It has been a year since TU Delft started a research on ‘Optimization of propulsion in and on water’. So it is time for an update! The target of the project at TU Delft is to find the optimal rowing stroke from a hydrodynamic perspective.
Replicating the oar blade movement in water
First of all the oar movement must be executed, otherwise there is nothing to measure. Ideally the whole oar blade movement is replicated in water. Of course one could let a crew row on the Bosbaan in Amsterdam and do field measurements. However, measuring hydrodynamic flows involves expensive, fragile and sometimes dangerous equipment (high-speed cameras, lenses, high power lasers, etc.) which all don’t do too well if used outdoors. Also the human factor, i.e. limited repeatability of the movement, is reason for concern. So instead of letting a crew perform the movement in real-life, a mechanical alternative is sought-after.
The oar blade movement is formed by a translational movement (the boat’s movement), and a rotational movement (the rotation of the oar). So in an earth-bound frame of reference (i.e. looking from a bridge at the oar blade down below) the oar blade follows a curl-shaped path (see figure right: data provided by VU Amsterdam). To replicate this path the possibilities of a towing tank carriage with a rotating oar as well as an industrial robotic arm are under investigation.
Measuring the hydrodynamics of the oar blade
Measuring flow was traditionally done by using pressure sensors (for instance pitot-tubes), but this only leads to a single point measurement. This all changed by the invention of PIV (Particle Image Velocimetry) enabling measurement of a flow field instead of a single point. By adding tiny visible particles to the water which follow the flow very well, the flow can be visualized as shown in video 1 (see below). In the video the water surface is seeded using coffee powder and a flow is created (two vortices) by dragging an object through the basin. A high-speed camera perpendicular to the water surface is used to capture the flow (figure on the left). The movement of the flow field (which is assumed to be the movement of the particles) is calculated from the images by calculating the displacement of the particles which leads to instantaneous snapshots like figure 3 (see below). From these velocity fields all kinds of interesting quantities such as the Z-vorticity (swirling strength in plane) can be calculated which also shows where vortices in the flow are present. An analysis of the flow field of the recorded movie is shown next to the original movie in video 2 (at least the part of the movie after which my hand left the field of view;: see below). The red areas are counter-clockwise rotating vortices while the blue areas correspond with clockwise rotating vortices.
Simulating the hydrodynamics of the oar blade
Experiments are relatively rigid. Most setups are limited in the different movements they can replicate while there are many variations possible in real-life. To somewhat predict the effects of movements which cannot be replicated experimentally computational fluid dynamics is used. To test the feasibility of the application of CFD in this research first a 2D model is made. Using Ansys Fluent a dynamic mesh is made and an oar blade following the blade path mentioned above is calculated. Again the Z-vorticity is calculated to see where the vortices are shed by the oar blade, shown in video 3 (see below).