Parametric-adjoint approach for the efficient optimization of flow-exposed geometries

Today, the optimization of ship hulls and appendages, including energy-saving devices, is typically undertaken by means of coupling parametric modelling (variable geometry) and Computational Fluid Dynamics (CFD). A relatively new approach is based on parameter-free solutions, solving the adjoint RANS equations for selected objective functions (like drag and lift). Combining parametric and parameter-free solutions is an emerging technique that helps to effectively optimize shapes without leaving the CAD domain of the model, making it easier to integrate in the overall design process. On the basis of the Computer Aided Engineering (CAE) software CAESES, a parametric- adjoint approach will be presented. The approach is built on concatenating so-called “design velocities” and “adjoint shape sensitivities”. Design velocities yield regions of influence from a pure geometric point of view within a given parametric model. Meanwhile, adjoint shape sensitivities show where and how changes of the surface affect the objective. Overlaying the surface distributions of both the design velocities and the adjoint shape sensitivities result in so-called “parametric sensitivities.” These help to understand the importance of all parameters wi hin the chosen model. This approach will be demonstrated on a practical hull form optimization example