Analysis of flow inside artificial lungs using computational fluid dynamics

Kenneth Gage, University of Pittsburgh

Photo of Kenneth Gage

Recent years have witnessed a dramatic increase in the use of computational fluid dynamics (CFD) to predict the performance of artificial organs. CFD provides spatial information not obtainable with more limited modeling techniques, allowing investigators to evaluate regional device performance. Experimental confirmation of the flow field is a critical component of model verification, yet few tools are capable of visualizing fluid flow inside complex, opaque devices such as membrane oxygenators. With its high spatial and temporal resolution, fluoroscopy is a promising imaging technique for visualizing flow in these systems. Combined with appropriate pressure field information, a thorough evaluation of the CFD flow model of an artificial lung should be possible.

The finite volume based CFD software package FLUENT5 (Fluent, Inc., Lebanon, NH) was used to predict the flow field inside a Maxima membrane oxygenator (Medtronic, Inc., Minneapolis, MN). Blood was modeled as a Newtonian fluid and the fiber bundle as an Ergun porous medium. Fluoroscopic imaging was performed with an angiographic unit (OEC-Diasonics, Salt Lake City, UT) to visualize fluid flow in the oxygenator. A bolus of contrast agent was injected into the oxygenator during steady flow conditions and a series of timed images generated from the resulting film. The optical flow in the image sequence was determined and used to approximate the velocity field within the oxygenator. In a separate experiment, pressure measurements were taken at critical points along planes of symmetry and interest under constant flow conditions. Measurements were taken from taps located on the outer housing of the device and used to generate a rough three dimensional map of the pressure distribution. An evaluation of the current models used in the artificial lung simulation is presented based on comparisons between the predicted and experimental results.

Abstract Author(s): Kenneth L. Gage