Numerical Modelling Capabilities

Numerical models are useful for testing a number of different layouts such that inappropriate designs can be discarded quickly and at an early stage. Once a preliminary design has been chosen a physical model can be constructed and tested. Minor changes can then be incorporated into the physical model such that an optimum design solution is reached.

An important feature of Ocean Energy Devices is that they will usually perform best when operating in waves that match a particular region of their scatter table. An optimum Wave Height and Period is deduced for each individual device via numerical analysis. This determines particular ‘high performance’ areas where waves of this type predominate and this converter is projected to perform well. Once these optimum values are determined the device can be modified and re-tested in the wave tank so that the high performance capabilities are matched to the deployment area.

Numerical modelling is a tool for calculating critical device characteristics so as to accurately predict offshore performance. Numerical methods are used for analysing wave disturbance, harbour resonance, device power take off, hydrodynamic parameters, fluid dynamics, meshing, turbulence modelling, fluid structure interaction, validating physical models.... Numerical modelling, especially when applied via modern high speed, high capacity computers, has the benefit of automating extensive computational effort. Powerful numerical modelling software is used for the optimisation of device development. Software packages including MatLab, Simulink, Mathematica, Ansys CFX, Workbench and ICEM, WAMIT, MultiSurf, Mike21, Microsoft Excel... are proficiently used and the results interpreted.

Numerical modelling has become a viable means of predicting wave interaction and wave transformation at the coastline. When areas being analysed are too large to be represented at an adequate scale in a physical model numerical models can still be used. Numerical models need to provide accurate prediciton, have a wide range of features in order to cover different applications, and to be computationally efficient, especially when a large area, data set or time frame is being studied.