Comparison of computational methods for hydrodynamic performance prediction of oscillating marine propulsors

J. A. Cole, G. Loubimov, M. Kinzel

Research output: Contribution to journalArticlepeer-review

3 Scopus citations


The biomimicry of fish fins has been suggested as an efficient and low-noise approach to propulsion for remotely operated and autonomous underwater vehicles. Propulsion of this type is achieved through the use of fin-like propulsors that undergo oscillatory motions. Two computational approaches of differing fidelity and computational space/time requirements have been used to calculate thrust and efficiency of oscillating marine propulsors in order to determine the region where each method is most appropriate for design applications. The higher fidelity and computational cost method is an unsteady Navier–Stokes approach. The lower fidelity and computational cost method is a lifting-surface potential flow method supplemented with strip theory to account for viscous effects. In comparison with two-dimensional experimental results, both methods were able to accurately capture magnitudes and trends in light to moderate thrust conditions (CT<0.7) assuming no dynamic stall occurs. Under these operating conditions, computational savings for three-dimensional analysis on the order of 5,000 processor-hours are achieved by using the lower fidelity method, making it the more appropriate choice for design applications. At high thrust conditions (CT>0.7) and in conditions where dynamic stall effects are prevalent (αmax>20, low Strouhal numbers), the higher fidelity method is required to achieve sufficient predictive accuracy.

Original languageEnglish (US)
Article number110002
JournalOcean Engineering
StatePublished - Dec 15 2021

All Science Journal Classification (ASJC) codes

  • Environmental Engineering
  • Ocean Engineering


Dive into the research topics of 'Comparison of computational methods for hydrodynamic performance prediction of oscillating marine propulsors'. Together they form a unique fingerprint.

Cite this