Entry Date:
June 6, 2011

Predictive Modeling Via CFD: Design and Performance Evaluation of Propulsion Devices

Principal Investigator George Karniadakis

Co-investigator Chryssostomos Chryssostomidis


In this project we investigate effective designs of new marine propellers and waterjets in order to evaluate their performance under realistic conditions using advanced Computational Fluid Dynamics (CFD) tools. To this end, we employ the parallel opensource code NEKTAR developed at Brown University to allow fast 3D simulations of rotating flows. NEKTAR is based on the spectral element method, which combines the high accuracy of spectral methods with the geometric flexibility of finite element methods.

To simulate propellers and waterjets, however, requires the use of stationary and moving meshes that increases computational complexity greatly and limits the utility of the CFD tools. To address this issue we have developed a new approach – called Smooth Profile Method (SPM) – that represents the moving sub-domains as equivalent effective forces but on a stationary and structured grid. This removes the tyranny of complex mesh generation and results in 1000-fold speed-up factors, hence allowing the CFD tool to be used effectively on the design stage but also on performing evaluation of the propeller at different flow conditions. Preliminary results using the combined NEKTAR-SPM code to simulate ducted propellers and waterjet systems have shown very promising results.

NEKTAR/SPM simulation of a three-blade propeller shows streamwise vorticity iso-contours at moderate rotational speeds. Flow is from left to right.

NEKTAR/SPM simulation of a waterjet system. Shown on the right are instantaneous iso-contours of the streamwise velocity field at moderate rotational speed.