CFD modeling of biomedical and mechanical systems involving rotating machinery

Bryan W Reuter, Purdue University


Computational fluid dynamics (CFD) studies have become a significant part of the design and development process. An important type of engineering problem which is becoming widely studied is the simulation of systems involving rotating machinery. This work explores two of the popular methods for simulating rotation within a CFD framework: Multiple Reference Frames (MRF) and Sliding Mesh models. Currently, MRF is more popular due to the significantly smaller computational demands, but it is limited to systems where the relative position of rotating components to stationary components is not important with respect to the parameters of interest. The main focus of the study is the development of a Ventricular Assist Device (VAD) for supporting patients born with Hypoplastic Left Heart Syndrome, a congenital heart defect which presents as a malformation of the left ventricle. This VAD is an axial blood pump known as a viscous impeller pump (VIP) and simulation results are validated against separately obtained data from Particle Image Velocimetry (PIV) as well as experimental hydraulic performance. The Sliding Mesh formulation was found to more accurately predict the pressure rise created by the VIP. The Sliding Mesh and MRF approaches perform similarly when comparing the velocity field with PIV data. The MRF method is recommended based on the much higher computational cost associated with obtaining time-averaged velocity profiles using the Sliding Mesh method. Additionally, four different turbulence models are employed: the standard k – epsilon, Realizable k – epsilon, standard k – omega, and k – omega Shear Stress Transport (SST). The choice of turbulence model had a significant effect on the resulting flow field and blood damage prediction. The Realizable k – epsilon model gave the best agreement with PIV profiles, most accurately predicted the pressure rise across the pump, and was able to capture mean flow features indicated by experiment.




Frankel, Purdue University.

Subject Area

Biomedical engineering|Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server