Project overview
Interest in wind assisted ship propulsion has increased dramatically in recent years, owing to the pressing need to
decarbonise our transport systems. The use of rigid wing sails has been found to be a promising approach, and only this
year a prototype full size wing sail has been mounted on a ship as part of the ‘Winds of Change’ Innovate UK project, of
which the University of Southampton is a project partner.
However, whilst the performance prediction and determination of optimal placement and operation of a single wing is
straightforward to predict using numerical methods, such as computational fluid dynamics (CFD), this is no longer the case
when large numbers of wings are employed.
For a single wing, the maximum thrust condition can be determined by parametric analysis, e.g. conducting simulations of
the wing at several orientations (‘angles of attack’) and selecting the scenario yielding the most thrust. A degree of freedom
is added when an independently actuated trailing-edge ‘flap’ is added, but the problem remains solveable with moderate
computational resource. The optimum wing angle will also vary both with the position of the wing on the ship, and with
the angle of the oncoming wind relative to the ship, as these parameters affect the wind ‘micro-climate’ that the wing sits
in but, again, the problem remains largely manageable. However, as the number of wings is increased, the number of
degrees of freedom of the problem increases by two per each additional wing, due to the need to determine both the
optimum wing angle and flap angle of the new wing, and it is not possible to know a priori what the best wing condition is,
since each wing will be located within its own wind microclimate, and the flow around each wing will influence that of
every other wing in a nonlinear fashion.
It is anticipated that large ships may need up to six wing sails, and if we optimistically assume five data points are required
per degree of freedom to determine the optimal performance using a parametric approach, a six wing configuration would
take 9.7 million times more data points to solve than a single wing configuration. Given that a computational prediction of
aerodynamic performance could take several hours to complete, this is clearly not feasible, and so we must either remove
degrees of freedom (e.g. by setting all wings to the same orientation), with a likely performance detriment, or else solve
the problem using a different approach.
decarbonise our transport systems. The use of rigid wing sails has been found to be a promising approach, and only this
year a prototype full size wing sail has been mounted on a ship as part of the ‘Winds of Change’ Innovate UK project, of
which the University of Southampton is a project partner.
However, whilst the performance prediction and determination of optimal placement and operation of a single wing is
straightforward to predict using numerical methods, such as computational fluid dynamics (CFD), this is no longer the case
when large numbers of wings are employed.
For a single wing, the maximum thrust condition can be determined by parametric analysis, e.g. conducting simulations of
the wing at several orientations (‘angles of attack’) and selecting the scenario yielding the most thrust. A degree of freedom
is added when an independently actuated trailing-edge ‘flap’ is added, but the problem remains solveable with moderate
computational resource. The optimum wing angle will also vary both with the position of the wing on the ship, and with
the angle of the oncoming wind relative to the ship, as these parameters affect the wind ‘micro-climate’ that the wing sits
in but, again, the problem remains largely manageable. However, as the number of wings is increased, the number of
degrees of freedom of the problem increases by two per each additional wing, due to the need to determine both the
optimum wing angle and flap angle of the new wing, and it is not possible to know a priori what the best wing condition is,
since each wing will be located within its own wind microclimate, and the flow around each wing will influence that of
every other wing in a nonlinear fashion.
It is anticipated that large ships may need up to six wing sails, and if we optimistically assume five data points are required
per degree of freedom to determine the optimal performance using a parametric approach, a six wing configuration would
take 9.7 million times more data points to solve than a single wing configuration. Given that a computational prediction of
aerodynamic performance could take several hours to complete, this is clearly not feasible, and so we must either remove
degrees of freedom (e.g. by setting all wings to the same orientation), with a likely performance detriment, or else solve
the problem using a different approach.
