VIRTUE: integrating CFD ship design

Novel ship concepts, increasing size and speed, and strong competition in the global maritime market require that a ship’s hydrodynamic performance be studied at the highest level of sophistication. All hydrodynamic aspects need to be considered so as to optimize trade-offs between resistance, propulsion (and cavitation), seakeeping or manoeuvring. VIRTUE takes a holistic approach to hydrodynamic design and focuses on integrating advanced CFD tools in a software platform that can control and launch multi-objective hydrodynamic design projects. In this paper current practice, future requirements and a potential software integration platform are presented. The necessity of parametric modelling as a means of effectively generating and efficiently varying geometry, and the added-value of advanced visualization, is discussed. An illustrating example is given as a test case, a container carrier investigation, and the requirements and a proposed architecture for the platform are outlined. 2 CURRENT PRACTICE AND FUTURE OF HYDRODYNAMIC DESIGN Today’s transportation demands often lead to ships which expand the limits of experience, either in terms of speed requirements, unconventional hull forms or very large designs with rather unusual main particulars. This is due to a rapidly changing maritime market and it is here where computational fluid dynamics offers its full potential for hydrodynamic design. While it is always desirable to obtain the best possible solution for a given problem this, unfortunately, is often prohibitively expensive with regard to time and budget. Model basins and design consultancies therefore develop, utilize and rely on a range of codes for hydrodynamic analysis to complement model test campaigns. The codes comprise (purpose built) potential flow methods, e.g. for wave resistance predictions or seakeeping analyses, as well as more complex and resource intensive RANSE methods. Up till now, manual processes prevail where just a few hull form modifications are studied and the evaluation and interpretation of results is mostly based on the intuition and experience of the specialists doing the work. For a resistance problem, for instance, this means that the wave pattern and the pressure distribution are assessed, that maximum wave elevation at the hull and in the near field are compared, and that local flow phenomena are inspected. All this is largely based on the skill of the consultant applying criteria that may be obvious and sensible to humans but which are sometimes qualitative and sometimes difficult to formalise. Running much more complex (and automated) optimisations for many hundreds and even thousands of designs, however, does not allow for any (substantial) human interaction and relies on the correct ranking of the competing variants. Being computationally efficient and sufficiently accurate it has been shown that potential flow codes could already be integrated very successfully into formal optimisation processes, see for instance (Valdenazzi et al., 2003) and (Harries, 2006). Rising challenges, however, call for future solutions which cover several, possibly conflicting, objectives and, eventually, the complete (hydrodynamic) life-cycle of a ship including its performance in relevant seaways, during (critical) manoeuvres and the like. A major drawback associated with the many different CFD methods needed to simulate the different flow phenomena of interest is the large variety of different (data) requirements. Presently, a substantial effort in handling, converting and providing the right data is required, be it geometry, computational parameters, grids, flow fields, visualisations or further assessments. Consequently, so as to perform such optimisations economically, the many special purpose tools used in the process need to be highly integrated and provide: o Ready availability of geometry for all analyses. o Sophisticated geometric modelling techniques to realize form variations. o Advanced visualisations to interpret, compare and confront comprehensive data sets. o Flexible access to a large range of different CFD methods meeting the required accu-