Research Project: Computational Studies of Fluid and Chemical Transport in Vascular Devices
Computational fluid dynamics (CFD) is a subset of computer modeling used to simulate the flow of fluids and the physical forces acting on the fluid. CFD is already being used to develop and prototype blood-contacting medical devices, such as prosthetic heart valves and ventricular assist devices (VADs). However, the use of CFD to demonstrate product safety in FDA premarket device applications and postmarket investigations has not been adequately validated. This is especially true in the final step of predicting biological responses (e.g., blood damage, thrombus formation) from the purely physical results (e.g., pressures, velocities, shear stresses) of the simulations. Even just the physical results generated by CFD are subject to considerable error as compared to experiment, as was demonstrated in a recent computational interlaboratory study sponsored by the CPI CFD/Blood Damage project organized by DSFM personnel. Participants' computations of a relative hemolysis index from the CFD simulations also showed a great deal of scatter. Thus it has not been conclusively shown that device safety can be reliably predicted by CFD. This project addresses this difficult problem by providing expertise to the CPI CFD/Blood Damage project, running our own comparisons of CFD to laboratory experiments in example cardiovascular medical devices, and comparing different modeling software under similar conditions. Ultimately the studies outlined in this project will lead to development of a Guidance Document for the use of CFD in device applications and investigations for industry and FDA personnel.
Finite Element Analysis (FEA) is another subset of computer modeling. The model consists of the detailed geometry of the device and the mechanical properties of the materials used, and then predicts stresses and strains in solid structures and materials from the applied external forces and deformations (e.g., as in a drug eluting stent). As with CFD, the method reduces costs by allowing virtual design and prototyping rather than actually building and testing each iteration. FEA can also predict failures due to unknown stresses by showing problem areas and allowing designers to see all of the stresses calculated within the device. The method is an essential part of fatigue and durability testing. Thus its usefulness in predicting safety in device applications has more of a successful history than does CFD.