Developing Fundamental Models of Colloid Transport and Absorption in Sand Filters

This work was undertaken as part of an Industrial Collaborative Awards in
Science and Engineering (iCASE) research programme, jointly funded by the National
Nuclear Laboratory (NNL) and the Engineering and Physical Sciences Research
Council (EPSRC). The aim was to probe the mechanisms of clogging of sand bed
filters using particle based computer simulation methods. Existing models take a top
down approach, making use of an empirical clogging parameter. Simulation holds the
prospect of relating this parameter to properties of the effluent and the sand bed.
The problem was approached using two computational methods: molecular
dynamics, and smooth particle applied mechanics. The molecular dynamics model
yielded successful results, qualitatively agreeing with existing experimental data with
regards to the rate of deposition within the bed, and the associated observed pressure
drop. The model was systematically explored by varying the nature of the colloidfluid-sand forces, the geometry and packing fraction of the sand bed, and the
concentration of the colloids. An investigation into the fractal nature of the deposits
was also performed, suggesting that a lower fractal dimension creates greater physical
hinderance to the flow. This serves as additional validation for existing theories.
The smooth particle model yielded less successful results. Substantial
parameterisation of the model was undertaken, however, the model still showed signs
of instability under certain conditions. Again, it produced qualitative agreement with
existing literature, but showed substantial deviation from the results gained from the
molecular dynamics model. Ultimately, further parameterisation of this model is
required to allow for a more effective comparison of the models.