray diffractometer and laser particle sizer. Results show that
under neutral percolating solution and fixed porosity
the initial hydraulic conductivity increases with the effective grain size. The initial hydraulic conductivity coefficient and the curvature coefficient follow a negative logarithmic relationship
while the initial hydraulic conductivity coefficient and the uniformity coefficient follow a logarithmic relationship. Under acidic conditions
hydraulic conductivity has a quadratic relationship with time; under neutral conditions
hydraulic conductivity remains largely unchanged. The increase in hydraulic conductivity intensifies as pH decreases. In acidic environments
the calcite content of coral sand increases
while aragonite
calcium and magnesium calcite
and rock-salt contents decrease; mineral changes are not obvious under neutral conditions. The percolating solution first passes through a layer with a significantly increased fine-particle content; under the same layer conditions
a lower pH leads to a more pronounced increase in the fraction of fine particles. We developed a preliminary regression-based model to predict the initial and final stabilized hydraulic conductivity. A machine-learning model incorporating time effects and percolating solution pH was also developed. These models can be used to assess reef-sand permeability and to analyze the evolution of freshwater lenses.