Dr.-Ing. Rodrigo Azcueta was delighted to attend and deliver his speech at the Wind Propulsion Conference, last week, in collaboration with the Royal Institution of Naval Architects at the International Maritime Organization in London. Two very positive and informative days were enjoyed with expert speakers, lots of networking and new contacts made.
Azcueta presented the conference paper and examples of the services provided by Cape Horn Engineering for performance prediction of all type of vessels using CFD, such as for resistance, propulsion, seakeeping, and manoeuvring simulations. Furthermore, he detailed the extensive research programme carried out using the SINTEC Ocean bulk carrier benchmark (SOBC-1) with comparison to towing tank results.
For the SOBC-1 benchmark, open water propeller simulations, resistance tests at model and full scale, propulsion tests with virtual disk and with rotating propeller, and finally 6 DOF simulations for motor-sailing conditions were presented and discussed.
Furthermore, the importance of verification and validation practices for EEXI/EEDI calculations, in particular when introducing ESDs and WASP devices to improve vessel performance, were highlighted. The requirement of demonstrating the qualification of the CFD solution by performing numerical uncertainty analysis following the ITTC and IACS recommendations was described. Finally, an investigation into the uses of Machine Learning with CFD for wing trim optimisation was presented.
All simulations presented were performed with the commercial CFD package STAR_CCM+ from Siemens Digital Industries in the latest version from mid 2022.
For the open water propeller simulations several strategies to set up the simulations were implemented and compared among each other and with cavitation tunnel results from SINTEF Ocean. The comparison of the simulated results to the experimental ones shows discrepancies of less than 2% on average.
For the towing resistance at model scale a perfect correlation (0.05% difference) with the measured total resistance in the tank was obtained with the k-epsilon model. Values for dynamic sinkage and trim are also well predicted. At full scale a correlation with the towing tank extrapolated results of between 0.6 to 3.7% was obtained, depending on the turbulence model used. Similarly, for the self-propulsion simulations, either with a simplified virtual disk model or with the actual rotating propeller, the agreement with SINTEF Ocean extrapolated results were withing 5%. However, in those cases there is always the uncertainty in the procedure that each towing tank uses for extrapolating the resistance to full scale.
Azcueta also presented results of Verification and Validation of recent projects for the calculation of the EEXI power curve, showing that a low level of numerical uncertainty of around 1% and correlation factors between simulation and sea trials of less than 3% are achievable. Furthermore, extracts of their Best Practice Guidelines and Demonstration of Qualification documents which are required by the ITTC and IACS were presented.
One of the main takeaways from the presentation was the introduction of Cape Horn Engineering’s latest workflow to directly assess the performance of different WASP devices with 6 Degrees-of-Freedom simulations. The 6 DOF simulations are basically all-in-one, simultaneous hydrodynamic and aerodynamic simulations where, together with the propeller, the thrust is generated by wings or any other WASP device on deck. The wind conditions above the water surface are modelled with an accurate wind profile, considering wind gradient. This results is a variable apparent wind speed and direction at different heights, due to the combination of an advancing ship and a wind velocity that varies with height. The wind forces on the WASP device, and on the hull and superstructure, induce drift and heel angles, and the rudder angle is adjusted during the simulation to balance the yaw moment of the whole system. The virtual disk model of the propeller adjusts the RPM to balance the drag and thrust forces. Thus, the vessel is motor-sailing and balanced in all six degrees-of-freedom. This type of simulations is much more accurate than running separate force models, for the hull and for the wings, and balancing the results with a Velocity Prediction Program. Thus, they can either be used to validate a VPP, or to directly compare different WASP technologies.
Finally, Azcueta presented the application of the 6 DOF simulations to three vessels. The first one is MV Regal, a well-known benchmark vessel for validation of CFD a ship scale. In this case, either two solid wings or two Fletner rotors were modelled and compared with the cases without any devices on deck. The second vessel is the SINTEF Ocean Bulk Carrier. In this case, two solid wings were modelled on deck and results also compared with the cases without the wings. Initially the wings were trimmed estimating the optimal sheeting angles. Following that a Machine Learning model of the wings’ interactions was used to find the optimum trim angles. After repeating the 6 DOF simulations the power savings improved by an extra 4% (the initial estimate was quite close to optimum). The last case presented was for an industry client; Humphreys Yacht Design with their patented FastRigs from the company Smart Green Shipping. Six FastRigs were modelled on deck of an existing Panamax bulk carrier and were analysed for six different wind strengths and directions. These simulations were compared with the cases with the exact same wind conditions but without the wings. The maximum power saving at 20 knots wind from the side was 54% and the average over the six conditions was 29%. It can be argued that optimising the wing trim angles, installing a variable pitch propeller and a hybrid power plant, the average power savings of 29% would increase even further.
There is little doubt that these levels of power savings are achievable, due to the nature of these type of simulations which considers all of the most important physical phenomena and interactions. It also directly compares the cases with and without the wings, and coming from an independent third-party subcontractor removes any bias. Figure 1 below shows the power savings in Watts and as percentages.