Maribot Vane - the KTH Autonomous Sailboat
The project includes methodology development, design, completion and demonstration of a autonomous sailboat intended as test platform for further technological and methodological development with the long-term goal to push the technology front for autonomous data acquisition forward.
The project addresses a growing need of ocean sensing for the benefit of maritime safety, environmental monitoring, research in climate research, meteorology, oceanography, etc. Possible modes of operation of the sailing craft are place-bound (loitering) long-term measurement, routine patrolling, sniffing of possible emissions, hydrographic survey or the like.
The project connects strongly to both completed and ongoing projects concerning hydrographic surveys in shallow waters as well as to other research in maritime robotics e.g. at the division of Underwater technology at KTH, University of Gothenburg, Stockholm University, FMV, FOI etc.
The project has high relevance, especially in light of the rapidly accelerating interest in automation and digitization of shipping. In the ongoing development towards a future scenario where a range of vessel types, (slow, fast, wind- driven, wave-driven, manned, unmanned) should co-exist and interact safely on the seas, we must now study the modes of integration of this type of autonomous vessel types.
The picture shows the sailing drone during a test in the Stockholm archipelago. One special feature is the rigid free-rotating rig which releases the vehicle from the need to sheet (control) of the rig relative angle to the boat hull. The rig adjusts its angle to the wind automatically increasing its robustness. Mechanical feedback of the relative angle between the rig and hull enables passive wind rudder feature, which further reduces the need for active control.
Aeroelastic Sailplane Modelling
The Centre for Naval Architecture at KTH is involved in the field of performance prediction of sailing yachts. As a further step in this field specific research has been made to model the aerodynamics of the yacht. A complex system such as sailing yachts calls for advanced methods in order to be accurately modeled. Since the sails are flexible and the rig is elastic the sails will alter their shape when loaded. Another sail shape generates a different pressure distribution. This motivates the need to also evaluate the structural response of the rig and sails.
Work has been performed at the KTH Centre for Naval Architecture to develop an aeroelastic tool for calculating the elastic response in sails and rig due to the aerodynamic pressure in upwind sailing conditions. The aerodynamic calculations are based on the Vortex Lattice Method and the structural response of calculated with Finite Element Method.
Sails are flexible and have generally a high in-plane stiffness meaning that large displacements must be accounted for but strains can be expected to remain small. The large displacements results in the need of a nonlinear formulation of the finite elements and the structural problem is solved using an iterative numerical solver routine.
Vortex ring elements are used in the VLM implementation and effort has been put to develop a realistic force free wake model. The VLM-implementation shows good correlation to wind tunnel tests and to other commercial VLM-codes.
Results from the work shows that there is a significant value in using a fully aeroelastic model when studying the behavior of yacht sails.
Velocity Prediction Program (VPP)
The development of reliable VPPs is a crucial part of optimizing yacht performance. The result of a VPP is a performance diagram (polar plot) that states the boats optimal (target) speed though the water as a function of the sailing conditions at best possible trimming. The implementation of a generic VPP, founded on the relevant and dominating physics of the yacht in its environment in Matlab is a compulsory part of the undergraduate program in Naval Architecture at KTH. The VPP is then used to optimize the performance of a yacht constrained by a set of rules. Empirical updating of the polar diagram is also addressed.
Weather Routing is another component of the project. The problem and art of finding the optimal route given a certain weather forecast is solved by a variation of deterministic dynamic programming. The routine enables quick onboard optimization of the route within a discretized geographic mesh. The route, steering commands, time table and isochrones are typical outputs. The sensitivity of the route to perturbations of the weather forecast in time, strength and space is possible within the optimizer.
Web based simulator for sailing boat performance
A web based sailing simulatior is developed. The simulator enables virtual sailing anywhere worldwide with any sailing boat. The simulations can be done in in true weather batch mode or in real time. Fictive races can be sailed where both strategy and boat performance are studied. Races can also be done parallell with true real life races such as Route-du_Rhum, Volvo Raound the World Race, Gotland Runt etc.
Onboard sail-race tactics software & student racing
An on boat sail race tactics tool is developed. The software reads data from the on board instruments such as GPS, wind sensors, log, compass and enables e.g. visualization of weather forecast files in the GRIB-format, route optimization, trend analysis, navigation, performance comparisons against VPP and scenario evaluation. A student endevour to participate in off shore sail racing is sponsored by KTH, Naval Architecture. The racing gives the students the possibility to apply knowledge and skills attained in courses in real life racing. Read the students story about the Gotland Runt race here.
IACC start optimization
As a special masters thesis project the pre start sequence of an Americas Cup (IACC) yacht was studied and optimized. A semi-empirical dynamic VPP was developed for the Victory Challenge team boats. Using the VPP in association with the time integration of a set of manouvering equations of motion, the time-to-burn was calculated. Hereby the time-burning pre start manouver could be optimized. A pdf report is available.