For this reason, the battle for Gold begun long ago in the offices and testing facilities of research centres, where the sports equipment used by the competitors in the Beijing Olympics has been constantly optimized and improved. Canoeing is one of the fields where extensive research has been performed into minimizing hull resistance using state of the art measurement and experimental techniques, backed extensively by Computational Fluid Dynamics simulation using CD-adapco software.

Fig. 1 The canoe hull in its static (left) and dynamic position in the flow.

The Polish company, Plastex Composite, recognized worldwide as one of the leading producers of competition canoes, has provided state-of-the-art equipment for Olympic Games and World Championships for many years. In 2005, during the 1st World Canoe Championships in Pozna, 56 out of 81 medals were won by competitors using Plastex boats. The boats themselves are designed by the company's owner, Ryszard Seruga, in cooperation with Tomasz Bugalski, Ph.D, from the Ship Design and Research Centre (CTO) S.A. - Poland.

In the spring of 2007, Plastex and CTO began to work on the shape of new canoes for the Olympic Games in Beijing. Optimization of the new design started with extensive investigation of the existing hull shapes in CTO's model basin, with the principal aim of determining the dependency of the canoe's performance on the basic design parameters and initial trim. Although such experiments provide a large amount of reliable data in a short timescale, they do not always illuminate the physical mechanisms that affect the performance of the hull. For this reason, the experimental research was widely supported with extensive CFD simulation, which is more suited to a detailed comparison of the influence of flow properties such as wave elevation and pressure distribution on the hull for different designs.

Fig. 2 K1, K2 and K4 canoes – free surface elevation.

Each CFD simulation considered a canoe hull towed at constant speed through calm water. The computational analyses yielded numerical data, such as hull resistance, as well as allowing the design team to visualize the flow field around the hulls, thereby helping them identify the mechanisms behind variations in physical performance, e.g. bow and stern wave height or wave interaction. After testing the existing boats, the best design was chosen based on analysis results and work on new designs began. By implementing CFD into the design process, timescales and costs have been significantly reduced. The viability of each new design was first tested numerically, so that only a small number of optimized designs were selected for manufacturing and testing in the model basin. Final tests were carried out in real conditions - with the professional competitor rowing along the basin. The simulations were carried out using the Volume Of Fluid (VOF) model for multiphase flows and the RNG k-epsilon turbulence model using specially constructed 1.5 million cell hexahedral meshes. The surface models of the existing canoes were obtained by digitally scanning the hulls, carried out using an ATOS II optical scanner and a TRIPOD photogrammetric system provided by GOM GmbH.

Fig. 3 Example of the computational mesh.

Due to the fact that the Olympic canoes travel at relatively high speed (of the order of 6m/s), it is absolutely necessary to take into account the dynamic trim and sinkage of the hull in the numerical analysis of the flow around it, requiring either experiment data, or if not available, adjusting the hull position during the CFD simulation until force and moment equilibrium is reached. Although this can be done iteratively, based on the hull hydrostatics, CTO uses an in-house, automated procedure for coupling the flow solver to the hull motion equations, allowing for accurate evaluation of the canoe's position. The computational mesh in this approach remains rigid - it moves together with the hull without relative motion of the nodes, which proved to be sufficiently accurate, robust and very simple - no re-meshing is required when the hull changes its position.

At present, the CFD analyses and model tests of the canoe's performance are limited to steady-state analyses - the hull is towed with constant speed and fixed centre of mass. Such a simplified approach allowed for effective optimization of the hull shapes based upon resistance with an identified 1% reduction - which could easily be the difference between Olympic Glory and ignominious defeat. The use of CFD methods allowed reduction in costs by limiting the number of designs tested and so reducing the need to manufacture many hull shapes. Further to this, identification of the flow phenomena by CFD allowed optimization to be carried out far quicker than previously possible.

It is very likely that the present shapes of the Olympic canoes are already very close to the absolute minimum resistance obtainable in steady flow. In the future, significant further development will only occur by optimizing the dynamic behaviour of the hull, which means taking into account all the phenomena encountered during a race - motion of the competitor and unsteady forces exerted on the hull. For that reason, the next step for CTO is to study 6 degrees of freedom (6-DOF) analyses to help aid hull optimization and to try to adjust the shape so as to minimize the loss of energy due to the hull motion and interaction with other canoes. CTO have already performed a trial simulation of the Wigley hull in head waves. In this simulation, the hull was free to pitch and heave (2-DOF motion), while sailing with constant speed and fixed zero drift angle results revealed good accuracy and robustness of the method.

Fig. 4 (Below) Example of the hull dynamics analysis – Wigley hull in waves.

Read more ...