Research Fellows
Ismael Sānchez Torreguitart
Early Stage Researcher 13 at Von Karman Institute
2. The proposed adaptive shape parameterization
A multilevel shape optimization method is proposed, which is assisted by a new parametric effectiveness indicator. The parametric effectiveness is the means to select the most appropiate CAD-based parameterization for the optimization. The multilevel strategy performs an initial rapid, low-dimensional exploration of the design space and, then, the algorithm progressively starts exploring higher-dimenisonal design spaces that approximate the continuous problem. By doing so, the progressive parameterization strategy has better efficiency.
Figure 4a and 4b show a reasonably good agreement between the performance sensitivities obtained with the discrete adjoint method and those obtained by finite differences when perturbing each design variable with a suitable step-size. Figure 5 shows the results of a single level and multilevel optimization of the LS89. Large aerodynamic improvements in the entropy generation are achieved whilst keeping the exit flow angle fixed. The trailing edge and axial chord length, which are kept fixed as manufacturing constraints. The optimization results show that the multilevel optimizations were more efficient than the single level optimization, even though they used the same number of design variables at the end of the multilevel optimizations.
