Oxygen transfer is an extremely important (perhaps the most important!) process in many biological systems. A number of medical devices have been developed to deliver oxygen (and remove carbon dioxide) to tissues (e.g. artificial lungs and artificial livers). In artificial lungs, complex matrices of gas permeable hollow fibres are used to promote blood side mixing and optimise oxygen uptake. However, the mathematical modelling of these devices has previously assumed that the perfused medium (blood) is a Newtonian fluid. While this assumption is adequate in a number of situations, novel devices for high rate transport involve flow in microchannnels and networks where the length scales make the Newtonian approximation invalid. In these cases it is essential to include non-Newtonian elements in the model.
The ultimate target of this work is therefore to optimise oxygen transport rates by theoretically modelling a number of different test device geometries using a non-Newtonian flow model and experimentally testing the most favourable device configurations. This approach could also be used for a number of different devices where it is necessary to transport of chemicals, drugs or proteins into or out of a fluid.
The above images show oxygen transfer (from the circles into the fluid) for two different values of fluid diffusion constant. Arrows indicate the direction of flow and colours indicate the concentration of oxygen.