A Hybrid Shape Optimization Approach for Complex Applications in Computational Fluid Dynamics
Abstract
A shape optimization tool for complex applications in Computational Fluid Dynamics combines a RANSE solver, hybrid genetic algorithms and artificial neural networks. Two approaches for mesh updating are presented, relying on linear and torsional spring analogies or spatial deformation laws. The influence of turbulence models during the optimization loop is analyzed for hull shape optimization at full scale.
Origin : Files produced by the author(s)