Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.632617
Title: Robust automated computational fluid dynamics analysis and design optimisation of rim driven thrusters
Author: Dubas, Aleksander
ISNI:       0000 0004 5362 2727
Awarding Body: University of Southampton
Current Institution: University of Southampton
Date of Award: 2014
Availability of Full Text:
Access through EThOS:
Full text unavailable from EThOS. Please try the link below.
Access through Institution:
Abstract:
The rim driven thruster is a novel electromagnetic marine propulsion device that uses a motor in its casing to drive a propeller by its rim. There are many interacting flow features posing a number of challenges when it comes to simulating the device with computational fluid dynamics. The primary concern is finding a suitable simulation method to capture the flow behaviour accurately, though a secondary challenge is created by the complex interactions creating a rugged design landscape that is difficult to optimise. A steady-state simulation method has been developed and a verification and validation process was conducted on a B4-70 standard series propeller as a baseline case. Results show a great sensitivity to computational domain size below a radial distance of five propeller diameters. The Re-Normalisation Group (RNG) k-e and k-w Shear Stress Transport (SST) turbulence models were compared and the k-w SST model was found to be the most robust due to its better handling of separation that occurs at low propeller advance ratios. To investigate the capture of rotor-stator interaction by the frozen rotor formulation an unsteady simulation method was developed. The unsteady method was also verified and validated, showing good agreement for a standard series propeller, and subsequently applied to rim driven thruster simulations. The results show the frozen rotor formulation does capture some variation and has reasonable agreement with thrust variation over one rotation, but does not predict the variation in torque accurately and thus is considered insufficient for rotor-stator interaction modelling. While the capture of rotor-stator interaction is flawed in frozen rotors, if the stators are omitted, the steady state simulation method is suitable for performance prediction. Given the computational cost of full unsteady simulation, steady state was chosen for the objective function calculation method for the design optimisation. A library of functions was written to robustly automate the geometry creation, mesh generation, solution and post-processing. An initial design study of the sensitivity of 13 parameters showed that the most significant variables were pitch distribution, thickness distribution and hub diameter. These were factored into a second design optimisation study of six parameters, using Kriging for surrogate modelling, to produce an improved rim driven thruster design. The improved design features a greater pitch at the tip exploiting the lack of tip-leakage experienced with rim drive. A high sensitivity of the hydrofoil to Reynolds number was discovered and exploited by increasing the blade thickness and pitch to make the blade section produce more force over a greater area of the blade. The open water efficiency of the improved design is 0.06 higher than the baseline design, showing the optimisation was a success.
Supervisor: Sharkh, Suleiman Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.632617  DOI: Not available
Keywords: TJ Mechanical engineering and machinery
Share: