A REDUCED ORDER MODEL OF PROPELLER UNSTEADY FORCES FOR COMPUTATIONAL FLUID DYNAMICS SIMULATIONS
Open Access
- Author:
- Koncoski, Jeremy J
- Graduate Program:
- Mechanical Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- July 16, 2010
- Committee Members:
- Dr Eric G Paterson, Thesis Advisor/Co-Advisor
Eric G Paterson, Thesis Advisor/Co-Advisor - Keywords:
- propeller
unsteady force
model
CFD
computational fluid dynamics - Abstract:
- Propellers generate unsteady forces and moments when operating in flowfields containing spatial and temporal variation. Time-accurate simulation of propeller unsteady forces using a viscous flow solver requires an expensive, full-wheel, dynamic grid of the rotating geometry. These computational costs can be reduced by modeling the propeller unsteady force response using a volumetric body force representation of the propeller. Existing body force models are only quasi-steady, defining propeller powering using open-water operating curves. Other methods specify the unsteady body force by linking to separate lifting-surface simulations of the propeller, which are expensive. Therefore, integrating a propeller unsteady-body-force model with a viscous flow solver can simplify gridding and reduce computational costs for maneuvering and sea-keeping simulations of ships and submarines, which are characterized by propeller inflow variations. This work calibrates an unsteady, sharp-edged gust response model for low-aspect ratio blades in sheared flow, typical of turbomachinery applications. The gust-response model is integrated with an existing, steady, body-force representation of a propeller in a viscous flow solver. The resulting unsteady-body-force model is validated against geometrically-resolved, unsteady simulations of a propeller in spatially-varying flow. The unsteady-body-force model correctly predicts propeller unsteady force and moment directionality based on inflow harmonic content. The magnitude and phase of forces and moments generated by the unsteady-body-force model compare well with geometrically-resolved, unsteady simulations, and match published results from unsteady lifting-surface codes. The unsteady-body-force model reduces computational costs and grid size by 44% relative to the geometrically-resolved, unsteady simulation. The current work yields a 50% improvement in accuracy over gust-response models for infinite wings, and eliminates effective wake calculations required by lifting-surface codes.